LiveKit Community

hey <@U040YGD65EC>, checkout this example app: <https://github.com/livekit/livekit-react/tree/master/example>

<@U01KZ61GV5Z> can you please give me some information about the URL and the token if i want to run the app in my local server

take a look at our guides @ <http://docs.livekit.io|docs.livekit.io>

Screen Shot 2022-08-30 at 8.21.39 AM.png

is it normal to have permission denied in that place or am missing something

Hey Ricky, that's not expected. Which platform are you building for? and what's the SDK version?

Hello david.
Really sorry for the late reply.
We send the metadata to room using livekit server sdk, and the consumer is android jet compose sdk :bow:

Hi ben. I would recommend benchmarking it yourself with the provided utility. what ultimately matters is how much work each server is doing. That's highly dependent on what you are doing within those rooms. resource consumption is drastically different between a 1on1 room, vs a 100 participant room

hi @david, thank you for you response. i was doing just that but was able to do it for audio only, i'm having this issue using the CLI: <https://github.com/livekit/livekit-cli/issues/84> and not sure how to get around it.

Hi Adri, for 1. can you elaborate on "audio only works for one of us" ? is one person not able to hear the other? can you reproduce this behavior with our demo app?

2. it's possible a STUN host isn't configured when this is set. What happens when you removed the custom TURN server from config?

Hi David, thank you for helping me.
1. We configured the demo and server to work only in private networks, without STUN/TURN (although in the config file it says that by default LiveKit uses the Google STUN servers). The problem we had was that the audio most of the times don't work for 1 of us (video works always). For example I join the room and when my friend joins I can't hear him but he can hear me. If we join in inverse order we have the same results. On the other hand sometimes we can hear each other.
2. With the configuration without STUN/TURN we can talk to each other and watch the video using the demo app <https://example.livekit.io/>

```port: 7880
rtc:
    tcp_port: 7881
    port_range_start: 50000
    port_range_end: 50200
    use_external_ip: true
keys:
    &lt;API_KEY&gt;: &lt;API_SECRET&gt;
logging:
    json: false
    level: info```

Where can I find in the Unity demo what Audio Device is using to get the audio track?

There is a selector for the Webcam but not for the Microphone

when you try the demo without your modifications, are you able to reproduce the issue with one user not able to hear another as well?

We did the modifications for easy to use and testing

The video always works but the audio sometimes we can hear each other and sometimes just one of us

We've also removed the algorithm to set the volume based on the distance to test it better (we tried with and without)

Hi Adri, it seems there is an issue with the Unity Demo if everything is working correctly on <https://example.livekit.io/>
I’ll update the client-sdk-js version of the Unity demo to see if that helps.
Are you able to reproduce the same issue on <https://unity.livekit.io> ?

Can it be related to the wrong microphone being used ?

We are having the same issue with the <https://unity.livekit.io>

What are the steps to reproduce it? I can’t reproduce the issue

<https://example.livekit.io/> works perfect

Yes :confused:
I think this is related to the spatial audio, auto-subscribe is enabled

We removed the spatial audio algorithm and we have the same issue

I made a fork disabling the spatial audio and susbcribing the audio tracks manually here <https://github.com/adcimon/livekit-unity-demo>

<https://github.com/adcimon/livekit-unity-demo/blob/main/unity-project/Assets/Scripts/GameManager.cs#L133>

Which browser do you use and what is your livekit version ?

<https://github.com/adcimon/livekit-unity-demo/blob/main/unity-project/Assets/Scripts/JoinHandler.cs#L163>

<https://github.com/adcimon/livekit-unity-demo/blob/main/unity-project/Assets/Scripts/Player.cs#L83>

We are having issues in Firefox and Chrome. LiveKit version is 1.1.2

You can try, I’m not sure we made a change related to a similar issue
cc <@U01KZ61GV5Z>

I just updated the <https://github.com/livekit/client-sdk-unity-web/releases/tag/v1.3.0|Unity SDK to client-sdk-js v1.3.0>, can you give it a try ?

I don't think we have any repro for this, so any steps so that we can replicate the issue would be helpful.

Hi, we just tried it and we are having the same issues

Our steps to get the issues are just to connect to the room with different webcams/mics and sometimes the same mic works and sometimes not

Maybe we can give you access to our demo but I don't know if that would be useful

from the same computer? or different machines?

when you say "different webcams/mics", what does that mean? are you choosing a different one each time?

I mean, if I join the room with my mic sometimes it doesn't work

We are joining with different mics to test it

We are considering using the js sdk and communicate the Unity WebGL build with JS, we only need audio

Hello again, I have now enabled webRTC on my Asterisk server and I need a way to connect it with livekit Server which is also installed on the same instance.

Hi Danny, can you describe what you mean by `not everyone can hear it, but only a part of it` ?
If some of the participants cannot hear the speaker, do they miss the first part of it? Or do they miss the entire thing?
What about the experience for other participants?

Hi~ david~ thanks your reply~
• Audio lost for some participants
• my situation is 20 people in same room.
• A participant only speaking with video and the others (19) audio/video mute
• One by one B or C participant speaking with mic unmute and video on 
• at that time, some participants cannot hear B voice entire part. but some participants can hear B’s voice. 
• If B participant web client is refresh, some participants can hear and some participants cannot hear his voice. 
• what should I check about it? I don’t know if it’s server or client side.

<@U01KZ61GV5Z> hello~ Could you help me? I found a log in livekit server. “channel is full” so many in log. should I increase a value “DefaultMessageChannelSize” 200 to 500? or Should I check another value? or vCore 32 machine is slow? I dont know exactly.

oh really? are you able to reproduce this with 1.2.0? we have gotten "channel is full" reports previously,  but those issues have been long addressed.

could you share information about how your server is configured? (i.e. are you using multi-node or single node? with redis?)

<@U01KZ61GV5Z> Thanks for your reply. I have 5 multi-node with redis. Livekit version 1.0.2 I will try to reproduce a issue in v1.2.0. Thanks

got it, do keep me posted if this happens with 1.2, we have yet to receive any reports about this on recent versions.

IIRC, previously certain users connecting via TCP on a slow connection could cause things to freeze up.

<@U01KZ61GV5Z> ok. Im using UDP with co-turn. but my company network have a so many firewalls.( PC web client in private network and public livekit server infra) it makes network so slow. this kind slow or private network factor can also make this issue? I will try to  reproduce it again soon.

yes, on older versions, slow client networks had been the cause of the TCP freezes. As mentioned, that's been resolved in 1.2

hey <@U040S287Q9E>, There are new hardware APIs here,
ht<tps://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 !

Hey <@U03PA7X4T1V>, this is a known issue that's not currently supported <https://github.com/livekit/client-sdk-react-native/issues/15>

Will there be any possibilities to have this feature enable? Any timeline for now? <@U01KZ61GV5Z>

I am also having the same issue. I would really appreciate if we could have this feature implemented sooner. Thanks you

audio output is implemented now, check out AudioSession.getAudioOutputs/selectAudioOutput. Additionally, audio output is set to speakerphone by default now.

<https://github.com/livekit/client-sdk-react-native/blob/main/src/audio/AudioSession.ts>

<@U01NZKTNE1H> thanks for the quick implementation. I did tested within the existing app, but I couldn’t get the speaker as output. I created an issue in GitHub as well. 

<@U01NZKTNE1H>  I still having the issue when trying AudioSession.

Hi Dheeraj! welcome to the community.

We would love your help on the content/blog side to help developers to build with LiveKit. I'll DM you!

I don't have experience with iOS sdk, but from my react-native sdk experience I have written a logic to stop streaming if I go to background and turn on stream when in foreground. But that was made me the same as you are going through. If you are stopping the streaming of local participant you should trigger an event again to add the media tracks of the participant again

<@U03PA7X4T1V> we shouldn't be doing this on purpose for this behavior but thanks a lot for your tip :pray:

<@U02MAJLGN65> I updated my version of swift package at now to 1.0.4

you changed something in this version for this problem?

Can you check if your Room itself is disconnecting ?
You need to publish audio track also for background mode to work correctly (iOS limitation)
Can you try that.

few seconds later in background, user disconnected

when I come back, some times reconnected, some time no

<@U02MAJLGN65> so i publish a video track for example, when I go to background , I need publish video track again , it's this?

<@U02MAJLGN65> when I disable the micro of the local participant, it disable the micro well but when I go to background it appears in the window that the micro is active, do you know how to solve this?

`viewModel.room.localParticipant?.setMicrophone(enabled: *false*).then(on: DispatchQueue.main, { localTrackPublication *in*`
        `sender.isSelected = !sender.isSelected`
      `})`

We decided this behavior that microphone indicator will be ON even after disabling the microphone , until you *unpublish* it.
(So when u enable it again there will be no lag and other issues)

&gt; so i publish a video track for example, when I go to background , I need publish video track again , it's this?
You shouldn't need to do this, if you have your mic published (have an active call). Can you confirm BG works when mic is published ?

<@U02MAJLGN65> yes , mic works when is published in BG

<@U02MAJLGN65> In short, tell me what you advise to do, or what is right to do in your opinion with the video and audio tracks, when we have the camera and the micro turned ON and we go to BG, so as not to have these problems. What's your general advice for Background?

I would appreciate if you answered me... I'm waiting for several days... now another topic... when I go to background the user disconnects? can't fix this?

<@U03UD008S7R> it's a bit hard to trace through the issues you're having. After rereading the thread, it sounds like there were/are two:

1. User is disconnected from the room when backgrounding your app
2. Visible microphone indicator
It sounds like the microphone is working while the app is backgrounded. How can the microphone be working in the background (which you confirmed) if the user is disconnected from the room?

If you could provide some more clarity, it would make it easier for us to help.

What might help: could you tell us what you're trying to accomplish? You want a user to be able to continue to stream video while your app is in the background?

<@U01KZ0SCDKL> Sorry if I wasn't very clear, I'll try to be clearer.
Point 1 - the user is disconnected from the room after 1-2 minutes, if I send the app to background
Point 2 - if I turn the microphone on and off, and then send the app to background, it appears in the status bar and gives a signal that the microphone is on.

yes, that's the goal, to continue streaming, with cameras on or off and microphones on or off.

<@U03UD008S7R> iOS is very strict about background modes, if you are not publishing an audio track or playing back audio, it may get killed by the OS.
See <https://developer.apple.com/documentation/xcode/configuring-background-execution-modes> on configuring your app with the correct modes.

Once microphone is published, we do not unpublish it (but to mute it). You can manually unpublish the track if you'd like with <https://docs.livekit.io/client-sdk-swift/LocalParticipant/#localparticipant.unpublish(publication:notify:)>

you'd also want to set <https://github.com/livekit/livekit/blob/master/config-sample.yaml#L173|external_tls: true>

then TLS termination will be the job of the load balancer, and LiveKit wouldn't require a TURN ssl cert

David is this also true if I’m going through the helm chart? I set it there

oh sorry, I missed it in your configuration. that should do it

Ok it wasn’t working using the latest helm chart

to be clear, you'd need to set up the L4 load balancer yourself

alb cannot terminate L4.. it's an L7 load balancer

<https://aws.amazon.com/elasticloadbalancing/features/>

Oh I do need a network load balancer. I suspected but wasn’t sure

<@U01KZ61GV5Z> I couldn’t manage to get it working with a network load balancer so I tried supplying a cert. It came up. Load balancer is in aws. But when I run the livekit connection test it still says can’t connect to turn server.

Any ideas?

I needed to read the fine print in your docs :sweat_smile:. But I used the default tls_port not 443 and properly setup the secretName by adding my ssl cert into the secrets manager in k8s

clipboard_1661857590240.png

When meeting is in progress, the participant list shows as below, is this a known issue?

Hi , Has someone installed monitoring for livekit?

 <https://github.com/livekit/livekit>
 <https://docs.livekit.io/deploy/>
 I saw these pages but I can't install it.

image.png

hi <@U03L2F74CF7>
as far as I understand, this is intended behavior, yes.
How are you pausing the upstream?

Would it be feasible to either
a) only start publishing when the moderator started the stream
b) disable the unmute buttons until the moderator started the stream

<@U02EU2ZPTQU> Is there an (admittedly clumsy) option (c)?

Use selective subscriptions: all participants can publish and mute/unmute (still should probably disable the buttons in the UI), but once the moderator starts the stream, programmatically subscribe to all the relevant streams from each participant?

yeah, nice, c) is arguably cleaner than the others :slightly_smiling_face:

<@U02EU2ZPTQU> by `pauseUpstream()`  method on TrackPublication right after they’re created.

```await this.publishAudioTrack();
await this.publishVideoTrack();

if (!this.hasStreamStarted) {
    // pause upstream untill the stream is started (waiting room closed)
    this.audioTrackPublication.pauseUpstream();
    this.videoTrackPublication.pauseUpstream();
}```
if I don’t publish them I’d have to display preview of the webcam differently. I was trying to keep the flow same for on/off stream.
Thinking about creating the LocalTrack without publishing it to LocalParticipant and doing so once the stream starts.

I see! then I’d suggest to use <@U01KZ0SCDKL>’s solution with selective subscriptions. it has the advantage that there’s not even a split second in which other participants could see each other, which is isn’t guaranteed with the publish -&gt; pauseUpstream workflow

I guess Livekit will then basically “unpublish” the track itself… nobody will subscribe to it so it’s not necessary to even send it to the server, right? :smile:

there is still an advantage when it comes to timing.
with selective subscriptions everyone publishes to the room first, so livekit server already knows about the publications.
when participants subscribe, the only thing that needs to happen is that livekit server forwards those tracks (that it already knows about) to the participants.
and as long as you have dynacast enabled, it will behave exactly the same as with “pauseUpstream” in terms of bandwidth consumption

Great to hear that :slightly_smiling_face: thanks for the advice

Hello <@U03SN0JBTJT>, no its not normal. you should be getting updates in realtime without a page refresh.

<https://example.livekit.io/> is a good example of the usage of LiveKit React.

You definitely don't need to refresh. It's likely you are not getting state updates when participants join.

are you using the <https://github.com/livekit/livekit-react/blob/master/packages/core/src/hooks/useRoom.ts#L26|useRoom> hook? the list of participants returned will be kept up-to-date automatically.

Thank you for your reply,
<@U01KZ61GV5Z> I will try useRoom :slightly_smiling_face:

I have tried subscribing the TrackPublished, LocalTrackPublished, TrackSubscribed, both on Room and on Participant, but i don't seem to get the correct tracks when entering a room that already have participants.

Anyone have some thoughts about this? Thanks

if you have autosubscribe set (the default), you don't need to manually subscribe to tracks.

It'd be best to set a listener on `TrackSubscribed`, for remote tracks they are always fired.

TrackPublished will not be fired for the second participant, publication events are only emitted for *new tracks* published after the participant has joined

Hi, i think i've understood it. On our app, the 2nd participant (the one that joins a previously created room) doesn't have to interact to the application/browser. But i think that (at least on chrome) sometimes the browser blocks video autoplay if the user doesn't interact with the page before. Does anyone ever had this behaviour? Is there a way to go around this? Thanks

are you attaching the audio/video tracks to the same video element? autoplay of video is allowed. but if audio is attached to the same element, then browsers would require user interaction first. Attaching them to different &lt;audio&gt; and &lt;video&gt; elements would be a way around it.

see <https://github.com/livekit/client-sdk-js#audio-playback|docs here> on handling user interaction requirement for audio

Hey Arush,
livekit/livekit-recorder was replaced with livekit/egress. We’ll need to run some tests on these potential fixes, but if you want to fix it ASAP the relevant code is here: <https://github.com/livekit/egress/blob/main/pkg/pipeline/input/builder_audio.go#L101>

<@U040XJTFZC4> Youtube audio issue is fixed in v1.4.0

that sounds great , but im assuming its still not fixed on livekit-recorder cuz its deprecated

could you tell me what i could do to fix the code for livekit recorder or how i could use egress for recording for a custom frontend

i had written business logic on top of livekit recorder but if i can replace it with egress for the same functionality that would be great and would love if you could help share some way to transition from recorder to egress

<@U040XJTFZC4> have you gone through the (recently updated) documentation for Egress yet?

<https://docs.livekit.io/guides/egress/|https://docs.livekit.io/guides/egress/>

I have , what i understood is egress is specifically for livekit server , whereas i want to to just record a web frontend just like you could in livekit recorder by passing the url in yaml file

<@U040XJTFZC4> thanks for that feedback, I’ll take a look through the egress docs and see if there’s opportunity for further clarification. Egress does support recording a web frontend, the same way the old recorder stack did. This might also be helpful as a higher-level overview of features Egress encompasses:
<https://blog.livekit.io/livekit-universal-egress-launch/>

but the document mandates for a livekit server, for me to host the egress service , the livekit-recorder was pretty simple to use

interestingly, we don’t say anything about a web browser here…
<https://docs.livekit.io/guides/egress/room-composite/>

thanks for drawing attention to this <@U040XJTFZC4>, i’ll mull over how to improve clarity :pray:

thanks for replying <@U01KZ0SCDKL>, just fixing the audio thing for the livekit recorder would actually solve my problem

well, we do say this:
```While it's a great starting point, you can easily create your own layout using standard web technologies that you are already familiar with```

which links to this: <https://docs.livekit.io/guides/egress/custom-template/>

i agree but why does the egress cluster need a live kit server

what my older implementation was that i would run livekit-recorder pods on my node and when ever i needed to livestream i would simply spin up a pod and livestream (note : i do not use any other livekit service only the recorder)

things worked brilliantly except when streaming to youtube where video was great but audio was garbled

what’s running the room though? (for clarity: i don’t have a deep understanding of how the old recorder worked)

<https://github.com/livekit/livekit-recorder> read the readme for this

so you’re simply using livekit-recorder as a standalone way to record and stream a webpage in chrome?

yess and i had manuually scaled it using k8s , sorry if that wasnt clear

i think that was the missing piece for me!

i was thinking if the fix isnt possible from your end cuz yall have deprecated the code , i would fix the code myself and maybe create a docker image from src code, but would have to put in some time :stuck_out_tongue: hence asked on slack

<https://github.com/livekit/egress/commit/24feb71960386a303467970de0b750e85df677ee|here> is the relevant fix in Egress. Most of the code is the same or similar, you would need to make these changes in the recorder

alternatively, you could switch to using Egress, change <https://github.com/livekit/egress/blob/main/pkg/pipeline/source/web.go#L237|this line> to `return nil`, and put the website in the `CustomBaseUrl` field of requests. That’s how I record other websites in testing

<@U01UB0LHTTQ> tried to fix the code on recorder myself(switch encoder from faac to opusenc ) but failed , the thing worked fine for local recording but broke on streaming to rtmp endpoint
```+ ./livekit-recorder
{"level":"error","ts":1663266960.3891566,"logger":"livekit-recorder","caller":"logger/logger.go:34","msg":"error building pipeline","error":"pad link: no common format","stacktrace":"<http://github.com/livekit/protocol/logger.Errorw\n\t/root/go/pkg/mod/github.com/livekit/protocol@v0.11.1-0.20211215011801-f77609470e70/logger/logger.go:34\ngithub.com/livekit/livekit-recorder/pkg/recorder.(*Recorder).Run\n\t/workspace/pkg/recorder/recorder.go:69\nmain.runRecorder\n\t/workspace/cmd/server/recorder.go:39\nmain.run\n\t/workspace/cmd/server/main.go:60\ngithub.com/urfave/cli/v2.(*App).RunContext\n\t/root/go/pkg/mod/github.com/urfave/cli/v2@v2.3.0/app.go:322\ngithub.com/urfave/cli/v2.(*App).Run\n\t/root/go/pkg/mod/github.com/urfave/cli/v2@v2.3.0/app.go:224\nmain.main\n\t/workspace/cmd/server/main.go:51\nruntime.main\n\t/usr/lib/go-1.18/src/runtime/proc.go:250|github.com/livekit/protocol/logger.Errorw\n\t/root/go/pkg/mod/github.com/livekit/protocol@v0.11.1-0.20211215011801-f77609470e70/logger/logger.go:34\ngithub.com/livekit/livekit-recorder/pkg/recorder.(*Recorder).Run\n\t/workspace/pkg/recorder/recorder.go:69\nmain.runRecorder\n\t/workspace/cmd/server/recorder.go:39\nmain.run\n\t/workspace/cmd/server/main.go:60\ngithub.com/urfave/cli/v2.(*App).RunContext\n\t/root/go/pkg/mod/github.com/urfave/cli/v2@v2.3.0/app.go:322\ngithub.com/urfave/cli/v2.(*App).Run\n\t/root/go/pkg/mod/github.com/urfave/cli/v2@v2.3.0/app.go:224\nmain.main\n\t/workspace/cmd/server/main.go:51\nruntime.main\n\t/usr/lib/go-1.18/src/runtime/proc.go:250>"}
{"level":"error","ts":1663266960.3893046,"logger":"livekit-recorder","caller":"logger/logger.go:34","msg":"recording failed","recordingID":"standalone","error":"pad link: no common format","stacktrace":"<http://github.com/livekit/protocol/logger.Errorw\n\t/root/go/pkg/mod/github.com/livekit/protocol@v0.11.1-0.20211215011801-f77609470e70/logger/logger.go:34\ngithub.com/livekit/livekit-recorder/pkg/service.LogResult\n\t/workspace/pkg/service/handler.go:132\nmain.runRecorder\n\t/workspace/cmd/server/recorder.go:40\nmain.run\n\t/workspace/cmd/server/main.go:60\ngithub.com/urfave/cli/v2.(*App).RunContext\n\t/root/go/pkg/mod/github.com/urfave/cli/v2@v2.3.0/app.go:322\ngithub.com/urfave/cli/v2.(*App).Run\n\t/root/go/pkg/mod/github.com/urfave/cli/v2@v2.3.0/app.go:224\nmain.main\n\t/workspace/cmd/server/main.go:51\nruntime.main\n\t/usr/lib/go-1.18/src/runtime/proc.go:250|github.com/livekit/protocol/logger.Errorw\n\t/root/go/pkg/mod/github.com/livekit/protocol@v0.11.1-0.20211215011801-f77609470e70/logger/logger.go:34\ngithub.com/livekit/livekit-recorder/pkg/service.LogResult\n\t/workspace/pkg/service/handler.go:132\nmain.runRecorder\n\t/workspace/cmd/server/recorder.go:40\nmain.run\n\t/workspace/cmd/server/main.go:60\ngithub.com/urfave/cli/v2.(*App).RunContext\n\t/root/go/pkg/mod/github.com/urfave/cli/v2@v2.3.0/app.go:322\ngithub.com/urfave/cli/v2.(*App).Run\n\t/root/go/pkg/mod/github.com/urfave/cli/v2@v2.3.0/app.go:224\nmain.main\n\t/workspace/cmd/server/main.go:51\nruntime.main\n\t/usr/lib/go-1.18/src/runtime/proc.go:250>"}
pad link: no common format```
sorry if this is getting annoying now :disappointed:

<@U040XJTFZC4> curious, why not the alternative approach <@U01UB0LHTTQ> mentions above? I’d recommend moving to the new Egress stuff anyway, since it’s under active development.

<@U01KZ0SCDKL> tbh deploying egress is a pain(cluster and redis etc) :_: also i anyways need to make changes to code and build a custom image for egress to work with normal websites and would require more labour
rather than doing all of this , if i can fix the garbled audio issue which is just a change in few LOCs would make my day , all i would need to do is build an image and use that

Hey Hansel, as for current browser support, I believe this functionality is only supported in Chrome tabs. With regards to Flutter, I'll defer to <@U02GAC58A79>.

Hi <@U01KZ0SCDKL>, Thanks for the reply. I currently trying to share screen in the <http://example.livekit.io|example.livekit.io>. But I couldn't Share the Audio too when I try to Share Screen with Chrome Tabs.

Google Meet.PNG

Usually the Google Meet, when we want to share a sound through Chrome Tabs they will have this checkbox

But from <http://example.livekit.io|example.livekit.io> there's no `Share tab audio` checkbox when sharing the Chrome Tabs

Hey Hansel, that's because we haven't implemented that feature into the LiveKit example, but it is indeed possible to do.

I looked at the flutter-sdk code createScreenShareTrack currently only captures the video track, the default setting is `audio: false`, I can create a new PR next week to support capturing audio for chrome tab

Okay, Thanks for your help <@U02GAC58A79>

<http://example.livekit.io|example.livekit.io> doesn't currently publish screen share audio. livekit SDKs are definitely capable though

<@U03S4JSQPFC> I just published a quick recipe here for accomplishing this with JS/TS – I’ll make a note to add a snippet for Flutter:
<https://docs.livekit.io/recipes/share-audio-browser-tab/>

Hi Everyone, any updates for implementing Share Audio in Flutter <@U02GAC58A79>? :pray:

hey <@U03S4JSQPFC>, It was actually merged today and I’m still preparing the new version for release.
<https://github.com/livekit/client-sdk-flutter/pull/158>

Wow, Thank You so much <@U02GAC58A79> :pray:

hi <@U03S4JSQPFC>, new version published, <https://pub.dev/packages/livekit_client/versions/1.1.2>

Hi, please take a look at our <https://docs.livekit.io/deploy/|deployment docs>.

The simplest way to deploy is to a single VM.

11.png

What does your config look like? specifically, did you set `use_external_ip` to true?

please post text snippets instead of screenshots. It's difficult to see and impossible to search through images.

Thanks for answer. It is solved. I think my config file was wrong.

Screenshot 2022-09-06 at 11.35.23 PM.png

Hey Shuvo, what do you mean by `manifest v2`?

<@U01KZ61GV5Z> I mean every chrome extension has a <https://www.json.org/|JSON>-formatted manifest file, named `manifest.json`, that provides important information. We were using manifest version 2 then we migrated to version 3.

I see, I'm not familiar with Chrome extensions, but I wonder if there are websocket / long connection limits imposed by Chrome on extensions.

we establish a websocket connection that is persistent, Chrome might not like that

Thanks <@U01KZ61GV5Z>. This there any way we can resolve this issue?

I'm not sure, are you able to avoid having to be connected to the room the entire time from the extension itself? it's worthwhile investigating to see if others have found ways of using long-running websockets within Chrome extensions

Hey <@U01KZ61GV5Z>, I’m able to connect to the room from the extension and everything seems working fine now, except sometimes the room resume it connection and auto reconnect again. Even if it disconnects and reconnects again the livekit and extension are working fine. I just want to know when livekit resume the connect or why it occurs this:
“resuming signal connection, attempt 0”

that's just a log line when it resumes connection.

Btw on M1 studio, will try on intel and report back

<https://github.com/livekit/client-sdk-flutter/issues/159> filled issue

here’s sample
<https://github.com/jascodes/livekit-RTCVideoRenderer-issue>
<@U02GAC58A79>

<https://codemagic.io/app/63179ea75fd30abed612350c/build/63179fb69df13ea7e4935299>

thanks for filing the bug report especially with a project to repro. we'll have a look.

<@U01KZ61GV5Z> dependency_overrides:
  flutter_webrtc: 0.9.6

fixes it for me as thr is relvant fix merged

I see, I think the issue is that you had to place an override in the first place.

seems that we need to update the dependencies in our pubspec.

hey <@U03T51VH7JS>,this is a hot fix <https://pub.dev/packages/livekit_client/versions/1.1.1-hotfix>

```livekit-RTCVideoRenderer-issue % git diff pubspec.yaml 
diff --git a/pubspec.yaml b/pubspec.yaml
index cbce037..ecbcede 100644
--- a/pubspec.yaml
+++ b/pubspec.yaml
@@ -34,7 +34,7 @@ dependencies:
   # The following adds the Cupertino Icons font to your application.
   # Use with the CupertinoIcons class for iOS style icons.
   cupertino_icons: ^1.0.2
-  livekit_client: ^1.1.1
+  livekit_client: 1.1.1-hotfix
 
 dev_dependencies:
   flutter_test:```

You can check the server side ice candidate gathered from sfu log, and client gathered candidate from `webrtc-internals`, to see if they are consistent as your network configuration.

<@U02GP9AENTH>, thanks! Is there a way to access logs via API, btw ?

From what I see in the code, this is not normal, the number of participants should not include hidden ones.

So it seems, for some reason, that the number of participants got out of sync.

Hey Valentin, that seems like a regression. I see the issue you've filed and will look into it

<@U03R01RQJ7K> it's possible your issue is the same as <https://github.com/livekit/livekit/issues/991>
If you wanna give it a shot, let me know if it works for you

I was able to reproduce this issue, filed issue <https://github.com/livekit/livekit/issues/991|here>

this has been fixed! <https://github.com/livekit/livekit/pull/992>

Thanks :+1:
I will test it in the next official release.

<@U01KZ61GV5Z> thanks we deployed the new server version and the num_participants has the value correct. Well done!

try before with hardcoded strings for the room name, room size and metadata. This should works just fine

is that the same URL you are using to connect to livekit rooms? do client connections work?

I am using <ws://mydomain:7880> for connect room on your example application. For `RoomServiceClient`  api I am trying to use <http://mydomain> .

for production set ups you would need to deploy it in front of SSL, otherwise browsers/clients may reject connections to it. <https://docs.livekit.io/deploy/#domain,-ssl-certificates,-and-load-balancer>

there might be a way to do so, but isn't performed by this helm chart at the moment. If you figure out a way to do it, we'd happily take a PR.