I implemented a call button for testing purposes on one of my webpages and it worked fine, but in recent updates of the webcall server it seems to be broken, see console output by Chrome browser: Uncaught TypeError: Cannot read properties of null (reading 'contentWindow') at window.onload (button.js:3:3) window.onload @ button.js:3 load (async) (anonymous) @ button.js:1 and Uncaught TypeError: Cannot set properties of null (setting 'value') at caller.js:642:27 at caller.js:1120:4 at xhr.onreadystatechange (client.js:315:4) (anonymous) @ caller.js:642 (anonymous) @ caller.js:1120 xhr.onreadystatechange @ client.js:315 XMLHttpRequest.send (async) ajaxFetch @ client.js:341 checkServerMode @ caller.js:1118 onload2 @ caller.js:630 window.onload @ caller.js:521 load (async) (anonymous) @ caller.js:115

it does not matter if I enter the ID or nickname, the error remains the same. Did the button behaviour change in any way different to the documentation maybe?

Thanks!

Hi, I can't find a file like "webcall_(yyyy-mm-ddThh+mm:ss).tgz" in the path "(home)/_backup/". I followed the steps on the site and it still doesn't appear (I've had the server running for over 48 hours).

Below is the content of my config.ini file.

hostname = (mydomain) httpsPort = 8068 wssPort = 8443 insecureSkipVerify = false logevents = attach thirtySecStats = true turnPort = 3739 turnRealm = (mydomain) turnIP = (myIP) backupScript = backupdb.sh backupPauseMinutes = 240

Then the following is the content of my backupdb.sh file.

export MYDATE=`date --iso-8601=hours` tar cfvz ~/_backup/webcall_db/_$MYDATE.tgz db/*

Then the following is what my terminal shows.

stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 callBackupScript sync db's (backupdb.sh) stats 0/0/0 calls:0/0 0/0 go:81 callBackupScript exec (backupdb.sh)... # callBackupScript backupdb.sh err=exit status 2 log=(db/rtccalls.db db/rtccontacts.db )stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 callBackupScript sync db's (backupdb.sh) stats 0/0/0 calls:0/0 0/0 go:81 callBackupScript exec (backupdb.sh)... # callBackupScript backupdb.sh err=exit status 2 log=(db/rtccalls.db db/rtccontacts.db db/rtchashedpw.db )stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 stats 0/0/0 calls:0/0 0/0 go:81 callBackupScript sync db's (backupdb.sh) stats 0/0/0 calls:0/0 0/0 go:81 callBackupScript exec (backupdb.sh)... # callBackupScript backupdb.sh err=exit status 2 log=(db/rtccalls.db )stats 0/0/0 calls:0/0 0/0 go:81

BTW I have put config.ini and backupdb.sh in the webcall folder.

Thank you.

I found the file transfer speed too slow.  I did the streaming test on Google's WebRTC and the speed was much better.  Another observation, sometimes the button to send files does not appear when in the browser.

Hi, thank you for webcall.

Is it possible to use webcall to have more than two participants in a call? I'm looking for a simple browser-based voice chat system.

Thanks.

I see only red light, I can see in Chrome browser the button "Hang up" but can't hang up. When clicking the button I can hear the busy signal, but can't hang up.

Hi,

I'm sorry if there is an obvious answer to this or if it has already been answered. I looked through the documentation and closed issues that I could find.

Does the webcall server support resetting, recovering, or changing user passwords? Or, is there some way to do this manually to the backing data store? The probability that I will forget or misplace a password approaches 1 as the amount of time I use webcall approaches infinity 😃

Thank you!

Very neat project! I was trying to run my own instance behind nginx as a reverse proxy. I followed the steps https://timur.mobi/webcall/install/

Using config.ini as such:

httpsPort = 0
hostname = hostname.url
wssPort = 8443

nginx config is same as provided on URL above (adjusted to point to proper webroot).

I also checked precedence to ensure it's not caught by any other location handlers... that said I can access the callee page, get assigned an ID but then when I'm taken to https://url/callee/123456789, I get a 404... "GET /callee/11727440667 HTTP/2.0" 404 107

I tried playing with try_files configs to point to full path to callee/index.html and some other things, but all to no avail.

There's nothing suspicious in my error.log, and I do not see activity in the webcall standard output (except registration calls which work fine)

What is it that handles routing URLs ending in numbers beyond the nginx confiugration? Please let me know if I can provide any additional information, would love to get this working. :)

Hi, When I use the video function, after clicking the "monitor" button, the message "monitor paused" will be displayed on my screen, but the other party (the person I am chatting with) will still see my video screen (turning head, moving, etc. ). Thank you.

Hi, When I set "Nickname" to "1" and call my friend, my friend's phone will show as "unkonw". When I set "Nickname" from "1" to "2" and call my friend again, my friend's phone will show "1". When I set "Nickname" from "2" to "3" and call my friend again, my friend's phone will show "2".

The "Nickname" displayed will be the "Nickname" of the previous call. Thank you.

Running as non-root user yields:

tls.LoadX509KeyPair err=(open tls.pem: permission denied)

I believe this is because it needs the certs to run the websocket server (let me know if im wrong) which is not and cannot (?) be proxied through nginx

Given that LetsEncrypt requires frequent renewals, and some people such as myself automate certbot using cron jobs, Thus, I opted to symlink the certs inside my webcall deployment so they remain unaffected when they change.

While I understand workarounds may exist like copying and altering the permissions of the files, they do require maintenance as well as leniency on security.

Perhaps if detected that EUID is 0 it can read TLS files into memory then de-escalate to a specified user from config.ini? Just first thing that came to mind.

Hello dear developer, thank you very much for your amazing project!

I got a ws-connect timeout problem when I was testing in a local area network.

what I plan to do

1. Build webcall server
2. Use NAT traversal service to use webcall on the internet.

what happened

1. Build webcall server on an orange pi zero 2 (a Raspberry Pi like device), succeed.
2. Webcall build succeed
3. Using this config.ini below, tested in localhost with http, works fine.

hostname = 127.0.0.1
httpsPort = 8068
wssPort = 8443
insecureSkipVerify = true

4. Using this config.ini below, tested in localhost with HTTP, self-signed certificate, works fine.

hostname = 127.0.0.1
httpsPort = 8068
wssPort = 8443
insecureSkipVerify = true

5. Using this config.ini below, test in LAN, not able to Go Online using a machine in LAN. Error log shows /login ws-connect timeout 30s remove xxxxxxxxxxx/xxxxxxxxxxx xxxxxxxxxxx rip=192.168.0.105:58573 ver= ua=xxxxxx

hostname = 192.168.0.112
httpsPort = 8068
wssPort = 8443
insecureSkipVerify = true

6. Tried many NAT traversal service, using this config.ini, same problem.

hostname = 127.0.0.1
httpsPort = 8068
wssPort = 8443
insecureSkipVerify = true

Environment

Orange pi zero 2 Debian 10 go1.17 linux/arm64 Firefox

There could be an alternative to create, on Android, a server on an internal network to talk to people who are on the same network, but they don't have the app. Could create a link with the data by creating a qrcode.

I used Jitsi Meet, but it doesn't allow having a single contact login and Wecall allows you to receive calls without the other person needing to install or create a login.  What I miss about WebCall is the ability to receive calls from dial-up (cell) numbers.  Another Jitsi Meet function is the ability to record and stream to YouTube and accept more than one person on the call.