Re: several messages about New Vocabularies in text/html

On Fri, 23 May 2008, Alexey Proskuryakov wrote:
> On May 23, 2008, at 12:24 PM, Ian Hickson wrote:
> 
> > There were numerous requests for this change and it was made some time ago
> 
> Indeed, | incorrectly assumed that 
> <https://bugs.webkit.org/show_bug.cgi?id=11448#c6> indicated HTML5 
> changing in the same direction as WebKit, so I didn't immediately go to 
> verify that.

At the time the spec did change in the same direction. It then changed 
again subsequently based on more feedback.


> > -- does the change actually break anything?
> 
> Certainly, this change would cause rendering differences, presumably 
> including line breaking ones, but I do not know if there are pages in 
> the wild that would be considered broken.
> 
> I could try implementing the current HTML5 mapping in WebKit to see if 
> bug reports arrive, but past experience tells that discovering issues 
> like this often takes longer than one release cycle, and not necessarily 
> because the impact is small. So, I'd prefer to avoid further jumping 
> back and forth implementation-wise, if possible.

Understood.


> > My understanding is that the original lang and rang codepoints were 
> > CJK characters only because no better characters existed.
> 
> Actually, they were not - the original mapping was to unified 
> characters, which were later deprecated in favor of CJK ones, while new 
> characters were added for other uses.
> 
> It seems that we are facing basically the same issue the Unicode 
> consortium faced with disunification of U+2329/U+232A, but resolving it 
> differently. That's confusing!

Agreed. David might have more input on this, he is the one now maintaining 
the file from which HTML5 autogenerates the entities table.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'