- From: Jon Bosak <bosak@atlantic-83.Eng.Sun.COM>
- Date: Thu, 21 Nov 1996 14:50:30 -0800
- To: preece@predator.urbana.mcd.mot.com
- CC: papresco@calum.csclub.uwaterloo.ca, msftrncs@htcnet.com, www-style@w3.org
[Scott Preece:] | It's still the case that standard DTDs are required if we are to get | the best use out of indexing/search systems, since the best searching | requires some awareness of the semantics of the data, which means the | indexer or the search engine must know something about the "meaning" | of the tags it finds. But it's not the (machine-readable portion of) the DTD that tells the indexer what that meaning is. In other words, if the indexer wants to give more weight to stuff it finds in title elements (headers), there's nothing in the DTD that tells the indexer that elements tagged TITLE are, in fact, titles. That information has to stand outside the machine-readable portion of the DTD somewhere. So the issue of semantics for indexers is orthogonal to the question of DTDs. This is one of the reasons that XML sanctions DTD-less Web documents. Jon ---------------------------------------------------------------------- Jon Bosak, Online Information Technology Architect, Sun Microsystems ---------------------------------------------------------------------- 2550 Garcia Ave., MPK17-101, | Best is he that inuents, Mountain View, California 94043 | the next he that followes Davenport Group::SGML Open::ANSI X3V1 | forth and eekes out a good ::ISO/IEC JTC1/SC18/WG8::W3C SGML ERB | inuention. ----------------------------------------------------------------------
Received on Thursday, 21 November 1996 17:54:21 UTC