On 17/12/17 00:56, Narcis Garcia wrote: > I like this proposal, if not interferes with any other feature, > performance or license.
For clarification: 1) The feature it adds (a new audio backend) does not interfere with any other features of the browser, except for one caveat; in the case where the user is running two different audio servers simultaneously, JACK and PulseAudio. In this case, IceCat would always connect to JACK with no possible way to select PulseAudio instead due to the hardcoded priority of JACK over PulseAudio. This little problem was solved in version 55 of firefox because I added a runtime about:config preference to select any audio backend manually. I don't see this as a big problem for IceCat 52.3.0 because anyone who runs JACK would probably not be using PulseAudio at the same time. Even if they did run both, I assume they prefer to have the audio routing through JACK if they were running jackd. In conclusion, I think it's better to have the option of JACK with this small caveat than not have it at all. 2) The performance of the browser is the same, but probably less CPU usage when JACK backend is in use instead of PulseAudio although I haven't measured it. Not to mention the new ability to route audio anywhere to and from the browser. 3) The licence is the same because the code is already part of Firefox upstream, it's just one compile time flag to enable it. Thanks, Damien -- http://gnuzilla.gnu.org
