🚀 go-pugleaf

A lot people responded to the original question about VDM vs Z
stated that one of their reasons for choosing either VDM or Z was
because of the language's ability to structure large specifications.
However nobody bother backing up their statements, i.e. nobody
said why say VDM was better for structuring.

I would really like to hear more arguments about why one is better
than the other in terms of structuring.  In particular I want to
know what mechanisms are available for building specifications 
incrmentally, (e.g. schema calculus). 

The next question is do you think object-orientation could 
assist the building of incremental specifications ?

--
Anthony Lee (Michaelangelo teenage mutant ninja turtle) (Time Lord Doctor) 
email: anthony@cs.uq.oz.au			    TEL:+(61)-7-365-2697 (w)
SNAIL: Dept Comp. Science, University of Qld, St Lucia, Qld 4072, Australia

In article <1991Jun18.152454.27474@fmg.bt.co.uk> elc@fmg.bt.co.uk (Elspeth Cusack) writes:

>Mark S Madsen <msm@computing.lancaster.ac.uk>
>, quoted by collin@hpindda.cup.hp.com (Collin
> Park @ HP Information Networks) said:

(I apologise in advance for any inconsistencies that may arise through
my not having ever seen Collin's quotation of my email to him.)

>These both sound like malicious rumours :-)  Standardising on VDM???
>Someone is joking with you.  I can tell you that BT (formerly British
>Telecom) has standardised on LOTOS and Z.

Sounds like my sort of self-assured tone, so it was probably I that
wrote it :-)

>This statement about BT interest in formal methods is a little
>misleading. It is true that there is research activity in both
 ^^^^^^^^^^
Oh, definitely it must indeed have been I writing, if the content is
misleading!

>LOTOS and Z,and that these are our `favoured candidates' (as
>researchers) for specifying distributed systems. But that is a
>long way from saying that the company has standardised on LOTOS
>and Z! BT also has an active community of SDL users in the
>protocol testing field, and considerable involvement with 
>software methodologies and development conventions.

My apologies to everyone at BT for misrepresenting them to the world.
(I am not apologising for misleading the net because the above quote
was sent in private email, and quoted without permission from me.)
I was clearly under a mistaken impression that LOTOS and Z were in
possession of official approval.

However (I love having the opportunity to begin a paragraph with that
word :-), let me emphasise that at least one point in my original
posting was correct, and is implied in Elspeth's posting:  VDM is by
no means the only formal technique in use in the UK with either
business or academic concerns.

>Elspeth Cusack

Hope you're all keeping well down there - say hi to Steve from me.

			Mark

-- 
______________________________________________________________________
      Mark S Madsen   +44-524-65201x3819    msm@comp.lancs.ac.uk
      Department of Computing, Lancaster University, LA1 4YR, UK
______________________________________________________________________

In article <2017@uqcspe.cs.uq.oz.au>, anthony@cs.uq.oz.au (Anthony Lee) writes:
> A lot people responded to the original question about VDM vs Z
> stated that one of their reasons for choosing either VDM or Z was
> because of the language's ability to structure large specifications.
> However nobody bother backing up their statements, i.e. nobody
> said why say VDM was better for structuring.

Another language ought to be mentioned here and that is Extended ML,
although it is not a "pure" specification language but a wide-spectrum 
language based on Standard ML.  Briefly, the module facilities of Standard 
ML are used to structure the specifications.  These can then be refined
into Standard ML code.  

> 
> The next question is do you think object-orientation could 
> assist the building of incremental specifications ?
> 

Myself and a colleague are looking into it for Object Z.

________________________________________________________________________________

Edmund Kazmierczak
email : eka@lfcs.ed.ac.uk
LFCS, Dept. of Computer Science
The King's Buildings
Universilty of Edinburgh
United Kingdom, EH9 3JZ