- From: Sylvain Galineau <sylvaing@microsoft.com>
- Date: Thu, 17 Jan 2013 18:29:39 +0000
- To: Øyvind Stenhaug <oyvinds@opera.com>, "www-style@w3.org" <www-style@w3.org>
[Øyvind Stenhaug:] > > > > [Øyvind Stenhaug:] > > >> <!doctype html> > >> <style> > >> @keyframes foo { from { margin-left: 0px; } to { margin-left: 600px; > >> } } @keyframes bar { from { margin-left: 0px; } to { margin-left: > >> 600px; } } p { animation-duration: 10s; animation-name: foo; } .first > >> { animation-name: foo, bar; } .second { animation-name: bar, foo; } > >> </style> <p>Moving right</p> > > >> So if all instances of 'bar' in this example is replaced by 'foo', > >> which case does that correspond to? > > > > When going from 'foo,foo' to 'foo', see above. I think it would be > > consistent for the behavior to be consistent with start with 'foo' and > > 2 durations. > > > > When starting with 'foo' and updating to 'foo,foo' then I would expect > > the second foo to 'win' i.e. like that bar in .first. > > OK, so when matching the old 'animation-name' value to a new one, repeated > names are matched up from left to right. Sounds sensible to me. > Yes; I think it makes sense for the updated animation-name value to be first 're-aligned' with the value(s) of the other animation-* properties so each animation name gets its proper duration, timing function etc. Still though, you are pointing to something that is somewhat underspecified: when animation-name is updated, under which conditions do those animations that were already in the list continue vs. restart? If I moved foo one slot down the name list where it will have the same duration but a different timing function, does it restart? What if everything is the same except its iteration count? Or its delay? I'll log a couple of spec bugs for this. Thank you!
Received on Thursday, 17 January 2013 18:30:54 UTC