Re: [css3-values][css-variables] definition of <value>

(12/04/24 4:40), Tab Atkins Jr. wrote:
> On Mon, Apr 23, 2012 at 1:27 PM, Kang-Hao (Kenny) Lu
> <kennyluck@csail.mit.edu> wrote:
>> (12/04/24 4:01), Tab Atkins Jr. wrote:
>>> Heya Kenny!
>>>
>>> Below are the resolutions of the issues you raised in this thread.
>>>
>>> Issue 12: cycle() and values that have commas
>>> Closed as OutOfScope pending WG resolution - we're proposing to punt
>>> cycle() to the next level so we can address these and other issues
>>> with cycle() more properly.
>>
>> Does this mean that <fallback> for attr() can contain commas? I have no
>> problem with this but I should warn that something like "attr(x, 50%,
>> 50%)" prevents us from having a third argument in the future (which I
>> don't consider a big problem), not to mention that it is a big ugly.
>>
>> Assuming the above interpretation of the current draft (i.e. attr() can
>> contain commas), I am satisfied with this as long as you editors are
>> satisfied, but If the idea is to make the interpreation ambiguous at
>> this level, I would like to request an explicit "undefined" about
>> <fallback> in this regard.
> 
> Yup, commas are allowed in the fallback. 

OK. I am satisfied then.

> We believe we can work around any issues in the future with adding more arguments.

But now I am confused by this sentence. How do you add more arguments to
attr() if <fallback> allows commas?


Cheers,
Kenny