Mozilla Strikes Again! (Firefox 52 Breaks Sound System)
July 19, 2017 at 1:40 AM #5846
Just tried watching Youtube and it seemed the sound on my laptop had stopped working. Then I found that the sound was working fine in other programs. Turns out that as of version 52, Firefox will only work with PulseAudio.
At least 11 bug reports have been filed. The above one has over 100 comments. Reading it (while enjoying a bowl of popcorn) reminded me of this bug report involving Systemd that we were talking about a few weeks ago. Interestingly Systemd’s developer also made Pulseaudio.
Firefox is still my primary browser but those folks at Mozilla sure are working hard at narrowing their target audience.
August 5, 2017 at 12:37 PM #5882
- This topic was modified 1 month ago by Christian.
You know what’s almost worse than
systemd‘s commandeering the
debugargument? The Ubuntu (maybe Debian) boot scripts used to commandeer that also. All I wanted to do is pass a debug argument to the kernel, and all the sudden I’m getting dumped into a shell instead of completing the boot process.
Despite all the positive things Poettering has contributed to FOSS, his seeming arrogance at times is astounding to me, as exemplified by his intransigence on the
debugkernel option and
Overall, PA is pretty good, especially its ability to regulate the level of each application wanting sound. It’ll even boost a source beyond 100%, so something with really low audio can be more audible. It also is helpful for routing sound too, for example to distribute it to several output devices. It’s just really sad to see that something like FF won’t work properly without it. There are a number of solution pages here and there which recommend uninstalling pulseaudio in order to get something to work.
The problem I have been having with PulseAudio is how it is invoked in Xubuntu sessions (via Upstart). What this means is, if I kill PA, it gets restarted automatically. I needed it gone so that JACK can work, so that Ardour can work. What I ended up doing was something like this in /usr/bin/pulseaudio:
It’s just weird that I have to go to such lengths to get things to work.August 10, 2017 at 12:03 AM #5900
I think the recommended solution is to give direct control of the sound card to JACK, and configure PA to access it through JACK. I recall getting this to work briefly when trying to use Rosegarden. The Arch wiki seems to have some information on this. That might be where I read it… can’t recall.
Speaking of bug reports and Mozilla, just came across this one from a few years back (I recall reading it before) while setting up Gmail SMTP in Thunderbird. In this case, at least a few Mozilla devs argued for doing the right thing (see Ben Bucksch’s comments), but were overwhelmed by those who wanted to enable Google to keep being evil.August 10, 2017 at 2:19 PM #5901
Google isn’t necessarily evil, they would just like a way to maintain multifactor auth. Since ordinarily IMAP has no intrinsic way of soliciting the other code, the weaker way of Google issuing an “app password” must be used. I don’t know of any way to restrict what that additional password is allowed to do, but the idea is that OAuth2 only allows specific uses of the obtained token. But I understand their reticence to implement something like that. They don’t want to implement something Google specific (or specific to anyone for that matter), they want to stick to published, widely accepted standards, such as standards track RFCs.
You must be logged in to reply to this topic.