On 11/03/2021 15:38, Michael Niedermayer wrote: > You explicitly asked me to comment, and also stated that all options you > listed suck:
Indeed, I was hoping for an insight that wasn't even worse hacks, though, or an opinion on which was Least Bad. Maybe I should have been more clear in my intention. > "So, basically, all options suck, and I thought some people may have opinions > on the least bad > option, here - specifically Michael, as the author of the original patch." All options will suck, because the files are broken. Maybe I didn't need to state it. > I just wanted to help and so I'm not disparaging that, apologies if it came out like that. I am frustrated. > I described a solution that will likely allow demuxing these files without > errors. Its more work and you would workaround the muxer bug in the demuxer. Indeed, but I do not agree with this approach, as it is fragile at best, and I do not think it belong in a demuxer, but the layer above. > Which is, lets call it unpopular. For me it is, at least. I was hoping others would chime in with opinions, too. > You are opposed to that ? then just pretend i never replied. I don't think adding hacks to a demuxer is the correct place for this, no. Whether the rest of the community shares that opinion, I do not know. > What i wrote was there to help not to restrict or object to anything. I didn't read it that way... > The fix for Ticket 385 is not nice and if that would be done better iam > all for it. But i dont think reverting that will magically fix these other > broken files. (Thats not an objection!) It just leaves it to the user > application to deal with the muxer bugs. Which is where I, personally, think it should be done. A demuxer should demux, not mess around with what it demuxes. This kind of stuff belongs in ffmpeg.c timestamp code, or generic timestamp code in avformat. > Iam not sure jb is interrested in this but ill leave him in CC You never know what j-b is interested in, he likes Go2Meeting, after all. - Derek _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
