Re: MathML in HTML: syntax for hypertext links

Jim Jewett wrote:
> Not at the moment; it may have even been in the xhtml 2 discussion
> list.  I'll try to remember to search later.


Thanks!

> And of course, if *you* are questioning it, the opinions may have changed in the last year or
> two.

Or UA implementor opinions are just not very monolithic on this.  Both 
are possible!

>> Really?  Here's some markup that produces nested <a> elements in at
>> least Gecko and Opera:
> 
>> http://software.hixie.ch/utilities/js/live-dom-viewer/?%3C!DOCTYPE%20html%3E%0A%3Chtml%3E%0A%20%20%3Cbody%3E%0A%20%20%20%20%3Ca%3E%3Cspan%3E%3Ctable%3E%3Ctr%3E%3Ctd%3E%3Ca%3E%3C%2Fa%3E%3C%2Ftd%3E%3C%2Ftr%3E%3C%2Ftable%3E%3C%2Fspan%3E
> 
> Im using Opera 9.64, and that it (says it) reparents the <table> as a
> sibling rather than a child of the <a>.

My apologies.  I mixed up my Opera and Safari windows.  Safari puts the 
<table> inside the <a> (and unlike Gecko it doesn't even need a <span> 
there to do it).
> 
>> Yes, true.  But HTML5 is defining the nested-href behavior anyway, since
>> it's trivial to get it in XHTML and not that hard to get it in HTML+DOM.
> 
> There is a difference between defining error correction and supporting
> in a way that encourages use.

Fully agreed, and I agree that nested links make for terrible user 
experience, in general.  That's a separate argument from them being hard 
to implement well, though, since the UA has to do something for them 
anyway because they can arise in existing pages.

-Boris

Received on Wednesday, 8 April 2009 17:18:40 UTC