Newsgroups: rec.arts.int-fiction
Path: news.duke.edu!newsgate.duke.edu!nntp-out.monmouth.com!newspeer.monmouth.com!news.maxwell.syr.edu!newsfeed.frii.net!easynews!uunet!dfw.uu.net!ash.uu.net!world!buzzard
From: buzzard@world.std.com (Sean T Barrett)
Subject: Re: Virtual Machine question...
Message-ID: <GAvMwo.Lv3@world.std.com>
Date: Tue, 27 Mar 2001 22:10:00 GMT
References: <99oeaa$1ajc@f1n1.spenet.wfu.edu> <CX4w6.281$fy3.117634@e3500-chi1.usenetserver.com> <99qmcp$1as$1@news.panix.com> <Nj6w6.863$ea6.149235@news1.frmt1.sfba.home.com>
Organization: The World Public Access UNIX, Brookline, MA
Lines: 23
Xref: news.duke.edu rec.arts.int-fiction:84651

Kevin Forchione <Kevin@lysseus.com> wrote:
>Perhaps it would be better to compile z-code to TADS-2 format.

I thought the problem people were considering was that of running
Tads-2 programs on machines that lack Tads-2 interpreters?  Z-code
is already widely portable.

Either way, though, I think this sort of cross-porting would
be more useful if the target format were designed for portability.
Tads2 and Tads3 do not have clearly-defined VMs which support a
variety of implementations and a variety of source languages.  For
an example of a high-level (as opposed to Zmachine or Glux) VM which does,
see my proposal for Slux at http://world.std.com/~buzzard/slux/slux.html
which I need to revise at some point to remove the 'set'
datatype.

Unfortunately, at this point in time there are exactly 0
compilers or interpreters that support Slux, but I think
it's much easier to implement an interpreter for and should
support retargetting from many languages, e.g. because it
supports definable inheritence rules.

Sean
