At 11:57 PM -0500 3/26/97, lee@sq.com wrote: >David Durand said: >> At any rate, I think we will need to keep going here, at least to get >> to the point that we can put an URN in XML and have it be legal. > >Early on in the URN development, the idea was that urn: would simply be >added as a URL scheme. In other words, wherever a URL was allowed, >you could put a URN. The notion that a URN was simply a kind of URL took a long time to kill, but it died because it's fundamentally not true. the URN prefix can be treated as a protocol by a parser, but it is semantically distinct and does not receive the same kind of processing. >If you adopt this attitude, the only change needed in the XML spec >would be to remind people of this fact, rhaps by saying URI everywhere >instead of URL, or whatever is the current politically correct way >of doing that. That would be required, but it's not reminding, it's a meaningful distinction. > >The difficulty comes if you want to give both a URN _and_ a URL for the >same thing -- a practice I consider dangerous at best. But if URNs >work as well as advertised, tht won't be necessary anyway :-) I think that it is a useful pragmatic way to address the resolution problems that are ineveitable, and enable additional semantics-based caching for smarter clients. We've had this discussion off-list a couple of times now. People will use URL backups for PUBLIC IDs properly -- not because they are virtuous, but because if they don't, their documents won't work. That's the way people learn how URLs work too. -- David _________________________________________ David Durand dgd@cs.bu.edu \ david@dynamicDiagrams.com Boston University Computer Science \ Sr. Analyst http://www.cs.bu.edu/students/grads/dgd/ \ Dynamic Diagrams --------------------------------------------\ http://dynamicDiagrams.com/ MAPA: mapping for the WWW \__________________________Received on Thursday, 27 March 1997 13:38:32 EST
This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 10:04:18 EDT