Re: [css3-transitions][css3-values] transition-duration's inital value is '0' without a unit

On Fri, Oct 22, 2010 at 7:35 PM, Sylvain Galineau
<sylvaing@microsoft.com> wrote:
>> From: L. David Baron [mailto:dbaron@dbaron.org]
>> Sent: Friday, October 22, 2010 7:06 PM
>> To: Sylvain Galineau
>> Cc: www-style list
>> Subject: Re: [css3-transitions][css3-values] transition-duration's
>> inital value is '0' without a unit
>>
>> On Friday 2010-10-22 23:40 +0000, Sylvain Galineau wrote:
>> > So either the former picks a unit or we extend the optional-unit-for-
>> zero exception of <length>
>> > to times for consistency. Although in the latter case it would seem
>> coherent to do the same for
>> > frequencies and angles as well. It's the bigger change but I lean
>> towards the latter. It'd be
>> > odd if width:0 was valid but not transition-duration:0 or
>> transform:rotate(0). (Although that
>> > ship may already have sailed, implementation-wise...).
>>
>> The gradient syntax depends on unitless zero not being allowed for
>> angles.  See the thread starting with
>> http://lists.w3.org/Archives/Public/www-style/2009Nov/0043.html .  I
>> believe we agreed on that at some point following the thread, though
>> I'm not sure.
>
> OK.
> 1. For <time> and transition-duration, both Firefox 4b6 and Chrome 7 return "0s". Can
> we simply make this the initial value for transition-duration in the spec ?

Yeah, we should.


> 2. For <angle>, my bad: I forgot that this would most likely always require a unit.
> The decision here seems to be whether existing transform implementations are wrong
> or make an exception for them.

They're wrong, but I think in the prior discussions we figured it was
probably okay to just say they're an exception at this point.  I'd
have to look up the old gradient discussions and check to be sure,
though.


> 3. For <frequency>, do we make 0 <frequency> unitless.

I doubt we need to.


> Given the gradient issue, I am now a bit warier of making 0 values unit-less since
> it could cause ambiguity in the future.

Right; I don't think we should make anything unitless that isn't
already so.  Gradients are the first place where there would have been
ambiguity, but I'd bet it wouldn't be the last.  Using a unit makes it
clearer what the value is meant for, anyway, which I think is worth
the marginal cost of (typically) 2 extra characters per value.

~TJ

Received on Monday, 25 October 2010 17:41:27 UTC