[Info-vax] Opportunity for VSI?

[Bill Gunshannon]

On 12/20/18 9:02 AM, Scott Dorsey wrote:
> Bill Gunshannon  <bill.gunshannon at gmail.com> wrote:
>> On 12/19/18 7:40 PM, Scott Dorsey wrote:
>>>
>>> It was the PL/1 of the 1980s.  So big that everyone programmed in a
>>> subset and so nobody could read anyone else's code...
>>
>> First, there are no Ada Subsets.  The owners of the language
>> specifically prohibit it.
> 
> The subset is in the programmers' heads.  The compiler was verified and
> implemented the full standard, but the programmers did not.
> 
>> Second, it is a typical item designed by committee.  BY the
>> time it was done  the people responsible for its creation in
>> the first place (The US Air Force) refused to use it even
>> with the DOD mandate.
> 
> It actually got traction there for stuff like realtime applications,
> nearly replacing languages like JOVIAL.  

Not with the USAF.  After the failure of creating Ada
they went back to Jovial over the objections of the DOD.

>                                             The problem is that the
> people writing realtime applications didn't understand the fancy string
> manipulation and text formatting functions.... they didn't need to, until
> all of a sudden they encountered someone's code that used them.

No, the problem was too much fluff and not enough substance.
The USAF wanted a lean, functional language for doing real-
time.  They got a boat anchor that thought it could do anything.

I once asked an Ada expert at a major government contracting
company I use to work for about the possibility of re-writing
Unix in Ada to eliminate all the well known security problems
(like buffer over-runs).  He told me it was doable but you
wouldn't like the result because the overhead would make it
a real dog.

bill