00:00 <iive> ah, a local variable

07:02 <fflogger> [editedticket] MasterQuestionable: Ticket #11268 ([ffmpeg] [Regression] "-use_wallclock_as_timestamps" ineffective) updated https://trac.ffmpeg.org/ticket/11268#comment:3

07:48 <fflogger> [editedticket] MasterQuestionable: Ticket #11653 ([avformat] -c copy does not copy track names) updated https://trac.ffmpeg.org/ticket/11653#comment:9

09:29 <fflogger> [editedticket] MasterQuestionable: Ticket #11668 ([ffmpeg] "-shortest" malfunction with audio much longer than video?) updated https://trac.ffmpeg.org/ticket/11668#comment:3

10:21 <fflogger> [editedticket] MasterQuestionable: Ticket #11671 ([undetermined] enable TCP keepalive socket options) updated https://trac.ffmpeg.org/ticket/11671#comment:2

12:17 <fflogger> [editedticket] RavenWorks: Ticket #11668 ([ffmpeg] "-shortest" malfunction with audio much longer than video?) updated https://trac.ffmpeg.org/ticket/11668#comment:4

13:53 <kasper93> I'm still curious why my fate worker didn't catch the APV issue. Only thing I can think of is that Windows Server is clearing stack or something, which does not happen on consumer kernels

13:54 <kasper93> (maybe it got lucky run) But the stack state was pretty consistent when run locally

14:45 <kasper93> BtbN: mingw-w64-libxml2 2.13.8-2 should fix the xmllint issue

14:45 <BtbN> hmm?

14:46 <kasper93> this https://fate.ffmpeg.org/report.cgi?time=20250720140254&slot=aarch64-clang-windows-native

14:46 <kasper93> is fixed now in msys2

14:48 <TheVibeCoder> YES!

14:49 <kasper93> :(

15:39 <BtbN> ah

15:39 <BtbN> Didn't even notice that

16:06 <kasper93> it's not relevant for ffmpeg, but actually was a real bug in libxml2

17:50 <fflogger> [editedticket] guest: Ticket #11671 ([undetermined] enable TCP keepalive socket options) updated https://trac.ffmpeg.org/ticket/11671#comment:3

18:35 <Lynne> TheVibeCoder: I can't find a single piece of data related to decoding in the prores decoder libs

18:36 <Lynne> I searched for codebook values in the opencl code, nothing, its just DSP, the .dll also has nothing, and not even apple's official decoder has a single variable from the codebooks

18:36 <Lynne> wtf is the data?

18:46 <TheVibeCoder> Lynne: your RE skills are 0

18:46 <TheVibeCoder> goto RE school lessons

18:46 <TheVibeCoder> first what tools you use for RE?

19:03 <Lynne> ghidra

19:03 <Lynne> but its irrelevant, I'm looking at the disassembly, not the decompilation

19:04 <Lynne> just looking for values like 1075 (hex and dec) or 16384

19:07 <TheVibeCoder> Lynne: you searched for codebook sequency say, 8 hex values, first 4 uint16 form codebook?

19:08 <Lynne> no, individual hex values

19:08 <Lynne> something uncommon present in the codebooks, like 1075 or 556

19:08 <TheVibeCoder> that will give you nothing

19:09 <Lynne> how so? if anything, I'd expect this to be the easiest way

19:10 <TheVibeCoder> https://0x0.st/8d7R.png

19:11 <TheVibeCoder> do you have those offsets in your ghidra window from this png?

19:12 <TheVibeCoder> obviously you will not have nice table and labels, that was added manually.

19:13 <TheVibeCoder> but you can infer a alot from offsets inside binary to find all those tables inside your ghidra session

19:18 <Lynne> oh come on

19:18 <Lynne> literally the only thing I didn't try, the h suffix

19:24 <TheVibeCoder> use seach memory

19:24 <TheVibeCoder> not displayed output of dissesmbly window

19:25 <TheVibeCoder> its slower

19:25 <TheVibeCoder> also search memory allows explicit search by hex/decimal/etc

22:57 <BtbN> compn: if you also want to familiarize yourself with mailman3, sign up on lists.ffmpeg.org, and I'll give you rights

22:58 <bracarino> Good night everyone. I'd like some help to resolve an issue related to "Non monotonic DTS" when processing a MPD input and outputting to mpegts or hls (this happens on both). The command I'm using is: ffmpeg -cenc_decryption_key KID -cenc_decryption_key KEY -i URL -c copy -f mpegts pipe:1. Thank you for any help

23:02 <BtbN> You waited barely 5 minutes before cross-posting, and also this is the wrong channel for that question.

23:21 <kasper93> oh, you can vote on responses

23:24 <BtbN> wait what?

23:24 <BtbN> in hyperkitty?

23:44 <cone-469> ffmpeg Jiasheng Jiang master:f7a6965308ab: fftools/cmdutils: Add av_freep() to avoid memory leak

23:44 <cone-469> ffmpeg Michael Niedermayer master:c0f1c3e18579: avcodec/ivi: Check luma/chroma mb_size

23:44 <cone-469> ffmpeg Michael Niedermayer master:051e0d7744db: avcodec/ffv1dec: Check k in get_vlc_symbol()

23:44 <cone-469> ffmpeg Michael Niedermayer master:2d72bf6fe079: avcodec/cfhd: Check idwt_buf size before allocation

23:44 <cone-469> ffmpeg Michael Niedermayer master:749aaf27448b: avcodec/smacker: cleanup on bet buffer failure

23:44 <cone-469> ffmpeg Michael Niedermayer master:8cdb47e47a7a: avformat/concatdec: Clip duration in one more case in get_best_effort_duration()

23:49 <kierank> haasn: you're on hn