[Ffmpeg-devel-irc] ffmpeg-devel.log.20140801

[burek]

[00:03] <wm4> hm is there any kind of specification on raw pgs files?
[00:04] <j-b> wm4: hahaha
[00:04] <j-b> I tried
[00:04] <j-b> it was a mess
[00:04] <wm4> :(
[00:05] <j-b> they used the same names of the old ones, of course, with different padding
[00:06] <Timothy_Gu_> michaelni: I really can't reproduce the bug. The very light green and yellow and messed-up table header styles suggest a cache problem. Try Ctrl+F5 on the fate page.
[00:13] <Timothy_Gu_> michaelni: or you can try to access tg-fate-test.ngrok.com, which is an ngrok tunnel to my laptop showing a mock fate page
[00:16] <michaelni> Timothy_Gu_, http://tg-fate-test.ngrok.com/ looks better
[00:17] <michaelni> Timothy_Gu_, seems it was a cache issue
[00:17] <Timothy_Gu_> I knew it ;)
[00:18] <Timothy_Gu_> so do you think the colors are OK or they need more adjustments?
[00:19] <michaelni> dunno but ill push it (well its already on fate.ffmpeg.org but not pushed to origin master)
[00:19] <michaelni> we/you can finetune the color later
[01:10] <llogan> i'll be gone friday-sunday sometime, so i'll be unavailable for mailing list stuff.
[01:56] <cone-766> ffmpeg.git 03Lou Logan 07master:d76675d9c015: doc/indevs: mention required configure options
[03:59] <BBB> ubitux: looks ok to me (e use ptrdiff_t not intptr_t total nag)
[04:58] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:d98e6c5d5d80: pgssubdec: Check RLE size before copying
[04:58] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:7db3a46a3d9d: Merge commit 'd98e6c5d5d80c1dfe0c30f2e73d41a3aea0b920d'
[08:31] <wolfie_> hi all
[09:40] <ubitux> ePirat: sorry about the delay for the icecast patch, i'll try to review & test on sunday
[09:40] <ubitux> ePirat: poke me saturday night or on sunday if i don't give any news
[10:59] <ePirat> ubitux, no problem, I had no internet for two days anyway due to thunder/lightning 
[12:37] <cone-766> ffmpeg.git 03Benoit Fouet 07master:66af2a01d275: h264: remove useless assignment.
[12:37] <cone-766> ffmpeg.git 03Benoit Fouet 07master:42f25fe00991: h264_mp4toannexb_bsf: always set idr_sps_pps_seen when SPS/PPS is seen.
[12:51] <cone-766> ffmpeg.git 03Carl Eugen Hoyos 07master:970c7d9b695a: Add forgotten APIchanges entry for probesize2.
[12:51] <cone-766> ffmpeg.git 03Carl Eugen Hoyos 07master:a80191379bb9: Fix standalone compilation of the WebVTT decoder.
[14:23] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:9e052df86d52: avfilter/vf_pullup: 2nd try to workaround gcc 4.4.3 bug on arm
[14:38] <wm4> michaelni: why bother???
[14:38] <Case> michael is a good guy who thinks about users
[14:39] <Case> or developers
[14:39] <wm4> even gcc apparently abandoned this release
[14:39] <wm4> and who the heck needs lavfi on arm...
[14:39] <ubitux> to deinterlate telecined content ofc
[14:39] <ubitux> deinterlace*
[14:41] <wm4> if "they" are stuck on gcc 4.4, I don't even want to know what environment that is
[14:49] <michaelni> wm4, why i bother to workaround/fix issues ? because id like things to work, the env is ubuntu 10.04 IIRC
[14:50] <wm4> randomly rearranging code to evade bugs in very old compilers is ugly
[14:50] <wm4> if ubuntu folks want to use/support that old crap, they should backport gcc bugfixes
[14:54] <michaelni> wm4, yeah, you are welcome to fix bugs, trolling that other do work, not so much
[14:55] <michaelni> unless you volunteer to maintain vf_pullup.c ?
[14:56] <michaelni> and if you want it backported, iam sure the gcc developers and ubuntu people would not be against a patch  either
[14:57] <wm4> if "fixing" it requires making the code less readable, I'm not sure if the fix is even in your own interest
[14:57] <michaelni> is it less readable ?
[14:58] <wm4> -#define ABS(a) ((a) > 0 ? (a) : -(a))
[14:58] <wm4> +#define ABS(a) (((a) ^ ((a) >> 31)) - ((a) >> 31))
[14:58] <wm4> (the new code even triggers UB, I think?)
[15:01] <michaelni> but its not new, that was the code before the 1st try to fix it
[15:01] <michaelni> just reverting what did not work
[15:02] <wm4> I see
[15:09] <cone-766> ffmpeg.git 03Diego Biurrun 07master:706208ef47bf: fate: Split fate-pixdesc tests and dispatch them through Make
[15:09] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:713dbe0655e4: Merge commit '706208ef47bffd525c982975d2756f7b2b220b8d'
[15:16] <cone-766> ffmpeg.git 03Diego Biurrun 07master:03c9f357a4c2: ppc: idctdsp: Immediately return if no AltiVec is available
[15:16] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:ad99837b638f: Merge commit '03c9f357a4c2307a7913cea2cbf0ba817e80beb6'
[15:21] <kierank> hmm so it's not swr crashing but something else in audio
[15:23] <kierank> or perhaps there is some asm i am missing
[15:23] <kierank> to delete from swr
[15:24] <kierank> oh wait fount some more
[15:24] <ubitux> libswresample/x86/*_init.c :p
[15:34] <kierank> crash appears to be in float dsp somewhere
[15:37] <trn> Hello.
[15:38] <trn> Quick question on using conditional formatting/evalution to the scale filter, seems x264 wants an even number.
[15:38] <nevcairiel> sounds like you want #ffmpeg, this channel is about developing ffmpeg itself
[15:39] <trn> nevcairiel: You are correct, I looked too soon.
[15:41] <trn> nevcairiel: But I also spoke too soon and see it's easy enough to solve my issue with a shell script and that can make the test script match exactly how I'm calling libavfilter anyway. :b
[15:41] <trn> so nevermind
[16:24] <cone-766> ffmpeg.git 03Diego Biurrun 07master:a786c8259daf: idct: Split off Xvid IDCT
[16:24] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:f54e01c24e03: Merge commit 'a786c8259dafeca9744252230b5d78f67810770c'
[16:29] <cone-766> ffmpeg.git 03Diego Biurrun 07master:5735552f1f17: pngenc: Drop pointless pointer cast in png_write_row()
[16:29] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:96cea6269ac3: Merge commit '5735552f1f17ea01dcbc99b08f54b5bf52176a8f'
[16:39] <cone-766> ffmpeg.git 03Diego Biurrun 07master:4da8cdbb91dd: tscc: Eliminate pointless variable indirections in decode_frame()
[16:39] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:79d0e164cfef: Merge commit '4da8cdbb91ddbac118b79076cad4dc28ba72e86f'
[16:45] <cone-766> ffmpeg.git 03Vittorio Giovara 07master:52269f48e835: Revert "vf_interlace: deprecate lowpass option"
[16:45] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:c87f645a757e: Merge commit '52269f48e835a52023656f8330262ae70e6061c4'
[16:55] <cone-766> ffmpeg.git 03Vittorio Giovara 07master:92a36a6b3338: pngdec: correctly indent macros
[16:55] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:5727d387fb24: Merge commit '92a36a6b3338df5f5cff573a486e182005367c24'
[17:03] <cone-766> ffmpeg.git 03Luca Barbato 07master:b2e059a1ffbd: pixdesc: K&R formatting cosmetics
[17:03] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:229a1e869429: Merge commit 'b2e059a1ffbdaaa985c6a7dcdd684fa034725238'
[17:56] <ubitux> i need some gnu make guru; i'm trying to do something very stupid so i hope you will encourage me
[17:56] <ubitux> i have a var N such as N=4, and i want a X=1<<$N
[17:56] <ubitux> in shell, you can do $((1<<$N)) to obtain the desired value
[17:57] <ubitux> and in make you can $(shell ...)
[17:57] <ubitux> but i'm unable to merge the two together
[17:57] <ubitux> anyone willing to give it a try? :)
[17:57] <cone-766> ffmpeg.git 03Vittorio Giovara 07master:7ab551f9fd9a: h264: prevent theoretical infinite loop in SEI parsing
[17:57] <cone-766> ffmpeg.git 03Michael Niedermayer 07master:293124cae80e: Merge commit '7ab551f9fd9a63586649a7df8790ddaeac55420f'
[17:58] <smarter> I think you need to escape $ as $$
[17:58] <ubitux> mmh ok it seems i simply need to $(shell echo $$((1 << $N)))
[17:58] <ubitux> smarter: yeah, i was trying to escape with \
[17:58] <ubitux> smarter: thx :)
[18:05] <cone-766> ffmpeg.git 03Benoit Fouet 07master:f9bd6d61bcc1: h264_mp4toannexb_bsf: rename first_idr to new_idr
[21:20] <cone-789> ffmpeg.git 03Michael Niedermayer 07master:7f852962dc6a: avcodec/mpeg12dec: print the amount of overread
[21:20] <cone-789> ffmpeg.git 03Michael Niedermayer 07master:f95298c91389: avcodec/mpeg12dec: fix support for interlaced mpeg2 with missing last slice
[21:43] <cone-789> ffmpeg.git 03Luca Barbato 07master:fa14804c83a7: flv: Index the audio stream
[21:43] <cone-789> ffmpeg.git 03Michael Niedermayer 07master:afd62b3f1845: Merge commit 'fa14804c83a7108a50c63d1f2180e30c75909529'
[00:00] --- Sat Aug  2 2014