On Jun 07, 2007, at 01:13, Jonas Sicking wrote: > This one is tricky for sure. I don't think it is all that tricky, sometimes you just have to let go. You don't want the AC spec to be specifying parsing strategies. You may want to *recommend* what to do depending on the parsing strategy (if tree-based and the document is not WF, AC is not processed; if stream-based and the document is not WF it doesn't matter, process AC as soon as you see it and *if* you keep reading the document then revoke it upon encountering a WF error), and you *may* want to point out that the streaming approach is best but I really wouldn't go any farther as I don't think there's any good decision here. IIRC the REX spec has that sort of weaseling somewhere, you might want to check it out. -- Robin Berjon - http://berjon.com/ ------------------------------------------------------------------------ RDF is like violence: if it doesn't work, use more!Received on Monday, 11 June 2007 15:06:54 GMT
This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:10:22 GMT