Ada Revision Process was RE: Extensible Enummerated types FW: When will next Ada revision be? (83, 95, ?)

Ada Revision Process was RE: Extensible Enummerated types FW: When will next Ada revision be? (83, 95, ?)

[Robert C. Leif, Ph.D.]

From: Bob Leif
To: Robert Dewar et al.

Your note concerns two questions: The first is the Ada 0Y process and the
second is the utility of Extensible Enumerated types. I will answer them
separately.

Ada 0Y Process:

I might first note, that I asked for the location of the documentation for
Ada 95 proposed addition of extensible enumerated types. I still do believe
that this was an appropriate place to start. However, after downloading what
I believe to be the location suggested by Tucker Taft, I still can not find
it.

You are quite correct concerning the formal process. However, I believe that
it makes sense to start with an informal discussion on Comp.Lang.Ada. This
has three virtues:

1) Since I am not a software expert, any syntax I suggest will probably, at
least, need to be refined.

2) I believe that introduction of enhancements or changes to Ada should be
as democratically decided as possible. Parenthetically, I vividly remember a
TriAda meeting where the audience voted not to have dual inheritance. This
proved to me that Ada was certainly not controlled by the military.

3) One of the reasons for Ada's apparent lack of commercial success is that
it has been technologically driven. When one posts a suggestion on
Comp.Lang.Ada, the resulting response to a very limited extent constitutes a
market survey or the equivalent of a focus group discussion. In order for
this magnificent technology to reach the greatest number of users, there has
to be some market driven activity.

Often in commercial endeavors, the marketing department makes difficult or
impossible demands on engineering. I have yet to read a posting from an Ada
marketing department saying, if we enhanced Ada by something, the Ada market
penetration would increase by some percentage.

Since one of the greatest virtues of Ada is interoperability, a consensus
often has to be reached before individual vendors enhance the language. I
might note that ACT by creating an 'Img attribute made an excellent
contribution, which I hope is included in the next revision of Ada.

Please post the URL where one can find the exact procedure for proposing an
enhancement to Ada.

I hope my fellow readers of Comp.Lang.Ada would also post their comments
concerning the Ada revision process.

-----Original Message-----
From: comp.lang.ada-admin@ada.eu.org
[mailto:comp.lang.ada-admin@ada.eu.org]On Behalf Of Robert Dewar
Sent: Friday, February 02, 2001 8:38 AM
To: comp.lang.ada@ada.eu.org
Subject: RE: Extensible Enummerated types FW: When will next Ada
revision be? (83, 95, ?)


In article <mailman.981098883.32366.comp.lang.ada@ada.eu.org>,
  comp.lang.ada@ada.eu.org wrote:
> From: Bob Leif
> To: Jeff Carter et al.

<<Bob's suggestions on how the ARG should proceed snipped>>

Bob, it seems like you are writing this without first finding
out what is going on right now, which might be a good idea,
since this procedure has been discussed and agreed on, and
is being actively used now, and is unlikely to change without
some very good arguments.

The current procedure is the following. For language extensions
an "amendment" class AI is prepared, this is processed by the
ARG into a fully formed, fully worked out, language extension
proposal which is then approved by the ARG, and subsequently
by WG-9.

Vendors are then free to implement these extensions. It is
typically likely that at least one vendor will do so, since
if it is the case that no vendors see any value in implementing
a feature (considering their customer needs), the proposal
is unlikely to fly in the first place.

The idea is that eventually any revision of Ada might include
some or all of these extensions in either their unmodfied form,
or perhaps a modified form (like the rest of the language at
that stage, they would be subject to revision).

In terms of the existing features in the language, as you know
a corrigendum has already been approved that would of course
be the starting point for a new standard.

In the case of this particular proposal (on which I have not
commented simply because I have seen no good argument for its
utility (it would be nice to see a convincing example of an
application that would benefit from this proposal, rather than
simply language proposals), the proper approach is to convince
at least one member of the ARG to prepare an AI, or
alternatively prepare a complete AI, and submit it for
consideration.


Sent via Deja.com
http://www.deja.com/

Re: Ada Revision Process

[Marc A. Criley]

"Robert C. Leif, Ph.D." wrote:
> 
 <snip>
> 
> Please post the URL where one can find the exact procedure for proposing an
> enhancement to Ada.
> 

See the section titled: "Instructions for Comment Submission" at

   http://www.ada-auth.org/arm-html/RM-0-3.html

Comments submitted in accordance with this procedure do enter the
pipeline for language revision consideration.  And language revisions
_really_do_ come out of this, I can personally vouch for that fact
(AI-00242).

Marc A. Criley
Senior Staff Engineer
Quadrus Corporatino
www.quadruscorp.com