Correct Safari data for HTML Element APIs #7402
Conversation
github-actions
bot
added
the
data:api
There was a problem hiding this comment.
This patch seems to have helped surface some instances of some ghost features in our data that should just be completely removed. I now feel like we should be looking into how we could somehow automate identifying other ghost features like this ones this patch has exposed. It seems like y’all might have the means to do that with mdn-bcd-collector stuff?
| @@ -268,10 +268,10 @@ | |||
| "version_added": true | |||
| }, | |||
| "safari": { | |||
| "version_added": true | |||
| "version_added": false | |||
There was a problem hiding this comment.
So I realize this patch is scoped to just updating the Safari data — but as far as I can see, all the data we have for this feature is wrong. There’s no HTMLAnchorElement.prototype.media in Blink or Gecko either — which is unsurprising since https://html.spec.whatwg.org/multipage/text-level-semantics.html#htmlanchorelement doesn’t actually define a media member for HTMLAnchorElement, not even as an obsolete member in https://html.spec.whatwg.org/multipage/obsolete.html#HTMLAnchorElement-partial. And there is actually no https://developer.mozilla.org/docs/Web/API/HTMLAnchorElement/media article.
So it seems to me we should just remove this entire feature.
There was a problem hiding this comment.
Agree that this should be removed. I think it's worth applying this change first, however, since (eventually) the feature would then show up in a listing of features not supported anywhere. Per our data it's still in Edge and Firefox, and we should check that before removal.
@vinyldarkscratch can you split this out into a separate PR?
There was a problem hiding this comment.
Just to confirm: by splitting out, are you talking about this change, or removing the feature?
There was a problem hiding this comment.
I mean doing the removal separately. Setting this to false here is fine.
There was a problem hiding this comment.
Gotcha, thanks for the clarification! Definitely, I'll get a PR for that.
| @@ -220,10 +220,10 @@ | |||
| "version_added": true | |||
| }, | |||
| "safari": { | |||
| "version_added": true | |||
| "version_added": false | |||
There was a problem hiding this comment.
There’s no HTMLAreaElement.prototype.hreflang in Blink or Gecko either — and https://html.spec.whatwg.org/multipage/image-maps.html#htmlareaelement doesn’t actually define a hreflang member for HTMLAreaElement, not even as an obsolete member in https://html.spec.whatwg.org/multipage/obsolete.html#HTMLAreaElement-partiall. And there is actually no https://developer.mozilla.org/en-US/docs/Web/API/HTMLAreaElement/hreflang article.
So it seems to me we should just remove this entire feature.
There was a problem hiding this comment.
The data still says Edge and Firefox support this. Not sure if that's right, though.
This is a capitalization typo: mdn/browser-compat-data#7402 (comment)
* Remove formEncType custom IDL This is a capitalization typo: mdn/browser-compat-data#7402 (comment) * Remove lowSrc custom IDL
|
I've reviewed again. The changes to api.HTMLImageElement.lowSrc and api.HTMLInputElement.formEncType were not correctly reverted, resulting in a null→true change remaining. I'll set them back to null and merge this PR. |
…Element.formEncType
This was implemented without a flag, and with the reflected IDL attribute, in WebKit 532.6: WebKit/WebKit@c205c4e https://github.com/WebKit/WebKit/blob/c205c4e7d40da38c5009df548c2316a358a4fca7/WebCore/Configurations/Version.xcconfig That maps to Chrome 5 + Safari 5 + iOS 4, which also matches exactly earlier updates based on collector test results: mdn#7243 mdn#7290 mdn#7402 The handled attribute tokens were "allow-same-origin", "allow-forms", and "allow-scripts", so update that data to match.
This was implemented without a flag, and with the reflected IDL attribute, in WebKit 532.6: WebKit/WebKit@c205c4e https://github.com/WebKit/WebKit/blob/c205c4e7d40da38c5009df548c2316a358a4fca7/WebCore/Configurations/Version.xcconfig That maps to Chrome 5 + Safari 5 + iOS 4, which also matches exactly earlier updates based on collector test results: mdn#7243 mdn#7290 mdn#7402 The handled attribute tokens were "allow-same-origin", "allow-forms", and "allow-scripts", so update that data to match. The "allow-popups" entry was bumped to pass the lint.
This was implemented without a flag, and with the reflected IDL attribute, in WebKit 532.6: WebKit/WebKit@c205c4e https://github.com/WebKit/WebKit/blob/c205c4e7d40da38c5009df548c2316a358a4fca7/WebCore/Configurations/Version.xcconfig That maps to Chrome 5 + Safari 5 + iOS 4, which also matches exactly earlier updates based on collector test results: #7243 #7290 #7402 The handled attribute tokens were "allow-same-origin", "allow-forms", and "allow-scripts", so update that data to match. The "allow-popups" entry was bumped to pass the lint.
This PR uses Safari (4-14) and Safari iOS (3-14) results from the mdn-bcd-collector project to correct the data for the HTML element APIs. All results were double-checked for accuracy.