- From: Jan Roland Eriksson <jrexon@newsguy.com>
- Date: Sun, 11 Aug 2002 13:46:32 +0200
- To: www-html@w3.org
- Cc: Lachlan Cannon <luminosity@members.evolt.org>
On Sun, 11 Aug 2002 20:44:34 +1000, you wrote: > > I agree, but would rather see <script src> removed completely > > in favour of a more sensible/consistent approach like > > <link rel="script" href="...">. >I'd vote in favour of this approach too. This has the benefits of >reducing bandwidth burden on people who use browsers that don't support >scripts, namely they're not loading all that extra script, just a small >link tag. Also, I'd imagine it'd be a lot easier for a parser to know >that what it was parsing now would all be xml... and that when it >followed this link to this resource it'd only be parsign a script. IMHO >all content of markedly different type.. eg, xml, css, js should always >have to be seperate. Baring in mind that XHTML 2.0 is not designed to be >bacwards compatible and that moving forward you'll probably want to stay >backwards compatible with XHTML 2.0, I'd recommend dropping the script >element now, or at the very least deprecating it. You may count me in on the support of eliminating the 'script' element from XHTML 2.0. However, XML has "inherited" the SGML method for how to reference external information in a way that the application level can get at it when/if it needs it and that is by inclusion of 'processing instructions' in the prolog of a document instance. We already have a W3C recommendation for how to set up a PI to reference external stylesheet data... <http://www.w3.org/TR/xml-stylesheet/> ...maybe a similar approach for how to reference scripts could be adopted? I mean, what would be the difference here... (besides the obvious that there is no recommendation for the use of a xml-script PI for XML) <?xml-stylesheet href="mystyle.css" type="text/css"?> <?xml-script href="myscript.js" type="text/javascript"?> Any way, going for the use of PI's to reference non structural external info would "pave the way" for a removal of the 'link' element too :)
Received on Sunday, 11 August 2002 07:47:48 UTC