- From: Ian Hickson <ian@hixie.ch>
- Date: Fri, 8 Apr 2005 14:11:27 +0000 (UTC)
On Fri, 8 Apr 2005, Henri Sivonen wrote: > Ian Hickson wrote: > > At the end of the day this would just be saying "in XML you can also > > do this". Avoiding those options for people who serialise to both XML > > and HTML is relatively easy, just like avoiding xml:base and MathML. > > Detecting stuff in non-XHTML namespaces is significantly easier than > detecting incompatible use of XHTML elements. Very true. But you have to do the checking anyway. Say someone wrote this (non-conforming) markup: ... <p> <input> Test </input> </p> ... You wouldn't be able to serialise it to HTML. Or this: <style> <em> { color: red; } </style> Again, you need special serialisation code for that. (In fact you need special code for <style> in general, since it's PCDATA in XHTML and CDATA in HTML.) Come to think of it, you also need special processing for <script> -- processing that actually changes the script, since for legacy UAs you need the non-namespaced methods but in XML you have to use the namespaced ones. So I don't think the situation is as clear cut as all that. Is catching <ul>s inside <p>s something that crosses the line? -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Friday, 8 April 2005 07:11:27 UTC