LiveKit Community

_if (isWeb()) {_
        _this.acquireAudioContext();_
      _}_

ok - so I spoke too soon - everything is not fine. I now don't get that error, but I do get the following:

disconnected from room
could not connect: could not establish signal connection Ai [Error]: could not establish signal connection
    at Promise.finally.connectFuture (C:\Users\mrbre\source\repos\IMP\demo_repo\WEBRtcSamples\IMServerSlotBase\node_modules\livekit-client\src\room\Room.ts:325:16)
    at processTicksAndRejections (node:internal/process/task_queues:96:5) {
  code: 1

I guess I am hoping that someone in here can tell me if 1. The client can run on a server environment 2. How to achieve that or 3. No - but here is an option to achieve this etc..

Thanks again in advance for any help on this. Cheers

I'm not sure how to move them to the other channel? I'm assuming by thread them, you want me to create a separate post and then only reply directly to that one?

I'm sorry - I don't use Slack for any purpose etc so I'm not up to speed on the correct etiquette, but of course I am will to learn !

hey <@U03UV4JFTK2> — no worries at all or apologies required. Just trying to keep things cleaner and easier for folks to parse.

appreciate you considering my ask! :pray:

I am trying to use the `setEnabled` in `remoteTrackPublication` , while it seems too heavy to do a mute?

Only mute the guy for myself, not the others in the room.

hi <@U03REEFGQQL>
`setEnabled` works perfectly for this, as it will also ensure that the server stops sending down data for this track (saves bandwidth).
If you would like to only control the audio track, there is a `setVolume` function on the participant, but this lacks a concept of `muted` state, so you’d have to check against volume being equal to 0

Hi, <@U02EU2ZPTQU> Thanks for your reply. `setEnabled` works perfectly now, while won't it won't trigger any events. Which property can I got the remoteParticipant's current audio/video status?

you can check for `publication.IsEnabled`

I am trying to use `IsMicEnabled` to determine whether should I setEnabld to true of false

What's the actual mean of `participant.IsMicrophoneEnabld` ?

this checks whether the remote participant has enabled/muted their microphone themselves

Another Q:
Is there any functions I can mute all room locally?

no worries at all!
no, currently there’s no such function built in. I imagine it depends a lot on the application logic what should happen after initially muting all participants locally. Like
• what happens to newly joined participants? are they also muted?
• is it possible to manually unmute one of the participants or does a `mutedAll` property take precedence


Ok, got that. That should be considered in applicatoin.

Hey <@U04078BB5QQ> can you elaborate on what your gstreamer pipeline looks like? are you processing both sinks within the same Gstreamer process? or is it separate?

I believe if they are the same, then the pipeline should wait for buffers to reach the end before pushing new frames;  that should serve as a method of synchronization.

though there will probably be at least a couple of frames of differences. I'm not sure how to make Gstreamer behave differently.

carbon.png

<@U01KZ61GV5Z> This is my current pipeline

I see, I'm not quite sure how to accomplish what you are looking for. I've only used Gstreamer lightly in the past.

From the livekit side, it should be possible to accept multiple streams. synchronization is the main challenge. You can feed each sample/frame to livekit via <https://pkg.go.dev/github.com/pion/webrtc/v3/pkg/media#pkg-types|media.Sample> and set its Timestamp. webrtc clients would use that to attempt to synchronize playback

<https://auth0.com/docs/secure/tokens/json-web-tokens|JWT tokens> gives us a pretty well battle tested security model that doesn't rely on having to call back to another authentication server.

It's <https://developers.google.com/identity/protocols/oauth2/service-account#jwt-auth|the same model> Google is opting for privileged access to user data.

&gt;  The token can be used in any way by the user.
I don't believe the tokens should be issued to be all-permissive. They should include the minimum set of claims to accomplish what the user needs to (i.e. connect to room `abcd` with identity `1234` ).

You should also set a short TTL on the token so it's effectively useless after a number of seconds. With these two measures in place, I'm not seeing the attack vector here.

Thanks <@U01KZ61GV5Z>. You may have implemented certain restrictions during the development of your service's client, such as the number of webcams displayed, the disablement of screenshare, and so on. However, if the same token is used in other clients, those restrictions can be easily avoided. I'm not concerned with the security model because I know it's adequate. I was only mentioning the usage case. I was considering validating the token from an external server because we can set different parameters in metadata and check them before passing. Of course we can check different way before generating token but in case further checking..

ok, I think you are really asking about finer grained permission controls (i.e. max video tracks/etc). That could be something that's an extension of the current JWT grant and enforced within the SFU.

I can provide another example. We have a user block option in plugNmeet. For various reasons, a moderator may refuse to allow a user to join a session but that user is not prohibited from generating a token. If the user is clever enough, he or she can use a different client to join the session. When we join a session in plugNmeet, we extract the token to see if the user is blocked.
If they are blocked, do not allow them to join. However, we cannot if the user uses a different client. We could simply stop users from joining if we had the ability to check the token from Livekit.

Shouldn't the user not be allowed to generate a token for the room they've been banned from? They can still be allowed to join other rooms

Yes, from application end we can do that. Just in case having another option. At present we're only checking during join/verify token.

Anyway <@U01KZ61GV5Z> thank you for creating such wonderful software. We're thrilled to be able to use it. I was sharing my idea, which is simply to extend the features without making too many changes.

thank you for the kind words and support Jibon. we appreciate you always being helpful in the community

Yes, you are correct. That is how Livekit is working now. Room will active when any user join. After that you'll be starting receive webhook events.

image.png

I debug liveserver find this error , How I solve this question?
envirment:
1, liveserver.
2, web test.

`receiveLwsProtocBufMessage(): unpackSignalResponse-&gt;trickle-&gt;target = IGNAL_TARGET__PUBLISHER unpackSignalResponse-&gt;trickle-&gt;candidateinit = {"candidate":"candidate:64545876 1 tcp 2124414975 113.90.236.18 7881 typ host tcptype passive","sdpMid":"","sdpMLineIndex":0,"usernameFragment":null}`

Hello everyone, Here is my new article, A Blueprint for Session Based Applications in Laravel,  where I expand on the use of LiveKit within a blueprint for creator economy, collaboration, and remote work applications. Based on previous work that I published here, and onward and forward to future additional WebRTC applications. I hope you enjoy it.  <https://yoramkornatzky.com/post/a-blueprint-for-session-based-applications-in-laravel>

could you be more specific? what are you trying to accomplish? what SDK &amp; platform?

There are new hardware APIs here,
<https://github.com/livekit/client-sdk-flutter/blob/main/lib/src/hardware/hardware.dart>

you can call API like this
```var audioOutpus = await Hardware.intance.audioOutputs();
Hardware.instanse.selectAudioOutput(audioOutput.first);```
or just call
```Hardware.instance.setSpeakerphoneOn(yesno);```

you guy's are legend
I appreciate your work from bottom of my heart.
Love your work
keep it up !

*error could not establish signal connection*

my access token its not working, please how do I resolve this?

There are a few things that you can do to drastically improve the audio quality, but there is a bit of a limit at the moment.

In particular, you want to adjust the AudioCaptureOptions to disable a bunch of the things that are good for voice, but bad for music (like echo cancelation and noise suppression). You'll also want to change the TrackPublishOptions that you are using to increase the bitrate. At the moment, I'm not hearing much improvement beyond 96-128kbps.

Take a look at this ticket: <https://github.com/livekit/livekit/issues/970>

Step 2 in the repro shows the settings that can be tuned.

<@U02C10ZEP8R> Thanks for answering !
Exactly I wanted to know if voice processing could be disabled...
And what you bring up on github about stereo support is also a concern for sure :wink:
Your github post contains a lot of answers, tks again!

Has someone installed monitoring for livekit?

I saw these pages but I can't install it.

hi Jake, you can deploy a copy on any DO droplet: <http://docs.livekit.io/deploy/vm/>

LK Cloud will also be available as an official option.

May be the following will fix the issue
```  const shouldDisconnect = useRef(false);
  useEffect(() =&gt; {
    roomState.connect(url, token, connectOptions).then((room) =&gt; {
      if (!room) {
        return;
      }
      if (onConnected &amp;&amp; room.state === ConnectionState.Connected) {
        onConnected(room);
      }
    });

    return () =&gt; {
      if (
        roomState.room?.state !== ConnectionState.Disconnected &amp;&amp;
        shouldDisconnect.current
      ) {
        roomState.room?.disconnect();
      } else {
        shouldDisconnect.current = true;
      }
    };
  }, []);```

Hi <@U02PDJ8R2A1>,
I think you have also commented about this on github, right?
It would be great if you could try upgrading to v1.3.1 for client-sdk-js and if the error persists open a new issue on github with some reproduction steps!

Hi <@U02EU2ZPTQU>
Yeah it's me and sorry for the double comment.
I just replied on github
Thanks

I'm still getting this issue on react, how did you resolve this? :)

hi <@U0440542S83> could you file an issue on github with reproduction steps?
it shouldn’t happen in the latest components version any more (paired with an up to date livekit-client version)

Hi <@U040Y8XJNR1> do you mean to publish out to live streaming services like Twitch/YT via RTMP, or egress HLS directly to viewers? In either case, yes, LiveKit can do both. :blush: 

Is there any repository that implemented both egress and RTMP.?

Our egress system handles both, or rather, RTMP out is part of egress.

<https://github.com/livekit/egress|https://github.com/livekit/egress>

Docs here (currently under revision), but should get you started:
<https://docs.livekit.io/guides/egress/|https://docs.livekit.io/guides/egress/>