- From: Tex Texin <tex@xencraft.com>
- Date: Sun, 07 Aug 2005 15:08:17 -0700
- To: fantasai <fantasai.lists@inkedblade.net>
- CC: Richard Ishida <ishida@w3.org>, 'WWW International' <www-international@w3.org>, Jony Rosenne <rosennej@qsm.co.il>
It is not clear to me that we need to insist on the document being coherent without css. There are already many other things you can do with css that have that defect. If I use the :before or :after with "content:" for example. Similarly, without positioning information that is in css styles, some documents become unsolved jigsaw puzzles. Language selectors can be used with css for other features, so although I see that bidi might take on additional risk, I wouldn't prohibit it and I would like for bidi to have the same benefits as other languages with respect to reduced file size and less markup, by leveraging styles. It would perhaps be good if these design goals (being coherent without css, not putting bidi in css, etc.) were collected and documented in one place and then put to a review with respect to how best to implement bidi. I have seen these in bits and pieces and mostly from discussion with others, but I am not aware of a real architectural overview of how bidi should work on the web. (Beyond the unicode bidi algorithm and the joint Unicode-W3C doc on markup on the web, neither of which is comprehensive with respect to bidi on the web.) The individual specs (CSS, HTML, etc.) make some comments but there isn't an integrated overview, and some of the specs are contradicted by recommendations elsewhere. (e.g. no bidi in css) Of course, I would be delighted to learn that there is such a doc. (I readily admit to not being up on the latest css3 in that regard.) If not, the W3C should work towards one. fantasai wrote: > > Tex Texin wrote: > > > > I would like to comment also, that for HTML, TABLE and other elements of > > HTML, I do see the need for the DIR attribute. I am not trying to have > > the bidi markup deprecated. I am more concerned with straight runs of > > text embedded in markup and I don't see why I should give up on WYSIWYG > > editing of that to satisfy the recommendation. > ... > > We should rope into this a discussion the use of the CSS bidi > > facilities. Last I looked bidi css were out of favor, but they work very > > well for me, and I think its fine to tie markup to language (when I am > > not using control codes! ;-) ) > > Indeed. CSS2.1 and CSS3 will be continuing in that line line of thinking. > BIDI embeddings are much more essential to a document's coherence than > other things that can be controlled with CSS. In theory, you should be > able to turn all CSS off and still be able to read the page. If you're > relying on CSS to do BIDI, that no longer holds true. > > This W3C I18N QA sums up the situation pretty well: > http://www.w3.org/International/questions/qa-bidi-css-markup > > The only paragraph I disagree with is > > | XHTML served as application/xhtml+xml, application/xml or text/xml > | is XML and so needs CSS to map the markup to the appropriate display > | behaviour, as described in 'General XML-based markup languages' above. > > which isn't really true: if you follow that line of thinking, you'd also > need to provide the default styling for every other element and attribute > as well as some way of noting the interactive behavior of links and form > elements. As long as the XHTML elements are namespaced correctly, browsers > should map the document's elements and attributes, including 'dir' and > <bdo>, to the correct behavior by themselves. > > ~fantasai -- ------------------------------------------------------------- Tex Texin cell: +1 781 789 1898 mailto:Tex@XenCraft.com Xen Master http://www.i18nGuy.com XenCraft http://www.XenCraft.com Making e-Business Work Around the World -------------------------------------------------------------
Received on Sunday, 7 August 2005 22:08:36 UTC