- From: Michael(tm) Smith <mike@w3.org>
- Date: Wed, 30 Apr 2008 15:16:21 +0900
- To: Chris Wilson <Chris.Wilson@microsoft.com>
- Cc: Thomas Broyer <t.broyer@gmail.com>, public-html@w3.org
- Message-ID: <20080430061619.GC19512@sideshowbarker>
Hi Chris, Just a heads-up about the following message with a question about parsing behavior in IE8. --Mike Thomas Broyer <t.broyer@gmail.com>, 2008-04-10 16:16 +0200: > On Thu, Apr 10, 2008 at 11:51 AM, Ian Hickson wrote: > > > > A common theme across some of the problems listed above, and discussed in > > detail in the e-mails below, is the issue of a generic syntax for non-HTML > > namespaces in text/html. It turns out to be _exceedingly_ hard to handle > > the giant mass of weirdness found in legacy content on the Web while > > adding new features. For example, this page: > > > > http://www.laroseweb.com/calcs/fans.php > > > > <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" > > "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> > > <html xmlns="http://www.w3.org/1999/xhtml" > > > <math xmlns="http://www.w3.org/1998/Math/MathML"> > > <math xmlns="&mathml;"> > > > > <head> > > [... continues with just a normal XHTML page] > > ...renders exactly the same in Firefox 2 and IE8b1 (don't know why, > probably because they come before <head>) > > > ...or this page: > > > > http://www.cocopahrv.com/map.html > > > > [...] > > <table width="592" border="0" cellspacing="0" cellpadding="0"> > > <tr align="left" valign="top"> > > <td width="592"><svg width="200" height="200" viewBox="0 0 > > 200 200"><img name="r1_c1" src="map/resortmap.jpg" width="600" > > height="465" border="0" alt="Cocopah RV and Golf Resort" > > usemap="#r1_c1Map"><map name="r1_c1Map"><area shape="rect" > > coords="4,2,298,230" href="map/lmp1a.jpg"><area shape="rect" > > coords="1,230,296,460" href="map/lmp1b.jpg"><area shape="rect" > > coords="297,4,598,231" href="map/lmp2a.jpg"><area shape="rect" > > coords="297,231,599,462" href="map/lmp2b.jpg"></map><rect > > x="100" y="100" width="10" height="20" style="fill: green"/> > > </svg> </td> > > </tr> > > </table> > > [...] > > Don't use xmlns="" so renders exactly the same in Firefox 2 and IE8b1 > (I've nly tested those two; not that I couldn't have tested more, but > I thought it was sufficient). > > > http://kr.blog.yahoo.com/yunneo2000/1111673 > > > http://puysl.com/view.htm > > http://albren.blogspot.com/2007/02/l05-traductor.html > > > On Tue, 10 Apr 2007, Sam Ruby wrote: > > > > > > Instead of "starts with x_", how about "contains a colon"? > > > > I considered that, but unfortunately, it forms a dichotomy of semantics, > > where when used in XML the DOM shows one set of names, and when used in > > text/html, it shows another (at least in legacy UAs, though we could > > hard-code behaviour for new UAs). > > IE already does this (report the thing following the first colon) > since version 5.5 (if there's a corresponding xmlns:<prefix> attribute > on the <html> element or a <?XML:NAMESPACE prefix=<prefix> /> > pseudo-processing-instruction) > If we follow the same rules (except for the <?XML:NAMESPACE />), we > probably won't break much pages... > > > On Sun, 9 Mar 2008, Sam Ruby wrote: > > > > > > > Establishing a pseudo-XML parsing scope for <svg> and math. > > > > > > IE8's approach seems to be "establish a pseudo-XML parsing scope for > > > unknown elements which contain an attribute named xmlns that happens to > > > match a list of known values", where the list of known values may vary > > > by user agent or installation. IMHO, that approach merits exploration. > > > > Unfortunately, exploration basically shows that IE8 doesn't do anything > > (at least, I couldn't reproduce it) with xmlns="", other than screw up the > > attribute parsing), > > It processes "/>" and puts the elements in a different "scope" (cf. > scopeName and tagUrn). > > > and a quick study of existing documents suggests that > > there's a great deal of content that wouldn't be handled very well at all > > by anything triggering off xmlns="", > > You might well be correct (Sam's blog fails to render in IE8b1, but it > seems to be a bug in their parser, probably because of the screwed-up > attribute parsing), though aside from those parser-bugs I guess > Microsoft has done some testing on their own... > Maybe someone from Microsoft could comment on this? Chris? > > > even if it were limited to unknown elements or known namespaces > > (e.g. see the examples at the top of this e-mail). > > They're all rendered "correctly" (i.e. the same as in Firefox) in IE8b1. > > Actually, the only problem I see with IE8's "enhanced namespace > support" is the lack of namespace support for attributes. > > > And fortunately, supporting namespaced elements (à la IE8 or > IE5.5/6/7) is not incompatible with your effort for supporting SVG and > MathML "out-of-the-box" (great job by the way). > > > My 2 c€nts... > > -- > Thomas Broyer -- Michael(tm) Smith http://people.w3.org/mike/ http://sideshowbarker.net/
Received on Wednesday, 30 April 2008 06:16:59 UTC