New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
telegram encryption key exchange ends with an error #27365
Comments
Are you sure about that? Such bugs are usually only on unofficial builds so I'm requesting a proof that you're using the official binary. The log.txt could be such a proof (how to get it is explained when you was creating the issue). |
Naturally, the binary file is official. We use only the official application from the official website |
Yes, it's called cheat code exactly because there's nowhere to enter. Just like when you enter a cheat code in a game. |
But games have a console for that. And here it is unclear how to enter it |
Just enter it blindly
Just upload the log.txt. |
I'm sorry, I've formatted the log files incorrectly. I'm attaching it. |
I attach the logs received after making unsuccessful calls. |
Same story. OS: Arch/EndeavourOS. Incoming and outgoing calls, "Encryption key exchange", ends/disconnects the call, showing duration of 20 seconds. In the logs standard garbage (three years can not fix for KDE, yep). When trying to close the client and reopen it, it shows up in the logs, though why it updates the configuration EVERY DAMN EXIT is a mystery: `20.01.2024 21:27 telegram-desktop D/tgvoip: === Updating voip config === 20.01.2024 21:27 telegram-desktop D/tgvoip: {"enable_vp8_encoder":true,"enable_vp8_decoder":true,"enable_vp9_encoder":true,"enable_vp9_decoder":true,"enable_h265_encoder":true,"enable_h265_decoder":true,"enable_h264_encoder":true,"enable_h264_decoder":true,"audio_frame_size":60,"jitter_min_delay_60":2,"jitter_max_delay_60":10,"jitter_max_slots_60":20,"jitter_losses_to_reset":20,"jitter_resync_threshold":0.5,"audio_congestion_window":1024,"audio_max_bitrate":20000,"audio_max_bitrate_edge":16000,"audio_max_bitrate_gprs":8000,"audio_max_bitrate_saving":8000,"audio_init_bitrate":16000,"audio_init_bitrate_edge":8000,"audio_init_bitrate_gprs":8000,"audio_init_bitrate_saving":8000,"audio_bitrate_step_incr":1000,"audio_bitrate_step_decr":1000,"use_system_ns":true,"use_system_aec":true,"force_tcp":false,"jitter_initial_delay_60":2,"adsp_good_impls":"(Qualcomm Fluence)","bad_call_rating":true,"use_ios_vpio_agc":false,"use_tcp":false,"audio_medium_fec_bitrate":20000,"audio_medium_fec_multiplier":0.1,"audio_strong_fec_bitrate":7000}` |
@AdrianusWest, Does it work fine in other desktop environments? |
@kartavenko1983 , I don't change the OS or desktop environments. Arch and KDE are always difficult. |
I'd suggest to ask the counterpart to update their client or switch to another one as tgvoip is a legacy protocol and is not tested nowadays I guess |
@ilya-fedin Versions on tablet, phones and more are fresh from the site, updated promptly. |
@AdrianusWest maybe it's Telegram X that is always lagging behind? |
@ilya-fedin telegram-desktop 4.14.8-1 - it's kind of ordinary. |
@AdrianusWest, Okay, thanks for the clarification. |
This version is not official and is not supported here. If you want support for this version, please refer to the distro bugtracker. |
@ilya-fedin I must be dumb, but this is the same repository with the very code that you and I are discussing here and now. It's not about another client, not about a fork, not about a neighboring repository - the source code is taken here. What do you mean, "This version is not official and is not supported here."!? |
Source code is not the only thing that could (and does) change the behavior. The way the client is built is also important and distributions never use the official way with pinned dependencies and build environment. |
@ilya-fedin Cool! )))) I'll contact the maintainer, of course, but the situation strikes me as extremely creepy and very ridiculous, especially considering that there are people sitting on github who are related to development in one way or another. Your answer reminds me of the classic "But everything runs on my computer!". |
There's sadly no manpower to test for every possible environment, sorry. If you can provide such manpower and maintain fixes for your distro build, please do it. If you can't, feel free to use the build with tested libraries combination from the official website. |
@ilya-fedin But we can't substitute one dependency for another, can we? The basic ones are immutable, regardless, aren't they? oO |
You can build with a different version of a dependency just fine and that's what distros do. The dependencies also do have configurations flags and the chances the distros build them with the same flags as tdesktop does are little. Some dependencies are patched (e.g. Qt) as they have plenty of bugs affecting tdesktop and ruining the UX while having them fixed upstream is unrealistic (due to bureaucracy, low priority of the bugs to the upstreams and etc). |
@ilya-fedin, what dependencies are needed for the normal operation of the telegram version from the official website? What should be installed in the system? |
@ilya-fedin, Just analyzing what is happening, I get the feeling that something is missing. |
Building instructions are linked in the readme. The key thing is building happening not in the system but in a container with all the required dependencies built the right way. |
Oh, oops, I've mis-read. There are not much runtime dependencies, the binary is almost static, as long as it launches, you have everything required. |
I'd guess something is wrong with the counterpart |
@ilya-fedin, and how can it help? |
I don't know what to answer, please check and report back |
@ilya-fedin, No, the problem persists. I specifically checked it using a desktop computer and a laptop and two different telegram accounts. I launched telegram on both devices using this command. The result is exactly the same. |
Provide the result of |
@ilya-fedin, grep: /opt/Telegram/Telegram: двоичный файл совпадает grep: /opt/Telegram/Telegram: binary file matches |
@kartavenko1983 the command couldn't have such a result, you have entered it wrong |
LC_ADDRESS=ru_RU.UTF-8 |
@ilya-fedin what should be the correct result of executing the command? |
That explains why it didn't help, you have an advanced locale setup |
Try |
Detected locale "C" with character encoding "ANSI_X3.4-1968", which is not UTF-8. |
on my system: $ env LC_ALL=C ./Telegram |
That's right. Do calls work now? If yes, you can try |
it helps, but not understand why not work with en_us or ru_ru |
@ilya-fedin, |
This sounds crazy, but redefining By default on Fedora
|
if LC_* aren't defined then the value from $LANG is used and undefining $LANG (the first command I asked to try) should be enough |
Linux's standard C++ library thinks converting integer to string should be locale-dependent |
@ilya-fedin How to fix it all, do not run it through the terminal every time? |
yes, I confirm on Fedora it enough |
This has to be done in source code. Do you want to build tdesktop from source? |
@ilya-fedin No |
Okay, thanks a lot, we'll be waiting |
after last update to beta Telegram crashes after accepting call, crash report has been sent automatically |
@VladimirMrzv, You need to create a separate problem, and specify the failure ID |
@kartavenko1983 No need, it'll be fixed later today. |
In version 4.14.11 beta, the calls worked correctly, we are waiting for the stable version to be updated |
last_call_log.txt
last_openal_log.txt
log.txt
log.txt
Steps to reproduce
Expected behaviour
A connection should be made and a voice communication session is taking place
Actual behaviour
telegram exchange of encryption keys during a call
it ends with an error, the call ends after 20 seconds
Operating system
linux
Version of Telegram Desktop
4.14.4
Installation source
Static binary from official website
Crash ID
No response
Logs
No response
The text was updated successfully, but these errors were encountered: