Voice Support? #121

This-guy-breaks-shit commented 2025年10月15日 15:33:37 +02:00 (Migrated from github.com)

Hi All,

Would it be within the scope of this bot to add voice bridge support?

While I currently lack any real understanding of how lightning currently functions, I'm aware of some potential obstacles:

• It may not be possible to simply plop a whole new major component on top of the existing system without some major overhauls
• Bandwidth and processing requirements could pose too much demand when hosting lightning for public use (though I guess you could just disable that component for all but selfhost/paid clients via config)

I'm fully expecting that this feature request is out of scope for the project, or at best a "far down the road" kind of thing. Still, the only stupid questions are the ones never asked.

Thanks for reading!

Hi All, Would it be within the scope of this bot to add voice bridge support? While I currently lack any real understanding of how lightning currently functions, I'm aware of some potential obstacles: - It may not be possible to simply plop a whole new major component on top of the existing system without some major overhauls - Bandwidth and processing requirements could pose too much demand when hosting lightning for public use (though I guess you could just disable that component for all but selfhost/paid clients via config) I'm fully expecting that this feature request is out of scope for the project, or at best a "far down the road" kind of thing. Still, the only stupid questions are the ones never asked. Thanks for reading!

jersey commented 2025年10月15日 21:14:56 +02:00

hi! voice support would be a pretty big thing to implement, and the resources required would require better hardware than the bot currently requires, but this is something i'm looking to implement in the eventual future. as for how this could be done on my end,

• discordgo has the dgvoice package which should simplify things
• guilded will be dead soon, and they never had voice support
• stoat's new voice support uses livekit, and that seems to have a go package
• i don't think this is possible with telegram
• element call seems horrendous but it could be implemented

as for when this could be done, this will be an eventually type of thing, but it'd be pretty cool

hi! voice support would be a pretty big thing to implement, and the resources required would require better hardware than the bot currently requires, but this is something i'm looking to implement in the eventual future. as for how this could be done on my end, - discordgo has the [dgvoice](https://github.com/bwmarrin/dgvoice) package which should simplify things - guilded will be dead soon, and they never had voice support - stoat's new voice support uses livekit, and that seems to have a [go package](https://github.com/livekit/server-sdk-go?tab=readme-ov-file#publishing-audio-from-pcm16-samples) - i don't think this is possible with telegram - element call seems horrendous but it could be implemented as for when this could be done, this will be an eventually type of thing, but it'd be pretty cool

This-guy-breaks-shit commented 2025年10月16日 13:35:43 +02:00 (Migrated from github.com)

Hi William, thanks for the fast reply!

This is actually quite a surprise, I wasn't expecting voice support to be possible at all let alone using existing packages (at least for some platforms). It's a shame that Telegram and Element are difficult/not possible, but supporting voice to begin with would be a huge leap.

I don't have boatloads of time on my hands and my Go experience is limited. That said, I do have experience with other C-based languages, so I might create a branch to have a crack at it in my own time (at least for Discord and Stoat). It's entirely likely that this is way out of my depth, but I'll let you know if something comes of it regardless.

Anyway, thanks again for your help. Hopefully I'll end up with something worth contributing.

Hi William, thanks for the fast reply! This is actually quite a surprise, I wasn't expecting voice support to be possible _at all_ let alone using existing packages (at least for some platforms). It's a shame that Telegram and Element are difficult/not possible, but supporting voice to begin with would be a huge leap. I don't have boatloads of time on my hands and my Go experience is limited. That said, I _do_ have experience with other C-based languages, so I might create a branch to have a crack at it in my own time (at least for Discord and Stoat). It's entirely likely that this is way out of my depth, but I'll let you know if something comes of it regardless. Anyway, thanks again for your help. Hopefully I'll end up with something worth contributing.

jersey commented 2025年10月20日 19:45:19 +02:00

totally agree that voice support would be a big leap, and i also lack boatloads of time (thanks, school!), so it's nice to see someone taking a stab at this! i'm not too worried about code quality for an initial implementation, since i can always clean it up (削除) instead of doing homework (削除ここまで).

also, element call shouldn't be as painful as i anticipated, since parts are based on livekit, with the usual random matrix insanity added on top

totally agree that voice support would be a big leap, and i also lack boatloads of time (thanks, school!), so it's nice to see someone taking a stab at this! i'm not too worried about code quality for an initial implementation, since i can always clean it up ~~instead of doing homework~~. also, element call shouldn't be as painful as i anticipated, since [parts are based on livekit](https://github.com/element-hq/element-call?tab=readme-ov-file#-overview), with the usual random matrix insanity added on top