Newsgroups: rec.arts.int-fiction
Path: nntp.gmd.de!news.ruhr-uni-bochum.de!news.rhrz.uni-bonn.de!RRZ.Uni-Koeln.DE!news.gtn.com!news.frankfurt.ecrc.net!news.ecrc.de!02-newsfeed.univie.ac.at!newsfeed.uk.ibm.net!news.ibm.net.il!arclight.uoregon.edu!nntp.primenet.com!netcom.com!erkyrath
From: erkyrath@netcom.com (Andrew Plotkin)
Subject: Re: Z-machine-  question about future.
Message-ID: <erkyrathDy5z4D.I0L@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
X-Newsreader: TIN [version 1.2 PL1]
References: <Pine.LNX.3.95.960922173222.1309A-100000@adamant.res.wpi.edu>
Date: Mon, 23 Sep 1996 02:33:49 GMT
Lines: 48
Sender: erkyrath@netcom13.netcom.com

George Caswell (timbuktu@adamant.res.wpi.edu) wrote:
>    Hi.  I've only barely begun learning Z-machine assembler...  I have a
> question about the Z-machine.  How far could it be extended without making
> the necessary modifications to Zcode tools to handle the changes
> impractical?  (I think that's what I want to ask...  I'm having some
> difficulty getting my meaning across right, I think...)  Where's the upper
> bound, and would it be worthwhile to make a Z-machine that breaks those
> bounds?  (Which, I think, would mean discarding parts of the current
> z-machine structure, re-working the thing into something that might or
> might not be a Z-machine...)

I've always thought that it should either be left alone, or completely 
redesigned. And I've already posted my position on that subject. :)

It is worthy of note that to the best of my knowledge, the Z-machine 
interpreters have only changed *twice* in the four-ish years I've been 
considering the subject. Once when Graham defined the V8 format; once, 
much less significantly, when a bug was found in the interpretation of 
the timed-input opcode. (No post-Infocom game has used timed input except 
my Tetris joke.)

This stability is good.

Also, we're not really pushing the limits of V8. _So Far_ used 31 
properties out of 62, and 33 attributes out of 48.

>    BTW, what's the best document currently out there for learning to
> program directly for the Z-machine?  I have the Z-machine spec, but it
> also spends a lot of time debating future changes to itself and defining
> what must and must not be done by interpreters and other such.

That's really about it. It's the first five chapters which have all the dirt.
It's not like there's a lot you need to know.

>   Mostly
> wondering if there was something more concise, for understanding the
> structure and programming of Z-code programs, the program module format
> introduced by Inform, (and calling conventions, I suppose) etc.  Thanks.

A lot of that is going to be Inform convention, not Z-machine hardwiring. 
Try the new Inform tech manual.

--Z

-- 

"And Aholibamah bare Jeush, and Jaalam, and Korah: these were the
borogoves..."
