From:	Stuart Moore <stuart_moore AT my-dejanews DOT com>
Newsgroups:	comp.os.msdos.djgpp,rec.arts.int-fiction
Subject:	Re: The Frotz interpreter - ASM problems!
Date:	Sat, 06 Mar 1999 11:14:12 -0800
Organization:	University College Worcester
Lines:	29
Message-ID:	<36E17E84.20C@my-dejanews.com>
References:	<36DD8703 DOT 248D AT my-dejanews DOT com> <36DDA2C1 DOT C9C627FA AT alcyone DOT com> <36DF4EA4 DOT 72F2 AT my-dejanews DOT com> <36E035B9 DOT 396187EF AT alcyone DOT com>
NNTP-Posting-Host:	194.83.249.77
Mime-Version:	1.0
X-Trace:	nellg.worc.ac.uk 920718994 29222 194.83.249.77 (6 Mar 1999 11:16:34 GMT)
X-Complaints-To:	usenet AT nellg DOT worc DOT ac DOT uk
NNTP-Posting-Date:	6 Mar 1999 11:16:34 GMT
X-Mailer:	Mozilla 3.04 (Win16; I)
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp
Reply-To:	djgpp AT delorie DOT com

Erik Max Francis wrote:
> 
> Stuart Moore wrote:
> 
> > Excellent question. Interpreting the Z-Machine's not >that< demanding
> > on
> > PC power, is it?
> 
> One need only resort to assembly in _extreme_ cases, say, where 90% of
> the CPU power is being expended in a tiny section of very essential
> code, say for a graphics engine -- but even then it's often not
> _required_, it will just speed things up.

I can't say Z-Machine graphics are especially demanding, can you?

> I am puzzled to understand why someone would use assembly in a Z machine
> interpreter, unless they simply already had a chunk of assembly code to
> do something that they didn't know how to do in C.  Even then it seems
> like a bizarre reason to put the dependency on assembly in there.

Blame Mark Howell and Stefan Jokisch, not me. Believe me, it's bizzare.
Stupid, in fact.

By the way - I've remembered the ASM is for byte-swapping now. Any help?

Bye,

--
Stuart Moore.

- Raw text -