Newsgroups: comp.sys.transputer
From: "Stephen Maudsley" <Stephen.Maudsley@btinternet.com>
Subject: Re: Post Mortem
Organization: esgem limited
Date: Tue, 17 Feb 1998 14:42:09 -0000
Mime-Version: 1.0
Content-Type: text/plain; harset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Message-ID: <6ccegb$mak$1@mendelevium.btinternet.com>


Jan Vorbrueggen wrote in message ...
>Richard Beton <richard.beton@roke.co.uk.no.spam.thankyou> writes:


<CUT>

>OTOH, careful use of some of occam's features, especially abbreviations,
>produces pretty good code, and the locality of variables in the
workspace due
>to nested scoping helps a lot, when compared to C, for instance. I've
never
>missed an optimiser that much, except for the numerically intensive
code,
>which I hand-optimsed in assembler...actually much easier than I
>expected. BTW, anybody know why this "mis-feature" was released, while
others
>"useful things" were not?


I think "released" is wrong - "escaped" is more accurate.

The assembler insert was originally a compiler hack to allow some of the
processor engineers to more accurately control instruction sequences as
part of the test regimes. Hence the name of the original construct GUY
after one of the engineers.

ISTR that it was essential for resetting channels (as you say elsewhere -
a non-occam requirement) for links that were operating in a noisy
environment.

---------------------
Stephen Maudsley      Stephen.Maudsley@btinternet.com
esgem limited
Tel/Fax: +44-1453-521626  Mobile: +44-370-810991
---------------------



