Use id attribute instead of name for ToC

[josh]

The name attribute is no longer considered a global attribute or part of the a element as of the XHTML/HTML5 spec. id has replaced this deprecated usage of the name attribute. name is now reserved only for "form controls".

Note that in XHTML 1.0, the name attribute of these elements is formally deprecated, and will be removed in a subsequent version of XHTML. - http://www.w3.org/TR/xhtml1/#h-4.10

Related

• Support non-English characters in anchor names #64 cc @jakedouglas @brianmario
• The a element - http://www.w3.org/TR/html5/text-level-semantics.html#the-a-element
• The id attribute - http://www.w3.org/TR/html5/dom.html#the-id-attribute
• The name attribute - http://www.w3.org/TR/html5/forms.html#naming-form-controls:-the-name-attribute

cc @jch @raganwald @mislav

[josh]

Hoping @jakedouglas or @brianmario can review this change. The escape utils called was removed in #64. It seems like nokogiri did some magic with the name attribute, but won't for the id.

[bkeepers]

Oh dang! That is really unfortunate. This is going to cause trouble with user content clashing with IDs used by the app. See github/markup#219 (comment)

Does HTML5 define any way for creating anchors to user-generated content without the potential of conflicting with the app? IMHO, id should really be reserved for implementors and not user-generated content.

[josh]

Gah, I really wish we could prefix these.

Another issue with name is that in IE, all the references are exported to the window object. So if you do <a name=foo>. window.foo will reference that element.

This can introduce other JS bugs. Mainly around feature detecting. Say you added a # Promise header to your md, you'd actually break the Promise js polyfill.

[josh]

Still can't merge as is. We're discussing some sort of prefixing solution.

[raganwald]

Another issue with name is that in IE, all the references are exported to the window object. So if you do <a name=foo>. window.foo will reference that element.

This can introduce other JS bugs. Mainly around feature detecting. Say you added a # Promise header to your md, you'd actually break the Promise js polyfill.

Why oh why oh why hasn't IE died in a 🔥 yet?

[josh]

Closing for now. Thinking about a different approach.

[mislav]

Yeah we can't let people inject IDs in the page. But we also can't break everyones URLs w/ fragment identifiers by prefixing all the headings. If we prefix, we need to handle old fragment identifiers in JS and perform scrolling manually. Not sure I want us to go down that road—another browser feature reinvented.