<     May 2017     >
Su Mo Tu We Th Fr Sa  
    1  2  3  4  5  6  
 7  8  9 10 11 12 13  
14 15 16 17 18 19 20  
21 22 23 24 25 26 27  
28 _2_9 30 31
00:03 hfb joined
02:02 <Compn> faLUCE : what do you mean asynch ?
02:02 chewey joined
02:08 iml_ joined
03:29 hfb joined
03:36 HarryS joined
04:15 ergZay joined
06:07 mrklintscher joined
06:19 microchip__ joined
06:20 ergZay joined
06:26 <bencoh> faLUCE: it's a pipeline framework and comes with its own ts demux (upipe-ts) and wrappers to demux/codec libraries (x264, av*, sws, ...)
06:36 ivanich joined
07:02 ivanich joined
07:19 kelnoky joined
07:29 mrklintscher joined
08:13 grublet joined
09:08 iive joined
09:08 iive joined
10:27 <faLUCE> bencoh: so, is the upipe ts a mpegts demuxer or a general purpose demuxer?
10:28 <faLUCE> Compn: with async, I mean an asychronous API, which can be used with a single-thread event-looped program
10:34 <Compn> ah
10:34 <Compn> no idea
10:34 <Compn> try asking on mailing list
10:36 <wm4> you can run a synchronous demuxer on a thread
10:39 <faLUCE> wm4: I don't want to use threads
10:40 <wm4> why not
10:42 <faLUCE> wm4: by the way, I got many stupid oppositions/silences about my API usage example from the ffmpeg-devel team in the . Someone, in the #ffmpeg channel noted that this was bad and tried to push my snippet again. Now, what is the current situation? It's that the doc/examples dir has still shitty examples, and mine (which is well coded, in good order, up to date with the 3.x API and useful) has been rejected.
10:43 <faLUCE> I did not say anything about that in the irc channel, but after someone made me see that, I have to say that this is a deprecable way of making a collaborative project.
10:46 <faLUCE> meanwhile, I coded my library: https://github.com/paolo-pr/laav, which allows to make multiple pipes in a very simple way, by wrapping libav
10:47 <faLUCE> with a single thread
10:48 <faLUCE> it's pretty absurd to use multiple threads, without taking advantage from the multi-core, only in order to split tasks into different functions
10:48 <faLUCE> then I'm looking for some asynchronous demuxer
10:53 <wm4> there are not many asynchronous demuxers, because it's really awkward and complex to write them in this way
10:53 <wm4> AFAIK you walked away yourself from ffmpeg-devel because you didn't want to rework your example anymore
10:53 <wm4> or something like this
10:55 <faLUCE> wm4: [12:53] <wm4> AFAIK you walked away yourself from ffmpeg-devel because you didn't want to rework your example anymore, <---- this is completely false. I did not receive any feedback, and it was nonsense to continue working in something that has not been considered
10:56 <faLUCE> wm4: I can only note that the doc/examples dir is full of shitty code. and people from the ffmpeg channel told me that this kind of silence is something new in the ffmpeg staff, and it is bad
10:57 <wm4> which people did
10:57 <bencoh> faLUCE: upipe-ts is a ts demuxer only
10:57 <faLUCE> someone else tried also to push again my example: http://ffmpeg.org/pipermail/ffmpeg-devel/2017-May/210971.html
10:57 <bencoh> (but upipe is a generic framework with a set of helper libraries including wrappers to other demuxers/codecs)
10:57 <faLUCE> wm4: please, I don't want to discuss about something which is CLEARLY stupid and bad.
10:59 <faLUCE> your observation are absolutely overkill. you, and other people, preferred to keep silence instead of saying that the example I provided is much better than the shit there's in the doc/examples dir
11:00 <faLUCE> this is the OPPOSITE of a collaborative project
11:00 <faLUCE> bencoh: thanks. Is there a matroska demuxer too?
11:00 <bencoh> "keeping silent" doesn't necessarily means "agree with the others", at least (especially?) on IRC ;)
11:01 <bencoh> faLUCE: nope, but I guess you could start writing one
11:01 <bencoh> (or plugging or porting any existing code)
11:02 <faLUCE> bencoh: "keep silent", in that case, meant only a BAD and STUPID behaviour in order to not "touch" some "gurus'" ideas
11:02 <faLUCE> bencoh: I could use libmatroska, but it would require time
11:03 <faLUCE> bencoh: but my first question is: does the upipe demuxer provide asynchronous I/O ?
11:03 <bencoh> I/O isn't even related to demuxer there
11:04 <faLUCE> bencoh: I mean: does it provide an asynchronous API?
11:04 <bencoh> everything is event-driven there
11:04 <faLUCE> bencoh: good. I have muxed frames as char* arrays
11:04 <faLUCE> do I have to fill a buffer and then call the demux function?
11:05 <faLUCE> or is there something like demux(char* data) ?
11:05 <bencoh> you don't call a demux function, you build a pipeline by connecting pipes together, and start feeding it using a data source (which is a specific type of pipe)
11:06 <bencoh> (or by inputting "packets" to your pipeline, but that's not really recommended)
11:06 <wm4> faLUCE: regarding collaborative project, the way it works is that your patch is reviewed, and then you're expected to make according modifications
11:06 <wm4> and if you don't, the patch might be rejected
11:07 <faLUCE> <bencoh> (or by inputting "packets" to your pipeline, but that's not really recommended) <--- this is what I want to do, because I receive input with libevent, from an existing pipe
11:07 <bencoh> (upipe-specific questions probably belong more to #upipe than #mplayer by the way)
11:07 microchip_ joined
11:08 <faLUCE> wm4: please don't say FALSE things. I said in the channel that I could do some other modifications. I also asked: "If I do them, will the patch be pushed" and did not receive ANY response.
11:08 <faLUCE> wm4: please don't say things that are NOT TRUE.
11:08 <faLUCE> bencoh: sorry, I did not know about the #upipe channel. Let's ask there
11:11 <faLUCE> wm4: but some guru clearly prefer to keep the shitty code in the doc/examples dir ( I looked again at many lines of the code and they are simply ridicolous )
11:13 <bencoh> while I do agree that code in doc/examples isn't exactly great, I doubt someone explicitely want to keep it in there
11:13 <wm4> to be fair, your example wasn't much better
11:13 <wm4> it was big and messy and did too much at once
11:13 mrklintscher joined
11:13 <wm4> (pretty much what's wrong with the other examples too)
11:13 <bencoh> but I suppose nobody really want to seriously maintain it, neither review modifications
11:13 <faLUCE> wm4: and this is false too.
11:13 <wm4> and yes, everyone was positive about replacing them with something better
11:14 <wm4> anyway, not going there, I don't need this kind of discussion
11:14 <wm4> feel free to continue spreading your slight slander of the ffmpeg project, I don't even care
11:16 <faLUCE> wm4: I really don't care about that too. But given that other people made me not what REALLY happened, and other people confirmed that I was NOT wrong, I say the reality of things. Be free to keep the shitty code in the doc/examples dir, and keep silent so to not offend gods and gurus
11:17 <faLUCE> wm4: that said, I go on whith my projects
11:19 <faLUCE> [13:13] <bencoh> while I do agree that code in doc/examples isn't exactly great, <--- it's only rubbish
11:20 <faLUCE> ridiculous code
11:22 <faLUCE> submission and slavery to "gods and gurus" of coding is something more stupid than ridiculous
11:23 <faLUCE> at least, wm4, don't say FALSE things.
11:25 <wm4> I don't know what you say sounds ridiculous, false, and offensive, so I'm adding you to my ignore list
11:26 <faLUCE> wm4: I don't care about that. as you said "feel free to continue... etc.". But please stop saying FALSE things: "you're expected to make according modifications". I did exactly that.
11:30 <faLUCE> you can ignore me and do what you want. But what is the result? the doc/examples dir of libav is a pile of crap. I could provide many good examples, useful for all the users. And the example I provided can be HONESTLY judged by everyone (http://ffmpeg.org/pipermail/ffmpeg-devel/2017-May/210971.html) . But it's better to serve the gurus and gods, so, be happy to leave the doc/examples dir in its ridiculous state
11:31 <faLUCE> this is the OPPOSITE of a collaborative project. This is the childish behaviour of people that think that making code is a sort of competition.
12:58 krabador joined
14:02 microchip_ joined
14:42 psychicist__ joined
14:55 mrklintscher joined
15:03 iml_ joined
15:09 hfb joined
15:14 johnny56 joined
15:23 kelnoky joined
15:25 mrklintscher joined
15:33 microchip_ joined
15:47 mrklintscher2 joined
16:13 mrklintscher joined
16:22 mrklintscher2 joined
16:57 ergZay joined
17:07 iive joined
17:07 iive joined
17:48 krabador joined
18:08 iive joined
18:08 iive joined
18:14 psychici1t__ joined
18:28 psychicist__ joined
19:35 iml_ joined
20:14 microchip_ joined
20:44 japh joined
20:44 japh joined
20:49 japh_ joined
21:30 microchip_ joined
21:53 MarcWeber joined
21:54 <MarcWeber> -ao jack port= ? using the same app 3 times showing up in qjackctl as my my-01 my-02 having in_0 and in_1 how can I tell mplayer to connect to the -02 instance?
22:23 microchip_ joined
23:44 debianuser joined