Public Release of AdaSAGE (Re: Why is the DOE selling AdaSage?)

Public Release of AdaSAGE (Re: Why is the DOE selling AdaSage?)

[Robert I. Eachus]

     Funny you should mention it.  At TriAda'92 there was a public
forum to discuss where to go with AdaSAGE.  The current intent is to
set up a consortium to manage future versions, have the Idaho National
Energy Laboratory (part of the DoE) actually do the configuration
management, and distribute AdaSAGE through several channels including
the Free Software Foundation.  There will be charges for physical
distribution, but use and electronic distribution is free.

     If you have comments on this plan send them before March 1st to
adasage@ajpo.sei.cmu.edu or Mr. Alden C. Wynn (FAX: 208-526-1405).
(Posting the snail mail address seems unnecessary...) For more details
on the plan look in the adasage directory on the ajpo machine.

     Does that make you feel any better Fred?


					Robert I. Eachus

with Standard_Disclaimer;
use  Standard_Disclaimer;
function Message (Text: in Clever_Ideas) return Better_Ideas is...
--

					Robert I. Eachus

with Standard_Disclaimer;
use  Standard_Disclaimer;
function Message (Text: in Clever_Ideas) return Better_Ideas is...

Re: Public Release of AdaSAGE (Re: Why is the DOE selling AdaSage?)

[agate!spool.mu.edu!howland.reston.ans.net!sol.ctr.columbia.edu!eff!world!]

With regards to the public release of Adasage:

>     Funny you should mention it.  At TriAda'92 there was a public
>forum to discuss where to go with AdaSAGE.  The current intent is to
>set up a consortium to manage future versions, have the Idaho National
>Energy Laboratory (part of the DoE) actually do the configuration
>management, and distribute AdaSAGE through several channels including
>the Free Software Foundation.  There will be charges for physical
>distribution, but use and electronic distribution is free.
>
>     If you have comments on this plan send them before March 1st to
>adasage@ajpo.sei.cmu.edu or Mr. Alden C. Wynn (FAX: 208-526-1405).
>(Posting the snail mail address seems unnecessary...) For more details
>on the plan look in the adasage directory on the ajpo machine.

    I know it sickens you guys to consider the use of capitalism in any
form, but this is another boneheaded idea when it comes to reuse.  It is
bad enough that we have have two DoD reuse centers (DSRO and ASSET);
know you want to start a third (and fourth...).

    If you care at all about learning from free markets (I think you guys
have been spending other people's money for too damn long), you would learn
that a growing phenomena in the business world is that of "superstores"
for commodities such as food, home entertainment, and surprise, surprise,
computers.

    The attraction about superstores is that it provides one-stop shopping
at reasonable (tho not necessarily the best) pricing and support.  For
example, I make frequent use (when I spend my own money) of a local CompUSA
computer superstore.  They have six of everything for both hardware and
software, so I can make all my purchases and comparision shop at the same
time, asking questions when I get stuck (like how to hook up a modem to
Windows for Workgroups).

    All of these benefits could be had for DoD software reuse.  BUT NO!!!!
The DoD has to set up a growing number of reuse centers that duplicate each
other (wasting tax dollars) and make it inconvenient for people looking for
reusable software by having to deal with multiple centers.  Not that the
DoD should be in the BUSINESS of reuse centers, but if you are going to do
it, do it right.

    Thus the idea of using INEL to support Adasage is a further waste of
tax dollars and more confirmation that discussions at Tri-Ada, when it comes
to non-technical issues of Ada, are nothing more than a bunch of people
looking for their next government handout.  I hate to think of the money
waste in time for having a public forum at Tri-Ada about Adasage, when a
five minute discussion on comp.lang.ada, like this one, is about as useful.

    One more thing about Adasage.  While I have never looked at it, the
following comments I received by email attest to the fact that Adasage is
not good enough to deserve any special attention.  If you are going to
do something stupid, at least do it with something very good.

    Feel free to pass these comments onto Mr. Wynn, though if he is asking
this question, he is part of the problem instead of part of the solution.

Gregory Aharonian
Source Translation & Optimization


>We got a copy of it where I used to work. IMHO it's got to be the most
>non-portable bunch of Ada code ever written. It only works on DOS and
>can only be compiled with the Alsys compiler because it makes extensive 
>use of vendor dependent interface pragmas. Before it would be of any use
>to someone else it would have to be extensively modified.  If it ever
>made it onto the net, it would give Ada critics a nuclear weapon.

>I don't know.  I think the technology content in Adasage is above
>average, at least if you compare it to things like STARS...





-- 
**************************************************************************
Greg Aharonian
Source Translation & Optimiztion
P.O. Box 404, Belmont, MA 02178

Re: Public Release of AdaSAGE (Re: Why is the DOE selling AdaSage?)

[Robert I. Eachus]

In article <SRCTRAN.93Feb11102357@world.std.com> srctran@world.std.com (Gregory
 Aharonian) writes:

  >    I know it sickens you guys to consider the use of capitalism in any
  > form, but this is another boneheaded idea when it comes to reuse.  It is
  > bad enough that we have have two DoD reuse centers (DSRO and ASSET);
  > know you want to start a third (and fourth...).

   First, there is NO reason or excuse in what I posted for such
vicious slander.  I work for MITRE.  Most of what we do involves
supporting the government in dealing with commercial contracts.  I
assure you I am not sickened by the idea of capitalism.  However, I
see no reason why the government should not freely release software
developed at public expense, and which has never been proprietary in
any sense of the word.

   Second, even though it is obvious that Greg posted well before
thinking, I feel compelled to respond and point out that everything he
said is wrong.

    1) AdaSAGE will be/is available through DoD and DoE reuse centers
as well as through the Free Software Foundation.

    2) Since INEL (a part of the DoE) developed SAGE over ten years
ago and has actively supported SAGE and AdaSAGE since, anything other
than acknowledging and continuing that role would be stupid.  INEL's
primary role will continue to be development and configuration
management.

    3) I have little experience with AdaSAGE and no recent experience,
but the ordering information lists versions for 80x86, RS/6000, Sun
SPARC, and AT&T 3B2, with PC versions coming in both Alsys and
Meridian versions.  Sounds portable (or at least significantly ported)
to me.

    4)  Last but not least, my only current involvement with AdaSAGE
is as a potential user who thought that the readers of this newsgroup
might be interested in this information.


					Robert I. Eachus

with Standard_Disclaimer;
use  Standard_Disclaimer;
function Message (Text: in Clever_Ideas) return Better_Ideas is...    
--

					Robert I. Eachus

with Standard_Disclaimer;
use  Standard_Disclaimer;
function Message (Text: in Clever_Ideas) return Better_Ideas is...

Re: Public Release of AdaSAGE (Re: Why is the DOE selling AdaSage?)

[Douglas Rupp]

In article <EACHUS.93Feb11194235@oddjob.mitre.org> eachus@oddjob.mitre.org (Rob
ert I. Eachus) writes:

		....  stuff deleted ...

>
>    3) I have little experience with AdaSAGE and no recent experience,
>but the ordering information lists versions for 80x86, RS/6000, Sun
>SPARC, and AT&T 3B2, with PC versions coming in both Alsys and
>Meridian versions.  Sounds portable (or at least significantly ported)
>to me.
>
>    4)  Last but not least, my only current involvement with AdaSAGE
>is as a potential user who thought that the readers of this newsgroup
>might be interested in this information.
>
>
>					Robert I. Eachus
>
		.... more stuff deleted ...

As I recall AdaSage is distributed in source code form, therefore
it seems to me that an Ada program which needs to have a different version
for each and every platform and compiler in existence just lends credence
to the comments in Greg's last post about lack of portability.

A properly designed and written program should have compiler and platform
dependencies isolated to as few packages as possible.  In general these
dependencies are the command line interface, which is different even
among compilers from the same vendor on different platforms;  and the
interface to the system runtime and C application libraries.  There is
no good reason for a system such as AdaSage to require a different version
for every platform/vendor.

Douglas B. Rupp
drupp@cs.washington.edu

Re: Public Release of AdaSAGE (Re: Why is the DOE selling AdaSage?)

[agate!spool.mu.edu!howland.reston.ans.net!sol.ctr.columbia.edu!The-Star.h]

In article <1993Feb12.173406.8810@beaver.cs.washington.edu> drupp@cs.washington
.edu (Douglas Rupp) writes:
>In article <EACHUS.93Feb11194235@oddjob.mitre.org> eachus@oddjob.mitre.org (Ro
bert I. Eachus) writes:
>
>		....  stuff deleted ...
>
>>
>>    3) I have little experience with AdaSAGE and no recent experience,
>>but the ordering information lists versions for 80x86, RS/6000, Sun
>>SPARC, and AT&T 3B2, with PC versions coming in both Alsys and
>>Meridian versions.  Sounds portable (or at least significantly ported)
>>to me.
>>
>>   [deletia]
>>
>>					Robert I. Eachus
>>
>		.... more stuff deleted ...
>
>As I recall AdaSage is distributed in source code form, therefore
>it seems to me that an Ada program which needs to have a different version
>for each and every platform and compiler in existence just lends credence
>to the comments in Greg's last post about lack of portability.
>
>A properly designed and written program should have compiler and platform
>dependencies isolated to as few packages as possible.  In general these
>dependencies are the command line interface, which is different even
>among compilers from the same vendor on different platforms;  and the
>interface to the system runtime and C application libraries.  There is
>no good reason for a system such as AdaSage to require a different version
>for every platform/vendor.
>

My experience with AdaSAGE has shown that is is very non-portable.
This is not to say it can't run on different platforms, just that
the effort to port the system is non-trivial.  

Perhaps most importantly is that the schema display program, THOR,
which displays the information from AdaSAGE, requires the use of
a PC (Intel-based system).  In spite of the bragging about the
portability, there are limitations to just _how_ portable AdaSAGE
really is.

DISCLAIMER: My knowledge of AdaSAGE is about 1.5 years old.  If
my above comments are outdated, please let me know.  Also, the last
I heard, THOR was written in Modula-2.

----------I don't speak for the AJPO or CAE-Link : Aren't we lucky? ;-) -----
David Weller,         | Space Station Training Facility: Like the real
CAE-Link,             | thing, only you can step outside for a breath
Space Technology Div. | of fresh air!
----I'm the Ultimate International Masochist: I speak Ada AND Esperanto!-----

Re: Public Release of AdaSAGE (Re: Why is the DOE selling AdaSage?)

[agate!spool.mu.edu!howland.reston.ans.net!paladin.american.edu!darwin.sur]

In article <1993Feb12.173406.8810@beaver.cs.washington.edu>,
drupp@cs.washington.edu (Douglas Rupp) writes:
|> There is
|> no good reason for a system such as AdaSage to require a different
|> version
|> for every platform/vendor.
|> 

It is more accurate to think of AdaSage as an application-development
toolkit, rather than a set of libraries.  AdaSage includes Thor,
which supports both screen and database accesses (and maybe reports
-- but it's been awhile since I did work with AdaSage).

So, I can see why they have different versions, depending on the
platform.

I'm not saying AdaSage is a good or bad tool, nor am I passing judgement
on the portability of the applications developed using AdaSage.

-- 
Scott McCoy	Harris ISD	Opinions expressed are my own.
Staff Eng - SW	Internet: 	smccoy@dw3g.ess.harris.com

Re: Public Release of AdaSAGE (Re: Why is the DOE selling AdaSage?)

[Jon Spear]

In article <1993Feb12.210553.19700@mlb.semi.harris.com> smccoy@dw3g.ess.harris.
com (Scott McCoy) writes:
>In article <1993Feb12.173406.8810@beaver.cs.washington.edu>,
>drupp@cs.washington.edu (Douglas Rupp) writes:
>|> There is
>|> no good reason for a system such as AdaSage to require a different
>|> version
>|> for every platform/vendor.
>
>It is more accurate to think of AdaSage as an application-development
>toolkit, rather than a set of libraries.  AdaSage includes Thor,
>which supports both screen and database accesses (and maybe reports
>-- but it's been awhile since I did work with AdaSage).
>
>So, I can see why they have different versions, depending on the
>platform.

I've gotten a little more info on AdaSAGE.  Here's a summary:
[Any typos are mine.]

What is it?  (From AdaSAGE Notes, Jan 1, 1993)

"AdaSAGE is an effective environment to develop software applications
primarily for the Management Information System (MIS), scientific, and
engineering domains.  This environment allows the user to build
applications through an interactive screen editor, called THOR.  Here,
options are presented to the applications developer based on reusable
modules in the AdaSAGE program library.  After all options are
selected for an application, the AdaSAGE environment builds the Ada
source code which can then be compiled to create the application.
Simple applications can be developed with limited knowledge of Ada or
software engineering.  Larger applications may require additional
programmers skilled in Ada snd software engineering.  A recent
application developed using AdaSAGE achieved 75% code reuse: 50% of
the application was built directly from the AdaSAGE program library;
25% of the application was imported form other reuse sources; only 25%
of the application was unique and requried new code development."

On what platforms is AdaSAGE available?

"Intel 8088, 8086, 80286, 80386, 80486 (Real and Protected Mode), AT&T
3B2, Sun SPARC, RS/6000 ...  Operating systems: MS-DOS, Unix, and
AIX."

Is AdaSAGE written in Ada?

"...In the Unix environment, AdaSAGE is written 100% in Ada.  In the
MS-DOS ... 5% of AdaSAGE is written in assembly ... Several of the
AdaSAGE utilities may be written in other languages (e.g., THOR is
written in Modula-2) ... The AdaSAGE system is comprised of over
300,000 lines of code representing over 250 person-years..."


I can't quite tell what AdaSAGE does from this high-level description,
but it sounds like it includes some basic relational database and
graphics functions (on some platforms) with an application generator
to use them.  But... it isn't a "code generator" it says. 


According to one paper (Draft of "AdaSAGE Evolution", 21 Jan 93),
"Distribution of the AdaSAGE source code will primarily be through the
Free Software Foundation as GNU AdaSAGE.  As such, GNU AdaSAGE is
free..."  But as of today, I can't find it at prep.ai.mit.edu or any
other FTP site (per an Archie search).  Coming soon to FTP near you...?

Floppy copies of the code for Alsys or Meridian are available if you
send the floppies.  It may be FTP-able from Asset, but you have to
fill out and mail/fax their account request form before you can look.

A support hotline is available (for a fee).  There's an Ada and
AdaSAGE BBS (804-444-7841 / DSN564-7841, 2400-16.8Kbps; HST/V.32bis).
And there are training video tapes as well as instructors available.

I'm still looking for more info, and will summarize if there is interest.

-Jon
-- 
| Jon L. Spear (spear@cs.nps.navy.mil) : With computers we can make        |
| Dept of Comp Sci, Code CS/Sj, x2830  : billions of misteaks each second. |