Re: More DoD use of everything but Ada

Re: More DoD use of everything but Ada

[David Tannen x8273]

Greg,

I am amazed and now better informed.  Kudos for providing useful and
interesting information in a non-flame-form.  This is the kind of data
you needed to give us so we can understand what it is you are trying
to accomplish.

David Tannen
tannen@tigger.geg.mot.com
tannend@source.asset.com

Re: More DoD use of everything but Ada

[John Cobarruvias]

In article <SRCTRAN.93Jun16161348@world.std.com> srctran@world.std.com
(Gregory Aharonian) writes:
>

I occasionaly read this group. Although at times I agree with Gregs posts,
I still wonder "Whats the point?"

So Greg, Whats the point? OK :
Ada is not being used as widely as Ada proponents believe it is,
Ada has problems
Ada may be dropped by the Air Force
Ada mandate is not right
Ada sucks

OK. now what? If you are trying to convince people that Ada is the root of
all computer evil, try writing your congressman, or someone who can help
you with your quest. But Usenet, especially comp.lang.ada isn't the place.
Maybe we need a new group like comp.lang.ada.complain.

Some of your ravings have merit, but so the hell what? Its beginning to
sound more like sour grapes than anything else.

>   In honor of Mike Feldman's periodic posts of Ada use outside the
>Mandated World, I thought I would start a competing list of non-Ada use
>inside the Mandated world.  If nothing else, this will be a simple
>example of language competition, a concept that escapes the Mandated
world,
>and has shunted Ada into a role as a niche language.  This list in time
>will also beg the question of what Ada is the more cost-effective for.
>
>AIR FORCE DRAIR ADVISER
>   The Air Force at the Oklahoma City Air Logistics Center at Tinker
>Air Force Base has developed the Deficiency Report Analysis Information
>Report (DRAIR), which tracks problem resolution for aircraft parts in the
>field.  It aims to standardize the DRAIR, reduce the time it takes to
>produce a DRAIR, and capture expertise from people knowledgable about
>specific aircraft.
>    DRAIR consists of about 7000 lines of C, 1600 lines of Unix shell
script,
>520 lines of SQL, 143 lines of Unify's RPT Report Writer Language and 603
>CLIPS rules.  CLIPS is a NASA expert shell with many benefits.  Use of AI
>was important to DRAIR.
>
>ARMY TURBINE ENGINE DIAGNOSTICS
>    The Army Research Lab and Ordance Center and School have developed
and
>are fielding (so much for soldiers lives) a lightweight, visual expert
system
>that diagnoses and suggests fixes for M1 tank engines.  TED uses the
expert
>system tool Adept from Symbologic, running under Windows and also using
>Excel and Asymmetrix's Toolbook to provide a graphical object-oriented
system
>that helps mechanics select repair parts.
>
>DARPA FUNDS TEMPORAL REASONING MODELS
>    The July 1993 issue of SIGART (like SIGADA, but for A.I.) Bulletin is
>devoted to temporal reasoning systems, a difficult problem in the field
of
>artificial intelligence.  And like all other advanced research in
computer
>science, none of this work is being done in Ada.  And what work is done,
>much is funded by those zealous defenders of the Ada Mandate, DARPA.
>Here, two DARPA funded efforts are mentioned. One is a system, Tachyon,
>being developed with funding from DARPA through the Air Force's Rome
>Laboratory.  The second is a system written in KRSL for testing temporal
>reasoning systems, funded by the Air Force's Rome Laboratory.
>
>ARMY INTEROPERABILITY NETWORK
>    The June 1993 issue of SIGNAL has an article on the Army CECOM's
>interoperability network, which uses nationwide digital communication
>network and off the shelf equipment to promote inexpensive remote
software
>testing.  It links up support centers at Monmouth, Redstone, Leavenworth,
>and Huachuca, as well as some contractor facilities.
>    The article does not mention which language the system is written in
>(again, why are these people so embarassed to be associated with Ada;
>it desparately needs the publicity), but given the extensive use of
>commercial communications equipment and off-the-shelf products, I would
have
>to assume that the system is Unix based and therefore has lots of C code.
>
>
>DARPA FUNDING OF OOP RESEARCH
>    In examining seven years of OOPSLA conference proceedings
(1986-1992),
>many of the systems developed were funded by government agencies.
>Collectively, DARPA/ONR/CECOM/AFOSR/ESD/ARMY/AIR FORCE funded 24 of 
>these projects - 22 of which were in Lisp/C++/Smalltalk, and 2 in Ada.
>[What is the OOP language of preference for the DoD?   Not Ada.]
>
>==========================================================================
====
>
>Please send any submissions for this list to me by email.  Help me make
>my list longer than Mike's.
>
>-- 
>**************************************************************************

> Greg Aharonian
> Source Translation & Optimization
> P.O. Box 404, Belmont, MA 02178
>




*************************************************************
John R. Cobarruvias, Texas A&M Class of '78, 
NASA Johnson Space Center Houston Tx.
(713)483-9357

"Your pain will be legendary" (Hellraiser I)
"And to think..................I hesitated" (Hellraiser II)
"These pins are killing me!" (Pinhead in Hellraiser IV)
*************************************************************

Re: More DoD use of everything but Ada

[cis.ohio-state.edu!pacific.mps.ohio-state.edu!math.ohio-state.edu!cs.utex]

In article <1993Jun17.152105.959@aio.jsc.nasa.gov>, Cobarruvias@asd2.jsc.nasa.g
ov (John Cobarruvias) scrawls...
>>AIR FORCE DRAIR ADVISER
>>   The Air Force at the Oklahoma City Air Logistics Center at Tinker
>>Air Force Base has developed the Deficiency Report Analysis Information
>>Report (DRAIR), which tracks problem resolution for aircraft parts in the
>>field.  It aims to standardize the DRAIR, reduce the time it takes to
>>produce a DRAIR, and capture expertise from people knowledgable about
>>specific aircraft.
>>    DRAIR consists of about 7000 lines of C, 1600 lines of Unix shell
>script,
>>520 lines of SQL, 143 lines of Unify's RPT Report Writer Language and 603
>>CLIPS rules.  CLIPS is a NASA expert shell with many benefits.  Use of AI
>>was important to DRAIR.
>>

There is a good point in here somewhere.  Do you remember the book
MegaTrends?  It talked about a trend from either-or choices to 
multiple choices.  As technology evolves and more specialized applications
appear, there will be more choices about how to deal with them.  

The idea that we can accept one language to support a growing number of 
different number of applications is naive.  

Ada has it's place.  Perhaps Ada 9x will expand Ada's role--it provides
multiple choices through annexes.

Why would you program an expert system in Ada?  Of course the tool use do use
could itself be written in Ada--CLIPS has been.

Re: More DoD use of everything but Ada

[Mark Bayern]

Greg,

About your list --
 
1.  The following probably doesn't belong on the list -- comp sci
research is not an engineering field.  I have always felt Ada is
intended to improve software engineering.  Its kinda' like the
difference between mechanical engineers and physicists.  The
engineers need to build stuff that works, the physicists are
looking to discover new things and new principles. 

>
>DARPA FUNDS TEMPORAL REASONING MODELS
>    The July 1993 issue of SIGART (like SIGADA, but for A.I.) Bulletin is
>devoted to temporal reasoning systems, a difficult problem in the field of
>artificial intelligence.  And like all other advanced research in computer
>science, none of this work is being done in Ada.  And what work is done,
>much is funded by those zealous defenders of the Ada Mandate, DARPA.
>Here, two DARPA funded efforts are mentioned. One is a system, Tachyon,
>being developed with funding from DARPA through the Air Force's Rome
>Laboratory.  The second is a system written in KRSL for testing temporal
>reasoning systems, funded by the Air Force's Rome Laboratory.
>


2.  Your list shouldn't be in competition with Mike's.  His is a
list of Ada use outside the mandated world.  It shows the
(unfortunately) slow acceptance Ada is getting out here.  However,
it does show that Ada is getting some use outside the mandated
world.  Your list shows that the mandate is not taken seriously by
those who should do so.  They should either learn how to work with
the mandate or get rid of it.  (I'd vote for the 2nd option.) 

>
>Please send any submissions for this list to me by email.  Help me make
>my list longer than Mike's.



Mark