Newsgroups: rec.arts.int-fiction
Path: gmd.de!urmel.informatik.rwth-aachen.de!uni-paderborn.de!zib-berlin.de!xlink.net!howland.reston.ans.net!spool.mu.edu!sgiblab!wetware!kaiwan.com!cinnamon
From: cinnamon@kaiwan.com (R. Dominick)
Subject: Re: Inform ported to 386 machines
Message-ID: <CHIo9M.Myq@kaiwan.com>
Organization: Coffeehouse Software
X-Newsreader: TIN [version 1.2 PL2]
References: <1993Dec2.181626.18576@vax.oxford.ac.uk>
Date: Sat, 4 Dec 1993 15:27:21 GMT
Lines: 39

Mathematical Institute, (0865) 2-73525 (nelson@vax.oxford.ac.uk) wrote:

> Volker has now tested Stephen's port and pronounced it healthy: it is
> in the archive in FTP.GMD.DE at...

Well, it's at least as healthy as my port of Inform to the '386,
'cause it seems to have the same bug, with an interesting twist:

My copy of Inform seems to generate different compressed text, but
runs perfectly under four different Infocom interpreters; the files
just differ in a binary way from the ones supplied by Graham.
Stephen's port produces files that differ from *mine*.  Neither mine
nor Stephen's compiled .z3 files will correctly complete a 'verify'
command, yet Graham's .z3 files will.  (This was tested under several
Unix and DOS Infocom interpreters.)

And there's also the matter of most inferences (even those as simple
as

>take
What do you want to take?) 

crashing the Z-machine before the 'What do...' message actually
appears, a quite serious bug in the supplied parser, IMHO.

PS:  If you have the djgpp program 'exe2aout', you can do the
following:

	exe2aout inform
	copy /b go32.exe+inform inform.exe

and you won't have to keep the copy of go32 in a separate file along
the path and all that.

--r


-- 
don't be alarmed, it's only a kiss
