Re: #refs in .../current-work/multipage/

Josh Sled wrote:
> While looking at [1], and clicking on the link for "[RFC3066]", I notice that
> the fragment is updated relative to the section [2], instead of to the
> relevant (though presently unfinished) "section-references.html".

That's because the refsRFC3066 target doesn't exist at all (yet), so the 
conversion script [1] doesn't know that it ought to point into the 
references section - it looks almost exactly like a broken link to any 
other section. In that situation, the script just prints a warning and 
then keeps the original (broken) relative link.

It could be hard-coded to assume any #refs* should link to the 
references section [and then it would only miss #CSSOM, which looks like 
a bug in the spec and it should be #refsCSSOM] - would it be worthwhile 
making that change, given that it's not currently going to link to 
useful information (while the references section is empty) and that the 
links are already going to automatically start working once the useful 
information is added?

(There are some other changes I'd like to make to the multipage version, 
like combining lots of the small sections into a single page, so I 
expect I'll try updating the script at some point, and so I'm interested 
in suggestions to improve the output.)


[1] http://html5.googlecode.com/svn/trunk/spec-splitter/spec-splitter.py

-- 
Philip Taylor
philip@zaynar.demon.co.uk

Received on Friday, 22 June 2007 15:04:42 UTC