grishka / libtgvoip

VoIP library for Telegram clients
The Unlicense
387 stars 156 forks source link

Current state of video calls? #93

Open fylux opened 5 years ago

fylux commented 5 years ago

What is the current state of support for video calls?

Vincymic commented 5 years ago

Damn right!

fylux commented 4 years ago

Any update?

Vincymic commented 4 years ago

Any update?

Forget about video calls now, especially from grishka, since he was fired from telegram months ago.

fylux commented 4 years ago

@Vincymic Are you aware of any other ongoing effort to support video calls?

grishka commented 4 years ago

@fylux there currently isn't anyone at Telegram to maintain this library

Vincymic commented 4 years ago

In fact, even for calls, it's been getting to the 2.4.4 version for months now. But is there any possibility that you will return to work for telegrams and maybe make video calls?

grishka commented 4 years ago

But is there any possibility that you will return to work for telegrams and maybe make video calls?

Not unless Pavel changes his stance on QA at least.

frzifus commented 4 years ago

@grishka, which QA point do you mean?

grishka commented 4 years ago

@frzifus I mean doing sensible QA at all. Especially regression tests. But then for calls, there isn't a good, exhaustive way of testing them prior to release because there's an infinite amount of combinations of devices, network conditions and network configurations. You have to collect feedback and logs, analyze that and act based on that. No way around. This is nothing but a dead end.

I still haven't quite figured out what to collect and how to analyze it. After all, I'm a developer. I write code. I make a terrible QA engineer, even though I naturally do some basic testing. I make an even more terrible data scientist and an absolutely abysmal DSP engineer. There's a reason why QA and data analysis is done by separate people in just about any other company.

Vincymic commented 4 years ago

I know you were already developing video calls, too bad you are not continuing the development ...

Il giorno mer 4 dic 2019 alle ore 17:19 Gregory K notifications@github.com ha scritto:

@frzifus https://github.com/frzifus I mean doing sensible QA at all. Especially regression tests. But then for calls, there isn't a good, exhaustive way of testing them prior to release because there's an infinite amount of combinations of devices, network conditions and network configurations. You have to collect feedback and logs, analyze that and act based on that. No way around. This https://contest.com/docs/voip is nothing but a dead end.

I still haven't quite figured out what to collect and how to analyze it. After all, I'm a developer. I write code. I make a terrible QA engineer, even though I naturally do some basic testing. I make an even more terrible data scientist and an absolutely abysmal DSP engineer. There's a reason why QA and data analysis is done by separate people in just about any other company.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/grishka/libtgvoip/issues/93?email_source=notifications&email_token=ALEU3MSNMF43NAR35JKLMELQW7KC3A5CNFSM4H6UFAXKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEF5SLCA#issuecomment-561718664, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALEU3MVMJS4WPT3GDDXCUY3QW7KC3ANCNFSM4H6UFAXA .