2.0 is broken due this error.
There is two way to fix:

1. Create an empty dir with a placeholder file. Empty dir not allowed in git

2. change script

-react-scripts build && rm -rf ../server/public/* && cp -r build/* ../server/public/
+react-scripts build && rm -rf ../server/public && cp -r build ../server/public

Which one do you prefer? I don't want to bug you too much: I will do the change.
I am just wondering which one should I pick..

should possible to minimize to Sidebar

Implement peer-2-peer mode. Be able to switch between modes either automatic or by some other mechanism.

• Different Last-N for desktop and mobile
• Dynamic Last-N based on load/network conditions.

Something to display on start without room name in URL.

Exit page showing after pressing exit button could be the same as start page?
Optionally with "go back to your old room" or last visited rooms view ...

• don't show if participants < 3
• set raiseHandStatus to false for all participants if participant number decreases to 2
• configuration option for enable/disable "raise hand" in config

Fade out button for going back from one fullscreen element.

support toggle camera when using mobile and support safari as well

Users should be able to change the layout of the room, at least:

• Make one or more videos larger than the rest
• One video in fullscreen

Implement file sharing using webtorrent or some other means

• integrate webtorrent.io
• do we need our own tracker or can this run without a tracker?
• drag and drop a file somewhere make a message in chat
• file upload button in chat
• progressbar

Currently, all information about which user was sent by the current user is lost, and all messages are converted to responses.

To fix this, I believe that a lot of work concerning how peers are handled server-side is needed. For example, the peer name is reset when loading the page. Ideally, the peer name should persist for the entire session, and probably somehow be connected to the OpenID identity given that it exists.

Any thoughts?

Change Chrome 72 to use navigator.MediaDevices.getDisplayMedia

Integrate JWT to secure participants from id-hijacking. This will also make it possible to authenticate users to enable functionality like muting, kicking etc.

Independent real time control panel for each room for use for moderators or from admin side

• participant list:
  • Change moderator role / admin role
  • Kickout
  • Mute all or arbitrary participants
• Room :
  • Lock/unlock
    ( - set pin)
  • LastN
  • Default Layout
  • Raise Hand
    ( - Delete)
    ( - Save )

Support local user languages based on browser locale

User ID are generated on client side!
Alternatives:

• change to socket.io? - too much work :(
• implement valid user id in protoo

To simplify deployment and testing, make a Docker container

What do you think?
Should we merge back to RC changes to devel and keep them sync?

[20:30:14] Plumber found unhandled error:
GulpUglifyError: unable to minify JavaScript
Caused by: SyntaxError: Unexpected token: keyword (const)
File: /root/multiparty-meeting/app/multiparty-meeting.js

please check and let me know solution

If Firefox has to use TCP (via TURN or directly to mediasoup-server) there will no mediastreams go to or from client.

Problem is that there are different ICE candidates used on server and client:

Firefox: receive:

firefox-receive.log

Firefox send:

firefox-send.log

mediasoup server:
21598298.log
44260674.log

As you can see Firefox has nominated and selected TCP candidates while mediasoup uses udp candidates.

Actually mediasoup sends streams to Firefox but firewall is blocking.
Transport for receive on mediasoup side is timing out with a dtls failure after 30 secs.
{"roomId":10882856,"peerName":"kgm2bxmq","transportId":44260674,"event":"transport.dtlsstatechange","dtlsState":"failed","timestamp":1540978250225}

router.js is buggy / hacky
-> replace with express and passportjs

If we want #67 without a massive increase to the bundle size, it seems to me that the most reasonable way to go ahead is to make the feature opt-in by a switch in the settings. Then, we can dynamically load the module when enabling the feature in the client.

However, I don't know how the support for code splitting is in Gulp. I think it's a bit easier to do in wbepack using modules such as react-loadable, so this issue is basically for discussing whether we would want to consider moving to webpack or not. I think moving to something like create-react-app would be nice, however there is no support for Stylus in CRA. Writing a custom webpack configuration is certainly viable, though.

Is this something to consider?

Implement screensharing with - at least - Chrome and Firefox.

• rename to ?
• make font size not relative to vmin but em
• fix chat vertical size in portrait mode
• replace "chat", "users" and "setting" with icons

Update Express to use http2.

can you add white board it will be useful for teaching

Screen sharing extension seems to be missing in Chrome Webstore :(

• Sidebar button either does not allow clicks or is incredibly inresponsive when open
• Sidebar is slow to open
• Me is a little to slow to actually drag around
• Peers have no vertical margins
• Displaying tooltips by mousemove is difficult
• Sidebar is not wide enough and the buttons are incredibly, should probably cover the entire screen below a certain screen size
• There is for some reason a vertical scrollbar
• Notifications overlap with the tooltip for the share link button
• Floating action buttons could be a bit bigger
• Consider disabling selection, as the Copy screen is a bit irritating and hard to remove sometimes
• Fullscreen button does not show a video
• Currently, Peer only shows tools when hovering

Currently, the chat (and the files tab, from the file sharing PR) scrolls to the bottom when the component mounts or the component props update. This works for scrolling to the bottom when the sidebar is first opened, and when a new message is received. However, there is a small UI problem here, IMO, as this behavior makes it very difficult to scroll up to see an earlier message if new messages are frequently being added. To me, this is pretty irritating. My suggested fix is to only scroll to the bottom provided the user already was scrolled to the bottom, but to do nothing if the user is scrolled up on the page. Then the user can manually scroll to the bottom again when done reading the chat history.

Support pause of video for all but the last-N speakers.

• make selected user a state in store
• push selected user to lastN speakers so that he gets visible (needs lastN)
• focus selected user in all peer related components (participantList, peer)
• let show up selected user in bigview of filmstrip layout

Hello, it seems I am unable to successfully setup the app. Although the app and the server are deployed with success, none of the clients, nor me with my local tests can receive audio or video from another client. Please, I need someone to even have access to my root server and help me find out why, or suggest me a way of getting any logs that can guide me through the process.

I am trying to make it work for over 15 days now and due to I didn't have any previous experience with webrtc, it's more than difficult to debug the installation.

ANY help would be appreciated.

Support for choosing between light/dark themes, and save it to users cookie.

• move local Hark from PeerView.jsx to higher level or RoomClient.js

Replace dataporten with OAuth or make it configurable..

Hi,
I have an very strange Error when i set the project on my own server.
First, My server is debian 8 . nodejs 8.11 and npm 6.4 ,
I have install the all depends by webtorrent-hybrid. use following commond
" apt-get install libgtk2.0-0 libxtst-dev libxss-dev libgconf2-dev libnss3 libasound2-dev xvfb"

I cloned the git branch "develop",And following your guid ,npm install everything. It seems everything is OK ..
also ,i changed the cert file to my let's encrypt wild cert.
But ,when I open the URL https://xxx.mydomain.club/ , always get error message :
"Failed to initialize native PeerConnection"

Also ,because i'm using chrome ,so i try to add "https://webrtc.github.io/adapter/adapter-latest.js" .
but no lucky . It's still same error.

I have attach the console log for your referrence.

Any help will be appreciate.

Raymond

BTW: I've also try these steps on other "clean" VHost . But ...you known the result.

noname.zssp.club-1540042472958.log

read / write access to database

https://github.com/facebook/create-react-app/blob/master/packages/react-scripts/template/README.md#making-a-progressive-web-app

• Add a manifest file and related configuration, so that the app can be added to phone homescreens
• Service workers to improve loading speed, regardless of if the page is installed as an app or not

• global fadeOut timer for: toolArea, Menu button, invite button and reappear onMouseMove
• local PeerView buttons should fade out independently and reappear onMouseOver

trustworthy user identity

• Login / Logout button / User info
• prevent same display name in one room
• marker for verified and anonymous users everywhere
  • in Participant list
  • Chat
  • Username on PeerView
• If in-logged: save a session token on clientside and serverside

The Web Audio autoplay policy will be re-enabled in Chrome 70 (October 2018) and it might require changes to hark (?). See https://goo.gl/7K7WLu for more information

needed for claim rooms and persistent room config

• mariaDB|postgres SQL?
• userTable (userID)
• userRooms (userID, roomName)
• roomTable (roomName, foreign key userID from userRooms,config:{'locked':false, 'pin':'',moderatorList:[],chatHistory:[]})