--- Log for 19.05.113 Server: leguin.freenode.net Channel: #rockbox --- Nick: logbot Version: Dancer V4.16 Started: 1 day and 21 hours ago 00.01.17 Quit benedikt93 (Quit: Bye ;)) 00.02.38 # bertrik: they base not on rockbox' trunk but on a branch in paumaury's tree. So I thought gerrit isn't the right choice. 00.07.28 # hmm, there's another possible crash with AA in opus 00.14.01 Join mrtux [0] (~colin@unaffiliated/mrtux) 00.14.26 Part dys ("Killed buffer") 00.22.41 # With what? 00.30.00 # just that some allocs are not checked for failures so if you have AA big enough so the following allocs fai but not big enough so that the codec bails out 00.31.44 # Still not parsing what AA stands for. 00.31.49 # album art 00.32.03 # Oh, I see. 00.32.19 # I didn't even know you guys supported that. 00.32.27 # probelm with the ogg code is that it allocs enough memory to assemble a packet and some people put images that are several hundred k in there 00.33.09 # we don't in opus afaik but we still have to deal with that huge packet in the codec 00.33.21 # I think we have a few other small bugs in the opus wrapper layer, like we use the same code for replaygain for opus and vorbis, but opus introduced a different tag for that IIRC 00.33.35 # bertrik: Yes, we did. 00.33.38 Quit kevku (Ping timeout: 260 seconds) 00.33.50 # bertrik: Though hopefully Opus files shouldn't even have ReplayGain tags. 00.34.06 # If you're applying the header gain, that should be enough. 00.34.47 # n1s: You're using libogg, right? 00.34.57 # derf: yes 00.35.00 # I think we apply that, but we'll probably also apply vorbis replaygain tags while we shouldn;t 00.35.18 # n1s: Recent versions check all of their allocations. 00.36.17 # derf: yeah the unchecked allocs are in our main codec thing, but that's a separate issue. the libogg issue is that it allocs memory for a whole packet with is effectively unlimited in size 00.36.18 *** Saving seen data "./dancer.seen" 00.36.35 # s/with/which/ 00.36.57 # so the allocs fail and the codec bails out 00.36.58 # Okay, I see. 00.37.05 # n1s: some other codec uses setjmp to return from codec_main() in case of alloc error 00.37.39 # I guess we should resync rockbox to upstream opus at some point (now?) 00.37.40 # kugel: yeah we don't actually need that here, just a regular if(!p) goto done: would do :) 00.38.00 # okay 00.38.12 # bertrik: i am planning to do that when they release the 1.1 version 00.38.16 # n1s: Anyway, this sounds like it's not a problem that should be fixed upstream. 00.38.35 # n1s: great 00.38.36 # n1s: Did you see the ARMv5E opts that got posted to the opus list? 00.39.33 # derf: maybe not but my grasp of this ogg code isn't too good so i could use some help :), nope haven't looked at the opus list 00.39.51 # n1s: Well, if you point me where I'm happy to take a look. 00.41.48 # derf: http://git.rockbox.org/?p=rockbox.git;a=blob;f=lib/rbcodec/codecs/opus.c lines around 388 is the upper level code, it loops callign ogg_stream_pagein untill it has a full packet 00.42.18 # ogg_stream_pagein reallocs its buffer if the new page doesn't fit 00.42.37 Quit Wardo (Read error: Connection reset by peer) 00.42.39 # then when the whole comment packet is assembled it's not used for anything 00.47.12 # but if the useless packet is too large we bail out anyway, it *should* not be hard to skip it but i can't really work out a clean way to do it 00.50.00 Quit bertrik (Remote host closed the connection) 00.50.50 # Right, so after the pagein call you can call int ogg_stream_check(ogg_stream_state *os){ 00.51.00 # Err, stupid clipboard. 00.51.30 # If that returns -1, then an allocation failed and the stream has been cleared. 00.52.28 # If you wanted to keep going, you'd have to re-initialize it. 00.52.54 # Do you guys actually have a free() that works? 00.53.13 # yes, and as of a few hours ago we use it :) 00.53.21 # Amazing. 00.54.34 # Of course, if you did re-initialize it, you're going to screw up the packetno counts. 00.55.43 # (packetno isn't actually stored in the stream... it's just a count of how many packets have been returned) 00.57.02 # iiuc that could end up with cases where this particular alloc succeeds but later ones fail 00.57.18 # Yes, that's also true. 00.59.34 # Not sure what you should do about that, though. 01.02.51 # Except pre-allocate as much as possible. 01.03.25 # Though it looks like ALLOC_STACK doesn't get called until the first call to opus_decode(). 01.03.38 # Would be easy enough to add one to opus_decoder_create(). 01.03.58 # yeah that's the last one in most cases, it would be trivial to just use a static alloc for it though 01.04.21 # What are the allocations that aren't being checked? 01.07.05 # well, ALLOC_STACK isn't checked and two early ones in our top level file opus.c 345 and 348 01.08.52 # Right, though that's before you pass in pages, so this couldn't affect that. 01.09.02 # The ALLOC_STACK one is a proble, though. 01.09.10 # *problem 01.09.37 # I probably need to audit our usage of that upstream anyway. 01.09.51 # Since I added a bunch more psuedostack usage. 01.10.26 # (to the encoder, which mostly shouldn't affect you guys, but also to PLC) 01.16.38 Join bertrik [0] (~quassel@rockbox/developer/bertrik) 01.18.05 # do you have any hints on how to reinit the stream in the correct place, i don't really see it 01.20.33 # if (ogg_stream_pagein(...) < 0 && ogg_stream_check(&os) < 0) { stream_init = 0; continue; } 01.20.47 # But as I said, that will screw up your reliance on packetno. 01.24.23 # hmm 01.25.25 # i'll need to experiment some then but wouldn't just setting it to packetno=2 as in neither header, nor comment be most likely to work? 01.25.43 # Yup, that should work fine. 01.25.55 # also what happens to whatever pages are left of the current packet? 01.26.17 # will it init to packet boundary 01.26.54 # Yes. 01.27.09 # ah, cool. thanks! 01.27.18 # See the "are we a 'continued packet' page? section of ogg_stream_pagein(). 01.44.27 # <[Saint]> My word that guy is a douche... 01.44.42 # <[Saint]> "you did not help me" 01.44.46 # <[Saint]> ...bah! 01.46.10 # <[Saint]> Translation: "you did not give me the answer I wanted and suggested I do something for myself. how dare you" 01.47.05 # <[Saint]> Apparently I should keep off the forum, all I seem to do lately is piss people off. Whether justified or not, it isn't pleasant. 01.47.46 Quit bertrik (Remote host closed the connection) 01.57.21 Join CaptainKewl [0] (~captainke@207-237-110-248.c3-0.nyr-ubr2.nyr.ny.cable.rcn.com) 02.02.36 Quit n1s (Quit: Ex-Chat) 02.06.07 Quit lebellium (Quit: ChatZilla 0.9.90 [Firefox 22.0/20130514181517]) 02.28.08 # Can anyone recommend a Rockboxed MP3 player that's small light and has bluetooth. Not a phone as I would like tactile buttons. Is there usch a beast? 02.29.28 # <[Saint]> Well, we don't have a BT stack....so.... 02.29.52 # <[Saint]> Whether it has bluetooth or not is irrelevant to Rockbox. 02.34.12 Quit akaWolf (Ping timeout: 245 seconds) 02.36.19 *** Saving seen data "./dancer.seen" 02.40.49 Join pamaury [0] (~quassel@rockbox/developer/pamaury) 02.51.15 # <[Saint]> https://www.youtube.com/v/Jqu4aNbdFp8 02.53.24 # <[Saint]> whoopsie. 02.57.37 Join maffe [0] (~Miranda@77-21-169-126-dynip.superkabel.de) 02.57.56 # hi 02.58.57 # http://www.rockbox.org/wiki/DocsIndex#General_Information => Known issues => 404 02.59.11 # I know one :o) 02.59.56 # I just tried Opus with various bitrate settings on an X5L 03.00.45 # <[Saint]> OPUS ISN'T INTENDED TO BE ANYWHERE NEAR USABLE ON ALL DEVICES YET. 03.00.52 # <[Saint]> bah! caps. 03.00.57 # :o 03.01.29 # <[Saint]> There is room for a lot of optimization still. 03.01.50 # decoding actually works fine up to about ~200 kb/s on this device 03.01.58 # <[Saint]> And if memory serves the X5L doesn't exactly have the greatest SoC. 03.02.33 # buuuuut it reacts reeeaaally slow to commands like pause 03.02.43 # <[Saint]> Yes. I would expect opus to get choppy around those bitrates. 03.02.53 # <[Saint]> FOr opus, that is quite a high bitrate indeed. 03.02.55 # takes about 20 s (guessed) even with low bitrates 03.02.57 # slowly to pause or unpause? 03.03.03 # pause 03.03.11 # and opening the menu and such things 03.03.18 # skipping tracks 03.03.36 # only thing I found which works instantly is volume control 03.03.58 # <[Saint]> when using opus around those bitrates, you're really missing any advantage it could give you over any other codec I think... 03.04.17 # <[Saint]> especially in the state it is at presently. 03.04.29 # I just wanted to test which bitrate I could theoretically use 03.05.02 # 48 kb/s already sounds pretty good 03.05.58 # I tried AAC a few years ago and could only use it up to 48 kb/s or so and even then only when it's not HE 03.07.01 # <[Saint]> Opus on Rockbox, and in general, is still very much a work in progress. 03.07.31 # <[Saint]> I think both parties have code to push back and forth still yet. ANd a lot of optimization on the Rockbox side could yet be done. 03.07.54 # <[Saint]> It is a very young codec. 03.08.29 # I just wanted to give it a try and give some feedback :-) 03.09.19 # <[Saint]> Oh, yes. Sure. And it is appreciated. :) 03.10.25 # I guess I’ll stick to Vorbis for the near future 03.13.01 # [Saint]: But it's relevant to the MP3 player features I'm looking for. 03.13.25 # <[Saint]> kiwicam: not if you want to use it with Rockbox it isn't. 03.13.26 # <[Saint]> No. 03.13.41 # <[Saint]> Unless you're planning on coding up a bt stack for us? 03.15.31 # So i guess what you're saying is that even if an MP3 player has bluetooth capabilities (for sending sound to another bluetooth device, such as bluetooth headphones) Rockbox can't make use of it. 03.15.49 # <[Saint]> Yes. That's exactly what I'm saying. 03.16.32 # <[Saint]> If you just want a "dumb" BT sender, you can buy such a module and use it with anything with a 3.5mm out. 03.17.03 # Got you. SO what about an Android device. Can you play the Rockbox app via bluetooth? 03.17.11 # <[Saint]> No. 03.17.32 # <[Saint]> In theory we probably could. But we don't. 03.17.54 # Rockbox for Android O_O 03.17.57 # TIL 03.18.35 # OK. So the only way around that is to either find an MP3 player that has some good non RB software, or connect the MP3 player to an external BT transmitter. 03.19.37 # <[Saint]> There's a third option too - but Mr Someone needs to do the work. There's probably a few good candidates for bt and network stacks out there in the wild. 03.20.47 # I work in a really noise environment and wear industrial earmuffs I use a sony MW1 to get my phone calls via the mic on the MW1 and the speaker in the earmuffs. I'd love to be able to listen to a RB'd player via bluetooth. 03.22.02 # <[Saint]> I used to have to wear such muffs as well. I ended up hardwiring a pair of IEMs and a 3.5mm input to them. ...but, yeah, offtopic. 03.22.28 # It's OK. I've got enough info now. 03.23.23 # Thanks 03.40.49 Quit pamaury (Ping timeout: 252 seconds) 03.48.36 Quit Cultist (Read error: Connection reset by peer) 03.51.14 Join Cultist [0] (~Cultist@c-71-194-185-109.hsd1.il.comcast.net) 04.06.44 Quit pystar89 (Read error: Operation timed out) 04.20.29 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn) 04.20.29 Quit pixelma (Disconnected by services) 04.20.29 Quit amiconn (Disconnected by services) 04.20.29 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma) 04.20.32 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma) 04.20.32 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn) 04.24.01 Quit Cultist (Read error: Connection reset by peer) 04.25.45 Join Cultist [0] (~Cultist@c-71-194-185-109.hsd1.il.comcast.net) 04.36.05 Quit prof_wolfff (Ping timeout: 256 seconds) 04.36.20 *** Saving seen data "./dancer.seen" 05.14.14 Quit TheSeven (Disconnected by services) 05.14.23 Join [7] [0] (~quassel@rockbox/developer/TheSeven) 05.21.32 Join amayer [0] (~amayer@h223.56.21.98.dynamic.ip.windstream.net) 06.01.54 Join dfkt [0] (dfkt@unaffiliated/dfkt) 06.02.33 Quit dfkt_ (Ping timeout: 245 seconds) 06.16.31 Join pystar89 [0] (~pystar89@ip-109-90-154-150.unitymediagroup.de) 06.36.21 *** Saving seen data "./dancer.seen" 06.56.07 Join joshin [0] (~josh@ip68-0-152-212.tc.ph.cox.net) 06.56.07 Quit joshin (Changing host) 06.56.07 Join joshin [0] (~josh@unaffiliated/joshin) 07.04.10 Join kevku [0] (~kevku@2001:470:27:773:0:feed:c0f:fee) 07.12.05 Part joshin ("You can't be serious") 07.25.42 Quit shamus (Read error: Connection reset by peer) 07.26.04 Join shamus [0] (~shmaus@ip-206-192-195-49.marylandheights.ip.cablemo.net) 07.41.10 Join Scall [0] (~chat@unaffiliated/scall) 07.48.05 Quit jhMikeS (Ping timeout: 264 seconds) 07.48.50 Quit SuperBrainAK (Quit: pbly gone to sleep (-.-)Zzz...) 07.51.02 Quit amiconn (Remote host closed the connection) 07.51.02 Quit pixelma (Remote host closed the connection) 07.51.55 Join pixelma [0] (pixelma@rockbox/staff/pixelma) 07.51.55 Join amiconn [0] (amiconn@rockbox/developer/amiconn) 08.18.41 Quit CaptainKewl (Quit: ( www.nnscript.com :: NoNameScript 4.22 :: www.esnation.com )) 08.22.12 Join pretty_function [0] (~sigBART@123.252.213.15) 08.36.22 *** Saving seen data "./dancer.seen" 09.15.27 Join n1s [0] (~n1s@nl118-168-30.student.uu.se) 09.15.27 Quit n1s (Changing host) 09.15.27 Join n1s [0] (~n1s@rockbox/developer/n1s) 09.24.07 Quit maffe (Quit: IRC ist obsolet!) 09.59.22 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf) 10.14.35 Join melmothX [0] (~melmoth@unaffiliated/melmothx) 10.15.02 Quit kiwicam (Remote host closed the connection) 10.17.04 Quit amayer (Quit: Leaving) 10.32.42 Join kiwicam [0] (~quassel@101.98.163.139) 10.35.10 Join bertrik [0] (~quassel@rockbox/developer/bertrik) 10.36.24 *** Saving seen data "./dancer.seen" 10.40.57 Quit kevku (Quit: KVIrc 4.3.1 Aria http://www.kvirc.net/) 11.01.49 Quit [Saint] (Read error: Connection reset by peer) 11.18.23 Join Horscht [0] (~Horscht@xbmc/user/horscht) 11.44.53 Join stoffel [0] (~quassel@pD9E40CC0.dip0.t-ipconnect.de) 11.53.01 # hmm, skipping left when playing an opus file doesn't restart from the beginning it skips directly to the previous track 12.01.32 Quit stoffel (Read error: Operation timed out) 12.09.07 Join midnightmagic [0] (~midnightm@unaffiliated/midnightmagic) 12.15.24 Join Wardo [0] (~Mirandaha@bpb01-1-88-162-4-186.fbx.proxad.net) 12.16.00 # just seeking back to the beginning does the same 12.16.36 # in fact it skips to the next track... 12.18.21 # uh, it sometimes skips forward and sometimes backwards 12.20.02 # anyway, found a simple 3 line solution to the album art problem 12.25.18 Quit pretty_function (Remote host closed the connection) 12.29.00 # what's that? 12.29.30 # the solution or the problem? :) 12.29.49 # the problem 12.30.43 # that the ogg code allocs memory for a whole packet and the comment packet is huge when there's album art embedded so the alloc fails and it bails out 12.31.53 # storing large album art in low bitrate opus files is silly though :P 12.32.14 # yes but we still get bug reports when they don't play :) 12.33.21 # so did you modify the ogg code? 12.34.37 # i just added a ogg_sync_reset to our decode loop if it's on packetno 1 (comment packet and it works fine 12.34.52 Join kevku [0] (~kevku@2001:470:27:773:0:feed:c0f:fee) 12.35.25 # ah, yes, uh, totally… 12.36.28 *** Saving seen data "./dancer.seen" 12.39.22 Join lebellium [0] (~chatzilla@lns-c10k-ld-02-m-212-194-176-149.dsl.sta.abo.bbox.fr) 12.40.40 # i like cheesecake, too 12.54.45 Join stoffel [0] (~quassel@pD9E40CC0.dip0.t-ipconnect.de) 13.02.07 Quit liar (Ping timeout: 256 seconds) 13.02.45 Join liar [0] (~liar@clnet-p09-185.ikbnet.co.at) 13.22.11 Join pamaury [0] (~quassel@rockbox/developer/pamaury) 13.26.15 Join mortalis [0] (~mortalis@77.108.98.176) 13.38.01 Quit Horscht (Quit: quit) 13.40.30 Quit melmothX (Quit: #) 13.52.54 Quit stoffel (Ping timeout: 276 seconds) 13.53.55 Join prof_wolfff [0] (~prof_wolf@62.83.50.196.dyn.user.ono.com) 14.02.38 Join kaputnik__ [0] (~kaputnik@port-92-206-153-78.dynamic.qsc.de) 14.30.24 Nick uwe__ is now known as uwe_ (~uwe_@dslb-088-064-210-138.pools.arcor-ip.net) 14.33.06 # Build Server message: 3New build round started. Revision a17d6de, 217 builds, 22 clients. 14.36.32 *** Saving seen data "./dancer.seen" 14.39.53 # Build Server message: 3Build round completed after 408 seconds. 14.44.21 Join melmothX [0] (~melmoth@unaffiliated/melmothx) 15.20.22 Join wodz [0] (~wodz@89-75-41-78.dynamic.chello.pl) 15.23.02 Quit dokan (Read error: Connection reset by peer) 15.23.06 Join dokan_ [0] (~minatani@ac250006.ppp.asahi-net.or.jp) 15.28.35 Quit prof_wolfff (Ping timeout: 256 seconds) 15.35.44 Join krabador [0] (~krabador@unaffiliated/krabador) 15.38.49 Quit akaWolf (Read error: Connection reset by peer) 15.39.13 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf) 16.02.32 Join wodz_ [0] (~wodz@89-75-41-78.dynamic.chello.pl) 16.02.45 Quit wodz_ (Client Quit) 16.04.14 # Zagor: (log) please take a look if http://pastie.org/7929468 makes sense. This hopefully will send build results to the clients. 16.05.44 # wodz: you could put it on gerrit and have that mail Zagor 16.06.06 # gevaerts: good idea 16.06.09 # will do 16.06.34 # People tend to forget that gerrit isn't just for the main rockbox repository :) 16.07.30 # gevaerts: hmm fatal: remote error: access denied or repository not exported: /www 16.07.40 # I have no idea what this means 16.09.07 # Did you do the necessary gerrit magic for that repository? 16.09.17 # ha, probably not 16.09.48 # Make sure you do the necessary s/rockbox/www/ in those :) 16.10.15 # sure 16.14.02 # g#474 16.14.54 # * gevaerts adds Zagor as a reviewer 16.20.04 Join maffe [0] (~Miranda@77-21-33-55-dynip.superkabel.de) 16.28.34 Quit DexterLB (Read error: Connection reset by peer) 16.33.51 Join DexterLB [0] (~dex@46.10.53.86) 16.36.35 *** Saving seen data "./dancer.seen" 16.37.38 Join pretty_function [0] (~sigBART@123.252.213.15) 16.38.14 Quit krabador (Ping timeout: 260 seconds) 17.00.14 Quit pamaury (Ping timeout: 240 seconds) 17.02.19 Join stoffel [0] (~quassel@pD9E40CC0.dip0.t-ipconnect.de)