- From: Roger B. Sidje <rbs@maths.uq.edu.au>
- Date: Thu, 21 Mar 2002 10:14:30 +1000
- To: David Carlisle <davidc@nag.co.uk>
- CC: mozilla-mathml@mozilla.org, www-math@w3.org
David Carlisle wrote: > er there is an accepted bug in IE6 which means it rejects the mathml dtd > so adding a DTD would limit the cross browser portability at the present > time. Hope IE6 can resolve the bug sooner rather than later so that we don't have to continuously work around their bugs. > > This will affect > > the rendering in m0.9.9+ since MathML data that used to be loaded > > at startup are now catalog driven. > > Oh as the output from XSLT never has any entity references, i thought > that mozilla's special casing of the mathml dtd wouldn't be required here > (and it seems to work on my mozilla?) m0.9.9 is fine. It will only affect m0.9.9+ (and onwards). [Note: there are PUA values in the Mozilla's mathml.dtd to handle surrogate and plane-1 characters via entities given that the MathML spec is ahead of the means at our disposal ATM, especially w.r.t. cross-platform support.] > Having read the bug report through, is it right that if, in the mozilla > case, I add a link element pointing to mozilla's copy of mathml.css > then everything would be fine? (If so, what URI can I use to find > mozilla's installed copy?) What about adding the DOCTYPE for Mozilla instead? More and more things are being made catalog driven (as per the catalog spec) to avoid doing unnecessary things upfront, especially those things that are deemed tangential for the majority of users. --- RBS
Received on Wednesday, 20 March 2002 19:14:44 UTC