Straw proposals -- get 'em while they're hot!

We seem to be finding quite enough to talk about just in replying to
each others' posts, so I hesitate to urge too much new material on
the work group.

But anyone seeking a tighter focus for discussion may wish to look at
an actual proposal for what XML might look like.  Straw proposals, as
Jon Bosak has suggested, are the best way of keeping an extended
email discussion on track.

And we do have straw proposals to discuss.  They have the virtue,
moreover, of being independent of our discussions, in the sense that
they all predate, in their essentials, the formation of this group,
and are fully worked out proposals -- or as fully worked out as the
inclinations of their authors have led them to be.

ISO 8879 itself defines two SGML subsets (Basic SGML and Minimal
SGML).  The others included in the summaries in the voting booth are:

  - SGML Lite, by Bert Bos
  - MGML (Minimal Generalized Markup Language) by Tim Bray
  - A Lexical Analyser for HTML and Basic SGML, by Dan Connolly
  - SGML Online, by Eliot Kimber
  - Normalised SGML, by Henry Thompson et al., (the standard format
    used by their NSL tools library)
  - PSGML (Poor Folks' SGML), developed by myself with the consultation
    of colleagues here at UIC
  - the TEI Interchange Format, developed by the metalanguage committee
    of the Text Encoding Initiative

All of these proposals are available on the World Wide Web; pointers
to the documentation are in the root document of the voting booth, i.e.


Having had the pleasure of studying all of these with some care, to
put together the summaries, I can assure you that they are all worth
reading.  They all agree (pretty much) on some things, and their
variation on other questions illustrates very well the way in which
the goals and assumptions of a subset can affect, and be affected by,
the technical decisions made.

-C. M. Sperberg-McQueen