- From: Mike Taylor <mike@indexdata.com>
- Date: Fri, 28 Mar 2003 21:58:54 GMT
- To: Theo.vanVeen@kb.nl
- CC: www-zig@w3.org, Theo.vanVeen@kb.nl
> Date: Fri, 28 Mar 2003 22:37:21 +0100 > From: "Theo van Veen" <Theo.vanVeen@kb.nl> > > > > Right. But to add some level of complexity: I do not mind having > > > names for brief and full DCX for example as long as it remains > > > clear that it is DCX. If different servers put different terms > > > in brief DCX I don't mind for the simple reason that what people > > > put in there will for 90% be something that I understand and the > > > other 10% I just ignore. > > > > But unless there's a registry (or profiling), there's no guarantee > > that my server agrees with your client that DCX is "Dublin Core, > > Extended". It might just as well be "Deep Custard, X-rated". > > Agreed. A schema saying "Qualified Dublin Core + <any /> for the > rest" will do, I think. although our applications are intelligent > enough to allow for agreements on a higher level. So what we seem to be converging on is the following agreement: "When we are requesting XML records, the element-set name can be construed to mean whatever the profile wants it to". But that's _always_ been true, whatever record syntax is requested. So what have we actually _done_ here? Anything? _/|_ _______________________________________________________________ /o ) \/ Mike Taylor <mike@indexdata.com> http://www.miketaylor.org.uk )_v__/\ Live fast, Die old. -- Listen to my wife's new CD of kids' music, _Child's Play_, at http://www.pipedreaming.org.uk/childsplay/
Received on Friday, 28 March 2003 16:59:17 UTC