On Sat, Aug 23, 2014 at 11:27:30AM -0400, Anne & Lynn Wheeler wrote:
> [email protected] (Tomasz Rola) writes:
> 
> > So, now the 5-6 years old anecdote about one contractor stating
> > that "Ada is obsolete" makes much more sense, even though at the
> > time I read it, it sounded rude and immoral. It doesn't really
> > matter anymore what language will be choosen for a project - well,
> > it may still be a problem if one prefers to write Lisp (MHO:
> > concise, elegant) over writing Java (MHO: overly talkative and
> > relying too much on external tools and cargo cult procedures like
> > refactoring - I guess almost nobody writes Java in Emacs
> > nowadays). But the source code is going to be verified in theorem
> > prover, automatically. And even the compiler does not need to be
> > trusted anymore, because one can compare exec file with source and
> > prove that one matches another.
> 
> ADA tends to still be used for "human rated" applications ... aka
> human lives at risk ... like commercial airplane control systems.
> http://en.wikipedia.org/wiki/Ada_%28programming_language%29

The original intention to post my previous message was that whatever
in the past could be achieved by self-disciplined mental effort and
learning "the right language for a job", nowadays can or will be
achieved by combining whatever fashionable language-du-jour with
automatic proof of correctness for the very same uses for which Ada
is/was being used.

My intention wasn't to say that I dislike Ada, because frankly I never
programmed in it. My intention was to say that IMHO certain thing is
going to happen to Ada, no matter who likes it or would like to learn
it.

[...]

> I've frequently pontificated that the original mainframe tcp/ip product
> was done in vs/pascal and had *NONE* of the pointer-related exploits
> common in c-language implementations. some past posts
> http://www.garlic.com/~lynn/subintegrity.html#buffer
> 
> part of the problem is c-language pointer values can be ambiquous which
> is not easily identifiable by source code analysis (there is currently
> thread in comp.arch about difficulties with language features that are
> abiquous or not stictly defined)

I like C, it allows me to do certain things with computer, some fancy
optimisations which I couldn't easily do in, say, Lisp. But C is about
the last language I would consider for writing software to fly
drones. Yet this is what DoD says is going to be used. C/C++ plus
Isabelle (theorem prover) means Ada is going out.

Ada is not going to dissapear overnight. And there will be plenty of
code old and some new in Ada. But right now I cannot see how the role
of Ada in a future could be bigger, actually what I see is that
certain languages I seriously dislike are going to be used more and
for more demanding tasks - if you can use C for things involving human
life (like they say they will), then you can use anything else,
including a fashionable language.

I am sceptical to this trend, like I have written already, but
apparently it is happenning right in the front of us, if one looks
carefully. Also, the idea that some guy who is unable to learn other
language will be behind software flying heavy stuff over my head (more
than a tonne? even a kilo can be serious if going off course or
dropping from high above) and perhaps the "correctness" of this
software would be achieved by trial-and-error, running consecutive
proofs and changing lines until it all checks ok, yes, this sounds
very unsettling.

-- 
Regards,
Tomasz Rola

--
** A C programmer asked whether computer had Buddha's nature.      **
** As the answer, master did "rm -rif" on the programmer's home    **
** directory. And then the C programmer became enlightened...      **
**                                                                 **
** Tomasz Rola          mailto:[email protected]             **

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to