> I think that the TIPSTER approach gives a flexibility not available > in SGML; however I don't think that is the whole of the issue that > could warrant discussion. SGML (or at least HTML its predecessor) > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ??! Umm, I take it you don't mean HyperText Markup Language, which could not have preceded SGML... or is my understanding way off here? The rest of the message re SGML as an encoding protocol makes sense to me, until: > You message seemed to be saying that TIPSTER is creating its own tag > set. Now that appears to be replicating work to me given the TEI tag > set is available. I can understand you wanting to make additions to > it but not why you might want to rewrite it entirely - or have I > misunderstood? > I think I haven't explained myself adequately -- which in part may be due to my unfamiliarity with TIPSTER, I'm afraid I'm more of an observer on a couple of projects here involving the technology. There may be a dnager here of mixing "levels", so maybe I'd better clarify one thing for the sake of it: TIPSTER is an alternative technology to *SGML*, as a means of "marking up" documents. The TIPSTER architecture itself (at least as far as my understanding goes) does not define any tags etc, in the same way that SGML does now. It's a formalism. *As far as I'm aware*, there are a number of "tags" which have been defined by the TIPSTER community, to meet their own needs (primarily those of the "defence" agencies (or "the spooks" as they're sometimes known! ;-) . These may be very different from the needs of the community that developed the TEI tagset! I don't believe there is any plan to replicate the effort of the TEI, there is simply no utility in that. One of the projects underway at Sheffield is aimed at creating a bridge between the SGML and TIPSTER formalisms -- I wondered if anyone else (in the TEI community) had either tried this, or had any interest in it. At an intuitive level I can't see any reason why the TEI *tagset* can't simply be used in a TIPSTER environment... it just wouldn't be in SGML. I hope this makes sense? Best wishes, mmalc.