- From: Nabil Layaida <layaida@sol.genie.uottawa.ca>
- Date: Thu, 27 Nov 1997 15:48:16 -0500
- To: "'Nabil Layaida'" <layaida@sol.genie.uottawa.ca>, "'Andrew n marshall'" <amarshal@usc.edu>
- Cc: "'www-multimedia@w3.org'" <www-multimedia@w3.org>
---------- From: Andrew n marshall Sent: Wednesday, November 26, 1997 3:15 PM To: 'Nabil Layaida' Cc: 'www-multimedia@w3.org' Subject: RE: Layout Revision Proposal: CSS and then some Hmmm... Adding a grouping in the form of another tag doesn't quite address the issue. It still is limited to temporal adjacency. For example, using standard TV and newscast models, how do you denote that two particular news reports (probably each implemented as a <PAR> tag) are siblings in terms of content organization when a commercial comes between them? For example: <SEQ id='newscast'> <PAR id='report1'> . . . </PAR> <PAR id='report2'> . . . </PAR> <SEQ id='sponsors'> . . . </SEQ> <PAR id='report3'> . . . </PAR> <PAR id='report4'> . . . </PAR> </SEQ> I did get the point, but what I meant goes into that same direction : To use mark-up for logical structuring (eg. Based on content only) and not on spatial or temporal aspects you require two things. 1) A logical markup of the document itself, for example tags for titles, sections, paragraphs, lists, ...... 2) Externalize the presentation aspects related to temporal or spatial information The argument for the comp tag is that you can use it as an aggregate to group elements logically, and independent from layout or synchronization. And based on these aggregates you can synchronize elements (possibly in different files) using only the names of these aggregates. Therefore, for the same content you will be able to apply two synchronization schemas just as CSS allows you to do right now. In that case you may be able to describe your document as follows, given that every element has an id, we'll have : Content file: <comp id='newscast'> <comp id='report1'> . . . </PAR> <comp id='report2'> . . . </PAR> <comp id='sponsors'> . . . </SEQ> <comp id='report3'> . . . </PAR> <comp id='report4'> . . . </PAR> </comp> Of course in a generalized mark-up instead of comps you may want to have section, sub-section, etc..... Synchronization 1: { syntax is not important here } Newscast : { seq : report1 report2 sponsors report3 report4 } Report1 : { par : ......} ...... Here is the same document without the commercial Synchronization 2: { syntax is not important here too } Newscast : { seq : report1 report2 report3 report4 } Report1 : { par : ......} ...... The advantage of having this is quite obvious. But, the reason that this was not taken in version 1, is that most of the multimedia content comes from separate files (audio, video images and even the text). So the content hierarchy will in fact have very little importance compared to what we have in text-based documents. I think for SGML in particular. Therefore we concentrated our efforts on the synchronization aspects while content is left to other standards like HTML for text, Mpeg, Jpeg, ......for media content. Nabil. Andrew n marshall student - artist - programmer http://www.media-electronica.com/anm-bin/anm "Everyone a mentor, Everyone a pupil"
Received on Thursday, 27 November 1997 16:46:42 UTC