- From: Rik Cabanier <cabanier@gmail.com>
- Date: Sat, 24 Nov 2012 11:10:54 -0800
- To: Boris Zbarsky <bzbarsky@mit.edu>
- Cc: www-style@w3.org
- Message-ID: <CAGN7qDAU=p6X8LSgEj9AYiOAEOiT8EosXx6bf7h=skemqa8coA@mail.gmail.com>
On Sat, Nov 24, 2012 at 5:29 AM, Boris Zbarsky <bzbarsky@mit.edu> wrote: > On 11/24/12 12:42 AM, Rik Cabanier wrote: > >> getComputedStyle in WebKit and FF is returning '100px' which is not the >> used value. (I assume by 'used value' you mean the value in the style >> sheet) >> > > Wrong assumption. The value in the stylesheet is the "specified value". > The "used value" is what you have once you've actually gone and done > layout and a few other things. See http://www.w3.org/TR/CSS21/** > cascade.html#value-stages<http://www.w3.org/TR/CSS21/cascade.html#value-stages> > > Indeed. The computed value is "auto" here. >> >> no, the '_resolved value'_ is auto according to the spec. >> > > The _computed_ value is "auto" according to http://www.w3.org/TR/CSS21/** > visudet.html#the-height-**property<http://www.w3.org/TR/CSS21/visudet.html#the-height-property> > > The _resolved_ value is 100px because for this property it's the used > value. The spec for determining the used value of height in this case is > http://www.w3.org/TR/CSS21/**visudet.html#root-height<http://www.w3.org/TR/CSS21/visudet.html#root-height>(because of the "overflow" styling). > > > The computed value is still what getComputedStyle() returns. >> > > No. It's just not. yes, that is where I disconnected. (Thanks for the explanation David!) GetComputedStyle returns the 'resolved' value - not the 'computed' value - which in this case is the 'used value' > > It seems up to the UA if it wants to return 'auto' or the actual height. >> > > No, it's not. > > > Maybe I'm reading the spec wrong or it is worded incorrectly... >> > > You're just reading it totally wrong, and I'm not quite sure how you're > managing it. Apart from the incorrect assumption about "used value", of > course. That was it :-) I'll correct my previous statement: I think transitions should happen with *resolved* values. I don't disagree but there has to be some sort of simple rule. >> > > Well, one simple rule is "no transitions to/from auto for height"... True. This thread started of with users asking for this feature so it would be nice if we can come up with some level of support which can be implemented and documented.
Received on Saturday, 24 November 2012 19:11:21 UTC