User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5) Gecko/20041231 Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.5) Gecko/20041231 It seams, that mozilla lower-cases nodeName of attribute. If an element has an attribute 'myAttribute', the nodeName of the attribute in mozilla is 'myattribute', but should be 'myAttribute'. Reproducible: Always Steps to Reproduce: 1.load attached test page 2. 3. Actual Results: Mozilla allerts nodeName of 'v:minLength' attributes in div and input. Both nodeNames are in lowercase. Expected Results: It should alert 'v:minLength' in both cases.

Actually, the nodename should be all uppercase for HTML. See http://w3.org/TR/2003/REC-DOM-Level-2-HTML-20030109/html.html#ID-5353782642 paragraph two.

This does depend on the patch in bug 103225. I can remove that dependency if desired.

mroz: If you want to use XHTML (where case of attributes and elements will be preserved) you should use application/xhtml+xml or text/xml as mimetype. text/html is always internpreted as HTML

How about nsDOMAttribute::GetLocalName?

Comment on attachment 170493 [details] [diff] [review] Patch to that effect There's an apparently unpublished errata to the DOM Level 2 HTML spec about this, it's at http://www.w3.org/DOM/Group/drafts/2000/11/DOM-Level-2-errata#html-2 (for my reference, as I don't think that's publicly reachable). That errata states: --- html-2. [error]. XHTML and the HTML DOM. The sentences: For instance, element and attribute names are exposed as all uppercase (for consistency) when used on an HTML document, regardless of the character case used in the markup. Since XHTML is based on XML, in XHTML everything is case sensitive, and element and attribute names must be lowercase in the markup. should read For instance, element names are exposed as all uppercase (for consistency) when used on an HTML document, regardless of the character case used in the markup. The names of attributes defined in HTML 4.01 are also exposed as all lowercase, regardless of the character case used in the markup, but for other attributes (i.e. ones that are not defined by HTML 4.01), the character casing is implementation dependent. Since XHTML is based on XML, in XHTML everything is case sensitive, and element and attribute names must be lowercase in the markup. --- So in other words we don't want this change. If anything we should identify valid HTML attributes and leave those unchanged (in the parser or wherever we lowercase attribute names today). I'm checking why this errata is not public yet...

Er... per that text the HTML 4.01 attributes _must_ be lowercased (as we do), and the other attributes _may_ be lowercased (as we do). So in other words, this bug is invalid, as far as I can see.

Yeah, um, that's what I meant, except that we *could* change our code to leave non-HTML attribute names alone and not mess with the character case in those names at all, as IE apparently does. Not high priority tho, as we're not in violation of the spec (errata) here at all.

I think that would more effort than is worth it (we'd need to hardcode a list of HTML attributes somewhere). Also, the spec is unclear as to what happens for "HTML attributes" that are on the wrong element (eg <div sRc="whatever">). So unless we plan to reverse-engineer IE...

Since nobody has arguments against Comment #9, and 3 months passed,I think I can safely mark the bug as WONTFIX