Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!zaphod.mps.ohio-state.edu!sunybcs!bowen
From: bowen@cs.Buffalo.EDU (Devon Bowen)
Newsgroups: comp.lang.modula2
Subject: Re: Re: Re: VDM & M2
Message-ID: <17936@eerie.acsu.Buffalo.EDU>
Date: 17 Feb 90 17:22:04 GMT
References: <9002170200.AA01479@ctc.contel.com>
Sender: nobody@acsu.Buffalo.EDU
Reply-To: bowen@cs.Buffalo.EDU (Devon Bowen)
Organization: State University of New York at Buffalo/Comp Sci
Lines: 24

In article <9002170200.AA01479@ctc.contel.com>, reid@CTC.CONTEL.COM (Tom
Reid x4505) writes:
> When VDM DOES stabilize and THERE IS a standard reference, THEN it is
> time to use it to formally define the language.  Unless convinced otherwise,
> I am leaning toward voting against the standard with the VDM left in unless
> a "reasonable" description of EXACTLY the variant of VDM used is included
> in the standard.

I agree entirely. When the draft arrived here we immediately set out to
try implementing it and we have had nothing but trouble since. Using a
non-standard language to describe a standard one is the type of thing
that anecdotes are made of.

What I'm wondering is how difficult is it to ammend a standard? The VDM
it currently includes is helpful (once you finally get your hands on one
of the references to learn it) but should not be considered the final
step. Is there a way the standard can be approved with the intention of
revising the VDM later? The difference here is that the implementor can
still get something working now but is aware the it will be changing in
the future once the semantic technology is in place.

On another note, any opinions on the FORWARD statement?

Devon