W3C home > Mailing lists > Public > public-webapi@w3.org > April 2008

Re: [selectors-api] NSResolver moving nodes between documents

From: Jonas Sicking <jonas@sicking.cc>
Date: Mon, 14 Apr 2008 17:26:29 -0700
Message-ID: <4803F635.3080709@sicking.cc>
To: Bjoern Hoehrmann <derhoermi@gmx.net>
CC: "Web APIs WG (public)" <public-webapi@w3.org>

Bjoern Hoehrmann wrote:
> * Jonas Sicking wrote:
>> This should be the easiest solution for spec writers and implementors 
>> alike. And I don't think users will care either way.
> 
> I do not think we need to specify anything in addition to what the draft
> already has, nor do I think we should define that prefixes must be re-
> solved before querying the DOM but then don't define in which order and
> how often each prefix may be resolved, but I don't care much either.

I think it would be a good idea to defined that the implementation 
should ensure to make all needed calls to NSResolver before starting to 
navigate the DOM to build the result.

> In
> any case, if we add something, please specify what happens if you switch
> between HTML and XML document with respect to case-sensitivity of names.

This is something that is in general very undefined by the DOM. What 
happens if you move a node from a HTML document to an XHTML document and 
then call getElementsByTagName on it. Do you get a case sensitive match 
or not?

I would prefer not to touch this area for this spec, but I guess I can 
be convinced as long as we can ensure we're consistent with how other 
APIs behave. It would be very bad both for users and implementations if 
some APIs switched to be case sensitive and some didn't.

/ Jonas
Received on Tuesday, 15 April 2008 00:29:02 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 15 April 2008 00:29:04 GMT