1627649 - Unable to use WebSocket in Thunderbird

Status: RESOLVED FIXED

(Thunderbird :: General, defect)

Description

[Booteille]

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:76.0) Gecko/20100101 Firefox/76.0

Steps to reproduce:

First, the following was tried on Thunderbird 75.0b3 and 68.6.0.

I tried to use WebSockets in a very simple WebExtension.

The background.js only have the following code:

console.log("a");
let ws = new WebSocket("wss://echo.websocket.org/");
console.log("b");

---

I also tried to open a WebSocket in the debugging console directly by just typing the following code:

let ws = new WebSocket("wss://echo.websocket.org");

I tried the same on Firefox 76.0a1 (2020-04-05) (64-bit). Both, background.js and console tries worked.

Actual results:

As a result, in the debugging console, I only got this result:

a background.js:1:9
: background.js:2
    <anonymous> moz-extension://aa11cfda-f35c-4d6d-ae9a-1f3595103d4f/background.js:2

As you can see, the error returned by the console is totally empty.
When trying to catch the Exception inside a try-catch block, the Exception returned is totally empty.

---

When entering the code directly in the console, I get another error:

undefined
NS_ERROR_FAILURE: Component returned failure code: 0x80004005 (NS_ERROR_FAILURE) [nsIInterfaceRequestor.getInterface]

Expected results:

I expect to get a fully working WebSocket, as it is the case with Firefox.

Comment 1

[Magnus Melin [:mkmelin]]

Actually, this is a general fault - websocket is not working in Thunderbird.

STR: load https://www.websocket.org/echo.html into the start page. Try pressing "Connect" - nothing happens.

Comment 2

[Magnus Melin [:mkmelin]]

Attachment: bug1627649_tb_websocket.patch

Sent to try now: https://treeherder.mozilla.org/#/jobs?repo=try-comm-central&revision=29f33dc9e3ec11df1d84eaa0809d439fea582c70

Comment 3

[Firefox Bug Husbandry Bot]

Because this bug's Severity has not been changed from the default since it was filed, and it's Priority is -- (Backlog,) indicating it has has not been previously triaged, the bug's Severity is being updated to -- (default, untriaged.)

Comment 4

[Ben Campbell]

Comment on attachment 9147424 [details] [diff] [review]
bug1627649_tb_websocket.patch

Review of attachment 9147424 [details] [diff] [review]:
-----------------------------------------------------------------

LGTM

Comment 5

[Magnus Melin [:mkmelin]]

Attachment: bug1627649_tb_websocket.patch

The earlier patch was not 100% right. I added some clarifying comments in the code

Comment 6

[Ben Campbell]

Comment on attachment 9148084 [details] [diff] [review]
bug1627649_tb_websocket.patch

Review of attachment 9148084 [details] [diff] [review]:
-----------------------------------------------------------------

I don't pretend to follow the nuances of contentpolicy (regarding your data:// changes), but mechanically it all looks fine to me.

::: mail/test/browser/content-policy/browser_generalContentPolicy.js
@@ +529,5 @@
>    assert_nothing_selected();
>  
>    for (let i = 0; i < TESTS.length; ++i) {
>      // Check for denied in mail
> +    dump("Doing test: " + TESTS[i].description + " ...\n");

Is there any policy/standard that for logging output for mochitests that should be followed here?

Comment 7

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/comm-central/rev/b5cb669844cd
Unable to use WebSocket in Thunderbird. r=benc DONTBUILD

Comment 8

[Jorg K (CEST = GMT+2)]

Comment on attachment 9148084 [details] [diff] [review]
bug1627649_tb_websocket.patch

[Approval Request Comment]
Regression caused by (bug #): No regression, never worked.
User impact if declined: Would be nice if add-ons could use this asap.
Risk to taking this patch (and alternatives if risky): Not risky to existing functionality, just allowing some additional URL schemes "ws" and "wss".

Comment 9

[Magnus Melin [:mkmelin]]

(In reply to Ben Campbell from comment #6)

Is there any policy/standard that for logging output for mochitests that should be followed here?

I think there's some other way, but I can't seem to find it atm. Subtests of the type used here are hard to debug without output since it's the same "checks" so line numbers do not help due to changing data.

Comment 10

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/comm-central/rev/936dbd8706b8
followup - use info() instead of dump() for the mochitest debugging. rs=me DONTBUILD

Comment 11

[Magnus Melin [:mkmelin]]

Are you able to verify it's working in the webextension you have with current nightly builds?

Comment 12

[Booteille]

(In reply to Magnus Melin [:mkmelin] from comment #11)

Are you able to verify it's working in the webextension you have with current nightly builds?

It seems to work properly! Thanks!

Comment 13

[Wayne Mery (:wsmwk)]

Comment on attachment 9148084 [details] [diff] [review]
bug1627649_tb_websocket.patch

Approved for beta, per comment 12 and Jorg's assessment.

Thanks

Comment 14

[Geoff Lankow (:darktrojan)]

77.0b3:
https://hg.mozilla.org/releases/comm-beta/rev/f3ae070191f10b7281025a43d69285bf769042d6 (both revisions rolled together)

Comment 15

[Wayne Mery (:wsmwk)]

Comment on attachment 9148084 [details] [diff] [review]
bug1627649_tb_websocket.patch

[Triage Comment]
Approved for ESR. Risks, etc per comment 8

Comment 16

[Rob Lemley [:rjl]]

Thunderbird 68.8.1:
https://hg.mozilla.org/releases/comm-esr68/rev/face673d2586