1251627 - Fix the exceptions XMLHttpRequest (XHR) send() throws to follow the spec more

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Boris Zbarsky [:bzbarsky]]

We should not be throwing NS_ERROR_FAILURE from DOM APIs.  Really, we should not.

Comment 1

[Boris Zbarsky [:bzbarsky]]

Attachment: Fix XMLHttpRequest.send() to follow the spec better in terms of the exceptions it throws

Comment 2

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

Comment on attachment 8724139 [details] [diff] [review]
Fix XMLHttpRequest.send() to follow the spec better in terms of the exceptions it throws

Review of attachment 8724139 [details] [diff] [review]:
-----------------------------------------------------------------

::: dom/workers/XMLHttpRequest.cpp
@@ +2282,5 @@
>  
> +  if (mStateData.mReadyState == 4) {
> +    // No one did anything to us while we fired abort events, so reset our state
> +    // to "unsent"
> +    mStateData.mReadyState = 0;

https://xhr.spec.whatwg.org/#the-abort%28%29-method says to do this unconditionally, no?

Comment 3

[Boris Zbarsky [:bzbarsky]]

> https://xhr.spec.whatwg.org/#the-abort%28%29-method says to do this unconditionally, no?

It does, but I think it's bogus.  I filed https://github.com/whatwg/xhr/issues/54 earlier today, with a testcase and an explanation of why I think the spec behavior in that testcase is bogus...

Comment 4

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/100ad61e04dc

Comment 5

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/100ad61e04dc