Hacker News new | ask | show | jobs
by lawl 1744 days ago
> Is that significantly different from making a new channel and labeling it "voice channel"? To me it doesn't seem to be, but I haven't tried it.

Yes it is. For matrix currently you'd need to have the discipline to always leave and join that text channel too when you join/leave a room. And another use case is having multiple rooms. Consider the scenario of four friends sitting in a channel and playing 2v2 games and they randomize team every round. With discord/mumble/teamspeak etc. this is extremely simple, you just switch channels with one (double)click.

Another one is that sometimes we have multiple sub-groups of our friend group sitting on the teamspeak server (e.g.) playing different games in different voice channels at the same time. Because you can see all the channels and who's currently actively talking in them, sometimes we just pop over to the other guys and ask if they're up for a round of game X. Sometimes they are, sometimes they're not and we just pop back into our channel.

And there's a lot more things that I personally consider absolutely mandatory, like push to talk, voice activation, individual volumes etc.

I honestly believe anyone who's doing VOIP seriously needs to look at how gaming VOIP has always done it with teamspeak/mumble etc. Discord has successfully copied this too, and imo all other VOIP is so far behind it's not even funny.

1 comments

I believe mumble/teamspeak have a similar model to Matrix though, in that every room is both a voice channel and a text channel?

Maybe there should be another feature request to auto join voice comms when entering a channel?

The text channels in mumble/teamspeak are pretty much irrelevant because matrix already has good text channels, the IRC part. The text channels there are just to post the odd link or something in mumble/ts.

There's a reason discord is winning, and i'm saying that as a matrix+mumble/ts user. They integrated modern IRC and modern (worse but free)/teamspeak nicely.

Honestly, I don't really want to file feature requests about this because if it's not as good as mumble i'm not going to use it for voice.

And making it as good as mumble is not trivial, because browser tech sucks. WebRTC is entirely inadequate imo. (I have messed around with POCs myself)