Undeprecate the unload event

[ddbeck]

Summary

unload is marked as deprecated, but it doesn't satisfy the guidelines.

Test results and supporting details

• Mark unload as deprecated #21480 set the value in the first place, on the basis of Chrome's deprecation, not the spec.
• Clarify cases where we would fire the "unload" event, or maybe deprecate it? whatwg/html#6026 is still open, where the question of unload deprecation does not have consensus, much less specification text.
• Permission-Policy: unload mozilla/standards-positions#691 (comment) seems to indicate that the effort to deprecate this event has been abandoned (closed without a position being given by Mozilla or WebKit).

Given the evidence here, I think it's much too early to mark this event deprecated.

Related issues

n/a

[github-actions]

Tip: Review these changes grouped by change (recommended for most PRs), or grouped by feature (for large PRs).

[Zakariathr22]

Please consider keeping the unload event undeprecated for as long as possible. Many existing applications still rely on it, and maintaining its availability helps avoid unnecessary breakage in real-world use cases.

[ddbeck]

@Zakariathr22 BCD does not decide whether a feature is deprecated (discouraged from use) or slated from removal from browsers. That work happens through the responsible standards body (WHATWG, in this case) or browser vendors, respectively.

[Zakariathr22]

@ddbeck Thanks for replying

My intent was mainly to highlight real-world dependency on the unload event and the impact that deprecation or removal can have on existing applications, so that this feedback can be considered in the appropriate standards or vendor discussions.

Could you please advise on the best channel or forum where this concern should be raised so it can be reviewed by the responsible standards body or browser vendors?

[ddbeck]

@Zakariathr22 I don't think anyone has proposed removal from the platform and breaking pages that use the unload event. Deprecation would only advise developers against using it, to help them consider other options instead (e.g., pagehide). That said, for this specific case, I believe whatwg/html#6026 would be the place to discuss with the spec editors.