humanum.arts.cuhk.edu.hk - Audio does not work
Categories
(Web Compatibility :: Site Reports, defect, P1)
Tracking
(Webcompat Priority:P3, Webcompat Score:4, firefox128 wontfix, firefox129 wontfix, firefox130 fix-optional)
Tracking | Status | |
---|---|---|
firefox128 | --- | wontfix |
firefox129 | --- | wontfix |
firefox130 | --- | fix-optional |
People
(Reporter: ctanase, Unassigned)
References
()
Details
(Keywords: regression, webcompat:platform-bug, webcompat:site-report, Whiteboard: [webcompat-source:web-bugs])
User Story
platform:windows,mac,linux,android impact:feature-broken configuration:general affects:few branch:release diagnosis-team:media user-impact-score:60
Attachments
(1 file)
535.05 KB,
image/png
|
Details |
Environment:
Operating system: Linux/Windwos 10
Firefox version: Firefox 128.0/130
Steps to reproduce:
- Go to https://humanum.arts.cuhk.edu.hk/Lexis/lexi-can/search.php?q=%F6i
- Click on the audio icon.
- Observe the behavior.
Expected Behavior:
The audio opens in a new tab and can be heard.
Actual Behavior:
A new tab is opened but it is blank and no audio plays.
Notes:
- Reproduces regardless of the status of ETP
- Reproduces in Firefox Nightly, and Firefox Release
- Does not reproduce in Chrome
Created from https://github.com/webcompat/web-bugs/issues/139784
Reporter | ||
Comment 1•1 year ago
|
||
Reporter | ||
Updated•1 year ago
|
Updated•1 year ago
|
Comment 2•1 year ago
|
||
https://humanum.arts.cuhk.edu.hk/Lexis/lexi-can/sound/oi3.wav gets a NS_BINDING_ABORTED in the network tab.
Randal, do you know what NS_BINDING_ABORTED means? It would be nice if dev tools gave more information on what might be going wrong here.
Comment 3•1 year ago
•
|
||
I did a mozregression and got it down to one day:
https://hg.mozilla.org/mozilla-central/pushloghtml?fromchange=61a5e77067ce509969ee4ea18bb849ce775ef615&tochange=72d7558735ea67f7ff149d69ec671f7b80dbb999
for this URL: https://humanum.arts.cuhk.edu.hk/Lexis/lexi-can/sound.php?s=oi3
Comment 4•1 year ago
|
||
I have a profile with logging showing what's cancelling that resource: https://share.firefox.dev/4dBZGki
Clicking on the Closing channel
marker shows the following stack trace:
nsObjectLoadingContent::CloseChannel()/home/icecold/mozilla-central/dom/base/nsObjectLoadingContent.cpp
nsObjectLoadingContent::LoadObject(bool, bool, nsIRequest*)/home/icecold/mozilla-central/dom/base/nsObjectLoadingContent.cpp
nsObjectLoadingContent::OnStartRequest(nsIRequest*)/home/icecold/mozilla-central/dom/base/nsObjectLoadingContent.cpp
nsObjectLoadingContent::OnStartRequest
mozilla::net::HttpChannelChild::DoOnStartRequest(nsIRequest*)/home/icecold/mozilla-central/netwerk/protocol/http/HttpChannelChild.cpp
mozilla::net::HttpChannelChild::OnStartRequest(mozilla::net::nsHttpResponseHead const&, bool const&, mozilla::net::nsHttpHeaderArray const&, mozilla::net::HttpChannelOnStartRequestArgs const&)/home/icecold/mozilla-central/netwerk/protocol/http/HttpChannelChild.cpp
So the cancel is coming from here
I haven't figured out the regressing commit.
Comment 5•1 year ago
|
||
Looking at the code, it could be related to bug 1595491 - but I'm not 100% sure, that's the cause.
Comment 6•1 year ago
|
||
I also looked at the page. The web page uses <object>
/ <embed>
to play sound. (So, this is unrelated to HTTPS-First / HTTPS auto upgrading).
Comment 7•1 year ago
|
||
From discussions with @valentin, he thinks this is ORB related. Note however that it still happens (doesn't work) if I disable ORB. It still could be ORB related, of course. He also tried turning off RCWN (race cache with network), and it still happens. (there were 2 RCWN patches in that range)
Updated•1 year ago
|
Updated•8 months ago
|
Updated•8 months ago
|
Comment 8•8 months ago
|
||
Given this is perhaps DOM or perhaps <object>/<embed>, and the cancel is coming from Dom code, moving diagnosis to dom
Comment 9•8 months ago
|
||
That content is audio/x-wav
, looks like we don't handle that in <object>.
Do you know if this is expected Andreas?
Comment 10•8 months ago
|
||
I guess we did not handle .wav files in embed/object because we support only linear PCM codecs and we tried to give plugins a chance to handle these files. But now NPAPI plugins are dead and this causes a real web-compat problem now. Should we revisit the decision?
Comment 11•8 months ago
|
||
Thanks Masatoshi!
I am adjusting the user story moving this to media team.
Comment 12•8 months ago
|
||
Fixing this could be as simple as adjusting https://searchfox.org/mozilla-central/source/dom/media/DecoderTraits.cpp#241-266 in this case to allow WAVE files. For historical reasons all files play but not WAVE, but I haven't tested this.
![]() |
||
Updated•8 months ago
|
Comment 13•8 months ago
|
||
/me believes webcompat:platform-bug
should be added
Description
•