I've been following this discussion on browser standards with interest, but one thought keeps nagging at me.
One of the mentioned benefits is this browser or that browser's XSLT parser and its standards compliance. I think that we all agree the standard compliance and also the promotion of open standards are good things. The little nagging thought comes with this discussion of XML transformation in the browser. Who actually requires this of their users? Many users have extremely out of date browsers which will display HTML with (sometimes if you are lucky) a bit of CSS. Many don't have browsers that have XSLT parsers built in. So it really isn't an issue, to me, because any site I design will always try to serve (X)HTML as a bare minimum.[1] The transformations of XML to HTML via XSLT may be done on the fly, but they certainly won't be done in the user's browser. I don't trust users to have any particular browser. So if the transformations (with my preference with apache's cocoon) are done on the server, then the XSLT parser of the browser doesn't really matter to me that much. Now, it is a shame that all browsers don't follow the same standard in the same way (esp. with regard to CSS), but I don't see that there is much I can do about that, only follow the spec. and hope the browsers finally sort things out.[2]
-James
[1] I hasten to add that I didn't design the Oxford Text Archive's website, and that we will be redesigning it to be entirely open standards compliant, once we have finished totally reorganising the back end and our matching workflow. [2] Ok, I use Firefox, and yes I could donate time and energy to the mozilla project to help it, but I don't really have any spare reserves of either.
--- Dr James Cummings, Oxford Text Archive, University of Oxford James dot Cummings at ota dot ahds dot ac dot uk