Anne van Kesteren

CSS WG: transitions and animations

Currently a part of the WG is discussing fonts. Petite caps in addition to small caps, computed value of 'font-weight', and other interesting details I do not have much feedback for. So I thought I would write something about earlier today. This morning we discussed transitions and animations. We figured out a way to solve the problem with animating to and from e.g. display:none. (I.e. when animating from display:none you want the 'display' property to have its value change immediately and when animating to display:none you want to change it at the end of the animation.) The discussion did not conclude, but the relevant drafts will hopefully soon be updated with the new proposal. The new proposal is something akin to a steps() function for discrete changes to property values in addition to making all properties animatable. (Hi typLAB!)

After that a brainstorming session followed about whether it would be possible to merge transitions and animations. howcome argued they had substantial property overlap:


Furthermore, some of the functionality limited to animations would potentially be interesting for transitions too, i.e. keyframes. There is a long thread on www-style too about transitions versus animations. As currently proposed they are quite different though and a proposal for how both can be addressed with a single syntax that meets all requirements has not been put forward. Having said that, dino is going to take another look as to whether this could be made to work.


  1. The most important thing I want for font CSS is "the size which lays out the text maximally large enough to fix in its containing box, with no wrapping." ARGH! So much time wasted!!!

    Posted by Mogden at

  2. If you could goad them into specifying 'font-variant' must be picked up in @font-face rules, the same way 'font-weight' and 'font-style' are, that'd be brilliant.

    Right now there's no good way to implement Small-Caps via @font-face with proper fallback because of this: see this font-face test case.

    Posted by James John Malcolm at