[Ffmpeg-devel-irc] ffmpeg-devel.log.20111112

[burek]

[00:01] <michaelni> Daemon404, its not "my ML" its everyones ML
[00:02] <iive> he just means the ml where you are subscribed ;)
[00:06] <michaelni> iive, iam subscribed to many MLs, quite a few more than i read regularely
[00:06] <Daemon404> michaelni, youre  was not referring specificallt to you
[00:06] <iive> well.. the one you always read and write :P
[00:06] <Daemon404> iy was referring to this room
[00:06] <Daemon404> poor english on my part.
[00:07] <michaelni> Daemon404, no i dont take offense so easily
[00:07] <beastd> It could be especially misunderstood as you were referring to michaelni personally earlier in that message
[00:08] <Daemon404> yes
[00:08] <Daemon404> i did say it was poor english on my part
[00:08] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r0e5b05134b 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[00:08] <CIA-18> ffmpeg: mandelbrot: fix typo that led to hardcoded max iteration count.
[00:08] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:08] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * raa198ad679 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[00:08] <CIA-18> ffmpeg: mandelbrot: make 2 variables const
[00:08] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:08] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * rdece017280 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[00:08] <CIA-18> ffmpeg: mandelbrot: center coordinate reference on the images center
[00:08] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:08] <beastd> Daemon404: I think everyone got what you meant. No problems...
[00:08] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * rddeb194b14 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[00:08] <CIA-18> ffmpeg: mandelbrot: add a end_scale and pts so we can zoom in.
[00:08] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:09] <Daemon404> :P
[00:12] <michaelni> Daemon404, where is the patch? i cant see it on "the" ML ?
[00:12] <Daemon404> im just about to hti enter on git send-email
[00:12] <Daemon404> ... wow i typo lots.
[00:13] <michaelni> sorry dont let me interrupt you ...
[00:13] <michaelni> no hurry, you will find the enter key one day
[00:13] <michaelni> its the big one at the right ;)
[00:13] <Daemon404> lol
[00:14] <Daemon404> sent.
[00:14] <JEEB> michaelni, unfortunately there are keyboards where the enter key is smaller >_>
[00:14] <Daemon404> and my shift key is bigger than my enter key!
[00:14] <JEEB> (which reminds me of the keyboards where backspace is the size of a thumb)
[00:14] <JEEB> had to use such a keyboard for a year in Japan :<
[00:15] <nevcairiel> that reminds me of those horrible mac keyboards
[00:15] <Daemon404> http://www.blogcdn.com/www.engadget.com/media/2006/10/10406-g11keyboard.jpg my kb
[00:15] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r5f36065ca6 10ffmpeg/libavformat/wtvenc.c: Use AC-3 as default codec for wtv.
[00:15] <Daemon404> i dont even game on pc either. it was a gift.
[00:15] <michaelni> it reminds me of my old notebook where some keys (nearly) stoped working
[00:15] <iive> there are actually 2 enter keys
[00:16] <iive> on regular 101+ key keyboards.
[00:22] <beastd> My right shift key has the same size as my e.g. "A" key. But my "<" key is only half that size and also on the outer right. 
[00:23] <drv> only keyboard worth using ;) http://upload.wikimedia.org/wikipedia/commons/5/57/ModelM.jpg
[00:23] <nevcairiel> drv: even that one has the horrible small enter key
[00:23] <nevcairiel> screw US layouts
[00:24] <michaelni> Daemon404, the code id looks wrong
[00:24] <Daemon404> code id?
[00:24] <michaelni> CODEC_ID_VBLE
[00:24] <michaelni> its value is nonsense
[00:24] <michaelni> MKBETAG( 0 ,'G','2','M') + 1
[00:25] <iive> drv: It felt so nostalgic ...
[00:25] <Daemon404> michaelni, i am sort of confused
[00:25] <Daemon404> where should it be
[00:25] <Daemon404> ffmpeg's codec id list with its g2m thing sort of confused me
[00:25] <michaelni> it should have the same value as in libav
[00:26] <michaelni> ideally
[00:26] <michaelni> if its still available
[00:26] <michaelni> ill fix it
[00:26] <Daemon404> which is directly under BMV
[00:26] <Daemon404> ok
[00:26] <Daemon404> thank you
[00:27] <Daemon404> go for it.
[00:27] <michaelni> Daemon404, also any specific reason why you didnt optimize things like i suggested?
[00:27] <Daemon404> i plan on doing the horiz_band stuff i na separate patch
[00:28] <michaelni> i mean vble_read_reverse_unary()
[00:28] <Daemon404> it is doen in a similar fashion to what you did
[00:28] <Daemon404> only with 9 bits
[00:28] <michaelni> it looks slow
[00:28] <Daemon404> (it wont every need more tahn 9)
[00:28] <Daemon404> you can try replacing it
[00:29] <michaelni> also it doesnt need to read 9
[00:29] <michaelni> 8 is enough
[00:29] <Daemon404> how so?
[00:29] <Daemon404> i check the last bit because teh bitstream might be corrupt
[00:31] <Daemon404> also, wouldnt len[] in your version be a pretty large lookup table?
[00:31] <Daemon404> like 256 long
[00:31] <michaelni> how large is the decoder ?
[00:32] <Daemon404> pretty small...?
[00:32] <michaelni> 192040 bytes ?
[00:32] <Daemon404> and for the other thing, i chose understandability/readability over optimization
[00:32] <michaelni> -rw-r--r-- 1 michael michael 192040 2011-11-12 00:29 libavcodec/vble.o
[00:32] <Daemon404> since it's a rather obscure decoder
[00:32] <Daemon404> oh you mena the object file?
[00:33] <Daemon404> you want to know how big my object file is?
[00:33] Action: Daemon404 confused
[00:35] <michaelni> iam just saying that the LUT could easily be smaller than the code that is used instead of it
[00:35] <Daemon404> oh, yes.
[00:36] <Daemon404> like i said i chose readability/understandability since it's a rather obscure codec that isnt really use in practice
[00:39] <Daemon404> michaelni, i dont have any problem optimizing it after in subsequent patches
[00:39] <Daemon404> just as a note.
[00:41] <iive> usually big lut also amount for big speed loss, as they use a lot of cpu cache. Well, my current cpu have megabytes of cache... but small embedded cpu are not that lucky.
[00:42] <Daemon404> and i do plan on implementing the horiz_band stuff
[00:42] <michaelni> iive we talk about 256 byte LUT
[00:42] <iive> oh, then not so big.
[00:42] <iive> not big at all.
[00:43] <Daemon404> btw, mike melansan gave me this: http://wiki.multimedia.cx/index.php?title=Small_FFmpeg_Tasks
[00:43] <Daemon404> i might work on one of those.
[00:43] <michaelni> Daemon404, ill fix the codec_id and commit after testing
[00:43] <Daemon404> ok
[00:43] <michaelni> where can i find a file to test
[00:43] <michaelni> ?
[00:44] <Daemon404> mike said he added it to teh archive
[00:44] <Daemon404> let me check
[00:44] <Daemon404> hmm not there yet
[00:44] <Daemon404> for now: http://japland.org/flowers.avi
[00:44] <michaelni> thx
[00:45] <michaelni> Daemon404, also as you want to maintain this, where is your git repo in which you want to maintain this code ?
[00:46] <Daemon404> ?
[00:47] <Daemon404> why do i need a public git repo?
[00:47] <Daemon404> perhaps you and i have have different definitions of what maintain means...
[00:48] <michaelni> maintain for me means you work on the code, review patches, fix bugs, (not neccessarily all of that)
[00:48] <Daemon404> yes, but why do i need a public git repo for that?
[00:49] <Daemon404> ML is on sufficient?
[00:49] <Daemon404> not*
[00:49] <michaelni> ML is not ideal
[00:49] <michaelni> a user cant checkout the ML to test your latest code
[00:50] <michaelni> you need to wait for a reply each time you do something on the ML
[00:50] <Daemon404> i generally dont send my stuff to ML, or even push my stuff to public git repos until im happy with it
[00:50] <Daemon404> im not very accustomed to making my Bad Code public until it;s no longer bad
[00:51] <Daemon404> or at least less bad
[00:51] <michaelni> iam not asking for your bad code to be public
[00:51] <Daemon404> is this so it is easier to pull from me?
[00:52] <michaelni> it is because it is easier for _you_ :)
[00:52] <Daemon404> lol
[00:52] <Daemon404> i find MLs easier to use... but i'm also very weird
[00:53] <michaelni> a ML is not a repository
[00:53] <Daemon404> i know
[00:53] <Daemon404> i dont see a benefit to me in having a public repository
[00:54] <Daemon404> i'd only push things i also felt were fit to send upstream anyway
[00:55] <michaelni> when you maintain a decoder then you are upstream for that decoder
[00:55] <Daemon404> i consider ffmpeg the upstream though
[00:55] <Daemon404> and i wouldnt make a change i didnt send
[00:56] <Daemon404> if it's gonna be trouble, you dont need to add me to the maintners file
[00:56] Action: Daemon404 is fine without it
[00:56] <beastd> Daemon404: no trouble
[00:56] <michaelni> Daemon404, its no problem to use the ML for me if you prefer but its not a good idea
[00:57] <michaelni> you can change the code in your repo when you spot a bug or want to improve it
[00:57] <michaelni> you cant do anything with a patch you posted
[00:57] <Daemon404> huh?
[00:59] <michaelni> with a repo you can do any kind of change you need to do, reoder commits split one, fix a typo, ask someone to test, ...
[00:59] <michaelni> reorder
[01:00] <michaelni> with a ML you must resubmit the patches
[01:00] <Daemon404> i get what you mean now.
[01:00] <Daemon404> ive never  really had problems with resubmitting patches myself, but i can certainly see how it is annoying
[01:00] <Daemon404> perhaps i finally have an excuse to get a github account
[01:01] <michaelni> :)
[01:02] <Daemon404> i do wish github's web interface wasnt mega-ugly.
[01:05] <beastd> Daemon404: It is just recommandations by michaelni. To be the maintainer you are not forced to have a public git repo. Just get one, if you like to have it. The important thing is you do the maintaining. if you do your changes in your local git repo and sent them to the ML once ready it's also fine as well.
[01:06] <Daemon404> not to be a prude
[01:06] <michaelni> beastd, very well said ...
[01:06] <Daemon404> but i to prefer the latter
[01:06] <Daemon404> s/to/do/
[01:07] <beastd> It's OK. If you like it better.
[01:09] <beastd> Other maintainer task are for example reading the ffmpeg-devel ML and commenting/reviewing the patches that change/influence the files you maintain.
[01:09] <Daemon404> im aware of that
[01:09] <Daemon404> i already read it as-is
[01:10] <beastd> fine fine
[01:10] <Daemon404> we use MLs internally at work, so im accustomed to it
[01:11] <Daemon404> with git nonetheless
[01:11] <beastd> yes, one doesn't exclude the other
[01:12] <beastd> Are you subscribed to ffmpeg-csvlog too?
[01:13] <Daemon404> just -devel
[01:13] <beastd> It would be good if you were subscribed to ffmpeg-cvslog too.
[01:14] <Daemon404> what is it? just git log?
[01:14] <beastd> yes
[01:14] <Daemon404> k
[01:14] <beastd> You don't need to read all the commits if you don't like. But you should watch out for reply messages to the commit messages.
[01:14] <beastd> Let me please explain...
[01:14] <Daemon404> i generally look at all of x264, ffmpeg, and libav's commits
[01:14] <Daemon404> to see if there is anything interesting
[01:15] <beastd> That's of course good anyway.
[01:15] <beastd> The important scenario is the following.
[01:16] <beastd> You made some changes and sent them to the ffdev ML or you ended a review round on the ffdev ML
[01:16] <beastd> then the patch gets pushed into the public ffmpeg repo
[01:17] <beastd> after that many developers will read the change on -cvslog mailing list and probably some have comments or spot errors
[01:17] <Daemon404> oh
[01:17] <beastd> then they will reply right on that list.
[01:24] <CIA-18> ffmpeg: 03Derek Buitenhuis 07master * r0c251257b5 10ffmpeg/ (9 files in 4 dirs): 
[01:24] <CIA-18> ffmpeg: VBLE Decoder
[01:24] <CIA-18> ffmpeg: Add a decoder for the VBLE Lossless Codec, which
[01:24] <CIA-18> ffmpeg: still has a cult following. Used to be popular
[01:24] <CIA-18> ffmpeg: several years ago on doom9.
[01:24] <CIA-18> ffmpeg: Signed-off-by: Derek Buitenhuis <derek.buitenhuis at gmail.com>
[01:24] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:24] <CIA-18> ffmpeg: 03Janne Grunau 07master * r6761b6b825 10ffmpeg/libavcodec/vble.c: 
[01:24] <CIA-18> ffmpeg: vble: remove vble_error_close
[01:24] <CIA-18> ffmpeg: It does not make much sense to factor the error handling to its own
[01:24] <CIA-18> ffmpeg: av_always_inline function. Fixes "format not a string literal and no
[01:24] <CIA-18> ffmpeg: format arguments" warning in the av_log.
[01:24] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:24] <Daemon404> and i realize, michaelni 
[01:24] <Daemon404> i need to fix libutvideodec
[01:24] <Daemon404> cause it sets the linesize.
[01:26] <michaelni> probably, no
[01:27] <michaelni> only if you use get_buffer() set CODEC_CAP_DR1
[01:27] <michaelni> are you not allowed top mess with linesize
[01:27] <Daemon404> oh
[01:27] <Daemon404> so im fine then
[01:27] <michaelni> if you allocate the frame yourself you can 
[01:27] <Daemon404> cool.
[01:27] <Daemon404> libutvideoenc is also done
[01:27] <Daemon404> but im waiting for a patch to libutvideo to make it upstream
[01:28] <Daemon404> (it's in the next release of libutvideo)
[01:43] <beastd> I am leaving now. Bye..
[01:44] <Daemon404> cya.
[02:19] <michaelni> Daemon404, just posted a bunch of patches to ffmpeg-dev for you to review :)
[02:20] <Daemon404> oh
[02:20] <Daemon404> that was quick
[02:22] <Daemon404> i wish thunderbird didnt suck at attachments
[02:42] <Daemon404> michaelni, reviewed.
[02:54] <michaelni> Daemon404, ill fix the issues and then push it, ok ?
[02:54] <Daemon404> ok
[02:54] <Daemon404> sounds good.
[03:02] <CIA-18> ffmpeg: 03Mike Melanson 07master * r45fcb86cf8 10ffmpeg/libavformat/aiff.h: 
[03:02] <CIA-18> ffmpeg: AIFF: add 'twos' FourCC for the mux/demuxer (big endian PCM audio)
[03:02] <CIA-18> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[03:02] <CIA-18> ffmpeg: 03Anton Khirnov 07master * re1e22851c1 10ffmpeg/libavutil/opt.c: AVOptions: don't return an invalid option when option list is empty
[03:02] <CIA-18> ffmpeg: 03Martin Storsjö 07master * rb01f5ba207 10ffmpeg/libavformat/http.c: (log message trimmed)
[03:02] <CIA-18> ffmpeg: http: Print an error message for Authorization Required, too
[03:02] <CIA-18> ffmpeg: The error was hidden before, to avoid showing an error on the
[03:02] <CIA-18> ffmpeg: first request where no auth has been provided, when the server
[03:02] <CIA-18> ffmpeg: indicates which authentication method to use.
[03:02] <CIA-18> ffmpeg: Now the error is printed if an authentication method was used,
[03:02] <CIA-18> ffmpeg: but failed.
[03:02] <CIA-18> ffmpeg: 03Martin Storsjö 07master * rb911518d1c 10ffmpeg/libavformat/ (http.c httpauth.c): 
[03:02] <CIA-18> ffmpeg: http: Handle proxy authentication
[03:02] <CIA-18> ffmpeg: Tested with both Basic and Digest authentication, and tested with
[03:02] <CIA-18> ffmpeg: both proxy authentication and authentication for the requested
[03:02] <CIA-18> ffmpeg: resource at the same time.
[03:02] <CIA-18> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[03:02] <CIA-18> ffmpeg: 03Sean McGovern 07master * r124e56454d 10ffmpeg/libswscale/utils.c: (log message trimmed)
[03:02] <CIA-18> ffmpeg: swscale: add padding to conversion buffer.
[03:02] <CIA-18> ffmpeg: Altivec does unaligned reads from this buffer in
[03:02] <CIA-18> ffmpeg: hscale_altivec_real(), and can thus read up to 16 bytes beyond
[03:23] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * rc5e55e214c 10ffmpeg/libavcodec/vble.c: 
[03:23] <CIA-18> ffmpeg: vble: change variable to int, its slightly faster and int is preferable if no specific size is needed.
[03:23] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:23] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * re3618cd4a8 10ffmpeg/libavcodec/vble.c: 
[03:23] <CIA-18> ffmpeg: vble: remove vble_read_reverse_unary(), the code is a bit simpler this way
[03:23] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:23] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r35fd7b2968 10ffmpeg/libavcodec/vble.c: 
[03:23] <CIA-18> ffmpeg: vble: use LUT for vble_read_reverse_unary()
[03:23] <CIA-18> ffmpeg: slightly faster
[03:23] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:23] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r177bcc1ad9 10ffmpeg/libavcodec/vble.c: 
[03:23] <CIA-18> ffmpeg: vble: remove unused variable len.
[03:23] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:24] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r2fea60c600 10ffmpeg/libavcodec/vble.c: 
[03:24] <CIA-18> ffmpeg: vble: remove len array, its unneeded
[03:24] <CIA-18> ffmpeg: also remove unneeded memset()
[03:24] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:24] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * rdde0af2df1 10ffmpeg/libavcodec/vble.c: 
[03:24] <CIA-18> ffmpeg: vble: move get_bits_left() check out of inner loop, we can perform the check completely before the loop.
[03:24] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[03:24] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * ref97d59fb3 10ffmpeg/libavcodec/vble.c: 
[03:24] <CIA-18> ffmpeg: vble: remove flags copy, its not used in any speed relevant code.
[03:24] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:56] <CIA-18> ffmpeg: 03Peter Ross 07master * r070d2d7589 10ffmpeg/libavcodec/jvdec.c: 
[04:56] <CIA-18> ffmpeg: jvdec: set alpha component
[04:56] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:56] <CIA-18> ffmpeg: 03Peter Ross 07master * r341d4903a6 10ffmpeg/libavcodec/cga_data.c: 
[04:56] <CIA-18> ffmpeg: set ff_cga_palette and ff_ega_palette alpha component
[04:56] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:59] Action: Daemon404 wonders if michaelni ever sleeps
[05:01] <michaelni> sleep? whats that?
[15:38] <Cervajz> Hello :)
[15:38] <Cervajz> Is it correct that av_seek_frame and av_seek_file doesn't work for MPEG TS? :)
[15:39] <Cervajz> It works only with BYTE seeking
[16:01] <michaelni> Cervajz, non byte should work with some TS
[16:01] <michaelni> TS can have timestamp discontinuities which makes seeking by time tricky
[16:02] <Cervajz> michaelni: Which timestamp? PTS?
[16:02] <nevcairiel> in my experience, most TS work fine, except if you have a DVB recording with its daily PTS wrap
[16:49] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r8ca891fc1a 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[16:49] <CIA-18> ffmpeg: mandelbrot: more interresting zoom coordinates borrowed from wikipedia.
[16:49] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[16:49] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r50b8f9388a 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[16:49] <CIA-18> ffmpeg: mandelbrot: detect cycles to speed up interior rendering.
[16:49] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[16:50] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r3de934c873 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[16:50] <CIA-18> ffmpeg: mandelbrot: increase default maxiter to 1024
[16:50] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[16:50] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r07f272bdbb 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[16:50] <CIA-18> ffmpeg: mandelbrot: use cache to avoid recalculating points.
[16:50] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[16:50] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r6c5dd858b4 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[16:50] <CIA-18> ffmpeg: mandelbrot: Only use cycle detection if previous pixel was an interior pixel
[16:50] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[16:50] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r40d6e30c55 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[16:50] <CIA-18> ffmpeg: mandelbrot: increase default maxiter
[16:50] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:23] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r905202ef6e 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[17:23] <CIA-18> ffmpeg: mandelbrot: increase cache size
[17:23] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:23] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * rfd8d03457a 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[17:23] <CIA-18> ffmpeg: mandelbrot: increase resolution, and decrease zoom speed to maintain the cpu requirement.
[17:23] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:23] <CIA-18> ffmpeg: 03Michael Niedermayer 07master * r32dd67162a 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[17:23] <CIA-18> ffmpeg: mandelbrot: Print a note if the cache is too small
[17:23] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r0c5fd6372e 10ffmpeg/libavcodec/pngdec.c: 
[20:57] <CIA-18> ffmpeg: Do not try to read 16bit gray png files with alpha channel.
[20:57] <CIA-18> ffmpeg: FFmpeg does not support gray16a.
[20:57] <CIA-18> ffmpeg: Fixes the crash in ticket #644.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r3fd9f08971 10ffmpeg/libavcodec/bfi.c: Set BFI palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * rf98722443f 10ffmpeg/libavcodec/bethsoftvideo.c: Set Bethesda VID palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r5be097c703 10ffmpeg/libavcodec/c93.c: Set Interplay C93 palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * rd6e14e24b0 10ffmpeg/libavcodec/bmv.c: Set Discworld II BMV palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r973f3d3e0e 10ffmpeg/libavcodec/cdgraphics.c: Set CD Graphics palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r320f0f04a1 10ffmpeg/libavcodec/dfa.c: 
[20:57] <CIA-18> ffmpeg: Fix Chronomaster DFA palette.
[20:57] <CIA-18> ffmpeg: The palette had a blue cast.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * re18f70c2d1 10ffmpeg/libavcodec/dfa.c: Set Chronomaster DFA palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * rdcbd18c615 10ffmpeg/libavcodec/dsicinav.c: Set Delphine Software International CIN palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r1567243e8f 10ffmpeg/libavcodec/dxa.c: Set ScummVM DXA palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * rf628134e77 10ffmpeg/libavcodec/eacmv.c: Set Electronic Arts CMV palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * rfd09cd08c0 10ffmpeg/libavcodec/tiertexseqv.c: Set Tiertex Limited SEQ palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r3b5733bcf9 10ffmpeg/libavcodec/tiff.c: Set TIFF default grascale palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r90f75f9804 10ffmpeg/libavcodec/vmdav.c: Set Sierra VMD palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r64176bc1a3 10ffmpeg/libavcodec/kmvc.c: Set KMVC palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r3bf54ab565 10ffmpeg/libavcodec/mmvideo.c: Set American Laser Games MM palette opaque.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r0e8da94c2f 10ffmpeg/ (libavcodec/avs.c tests/ref/fate/creatureshock-avs): Increase AVS palette dynamics.
[20:57] <CIA-18> ffmpeg: 03Carl Eugen Hoyos 07master * r7b40f46fe2 10ffmpeg/libavcodec/avs.c: Set AVS palette opaque.
[21:03] <CIA-18> ffmpeg: 03Giorgio Vazzana 07master * r6024bdb21a 10ffmpeg/libavfilter/vsrc_mandelbrot.c: 
[21:03] <CIA-18> ffmpeg: mandelbrot: fix sscanf format string
[21:03] <CIA-18> ffmpeg: Replace ',' with ':'
[21:03] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:03] <CIA-18> ffmpeg: 03Dominique Leuenberger 07master * r8a25d0a860 10ffmpeg/libavcodec/j2kenc.c: 
[21:03] <CIA-18> ffmpeg: Trivial: j2kenc: fix compiler warning
[21:03] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:03] <CIA-18> ffmpeg: 03Dominique Leuenberger 07master * r05423d3afd 10ffmpeg/RELEASE: 
[21:03] <CIA-18> ffmpeg: RELEASE: bump to 0.8.6.git after the release
[21:03] <CIA-18> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[22:41] <Daemon404> michaelni, have you spent much time in IDA Pro?
[23:35] <michaelni> Daemon404, its long time ago that i last used it ...
[23:35] <Daemon404> alright, np then.
[00:00] --- Sun Nov 13 2011