- From: Mark Baker <distobj@acm.org>
- Date: Mon, 29 Jul 2002 21:35:19 -0400
- To: Joshua Allen <joshuaa@microsoft.com>
- Cc: www-tag@w3.org
On Mon, Jul 29, 2002 at 02:01:52PM -0700, Joshua Allen wrote: > I would recommend not mentioning HTTP at all. I think HTTP plays an important role in aspects of what's being discussed here. Namely, that its built-in accomodation for dealing with types of variation in resource representations, provides resource publishers with some tools to help maintain uniformity for their URIs. Specifically, the Vary header, and the use of content negotiation (and perhaps other features that aren't occuring to me at the moment). Using TimB's Sect 1.1 example, the URI http://weather.yahoo.com/forecast/MXOA0069.html could have been an FTP URI (FTP is also well deployed), but FTP doesn't handle any type of variability very well. So if somebody wanted Oaxaca's weather report in PDF, the resource publisher would have to provide a separate FTP URI for that information, whereas a single HTTP URI would suffice. IMO, it's these types of features that make HTTP URIs able to identify anything, not just in theory - because any scheme can do that - but in practice. The relevant TimBL design issue is; http://www.w3.org/DesignIssues/Generic.html MB -- Mark Baker, CTO, Idokorro Mobile (formerly Planetfred) Ottawa, Ontario, CANADA. distobj@acm.org http://www.markbaker.ca http://www.idokorro.com
Received on Monday, 29 July 2002 21:22:50 UTC