Iñaki Baz Castillo comments

Results 550 comments of


                                            Iñaki Baz Castillo

Deprecate `observer` properties

Close() methods should never fail, that's my point. That's why I said that close() must be idempotent. If they "fail" that's a bug in the library or app. And whatever...

Deprecate `observer` properties

I'll expose a different point of view: - parent calls "await child.close()". - it rejects due to some error. - now what, should parent call close() again? maybe with more...

Deprecate `observer` properties

Coming back to mediasoup, have you identified any case in which close() fails or doesn't do what it's supposed to do? :)

Deprecate `observer` properties

Will reconsider this in v4.

Fix "SimulcastConsumer cannot switch layers if initial tsReferenceSpa…

Actually we had something a bit different in mind. Unfortunately we are not gonna be able to review this PR no our planned approach for a while. Will notify here...

🦀 Similarities and differences between Rust port and original TypeScript version

So the idea behind this issue is to document future/temporal differences between the TypeScript and Rust layers, right?

MSVC warnings around negating unsigned types

Thanks, will check tomorrow.

MSVC warnings around negating unsigned types

I didn't check it tomorrow

MSVC warnings around negating unsigned types

@nazar-pc we don't really know how to fix these warnings. Trying things like this: ```diff diff --git a/worker/src/RTC/RTCP/Sdes.cpp b/worker/src/RTC/RTCP/Sdes.cpp index 05ada2586..86b868df8 100644 --- a/worker/src/RTC/RTCP/Sdes.cpp +++ b/worker/src/RTC/RTCP/Sdes.cpp @@ -156,7 +156,7 @@...

signed-to-unsigned overflow in send_side_bandwidth_estimation.cc

@sarumjanuch can you consider these changes into your ongoing PR https://github.com/versatica/mediasoup/pull/922? UPDATE: Ignore, since it's already done in your PR: https://github.com/versatica/mediasoup/pull/922/files#diff-531cc543a44ef9ae28d7eaea4e06c92879598eeac34186fcf2c5c225c3ad6833R395