Re: IBM still undermining commercial Ada

Re: IBM still undermining commercial Ada

[munck]

In INFO-ADA Digest V93 #63, Gregory Aharonian says

>   For a long time, I have stressed that Ada is dead in the
>commercial world (non-mandated world) as long as corporate IBM
>refuses to acknowledge and support Ada, ...

It seems fairly certain to me that Ada will be around long after IBM
is gone.  They just made a "major" product announcement of new huge,
raised-floor, special-room _mainframes_.  Jeez!


>   An article in the January 25, 1993, Computerworld, ... Titled
> "Has IBM Scored", 
>	Late last summer, IBM downsized Repository Manager and AD/CYCLE,
>	its mainframe CASE strategy. ... instead, IBM is
>	hard working on a LAN strategy called AD Platform. ...
>
>	OPERATING SYSTEM INDEPENDENCE - Initially, AD Platform will
>	run on AIX and OS/2;
>	TOOL SERVICES - The repository consists of C++ code layered on
>	top of Versant.
>
>    As this shows, Ada is still not getting any recognition at IBM.

I don't see that this shows any such thing.  CASE technology is
EXTREMELY language-independent; the only parts of a full SEE that have
to know what application language is being used are a small handful of
the tools (compiler, source editor, linker, debugger).  Tools are
almost always third-party.

Also, it doesn't matter at all that the platform (framework) is
implemented in C++.  A minor question is whether there is an Ada
binding provided, but that is only of interest to tool builders who
want to write their tools in Ada.  The DoD is actively indifferent to
what language the framework and tools are written in.


> ...  Sure some of the tools for third parties
>will have Ada capabilities, but all of IBM's might will be put behind
>their OO languages of the future, C++, Smalltalk and (ugh) Object
>Cobol.

Isn't it called "Add 1 to COBOL;" ?  Anyway, who cares what language
IBM picks?  I don't see PL/1 (or PL/O-O) in the list.  Their choice of
languages has never had much influence on the world.  Now, DoD's
choice of COBOL did have quite a bit of influence on commercial
programming...

Also, consider their choice of OS:  AIX and OS/2.  


> It seems then that this country can fund a better company to lead
the STARS program, given this continual internal corporate hyporcrisy,
and given IBM mismanagement of such STARS activities as ASSET ( C.L. -
have you ever heard about the concept of COMPONENT COST???? Wake up)

Quite a bit of misinformation there, starting with the idea that IBM
"leads" STARS.  That will be a surprise to DARPA.  It sounds like you
get your information from the trade rags, most of which are completely
unable to get a handle on CASE activities.

Bob Munck, Deputy Architect STARS
           PCIS Expert