Bug 168768
| Summary: | [XHR] Consider not canonicalizing known header names | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Anne van Kesteren <annevk> |
| Component: | DOM | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED CONFIGURATION CHANGED | ||
| Severity: | Normal | CC: | achristensen, ahmad.saleem792, ap, cdumez, rbuis, rniwa, youennf |
| Priority: | P2 | ||
| Version: | Safari Technology Preview | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Bug Depends on: | |||
| Bug Blocks: | 171031 | ||
Anne van Kesteren
Test: http://w3c-test.org/XMLHttpRequest/getallresponseheaders-cl.htm.
(I keep having the feelings I'm getting the component wrong, but there's really not that many components.)
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
youenn fablet
I am not sure the component really matters.
Prefixing the bug title with [XHR] might be good enough?
Alexey Proskuryakov
Is there some reason for this proposed change? Pretty sure all browsers used to agree, there were web compatibility reasons, and canonicalizing just makes sense.
Anne van Kesteren
I've found that Chrome and Edge don't agree with Firefox and Safari and the way Firefox does interning is problematic, so we're likely aligning with Chrome.
(Thanks for the component/summary advice.)
Ahmad Saleem
If I search for "Get All Response Headers" on WPT.FYI, Safari Technology Preview 155 passes all tests:
https://wpt.fyi/results/xhr?label=master&label=experimental&aligned&view=subtest&q=get%20all%20response%20headers
Do we need anything more here? Thanks!
Anne van Kesteren
Thanks Ahmad, looks like this got sorted at some point. (The `-cl.htm` got merged into the `.htm` file in https://github.com/web-platform-tests/wpt/commit/b32391a796893529fa425c45b9e8b96d58d68440 and then it was moved and rewritten some, but it kept testing the same thing.)