[01:42:01 CEST] <kierank> michaelni: thx [04:53:18 CEST] <pross> these fuzzing services provided to open source projects seem like a cheap way to collect large amounts of training data [10:09:02 CEST] <cone-241> ffmpeg 03Michael Niedermayer 07master:2603f25d3264: avcodec/bink: Reorder operations in init to avoid memleak on error [10:09:02 CEST] <cone-241> ffmpeg 03Amir Pauker 07master:edfced8c045c: avutil: add FF_DECODE_ERROR_CONCEALMENT_ACTIVE flag for AVFrame.decode_error_flags [10:09:02 CEST] <cone-241> ffmpeg 03Limin Wang 07master:1c3ed1189313: doc/filters.texi: Fix the confusing description for find_rect and cover_rect command [13:00:58 CEST] <JEEB> should I make a patch to disable the connection re-usage from TLS with schannel? [13:01:08 CEST] <JEEB> since it seems like schannel performance just goes way down when that's enabled [13:11:27 CEST] <nevcairiel> schannel overall has an issue where it breaks down on some connections, and the API is so convoluted that I couldn't figure out why, hence I'm using gnutls now :D [13:12:33 CEST] <JEEB> :D [13:12:38 CEST] <JEEB> yea, I've seen various people do that [13:13:47 CEST] <nevcairiel> i implemented it roughly base on the implementation in curl i think, but not sure if I made a snafu or whatever, but I couldn#t figure it out [16:53:27 CEST] <stuartm> Out from -progress includes the following issue: [16:53:31 CEST] <stuartm> out_time_us=51440078 [16:53:32 CEST] <stuartm> out_time_ms=51440078 [16:54:19 CEST] <stuartm> Is there some legacy, intentional reason that it's outputting the millisecond value in microseconds? [16:55:01 CEST] <stuartm> Or should I just open a ticket with patch? [17:17:11 CEST] <nevcairiel> It was a bug, and cannot be changed without breaking scripts that rely on the current behavior, hence the correct second one was added, so the old one can be removed or fixed at a later date [00:00:00 CEST] --- Mon Jun 17 2019