Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!mcsun!ukc!icdoc!gould!iwm From: iwm@doc.ic.ac.uk (Ian Moor) Newsgroups: comp.text.sgml Subject: Re: What would it take to implement SGML in TeX? Message-ID: Date: 9 Oct 90 17:18:20 GMT References: <1990Sep24.174222.22487@terminator.cc.umich.edu> <8935@jarthur.Claremont.EDU> Sender: iwm@doc.ic.ac.uk Followup-To: comp.text.sgml Organization: /usr/staff/courman/iwm/.organization Lines: 23 In-reply-to: dhosek@frigga.claremont.edu's message of 8 Oct 90 18:31:03 GMT I used to use GML (which I presume is (was) related to SGML), from Waterloo, this was implemented in Waterloo Script (somewhat like runoff). They added one feature to Script to allow commands to be anywhere in the text, not just line beginnings. It should be possible to use TeX, which is rather more versatile. In my opinion, the real problem is error reporting. Assuming that the implementation is done by use of macros (\def), SMGL errors will often result in reports of errors by TeX in terms of the generated code. This can be seen to some degree in LaTeX, but I suspect that SMGL, being more configurable might produce more confusing reports. Pick a versatile lexer/parser generator or perhaps a Macroprocessor like ML/1 and do a translation to TeX. -- Ian W Moor ARPA: iwm@doc.ic.ac.uk JANET: iwm@uk.ac.ic.doc Department of Computing, (The skin is mightier than the banana) Imperial College. 180 Queensgate London SW7 UK.