**
Hi Listers,

I would appreciate your input to the following question:

"What advantage does Remedy offer in developing an application against using something like Visual Basic or Oracle Forms."

Thank you for your input.

Raj

This posting was submitted via the Web interface

**
**

Ok, I’ll bite first.



Oracle Forms/VB are true 4GL’s (Generation Language) and Remedy can be viewed as ‘4 and half’ GL. The distinction that I make is the granularity of control the developer has. With a 4GL you have complete control of the environment. That control comes with a cost. The cost is longer development and debugging time. Remedy offers 90% of the control of a 4GL without the lengthy development time.



That’s my 50k foot view.



Peter McMullen

QMX Support Services

Voice: 250.701.3748



-----Original Message-----
From: Action Request System discussion list(ARSList) [mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms



Hi Listers,

I would appreciate your input to the following question:

"What advantage does Remedy offer in developing an application against using something like Visual Basic or Oracle Forms."

Thank you for your input.

Raj

This posting was submitted via the Web This posting was submitted via the Web interface interface

**
Peter,

Wouldn't that be 3 and a half GL or less ? Semantics.....

Regards...Gidd

-----Original Message-----
From: Action Request System discussion list(ARSList) [mailto:ARSLIST@ARSLIST.ORG]On Behalf Of Peter McMullen
Sent: Friday, February 07, 2003 9:25 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


** **

Ok, I’ll bite first.



Oracle Forms/VB are true 4GL’s (Generation Language) and Remedy can be viewed as ‘4 and half’ GL. The distinction that I make is the granularity of control the developer has. With a 4GL you have complete control of the environment. That control comes with a cost. The cost is longer development and debugging time. Remedy offers 90% of the control of a 4GL without the lengthy development time.



That’s my 50k foot view.



Peter McMullen

QMX Support Services

Voice: 250.701.3748



-----Original Message-----
From: Action Request System discussion list(ARSList) [mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms



Hi Listers,

I would appreciate your input to the following question:

"What advantage does Remedy offer in developing an application against using something like Visual Basic or Oracle Forms."

Thank you for your input.

Raj

This posting was submitted via the Web This posting was submitted via the Web interface interface

This posting was submitted via the Web interface

**
The vision at Remedy is to be cross-platform web application development framework. Version 5 is a big step in that direction. Being cross-platform, you can't expect to have all the flexibility and control you would have in developing for a specific platform.

Even Visual Basic will eventually be dropped in favor of VB.NET, which takes us back to web application's development. So eventually all application development will be for the web, and client server technologies will slowly and gradually fade away. It will take time, but that's the end game.

If you develop something in VB now, you may need to migrate to VB.NET in the mid term. Microsoft will do whatever they can to force application migrations to .NET, including dropping support for "older" technologies. As an FYI, did you know that Microsoft had initial plans to drop Win2K support in April 2003? That's two months from now!! Here at Target, there is still a lot of NT yet! We haven't even migrated to W2K entirely yet!!

Guillaume


-----Original Message-----
From: Raj Rajakumar [mailto:rajkn@MSN.COM]
Sent: Friday, February 07, 2003 10:02
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms


**
Hi Listers,

I would appreciate your input to the following question:

"What advantage does Remedy offer in developing an application against using something like Visual Basic or Oracle Forms."

Thank you for your input.

Raj

This posting was submitted via the Web interface

This posting was submitted via the Web interface

**
Gidd and Peter,

This is a great point. Can we find a consensus on what generation Remedy is as a development language? I've been going under the assumption it is either a 4th generation language or a code generator, a distinction which is extremely important when comparing metrics data for our project against dissimilar projects. How do others generate metrics for their Remedy project? We're using LSLOC to generate metrics for our project. Currently, we sit at about 27 million LSLOC for 4 developers over 4 years of developing with Remedy.

If anyone can give us some feedback on how they're getting numbers, I'd really appreciate it.

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516



-----Original Message-----
From: Gidd Calden [mailto:gidd@BUOYANTSOLUTIONS.NET]
Sent: Friday, February 07, 2003 11:43
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


**
Peter,

Wouldn't that be 3 and a half GL or less ? Semantics.....

Regards...Gidd

-----Original Message-----
From: Action Request System discussion list(ARSList) [mailto:ARSLIST@ARSLIST.ORG]On Behalf Of Peter McMullen
Sent: Friday, February 07, 2003 9:25 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


** **

Ok, I'll bite first.



Oracle Forms/VB are true 4GL's (Generation Language) and Remedy can be viewed as '4 and half' GL. The distinction that I make is the granularity of control the developer has. With a 4GL you have complete control of the environment. That control comes with a cost. The cost is longer development and debugging time. Remedy offers 90% of the control of a 4GL without the lengthy development time.



That's my 50k foot view.



Peter McMullen

QMX Support Services

Voice: 250.701.3748



-----Original Message-----
From: Action Request System discussion list(ARSList) [mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms



Hi Listers,

I would appreciate your input to the following question:

"What advantage does Remedy offer in developing an application against using something like Visual Basic or Oracle Forms."

Thank you for your input.

Raj

This posting was submitted via the Web This posting was submitted via the Web interface interface

This posting was submitted via the Web interface
This posting was submitted via the Web interface

**
**

Depends on your perspective !!!



Peter McMullen

QMX Support Services

Voice: 250.701.3748



-----Original Message-----
From: Action Request System discussion list(ARSList) [mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Gidd Calden
Sent: Friday, February 07, 2003 9:43 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms



Peter,



Wouldn't that be 3 and a half GL or less ? Semantics.....



Regards...Gidd

-----Original Message-----
From: Action Request System discussion list(ARSList) [mailto:ARSLIST@ARSLIST.ORG]On Behalf Of Peter McMullen
Sent: Friday, February 07, 2003 9:25 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms

** **

Ok, I’ll bite first.



Oracle Forms/VB are true 4GL’s (Generation Language) and Remedy can be viewed as ‘4 and half’ GL. The distinction that I make is the granularity of control the developer has. With a 4GL you have complete control of the environment. That control comes with a cost. The cost is longer development and debugging time. Remedy offers 90% of the control of a 4GL without the lengthy development time.



That’s my 50k foot view.



Peter McMullen

QMX Support Services

Voice: 250.701.3748



-----Original Message-----
From: Action Request System discussion list(ARSList) [mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms



Hi Listers,

I would appreciate your input to the following question:

"What advantage does Remedy offer in developing an application against using something like Visual Basic or Oracle Forms."

Thank you for your input.

Raj

This posting was submitted via the Web This posting was submitted via the Web interface interface

This posting was submitted via the Web This posting was submitted via the Web interface interface

Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters -
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime to
master."

----- Original Message -----
From: Spaeth Jennifer Contractor HQ AFWA
To: 'arslist@ARSLIST.ORG'
Sent: Friday, February 07, 2003 12:12 PM
Subject: RE: Remedy Vs VB/Oracle Forms


**
Gidd and Peter,

This is a great point. Can we find a consensus on what generation Remedy is
as a development language? I've been going under the assumption it is
either a 4th generation language or a code generator, a distinction which is
extremely important when comparing metrics data for our project against
dissimilar projects. How do others generate metrics for their Remedy
project? We're using LSLOC to generate metrics for our project. Currently,
we sit at about 27 million LSLOC for 4 developers over 4 years of developing
with Remedy.

If anyone can give us some feedback on how they're getting numbers, I'd
really appreciate it.

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516


-----Original Message-----
From: Gidd Calden [mailto:gidd@BUOYANTSOLUTIONS.NET]
Sent: Friday, February 07, 2003 11:43
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


**
Peter,

Wouldn't that be 3 and a half GL or less ? Semantics.....

Regards...Gidd
-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG]On Behalf Of Peter McMullen
Sent: Friday, February 07, 2003 9:25 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


** **
Ok, I'll bite first.

Oracle Forms/VB are true 4GL's (Generation Language) and Remedy can be
viewed as '4 and half' GL. The distinction that I make is the granularity
of control the developer has. With a 4GL you have complete control of the
environment. That control comes with a cost. The cost is longer
development and debugging time. Remedy offers 90% of the control of a 4GL
without the lengthy development time.

That's my 50k foot view.

Peter McMullen
QMX Support Services
Voice: 250.701.3748

-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj

UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com This posting was submitted via the Web

UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com This posting was submitted via the Web interface
interface

UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com This posting was submitted via the Web interface

UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com This posting was submitted via the Web interface

Guillaume.Rheault wrote:

[...]

> If you develop something in VB now, you may need to migrate to VB.NET in
> the mid term. Microsoft will do whatever they can to force application
> migrations to .NET, including dropping support for "older" technologies.
> As an FYI, did you know that Microsoft had initial plans to drop Win2K
> support in April 2003? That's two months from now!! Here at Target,
> there is still a lot of NT yet! We haven't even migrated to W2K entirely
> yet!!

I think you'll find that they may reverse that decision at some point.
You see, if you're being forced to move from the relatively stable Win2K
environment to XP, then you're being forced from a known good platform
to one that costs more, requires more system resources to run, and has
an even more restrictive licencing policy.

On the other hand, Linux doesn't require licences, is now known to be
more stable than Win2K even, and happily out-performs Win2K and XP on
similar hardware.

The Linux environment is also quite a nice one to use; lots of scripting
languages built in, nice CRON system, firewalling built into the modern
kernels, and it is still sufficiently UNIX-like that Solaris engineers
can get along with it, and sufficiently Windows-like that most users get
on with the desktops fairly well, too.

So, Microsoft is just starting to get a little nervous about Linux.
Sufficiently nervous that I think they'll reconsider phasing out Win2K,
since such a move might give quite a few IT managers the push they've
been needing to move to Linux, especially in the modern economic climate
where cutting costs is rather important.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606

Hi All!

After having written enough code with "real programming languages" (i.e.:
2GL's & 3GL's) to kill a small forest for paper to print all of it out, I
would agree with Will that one might not be able to label oneself a true
"Computer Programmer" after having "Written" a set of Remedy apps using the
basic Remedy functionality.

However, In my opinion, Remedy IS a 4GL.

By the definitions on the dictionary.com site at
http://dictionary.reference.com/search?q=fourth%20generation%20language:

"...
An "application specific" language. The term was invented by
Jim Martin to refer to non-procedural high level languages
built around database systems. The first three generations
were developed fairly quickly, but it was still frustrating,
slow, and error prone to program computers, leading to the
first "programming crisis", in which the amount of work that
might be assigned to programmers greatly exceeded the amount
of programmer time available to do it. Meanwhile, a lot of
experience was gathered in certain areas, and it became clear
that certain applications could be generalised by adding
limited programming languages to them.
..."

Now:
1) Remedy is "application specific"
i.e.: Worflow automation
2) Remedy is "non-procedural"
i.e.: No extensive million Lines Of Code to be typed manually.
3) Remedy is "high level"
i.e.: Drag-&-Drop form creation without having to manually create
supporting data storage structures on disk or DB.
4) Remedy is "built around database systems"
i.e.: See Remedy Compatibility Matrix.
5) Remedy is "generalised"
i.e.: Generalised functions available for application generation in
tems of active link/filter triggers & actions.


Just my 0.02.


Best Regards,
Theo J. Fondse
Remedy Developer
Siemens Business Services

Telephone: +27 11 652 7477
Mobile: +27 83 773 3075
e-Mail: theo.fondse@ars.smc.siemens.co.za
Web: http://www.siemens.co.za/
===========================================================================


-----Original Message-----
From: William H. "Will" Du Chene [mailto:wduchene@GENESISWIRELESS.COM]
Sent: Saturday, February 08, 2003 5:53 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters -
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime to
master."

----- Original Message -----
From: Spaeth Jennifer Contractor HQ AFWA
To: 'arslist@ARSLIST.ORG'
Sent: Friday, February 07, 2003 12:12 PM
Subject: RE: Remedy Vs VB/Oracle Forms


**
Gidd and Peter,

This is a great point. Can we find a consensus on what generation Remedy is
as a development language? I've been going under the assumption it is
either a 4th generation language or a code generator, a distinction which is
extremely important when comparing metrics data for our project against
dissimilar projects. How do others generate metrics for their Remedy
project? We're using LSLOC to generate metrics for our project. Currently,
we sit at about 27 million LSLOC for 4 developers over 4 years of developing
with Remedy.

If anyone can give us some feedback on how they're getting numbers, I'd
really appreciate it.

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516


-----Original Message-----
From: Gidd Calden [mailto:gidd@BUOYANTSOLUTIONS.NET]
Sent: Friday, February 07, 2003 11:43
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


**
Peter,

Wouldn't that be 3 and a half GL or less ? Semantics.....

Regards...Gidd
-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG]On Behalf Of Peter McMullen
Sent: Friday, February 07, 2003 9:25 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


** **
Ok, I'll bite first.

Oracle Forms/VB are true 4GL's (Generation Language) and Remedy can be
viewed as '4 and half' GL. The distinction that I make is the granularity
of control the developer has. With a 4GL you have complete control of the
environment. That control comes with a cost. The cost is longer
development and debugging time. Remedy offers 90% of the control of a 4GL
without the lengthy development time.

That's my 50k foot view.

Peter McMullen
QMX Support Services
Voice: 250.701.3748

-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj

UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com This posting was submitted via the Web

UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com This posting was submitted via the Web interface
interface

UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com This posting was submitted via the Web interface

UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com This posting was submitted via the Web interface



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

I think (at least in an environment like ours, where processes frequently
evolve and change) that a good Remedy implementation SHOULD be
"procedural". We've written, and use, self contained "modules" that can
be called on by any of our Remedy applications, for example to manage
authorisations, email notifications, etc. Applications for these tasks can
easily be designed and implemented "black box" style, with clearly defined
data inputs and a clearly defined end-result or data outputs. That's
procedural, even if it doesn't say so on the tin.

Hard coding every part of the business process, in my opinion, can lead to
vast amounts of work down the line. Remedy to me is a programming
environment without the useless details - you can do pretty much anything
with it, but there's no compiler to throw 500 error messages because you've
dared to put a semicolon in the wrong place.


Jonathan Hall
Remedy Systems Manager
Yell Ltd.






"Fondse, Theo" @ARSLIST.ORG> on
10/02/2003 12:42:15

To: ARSLIST@ARSLIST.ORG
cc:
bcc:
From: "Action Request System discussion list(ARSList)"

JMHO,

I have read all of the post on this thread through Theo Fondse's and
Jonathan Hall's, and I think there is a bit of a misnomer in the response
below. The reply below implies that ARS development (or coding if you
prefer) is "easy". While I agree that it is simple to create an
application, I would contend that there is a big difference between a "good
application" and "any application". (Jonathan's email speaks to some of the
design/implementation choices in ARS.)

More to the topic at hand....
Some languages are "easy" because the syntax is very forgiving(allow
just about anything), and some languages are easy because they offer so few
options. (AKA: one way to do a task and it is obvious how to do that.) On
the other hand some IDE's make some languages much easier. That is a
function of the IDE and how well it maps the programming language
constructs into human behavioral patterns (IMHO.) and not representative of
the underlaying language.

I am not from the classical Computer Science background. (In fact I
come from a BS in Mechanical Engineering.) So, my "definitions" are very
week about what makes a programming language. However, I also think the
specific example of "parsing a line from a file" and "socket programming"
are poor tests to use to determine the validity of a programming language.
(Just take a look at some of the first programming languages for example.
Or even Assembler, Or MS Batch scripts.... or GW Basic... The list goes
on... Do you really think that those specific tasks were easy/do able in
those languages?)
I do program in several languages and have found that each language
has it's own strengths. (Just like spoke language.) Some are faster than
others, some are more compact, and others still are "easier to learn".

William does have a perspective. However I want to try to add some flavor
to his statements.

I view the Admin tool (and parts of the User tool) as an IDE. (An IDE
specific to the ARS "object language".) The point of the Admin tool is to
avoid syntax/compiler problems. It will not let you define an ARS object
that is "invalid". (Actually the C API really does that job. However the
IDE provides a GUI to the API. So I view them as one and the same. The
difference is merely a translation of "coding interface" to "GUI
interface". True, the GUI requires less understanding and grants less
control than the coding interface, but it is also much faster and less
syntax error prone to develop in too. But I digress....) The Admin tool
allows us to think in terms of ARS objects (Forms, Active Links, Menus,
Filters, Groups, etc...) instead of the language that implements the
"coding interface". These constructs do not map well to any "programming
language" as best I can tell. After all an object is an encapsulated
"thing". The "thing" could be anything from a socket connection, to a
Integer variable, to a reference to another object.


IMHO:

Think in terms of the "English language". A variable is a single
letter of the alphabet. It has a value of a "set of sounds" and is a member
of one of two types. (consent, vowel) Multiple variables combined together
produce a "word". Which is a language specific object. That object has a
greater meaning that the "sum" of the variables that compose the object.
String together a set of those objects and you get a sentence. Again a new
language specific object with a greater meaning than the sum of the
components. There are rules that define how to combine the variables and
the objects. (Grammar) So anything that exhibits those traits would be a
"language" in my book.

The bottom line for me to decide if something is a "programming
language" is based being able to derive a "variable construct". IMHO,
anything that has a "variable" is a programming language. (Yes, I think
Mathematics really is a programming language too. However it is more
abstract and compact language than any "human or computer based" language
that I know. Again, JMHO. :) Without a variable you can not have a
programming language. As you have no way to distinguish between multiple
"things" that have value. Ideally any language has a grammar that also
specifies how variables relate to each other and how they can be combined
to make "greater meanings".

To that end, I see ARS as having variables. They are more commonly
known as "ARS fields". Why do I think an ARS field is a variable? (Glad you
asked.) A variable is a storage unit that holds a value. Normally the
storage unit is known by a name, and may have an associated "data type" as
well. The data type is completely optional for my definition as some
languages are "type independent". A variables value is normally mutable and
can contain three basic conditions. (undefined[does not exist therefore can
not have a value], NULL[exists, but has no value] and NOT NULL[exists and
has a value]) The exact implementation of what NOT NULL means depends on
the data type of the variable. (If a language is "type independent" then
the variables still have a type, it is simply all the same type and is
language specific.) BTW: ARS is a "type dependent" language. (ARS fields
come in "Data Types" that are ARS constructs like: Character, Integer,
Real, Decimal, Attachment, Flashboard, etc...
ARS also provides "rules" about how the objects are combined to make
a "larger object" (form/Application with more meaning. These rules are
enforced at the API level, however they manifest in the IDE as
"options"/"settings" for the ARS objects.


To me the value of ARS is that they have created objects that make
application development much faster (and easier) than starting at first
principals.

However, that can be a two sided sword. Because the down side is that
you have to use ARS objects. (At this point ARS is only marginally
extendable. The ARF API/ ARDBC API opens up a good deal of extendibility.
However there is much more that is needed before I would declare ARS as
"extensible". For now I would have to define ARS as able to accept
"plug-ins". (A Plug-in has to meet a defined standard interface. While an
extension (IMHO, a requirement to declare a language as be an extensible)
can encapsulate and/or add to existing interfaces of the language. One adds
new "features" to an existing interface. The other "alters" the base
interface/language.)

As a finishing note:
http://dictionary.reference.com/search?q=fourth%20generation%20language:
has the following at the end of it's description.

"

Some other successful 4th-generation languages are: database
query languages, e.g. SQL; Focus, Metafont,
PostScript, RPG-II, {S}, IDL-PV/WAVE, Gauss,
Mathematica and data-stream languages such as AVS,
APE, Iris Explorer.


"

If SQL is a language then ARS is a language. (Especially if you simply view
ARS as a "subset"/"superset" of SQL. I think it is much more than that, but
it is definitely at least that.)

--
Carey Matthew Black
Remedy Skilled Professional (RSP)
ARS=Action Request System(Remedy)

Solution = People + Process + Tools
Fast, Accurate, Cheap.... Pick two.
Never ascribe to malice, that which can be
explained by incompetence.



"William H.
\"Will\" Du Chene" To: ARSLIST@ARSLIST.ORG
RELESS.COM> Subject: Re: Remedy Vs VB/Oracle Forms
Sent by: "Action
Request System
discussion
list(ARSList)"
G>


02/07/2003 09:53 PM
Please respond to
arslist






Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with
a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type
of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will
allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use
it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters
-
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I
can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime
to
master."


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against
using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Thanks to all who replied.

Whether Remedy is or is not a "language" was not the point of my earlier
post. My sincerest apologies for offending the programming purists.

I'm trying to gather metrics on our development effort. In order to compare
our metrics meaningfully to the metrics of another project, say a project
that is programming with a "real" language, I am using Logical Source Lines
of Code (LSLOC) to establish a common ground.

Has anyone else attempted to gather metrics on their Remedy development? If
so, how are you going about it? Our team is assigning a difficulty level to
each piece of workflow, multiplying it by a constant for that piece of
workflow, and figuring in a complexity factor for the level of the
programming language.

For example, I create active link ZZZMyActiveLink. It is a very complex
active link, with a long qualification string and 15 actions, so I assign it
a complexity factor of 17. Active links have a constant factor of 9. So
9x17=153. Since we have assumed that Remedy is a 4th generation "language,"
we have used a programming language conversion factor of 20 and an overall
complexity factor of 1.07 for our project. So ZZZMyActiveLink equates to
3274 LSLOC.

Doing the same thing in C, a third generation language with a programming
language conversion factor of 15, should give me the same number of LSLOC,
because it would require more lines of code in C.

Now that I've explained the nature of my inquiry more concisely, can anyone
provide feedback on our method?

Many thanks,

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516




-----Original Message-----
From: Carey Matthew Black [mailto:matt.black@VERIZON.COM]
Sent: Monday, February 10, 2003 09:44
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


JMHO,

I have read all of the post on this thread through Theo Fondse's and
Jonathan Hall's, and I think there is a bit of a misnomer in the response
below. The reply below implies that ARS development (or coding if you
prefer) is "easy". While I agree that it is simple to create an
application, I would contend that there is a big difference between a "good
application" and "any application". (Jonathan's email speaks to some of the
design/implementation choices in ARS.)

More to the topic at hand....
Some languages are "easy" because the syntax is very forgiving(allow
just about anything), and some languages are easy because they offer so few
options. (AKA: one way to do a task and it is obvious how to do that.) On
the other hand some IDE's make some languages much easier. That is a
function of the IDE and how well it maps the programming language
constructs into human behavioral patterns (IMHO.) and not representative of
the underlaying language.

I am not from the classical Computer Science background. (In fact I
come from a BS in Mechanical Engineering.) So, my "definitions" are very
week about what makes a programming language. However, I also think the
specific example of "parsing a line from a file" and "socket programming"
are poor tests to use to determine the validity of a programming language.
(Just take a look at some of the first programming languages for example.
Or even Assembler, Or MS Batch scripts.... or GW Basic... The list goes
on... Do you really think that those specific tasks were easy/do able in
those languages?)
I do program in several languages and have found that each language
has it's own strengths. (Just like spoke language.) Some are faster than
others, some are more compact, and others still are "easier to learn".

William does have a perspective. However I want to try to add some flavor
to his statements.

I view the Admin tool (and parts of the User tool) as an IDE. (An IDE
specific to the ARS "object language".) The point of the Admin tool is to
avoid syntax/compiler problems. It will not let you define an ARS object
that is "invalid". (Actually the C API really does that job. However the
IDE provides a GUI to the API. So I view them as one and the same. The
difference is merely a translation of "coding interface" to "GUI
interface". True, the GUI requires less understanding and grants less
control than the coding interface, but it is also much faster and less
syntax error prone to develop in too. But I digress....) The Admin tool
allows us to think in terms of ARS objects (Forms, Active Links, Menus,
Filters, Groups, etc...) instead of the language that implements the
"coding interface". These constructs do not map well to any "programming
language" as best I can tell. After all an object is an encapsulated
"thing". The "thing" could be anything from a socket connection, to a
Integer variable, to a reference to another object.


IMHO:

Think in terms of the "English language". A variable is a single
letter of the alphabet. It has a value of a "set of sounds" and is a member
of one of two types. (consent, vowel) Multiple variables combined together
produce a "word". Which is a language specific object. That object has a
greater meaning that the "sum" of the variables that compose the object.
String together a set of those objects and you get a sentence. Again a new
language specific object with a greater meaning than the sum of the
components. There are rules that define how to combine the variables and
the objects. (Grammar) So anything that exhibits those traits would be a
"language" in my book.

The bottom line for me to decide if something is a "programming
language" is based being able to derive a "variable construct". IMHO,
anything that has a "variable" is a programming language. (Yes, I think
Mathematics really is a programming language too. However it is more
abstract and compact language than any "human or computer based" language
that I know. Again, JMHO. :) Without a variable you can not have a
programming language. As you have no way to distinguish between multiple
"things" that have value. Ideally any language has a grammar that also
specifies how variables relate to each other and how they can be combined
to make "greater meanings".

To that end, I see ARS as having variables. They are more commonly
known as "ARS fields". Why do I think an ARS field is a variable? (Glad you
asked.) A variable is a storage unit that holds a value. Normally the
storage unit is known by a name, and may have an associated "data type" as
well. The data type is completely optional for my definition as some
languages are "type independent". A variables value is normally mutable and
can contain three basic conditions. (undefined[does not exist therefore can
not have a value], NULL[exists, but has no value] and NOT NULL[exists and
has a value]) The exact implementation of what NOT NULL means depends on
the data type of the variable. (If a language is "type independent" then
the variables still have a type, it is simply all the same type and is
language specific.) BTW: ARS is a "type dependent" language. (ARS fields
come in "Data Types" that are ARS constructs like: Character, Integer,
Real, Decimal, Attachment, Flashboard, etc...
ARS also provides "rules" about how the objects are combined to make
a "larger object" (form/Application with more meaning. These rules are
enforced at the API level, however they manifest in the IDE as
"options"/"settings" for the ARS objects.


To me the value of ARS is that they have created objects that make
application development much faster (and easier) than starting at first
principals.

However, that can be a two sided sword. Because the down side is that
you have to use ARS objects. (At this point ARS is only marginally
extendable. The ARF API/ ARDBC API opens up a good deal of extendibility.
However there is much more that is needed before I would declare ARS as
"extensible". For now I would have to define ARS as able to accept
"plug-ins". (A Plug-in has to meet a defined standard interface. While an
extension (IMHO, a requirement to declare a language as be an extensible)
can encapsulate and/or add to existing interfaces of the language. One adds
new "features" to an existing interface. The other "alters" the base
interface/language.)

As a finishing note:
http://dictionary.reference.com/search?q=fourth%20generation%20language:
has the following at the end of it's description.

"

Some other successful 4th-generation languages are: database
query languages, e.g. SQL; Focus, Metafont,
PostScript, RPG-II, {S}, IDL-PV/WAVE, Gauss,
Mathematica and data-stream languages such as AVS,
APE, Iris Explorer.


"

If SQL is a language then ARS is a language. (Especially if you simply view
ARS as a "subset"/"superset" of SQL. I think it is much more than that, but
it is definitely at least that.)

--
Carey Matthew Black
Remedy Skilled Professional (RSP)
ARS=Action Request System(Remedy)

Solution = People + Process + Tools
Fast, Accurate, Cheap.... Pick two.
Never ascribe to malice, that which can be
explained by incompetence.



"William H.
\"Will\" Du Chene" To:
ARSLIST@ARSLIST.ORG
RELESS.COM> Subject: Re: Remedy Vs
VB/Oracle Forms
Sent by: "Action
Request System
discussion
list(ARSList)"
G>


02/07/2003 09:53 PM
Please respond to
arslist






Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with
a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type
of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will
allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use
it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters
-
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I
can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime
to
master."


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against
using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

The point about extendability is an interesting one. I personally find it
an enjoyable challenge to push the boundaries of ARS and other
environments, while NOT using any extensions. I frequently see perfectly
good suggested solutions to questions on this loop, which use SQL or
similar to achieve their ends, yet can see an all-ARS way of doing the same
thing.

I inherited a 3-year old ARS system several years ago, and one day will
possibly pass it on to someone else. Using ARS forms and workflow (great
point about variables below) for the solution will tend to allow better
maintanability and reliability (less external components to cause
problems). Having used Acorn's range of computers through the late
eighties and early nineties, I seem to have picked up a "reduced
instruction set" mentality which I apply to the likes of Remedy and Access
(why use VBA when an Access macro will do?).

The use of external code through the API is obviously highly appropriate in
certain areas, like customisation of interfaces and integration with other
applications at an API level, and I've seen some very well thought out
products and solutions in this area (hello UWIP!).

But as I've said, the point about system variables below is a great one.
There is little that can't be coded in Remedy with a bit of thought, and by
maximising the use of Remedy's own code, you can build a very maintainable
and adaptable system. I've done enough C coding in my shortish career
(academic and professional) to know that it can be a complete pain in the
backside, from attempting to translate someone else's shonky code, to
working out which misplaced semicolon has caused the 5400 errors in your
compiler report. I'd personally rather get on with the programming and
ignore the semantic details. I fully regard myself as a programmer, and
the fact that I'm using Remedy rather than Oracle/Forms or C++ or any other
environment doesn't cheapen that in my opinion.

Jonathan Hall
Remedy Systems Manager, Yell Ltd.






Carey Matthew Black @ARSLIST.ORG> on 10/02/2003
15:43:59

To: ARSLIST@ARSLIST.ORG
cc:
bcc:
From: "Action Request System discussion list(ARSList)"

**

Actually, the distinction between first/second/third/fourth generation languages is key to your LSLOC metrics. Perhaps I may offer more traditional definitions?

1st Generation: Machine code (octal/hex instructions DIRECTLY executed by the CPU)
2nd Generation: Assembler (mnemonic-based, with a one-to-one correspondance between mnemonic and machine code)
3rd Generation: "English-like" and, for the first time, portable between computers (via compilers/interpreters)
4th Generation: Designed for non- or semi-programmers and typically associated with a specific DBMS

If you're referring to T.J. McCabe's complexity measures (IEEE, 1976), then your method is sound (but may need further tuning as you compare resources planned against actual resources consumed over several projects). By the way, some useful documents directly related to your metrics issues are:

- T.J. McCabe's "A Complexity Measure" from the IEEE Transactions on Software Engineering
- Mitre/USAF document ESD-TR-88-001, "Software Management Metrics"

You might also take a look at CeBASE ( http://www.cebase.org/www/home/index.htm).

Another model for software metrics cross-validation involves Function Points. Some URLs:

http://www.ifpug.org/
http://www.spr.com/
http://www.functionpoints.com/

Please feel free to call or email (off the ARSlist) if you have any follow-up questions!

-- Bing

Bradford Bingel ("Bing")
ITM3 - San Francisco Bay Area
bing@itm3.com (email)
925-260-6394 (mobile)

-----Original Message-----
From: Action Request System discussion list(ARSList)
[ mailto:ARSLIST@ARSLIST.ORG]On Behalf Of Spaeth Jennifer Contractor HQ
AFWA
Sent: Monday, February 10, 2003 8:41 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Thanks to all who replied.

Whether Remedy is or is not a "language" was not the point of my earlier
post. My sincerest apologies for offending the programming purists.

I'm trying to gather metrics on our development effort. In order to compare
our metrics meaningfully to the metrics of another project, say a project
that is programming with a "real" language, I am using Logical Source Lines
of Code (LSLOC) to establish a common ground.

Has anyone else attempted to gather metrics on their Remedy development? If
so, how are you going about it? Our team is assigning a difficulty level to
each piece of workflow, multiplying it by a constant for that piece of
workflow, and figuring in a complexity factor for the level of the
programming language.

For example, I create active link ZZZMyActiveLink. It is a very complex
active link, with a long qualification string and 15 actions, so I assign it
a complexity factor of 17. Active links have a constant factor of 9. So
9x17=153. Since we have assumed that Remedy is a 4th generation "language,"
we have used a programming language conversion factor of 20 and an overall
complexity factor of 1.07 for our project. So ZZZMyActiveLink equates to
3274 LSLOC.

Doing the same thing in C, a third generation language with a programming
language conversion factor of 15, should give me the same number of LSLOC,
because it would require more lines of code in C.

Now that I've explained the nature of my inquiry more concisely, can anyone
provide feedback on our method?

Many thanks,

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516




-----Original Message-----
From: Carey Matthew Black [ mailto:matt.black@VERIZON.COM]
Sent: Monday, February 10, 2003 09:44
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


JMHO,

I have read all of the post on this thread through Theo Fondse's and
Jonathan Hall's, and I think there is a bit of a misnomer in the response
below. The reply below implies that ARS development (or coding if you
prefer) is "easy". While I agree that it is simple to create an
application, I would contend that there is a big difference between a "good
application" and "any application". (Jonathan's email speaks to some of the
design/implementation choices in ARS.)

More to the topic at hand....
Some languages are "easy" because the syntax is very forgiving(allow
just about anything), and some languages are easy because they offer so few
options. (AKA: one way to do a task and it is obvious how to do that.) On
the other hand some IDE's make some languages much easier. That is a
function of the IDE and how well it maps the programming language
constructs into human behavioral patterns (IMHO.) and not representative of
the underlaying language.

I am not from the classical Computer Science background. (In fact I
come from a BS in Mechanical Engineering.) So, my "definitions" are very
week about what makes a programming language. However, I also think the
specific example of "parsing a line from a file" and "socket programming"
are poor tests to use to determine the validity of a programming language.
(Just take a look at some of the first programming languages for example.
Or even Assembler, Or MS Batch scripts.... or GW Basic... The list goes
on... Do you really think that those specific tasks were easy/do able in
those languages?)
I do program in several languages and have found that each language
has it's own strengths. (Just like spoke language.) Some are faster than
others, some are more compact, and others still are "easier to learn".

William does have a perspective. However I want to try to add some flavor
to his statements.

I view the Admin tool (and parts of the User tool) as an IDE. (An IDE
specific to the ARS "object language".) The point of the Admin tool is to
avoid syntax/compiler problems. It will not let you define an ARS object
that is "invalid". (Actually the C API really does that job. However the
IDE provides a GUI to the API. So I view them as one and the same. The
difference is merely a translation of "coding interface" to "GUI
interface". True, the GUI requires less understanding and grants less
control than the coding interface, but it is also much faster and less
syntax error prone to develop in too. But I digress....) The Admin tool
allows us to think in terms of ARS objects (Forms, Active Links, Menus,
Filters, Groups, etc...) instead of the language that implements the
"coding interface". These constructs do not map well to any "programming
language" as best I can tell. After all an object is an encapsulated
"thing". The "thing" could be anything from a socket connection, to a
Integer variable, to a reference to another object.


IMHO:

Think in terms of the "English language". A variable is a single
letter of the alphabet. It has a value of a "set of sounds" and is a member
of one of two types. (consent, vowel) Multiple variables combined together
produce a "word". Which is a language specific object. That object has a
greater meaning that the "sum" of the variables that compose the object.
String together a set of those objects and you get a sentence. Again a new
language specific object with a greater meaning than the sum of the
components. There are rules that define how to combine the variables and
the objects. (Grammar) So anything that exhibits those traits would be a
"language" in my book.

The bottom line for me to decide if something is a "programming
language" is based being able to derive a "variable construct". IMHO,
anything that has a "variable" is a programming language. (Yes, I think
Mathematics really is a programming language too. However it is more
abstract and compact language than any "human or computer based" language
that I know. Again, JMHO. :) Without a variable you can not have a
programming language. As you have no way to distinguish between multiple
"things" that have value. Ideally any language has a grammar that also
specifies how variables relate to each other and how they can be combined
to make "greater meanings".

To that end, I see ARS as having variables. They are more commonly
known as "ARS fields". Why do I think an ARS field is a variable? (Glad you
asked.) A variable is a storage unit that holds a value. Normally the
storage unit is known by a name, and may have an associated "data type" as
well. The data type is completely optional for my definition as some
languages are "type independent". A variables value is normally mutable and
can contain three basic conditions. (undefined[does not exist therefore can
not have a value], NULL[exists, but has no value] and NOT NULL[exists and
has a value]) The exact implementation of what NOT NULL means depends on
the data type of the variable. (If a language is "type independent" then
the variables still have a type, it is simply all the same type and is
language specific.) BTW: ARS is a "type dependent" language. (ARS fields
come in "Data Types" that are ARS constructs like: Character, Integer,
Real, Decimal, Attachment, Flashboard, etc...
ARS also provides "rules" about how the objects are combined to make
a "larger object" (form/Application with more meaning. These rules are
enforced at the API level, however they manifest in the IDE as
"options"/"settings" for the ARS objects.


To me the value of ARS is that they have created objects that make
application development much faster (and easier) than starting at first
principals.

However, that can be a two sided sword. Because the down side is that
you have to use ARS objects. (At this point ARS is only marginally
extendable. The ARF API/ ARDBC API opens up a good deal of extendibility.
However there is much more that is needed before I would declare ARS as
"extensible". For now I would have to define ARS as able to accept
"plug-ins". (A Plug-in has to meet a defined standard interface. While an
extension (IMHO, a requirement to declare a language as be an extensible)
can encapsulate and/or add to existing interfaces of the language. One adds
new "features" to an existing interface. The other "alters" the base
interface/language.)

As a finishing note:
http://dictionary.reference.com/search?q=fourth%20generation%20language:
has the following at the end of it's description.

"

Some other successful 4th-generation languages are: database
query languages, e.g. SQL; Focus, Metafont,
PostScript, RPG-II, {S}, IDL-PV/WAVE, Gauss,
Mathematica and data-stream languages such as AVS,
APE, Iris Explorer.


"

If SQL is a language then ARS is a language. (Especially if you simply view
ARS as a "subset"/"superset" of SQL. I think it is much more than that, but
it is definitely at least that.)

--
Carey Matthew Black
Remedy Skilled Professional (RSP)
ARS=Action Request System(Remedy)

Solution = People + Process + Tools
Fast, Accurate, Cheap.... Pick two.
Never ascribe to malice, that which can be
explained by incompetence.



"William H.
\"Will\" Du Chene" To:
ARSLIST@ARSLIST.ORG
RELESS.COM> Subject: Re: Remedy Vs
VB/Oracle Forms
Sent by: "Action
Request System
discussion
list(ARSList)"
G>


02/07/2003 09:53 PM
Please respond to
arslist






Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with
a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type
of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will
allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use
it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters
-
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I
can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime
to
master."


-----Original Message-----
From: Action Request System discussion list(ARSList)
[ mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against
using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com





This posting was submitted via the Web interface

Jennifer,

I can guarantee you are not the first one to tackle these estimations
problem. Any consulting company, at some point, needs to deal with this
problem one way or the other, specially when they are responding to an RFP,
or any kind of engagement proposal.

Since Remedy applications tend to be enterprise wide applications, the
problem is even though you may have excellent requirements and consensus
from the user community on business requirements and functionality, most
often than not, you'll see those requirements "mature" over time.

So at the end, in many many cases, the project took longer not because of
development issues, but because the **true** requirements were not fully
uncovered at the beginning... And I am not even talking referring to scope
creep! Hence the difficulty in estimating the effort...

Guillaume

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Monday, February 10, 2003 03:25
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Thanks, Bing and Guillaume,

We are keeping track of lots of things in our metrics, including the ones
you suggested below, Guillaume. I just wondered if anyone has been trying
to perform function point analysis for Remedy. We have been keeping track
of these metrics since November 1st, so we're gathering historical data.
I'm asking these questions of the developer community now because we have
enough historical data at this point to suggest our estimates are a wee off.
I had hoped that in the 2000+ readers of the ARSlist that I wasn't the first
person to tackle this beast, but it's starting to look as if I may be. :-/

Thanks for the feedback.

Jennifer

-----Original Message-----
From: Guillaume.Rheault [mailto:Guillaume.Rheault@TARGET.COM]
Sent: Monday, February 10, 2003 15:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Jennifer,

To the best of my knowledge, not even Remedy has come up with a good enough
scientific method to estimate development efforts. Besides the pure
technical challenge, you need to consider the following factors:

- Quality of requirements and design documentation provided to the developer
- Overall experience of the developer
- Particular experience of the developer in the task at hand

I don't want to discourage you in your efforts to come up with a formula,
but in many cases, history data is the best thing you can rely on.

I believe in order to come up with a comprehensive formula, I think you
should somehow incorporate the three points I mentioned, in some kind of
factor

Guillaume


-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Monday, February 10, 2003 10:41
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Thanks to all who replied.

Whether Remedy is or is not a "language" was not the point of my earlier
post. My sincerest apologies for offending the programming purists.

I'm trying to gather metrics on our development effort. In order to compare
our metrics meaningfully to the metrics of another project, say a project
that is programming with a "real" language, I am using Logical Source Lines
of Code (LSLOC) to establish a common ground.

Has anyone else attempted to gather metrics on their Remedy development? If
so, how are you going about it? Our team is assigning a difficulty level to
each piece of workflow, multiplying it by a constant for that piece of
workflow, and figuring in a complexity factor for the level of the
programming language.

For example, I create active link ZZZMyActiveLink. It is a very complex
active link, with a long qualification string and 15 actions, so I assign it
a complexity factor of 17. Active links have a constant factor of 9. So
9x17=153. Since we have assumed that Remedy is a 4th generation "language,"
we have used a programming language conversion factor of 20 and an overall
complexity factor of 1.07 for our project. So ZZZMyActiveLink equates to
3274 LSLOC.

Doing the same thing in C, a third generation language with a programming
language conversion factor of 15, should give me the same number of LSLOC,
because it would require more lines of code in C.

Now that I've explained the nature of my inquiry more concisely, can anyone
provide feedback on our method?

Many thanks,

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516




-----Original Message-----
From: Carey Matthew Black [mailto:matt.black@VERIZON.COM]
Sent: Monday, February 10, 2003 09:44
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


JMHO,

I have read all of the post on this thread through Theo Fondse's and
Jonathan Hall's, and I think there is a bit of a misnomer in the response
below. The reply below implies that ARS development (or coding if you
prefer) is "easy". While I agree that it is simple to create an
application, I would contend that there is a big difference between a "good
application" and "any application". (Jonathan's email speaks to some of the
design/implementation choices in ARS.)

More to the topic at hand....
Some languages are "easy" because the syntax is very forgiving(allow
just about anything), and some languages are easy because they offer so few
options. (AKA: one way to do a task and it is obvious how to do that.) On
the other hand some IDE's make some languages much easier. That is a
function of the IDE and how well it maps the programming language
constructs into human behavioral patterns (IMHO.) and not representative of
the underlaying language.

I am not from the classical Computer Science background. (In fact I
come from a BS in Mechanical Engineering.) So, my "definitions" are very
week about what makes a programming language. However, I also think the
specific example of "parsing a line from a file" and "socket programming"
are poor tests to use to determine the validity of a programming language.
(Just take a look at some of the first programming languages for example.
Or even Assembler, Or MS Batch scripts.... or GW Basic... The list goes
on... Do you really think that those specific tasks were easy/do able in
those languages?)
I do program in several languages and have found that each language
has it's own strengths. (Just like spoke language.) Some are faster than
others, some are more compact, and others still are "easier to learn".

William does have a perspective. However I want to try to add some flavor
to his statements.

I view the Admin tool (and parts of the User tool) as an IDE. (An IDE
specific to the ARS "object language".) The point of the Admin tool is to
avoid syntax/compiler problems. It will not let you define an ARS object
that is "invalid". (Actually the C API really does that job. However the
IDE provides a GUI to the API. So I view them as one and the same. The
difference is merely a translation of "coding interface" to "GUI
interface". True, the GUI requires less understanding and grants less
control than the coding interface, but it is also much faster and less
syntax error prone to develop in too. But I digress....) The Admin tool
allows us to think in terms of ARS objects (Forms, Active Links, Menus,
Filters, Groups, etc...) instead of the language that implements the
"coding interface". These constructs do not map well to any "programming
language" as best I can tell. After all an object is an encapsulated
"thing". The "thing" could be anything from a socket connection, to a
Integer variable, to a reference to another object.


IMHO:

Think in terms of the "English language". A variable is a single
letter of the alphabet. It has a value of a "set of sounds" and is a member
of one of two types. (consent, vowel) Multiple variables combined together
produce a "word". Which is a language specific object. That object has a
greater meaning that the "sum" of the variables that compose the object.
String together a set of those objects and you get a sentence. Again a new
language specific object with a greater meaning than the sum of the
components. There are rules that define how to combine the variables and
the objects. (Grammar) So anything that exhibits those traits would be a
"language" in my book.

The bottom line for me to decide if something is a "programming
language" is based being able to derive a "variable construct". IMHO,
anything that has a "variable" is a programming language. (Yes, I think
Mathematics really is a programming language too. However it is more
abstract and compact language than any "human or computer based" language
that I know. Again, JMHO. :) Without a variable you can not have a
programming language. As you have no way to distinguish between multiple
"things" that have value. Ideally any language has a grammar that also
specifies how variables relate to each other and how they can be combined
to make "greater meanings".

To that end, I see ARS as having variables. They are more commonly
known as "ARS fields". Why do I think an ARS field is a variable? (Glad you
asked.) A variable is a storage unit that holds a value. Normally the
storage unit is known by a name, and may have an associated "data type" as
well. The data type is completely optional for my definition as some
languages are "type independent". A variables value is normally mutable and
can contain three basic conditions. (undefined[does not exist therefore can
not have a value], NULL[exists, but has no value] and NOT NULL[exists and
has a value]) The exact implementation of what NOT NULL means depends on
the data type of the variable. (If a language is "type independent" then
the variables still have a type, it is simply all the same type and is
language specific.) BTW: ARS is a "type dependent" language. (ARS fields
come in "Data Types" that are ARS constructs like: Character, Integer,
Real, Decimal, Attachment, Flashboard, etc...
ARS also provides "rules" about how the objects are combined to make
a "larger object" (form/Application with more meaning. These rules are
enforced at the API level, however they manifest in the IDE as
"options"/"settings" for the ARS objects.


To me the value of ARS is that they have created objects that make
application development much faster (and easier) than starting at first
principals.

However, that can be a two sided sword. Because the down side is that
you have to use ARS objects. (At this point ARS is only marginally
extendable. The ARF API/ ARDBC API opens up a good deal of extendibility.
However there is much more that is needed before I would declare ARS as
"extensible". For now I would have to define ARS as able to accept
"plug-ins". (A Plug-in has to meet a defined standard interface. While an
extension (IMHO, a requirement to declare a language as be an extensible)
can encapsulate and/or add to existing interfaces of the language. One adds
new "features" to an existing interface. The other "alters" the base
interface/language.)

As a finishing note:
http://dictionary.reference.com/search?q=fourth%20generation%20language:
has the following at the end of it's description.

"

Some other successful 4th-generation languages are: database
query languages, e.g. SQL; Focus, Metafont,
PostScript, RPG-II, {S}, IDL-PV/WAVE, Gauss,
Mathematica and data-stream languages such as AVS,
APE, Iris Explorer.


"

If SQL is a language then ARS is a language. (Especially if you simply view
ARS as a "subset"/"superset" of SQL. I think it is much more than that, but
it is definitely at least that.)

--
Carey Matthew Black
Remedy Skilled Professional (RSP)
ARS=Action Request System(Remedy)

Solution = People + Process + Tools
Fast, Accurate, Cheap.... Pick two.
Never ascribe to malice, that which can be
explained by incompetence.



"William H.
\"Will\" Du Chene" To:
ARSLIST@ARSLIST.ORG
RELESS.COM> Subject: Re: Remedy Vs
VB/Oracle Forms
Sent by: "Action
Request System
discussion
list(ARSList)"
G>


02/07/2003 09:53 PM
Please respond to
arslist






Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with
a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type
of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will
allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use
it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters
-
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I
can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime
to
master."


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against
using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Jennifer,

To the best of my knowledge, not even Remedy has come up with a good enough
scientific method to estimate development efforts. Besides the pure
technical challenge, you need to consider the following factors:

- Quality of requirements and design documentation provided to the developer
- Overall experience of the developer
- Particular experience of the developer in the task at hand

I don't want to discourage you in your efforts to come up with a formula,
but in many cases, history data is the best thing you can rely on.

I believe in order to come up with a comprehensive formula, I think you
should somehow incorporate the three points I mentioned, in some kind of
factor

Guillaume


-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Monday, February 10, 2003 10:41
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Thanks to all who replied.

Whether Remedy is or is not a "language" was not the point of my earlier
post. My sincerest apologies for offending the programming purists.

I'm trying to gather metrics on our development effort. In order to compare
our metrics meaningfully to the metrics of another project, say a project
that is programming with a "real" language, I am using Logical Source Lines
of Code (LSLOC) to establish a common ground.

Has anyone else attempted to gather metrics on their Remedy development? If
so, how are you going about it? Our team is assigning a difficulty level to
each piece of workflow, multiplying it by a constant for that piece of
workflow, and figuring in a complexity factor for the level of the
programming language.

For example, I create active link ZZZMyActiveLink. It is a very complex
active link, with a long qualification string and 15 actions, so I assign it
a complexity factor of 17. Active links have a constant factor of 9. So
9x17=153. Since we have assumed that Remedy is a 4th generation "language,"
we have used a programming language conversion factor of 20 and an overall
complexity factor of 1.07 for our project. So ZZZMyActiveLink equates to
3274 LSLOC.

Doing the same thing in C, a third generation language with a programming
language conversion factor of 15, should give me the same number of LSLOC,
because it would require more lines of code in C.

Now that I've explained the nature of my inquiry more concisely, can anyone
provide feedback on our method?

Many thanks,

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516




-----Original Message-----
From: Carey Matthew Black [mailto:matt.black@VERIZON.COM]
Sent: Monday, February 10, 2003 09:44
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


JMHO,

I have read all of the post on this thread through Theo Fondse's and
Jonathan Hall's, and I think there is a bit of a misnomer in the response
below. The reply below implies that ARS development (or coding if you
prefer) is "easy". While I agree that it is simple to create an
application, I would contend that there is a big difference between a "good
application" and "any application". (Jonathan's email speaks to some of the
design/implementation choices in ARS.)

More to the topic at hand....
Some languages are "easy" because the syntax is very forgiving(allow
just about anything), and some languages are easy because they offer so few
options. (AKA: one way to do a task and it is obvious how to do that.) On
the other hand some IDE's make some languages much easier. That is a
function of the IDE and how well it maps the programming language
constructs into human behavioral patterns (IMHO.) and not representative of
the underlaying language.

I am not from the classical Computer Science background. (In fact I
come from a BS in Mechanical Engineering.) So, my "definitions" are very
week about what makes a programming language. However, I also think the
specific example of "parsing a line from a file" and "socket programming"
are poor tests to use to determine the validity of a programming language.
(Just take a look at some of the first programming languages for example.
Or even Assembler, Or MS Batch scripts.... or GW Basic... The list goes
on... Do you really think that those specific tasks were easy/do able in
those languages?)
I do program in several languages and have found that each language
has it's own strengths. (Just like spoke language.) Some are faster than
others, some are more compact, and others still are "easier to learn".

William does have a perspective. However I want to try to add some flavor
to his statements.

I view the Admin tool (and parts of the User tool) as an IDE. (An IDE
specific to the ARS "object language".) The point of the Admin tool is to
avoid syntax/compiler problems. It will not let you define an ARS object
that is "invalid". (Actually the C API really does that job. However the
IDE provides a GUI to the API. So I view them as one and the same. The
difference is merely a translation of "coding interface" to "GUI
interface". True, the GUI requires less understanding and grants less
control than the coding interface, but it is also much faster and less
syntax error prone to develop in too. But I digress....) The Admin tool
allows us to think in terms of ARS objects (Forms, Active Links, Menus,
Filters, Groups, etc...) instead of the language that implements the
"coding interface". These constructs do not map well to any "programming
language" as best I can tell. After all an object is an encapsulated
"thing". The "thing" could be anything from a socket connection, to a
Integer variable, to a reference to another object.


IMHO:

Think in terms of the "English language". A variable is a single
letter of the alphabet. It has a value of a "set of sounds" and is a member
of one of two types. (consent, vowel) Multiple variables combined together
produce a "word". Which is a language specific object. That object has a
greater meaning that the "sum" of the variables that compose the object.
String together a set of those objects and you get a sentence. Again a new
language specific object with a greater meaning than the sum of the
components. There are rules that define how to combine the variables and
the objects. (Grammar) So anything that exhibits those traits would be a
"language" in my book.

The bottom line for me to decide if something is a "programming
language" is based being able to derive a "variable construct". IMHO,
anything that has a "variable" is a programming language. (Yes, I think
Mathematics really is a programming language too. However it is more
abstract and compact language than any "human or computer based" language
that I know. Again, JMHO. :) Without a variable you can not have a
programming language. As you have no way to distinguish between multiple
"things" that have value. Ideally any language has a grammar that also
specifies how variables relate to each other and how they can be combined
to make "greater meanings".

To that end, I see ARS as having variables. They are more commonly
known as "ARS fields". Why do I think an ARS field is a variable? (Glad you
asked.) A variable is a storage unit that holds a value. Normally the
storage unit is known by a name, and may have an associated "data type" as
well. The data type is completely optional for my definition as some
languages are "type independent". A variables value is normally mutable and
can contain three basic conditions. (undefined[does not exist therefore can
not have a value], NULL[exists, but has no value] and NOT NULL[exists and
has a value]) The exact implementation of what NOT NULL means depends on
the data type of the variable. (If a language is "type independent" then
the variables still have a type, it is simply all the same type and is
language specific.) BTW: ARS is a "type dependent" language. (ARS fields
come in "Data Types" that are ARS constructs like: Character, Integer,
Real, Decimal, Attachment, Flashboard, etc...
ARS also provides "rules" about how the objects are combined to make
a "larger object" (form/Application with more meaning. These rules are
enforced at the API level, however they manifest in the IDE as
"options"/"settings" for the ARS objects.


To me the value of ARS is that they have created objects that make
application development much faster (and easier) than starting at first
principals.

However, that can be a two sided sword. Because the down side is that
you have to use ARS objects. (At this point ARS is only marginally
extendable. The ARF API/ ARDBC API opens up a good deal of extendibility.
However there is much more that is needed before I would declare ARS as
"extensible". For now I would have to define ARS as able to accept
"plug-ins". (A Plug-in has to meet a defined standard interface. While an
extension (IMHO, a requirement to declare a language as be an extensible)
can encapsulate and/or add to existing interfaces of the language. One adds
new "features" to an existing interface. The other "alters" the base
interface/language.)

As a finishing note:
http://dictionary.reference.com/search?q=fourth%20generation%20language:
has the following at the end of it's description.

"

Some other successful 4th-generation languages are: database
query languages, e.g. SQL; Focus, Metafont,
PostScript, RPG-II, {S}, IDL-PV/WAVE, Gauss,
Mathematica and data-stream languages such as AVS,
APE, Iris Explorer.


"

If SQL is a language then ARS is a language. (Especially if you simply view
ARS as a "subset"/"superset" of SQL. I think it is much more than that, but
it is definitely at least that.)

--
Carey Matthew Black
Remedy Skilled Professional (RSP)
ARS=Action Request System(Remedy)

Solution = People + Process + Tools
Fast, Accurate, Cheap.... Pick two.
Never ascribe to malice, that which can be
explained by incompetence.



"William H.
\"Will\" Du Chene" To:
ARSLIST@ARSLIST.ORG
RELESS.COM> Subject: Re: Remedy Vs
VB/Oracle Forms
Sent by: "Action
Request System
discussion
list(ARSList)"
G>


02/07/2003 09:53 PM
Please respond to
arslist






Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with
a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type
of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will
allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use
it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters
-
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I
can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime
to
master."


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against
using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Thanks, Bing and Guillaume,

We are keeping track of lots of things in our metrics, including the ones
you suggested below, Guillaume. I just wondered if anyone has been trying
to perform function point analysis for Remedy. We have been keeping track
of these metrics since November 1st, so we're gathering historical data.
I'm asking these questions of the developer community now because we have
enough historical data at this point to suggest our estimates are a wee off.
I had hoped that in the 2000+ readers of the ARSlist that I wasn't the first
person to tackle this beast, but it's starting to look as if I may be. :-/

Thanks for the feedback.

Jennifer

-----Original Message-----
From: Guillaume.Rheault [mailto:Guillaume.Rheault@TARGET.COM]
Sent: Monday, February 10, 2003 15:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Jennifer,

To the best of my knowledge, not even Remedy has come up with a good enough
scientific method to estimate development efforts. Besides the pure
technical challenge, you need to consider the following factors:

- Quality of requirements and design documentation provided to the developer
- Overall experience of the developer
- Particular experience of the developer in the task at hand

I don't want to discourage you in your efforts to come up with a formula,
but in many cases, history data is the best thing you can rely on.

I believe in order to come up with a comprehensive formula, I think you
should somehow incorporate the three points I mentioned, in some kind of
factor

Guillaume


-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Monday, February 10, 2003 10:41
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Thanks to all who replied.

Whether Remedy is or is not a "language" was not the point of my earlier
post. My sincerest apologies for offending the programming purists.

I'm trying to gather metrics on our development effort. In order to compare
our metrics meaningfully to the metrics of another project, say a project
that is programming with a "real" language, I am using Logical Source Lines
of Code (LSLOC) to establish a common ground.

Has anyone else attempted to gather metrics on their Remedy development? If
so, how are you going about it? Our team is assigning a difficulty level to
each piece of workflow, multiplying it by a constant for that piece of
workflow, and figuring in a complexity factor for the level of the
programming language.

For example, I create active link ZZZMyActiveLink. It is a very complex
active link, with a long qualification string and 15 actions, so I assign it
a complexity factor of 17. Active links have a constant factor of 9. So
9x17=153. Since we have assumed that Remedy is a 4th generation "language,"
we have used a programming language conversion factor of 20 and an overall
complexity factor of 1.07 for our project. So ZZZMyActiveLink equates to
3274 LSLOC.

Doing the same thing in C, a third generation language with a programming
language conversion factor of 15, should give me the same number of LSLOC,
because it would require more lines of code in C.

Now that I've explained the nature of my inquiry more concisely, can anyone
provide feedback on our method?

Many thanks,

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516




-----Original Message-----
From: Carey Matthew Black [mailto:matt.black@VERIZON.COM]
Sent: Monday, February 10, 2003 09:44
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


JMHO,

I have read all of the post on this thread through Theo Fondse's and
Jonathan Hall's, and I think there is a bit of a misnomer in the response
below. The reply below implies that ARS development (or coding if you
prefer) is "easy". While I agree that it is simple to create an
application, I would contend that there is a big difference between a "good
application" and "any application". (Jonathan's email speaks to some of the
design/implementation choices in ARS.)

More to the topic at hand....
Some languages are "easy" because the syntax is very forgiving(allow
just about anything), and some languages are easy because they offer so few
options. (AKA: one way to do a task and it is obvious how to do that.) On
the other hand some IDE's make some languages much easier. That is a
function of the IDE and how well it maps the programming language
constructs into human behavioral patterns (IMHO.) and not representative of
the underlaying language.

I am not from the classical Computer Science background. (In fact I
come from a BS in Mechanical Engineering.) So, my "definitions" are very
week about what makes a programming language. However, I also think the
specific example of "parsing a line from a file" and "socket programming"
are poor tests to use to determine the validity of a programming language.
(Just take a look at some of the first programming languages for example.
Or even Assembler, Or MS Batch scripts.... or GW Basic... The list goes
on... Do you really think that those specific tasks were easy/do able in
those languages?)
I do program in several languages and have found that each language
has it's own strengths. (Just like spoke language.) Some are faster than
others, some are more compact, and others still are "easier to learn".

William does have a perspective. However I want to try to add some flavor
to his statements.

I view the Admin tool (and parts of the User tool) as an IDE. (An IDE
specific to the ARS "object language".) The point of the Admin tool is to
avoid syntax/compiler problems. It will not let you define an ARS object
that is "invalid". (Actually the C API really does that job. However the
IDE provides a GUI to the API. So I view them as one and the same. The
difference is merely a translation of "coding interface" to "GUI
interface". True, the GUI requires less understanding and grants less
control than the coding interface, but it is also much faster and less
syntax error prone to develop in too. But I digress....) The Admin tool
allows us to think in terms of ARS objects (Forms, Active Links, Menus,
Filters, Groups, etc...) instead of the language that implements the
"coding interface". These constructs do not map well to any "programming
language" as best I can tell. After all an object is an encapsulated
"thing". The "thing" could be anything from a socket connection, to a
Integer variable, to a reference to another object.


IMHO:

Think in terms of the "English language". A variable is a single
letter of the alphabet. It has a value of a "set of sounds" and is a member
of one of two types. (consent, vowel) Multiple variables combined together
produce a "word". Which is a language specific object. That object has a
greater meaning that the "sum" of the variables that compose the object.
String together a set of those objects and you get a sentence. Again a new
language specific object with a greater meaning than the sum of the
components. There are rules that define how to combine the variables and
the objects. (Grammar) So anything that exhibits those traits would be a
"language" in my book.

The bottom line for me to decide if something is a "programming
language" is based being able to derive a "variable construct". IMHO,
anything that has a "variable" is a programming language. (Yes, I think
Mathematics really is a programming language too. However it is more
abstract and compact language than any "human or computer based" language
that I know. Again, JMHO. :) Without a variable you can not have a
programming language. As you have no way to distinguish between multiple
"things" that have value. Ideally any language has a grammar that also
specifies how variables relate to each other and how they can be combined
to make "greater meanings".

To that end, I see ARS as having variables. They are more commonly
known as "ARS fields". Why do I think an ARS field is a variable? (Glad you
asked.) A variable is a storage unit that holds a value. Normally the
storage unit is known by a name, and may have an associated "data type" as
well. The data type is completely optional for my definition as some
languages are "type independent". A variables value is normally mutable and
can contain three basic conditions. (undefined[does not exist therefore can
not have a value], NULL[exists, but has no value] and NOT NULL[exists and
has a value]) The exact implementation of what NOT NULL means depends on
the data type of the variable. (If a language is "type independent" then
the variables still have a type, it is simply all the same type and is
language specific.) BTW: ARS is a "type dependent" language. (ARS fields
come in "Data Types" that are ARS constructs like: Character, Integer,
Real, Decimal, Attachment, Flashboard, etc...
ARS also provides "rules" about how the objects are combined to make
a "larger object" (form/Application with more meaning. These rules are
enforced at the API level, however they manifest in the IDE as
"options"/"settings" for the ARS objects.


To me the value of ARS is that they have created objects that make
application development much faster (and easier) than starting at first
principals.

However, that can be a two sided sword. Because the down side is that
you have to use ARS objects. (At this point ARS is only marginally
extendable. The ARF API/ ARDBC API opens up a good deal of extendibility.
However there is much more that is needed before I would declare ARS as
"extensible". For now I would have to define ARS as able to accept
"plug-ins". (A Plug-in has to meet a defined standard interface. While an
extension (IMHO, a requirement to declare a language as be an extensible)
can encapsulate and/or add to existing interfaces of the language. One adds
new "features" to an existing interface. The other "alters" the base
interface/language.)

As a finishing note:
http://dictionary.reference.com/search?q=fourth%20generation%20language:
has the following at the end of it's description.

"

Some other successful 4th-generation languages are: database
query languages, e.g. SQL; Focus, Metafont,
PostScript, RPG-II, {S}, IDL-PV/WAVE, Gauss,
Mathematica and data-stream languages such as AVS,
APE, Iris Explorer.


"

If SQL is a language then ARS is a language. (Especially if you simply view
ARS as a "subset"/"superset" of SQL. I think it is much more than that, but
it is definitely at least that.)

--
Carey Matthew Black
Remedy Skilled Professional (RSP)
ARS=Action Request System(Remedy)

Solution = People + Process + Tools
Fast, Accurate, Cheap.... Pick two.
Never ascribe to malice, that which can be
explained by incompetence.



"William H.
\"Will\" Du Chene" To:
ARSLIST@ARSLIST.ORG
RELESS.COM> Subject: Re: Remedy Vs
VB/Oracle Forms
Sent by: "Action
Request System
discussion
list(ARSList)"
G>


02/07/2003 09:53 PM
Please respond to
arslist






Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with
a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type
of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will
allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use
it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters
-
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I
can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime
to
master."


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against
using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Matt, Jonathan:
You have some valid points. My apologies if I offended anyone by my
reference to "true computer programmer". That was not the intent.
I believe that 4GL's have made application development much easier which
allowed non-programmers to also be able to put applications together and
therefore the effort behind writing a COBOL, C++, Java Applet or even VB
application is diminished, generalised and not appreciated.
That is why I believe that there should be a distinction between a 4GL
(e.g.: Remedy) developer and a 3GL (e.g. C++) programmer because the
skillset, natural interest and -temperaments required are different. This is
not to say that one is a lesser person/professional by virtue of being a
Remedy developer. One can also be both depending on which hat you put on.
Like you, I am both.

I also believe that those who have been the victim of a single misplaced or
missing semicolon in a C++ program or a missing single full stop in the
working storage section of a COBOL program causing 500+ errors after
compiling the source code are best equipped to fully appreciate the
advantages that comes with Remedy via it's 4GL interface.

I want to drive the point home that Remedy is a "Language" per say. The
difference is that it is (IMHO) a Fourth Generation Language (4GL). It saves
a great deal of time and effort to produce a working, neat and usable system
for a specific purpose. Take your own current Remedy implementation as an
example. Just imagine recreating the entire system using C++ or COBOL.....?
:-)


Remedy is my "Language of choice"



Best Regards,
Theo J. Fondse
Remedy Developer
Siemens Business Services

Telephone: +27 11 652 7477
Mobile: +27 83 773 3075
e-Mail: theo.fondse@ars.smc.siemens.co.za
Web: http://www.siemens.co.za/
===========================================================================


-----Original Message-----
From: Jonathan Hall [mailto:Jonathan.Hall@YELLGROUP.COM]
Sent: Monday, February 10, 2003 6:10 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


The point about extendability is an interesting one. I personally find it
an enjoyable challenge to push the boundaries of ARS and other
environments, while NOT using any extensions. I frequently see perfectly
good suggested solutions to questions on this loop, which use SQL or
similar to achieve their ends, yet can see an all-ARS way of doing the same
thing.

I inherited a 3-year old ARS system several years ago, and one day will
possibly pass it on to someone else. Using ARS forms and workflow (great
point about variables below) for the solution will tend to allow better
maintanability and reliability (less external components to cause
problems). Having used Acorn's range of computers through the late
eighties and early nineties, I seem to have picked up a "reduced
instruction set" mentality which I apply to the likes of Remedy and Access
(why use VBA when an Access macro will do?).

The use of external code through the API is obviously highly appropriate in
certain areas, like customisation of interfaces and integration with other
applications at an API level, and I've seen some very well thought out
products and solutions in this area (hello UWIP!).

But as I've said, the point about system variables below is a great one.
There is little that can't be coded in Remedy with a bit of thought, and by
maximising the use of Remedy's own code, you can build a very maintainable
and adaptable system. I've done enough C coding in my shortish career
(academic and professional) to know that it can be a complete pain in the
backside, from attempting to translate someone else's shonky code, to
working out which misplaced semicolon has caused the 5400 errors in your
compiler report. I'd personally rather get on with the programming and
ignore the semantic details. I fully regard myself as a programmer, and
the fact that I'm using Remedy rather than Oracle/Forms or C++ or any other
environment doesn't cheapen that in my opinion.

Jonathan Hall
Remedy Systems Manager, Yell Ltd.






Carey Matthew Black @ARSLIST.ORG> on 10/02/2003
15:43:59

To: ARSLIST@ARSLIST.ORG
cc:
bcc:
From: "Action Request System discussion list(ARSList)"



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

No offense to those that have tried, but I personally don't think it
feasible to define one's development based on an objective, mathematical
computation such as Logical Lines of Code. And not because I think the
ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
to ARSystem development. Don't get me wrong, I like Math. But I can develop
a solution using many workflow objects, or few. And my approach can dictate
the complexity of the workflow.

For example, I could create several character menus for an application, or I
could create search menus, and store the menu data in a form. I could
create one large form with multiple fields and functions, or several forms
with workflow that push data between them. As a more real example, look at
how the HelpDesk application handles automatic assignment - there are
multiple forms and a passel of links and filters involved. But most of us
can look at that and come up with what we consider a "better" way, with
fewer, less complicated workflow.

So when attempting to define how well a developer or team is doing, I think
you have to look at subjective means, such as how well they define
requirements, or functional specifications. Or look at development time, or
satisfaction. (Someone came up with the requirements, so how well are they
satisfied with the result?) How many "redo's"? How many milestones or
deadlines do they miss, etc.

I hate to be the negative one, but this is application development; It's a
dark art.

-Drake

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Monday, February 10, 2003 11:41 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Thanks to all who replied.

Whether Remedy is or is not a "language" was not the point of my earlier
post. My sincerest apologies for offending the programming purists.

I'm trying to gather metrics on our development effort. In order to compare
our metrics meaningfully to the metrics of another project, say a project
that is programming with a "real" language, I am using Logical Source Lines
of Code (LSLOC) to establish a common ground.

Has anyone else attempted to gather metrics on their Remedy development? If
so, how are you going about it? Our team is assigning a difficulty level to
each piece of workflow, multiplying it by a constant for that piece of
workflow, and figuring in a complexity factor for the level of the
programming language.

For example, I create active link ZZZMyActiveLink. It is a very complex
active link, with a long qualification string and 15 actions, so I assign it
a complexity factor of 17. Active links have a constant factor of 9. So
9x17=153. Since we have assumed that Remedy is a 4th generation "language,"
we have used a programming language conversion factor of 20 and an overall
complexity factor of 1.07 for our project. So ZZZMyActiveLink equates to
3274 LSLOC.

Doing the same thing in C, a third generation language with a programming
language conversion factor of 15, should give me the same number of LSLOC,
because it would require more lines of code in C.

Now that I've explained the nature of my inquiry more concisely, can anyone
provide feedback on our method?

Many thanks,

Jennifer Spaeth
SAIC
jennifer.spaeth@afwa.af.mil
402.294.7516




-----Original Message-----
From: Carey Matthew Black [mailto:matt.black@VERIZON.COM]
Sent: Monday, February 10, 2003 09:44
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


JMHO,

I have read all of the post on this thread through Theo Fondse's and
Jonathan Hall's, and I think there is a bit of a misnomer in the response
below. The reply below implies that ARS development (or coding if you
prefer) is "easy". While I agree that it is simple to create an
application, I would contend that there is a big difference between a "good
application" and "any application". (Jonathan's email speaks to some of the
design/implementation choices in ARS.)

More to the topic at hand....
Some languages are "easy" because the syntax is very forgiving(allow
just about anything), and some languages are easy because they offer so few
options. (AKA: one way to do a task and it is obvious how to do that.) On
the other hand some IDE's make some languages much easier. That is a
function of the IDE and how well it maps the programming language
constructs into human behavioral patterns (IMHO.) and not representative of
the underlaying language.

I am not from the classical Computer Science background. (In fact I
come from a BS in Mechanical Engineering.) So, my "definitions" are very
week about what makes a programming language. However, I also think the
specific example of "parsing a line from a file" and "socket programming"
are poor tests to use to determine the validity of a programming language.
(Just take a look at some of the first programming languages for example.
Or even Assembler, Or MS Batch scripts.... or GW Basic... The list goes
on... Do you really think that those specific tasks were easy/do able in
those languages?)
I do program in several languages and have found that each language
has it's own strengths. (Just like spoke language.) Some are faster than
others, some are more compact, and others still are "easier to learn".

William does have a perspective. However I want to try to add some flavor
to his statements.

I view the Admin tool (and parts of the User tool) as an IDE. (An IDE
specific to the ARS "object language".) The point of the Admin tool is to
avoid syntax/compiler problems. It will not let you define an ARS object
that is "invalid". (Actually the C API really does that job. However the
IDE provides a GUI to the API. So I view them as one and the same. The
difference is merely a translation of "coding interface" to "GUI
interface". True, the GUI requires less understanding and grants less
control than the coding interface, but it is also much faster and less
syntax error prone to develop in too. But I digress....) The Admin tool
allows us to think in terms of ARS objects (Forms, Active Links, Menus,
Filters, Groups, etc...) instead of the language that implements the
"coding interface". These constructs do not map well to any "programming
language" as best I can tell. After all an object is an encapsulated
"thing". The "thing" could be anything from a socket connection, to a
Integer variable, to a reference to another object.


IMHO:

Think in terms of the "English language". A variable is a single
letter of the alphabet. It has a value of a "set of sounds" and is a member
of one of two types. (consent, vowel) Multiple variables combined together
produce a "word". Which is a language specific object. That object has a
greater meaning that the "sum" of the variables that compose the object.
String together a set of those objects and you get a sentence. Again a new
language specific object with a greater meaning than the sum of the
components. There are rules that define how to combine the variables and
the objects. (Grammar) So anything that exhibits those traits would be a
"language" in my book.

The bottom line for me to decide if something is a "programming
language" is based being able to derive a "variable construct". IMHO,
anything that has a "variable" is a programming language. (Yes, I think
Mathematics really is a programming language too. However it is more
abstract and compact language than any "human or computer based" language
that I know. Again, JMHO. :) Without a variable you can not have a
programming language. As you have no way to distinguish between multiple
"things" that have value. Ideally any language has a grammar that also
specifies how variables relate to each other and how they can be combined
to make "greater meanings".

To that end, I see ARS as having variables. They are more commonly
known as "ARS fields". Why do I think an ARS field is a variable? (Glad you
asked.) A variable is a storage unit that holds a value. Normally the
storage unit is known by a name, and may have an associated "data type" as
well. The data type is completely optional for my definition as some
languages are "type independent". A variables value is normally mutable and
can contain three basic conditions. (undefined[does not exist therefore can
not have a value], NULL[exists, but has no value] and NOT NULL[exists and
has a value]) The exact implementation of what NOT NULL means depends on
the data type of the variable. (If a language is "type independent" then
the variables still have a type, it is simply all the same type and is
language specific.) BTW: ARS is a "type dependent" language. (ARS fields
come in "Data Types" that are ARS constructs like: Character, Integer,
Real, Decimal, Attachment, Flashboard, etc...
ARS also provides "rules" about how the objects are combined to make
a "larger object" (form/Application with more meaning. These rules are
enforced at the API level, however they manifest in the IDE as
"options"/"settings" for the ARS objects.


To me the value of ARS is that they have created objects that make
application development much faster (and easier) than starting at first
principals.

However, that can be a two sided sword. Because the down side is that
you have to use ARS objects. (At this point ARS is only marginally
extendable. The ARF API/ ARDBC API opens up a good deal of extendibility.
However there is much more that is needed before I would declare ARS as
"extensible". For now I would have to define ARS as able to accept
"plug-ins". (A Plug-in has to meet a defined standard interface. While an
extension (IMHO, a requirement to declare a language as be an extensible)
can encapsulate and/or add to existing interfaces of the language. One adds
new "features" to an existing interface. The other "alters" the base
interface/language.)

As a finishing note:
http://dictionary.reference.com/search?q=fourth%20generation%20language:
has the following at the end of it's description.

"

Some other successful 4th-generation languages are: database
query languages, e.g. SQL; Focus, Metafont,
PostScript, RPG-II, {S}, IDL-PV/WAVE, Gauss,
Mathematica and data-stream languages such as AVS,
APE, Iris Explorer.


"

If SQL is a language then ARS is a language. (Especially if you simply view
ARS as a "subset"/"superset" of SQL. I think it is much more than that, but
it is definitely at least that.)

--
Carey Matthew Black
Remedy Skilled Professional (RSP)
ARS=Action Request System(Remedy)

Solution = People + Process + Tools
Fast, Accurate, Cheap.... Pick two.
Never ascribe to malice, that which can be
explained by incompetence.



"William H.
\"Will\" Du Chene" To:
ARSLIST@ARSLIST.ORG
RELESS.COM> Subject: Re: Remedy Vs
VB/Oracle Forms
Sent by: "Action
Request System
discussion
list(ARSList)"
G>


02/07/2003 09:53 PM
Please respond to
arslist






Oy!

Ok folks, I think that before you can define what generation a language is,
that you first have to define that something *is* a programming language.

This being said, I think that it is safe to say that the rest of my message
here is more techni-rant than anything else - yet it is a techni-rant with
a
point.

Depending on whose interpretation that you happen to lean towards, exactly
what constitutes a language per se is somewhat open to intreptation. As far
as I understand the theory, there are two primary branches to what the
industry terms as a "programming language." Each branch of the language
family is subdivided further ( It would be, of course ).

The first is called an FPL, or Functional Programming Language. This type
of
language is based upon the composition of functions to solve problems. In
such a language, the functions are relational, meaning that when the
functions are invoked, they recieve and return values to parent which
invoked them.

The second type of language is known as a DPL, or Declartive Programming
Language. This division of languages is geared toward static, unchanging,
ever-so-stolid rules on what needs to be processed. In short, the two
divisions can be tought of like this : DPL's describe what to process and
FPL's describe how to process it.

Admittedly, I am over simplifying the entire debate ( and my definitions )
but the point is still the same : Where does the AR System fit into any of
this? If I am "writing code" in the admin tool, am I really "programming?"

If we define a language - still further - by it's extensibility - ergo it's
applicability to specific progamming tasks - how well does the AR System
compare? Does an "active link" really provide me with a method of doing
something as simple as opening a text file and reading the first few lines?
Well... No it doesn't. Is there some "action" in a "filter" which will
allow
me to open up a socket and recieve results from a process on a server?
Sorry, but again the answer is no. I am left without a "language" specific
option for doing this. Where is the "code?" Where is the syntax of the
language?

To accomplish all of these things and more, I have to step outside the AR
System and put something together in *another language* so that I can use
it
in *conjunction* with the AR System.

IMHO, writing code is characterized by the painstaking act of writing (
many ) lines of code that are syntactically correct in an accepted,
extensible language. ( That and loosing hair, sleep and ones sanity after a
14 hour stint in front of a C++ IDE on a project due yesterday. ) Normally,
( with a few exceptions ) mastering a programming language takes much time.
Somehow the act of stringing together a series of active links and filters
-
again IMHO - does not automatically make one a computer programmer. A
developer, most certainly, but not a programmer.

Don't get me wrong here, I believe to the AR System to be a *wonderful*
tool. I recommend it's use to everyone and it is my development platform of
choice. I have built and deployed a number of *solutions* ( not programs )
based upon it, and hope to keep working with it for many years to come. I
just seem to have a slight difference of opinion here.

Will Du Chene
william.du.chene@state.mn.us

AR System 4.5
Sybase 11.9.2

PS : I post this to the list only because it is a matter of convience. I
can
get all the interested parties in one shot. Flame / spam me for it off the
list, please.

"C and C++ are an artform : They may take a week to learn, but a lifetime
to
master."


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Raj Rajakumar
Sent: Friday, February 07, 2003 8:02 AM
To: ARSLIST@ARSLIST.ORG
Subject: Remedy Vs VB/Oracle Forms

Hi Listers,
I would appreciate your input to the following question:
"What advantage does Remedy offer in developing an application against
using
something like Visual Basic or Oracle Forms."
Thank you for your input.
Raj




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

I think the discussion that resulted from where Remedy stands in contrast
to other programming languages is quite interesting.

Everyone who commented on this issue (Theo Fondse, Jonathan Hall, JMHO , Mr
"BING")
is correct on the issues pointed out.

I especially agree with Jonathan in that Remedy should be procedural. One
can
superimpose programming methedologies within Remedy, but before one can do
this you must know what they are and have had first hand experience in using
them.

My official title is a Remedy Consultant, but if it was not for my prior
exposure in languages ranging from Assember > Pascal > C
> Cobol > Delphi etc etc I don't think I would have been able to solve
half of the challanges I have faced over the last 7 years implementing
"pure" Remedy solutions.

I have worked with Remedy Developers (4+ years experience) who look cross
eye at you when you start to explain a sertain design approach that will
save bandwidth over a network because it will save on the amount of
bytes flying around...or why I'm trying to squeeze execution time of
a mainframe api down to .02 seconds, why is 5 seconds not enough ? Then
you have to explain what impact 5 seconds will have when you process 10 000+
per day and what effect it will have on the sever performance.

My college tutor always told us. We want to make out of you programmers,
no C or VB or Cobol programmers, but just programmers. This means you
should be able to use any tool (language) BEST Suited for the job.

I firmly believe if you decide to use only one tool (language incl Remedy)
to solve problems it was not intended for, you are making life difficult
for yourself...not to mention those poor people who must take over from you.

Personally I have spend far too many nights trying to figure out how a
Remedy system was thrown together for there is no proper documentation,
no proper flow diagrams and the trigger spagetti code will make the
chaos theory look organised (Ok, maybe not THAT bad)

I still code my solutions on paper first, work out all the gotchas and of
coarse, it is based on how Remedy can handle "lowlevel stuff" etc. What a
joy
to implement it and then have it work close to what was aimed for first
time...and I use to tell my lecturer that designing was a waste of time was
very ignorant back then>

Maybe it's because IT got to be something you do because their *was* a lot
of money in it, and not because it was that person's driving passion to
solve problems.

Timelimits on projects is another major headache !!

Well, just my 2c

Regards,

- Johann Groenewald
CRM Consultant

Fusion Business Solutions
Ireland







-----Original Message-----
From: Fondse, Theo [mailto:Theo.Fondse@ARS.SMC.SIEMENS.CO.ZA]
Sent: 11 February 2003 08:15
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms
Importance: High


Matt, Jonathan:
You have some valid points. My apologies if I offended anyone by my
reference to "true computer programmer". That was not the intent.
I believe that 4GL's have made application development much easier which
allowed non-programmers to also be able to put applications together and
therefore the effort behind writing a COBOL, C++, Java Applet or even VB
application is diminished, generalised and not appreciated.
That is why I believe that there should be a distinction between a 4GL
(e.g.: Remedy) developer and a 3GL (e.g. C++) programmer because the
skillset, natural interest and -temperaments required are different. This is
not to say that one is a lesser person/professional by virtue of being a
Remedy developer. One can also be both depending on which hat you put on.
Like you, I am both.

I also believe that those who have been the victim of a single misplaced or
missing semicolon in a C++ program or a missing single full stop in the
working storage section of a COBOL program causing 500+ errors after
compiling the source code are best equipped to fully appreciate the
advantages that comes with Remedy via it's 4GL interface.

I want to drive the point home that Remedy is a "Language" per say. The
difference is that it is (IMHO) a Fourth Generation Language (4GL). It saves
a great deal of time and effort to produce a working, neat and usable system
for a specific purpose. Take your own current Remedy implementation as an
example. Just imagine recreating the entire system using C++ or COBOL.....?
:-)


Remedy is my "Language of choice"



Best Regards,
Theo J. Fondse
Remedy Developer
Siemens Business Services

Telephone: +27 11 652 7477
Mobile: +27 83 773 3075
e-Mail: theo.fondse@ars.smc.siemens.co.za
Web: http://www.siemens.co.za/
===========================================================================


-----Original Message-----
From: Jonathan Hall [mailto:Jonathan.Hall@YELLGROUP.COM]
Sent: Monday, February 10, 2003 6:10 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


The point about extendability is an interesting one. I personally find it
an enjoyable challenge to push the boundaries of ARS and other
environments, while NOT using any extensions. I frequently see perfectly
good suggested solutions to questions on this loop, which use SQL or
similar to achieve their ends, yet can see an all-ARS way of doing the same
thing.

I inherited a 3-year old ARS system several years ago, and one day will
possibly pass it on to someone else. Using ARS forms and workflow (great
point about variables below) for the solution will tend to allow better
maintanability and reliability (less external components to cause
problems). Having used Acorn's range of computers through the late
eighties and early nineties, I seem to have picked up a "reduced
instruction set" mentality which I apply to the likes of Remedy and Access
(why use VBA when an Access macro will do?).

The use of external code through the API is obviously highly appropriate in
certain areas, like customisation of interfaces and integration with other
applications at an API level, and I've seen some very well thought out
products and solutions in this area (hello UWIP!).

But as I've said, the point about system variables below is a great one.
There is little that can't be coded in Remedy with a bit of thought, and by
maximising the use of Remedy's own code, you can build a very maintainable
and adaptable system. I've done enough C coding in my shortish career
(academic and professional) to know that it can be a complete pain in the
backside, from attempting to translate someone else's shonky code, to
working out which misplaced semicolon has caused the 5400 errors in your
compiler report. I'd personally rather get on with the programming and
ignore the semantic details. I fully regard myself as a programmer, and
the fact that I'm using Remedy rather than Oracle/Forms or C++ or any other
environment doesn't cheapen that in my opinion.

Jonathan Hall
Remedy Systems Manager, Yell Ltd.






Carey Matthew Black @ARSLIST.ORG> on 10/02/2003
15:43:59

To: ARSLIST@ARSLIST.ORG
cc:
bcc:
From: "Action Request System discussion list(ARSList)"



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can develop
> a solution using many workflow objects, or few. And my approach can dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time, or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606

LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated because
of the exacting nature of their work tasks. What kind of mystery is that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of
our productivity using well-defined industry standards, not some qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in
1976, (see Brad Bingel's post earlier in this thread) surely we can do at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time,
or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Sounds like we've found our author! Though some good points have been made by all, I do agree that programming in Remedy isn't all that hard to do at a low or even mediocre level. Doing it really well does take some skill and smarts, but it isn't brain surgery.

I'd love to see the quantification of the work complexity that you talked about. It would make for some interesting ROI discussions, both at an architectural and project level; for instance, what true value is there in being able to quantify the complexity of the work?

Rick

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated because
of the exacting nature of their work tasks. What kind of mystery is that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of
our productivity using well-defined industry standards, not some qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in
1976, (see Brad Bingel's post earlier in this thread) surely we can do at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time,
or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Wow, flame retardant suit on. I did not think that my original response
to this questions would illicit such a response. My original intent was
to illustrate a point, to put a framework to wrap one's mind around
Remedy as a solution development platform. Sure, it's easy to 'code' in
remedy. But just like SQL, it's real easy write BAD stuff. All
software development efforts can, and should, be measured for
productivity and ROI. Remedy, Delphi, PowerBuilder, JAVA, whatever.
There are products that run through code to measure quality. One such
product is Jauditor. It will spit out pages of statistics on variable
usage, memory, etc., and give a rating for JAVA code.

Peter McMullen
QMX Support Services
Voice: 250.701.3748


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Spaeth Jennifer Contractor HQ
AFWA
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms

LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated
because
of the exacting nature of their work tasks. What kind of mystery is
that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure
of
our productivity using well-defined industry standards, not some
qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement
in
1976, (see Brad Bingel's post earlier in this thread) surely we can do
at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective,
mathematical
> computation such as Logical Lines of Code. And not because I think
the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much
subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an
application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several
forms
> with workflow that push data between them. As a more real example,
look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most
of us
> can look at that and come up with what we consider a "better" way,
with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development
time,
or
> satisfaction. (Someone came up with the requirements, so how well are
they
> satisfied with the result?) How many "redo's"? How many milestones
or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES
at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com

I'm not trying to flame you, I'm just getting frustrated. Sorry if I came
on a little strong. :(

Jennifer

-----Original Message-----
From: Peter McMullen [mailto:pmcmullen@QMXS.COM]
Sent: Tuesday, February 11, 2003 13:31
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Wow, flame retardant suit on. I did not think that my original response
to this questions would illicit such a response. My original intent was
to illustrate a point, to put a framework to wrap one's mind around
Remedy as a solution development platform. Sure, it's easy to 'code' in
remedy. But just like SQL, it's real easy write BAD stuff. All
software development efforts can, and should, be measured for
productivity and ROI. Remedy, Delphi, PowerBuilder, JAVA, whatever.
There are products that run through code to measure quality. One such
product is Jauditor. It will spit out pages of statistics on variable
usage, memory, etc., and give a rating for JAVA code.

Peter McMullen
QMX Support Services
Voice: 250.701.3748


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Spaeth Jennifer Contractor HQ
AFWA
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms

LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated
because
of the exacting nature of their work tasks. What kind of mystery is
that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure
of
our productivity using well-defined industry standards, not some
qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement
in
1976, (see Brad Bingel's post earlier in this thread) surely we can do
at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective,
mathematical
> computation such as Logical Lines of Code. And not because I think
the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much
subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an
application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several
forms
> with workflow that push data between them. As a more real example,
look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most
of us
> can look at that and come up with what we consider a "better" way,
with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development
time,
or
> satisfaction. (Someone came up with the requirements, so how well are
they
> satisfied with the result?) How many "redo's"? How many milestones
or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES
at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Indeed, but we know different, right? I think it's a great database tool,
period . . .IMHO. :)

Jennifer

-----Original Message-----
From: Guillaume.Rheault [mailto:Guillaume.Rheault@TARGET.COM]
Sent: Tuesday, February 11, 2003 16:41
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


I am not against the Remedy applications by any means.
However, in many occasions, I never heard the Remedy salesmen
making/stressing the point that Remedy is an application development
platform...
As a matter of fact, most non-Remedy people out there think of it as a "help
desk tool"

Guillaume

I am not against the Remedy applications by any means.
However, in many occasions, I never heard the Remedy salesmen
making/stressing the point that Remedy is an application development
platform...
As a matter of fact, most non-Remedy people out there think of it as a "help
desk tool"

Guillaume


-----Original Message-----
From: Rick Cook [mailto:rcook@DENALIAI.COM]
Sent: Tuesday, February 11, 2003 03:32
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Though I think that Doug, et al, did to at least some degree know what they
had early on, applications are sold more because although techies like us
can see the value in a development platform, the people that write the
checks need (most of the time) to see something they can understand before
they'll approve the money. It also looks good to be able to say that your
company (meaning Remedy) has some product features that can be benchmarked
against competitors.

Sort of like Pilsbury selling flour to those who wish to bake their own
cakes from scratch, and cake mixes to those who lack the skill or the
inclination to make from scratch. Both are valuable to different market
segments and therefore increase profitability.

Rick

-----Original Message-----
From: Guillaume.Rheault [mailto:Guillaume.Rheault@TARGET.COM]
Sent: Tuesday, February 11, 2003 1:13 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


>Did the original developers
>know what they had? I don't think so. I think they found out by accident.

I think that is too bold a statement...
I believe a small core group at Remedy had their vision straight, and new
the end game of this product. However, you need to start somewhere,
marketing wise. So they targeted the help desk market first...

But I agree with you that many many people at Remedy itself (specially the
sales force, way too focused on selling Remedy applications licenses) didn't
really understand the vision up until recently, thanks to BMC efforts on
making people understand what Remedy is all about

Guillaume

-----Original Message-----
From: Frank, Gordon M. [mailto:Gordon.Frank@JHUAPL.EDU]
Sent: Tuesday, February 11, 2003 02:54
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


I consider the Remedy ARS toolset a 4.5GL. In the traditional sense, it is
not a 5GL because to me a 5GL implies some Artificial Intelligence
mechanism.

One person I work with described Remedy ARS with this analogy. In the past
automobile manufacturers put "lug nuts" on cars by hand. As automation came
around, "lug nuts" were now put on by robots. I view Remedy ARS as a robot.
It allows a developer to do more because he doesn't have to worry about the
primitive operations of putting on a "lug nut" or fighting with a 3GL or
4GLs syntax. I always thought this was an excellent analogy.

One of the reasons I feel Remedy ARS is a 4.5GL, is closes the options to
choose from. If the operation is on the Client, it's probably an Active
Link. If the operation is on the Server, it's probably a Filter. This binary
decision makes development constrained and therefore forces programmers not
try fancy methods. I believe one of the problems of things like PowerBuilder
and other 4GLs, is that you can "object" everything. Remedy ARS stops this,
because you only can do it one way or the other.

It also forces reuse. Because Active Links and Filters are the only items
available (Escalations are timed Filters in from my point of view),
developers concentrate on reusing object on several applications.

Mr. Dijkstra (God rest his soul) would have loved Remedy ARS. Dijkstra
talked about "D-Structures". "D-Stuctures" are the basics of "Structured
Programming" and "Control Structures." A definite beginning and a definite
ending (Begin/End).

Isn't Remedy ARS the perfect "D-Structure?"

If an object needs work, Disable it! Testing and integrating new objects is
very easy with the Remedy ARS system.

Now ask me if the management knew or know what they had done! Remedy ARS was
traditionally a "Help Desk" oriented product. Did the original developers
know what they had? I don't think so. I think they found out by accident.
Recently they have been selling Remedy ARS as a development or workflow
toolkit. I think BMC may be helping with this viewpoint.

That's my two cents,
Gordon M. Frank
Johns Hopkins University Applied Physics Laboratory

-----Original Message-----
From: Rick Cook [mailto:rcook@DENALIAI.COM]
Sent: Tuesday, February 11, 2003 1:52 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Sounds like we've found our author! Though some good points have been made
by all, I do agree that programming in Remedy isn't all that hard to do at a
low or even mediocre level. Doing it really well does take some skill and
smarts, but it isn't brain surgery.

I'd love to see the quantification of the work complexity that you talked
about. It would make for some interesting ROI discussions, both at an
architectural and project level; for instance, what true value is there in
being able to quantify the complexity of the work?

Rick

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated because
of the exacting nature of their work tasks. What kind of mystery is that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of
our productivity using well-defined industry standards, not some qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in
1976, (see Brad Bingel's post earlier in this thread) surely we can do at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time,
or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

Gordon,

Remedy actually started out as a developer's toolkit. In the early days, they sold ARS and you had to design your own use for it. The original idea was to get in the door as a helpdesk tool and then climb the "Stairway to Heaven" by creating other applications. In Remedy itself, they used it in all departments ( Sales, facilities, marketing, etc). As time moved on they decided to create "packaged" applications ( HelpDesk, Change Management, etc). Then they bought Baystone for their CRM product. The Remedy ARS is the base for ALL the applications and still has the ability to be a developer's toolkit.

I have always thought of ARS as an easy way to get data into and out of a database without having to know all intricacies of each Database. I can create a Remedy app that works the same ( to the user) on any DB (Oracle, Sybase, etc).

Roger

-----Original Message-----
From: Frank, Gordon M. [mailto:Gordon.Frank@JHUAPL.EDU]
Sent: Tuesday, February 11, 2003 12:54 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


I consider the Remedy ARS toolset a 4.5GL. In the traditional sense, it is not a 5GL because to me a 5GL implies some Artificial Intelligence mechanism.

One person I work with described Remedy ARS with this analogy. In the past automobile manufacturers put "lug nuts" on cars by hand. As automation came around, "lug nuts" were now put on by robots. I view Remedy ARS as a robot. It allows a developer to do more because he doesn't have to worry about the primitive operations of putting on a "lug nut" or fighting with a 3GL or 4GLs syntax. I always thought this was an excellent analogy.

One of the reasons I feel Remedy ARS is a 4.5GL, is closes the options to choose from. If the operation is on the Client, it's probably an Active Link. If the operation is on the Server, it's probably a Filter. This binary decision makes development constrained and therefore forces programmers not try fancy methods. I believe one of the problems of things like PowerBuilder and other 4GLs, is that you can "object" everything. Remedy ARS stops this, because you only can do it one way or the other.

It also forces reuse. Because Active Links and Filters are the only items available (Escalations are timed Filters in from my point of view), developers concentrate on reusing object on several applications.

Mr. Dijkstra (God rest his soul) would have loved Remedy ARS. Dijkstra talked about "D-Structures". "D-Stuctures" are the basics of "Structured Programming" and "Control Structures." A definite beginning and a definite ending (Begin/End).

Isn't Remedy ARS the perfect "D-Structure?"

If an object needs work, Disable it! Testing and integrating new objects is very easy with the Remedy ARS system.

Now ask me if the management knew or know what they had done! Remedy ARS was traditionally a "Help Desk" oriented product. Did the original developers know what they had? I don't think so. I think they found out by accident. Recently they have been selling Remedy ARS as a development or workflow toolkit. I think BMC may be helping with this viewpoint.

That's my two cents,
Gordon M. Frank
Johns Hopkins University Applied Physics Laboratory

-----Original Message-----
From: Rick Cook [mailto:rcook@DENALIAI.COM]
Sent: Tuesday, February 11, 2003 1:52 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Sounds like we've found our author! Though some good points have been made by all, I do agree that programming in Remedy isn't all that hard to do at a low or even mediocre level. Doing it really well does take some skill and smarts, but it isn't brain surgery.

I'd love to see the quantification of the work complexity that you talked about. It would make for some interesting ROI discussions, both at an architectural and project level; for instance, what true value is there in being able to quantify the complexity of the work?

Rick

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA [mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that machinists are a privileged group whose efforts cannot be duplicated because of the exacting nature of their work tasks. What kind of mystery is that? My grandmother could code in Remedy. (May she rest in peace.) It's a skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of our productivity using well-defined industry standards, not some qualitative seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in 1976, (see Brad Bingel's post earlier in this thread) surely we can do at least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective,
> mathematical computation such as Logical Lines of Code. And not
> because I think the ARSystem is a 3GL, 4GL, or 4.53687GL. There is
> just too much subjectivity to ARSystem development. Don't get me
> wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an
> application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several
> forms with workflow that push data between them. As a more real
> example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most
> of us can look at that and come up with what we consider a "better"
> way, with fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task simply by varying how you go about doing it. My background as a Perl programmer reflects this; I have encountered good and bad coding styles, but the ones that stick in my mind are the fantastically-complicated object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very different opinion would become apparent; the overly-complicated code would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development
> time,
or
> satisfaction. (Someone came up with the requirements, so how well are
> they satisfied with the result?) How many "redo's"? How many
> milestones or deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else even comes close to being useful; deadlines missed can indicate poor developers, OR it can indicate poor management, or merely an especially nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing daniel.r.holdsworth@man.ac.uk 0161 275 0606

Not a problem. The world would be pretty boring without strong
personalities. I should know, I have one too!

Peter McMullen
QMX Support Services
Voice: 250.701.3748


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Spaeth Jennifer Contractor HQ
AFWA
Sent: Tuesday, February 11, 2003 1:05 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms

I'm not trying to flame you, I'm just getting frustrated. Sorry if I
came
on a little strong. :(

Jennifer

-----Original Message-----
From: Peter McMullen [mailto:pmcmullen@QMXS.COM]
Sent: Tuesday, February 11, 2003 13:31
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Wow, flame retardant suit on. I did not think that my original response
to this questions would illicit such a response. My original intent was
to illustrate a point, to put a framework to wrap one's mind around
Remedy as a solution development platform. Sure, it's easy to 'code' in
remedy. But just like SQL, it's real easy write BAD stuff. All
software development efforts can, and should, be measured for
productivity and ROI. Remedy, Delphi, PowerBuilder, JAVA, whatever.
There are products that run through code to measure quality. One such
product is Jauditor. It will spit out pages of statistics on variable
usage, memory, etc., and give a rating for JAVA code.

Peter McMullen
QMX Support Services
Voice: 250.701.3748


-----Original Message-----
From: Action Request System discussion list(ARSList)
[mailto:ARSLIST@ARSLIST.ORG] On Behalf Of Spaeth Jennifer Contractor HQ
AFWA
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms

LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated
because
of the exacting nature of their work tasks. What kind of mystery is
that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure
of
our productivity using well-defined industry standards, not some
qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement
in
1976, (see Brad Bingel's post earlier in this thread) surely we can do
at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective,
mathematical
> computation such as Logical Lines of Code. And not because I think
the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much
subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an
application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several
forms
> with workflow that push data between them. As a more real example,
look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most
of us
> can look at that and come up with what we consider a "better" way,
with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development
time,
or
> satisfaction. (Someone came up with the requirements, so how well are
they
> satisfied with the result?) How many "redo's"? How many milestones
or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES
at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com




UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES
at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org
(Support: mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT
SERVICES at www.QMXS.com

Though I think that Doug, et al, did to at least some degree know what they had early on, applications are sold more because although techies like us can see the value in a development platform, the people that write the checks need (most of the time) to see something they can understand before they'll approve the money. It also looks good to be able to say that your company (meaning Remedy) has some product features that can be benchmarked against competitors.

Sort of like Pilsbury selling flour to those who wish to bake their own cakes from scratch, and cake mixes to those who lack the skill or the inclination to make from scratch. Both are valuable to different market segments and therefore increase profitability.

Rick

-----Original Message-----
From: Guillaume.Rheault [mailto:Guillaume.Rheault@TARGET.COM]
Sent: Tuesday, February 11, 2003 1:13 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


>Did the original developers
>know what they had? I don't think so. I think they found out by accident.

I think that is too bold a statement...
I believe a small core group at Remedy had their vision straight, and new
the end game of this product. However, you need to start somewhere,
marketing wise. So they targeted the help desk market first...

But I agree with you that many many people at Remedy itself (specially the
sales force, way too focused on selling Remedy applications licenses) didn't
really understand the vision up until recently, thanks to BMC efforts on
making people understand what Remedy is all about

Guillaume

-----Original Message-----
From: Frank, Gordon M. [mailto:Gordon.Frank@JHUAPL.EDU]
Sent: Tuesday, February 11, 2003 02:54
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


I consider the Remedy ARS toolset a 4.5GL. In the traditional sense, it is
not a 5GL because to me a 5GL implies some Artificial Intelligence
mechanism.

One person I work with described Remedy ARS with this analogy. In the past
automobile manufacturers put "lug nuts" on cars by hand. As automation came
around, "lug nuts" were now put on by robots. I view Remedy ARS as a robot.
It allows a developer to do more because he doesn't have to worry about the
primitive operations of putting on a "lug nut" or fighting with a 3GL or
4GLs syntax. I always thought this was an excellent analogy.

One of the reasons I feel Remedy ARS is a 4.5GL, is closes the options to
choose from. If the operation is on the Client, it's probably an Active
Link. If the operation is on the Server, it's probably a Filter. This binary
decision makes development constrained and therefore forces programmers not
try fancy methods. I believe one of the problems of things like PowerBuilder
and other 4GLs, is that you can "object" everything. Remedy ARS stops this,
because you only can do it one way or the other.

It also forces reuse. Because Active Links and Filters are the only items
available (Escalations are timed Filters in from my point of view),
developers concentrate on reusing object on several applications.

Mr. Dijkstra (God rest his soul) would have loved Remedy ARS. Dijkstra
talked about "D-Structures". "D-Stuctures" are the basics of "Structured
Programming" and "Control Structures." A definite beginning and a definite
ending (Begin/End).

Isn't Remedy ARS the perfect "D-Structure?"

If an object needs work, Disable it! Testing and integrating new objects is
very easy with the Remedy ARS system.

Now ask me if the management knew or know what they had done! Remedy ARS was
traditionally a "Help Desk" oriented product. Did the original developers
know what they had? I don't think so. I think they found out by accident.
Recently they have been selling Remedy ARS as a development or workflow
toolkit. I think BMC may be helping with this viewpoint.

That's my two cents,
Gordon M. Frank
Johns Hopkins University Applied Physics Laboratory

-----Original Message-----
From: Rick Cook [mailto:rcook@DENALIAI.COM]
Sent: Tuesday, February 11, 2003 1:52 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Sounds like we've found our author! Though some good points have been made
by all, I do agree that programming in Remedy isn't all that hard to do at a
low or even mediocre level. Doing it really well does take some skill and
smarts, but it isn't brain surgery.

I'd love to see the quantification of the work complexity that you talked
about. It would make for some interesting ROI discussions, both at an
architectural and project level; for instance, what true value is there in
being able to quantify the complexity of the work?

Rick

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated because
of the exacting nature of their work tasks. What kind of mystery is that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of
our productivity using well-defined industry standards, not some qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in
1976, (see Brad Bingel's post earlier in this thread) surely we can do at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time,
or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

>Did the original developers
>know what they had? I don't think so. I think they found out by accident.

I think that is too bold a statement...
I believe a small core group at Remedy had their vision straight, and new
the end game of this product. However, you need to start somewhere,
marketing wise. So they targeted the help desk market first...

But I agree with you that many many people at Remedy itself (specially the
sales force, way too focused on selling Remedy applications licenses) didn't
really understand the vision up until recently, thanks to BMC efforts on
making people understand what Remedy is all about

Guillaume

-----Original Message-----
From: Frank, Gordon M. [mailto:Gordon.Frank@JHUAPL.EDU]
Sent: Tuesday, February 11, 2003 02:54
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


I consider the Remedy ARS toolset a 4.5GL. In the traditional sense, it is
not a 5GL because to me a 5GL implies some Artificial Intelligence
mechanism.

One person I work with described Remedy ARS with this analogy. In the past
automobile manufacturers put "lug nuts" on cars by hand. As automation came
around, "lug nuts" were now put on by robots. I view Remedy ARS as a robot.
It allows a developer to do more because he doesn't have to worry about the
primitive operations of putting on a "lug nut" or fighting with a 3GL or
4GLs syntax. I always thought this was an excellent analogy.

One of the reasons I feel Remedy ARS is a 4.5GL, is closes the options to
choose from. If the operation is on the Client, it's probably an Active
Link. If the operation is on the Server, it's probably a Filter. This binary
decision makes development constrained and therefore forces programmers not
try fancy methods. I believe one of the problems of things like PowerBuilder
and other 4GLs, is that you can "object" everything. Remedy ARS stops this,
because you only can do it one way or the other.

It also forces reuse. Because Active Links and Filters are the only items
available (Escalations are timed Filters in from my point of view),
developers concentrate on reusing object on several applications.

Mr. Dijkstra (God rest his soul) would have loved Remedy ARS. Dijkstra
talked about "D-Structures". "D-Stuctures" are the basics of "Structured
Programming" and "Control Structures." A definite beginning and a definite
ending (Begin/End).

Isn't Remedy ARS the perfect "D-Structure?"

If an object needs work, Disable it! Testing and integrating new objects is
very easy with the Remedy ARS system.

Now ask me if the management knew or know what they had done! Remedy ARS was
traditionally a "Help Desk" oriented product. Did the original developers
know what they had? I don't think so. I think they found out by accident.
Recently they have been selling Remedy ARS as a development or workflow
toolkit. I think BMC may be helping with this viewpoint.

That's my two cents,
Gordon M. Frank
Johns Hopkins University Applied Physics Laboratory

-----Original Message-----
From: Rick Cook [mailto:rcook@DENALIAI.COM]
Sent: Tuesday, February 11, 2003 1:52 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Sounds like we've found our author! Though some good points have been made
by all, I do agree that programming in Remedy isn't all that hard to do at a
low or even mediocre level. Doing it really well does take some skill and
smarts, but it isn't brain surgery.

I'd love to see the quantification of the work complexity that you talked
about. It would make for some interesting ROI discussions, both at an
architectural and project level; for instance, what true value is there in
being able to quantify the complexity of the work?

Rick

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated because
of the exacting nature of their work tasks. What kind of mystery is that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of
our productivity using well-defined industry standards, not some qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in
1976, (see Brad Bingel's post earlier in this thread) surely we can do at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time,
or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

I consider the Remedy ARS toolset a 4.5GL. In the traditional sense, it is
not a 5GL because to me a 5GL implies some Artificial Intelligence
mechanism.

One person I work with described Remedy ARS with this analogy. In the past
automobile manufacturers put "lug nuts" on cars by hand. As automation came
around, "lug nuts" were now put on by robots. I view Remedy ARS as a robot.
It allows a developer to do more because he doesn't have to worry about the
primitive operations of putting on a "lug nut" or fighting with a 3GL or
4GLs syntax. I always thought this was an excellent analogy.

One of the reasons I feel Remedy ARS is a 4.5GL, is closes the options to
choose from. If the operation is on the Client, it's probably an Active
Link. If the operation is on the Server, it's probably a Filter. This binary
decision makes development constrained and therefore forces programmers not
try fancy methods. I believe one of the problems of things like PowerBuilder
and other 4GLs, is that you can "object" everything. Remedy ARS stops this,
because you only can do it one way or the other.

It also forces reuse. Because Active Links and Filters are the only items
available (Escalations are timed Filters in from my point of view),
developers concentrate on reusing object on several applications.

Mr. Dijkstra (God rest his soul) would have loved Remedy ARS. Dijkstra
talked about "D-Structures". "D-Stuctures" are the basics of "Structured
Programming" and "Control Structures." A definite beginning and a definite
ending (Begin/End).

Isn't Remedy ARS the perfect "D-Structure?"

If an object needs work, Disable it! Testing and integrating new objects is
very easy with the Remedy ARS system.

Now ask me if the management knew or know what they had done! Remedy ARS was
traditionally a "Help Desk" oriented product. Did the original developers
know what they had? I don't think so. I think they found out by accident.
Recently they have been selling Remedy ARS as a development or workflow
toolkit. I think BMC may be helping with this viewpoint.

That's my two cents,
Gordon M. Frank
Johns Hopkins University Applied Physics Laboratory

-----Original Message-----
From: Rick Cook [mailto:rcook@DENALIAI.COM]
Sent: Tuesday, February 11, 2003 1:52 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Sounds like we've found our author! Though some good points have been made
by all, I do agree that programming in Remedy isn't all that hard to do at a
low or even mediocre level. Doing it really well does take some skill and
smarts, but it isn't brain surgery.

I'd love to see the quantification of the work complexity that you talked
about. It would make for some interesting ROI discussions, both at an
architectural and project level; for instance, what true value is there in
being able to quantify the complexity of the work?

Rick

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated because
of the exacting nature of their work tasks. What kind of mystery is that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of
our productivity using well-defined industry standards, not some qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in
1976, (see Brad Bingel's post earlier in this thread) surely we can do at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time,
or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com

An excellent analogy Rick. Never thought of it that way.

Scott

-----Original Message-----
From: Rick Cook [mailto:rcook@DENALIAI.COM]
Sent: Tuesday, February 11, 2003 4:32 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Though I think that Doug, et al, did to at least some degree know what they
had early on, applications are sold more because although techies like us
can see the value in a development platform, the people that write the
checks need (most of the time) to see something they can understand before
they'll approve the money. It also looks good to be able to say that your
company (meaning Remedy) has some product features that can be benchmarked
against competitors.

Sort of like Pilsbury selling flour to those who wish to bake their own
cakes from scratch, and cake mixes to those who lack the skill or the
inclination to make from scratch. Both are valuable to different market
segments and therefore increase profitability.

Rick

-----Original Message-----
From: Guillaume.Rheault [mailto:Guillaume.Rheault@TARGET.COM]
Sent: Tuesday, February 11, 2003 1:13 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


>Did the original developers
>know what they had? I don't think so. I think they found out by accident.

I think that is too bold a statement...
I believe a small core group at Remedy had their vision straight, and new
the end game of this product. However, you need to start somewhere,
marketing wise. So they targeted the help desk market first...

But I agree with you that many many people at Remedy itself (specially the
sales force, way too focused on selling Remedy applications licenses) didn't
really understand the vision up until recently, thanks to BMC efforts on
making people understand what Remedy is all about

Guillaume

-----Original Message-----
From: Frank, Gordon M. [mailto:Gordon.Frank@JHUAPL.EDU]
Sent: Tuesday, February 11, 2003 02:54
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


I consider the Remedy ARS toolset a 4.5GL. In the traditional sense, it is
not a 5GL because to me a 5GL implies some Artificial Intelligence
mechanism.

One person I work with described Remedy ARS with this analogy. In the past
automobile manufacturers put "lug nuts" on cars by hand. As automation came
around, "lug nuts" were now put on by robots. I view Remedy ARS as a robot.
It allows a developer to do more because he doesn't have to worry about the
primitive operations of putting on a "lug nut" or fighting with a 3GL or
4GLs syntax. I always thought this was an excellent analogy.

One of the reasons I feel Remedy ARS is a 4.5GL, is closes the options to
choose from. If the operation is on the Client, it's probably an Active
Link. If the operation is on the Server, it's probably a Filter. This binary
decision makes development constrained and therefore forces programmers not
try fancy methods. I believe one of the problems of things like PowerBuilder
and other 4GLs, is that you can "object" everything. Remedy ARS stops this,
because you only can do it one way or the other.

It also forces reuse. Because Active Links and Filters are the only items
available (Escalations are timed Filters in from my point of view),
developers concentrate on reusing object on several applications.

Mr. Dijkstra (God rest his soul) would have loved Remedy ARS. Dijkstra
talked about "D-Structures". "D-Stuctures" are the basics of "Structured
Programming" and "Control Structures." A definite beginning and a definite
ending (Begin/End).

Isn't Remedy ARS the perfect "D-Structure?"

If an object needs work, Disable it! Testing and integrating new objects is
very easy with the Remedy ARS system.

Now ask me if the management knew or know what they had done! Remedy ARS was
traditionally a "Help Desk" oriented product. Did the original developers
know what they had? I don't think so. I think they found out by accident.
Recently they have been selling Remedy ARS as a development or workflow
toolkit. I think BMC may be helping with this viewpoint.

That's my two cents,
Gordon M. Frank
Johns Hopkins University Applied Physics Laboratory

-----Original Message-----
From: Rick Cook [mailto:rcook@DENALIAI.COM]
Sent: Tuesday, February 11, 2003 1:52 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Sounds like we've found our author! Though some good points have been made
by all, I do agree that programming in Remedy isn't all that hard to do at a
low or even mediocre level. Doing it really well does take some skill and
smarts, but it isn't brain surgery.

I'd love to see the quantification of the work complexity that you talked
about. It would make for some interesting ROI discussions, both at an
architectural and project level; for instance, what true value is there in
being able to quantify the complexity of the work?

Rick

-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Tuesday, February 11, 2003 10:34 AM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated because
of the exacting nature of their work tasks. What kind of mystery is that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of
our productivity using well-defined industry standards, not some qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in
1976, (see Brad Bingel's post earlier in this thread) surely we can do at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time,
or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com


*****
"The information transmitted is intended only for the person or entity to
which it is addressed and may contain confidential, proprietary, and/or
privileged material. Any review, retransmission, dissemination or other use
of, or taking of any action in reliance upon, this information by persons or
entities other than the intended recipient is prohibited. If you received
this in error, please contact the sender and delete the material from all
computers."

Well you certainly put me in my place with that widely applicable analogy.
But while I'm off reconsidering my career choices, consider that not
everything in life, or computers, can be accurately broken down into a
confluence of numbers.

You asked if anyone has attempted to gather metrics on Remedy development,
and I simply said I didn't think that was feasible. Feel free to use
whatever method you like, and label others' methods however you like.



-----Original Message-----
From: Spaeth Jennifer Contractor HQ AFWA
[mailto:jennifer.spaeth@AFWA.AF.MIL]
Sent: Tuesday, February 11, 2003 1:34 PM
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


LOL.

Calling Remedy development a dark art is like the AFL-CIO claiming that
machinists are a privileged group whose efforts cannot be duplicated because
of the exacting nature of their work tasks. What kind of mystery is that?
My grandmother could code in Remedy. (May she rest in peace.) It's a
skill, not a God-given talent.

We're a CMM level 3 organization. We are creating a quantities measure of
our productivity using well-defined industry standards, not some qualitative
seat-of your pants, Ouija-board system of software development.

If T.J. McCabe could write a book about software complexity measurement in
1976, (see Brad Bingel's post earlier in this thread) surely we can do at
least as well 25 years later, regardless of the language.

Jennifer

-----Original Message-----
From: Dan Holdsworth [mailto:daniel.r.holdsworth@MAN.AC.UK]
Sent: Tuesday, February 11, 2003 11:02
To: ARSLIST@ARSLIST.ORG
Subject: Re: Remedy Vs VB/Oracle Forms


Aaron Keller wrote:
> No offense to those that have tried, but I personally don't think it
> feasible to define one's development based on an objective, mathematical
> computation such as Logical Lines of Code. And not because I think the
> ARSystem is a 3GL, 4GL, or 4.53687GL. There is just too much subjectivity
> to ARSystem development. Don't get me wrong, I like Math. But I can
develop
> a solution using many workflow objects, or few. And my approach can
dictate
> the complexity of the workflow.
>
> For example, I could create several character menus for an application, or
I
> could create search menus, and store the menu data in a form. I could
> create one large form with multiple fields and functions, or several forms
> with workflow that push data between them. As a more real example, look
at
> how the HelpDesk application handles automatic assignment - there are
> multiple forms and a passel of links and filters involved. But most of us
> can look at that and come up with what we consider a "better" way, with
> fewer, less complicated workflow.

In other words, it is quite possible to vary the difficulty of a task
simply by varying how you go about doing it. My background as a Perl
programmer reflects this; I have encountered good and bad coding styles,
but the ones that stick in my mind are the fantastically-complicated
object-orientation addict styles.

If you assess these sorts of styles of coding, then by the methods
described, the coders would come out looking like Perl geniuses.

If you try reviewing the code by some form of peer review, a very
different opinion would become apparent; the overly-complicated code
would universally be considered vile and horrible to maintain.

> So when attempting to define how well a developer or team is doing, I
think
> you have to look at subjective means, such as how well they define
> requirements, or functional specifications. Or look at development time,
or
> satisfaction. (Someone came up with the requirements, so how well are they
> satisfied with the result?) How many "redo's"? How many milestones or
> deadlines do they miss, etc.

The best way to do it is with some form of peer review. Nothing else
even comes close to being useful; deadlines missed can indicate poor
developers, OR it can indicate poor management, or merely an especially
nasty problem.

Mathematics doesn't come close to measuring that. Peer review isn't so
hot either, but it is probably the least worst option.


--
Dr Dan Holdsworth
Remedy ARS Administrator, Manchester Computing
daniel.r.holdsworth@man.ac.uk
0161 275 0606



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com



UNSUBSCRIBE or access ARSList Archives at http://www.ARSLIST.org (Support:
mailto:support@arslist.org ) ARSList is hosted by QMX SUPPORT SERVICES at
www.QMXS.com