[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [office-metadata] RDFa model and xml:id
Elias, Elias Torres wrote: >Patrick Durusau <patrick@durusau.net> wrote on 12/13/2006 02:10:33 PM: > > > >>Elias, >> >>Elias Torres wrote: >><snip> >> >> >> >>>>Only if you are careful to select the outside <div about="party1">. How >>>>do you guarantee that all select operations are going to do that? >>>> >>>>I think I understand what you mean by "in-context" metadata but I think >>>>you are missing the "correct" select issue that is required to preserve >>>>the right context. >>>> >>>> >>>> >>>> >>>I think you are somehow thinking that users see XML markup and are >>>selecting parts of it. The implementation can make use of the RDFa >>>processing rules to resolve in-context metadata to capture a large set >>> >>> >of > > >>>the selection without losing it. People don't see in the screen <div> >>> >>> >and > > >>>place their cursor after the i in div. It might be the case that we >>> >>> >might > > >>>lose one property value, but hey nothing is perfect, however, everything >>>else below it can be re-used. >>> >>> >>> >>> >>> >>No, actually that is not what the processing rules, at least as of 20 >>October 2006. Under section 4.1 it says: >> >>"The [RDFa element] under consideration at any time is the [current >>statement], and its parent element is the [context statement]." >> >>As I read the rule, only the present element and its parent are captured >>for context. That may or may not be sufficient. It is certainly an >>arbitrary limit that is being touted as a "context" solution. >> >>Better than nothing I suppose but only just. What happens if parent + 1 >>provides needed context? >> >>Oh, and you are presuming that there is no intervening content between >>the parent and its child [current statement]. I would be surprised if I >>tried to select some text and the selection expanded to include material >>I did not want. >> >>Assuming we are not limtiing the metadata that can be supplied, I could >>even point to a parent element in the metadata to control how much >>context is selected when a copy-n-paste operation is being performed. >>Which would have the same select more content problem I just noted for >> >> >RDFa. > > >>Suspect the select and preserve metadata is actually a hard problem. I >>can't say I am impressed with the parent/[current statement] default. >>Fortunately, since RDFa is not a standard, we can define whatever rules >>we think are appropriate. >> >> > >This is good feedback for us in the RDFa working group since your reading >of the text (at least the text you pasted, no-context :)) is not exactly >right. > > > Standards, good ones anyway, say anything only once and then hopefully correctly. ;-) Context would not have helped in this situation. >The statement you pasted is correct. The only thing that you are missing is >recursion. When you go to the parent, you re-use the same statement again >and again until you find a parent. > ><div about="foo"> > <div> > <div> > <div property="description"></div> > </div> > </div> ></div> > >In RDFa, that description is about <#foo>. > >Does that help? > > > Well, it did make me look for recursion anyway. ;-) Actually one does not look for the parent but for the subject (which is held as an attribute in a parent). See 4.3, where I find recursion to determine the "subject." The about attribute. And it imposes a requirement that we traverse the DOM tree to find the about attribute. So that does specify how far I must go in the tree, defaults to the document if nothing found in between. I would have to ask one of the engineers what the cost of traversing the DOM tree would be versus simply having the required data in a metadata statement. Noting that one of the tradeoffs would be that if all the RDF triples are in one or more metadata files, you don't have to process the content.xml file unless you have some compelling reason to do so. Nor do you have to walk the DOM tree. We are, afterall, specifying the rules and if we don't want to allow syntax that could cause us to hunt for the about attribute, we are not obligated to do so. Hope you are having a great day! Patrick >-Elias > > > >>Hope you are having a great day! >> >>Patrick >> >>-- >>Patrick Durusau >>Patrick@Durusau.net >>Chair, V1 - Text Processing: Office and Publishing Systems Interface >>Co-Editor, ISO 13250, Topic Maps -- Reference Model >>Member, Text Encoding Initiative Board of Directors, 2003-2005 >> >>Topic Maps: Human, not artificial, intelligence at work! >> >> >> >> > > > > > > -- Patrick Durusau Patrick@Durusau.net Chair, V1 - Text Processing: Office and Publishing Systems Interface Co-Editor, ISO 13250, Topic Maps -- Reference Model Member, Text Encoding Initiative Board of Directors, 2003-2005 Topic Maps: Human, not artificial, intelligence at work!
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]