Make CSS is animatable data consistent - #30417
Conversation
|
Tip: Review these changes grouped by change (recommended for most PRs), or grouped by feature (for large PRs). |
ddbeck
left a comment
There was a problem hiding this comment.
Thanks for starting this, Chris! I think this will be a nice improvement. I've got one inline suggestion.
Additionally, I'd say that I don't love the @keyframe text, though I don't feel super strongly about it. I kinda think "Animatable and transitionable" gets the point across, but maybe others will have a different take.
For the guideline text, that will need to go into a new file: docs/data-guidelines/css.md. It'll look something like this:
# Data guidelines for CSS features
This file contains guidelines that are specific to CSS features (`css/`).
## Animatable and transitionable
Use the key name `is_animatable` to…
<!-- whatever else you have to say here, ideally imperative mood -->
This guideline was proposed in [#30417](https://github.com/mdn/browser-compat-data/pull/30417/).There will be just one guideline in it, to start. That's fine.
There was a problem hiding this comment.
AFAICT, the only difference here is between Chrome 116 and 117, so I think here we ought to have one "normal" is_animatable feature, where Chrome is partial at 116 and fully implemented at 117.
If that makes sense, we can dispense with the special case of is_transitionable.
There was a problem hiding this comment.
This is a nice idea. I've had a go at updating it.
Notes:
- On the 116 partial data, I've put a note "
@keyframeanimatable only" - For the overall description, I've put a slightly different form than previous ones we've discussed — "
@keyframeanimatable; transitionable when settingtransition-behavior: allow-discrete". This is because the keyframe animatable part doesn't requiretransition-behavior: allow-discreteto be set, but transitions do. I didn't want people thinkingtransition-behavior: allow-discretewas required for both.
The Maybe it's not that important. Perhaps we should get a third opinion on this. |
I've added the data guideline too. |
Co-authored-by: Daniel D. Beck <daniel@ddbeck.com>
Co-authored-by: Daniel D. Beck <daniel@ddbeck.com>
ddbeck
left a comment
There was a problem hiding this comment.
I'm happy with this, but this needs more eyes from owners. I've put it on the agenda for the next BCD meeting.
Summary
As per #30358 (comment), This PR aims to make CSS data more consistent in cases where a feature (property or type) started its life off not animatable, but was made animatable later on.
In such cases, I am using the following keys:
is_animatablewhen the data point describes when the feature was made@keyframeanimatable, or@keyframeanimatable and transitionable.is_transitionablewhen the data point describes when the feature was made transitionable.And the following
descriptionstructure:@keyframeanimatable" when the data point describes when the feature was made@keyframeanimatable.@keyframeanimatable and transitionable" when the data point describes when the feature was made@keyframeanimatable and transitionable.The
descriptionalso has an optional suffix: "when x", where x is a particular condition under which the feature is animatable and/or transitionable. For example — "@keyframeanimatable and transitionable when setting inset properties"Notes:
text-combine-uprightstarted off animatable in Fx, but that was a mistake, and it was corrected per spec later on. I don't think this is worth adding a separate data point for, so I've just left that one as a note.Test results and supporting details
Related issues