Re: Standadised OO Language

[Brian Rogoff <bpr@shell5.ba.best.com>]

On Fri, 13 Feb 1998, Howard W. LUDWIG wrote:
> Perhaps some specific dates and details would help:
> 
> CLOS (Common Lisp Object System)
>   1994-12-08  ANSI X3.226:1994
>   ----------  has not become an ISO standard

To be precise (this is the Ada newsgroup after all :-), one should say 
"Common Lisp", not CLOS, as CLOS is just an object system sitting on top 
of the objectless ANSI CL subset. 

> ISLISP (which I know very little about and haven't kept up with any
> details, 
>         other than it is supposed to be neither a subset nor a superset
> of CLOS!!!)
>   1997-??-??  ISO 13816:1997
>   ----------  has not become an ANSI standard

Never finished, supposedly based on EuLisp, a nice clean Lisp which never
went anywhere. Its a sort of reaction to the fact that Common Lisp was 
originally an all-American, or rather all-USA grass roots effort to
synthesize a Lisp out of the extant US Lisps. EuLisp was a brand new
design somewhere between Scheme and Common Lisp in the design space.
Some of the story is in the HOPL-II proceedings.

> [Thus, Lisp has the dubious distinction of being one of the few
> languages for 
> which both an ANSI standard and an ISO standard exist but they never
> have matched 
> nor does there seem to be any attempt to harmonize them.]

Last time I looked ISO Lisp wasn't finished, so I'm not sure if this is 
true. I'll look at the ISO web page when I have a web connection. 
> 
> Ada 95 has been a published ISO standard for 3 years now (well, after
> two 
> more days, actually :), and this version is a revision of a previous ISO 
> standard.  Mature compilers from several vendors for several
> [understatement] 
> platforms are available in a _validated_ form.  The validation process
> for 
> Ada 95 compilers has been growing in rigor, following a transition time
> of 
> relative laxness from Ada 83 to Ada 95.  GNAT, which runs on a wide
> variety 
> of platforms, implements the full (including all optional annexes)
> language 
> and passes _all_ relevant validation tests.
> There is no mechanism in place to assure compliance  of compilers to the
> standard (like Ada validation).  

This is a big plus for Ada, much more important IMO than whether an ANSI
or ISO standard exists.

> Since it is expected 
> that a revision to the C standard will be approved in 1999 or 2000 and 
> C++ has been intricately coupled in structure with C [which is a whole 
> other issue and a can of worms in itself], the C++ folks are already 
> talking about mechanisms for C++ compiler vendors to modify their C++ 
> compilers to handle the revisions to the C standard without adjusting 
> the C++ standard!?#!

Excellent point, which I hadn't really thought of. Of course, it may also
entail some slight modification to Annex B of Ada 95. 

As Robert Dewar pointed out, none of this is likely to sway anyone who has
already decided that they "hate" language X, where X is Ada, C++, Lisp, or 
whatever. I know I prefer Ada over C++, but C++ has some nice ideas too.

-- Brian