- From: Simon St.Laurent <simonstl@simonstl.com>
- Date: Mon, 26 Mar 2012 12:42:02 -0400
- To: public-cssselfrags@w3.org
[Sorry - still recovering from a crazy mixture of sneezing and SXSW.] On 3/9/12 12:30 PM, Robin Berjon wrote: > On Mar 9, 2012, at 17:37 , Chris Lilley wrote: >> RB> """ RB> If a location-set contains multiple nodes, the >> processor should RB> consider that this fragment identifier is >> pointing to all of them RB> simultaneously (e.g. for :target >> processing for instance). >> >> That part is good. >> >> RB> However, the identification of multiple results may not be RB> >> appropriate in some contexts. In these contexts, the processor RB> >> must return only the first result in document order. RB> """ >> >> That part is wooly and largely negates the clarity of the first >> part. > > I was wondering if something specific was being considered there, or > if we could just drop it. I think it would be fine for the processor > to always return multiple results, and if there are contexts that > only want the first (or the last, or the seventh from the back) then > they can define that additional constraint themselves. I think what we were attempting to do here was preserve the option of multiple results - which enable some powerful things like XLink processing and interfaces beyond "GOTO X LOCATION" - while not terrifying developers who get to implement these things. Would it be more sensible to say that the processor should return all results, but the recipient can do whatever it wants with them? Mostly likely, the first result in document order will be privileged... Thanks, -- Simon St.Laurent http://simonstl.com/
Received on Monday, 26 March 2012 16:43:18 UTC