Re: Is the "Ada mandate" being reconsidered?

Re: Is the "Ada mandate" being reconsidered?

[Thomas C. Timberlake]

In article <4mq7mg$8hs@jake.probe.net>, hddodson@probe.net (Howard Dodson)
wrote:

[snip]
> Does anyone know any more about this committee or have any background
> as to what its charter is?  How do its findings affect whether future
> DoD software projects must be developed in Ada?
> 
> Any information about this committee would be greatly appreciated.
> 

I have the following from a person on the committee:

A homepage has been created for the NRC "Review of the Past and Present Contexts
for Using Ada Within the Department of Defense" project. It is located
within the Computer Science and Telecommunications Board website, under
"Work in Progress."  The content consists of the public information
concerning the project (overview, committee members, detailed description).

The CSTB address is:

     http://www2.nas.edu/cstbweb/index.html

-- 
Tom Timberlake                                                 Member, Team Ada
Boeing Defense & Space Group                  Software Systems         
timberlake@xavier.ds..boeing.com
Member Team Ada

Re: Is the "Ada mandate" being reconsidered?

[David Weller]

In article <4mq7mg$8hs@jake.probe.net>,
Howard Dodson <hddodson@probe.net> wrote:
>At the Software Technology Conference held in Salt Lake City last month,
>one of the speakers (Col. Chadwick, USMC) said that a committee was
>being formed, including Dr. Barry Boehm, to consider whether Ada should
>be a mandate for software development for the DoD, and that their 
>findings were due in October '96.  
>
>Does anyone know any more about this committee or have any background
>as to what its charter is?  How do its findings affect whether future
>DoD software projects must be developed in Ada?
>
>Any information about this committee would be greatly appreciated.
>

Not a problem, for those of you wanting to read the greasy details,
here's the web page:  http://www2.nas.edu/cstbweb/index.html

(Actually, the direct link is: http://www2.nas.edu/cstbweb/2252.html )

For those of you wanting the Reader's Digest synopsis:
(From the Scope statement for the study:)
"A study committee will examine the original and current rationales for
using Ada as the standard programming language for DOD weapons systems 
and information systems and will evaluate alternative means of achieving 
the objectives originally desired from Ada: to enhance discipline in DOD 
software engineering activities and to reduce the proliferation of
programming languages in DOD systems. In doing so, the committee will
consider the ways in which the environment for software engineering has 
changed in the two decades since the Ada program was originally established."


This is a perfectly reasonable thing to do, in my opinion.  It stands
to reason that the government should periodically reevaluate a
preferred language to determine if it should remain a preferred
language, and to decide what role it should have.

The panel is composed of many professionals that are highly respected
in the computer science community (see the web page).  I, for one,
will have absolutely no problem supporting the recommendations of the
panel.


-- 
    Visit the Ada 95 Booch Components Homepage: www.ocsystems.com/booch
           This is not your father's Ada -- lglwww.epfl.ch/Ada

Is the "Ada mandate" being reconsidered?

[Howard Dodson]

At the Software Technology Conference held in Salt Lake City last month,
one of the speakers (Col. Chadwick, USMC) said that a committee was
being formed, including Dr. Barry Boehm, to consider whether Ada should
be a mandate for software development for the DoD, and that their 
findings were due in October '96.  

Does anyone know any more about this committee or have any background
as to what its charter is?  How do its findings affect whether future
DoD software projects must be developed in Ada?

Any information about this committee would be greatly appreciated.

Howard Dodson
Unisys Corporation
10843 Old Mill Road
Omaha, NE  68144

Telephone:  (402) 334-4080

Re: Is the "Ada mandate" being reconsidered?

[Tucker Taft]

Howard Dodson (hddodson@probe.net) wrote:
: At the Software Technology Conference held in Salt Lake City last month,
: one of the speakers (Col. Chadwick, USMC) said that a committee was
: being formed, including Dr. Barry Boehm, to consider whether Ada should
: be a mandate for software development for the DoD, and that their 
: findings were due in October '96.  

: Does anyone know any more about this committee or have any background
: as to what its charter is?  

For public information on this committee, see:

    http://www2.nas.edu/cstbweb/

under "Work in Progress" (or cstbweb/2252.html if you want to go there
directly).

: ... How do its findings affect whether future
: DoD software projects must be developed in Ada?

The committee will make recommendations; the DoD will 
decide how to act on them.

: Any information about this committee would be greatly appreciated.

In general, the committee is very informed about Ada;
the above web site includes a list of the committee members.

: Howard Dodson
: Unisys Corporation
: 10843 Old Mill Road
: Omaha, NE  68144

: Telephone:  (402) 334-4080

-Tucker Taft   stt@inmet.com   http://www.inmet.com/~stt/
Intermetrics, Inc.  Cambridge, MA  USA

Re: Is the "Ada mandate" being reconsidered?

[Robert Munck]

On Wed, 08 May 1996 20:12:19 -0400, Theodore E. Dennison
<dennison@escmail.orl.mmc.com> wrote:
>I just read through the descripion of their review. Important points:
>  o   They seem to be taking a fair amount of care to prevent the 
>      committee from being "stacked" one way or the other.

Except, of course, that there's hardly anyone on the committee
who has come within a mile of developing applications software
for the DoD in the last decade.  They lean heavily toward university
types and vendors of development software, the kind of people who
believe "common knowledge" fairy tales about reuse, COTS, rapid
prototyping, and O-O.

Just once, I'd like to see a committee like this have a substantial
number of people who have been grunt programmers, line managers,
and second-level managers on real multi-year, multi-contractor DoD
development projects. The professorial types can do brilliant work,
and quite a few of these have made important advances in the state
of the art, but there seems to be a huge difference between the 
state of the art in which they work and the state of the practice in
which real DoD software gets written.  It's not just that one lags the
other; they seem to be going off in different directions.

For example, this group is apt to be quite taken with the 
wonderful logic of COTS and know nothing about the practical
nightmares of frequent package, OS, and platform upgrades,
user "support," and vendor instability.

I was involve in some work years ago at NRL where they brought
in the kind of real-world experienced people I'm talking about to
do some work on development environments.  It was fantastic;
they really knew what works and what doesn't from painful
experience.  Unfortunately, that kind of person is much too 
valuable doing the real work to spend time sitting on committees.

Bob Munck@acm.org

Re: Is the "Ada mandate" being reconsidered?

[Theodore E. Dennison]

Robert Munck wrote:
> 
> On Wed, 08 May 1996 20:12:19 -0400, Theodore E. Dennison
> <dennison@escmail.orl.mmc.com> wrote:
> >I just read through the descripion of their review. Important points:
> >  o   They seem to be taking a fair amount of care to prevent the
> >      committee from being "stacked" one way or the other.
> 
> Except, of course, that there's hardly anyone on the committee
> who has come within a mile of developing applications software
> for the DoD in the last decade.  They lean heavily toward university

I was all set to argue with you, until I got to this paragraph -

> For example, this group is apt to be quite taken with the
> wonderful logic of COTS and know nothing about the practical
> nightmares of frequent package, OS, and platform upgrades,
> user "support," and vendor instability.

You could not have described my current hell more succintly. Except
you left out hardware/hardware and hardware/software compatability
issues. And hardware/hardware/software compatability, and ...

    (Aieeeee!)

-- 
T.E.D.          
                |  Work - mailto:dennison@escmail.orl.mmc.com  |
                |  Home - mailto:dennison@iag.net              |
                |  URL  - http://www.iag.net/~dennison         |

Re: Is the "Ada mandate" being reconsidered?

[Ken Garlington]

Robert Munck wrote:
> 
> Except, of course, that there's hardly anyone on the committee
> who has come within a mile of developing applications software
> for the DoD in the last decade.

I believe Maretta Holden would be an exception to this rule, and based
on personal experience, I don't think she'll have any trouble presenting
the DoD contractor view of things!

Furthermore, based on my conversations with other members of this committee,
I believe the "practical" point of view is reasonably well represented.
Also, the committee has had inputs from DoD (which had inputs from DoD
contractors), etc. so they are certainly not working in a vacuum.

-- 
LMTAS - "Our Brand Means Quality"

Re: Is the "Ada mandate" being reconsidered?

[Robert Munck]

On Mon, 13 May 1996 07:27:37 +0000, Ken Garlington
<garlingtonke@lmtas.lmco.com> wrote:

>Robert Munck wrote:
>>... hardly anyone on the committee ... has come within a mile of developing
>> applications software for the DoD in the last decade.
>
>I believe Maretta Holden would be an exception to this rule, and based
>on personal experience, I don't think she'll have any trouble presenting
>the DoD contractor view of things!

Maretta was the main reason that I qualified my statement.  However,
knowing how the Boeing management structure works, I'd still be
inclined to bet that she hasn't been a designer, programmer, or first-
or second-level manager on a project developing production software
for the DoD for a very long time.  Heck, that's the standard thing
that happens to good people.

My plea is for people who actually know how projects work, not people
who read status reports, listen to briefings, read (and write)
scholarly academic papers, and more or less believe what they hear.


>Also, the committee has had inputs from DoD (which had inputs from DoD
>contractors), etc. so they are certainly not working in a vacuum.

That gave me a good laugh.  By the time status reports work their way
up the management chain from the people who know what the status
is, are "vetted" by Marketing, go to the PM office and up that
service's bureaucracy, have all possible problems and deficiencies
cleaned out, and go to DoD, a vacuum would be crowded with
information by comparison.

Remember, this is the community that, after discovering that
maintenance forms the great majority of their software costs,
declared that COTS software would be used whenever possible,
a strategy that obviously reduces initial acquisition costs at
the expense of greatly increased maintenance.

Bob Munck@acm.org

Re: Is the "Ada mandate" being reconsidered?

[Tucker Taft]

Robert Munck (pp000166@interramp.com) wrote:
: On Mon, 13 May 1996 07:27:37 +0000, Ken Garlington
: <garlingtonke@lmtas.lmco.com> wrote:

: >Robert Munck wrote:
: >>... hardly anyone on the committee ... has come within a mile of developing
: >> applications software for the DoD in the last decade.
: >
: >I believe Maretta Holden would be an exception to this rule, and based
: >on personal experience, I don't think she'll have any trouble presenting
: >the DoD contractor view of things!

: My plea is for people who actually know how projects work, not people
: who read status reports, listen to briefings, read (and write)
: scholarly academic papers, and more or less believe what they hear.

I think that the committee is pretty well balanced
between "trench fighters" and "big picture" folks.
Admittedly, not all of us have done DoD software development,
but pretty much all of us have been in the trenches developing
software, and are quite familiar with the sorry state
of many "off-the-shelf" software products and components.

In any case, at this point the Ada community might want to focus
on providing constructive, well documented, input to the committee.

: ...
: Bob Munck@acm.org

-Tucker Taft   stt@inmet.com   http://www.inmet.com/~stt/
Intermetrics, Inc.  Cambridge, MA  USA

Re: Is the "Ada mandate" being reconsidered?

[Robert Munck]

On Tue, 14 May 1996 17:53:29 GMT, stt@henning.camb.inmet.com (Tucker
Taft) wrote:
>Admittedly, not all of us have done DoD software development,
>but pretty much all of us have been in the trenches developing
>software,

But that's the difference that's important.  There is a huge
gap between developing s/w in a commercial, university,
or vendor environment and doing so as a bare-bones DoD contractor
on a production application project. For one thing, I'll bet you that
the people you've worked with at Intermetrics, and that I worked
with at SofTech, were paid twice as much and were ten times as
good as equivalent programmers for big aerospace contractors.
DoD projects force contractors to use the cheapest people possible.

Other differences include the tremendous rate of requirements
changes, with your own company's management and marketing
encouraging it and doing nothing to protect you from it. Also,
the institutional barriers, government and contractor alike, that
make it absolutely impossible to do any kind of real s/w reuse,
rapid prototyping, fads like cleanroom or O-O. The schedules
written with no connection to reality or possibility. Etc. etc.

These things have all had a lot more to do with the success or
failure of Ada over the last 15 years than anything like language
features, development tools, or lack of commercial success.

>and are quite familiar with the sorry state
>of many "off-the-shelf" software products and components.

Heck, the committee's own charter makes COTS an untouchable:

     "While commercial off-the-shelf (COTS) software is considered
      the first choice for military systems, ..."

After all, some of the people who developed this policy are still
in place and well short of retirement.


>In any case, at this point the Ada community might want to focus
>on providing constructive, well documented, input to the committee.

Like what?  What kind of things would you like to see?  I'm completely
at a loss.  What kind of input could there be, other than anecdotes?
Are you guys going to have access to the kinds of financial data,
project post-mortems, long-term maintenance costs, successful and
failed projects, etc. that would make a quantative, bottom-line
decision possible?

Bob Munck@acm.org

Re: Is the "Ada mandate" being reconsidered?

[Roy M. Bell]

I have just finished reading the web pages mentioned in earlier articles
with the title of: "Review of the Past and Present Contexts for Using
Ada Within the Department of Defense". I could tell the that the people
who wrote this synopsis were trying to be fair because they mentioned a
number of good things about Ada and some are fairly recent.

In my opinion the most important fact comes in the last sentence of the
first paragraph where they say the following: "Since 1977, the number of
programming languages used in DOD systems has declined from an estimated
450 to 37, four of which (Ada, C, Cobol, and Fortran) account for three
quarters of all source code."

I find the most disturbing statement in the middle of the second
paragraph to be the following: "Today, as commercial firms transition to
client-server based technologies and object-oriented programming
languages, they are shifting to C++, which is seen as a natural
extension of their experience with C."

I was under the impression the sales of C++ products and services to be
declining. Does anyone have any figures? I was under the impression that
the fastest growth was in the non-traditional 3rd generational languages
such as HTML, VRML, and graphical front ends for databases.

It is very common to see a custom language or custom application beat a
traditional 3rd generational language in specific small-scale
circumstances, but this doesn't mean that the mandated should be lifted.

The fifth paragraph contains the following: "In addition, the lack of
suitable software development tools, software libraries, and bindings
can inhibit use of a programming language. These elements enable
programmers to rapidly generate reliable code, reuse proven building
blocks of code written in other languages, and link with a wide range of
system software utilities." This is a fair point. I would also add that
the lack of a standard for C++ also makes reuse difficult.

Re: Is the "Ada mandate" being reconsidered?

[Peggy Byers]

I work for a large defense contractor developing standard
transportation systems for the US Army.  I am currently working in C,
but I am indebted to Ada and a few good Ada books for taking me from a
COBOL/4GL coder to a pretty fair software engineer.  Another system
within our project developes for MS-DOS Character mode using Ada and
AdaSage.  The user community is asking for a GUI version running under
Windows NT.  They are also suggesting a language other than Ada
because it is there understanding that "the Ada mandate has been
lifted".  Do they know something I don't know?  Who is responsible for
deciding whether the Ada mandate stays in place?  

Mike Byers
Computer Science Corp.
Fort Lee, VA

byersm@lee-dns.army.mil
byersmp@erols.com

Re: Is the "Ada mandate" being reconsidered?

[Robert Dewar]

Mike, the policy that Ada must be used for DoD development has not changed.
It is of course under constant review to make sure that it is applied
appropriately, but there is no change in policy that would affect your
project at the current time. The understanding that there has been such
a change is just plain wrong.

Re: Is the "Ada mandate" being reconsidered?

[David Weller]

In article <4peu0v$rfq@news15.erols.com>,
Peggy Byers <byersmp@erols.com> wrote:
>Windows NT.  They are also suggesting a language other than Ada
>because it is there understanding that "the Ada mandate has been
>lifted".  Do they know something I don't know?  Who is responsible for
>deciding whether the Ada mandate stays in place?  
>

Your customers are wrong and ill-informed (*gasp*!  Could it be? :-)

The "mandate" is under review, but that's a LONG way from "has been
lifted", and it's _definitely_ not the way I would bet.

More info can be found at: http://www2.nas.edu/cstbweb/21b6.html

-- 
    Visit the Ada 95 Booch Components Homepage: www.ocsystems.com/booch
           This is not your father's Ada -- lglwww.epfl.ch/Ada

Re: Is the "Ada mandate" being reconsidered?

[Ken Garlington]

Mike Byers wrote:
> 
> Who is responsible for
> deciding whether the Ada mandate stays in place?

As I understand it, the final DoD decision happens in Mr. Paige's 
office. Here's a POC: 

    Leonard, Connie
    connie.leonard@osd.mil

> Mike Byers
> Computer Science Corp.
> Fort Lee, VA
> 
> byersm@lee-dns.army.mil
> byersmp@erols.com

-- 
LMTAS - "Our Brand Means Quality"

Re: Is the "Ada mandate" being reconsidered?

[Tucker Taft]

Peggy Byers (byersmp@erols.com) wrote:
: I work for a large defense contractor developing standard
: transportation systems for the US Army.  I am currently working in C,
: but I am indebted to Ada and a few good Ada books for taking me from a
: COBOL/4GL coder to a pretty fair software engineer.  Another system
: within our project developes for MS-DOS Character mode using Ada and
: AdaSage.  The user community is asking for a GUI version running under
: Windows NT.  They are also suggesting a language other than Ada
: because it is there understanding that "the Ada mandate has been
: lifted".  Do they know something I don't know?  Who is responsible for
: deciding whether the Ada mandate stays in place?  

More interesting, perhaps, is why they want to use a different language.
Can you identify some of the reasons (independent of whether they
are "good" or "bad")?  The only way Ada will succeed in the long run
is if "they" (whoever "they" may be in any particular case) choose
Ada on its merits, not on the basis of some "mandate."

Even in a case where the mandate applies, I would encourage people
to understand, and publicize if possible, the considerations involved
in making a language selection.  Only by truly understanding these
considerations can progress be made toward making Ada a more attractive
*choice*.  In my experience, if you focus on the underlying reasons,
there are better possibilities to affect the decision, whereas just 
hammering on the "mandate" or equivalent top-down considerations, you
run the risk of creating an unhappy and unsatisfiable customer/user.

: Mike Byers
: Computer Science Corp.
: Fort Lee, VA

: byersm@lee-dns.army.mil
: byersmp@erols.com

-Tucker Taft   stt@inmet.com   http://www.inmet.com/~stt/
Intermetrics, Inc.  Cambridge, MA  USA

Re: Is the "Ada mandate" being reconsidered?

[James Krell]

In article <4peu0v$rfq@news15.erols.com>, byersmp@erols.com says...
>
>I work for a large defense contractor developing standard
>transportation systems for the US Army.  I am currently working in C,
>but I am indebted to Ada and a few good Ada books for taking me from a
>COBOL/4GL coder to a pretty fair software engineer.  Another system
>within our project developes for MS-DOS Character mode using Ada and
>AdaSage.  The user community is asking for a GUI version running under
>Windows NT.  They are also suggesting a language other than Ada
>because it is there understanding that "the Ada mandate has been
>lifted".  Do they know something I don't know?  Who is responsible for
>deciding whether the Ada mandate stays in place?  
>
Whether or not the Ada mandate stays in place is irrelevant.  In general,
the mandate is ignored throughout the Dod.  You can pretty much write
the code in whatever you wish to.

Re: Is the "Ada mandate" being reconsidered?

[Paul Whittington]

Peggy Byers wrote:
> ... Another system within our project developes for MS-DOS Character
> mode using Ada and AdaSage.  The user community is asking for a GUI
> version running under Windows NT. 

The AdaSAGE development at the Idaho National Engineering Laboratory
has released a new version of AdaSAGE that provides complete GUI 
development capabilities for Windows NT.  Call the AdaSAGE hotline
at (208)526-0656 for information on obtaining this release.

> They are also suggesting a language other than Ada because it is 
> there understanding that "the Ada mandate has been lifted".  Do they
> know something I don't know?  

Why would the user community care what language is used to develope
an application?  

No the Ada mandate has not been lifted, and even if it had Ada is still
the best language to use for the development of applications if said
applications need to be developed for the least cost, and be reliable
and maintainable.  See the "Home of the brave Ada programmers" WWW page
for information on relative costs of Ada and C/C++ development.

> Who is responsible for deciding whether the Ada mandate stays in
> place?

DISA

Re: Is the "Ada mandate" being reconsidered?

[Jon S Anthony]

In article <4peu0v$rfq@news15.erols.com> byersmp@erols.com (Peggy Byers) writes:

> I work for a large defense contractor developing standard
> transportation systems for the US Army.  I am currently working in C,
> but I am indebted to Ada and a few good Ada books for taking me from a
> COBOL/4GL coder to a pretty fair software engineer.  Another system
> within our project developes for MS-DOS Character mode using Ada and
> AdaSage.  The user community is asking for a GUI version running under
> Windows NT.  They are also suggesting a language other than Ada
> because it is there understanding that "the Ada mandate has been
> lifted". 

What a great piece of "reasoning" and "analysis", your customer has
there.  Switch (and piss away large sums of money in the process) just
because "maybe no one watching".  Forget the fact that you could get
to the purported goal simpler, easier and cheaper using the tools you
are already using (AdaSage has GUI capability).  Of course, this sort
of idiocy and waste doesn't surprise me much as your customer appears
to be in the government.  But it does piss me off because they will be
throwing away some of _my_ money in the course of fulfilling their
quest for waste "justified" by stupidity.

/Jon
-- 
Jon Anthony
Organon Motives, Inc.
1 Williston Road, Suite 4
Belmont, MA 02178

617.484.3383
jsa@organon.com

Re: Is the "Ada mandate" being reconsidered?

[Jim Kingdon]

> The "mandate" is under review, but that's a LONG way from "has been
> lifted", and it's _definitely_ not the way I would bet.
> More info can be found at: http://www2.nas.edu/cstbweb/21b6.html

I agree.  One can usually make a pretty good guess about what this
sort of committee will say by looking at who is on it.  And I noticed
quite a few Ada people listed.  Ergo.....

Re: Is the "Ada mandate" being reconsidered?

[Michael Levasseur]

This weeks "Military & Aerospace Electronics" publications has
an excellent called "Has Ada software language fallen from
grace inside DOD?".

To paraphase it, it states the the movement to COTS, the loss
of credible compiler companies, company management, and DOD
management are all contributing to the demise of the Ada 
mandate.

Alot of these issues are very relevant.

I'll cover these one by one:
COTS - These movement to COTS hass been including move and
more code that has been written in C++ as well 4GL stuff.
Forgetting the wisdom of using COTS on DoD software, interfacing
Ada to C is farely painless. Interfacing Ada to 4GL, commercial
applications, C++ or JAVA are all a major undertaking.

The loss of credible compiler companies - As the number of credible
compiler companies shinks and DoD software budgets continue to
shink getting a vendor for the particular platform are harder and
also more expensive. Ada has not and probably never will overcome
the stigma of being developed by the Government.

Company Management - Many DoD companies now don't really worry
about complying with the DoD directive. They just tell the
customer that inorder to meet timeing and space requirements
C or C++ must be used. This is usually bogus, Ada can usually
satify the requirements. Yes the Ada code must be written tight
and with some forethought, but bad Ada is just as bad as bad C or 
C++.

DoD Management - The DoD lets the companies get away with this
and as DoD's clout diminishes as DoD companies start to work on 
comercial companies as well as software engineers don't want to
chase a skill/tool that is going to die!!!

I've been programming in Ada and C for the last 10 years. I
personally believe that Ada is better for software development.
More maintainable, better information hiding and encapsulation.
Unfortunately, the economics law called "the law of diminishing
returns" this law basically says that the old saying
"if you build a better mouse trap the world will beat a path to
your door" is incorrect. Although Ada is better, C and C++ will
probably be the winner. Remember Beta vs. VHS or IBM vs. MAC.

This is all my 0.02 worth.....

Re: Is the "Ada mandate" being reconsidered?

[Theodore E. Dennison]

Michael Levasseur wrote:
> I'll cover these one by one:
> COTS - These movement to COTS hass been including move and
> more code that has been written in C++ as well 4GL stuff.
> Forgetting the wisdom of using COTS on DoD software, interfacing
> Ada to C is farely painless. Interfacing Ada to 4GL, commercial
> applications, C++ or JAVA are all a major undertaking.

I have recently been directly involved in a DoD effort using large
amounts of COTS hardware and software, and interfacing Ada to it has 
literally been the LEAST of our problems. Writing Ada bindings, even
high level ones, is really not all that difficult for an experienced
Ada developer. There are even some tools available to do it for you.
Getting COTS hardware/software combinations that will work 
together, now THAT has been a nightmare!

> The loss of credible compiler companies - As the number of credible
> compiler companies shinks and DoD software budgets continue to
> shink getting a vendor for the particular platform are harder and
> also more expensive.

I have seen neither of these. I suppose it depends on your 
definition of "credible". The major players, Alsys and Rational seem
to be getting stonger than ever. And the last few months have seen
the emergence of ACT.

The platform I am currently using is now supported by no less than
4 different Ada vendors. When I started 6 months ago, that number 
was 2! I guess you are seeing different trends than I am.

I have also not noticed a large upward movement in prices. Either
way, the price of our Ada compiler roughly in line with, and in 
many cases less than the price of all our COTS software.

In any event, you can now get an Ada compile for FREE.

>                      Ada has not and probably never will overcome
> the stigma of being developed by the Government.

That is perhaps true. I find it odd that this stigma would apply
within the DoD, though.

> I've been programming in Ada and C for the last 10 years. I
> personally believe that Ada is better for software development.
> More maintainable, better information hiding and encapsulation.
> Unfortunately, the economics law called "the law of diminishing
> returns" this law basically says that the old saying
> "if you build a better mouse trap the world will beat a path to
> your door" is incorrect. Although Ada is better, C and C++ will
> probably be the winner. Remember Beta vs. VHS or IBM vs. MAC.

In my experience, C or FORTRAN are usually pushed the strongest 
by high-level engineers who have never used Ada much, and don't 
see any good reason to change from "their" language. They will
use ANY argument handy to justify this. As time goes by I am
starting to see some engineers who have good Ada experience
promoted into these these positions, and am seeing more jobs
proposed in Ada.

So my outlook is a bit more optimistic than yours...but I'm 
learning Java, just in case. :-)

-- 
T.E.D.          
                |  Work - mailto:dennison@escmail.orl.mmc.com  |
                |  Home - mailto:dennison@iag.net              |
                |  URL  - http://www.iag.net/~dennison         |

Re: Is the "Ada mandate" being reconsidered?

[Ken Garlington]

Michael Levasseur wrote:
> 
> COTS - These movement to COTS hass been including move and
> more code that has been written in C++ as well 4GL stuff.

What does writing custom code in C++ or 4GL have to do with COTS?
(Note that, for many years, 4GL has been a DoD preference -- when
Ada is not possible. This isn't new.)

> The loss of credible compiler companies - As the number of credible
> compiler companies shinks and DoD software budgets continue to
> shink getting a vendor for the particular platform are harder and
> also more expensive.

However, when _I_ look at the data, the number of supported platforms 
continues to grow, and more platforms have free compilers available 
today than they did 5 years ago. Could you post the source of the data 
you're using to justify this statement?

> Ada has not and probably never will overcome
> the stigma of being developed by the Government.

Is there not an exquisite irony in posting this statement on the
Internet?

> Company Management - Many DoD companies now don't really worry
> about complying with the DoD directive.
> 
> DoD Management - The DoD lets the companies get away with this

This has been said since the beginning of Ada. The anecdotal
evidence in support of this doesn't appear to be any greater now
than it was 5 years ago. Why would this suddenly cause the "demise"
of Ada? (Interestingly, however, every time I see a service listing
their major software projects, the list of Ada projects continues
to grow...)

> Although Ada is better, C and C++ will
> probably be the winner.

No, obviously Visual Basic will be the winner. No wait, Java will be the
winner! No... :)

-- 
LMTAS - "Our Brand Means Quality"

Re: Is the "Ada mandate" being reconsidered?

[Fergus Henderson]

Tom Robinson <robinson@gdesystems.com> writes:
>[someone writes]:
>>Thomson's ObjectAda compiler is dirt cheap for personal
>>use and the professional version is cheaper than a "professional" C++
>>package.  Also, Gnat is on all sorts of platforms.
>
>Is it really?  When I look at the Ada 95 validated compiler list it 
>looks pretty small to me.  So you're saying that gnat is available as
>long as I am willing to pay for a validation and arrange for maintenence
>or do it myself.  

If the competition is C++ compilers, then I don't see what validation
has to do with it.  I mean it's not as if you are going to find any
validated C++ compilers.

--
Fergus Henderson <fjh@cs.mu.oz.au>   |  "I have always known that the pursuit
WWW: <http://www.cs.mu.oz.au/~fjh>   |  of excellence is a lethal habit"
PGP: finger fjh@128.250.37.3         |     -- the last words of T. S. Garp.

Re: Is the "Ada mandate" being reconsidered?

[Jon S Anthony]

In article <4pk5sm$i7k@gde.GDEsystems.COM> Michael Levasseur <levass@gdesystems.com> writes:

> Alot of these issues are very relevant.

No doubt.


> Ada to C is farely painless. Interfacing Ada to 4GL, commercial
> applications, C++ or JAVA are all a major undertaking.
                       ^^^^
Interfacing to Java is cake.  What are you talking about?


> The loss of credible compiler companies - As the number of credible
> compiler companies shinks and DoD software budgets continue to
> shink getting a vendor for the particular platform are harder and
> also more expensive. 

Well, this sounds good, it's just wrong.  Gnat is free and of better
overall quality than all but the very best Ada83 compilers.  I would
say it is about on par with the best (ahead here, behind there - over
all a wash).  Thomson's ObjectAda compiler is dirt cheap for personal
use and the professional version is cheaper than a "professional" C++
package.  Also, Gnat is on all sorts of platforms.  And since
ObjectAda uses the Intermetrics AdaMagic frontend, it will likely be
all over the place too (or at least highly compatible counterparts
based on the same frontend).

No, overall, this situation is _vastly_ better today than before.


> Ada has not and probably never will overcome
> the stigma of being developed by the Government.

Very legitimate problem.


> Company Management - Many DoD companies now don't really worry
> about complying with the DoD directive. They just tell the
> customer that inorder to meet timeing and space requirements
> C or C++ must be used. This is usually bogus, Ada can usually

Agreed.  There seems to be a real "Mommy! Mommy! Me too! Me too!"
mentality in these people, and that sort of infantilism(wd?) is hard
to overcome.  You certainly can't use reason or logic or facts...

> DoD Management - The DoD lets the companies get away with this
> and as DoD's clout diminishes as DoD companies start to work on 
> comercial companies as well as software engineers don't want to
> chase a skill/tool that is going to die!!!

While there is certainly a problem of perception here, I don't think
this is that big of a deal.  There is evidence of Ada getting some
good press and rational people are taking a real look.  The irrational
incompetents are the last thing you want glomming on to it.  They've
even managed to make C++ look worse than it is (no mean feat...)


> your door" is incorrect. Although Ada is better, C and C++ will
> probably be the winner. Remember Beta vs. VHS or IBM vs. MAC.

If you are speaking of sheer overall popularity, "will" can be
replaced by "is" and C++ can be dropped.  Nothing is anywhere near as
"popular" as C (including C++).  Second, we are not talking mass
market economics here anyway.  The VHS/Beta and IBM/Mac analogies
aren't really relevant.  The man in the street isn't going to be
"programming" in C any more than Ada.

/Jon

-- 
Jon Anthony
Organon Motives, Inc.
1 Williston Road, Suite 4
Belmont, MA 02178

617.484.3383
jsa@organon.com

Re: Is the "Ada mandate" being reconsidered?

[Tom Robinson]

>> The loss of credible compiler companies - As the number of credible
>> compiler companies shinks and DoD software budgets continue to
>> shink getting a vendor for the particular platform are harder and
>> also more expensive. 
>
>Well, this sounds good, it's just wrong.  Gnat is free and of better
>overall quality than all but the very best Ada83 compilers.

I am interested in what measure you are using for quality.  From the small
snipits I have read on the net it is not clear that the generated code
quality  of gnat is quite up to current standards yet.  Am I missing something
here?  Has anyone done ACEC or PIWG comparisons yet?

>Thomson's ObjectAda compiler is dirt cheap for personal
>use and the professional version is cheaper than a "professional" C++
>package.  Also, Gnat is on all sorts of platforms.

Is it really?  When I look at the Ada 95 validated compiler list it 
looks pretty small to me.  So you're saying that gnat is available as
long as I am willing to pay for a validation and arrange for maintenence
or do it myself.  

>And since
>ObjectAda uses the Intermetrics AdaMagic frontend, it will likely be
>all over the place too (or at least highly compatible counterparts
>based on the same frontend).
>
>No, overall, this situation is _vastly_ better today than before.
>
>

I don't understand how you can say that the situation is _vastly_ better
based on what you have stated.  It is a fact that the Ada 83 compiler
choices have been shrinking.  All one has to do is look at the list of
Validated compilers and realize that many companies that did Ada 83 
validations (Alsys, TeleSoft, Verdix, Meridian, Systeam, (off the top of my
head)) are no longer operating under those names.  They have merged with
other companies or gone out of business.  Other companies have 
announced that they will use non-proprietary front ends for their
product offerings (DEC).  Other OEMs have transitioned from directly
offering Ada products to offering them through a 3rd party (IBM->OCSystems).

I think that the Ada market is in transition.  As an Ada user the transition
to Ada 95 seems like it might be a bit tricky.  For large projects it looks
like there are going to be more decisions that I (as a buyer of Ada technology)
will need to make:

  (1) Go with gnat.  Do my own maintenence of the "free" compiler.
  (2) Go with gnat.  Pay someone *ACT?* to provide support.
  (3) Buy a product from a vendor that does not have direct support
      for fixing their own front end problems (AdaMagic).
  (4) Buy a traditional type of Ada product from the few vendors that
      still exist (Rational?).

Tom Robinson

Re: Is the "Ada mandate" being reconsidered?

[Tucker Taft]

Tom Robinson (robinson@gdesystems.com) wrote:
: ...
: I think that the Ada market is in transition.  

I think the same statement can be made about the whole compiler
industry, and especially the C++ market, by the way.

: ... As an Ada user the transition
: to Ada 95 seems like it might be a bit tricky.  For large projects it looks
: like there are going to be more decisions that I (as a buyer of Ada 
: technology) will need to make:

:   (1) Go with gnat.  Do my own maintenence of the "free" compiler.
:   (2) Go with gnat.  Pay someone *ACT?* to provide support.
:   (3) Buy a product from a vendor that does not have direct support
:       for fixing their own front end problems (AdaMagic).

Note that this approach to having multiple compiler vendors
use a common front end has a long history in the C, C++, and Fortran 
marketplace, and is even increasing in the C++ marketplace with the 
growing predominance of Edison Design Group's C++ front end.

The justification is simply that the more people who use the same
front end, the fewer bugs will be left in it for *you* to stumble
over.  In fact, the proliferation of Ada 83 front ends I believe
was one of the major problems with the Ada 83 marketplace.  Having
to maintain all of those front ends through the multiple ACVC releases
was extremely expensive for the collective Ada 83 client base.
Having to deal with the peculiarities of different front ends when
moving from target to target added to a user's porting costs.  And by
having multiple front ends, it was more difficult for all the vendors
to agree on common pragmas, attributes, etc.  

With the new, smaller number of front ends, more coordination 
is likely (e.g. the GNAT and the Intermetrics folks keep in 
close touch on pragmas and attributes, etc), and you can
more likely find another compiler based on the same front end on 
your next target.

As an anecdote, there are several users who have found that 
the biggest expense in porting from Ada 83 to Ada 95 is the inevitable
expense in porting from front-end to front-end, not in accommodating 83=>95 
language changes.  Having lots of front ends out there is not 
necessarily a "good thing" given limited overall resources...

:   (4) Buy a traditional type of Ada product from the few vendors that
:       still exist (Rational?).

Heaven forbid ;-).

: Tom Robinson

-Tucker Taft   stt@inmet.com   http://www.inmet.com/~stt/
Intermetrics, Inc.  Cambridge, MA  USA

Re: Is the "Ada mandate" being reconsidered?

[Tom Robinson]

fjh@mundook.cs.mu.OZ.AU (Fergus Henderson) wrote:
>Tom Robinson <robinson@gdesystems.com> writes:
>>[someone writes]:
>>>Thomson's ObjectAda compiler is dirt cheap for personal
>>>use and the professional version is cheaper than a "professional" C++
>>>package.  Also, Gnat is on all sorts of platforms.
>>
>>Is it really?  When I look at the Ada 95 validated compiler list it 
>>looks pretty small to me.  So you're saying that gnat is available as
>>long as I am willing to pay for a validation and arrange for maintenence
>>or do it myself.  
>
>If the competition is C++ compilers, then I don't see what validation
>has to do with it.  I mean it's not as if you are going to find any
>validated C++ compilers.
>

Ah, well perhaps I am too tied to the "old" Ada business.  It used to be that
before you could claim you even had a product you would perform a validation
on the compiler.  This put you on the "validated compilers list", a form
of advertising.  It was also recognized that validation was merely the first
step of producing an Ada product.  I mean validation doesn't even require you
to produce a debugger!  It doesn't address the quality of the generated code
at all.  But, validation does at least test that the compiler does successfully
process the Ada language (at least to some minimal level).

Look, I think the validation process is a good thing.  It has a third party
verify that an Ada vendors claims about having run the ACVC and passed the
suite are true.  That should not be that big a deal.  I also believe that the
validation process is one of the things that Ada has going for it.  I am not
really in touch with C++, but I have heard that one of its problems is that
some of the language features are implemented inconsistently by the C++
vendor community.

And, *I thought*, that the DOD was required to use validated compilers.  But
I could be wrong on that.  But *if it does*, then in order to sell to the DOD
companies would need to be on the validation list.  Since the list is extremely
small when compared to Ada 83 I use it as the measure of how far along the
Ada 95 market is today.  One measure of how successful Ada 95 is will be how
fast that list grows in the next 12 months as the gnat and AdaMagic based
compilers start hitting the market.

Tom Robinson

Re: Is the "Ada mandate" being reconsidered?

[Michael Levasseur]

Theodore E. Dennison wrote:

> I have recently been directly involved in a DoD effort using large
> amounts of COTS hardware and software, and interfacing Ada to it has
> literally been the LEAST of our problems. Writing Ada bindings, even
> high level ones, is really not all that difficult for an experienced
> Ada developer. There are even some tools available to do it for you.
> Getting COTS hardware/software combinations that will work
> together, now THAT has been a nightmare!

Have you been building Ada bindings to C++? I've built plenty of
binding to C source code, i.e. VxWorks. I don't know of a way to
bind to C++ objects or C++ methods.

> I have seen neither of these. I suppose it depends on your
> definition of "credible". The major players, Alsys and Rational seem
> to be getting stonger than ever. And the last few months have seen
> the emergence of ACT.

Well one of the companies you sight "Alsys" no longer exists. They 
are part of "Telesoft"... Oh no now both "Alsys" and "Telesoft" no longer
exist they're "Thomson". Well there's "Verdix" oh wait that's part of
"Rational". From conversations with "Rational" what they're working
on is their C++ compiler. My information comes from the list of
validated compilers. It is clear that the number of Ada compiler
venders have been shrinking. There aren't alot of companies entering
the Ada compiler market. There is also some fallout from the TI
aquisition of Tartan. I'm fairly sure that the product line of
Tartan will be shrank.

> In any event, you can now get an Ada compile for FREE.

Yes, but do you know of any major DoD projects that are be developed
using GNAT? When you buy a FREE compiler you get what you pay for.


Remember what the "Ada Mandate" was created to do. It was created to
reduce the DoD software maintanence of over 200 langauges down to
just one. The commercial world has now paired the major languages
down to a handfull. It is now time to drop the Ada Mandate! If
Ada is as strong as everyone says, clearly it will thrive and survive.
I personnally believe that if that "Ada Mandate" was removed,
Ada would take it's place in history like Jovial, Pascal, and Atlas...

These are just my humble opinion from an Ada Programmer, yes I'm
an Ada programmer!!!

Re: Is the "Ada mandate" being reconsidered?

[Jon S Anthony]

In article <4pn0rs$mbe@gde.GDEsystems.COM> Tom Robinson <robinson@gdesystems.com> writes:

> >Well, this sounds good, it's just wrong.  Gnat is free and of better
> >overall quality than all but the very best Ada83 compilers.
>
> I am interested in what measure you are using for quality.  From the
> small snipits I have read on the net it is not clear that the
> generated code quality of gnat is quite up to current standards yet.
> Am I missing something here?  Has anyone done ACEC or PIWG
> comparisons yet?

I am speaking of correctness, robustness, flexibility, fullness of
implementation, error messages and error recovery, integration with
widely available and common tools, and speed of compilation.  Actually
in terms of speed I don't think any of the Ada83 compilers can even
come close (well, discounting certain incremental compilation
situations in something like Apex).  No way.

The generated code quality is often about the same as for GCC C
compiler.  This is really quite good, but there are Ada constructs at
present which can derail things.  But this was true of Ada83 compilers
as well.  For me that is "good enough" for now.  If you are in some
sort of "hard real-time" situation where the code has to be as good
as possible, then I would be very careful or looking elsewhere.


> >Thomson's ObjectAda compiler is dirt cheap for personal
> >use and the professional version is cheaper than a "professional" C++
> >package.  Also, Gnat is on all sorts of platforms.
> 
> Is it really?  When I look at the Ada 95 validated compiler list it 
> looks pretty small to me.  So you're saying that gnat is available as
> long as I am willing to pay for a validation and arrange for maintenence
> or do it myself.  

Yes, really.  Who cares about validation?  Maybe you do, but I don't.
I _like_ the fact that there is an ACVC for all sorts of QA reasons.
But the actual stamp means nothing to me.  I'll tell you a little
secret: C/C++ programmers don't care about AJPO validation
certificates either.  For me (and I am sure most "usual" commercial
developers), if I knew the thing passed the test suites (ACVC and any
internal ones - especially ones that _I_ submitted) on _any_ common
platform, that would be quite sufficient.  And even if it failed a
few "goofy" ACVC tests, that wouldn't break my heart either (even
though that would prevent an official stamp).




> >And since
> >ObjectAda uses the Intermetrics AdaMagic frontend, it will likely be
> >all over the place too (or at least highly compatible counterparts
> >based on the same frontend).
> >
> >No, overall, this situation is _vastly_ better today than before.
> >
> >
> 
> I don't understand how you can say that the situation is _vastly_ better
> based on what you have stated.  It is a fact that the Ada 83 compiler
> choices have been shrinking.

I can state it very easily because it is true.  People can use obtain
and use very high quality compilers for all sorts of common platforms
(PCs in particular) for little or nothing.  In my book that is indeed
_VASTLY_ better than 5 or 6 years ago where you had it on only a
couple common platforms and to get it would cost a fortune.  I certainly
would not be using it if the situation were as before - I just plain
couldn't!!


>  All one has to do is look at the list of Validated compilers and
> realize that many companies that did Ada 83 validations (Alsys,
> TeleSoft, Verdix, Meridian, Systeam, (off the top of my head)) are
> no longer operating under those names.  They have merged with other
> companies or gone out of business

So?  Who cares?  The same thing is happening in the compiler biz no
matter the language.  The question is, are the products of today
better fitting into the market place?  From where I sit the answer
isn't just "yes" it's "HELL YES!"


> announced that they will use non-proprietary front ends for their
> product offerings (DEC).

One hell of a good idea, if you ask me.  A smart idea which allows for
putting true value added where it means something - not just
duplicating the same effort over and over again.  Intelligent reuse
with continual improvement!  Imagine that - in software!  Go figure!
BTW, where do you think most of the C and C++ compilers have come
from?  Hmmmm?


>  Other OEMs have transitioned from directly
> offering Ada products to offering them through a 3rd party (IBM->OCSystems).

Great idea.  Excellent idea.  Why the hell should IBM (a _hardware_
company) be making Ada (or any) compilers?

 
> I think that the Ada market is in transition.  As an Ada user the
> transition to Ada 95 seems like it might be a bit tricky.  For large
> projects it looks like there are going to be more decisions that I
> (as a buyer of Ada technology) will need to make:

Sure, it is in transition.  And that is a VERY good thing.  If it
weren't it would be dead.  I don't really understand your concerns
here, but I'm sure that for you they are legitimate.

For us, the decision was obvious.  One of the things we did here was
move a "large" (around 200KSLOC) legacy '83 program over to Ada95 with
GNAT (and yes, we have ACT support - which BTW has been much better
than what I am used to from other software vendors - be they compiler
shops or whatever).  This was from a Verdix based project.  The
situation is now so much better than before, that it is simply
wog-boggling.  We can and we have moved it to several other platforms
- including Win95.

No, for the Ada world, the situation is so much better than a few
years ago, that it is positively stupefying.


/Jon
-- 
Jon Anthony
Organon Motives, Inc.
1 Williston Road, Suite 4
Belmont, MA 02178

617.484.3383
jsa@organon.com

Re: Is the "Ada mandate" being reconsidered?

[Jon S Anthony]

In article <Dsx9Dt.y3.0.-s@inmet.camb.inmet.com> stt@henning.camb.inmet.com (Tucker Taft) writes:

> Note that this approach to having multiple compiler vendors
> use a common front end has a long history in the C, C++, and Fortran 
> marketplace, and is even increasing in the C++ marketplace with the 
> growing predominance of Edison Design Group's C++ front end.

Exactly.  This is precisely the sort of thing that we should all be
_ecstatic_ about: A first step toward true interoperable reuse of
software, "OK, fine - I'll buy your frontend because it is better than
ACME's, and put our special purpose backend for the DELTA 3000
embedded chip on it.  Great!  All your current users will
automatically be quite familiar with the look and feel and use of the
thing."

The HW guys have known this "secret" for ages.


> The justification is simply that the more people who use the same
> front end, the fewer bugs will be left in it for *you* to stumble
> over.

Exactly.

> In fact, the proliferation of Ada 83 front ends I believe
> was one of the major problems with the Ada 83 marketplace.

Abso-fraggen-lutely.

>[spot on stuff snipped]


> With the new, smaller number of front ends, more coordination 
> is likely (e.g. the GNAT and the Intermetrics folks keep in 
> close touch on pragmas and attributes, etc), and you can
> more likely find another compiler based on the same front end on 
> your next target.

Yes, the benefits of this really can't be overestimated at present.


> As an anecdote, there are several users who have found that 
> the biggest expense in porting from Ada 83 to Ada 95 is the inevitable
> expense in porting from front-end to front-end, not in accommodating 83=>95 
> language changes

Quite true.  For us this accounted for _all_ the significant problems
in moving a piece of legacy code to Ada95.


>  Having lots of front ends out there is not 
> necessarily a "good thing" given limited overall resources...

You are being too kind.  It would be the kiss of death for Ada.


/Jon
-- 
Jon Anthony
Organon Motives, Inc.
1 Williston Road, Suite 4
Belmont, MA 02178

617.484.3383
jsa@organon.com

Re: Is the "Ada mandate" being reconsidered?

[Robert Dewar]

Tom Robinson says

"I am interested in what measure you are using for quality.  From the small
snipits I have read on the net it is not clear that the generated code
quality  of gnat is quite up to current standards yet.  Am I missing something
here?  Has anyone done ACEC or PIWG comparisons yet?"

As for most compilers, GNAT does some things well, and some things not
as well. For some programs, GNAT is the fastest Ada compiler around (in
terms of generated code), these are programs for which the backend of GCC
(which is very highly optimized) can generate really good code.

For some Ada specific constructs (notably check generation, aggregates,
exception handling and unconstrained array handling), there is a way to go,
but look to see major improvements in GNAT in all these areas in the near
future.

If you benchmark GNAT against other compilers, you will find that some
things are faster, some are slower. Overall, the results are pretty
comparable, which is quite interesting considering that in these cases
you are typically comparing an old technology where a LOT of effort has
been put into improving the quality of code in general (and the PIWG's
in particular) with a new technology where so far almost no effort has
been put into improving efficiency of generated code (we have been much
more concerned with functionality and reliability).

Robert Dewar
ACT

Re: Is the "Ada mandate" being reconsidered?

[Robert Dewar]

Tom said

"Ah, well perhaps I am too tied to the "old" Ada business.  It used to be that
before you could claim you even had a product you would perform a validation
on the compiler.  This put you on the "validated compilers list", a form
of advertising.  It was also recognized that validation was merely the first
step of producing an Ada product.  I mean validation doesn't even require you
to produce a debugger!  It doesn't address the quality of the generated code
at all.  But, validation does at least test that the compiler does successfully
process the Ada language (at least to some minimal level).
"

Yes, you are too tied to the old Ada business.

Sure validation is important, but during the transition period to Ada 95,
it is less important than in the traditional Ada 83 setting.

Also validation does not mean so much during the transitional period. Of
14 compilers in the validated compiler list right now, several (6?) pass
NONE of the Ada 95 tests, they are Ada 83 compilers only. 

If you need an Ada 95 compiler, you have to figure out what the best choice
is. In some cases, a non-validated compiler may be the best choice. For
example, you may do better to get a full language compiler that is not
validated yet, than a validated compiler which does not implement any
of the Ada 95 features yet.

Eventually (a year from now), the validation procedures will fall more in
line with what you are used to, but right now, validation does not have quite
the same significance that you are used to.

This deinitely causes some confusion! There are government projects that
require Ada 95 and require a validated compiler, and will find that
right now, if they stick to the valiated compiler list, they will be
forced to choose a compiler that implements NONE of the Ada 95 features.

I expect that a lot more compilers will show up on the list soon. We could
validate a lot of targets with GNAT very soon if that were our highest
priority, but we will probably wait will we can do a bunch together
(we will certainly be validating the Alpha VMS version later this year,
and probably several others).

RObert Dewar
ACT

Re: Is the "Ada mandate" being reconsidered?

[Robert Dewar]

Tom Robinson said

"And, *I thought*, that the DOD was required to use validated compilers.  But
I could be wrong on that.  But *if it does*, then in order to sell to the DOD
companies would need to be on the validation list.  Since the list is extremely
small when compared to Ada 83 I use it as the measure of how far along the
Ada 95 market is today.  One measure of how successful Ada 95 is will be how
fast that list grows in the next 12 months as the gnat and AdaMagic based
compilers start hitting the market."

Using this list for this purpose is misleading, since, as I noted in my
previous message, the list includes compilers that are not Ada 95 compilers
at all, and does not includes most of the dozens of ports of GNAT, which
is the most complete Ada 95 implementation around.

During the transition period, a significant number of DoD projects are
using non-validated compilers, including GNAT. In particular, Airfields,
the first deployed Ada 95 DoD application (if you know an earlier one,
speak up :-) used GNAT on a non-validated platform (Solaris).

Sure it will be interesting to see how fast the list grows, but much more
interesting is to see how fast the list of projects using Ada 95 grows,
which is not necessarily the same measure at all.

Remember that the validation suite that corresponds to the ACVC suites
you are used to (which cover the whole language and require 100%
compliance) will not be in use for another year yet! Your viewpoint
may be appropriate for a year from now, but during the transition period
the situation is quite different.

P.S. for a more detailed explanation of how Ada 95 validation works,
you can visit the Thomson home page. They have a page discussing the
important things to look out for in evaluating Ada 95 validations.
It's not completely general, because (not surprisingly :-) it is most
concerned with contrasting the Thomson validation with others, but it
is useful reading on the subject!

Re: Is the "Ada mandate" being reconsidered?

[Tom Robinson]

>
>As an anecdote, there are several users who have found that 
>the biggest expense in porting from Ada 83 to Ada 95 is the inevitable
>expense in porting from front-end to front-end, not in accommodating 83=>95 
>language changes.
>

I certainly agree with this!

> Having lots of front ends out there is not 
>necessarily a "good thing" given limited overall resources...
>

And, all I have been saying is that time will tell.  I simply believe that
it is too early to start proclaiming that between Adamagic and Gnat we now
have a much better situation today than we did under Ada 83.

I guess I'm just a little cynical.  I still remember sitting through many
company meetings of a now defunct Ada vendor and hearing all the marketing 
hype about how Ada was going to be the next dominant programming language.
(I think at the time it was being compared to Cobol -- at least from a
language market share standpoint).  The net result of alot of this hype was
that many private investors put up money to fund Ada 83 development.  Many
computer manufacturers also put up money to make sure that Ada was available
on their platforms.  And we now have a situation where many of those companies
are gone and many of those manufacturers are getting out.

But, I am ready to be pleasantly suprised :)

Tom Robinson

Re: Is the "Ada mandate" being reconsidered?

[Jim Kingdon]

> Yes, but do you know of any major DoD projects that are be developed
> using GNAT? When you buy a FREE compiler you get what you pay for.

I should probably let ACT comment on details, if there is anything
which merits a public comment at this time, but suffice it to say that
your assumption about whether GNAT is used on major DoD projects is,
er, uh, questionable.

I partly agree with "you get what you pay for", but free software
businesses such as ACT are happy to take your money :-).

Re: Is the "Ada mandate" being reconsidered?

[Jon S Anthony]

In article <4ppb89$gbq@gde.GDEsystems.COM> Michael Levasseur <levass@gdesystems.com> writes:

> Theodore E. Dennison wrote:
>...
> > In any event, you can now get an Ada compile for FREE.
> 
> Yes, but do you know of any major DoD projects that are be developed
> using GNAT? When you buy a FREE compiler you get what you pay for.

Hmmm, in this case what you get is something better (in some cases a
_LOT_ better) than most of the Ada83 compilers that cost 10s of
thousands of dollars per seat.  The trick is to get support with
it (which as I've said has been for me much better than what I've
seen from others software vendors).


> Remember what the "Ada Mandate" was created to do. It was created to
> reduce the DoD software maintanence of over 200 langauges down to
> just one. The commercial world has now paired the major languages
> down to a handfull. It is now time to drop the Ada Mandate! If
> Ada is as strong as everyone says, clearly it will thrive and survive.

Yes, I believe this may be the right thing to do.  It might even help
remove the stigma of government from the language.  It is a tricky and
non obvious issue, but overall, this probably would be a good idea.


> I personnally believe that if that "Ada Mandate" was removed,
> Ada would take it's place in history like Jovial, Pascal, and Atlas...

I don't think so - these examples were either not commercially viable
or too limited in scope.  Another way of looking at this is to look at
the Eiffel market.  While not jumbo, there are signs that there is a
truly viable market emerging there even though Eiffel does not yet
have the level of support and use that Ada currently does.  So, this
is kind of a "proof" of concept that it is possible to market a good
software engineering language.  The real key is to hook enough
_commercial_ development into seeing the significantly better
capabilities of Ada over C (or C++) and that it has sufficient support
so that projects have good assurance that they will realize those
benefits.  Whether the government has enough brains to follow this is
really irrelevant (my guess is they don't - being knee jerk
"fadophiles").

/Jon

-- 
Jon Anthony
Organon Motives, Inc.
1 Williston Road, Suite 4
Belmont, MA 02178

617.484.3383
jsa@organon.com

Re: Is the "Ada mandate" being reconsidered?

[Jon S Anthony]

In article <4ppceg$gha@gde.GDEsystems.COM> Tom Robinson <robinson@gdesystems.com> writes:

> small when compared to Ada 83 I use it as the measure of how far
> along the Ada 95 market is today.  One measure of how successful Ada
> 95 is will be how fast that list grows in the next 12 months as the
> gnat and AdaMagic based compilers start hitting the market.

I don't think this is a very good measure.  I would guess (don't know
for sure, but would be surprised otherwise) that well over half of all
C (and C++) is done using only three compilers: MSC++, Borland C/C++
and Gnu C (with g++).  It wouldn't surprise me if these accounted for
three quarters of the code.  Even at only half of all such code, that
is a _very_ successful market.  With only a couple three vendors.

The real measure for how successful Ada95 is and will be, will be
reflected in how many _customers_ ACT (including Digital and SGI),
Thompson, Intermetrics, and OC Systems have and will get.  Even if
they just plain quit validating altogether.  As for the goofy
government, well, I don't see why they should be wigged out over using
an unvalidated Ada compiler when they are perfectly willing to use
MSC++.  If "you're" (rhetorical you) going to waste time obtaining
waivers, why not get a waiver for validation?

/Jon
-- 
Jon Anthony
Organon Motives, Inc.
1 Williston Road, Suite 4
Belmont, MA 02178

617.484.3383
jsa@organon.com

Re: Is the "Ada mandate" being reconsidered?

[Theodore E. Dennison]

Michael Levasseur wrote:
> 
> Theodore E. Dennison wrote:
> 
> > I have recently been directly involved in a DoD effort using large
> > amounts of COTS hardware and software, and interfacing Ada to it has
> > literally been the LEAST of our problems. Writing Ada bindings, even
...
> > Getting COTS hardware/software combinations that will work
> > together, now THAT has been a nightmare!
> 
> Have you been building Ada bindings to C++? I've built plenty of
> binding to C source code, i.e. VxWorks. I don't know of a way to
> bind to C++ objects or C++ methods.

Nope. Haven't needed to.(That was my point, vis-a-vis COTS). 


> > I have seen neither of these. I suppose it depends on your
> > definition of "credible". The major players, Alsys and Rational seem
> > to be getting stonger than ever. And the last few months have seen
> > the emergence of ACT.
> 
> Well one of the companies you sight "Alsys" no longer exists. They
> are part of "Telesoft"... Oh no now both "Alsys" and "Telesoft" no longer
> exist they're "Thomson". Well there's "Verdix" oh wait that's part of
> "Rational". From conversations with "Rational" what they're working

That was 2 years ago. The current trend is quite different (If you can even
call something a "trend" based on such small numbers).

> Yes, but do you know of any major DoD projects that are be developed
> using GNAT? When you buy a FREE compiler you get what you pay for.

I do know of a couple being proposed using it (remember, GNAT is validated
on some platforms, including SGI's). 

Frankly, I have found that "you get what you pay for" adage quite untrue.
In my experience, "you pay what the vendor thinks you can afford (or a 
little more)" would be more accurate. 

-- 
T.E.D.          
                |  Work - mailto:dennison@escmail.orl.mmc.com  |
                |  Home - mailto:dennison@iag.net              |
                |  URL  - http://www.iag.net/~dennison         |

Re: Is the "Ada mandate" being reconsidered?

[Robert Dewar]

Michael said:

"> Yes, but do you know of any major DoD projects that are be developed
> using GNAT? When you buy a FREE compiler you get what you pay for."

A number of major projects are being developed using GNAT, including DoD
projects, e.g. one component of the Aegis project is using GNAT. Also the
first deployed Ada 95 DoD project, Airfields, used GNAT.

One thing to remember about the FREE in free software is that it primarily
about freedom of distribution not about the price -- indeed there is nothing
to stop anyone, including us, from charging for GNAT, although we choose
not to as long as it is obtained by FTP. As Richard Stallman says, the
free is for free as in free speach, not free as in free beer. 

Any serious project has to make sure the tools it is using are supported.
One interesting possibility opened up by the free software model is to
maintain the softare yourself, since you have full access to the sources.
However, that does not make sense for most projects, and serious DoD
projects that I know about get support from SGI or ACT (of course I
would not necessarily know about some secret project that had decided
to do its own maintenance).

Do not assume that because software is free software it is inferior to
proprietary software for which you have to pay. The fact of the matter
is that there is good and bad free software and good and bad proprietary
software. You have to evaluate the particular product in question and
see if it meets your needs, something that is true of any software in
any context. Of course it is convenient that you can do at least a 
preliminary evaluation of free software rather easily without having
to sign contracts and pay money -- another advantage of free software.

As for validation, it turns out that even in the DoD during this
transitional period, validation is less important than making sure
the compiler you are using supports the Ada 95 features you need
(validation does not guarantee that AT ALL at the moment), and
that in general it meets your needs.

Of the two projects I mentioned above, Airfields used the Solaris version
of GNAT which is not yet validated, and the Aegis project is using the
SGI version, which has been validated (was in fact the first general
purpose Ada 95 validation, and is still the only validation of a compiler
covering all the annexes).

We expect many additional versions of GNAT, including the SOlaris version
to be validated sometime soon. The order in which we carry out validations
depends on the needs and requirements of our customers.

Robert Dewar
ACT

Re: Is the "Ada mandate" being reconsidered?

[Ken Garlington]

Michael Levasseur wrote:

> The commercial world has now paired the major languages
> down to a handfull.

Could you list this "handful"? From what I read in the commercial magazines, the 
list of languages is constantly _growing_. What's more, very few languages in the 
commericial world are disappering (e.g., COBOL and FORTRAN are still going strong, 
and variations of both continue to be introduced).

> It is now time to drop the Ada Mandate! If
> Ada is as strong as everyone says, clearly it will thrive and survive.
> I personnally believe that if that "Ada Mandate" was removed,
> Ada would take it's place in history like Jovial, Pascal, and Atlas...

You mean, yet another language for DoD to maintain, along with Objective-C,
Visual Basic, Concurrent-Visual-C++, etc. etc.?

I don't see how this helps...

> These are just my humble opinion from an Ada Programmer, yes I'm
> an Ada programmer!!!

Unfortunately, the Ada policy is designed to address non-programmer concerns,
like language proliferation. That's why it's generally uninteresting to talk about 
the technical strengths of Ada vs. {whatever} when discussing changes to Ada 
policy. The discussion is in the wrong domain!

-- 
LMTAS - "Our Brand Means Quality"

Re: Is the "Ada mandate" being reconsidered?

[Theodore E. Dennison]

Robert Dewar wrote:
> using non-validated compilers, including GNAT. In particular, Airfields,
> the first deployed Ada 95 DoD application (if you know an earlier one,
> speak up :-) used GNAT on a non-validated platform (Solaris).

Would summer of '93 beat it?

Of course said project should probably have used at least ONE new feature to
qualify...


-- 
T.E.D.          
                |  Work - mailto:dennison@escmail.orl.mmc.com  |
                |  Home - mailto:dennison@iag.net              |
                |  URL  - http://www.iag.net/~dennison         |

Re: Is the "Ada mandate" being reconsidered?

[Theodore E. Dennison]

Theodore E. Dennison wrote:
> 
> Robert Dewar wrote:
> > using non-validated compilers, including GNAT. In particular, Airfields,
> > the first deployed Ada 95 DoD application (if you know an earlier one,
> > speak up :-) used GNAT on a non-validated platform (Solaris).
> 
> Would summer of '93 beat it?

Errr. That's summer of '94.

(Math not Zathras strength)



-- 
T.E.D.          
                |  Work - mailto:dennison@escmail.orl.mmc.com  |
                |  Home - mailto:dennison@iag.net              |
                |  URL  - http://www.iag.net/~dennison         |

Re: Is the "Ada mandate" being reconsidered?

[Joe Gwinn]

In article <4ppb89$gbq@gde.GDEsystems.COM>, Michael Levasseur
<levass@gdesystems.com> wrote:

> Theodore E. Dennison wrote:
> 
> > I have recently been directly involved in a DoD effort using large
> > amounts of COTS hardware and software, and interfacing Ada to it has
> > literally been the LEAST of our problems. Writing Ada bindings, even
> > high level ones, is really not all that difficult for an experienced
> > Ada developer. There are even some tools available to do it for you.
> > Getting COTS hardware/software combinations that will work
> > together, now THAT has been a nightmare!
> 
> Have you been building Ada bindings to C++? I've built plenty of
> binding to C source code, i.e. VxWorks. I don't know of a way to
> bind to C++ objects or C++ methods.

For mixed-language developments, covering Ada plus C and/or fortran for
instance, I would suggest people consider Green Hills, who I have not
heard mentioned in these discussions.  They have done quite well in our
evaluations, although no selection decision has been made yet.  

With COTS and NDI, and integration with various GUIs and operating
systems, I suspect that many current and future systems will be of
necessity mixed-language at least in the necessary-but-boring
underpinnings.


> > I have seen neither of these. I suppose it depends on your
> > definition of "credible". The major players, Alsys and Rational seem
> > to be getting stonger than ever. And the last few months have seen
> > the emergence of ACT.
> 
> Well one of the companies you sight "Alsys" no longer exists. They 
> are part of "Telesoft"... Oh no now both "Alsys" and "Telesoft" no longer
> exist they're "Thomson". Well there's "Verdix" oh wait that's part of
> "Rational". From conversations with "Rational" what they're working
> on is their C++ compiler. My information comes from the list of
> validated compilers. It is clear that the number of Ada compiler
> venders have been shrinking. There aren't alot of companies entering
> the Ada compiler market. There is also some fallout from the TI
> aquisition of Tartan. I'm fairly sure that the product line of
> Tartan will be shrank.

For large-scale use, only a few vendors survive, and consolidations will
likely continue.  The Ada market is clearly shrinking.  I submit the only
metric that counts is aggregate revenue to the Ada compiler and tool
vendors, as the rate of further development of Ada depends on the size of
their food supply.  I don't have the revenue figures, but I bet someone on
this newsgroup does, and ask that the figures from 1983 to present be
posted.


> > In any event, you can now get an Ada compile for FREE.
> 
> Yes, but do you know of any major DoD projects that are be developed
> using GNAT? When you buy a FREE compiler you get what you pay for.

I would comment that I have seen large ATC projects use Gnu C as their
main language, with some success, so it isn't obvious that one could not
use GNAT.  However, Gnu C is a great deal older and morermature then GNAT,
and has better support in general, simply because of its ten to a hundred
times larger user base.


> Remember what the "Ada Mandate" was created to do. It was created to
> reduce the DoD software maintanence of over 200 langauges down to
> just one. The commercial world has now paired the major languages
> down to a handfull. It is now time to drop the Ada Mandate! If
> Ada is as strong as everyone says, clearly it will thrive and survive.
> I personnally believe that if that "Ada Mandate" was removed,
> Ada would take it's place in history like Jovial, Pascal, and Atlas...

My experience has been that management attempts to dictate technical
decisions based on techno-political correctness usually results in severe
resistance -- the doers know from experience that the managers will be
long gone onto yet another techno-political fad while the doers are still
struggling to recover from the last one.  So, the doers judge it better to
stop the fad at the door.  

The managers fear to directly make the technical decision themselves,
because they sense that they will end up wearing it like an albatross when
it fails.  So, they badger the doers in the hope that they will "accept
the challenge".

Joe Gwinn

Re: Is the "Ada mandate" being reconsidered?

[Richard Riehle]

On 14 Jun 1996, Jim Kingdon wrote:

> > Yes, but do you know of any major DoD projects that are be developed
> > using GNAT? When you buy a FREE compiler you get what you pay for.
>
> I should probably let ACT comment on details, if there is anything
> which merits a public comment at this time, but suffice it to say that
> your assumption about whether GNAT is used on major DoD projects is,
> er, uh, questionable.

  If I may attempt a answer to this based on experience with clients
  who are initiating Ada 95 projects ...

  GNAT, by itself, is not enough. So far, my clients have been able to
  take advantage of the ports of GNAT as integrated into development
  environments and debugers, etc.  It turns out that the GNAT compiler
  is a pretty good "engine" for these environments. The work that ACT
  does to make the GNAT engine correspond to the targeted platform works
  out pretty well, too.

  The jury is still out on the comercial compilers since many of them
  are only partial implementations of the standard.  As this changes, and
  as the commercial compiler publishers fine tune their products, one
  of GNAT's major benefits will become manifest: it raises the standard
  against which the commercial compilers must compete.

  Meanwhile, real projects are being developed in Ada 95 using GNAT for
  the development, but which may be deployed using some entirely different
  compiler.  This is one of the virtues of Ada.

  Richard Riehle

Re: Is the "Ada mandate" being reconsidered?

[Robert Dewar]

Richard Riehle says

"  Meanwhile, real projects are being developed in Ada 95 using GNAT for
  the development, but which may be deployed using some entirely different
  compiler.  This is one of the virtues of Ada."

Certainly it is true that projects can be developed on one compiler and
deployed on another. For example, we have several customers planning to
develop using Rational Apex, and deploy using GNAT, and I would guess that
all sorts of other combinations make sense in some environment or other.

As to GNAT not being enough on its own, of course it isn't. No compiler
is usable entirely on its own, but instead is used in conjunction with
a suite of development tools suitable to the environment. What 
distinguishes GNAT from some other Ada systems is that since GNAT has
a very standard compilation model and view of the world, it works fine
with lost of standard tools, so you can pick and choose what tools
you need. Some GNAT users use EMACS as an integrating environment for
the compiler, GDB, RCS, GNATF and other tools that embed nicely into
EMACS, others use SGI Workshop, others use other available tools (e.g.
several of our users have found Purify and Gprof useful, both of which
work with GNAT).

From a users point of view, it seems to me that the advantageous thing
is to have as many choices as possible for both Ada compilers and
Ada development tools!

Robert Dewar
ACT

Re: Is the "Ada mandate" being reconsidered?

[Carl Bowman]

Sections snipped for brevity.

NOTE: In this post I am not speaking for either the AJPO or the AdaIC.
No part of this post may be considered DoD, AJPO or AdaIC policy.  If
you have questions about such policy, send email to me and I will
forward the email to the appropriate party, or you may contact the
AJPO or AdaIC directly.

In article <4ppceg$gha@gde.GDEsystems.COM>,
Tom Robinson  <robinson@gdesystems.com> wrote:
>fjh@mundook.cs.mu.OZ.AU (Fergus Henderson) wrote:
>>Tom Robinson <robinson@gdesystems.com> writes:
>>>[someone writes]:

...Snip!

>>>
>>>Is it really?  When I look at the Ada 95 validated compiler list it 
>>>looks pretty small to me.  So you're saying that gnat is available as
>>>long as I am willing to pay for a validation and arrange for maintenence
>>>or do it myself.  
>>

When Ada83 was young, we would often suggest potential users should
contact vendors with similar compilers and arrange derived
validations.  I think such requests from the user community alerted
vendors to which compilers to prioritize on their validation list.  I
don't think in EVERY case the cost of validation was paid for by the
user, but I plead ignorance in this matter.

...Snip!

>>
>
>Ah, well perhaps I am too tied to the "old" Ada business.  It used to be that
>before you could claim you even had a product you would perform a validation
>on the compiler.  This put you on the "validated compilers list", a form
>of advertising.  

This is news to me!  There were plenty of conferences I attended where
a vendor would announce a new Ada compiler was validated that had not
completed the full validation process.  I would have the official list
with me!

...Snip!

>And, *I thought*, that the DOD was required to use validated compilers.  But
>I could be wrong on that.  But *if it does*, then in order to sell to the DOD
>companies would need to be on the validation list.  Since the list is 
>extremely
>small when compared to Ada 83 I use it as the measure of how far along the
>Ada 95 market is today.  One measure of how successful Ada 95 is will be how
>fast that list grows in the next 12 months as the gnat and AdaMagic based
>compilers start hitting the market.
>
>Tom Robinson
>

For the moment, Ada 83 compilers are still validated compilers.  See
Hon. Emmett Paige, Jr.'s memo for guidance on use of non-validated
compilers that plan to become validated.  (AdaIC form P145) You can
start development with a compiler that is not currently validated, as
long as delivery is with a validated compiler.

PERSONALLY ( and some of my colleagues will surely disagree) - I don't
see the growth of the validated compilers list as a measure of Ada
95's success at all.  That is mearly a measure of the Ada 95 compilers
that have been officially validated.  

Nothing would please me more than to see Ada 95 used universally -
validated when conditions require, and non-validated when acceptable.
Why in the world use validated compilers for education, personal use,
or in-house commercial products where portability and standardization
are not an issue?  A few compiler companies had non-validatd education
versions of their compilers.  I can tell you non-validated GNAT is
doing a lot for Ada's re-emergence.

I think many commercial users would use a compiler that guaranteed
compliance and support, but didn't officially validate.

If we use validation as a measure of success, surely C and C++ must be
dismal failures.

- Carl Bowman

NOTE: In this post I am not speaking for either the AJPO or the AdaIC.
No part of this post may be considered DoD, AJPO or AdaIC policy.  If
you have questions about such policy, send email to me and I will forward
the email to the appropriate party, or you may contact the AJPO or
AdaIC directly.

Re: Is the "Ada mandate" being reconsidered?

[Bob Kitzberger]

[I hope my attribution line is correct...]

In article <4ppb89$gbq@gde.GDEsystems.COM>, Michael Levasseur
<levass@gdesystems.com> wrote:

> "Rational". From conversations with "Rational" what they're working
> on is their C++ compiler. 

I'm not sure who you talked to at Rational, but this is very misleading.
Yes, we are working on our C++ compiler.  And our Ada 83 and Ada 95
compilers, and our kernel, and our OOD tools, and our CMVC tools,
and our test automation tools, and our process automation tools,
and our document automation tools, and our safety-critical tools ...

We believe that a comprehensive toolset, not only a compiler, is 
required to build any significant application.

--
Bob Kitzberger	      Rational Software Corporation       rlk@rational.com
http://www.rational.com http://www.rational.com/pst/products/testmate.html

Re: Front Ends (was: Re: Is the "Ada mandate" being reconsidered?)

[Tom Robinson]

>Dave Wood <dpw@thomsoft.com> wrote:
>>Tom Robinson wrote:
>>
>>>   (3) Buy a product from a vendor that does not have direct support
>>>       for fixing their own front end problems (AdaMagic).
>>
>>Tom,
>>
>>The above referring presumably to TSP's ObjectAda, the statement
>>is flat-out incorrect and the product of mistaken assumptions
>>about the nature of our rights and plans WRT AdaMagic.

I don't believe I ever mentioned ObjectAda, you did.  I should not have even
used AdaMagic as an example.  You are correct in that I don't know "the nature"
of your rights and plans.  I was more referring to the tendency to take a
variety of third party software components, integrate them into a single
box and sell it as a product.  As a customer, if I know I am going to be
content with the product "as is" then this is fine.  However, if I am going
to make demands on the vendor that actually require them to understand and
service the components they ship, then I would be foolish not to try to
understand the level of support they are capable of delivering.  

And this is where my personal *bias* comes in.  I *have a belief* that it
is more difficult for a customer to get problems fixed in third party software.
Why?  There are two ways to *get a fix* in this situation.  Either the vendor
fixes the problem directly or they get the third party to fix the problem.
In case one we (the customers) are relying on the vendor developing expertise
in software that they did not want to spend the money and time to develop.  It 
is most likely that the *expertise* required to provide problem solutions will
be developed as the problems (my problems) are solved.  In case two we are
relying on the selling vendor having a good enough relationship with the
third party vendor to insure that the problems (my problems) are solved and
also to coordinate to provide me with emergency updates.

So I guess (since you brought it up) I would advise customers that
they need to understand the relationships and "nature" of the agreements
that are used to create products that are integrated from multiple sources.
Which really goes back to my original post.  My only point was that there
are multiple types of choices one can make in the Ada 95 market.  As a
customer I need to understand that each of those choices has different
implications on how *I* will do business.

Tom Robinson

Re: Front Ends (was: Re: Is the "Ada mandate" being reconsidered?)

[Jon S Anthony]

In article <4sm455$338@gde.GDEsystems.COM> Tom Robinson <robinson@gdesystems.com> writes:

> Why?  There are two ways to *get a fix* in this situation.  Either
> the vendor fixes the problem directly or they get the third party to
> fix the problem.  In case one we (the customers) are relying on the
> vendor developing expertise in software that they did not want to
> spend the money and time to develop.  It is most likely that the
> *expertise* required to provide problem solutions will be developed
> as the problems (my problems) are solved.  In case two we are
> relying on the selling vendor having a good enough relationship with
> the third party vendor to insure that the problems (my problems) are
> solved and also to coordinate to provide me with emergency updates.

I think that this particular attitude/concern is another example of
the anamolous aspect of software as "product" and why it has such a
hard time moving beyond a kind of "blacksmith" level.  Note that most
vendors seem to express the mirror image of this from "their side of
the desk".

The computer you are using to compose your messages on has parts from
many different vendors.  If it's a PC, it's very likely that the
vendor with the name on the front didn't make any of the parts and
"just integrated" them to produce the complete system.  Same thing
with automobiles, airplanes, TVs, stereos, coffee makers, weed
wackers, etc., etc., etc.  I would bet you don't find yourself
worrying about the fact that Honda (or whoever) doesn't make the chips
in its ECUs when you think about purchasing a car.  You might wonder
which _ones_ they use, but not that they didn't make them.

Now, in software, this sort of thing is rare to non-existent.  The
obvious exceptions revolving around "commercial libraries" like
RogueWave.  How would you view the use of these in a "third party"
product you were looking at?

Still, most SW is all "custom" (including COTS - you don't believe MS
buys an outliner from Supreme SW, a GUI from WindozeRUS, and adds the
other bits to create Word, right?) and does not have the "problem" you
are worried about.  But this typically means that the true value added
for any given piece of software is only a small percentage of the
whole effort.  This is an old story.  It's a variant of the reuse
story, not the same though, as the single vendor may be reusing a lot
of its own stuff - not likely, but it does happen. And like the reuse
story there are certainly many issues to it.  One of which seems to be
a kind of chicken-or-egg thing with customers (as evidenced by your
concerns).

The AdaMagic frontend and ObjecAda example is quite interesting and
very appealing to me.  For one thing it is rather unique even for the
rare cases of software composed of "integrated parts".  This is
because most "integrated parts" examples are either very low level (a
RogueWave class or two was used) or the parts are really standalone
systems that were whacked together to form a larger system (the land
of the custom system integrator...)  But here, the two obvious bits
the frontend and the Windoze/Intel backend (and there are clearly many
others), are quite highlevel but are not "standalones".  Further, they
are produced by two completely different vendors.  And even more
interesting is that the result is one of the very few SW products
whose construction mirrors (at least to some extent) what happens in
non-SW products.  Overall quality and capability in such instances
will typically be much better than what any single vendor could do.

It's not clear how to make more of this sort of thing happen, but it
seems quite clear (IMO) that it is the sort of thing that needs to
happen if we are going to go beyond, "Hey, Emil!  When ya get a
chance, hows 'bout workin' me up some new spurs?"

/Jon
-- 
Jon Anthony
Organon Motives, Inc.
1 Williston Road, Suite 4
Belmont, MA 02178

617.484.3383
jsa@organon.com