@notme, could it be ready to test tomorrow evening?
If you are speaking about jitsi (that is how I understood Maria in another thread), be aware that besides some server configuration, a single non-chromium-browser-user can potentially ruin the experience (also, for everybody else!) at the moment. I dont want to go into details here.
Point is: if you want to have a pleasant experience and rule out some weirdness, have your users use chromium (not safari, not edge, not firefox, not opera). Its a tough hit, because Firefox is kind of the only contender left in the browser-market; the jitsi-Team is working on it - but compatibility will not arrive tomorrow - I guess it will take weeks or months.
Chrome of course would work as well (Chromium is Google Chrome minus some of the tell-google-exactly-what-i-do-functionality).
Thank you for that insight
We should also test that info about FF. Easy enough.
The server will be ready for testing 14:00 today. Other groups are also testing at the same time and I’ll be monitoring stats.
meet.complete.fi/crowdtesting is the common test room. Feel free to create /anyroom you like.
I’ll probably remove the server in the evening. There is a chance for another test on Friday. Let me know when you are testing and with how may people so I can inform everyone how that effects performance. (I’ll join your room if you like)
We can test JWT later.
We had a test group of 6 people. About 1.5GB data was transferred over about 1h. The bandwidth hovered around 10mb for the 5-6 users during this time. Everything worked surprisingly well during the test. Would be fun to see how far we can scale this. I suspect it will be much more trouble for individual users after the bandwidth goes above 10mb.
I’ll destroy the server this evening. I can set it up on Friday again if you like. Or you can suggest a better time for a test day. I missed your earlier timing question.
If you want to test JWT, let me know before I do setup, as I need to prep the configuration.
You can configure the preferred and max resolution that the videobridge will relay and thus influence the bandwidth (and clients CPU) used.
The FF issues are real. There is kind of a meta-issue at github: https://github.com/jitsi/jitsi-meet/issues/4758 (but there are many more). Testing is difficult because not necessarily reproducible. Also, fixes are not trivial, but sure enough will come.
By “non-chromium-browser”, do you mean actual Chromium, or any browser based on Chromium?
Asking because I thought both Edge and Opera were Chromium-based; and so it’s Brave, which I use as an alternative to Firefox sometimes and was thinking of using it for Jitsi Meet, so shall I assume it would have issues?
I meant Chromium-based. Jitsi makes big progress on improving the situation though. It should already work smoother and more stable with Firefox in latest versions. As far as I understand, to optimally use some features of WebRTC/simulcast/EEC (?) Firefox itself needs to be fixed though. There is the mega-issue on the github for the technically interested.