Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2009-04-23

00:02:36 Join wincent [0] (
00:03:49 Quit bertrik ("Leaving")
00:04:09MTsaratoga : Hi :) .. did you see the logs or the project's wiki ?
00:04:16 Quit BeholdMyGlory (Remote closed the connection)
00:07:29 Quit Toored (Read error: 110 (Connection timed out))
00:07:40 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
00:08:32 Quit stripwax ("")
00:09:40JdGordon|Bagder: ping?
00:10:45JdGordon|Zagor: can you fiddle with build servers or only Bagder ?
00:11:38ZagorJdGordon|: I can but I tend to leave it to bagder. he knows the system better.
00:12:35JdGordon|can you tell me what its expecting for my box so i can set ip up properly? (ip/port?)
00:14:07ZagorJdGordon|: you know of , right?
00:14:34Zagoras for ip/port, anything that has ssh up is fine
00:15:05 Quit PyroBor (Remote closed the connection)
00:16:17 Quit XavierGr (Remote closed the connection)
00:18:00 Join hillshum [0] (n=hillshum@unaffiliated/hillshum)
00:19:16hillshumWas the e200v2/fuze scrollwheel fixed already?
00:19:22JdGordon|Zagor: yeah, I mean, its all setup already, but i've changed ip's and stuff and will just setup ports again the way its setup for already on the buildmaster
00:20:14*JdGordon| assumes the build server list isnt public other than whats beeing shown on that wiki page?
00:21:43 Join CaptainKwel [0] (i=2669ecc2@gateway/web/ajax/
00:21:48Zagorwell, it's not exactly a secret: :-)
00:25:00MTsaratoga , linuxstb : I'm testing the decode function, Is there a way to play the raw decoded audio frames ?
00:25:35MTor maybe any way for checking whether the output is correct ?
00:29:19linuxstbMT: I would use the "play" utility (part of sox)
00:29:48 Join BryanJacobs [0] (n=braujac@
00:30:35 Quit Zagor ("Clint excited")
00:30:49MTlinuxstb : will try it, thanks.
00:31:38 Quit bluebrother ("zzz")
00:32:10 Quit flydutch ("/* empty */")
00:34:21***Saving seen data "./dancer.seen"
00:34:32 Quit CaptainKwel (" ajax IRC Client")
00:35:23 Quit midijunkie ("?(???~~)?")
00:36:10 Quit gevaerts (Nick collision from services.)
00:36:41 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
00:37:41 Nick fxb is now known as fxb__ (
00:39:20 Quit robin0800 (Remote closed the connection)
00:42:06 Join robin0800 [0] (
00:43:09 Quit lymeca (Read error: 113 (No route to host))
00:43:12 Quit hillshum ("Leaving")
00:51:51 Quit evilnick_7 (" ajax IRC Client")
00:57:04 Quit Lynx_ (" Try HydraIRC -> <-")
00:59:38 Quit tvelocity (Remote closed the connection)
01:01:00 Quit kugel ("ChatZilla 0.9.84-rdmsoft [XULRunner 1.9/2009030810]")
01:01:25 Quit evilnick_230 (" ajax IRC Client")
01:02:21 Quit Seed ("cu, Andre")
01:04:47 Join bs66_ [0] (
01:05:22 Quit bs66_1 (Read error: 110 (Connection timed out))
01:06:00 Quit barrywardell ()
01:08:34 Part mcuelenaere ("Leaving.")
01:10:41 Join kugel [0] (i=kugel@rockbox/developer/kugel)
01:14:02 Join _Auron_ [0] (
01:18:26 Quit robin0800 ("No Ping reply in 90 seconds.")
01:19:00 Join robin0800 [0] (
01:19:03 Join krazykit [0] (n=kkit@
01:19:39 Part krazykit ("Connection reset by beer")
01:20:38 Join steeven [0] (n=sergei@unaffiliated/tuxmp)
01:21:31steevenI've read a lot about rockbox but I'm having trouble figuring something out
01:23:20kkurbjunwhat's the trouble?
01:23:55steevenwith rockbox on a capable fifth generation+ ipod, will the device lose the ability to send digital audio to a car audio receiver?
01:24:58steevenI can't really find much information on this
01:24:59kkurbjunthat depends, you can always boot back into the apple firmware, rockbox supports some accessories, but I can't say for sure whether your receiver will work
01:26:01steevenwell, some receivers have the capability to control an ipod and play audio from it using the receiver's own DAC
01:28:14kkurbjunYou might try looking at the ipod accessory page in the wiki, I'm not really sure if it would work in rockbox or not, but if it doesn't you can always go back to the apple firmware
01:28:35kkurbjunyou just hold the menu button I believe while it is booting to go back to the apple software
01:29:26steevenwell, the problem is I don't have an ipod or an ipod-compatible receiver
01:30:08steevenI'm only interested in getting them if I can easily play lossless audio sent to the receiver digitally
01:30:13BryanJacobssteeven: is "digital audio" all that important?
01:30:44BryanJacobsyou could just use the line out - the cable you'd be using is only a few feet long tops
01:30:54steevensure is, no portable music player has a DAC designed to power an enthusiast audio system
01:31:10steeventhe problem is the line out is designed for headphones
01:31:16BryanJacobsyou don't have an amp in an "enthusiast" system?
01:31:28 Quit Thundercloud (Remote closed the connection)
01:31:29BryanJacobsno, the line out is not designed for headphones
01:31:37BryanJacobsline out is preamped DAC output
01:32:19scorchesteeven: many a device's DAC is just are either thinking of the amp, or you are misinformed
01:32:39steeventhe line out? is there some kind of adapter cable that gives you that connection on an ipod?
01:33:05 Join Thundercloud [0] (i=thunderc@
01:33:19BryanJacobssteeven: it's on the cradle
01:33:46 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
01:33:57steevenwell like I said, I don't have an ipod yet ;)
01:34:21steevenwhere can i find out the voltage of this connector?
01:34:35BryanJacobssteeven: not sure you understand the problem
01:34:49 Join BlakeJohnson86 [0] (
01:35:03scorchesteeven: most of the headunits you are talking about either communicate with the ipod through what we call advanced features of the apple accessory protocol (which we havent implemented yet, though more simple AAP bits work), or simply use the DAP as a portable hard drive and just read the files off it
01:35:48BryanJacobssteeven: what I was talking about is a system with an "aux in" jack to which you may connect the ipod's line out; such a system has its own amplifier and so the input voltage doesn't matter.
01:36:30 Quit timc (Remote closed the connection)
01:36:37 Quit Thundercloud (Remote closed the connection)
01:37:17steevenso I should just disregard the fact that the system is designed for 2V aux inputs and pretend like giving it a 200mV signal is going to be close to adequate?
01:38:19BryanJacobsyou're saying your amp has a *minimum* input?
01:38:43BryanJacobsit's not powered?
01:39:09steevenI'm saying it has a specific sensitivity to input and thus has an optimal input range
01:39:47steevenhehe, maybe the connector is exactly what I need. I'm not saying it isn't
01:40:11steevenit's kind of odd that the manufacturer doesn't specify this on their website though
01:41:06BryanJacobsto answer your original question: if your system has an iPod connector which uses the DAP as a UMS device, then rockbox will work fine. If it has a connector using the AAP, Rockbox probably won't work yet. If it has an aux in, you MIGHT <I'm skeptical> need a preamp but it will work
01:42:00steevenkeep in mind that if it's extremely low voltage it will be more susceptible to noise from the mess of cables it'll be required to run next to
01:42:16steevenwell, that's helpful, thank you
01:42:25BryanJacobssteeven: no problem
01:42:28steevenis AAP support something that's being actively developed?
01:43:07scorcheas i said, simple functionality of the AAP has been implemented...advanced features havent been
01:43:07steevenit's seeing quite widespread adoption in automotive audio systems
01:43:34steevenoh, I forgot to link the receiver!
01:43:38 Join Traveler2 [0] (
01:44:36Traveler2I took a look at plugin.h and am wondering if strncpy stops reading at nextline?
01:44:54BryanJacobssteeven: hmm, says it requires an FAT iPod
01:45:28steeventhe frustrating part is the fact that the receiver can decode ALAC from an iPod, but *not* from a usb mass storage device (which can be directly plugged into the unit)
01:46:02scorchesteeven: that says to me that it uses the advanced features of AAP (wont work with rockbox), but will also read the files from the hard drive (will work)
01:46:13MTlinuxstb : sorry, tried a little with sox play but can't play the raw decoded data.
01:47:00steevenin that case I'd be much better off with a 1.5tb hard drive, an enclosure, and a 12v power supply with inline surge suppressor
01:47:00Unhelpfulcaring about the format makes me think MSC
01:47:02MTI converted it to wav using sox and then played the output file.
01:47:23steevenI could get two for the price of an iPod, heh
01:47:38BryanJacobssteeven: this receiver has three modes, see page 16 of the manual
01:48:13MTlinuxstb : but it's just rubbish, so I want to try playing the decoded data directly without conversion ..
01:50:19steevenBryanJacobs: the depressing part is that it doesn't say it can decode ALAC from any input besides an iPod
01:50:47BryanJacobssteeven: it doesn't say it can do ALAC at all
01:51:17BryanJacobssteeven: and it explicitly says it does NOT support WMA lossless
01:51:42 Quit [8bit]bubsy ("Don't panic.")
01:51:47BryanJacobssteeven: in fact, it lists AAC max bitrate as 320k
01:51:55steevenALAC isn't really AAC
01:51:57steevenhold on a second
01:53:18steevenhrm trying to find my source on that. oh please be a feature that actually exists XD
01:53:27 Join apo_ [0] (
01:53:32scorchewe know the difference between ALAC and AAC...
01:53:38steevenit was touted at when it was introduced at CES
01:53:39 Quit JdGordon| (" ajax IRC Client")
01:53:46 Quit pixelma (Nick collision from services.)
01:53:46 Join pixelma_ [50] (n=pixelma@rockbox/staff/pixelma)
01:53:48 Quit amiconn (Nick collision from services.)
01:53:53 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
01:54:01 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
01:54:02steevenscorche: of course, I was just saying that "320kbps max AAC" doesn't necessarily mean "no ALAC"
01:54:06 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
01:54:12scorchesteeven: if it doesnt list it as supported, then it likely cant read those files...
01:54:20scorchesure it does
01:54:31BryanJacobssteeven: "The unit can control a Portable audio player with the USB interface. Playable audio file formats are MP3 and WMA."
01:54:43scorcheit says "MP3/AAC/WMA Playback"
01:55:13BryanJacobssteeven: "The unit can play back “mp3,” “wma” or “m4a” file extensions."
01:55:28scorchethats what i said
01:55:29 Quit faemir ("Leaving")
01:55:44steevenundocumented feature :v
01:55:48BryanJacobsscorche: just quoting the manual directly
01:56:09BryanJacobssteeven: don't bet on it
01:59:00BryanJacobssteeven: googling "alac iDA-X305" gives three results; "lossless iDA-X305" gives more but filled with speculation. I suggest you just use the aux in and give up on the all-digital audio path idea... Are you planning on using this while parked in a soundproof garage?
01:59:18 Quit HellDragon (Client Quit)
01:59:35scorchethis conversation has stopped being about rockbox, and is better off moved somewhere else...
02:00:01steevendoes rockbox work with the cowon q5w
02:00:42scorchesteeven: is it listed on the front page?
02:00:59 Join HellDragon [0] (
02:01:13 Part toffe82
02:01:35steevenhmm, I'll ask less stupid questions ;)
02:01:54steevendoes rockbox work on any device with spdif output? :)
02:02:09scorcheiriver H100 series
02:02:26BryanJacobsscorche: are those still for sale?
02:02:32scorchesure...on ebay...
02:03:42steevenhmm, interesting.
02:04:45 Quit apo (Read error: 110 (Connection timed out))
02:07:21steevenalpine made a stereo receiver with toslink inputs... and the h100 series with rockbox can play FLAC
02:07:54steevenwhich could mean no painstaking transcoding of my entire library
02:08:45 Quit MT (Read error: 104 (Connection reset by peer))
02:09:32steevento be honest, i understand the tiny return on investment of playing lossless music in an environment with such a high noise floor
02:10:30steevenI just want a single format :)
02:11:51 Join Rob2222 [0] (
02:12:10BryanJacobssteeven: heh, my summer of code is about this
02:12:21steevenoh yeah?
02:12:34steevenmusic specifically?
02:12:39BryanJacobssteeven: Wavpack hybrid coding for Rockbox
02:12:55BryanJacobsright now the lossy portion works
02:18:17steevenah there's the catch: iriver h140s are several hundred dollars, alpine toslink DSP is ~$600, compatible head unit is $600, and it's all six year old technology XD
02:18:21 Quit robin0800 ("No Ping reply in 90 seconds.")
02:20:00 Join robin0800 [0] (
02:20:17steevenokay, I'm still totally on-rockbox-topic here
02:21:17scorchesix year old technology doesnt mean it is bad...
02:21:31scorcheeither way, it seems you have a choice to make...
02:22:45steevenI want to get a handle on the options first. I suppose digital isn't that important if I can get a good preamplified signal to the receiver
02:22:50steevenso my next question is
02:23:13steevenis there a device rockbox works with that can act as a usb host to a removable storage device such as a external hard drive?
02:23:47BryanJacobssteeven: even my PDA can do that
02:23:48scorcherockbox doesnt yet support USB host...there are devices capable of it, but no one has made it work yet
02:24:09BryanJacobsoh, you mean "such that rockbox is the host to..."
02:24:47Unhelpfullook for an adapter with a long ribbon that connects to a microsd "card" :P
02:26:14steevenah, I see.
02:28:46 Quit Rob2223 (Read error: 110 (Connection timed out))
02:30:23 Join lymeca [0] (
02:34:24***Saving seen data "./dancer.seen"
02:38:08Traveler2why does read_line() return an integer?
02:46:28Traveler2more importantly, is there a method that returns whether there is another line in a fie?
02:48:31UnhelpfulTraveler2: i'd expect you do that by *trying* to read another line.
02:48:55BryanJacobsTraveler2: and that answers your first question :-)
02:50:52Traveler2so, does read_line return -1 if there isn't another line?
02:51:42BryanJacobsTraveler2: returns non-true, IIRC
02:52:22BryanJacobsTraveler2: ahh, here we are, apps/misc.c line 155: if an error occurs, returns -1
02:53:07Traveler2alright, thanks :D
02:58:46*kugel would expect EOF as return, but well
02:59:15kugelof course, EOF may happen to be -1
02:59:24steeventhanks again folks
02:59:25 Part steeven
02:59:47*BryanJacobs would just use if (!read_line(foo))
02:59:58kugelthat doesn't work as you see
03:00:18BryanJacobsah, yes
03:00:24BryanJacobsif (read_line(foo) <= 0)
03:00:32kugelunless you mean "if (read_line(foo)) printf("there's another line");"
03:00:59kugel<= ?
03:01:22BryanJacobs<= 0 catches the "I read 0 bytes because nothing's left" and the "there's an error" cases
03:01:41*BryanJacobs assumes the file has no blank lines
03:02:05kugelif it's 0, it already reached EOF before that call, which possibly means your check was unsufficient in the call before :)
03:02:42EternalRainsIn WPS, is the Shuffle (%ps) conditional always set when the Repeat mode (%mm) is set to shuffle?
03:02:52BryanJacobskugel: <shrug> any time the call returns >0 you may use the result
03:03:22kugelchecking for EOF gives better readable code though
03:03:37*BryanJacobs agrees
03:03:51*kugel is off, bye
03:03:53 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]")
03:06:48 Quit daurn (Remote closed the connection)
03:06:52 Quit fyrestorm (Read error: 104 (Connection reset by peer))
03:08:48 Quit robin0800 ("No Ping reply in 90 seconds.")
03:09:39 Quit midkay (Read error: 60 (Operation timed out))
03:11:24 Join robin0800 [0] (
03:13:12 Join CaptainKewl [0] (
03:18:25 Join jmillikin [0] (
03:24:10 Quit robin0800 (Remote closed the connection)
03:25:34 Quit Traveler2 ("Java user signed off")
03:27:40 Join fyrestorm [0] (
03:57:24 Quit lymeca (Success)
04:00:16Unhelpfulamiconn: fyi r20710 vanilla does not appear to decode chroma-subsampled jpeg correctly - this would correspond with what the confusing segment of code i referenced earlier. i'll try to make sure that the core decoder handles this correctly, but i'm not sure how best to fix the plugin decoder without it getting a bit more bloated
04:01:46 Quit Conic (Read error: 104 (Connection reset by peer))
04:03:41 Quit EternalRains (Read error: 104 (Connection reset by peer))
04:13:52 Join gregorovius [0] (n=diego@
04:14:40 Join robin0800 [0] (
04:17:06 Quit gregorovius (Client Quit)
04:34:25***Saving seen data "./dancer.seen"
04:35:59 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]")
04:41:13 Quit dmb (Read error: 104 (Connection reset by peer))
04:42:14 Part wincent ("Kopete 0.12.7 :")
04:45:40 Join StealthyXIIGer [0] (
04:46:28 Quit RoC_MasterMind ("Leaving")
04:49:19 Quit robin0800 (Read error: 110 (Connection timed out))
05:00:14 Quit CaptainKewl ("( :: NoNameScript 4.02 :: )")
05:06:14 Join robin0800 [0] (
05:06:59 Join RoC_MasterMind [0] (
05:10:00 Quit robin0800 (Read error: 104 (Connection reset by peer))
05:16:51 Join robin0800 [0] (
05:22:17 Quit l401 (Read error: 113 (No route to host))
05:38:01 Quit Horscht ("Verlassend")
05:46:40 Quit RoC_MasterMind (Read error: 104 (Connection reset by peer))
05:47:02 Join RoC_MasterMind [0] (
05:50:01jmillikinHow do I find out how much RAM is in use, in the simulator? I'm working on task 7287 (sort-tags), and I want to make sure the sort indexes aren't being loaded during normal browsing.
05:51:00 Quit RoC_MasterMind (Remote closed the connection)
05:53:43Unhelpfuljmillikin: i think a better solution might be to DEBUGF the filename when an index file is loaded...
05:54:22Unhelpfulalso, thanks for your work on this, it's a feature i've wanted and not been motivated or knowledgeable enough to do :)
05:57:43jmillikinHappy to help, I'm glad the RB source is so clean that I can be useful with only a few days of experience in it.
05:58:41 Join ajonat [0] (n=ajonat@
06:01:04 Nick alexbobp is now known as socat (
06:05:45 Nick socat is now known as alexbobp (
06:08:39 Join timc [0] (n=aoeu@
06:28:10 Join bubsy [0] (i=Bubsy@
06:34:27***Saving seen data "./dancer.seen"
06:36:28robin0800anyone know if this is the sansa linux USB problem
06:37:15LloreanThis is #Rockbox, why are you asking a sansa linux question here?
06:38:41robin0800Lloren it is about Rockbox but an e200 not c200 please read
06:39:49LloreanYour question isn't about Rockbox...
06:39:55 Join laburd [0] (
06:40:19LloreanYou said "anyone know if this is the sansa linux USB problem"
06:40:31laburdHello everyone
06:40:33LloreanDo you mean "Anyone know if this is the problem the Sansa has with Rockbox USB in Linux"?
06:40:44laburdI have sansa c240 using linux
06:40:48laburdyou want me to test something
06:42:51robin0800Llorean: it is the rockbox usb mode with Jaunty not working in gnome
06:43:28Lloreanrobin0800: That thread seems to make plenty of reference to Rockbox.
06:43:35LloreanWhat, exactly, are you asking?
06:44:28robin0800Llorean: Would this also apply to c200
06:45:12LloreanThe c200 has a different set of USB IDs.
06:45:25LloreanIt's possible that the same bug exists there, but there's nothing in that thread that mentions it.
06:45:34laburdI can try something
06:45:36laburdI am running jaunty
06:45:41laburdYou want me to test c240 ?
06:46:26Unhelpfulrobin0800: the problem described there looks to me like a distribution problem, not a rockbox problem.
06:47:32 Join hittudiv [0] (n=hittudiv@
06:48:35robin0800Llorean: I don't think its the same unless we have The wrong USB Id's remembe c200 OF does work
06:48:55LloreanI didn't know the c200 OF worked.
06:49:11LloreanIf it's different, and you don't think it's the same problem, there wasn't much point in asking.
06:51:49robin0800Llorean: only the description is exactly what I see and trying many times it might evently connect
06:54:06 Quit hittudiv ("Leaving")
06:55:57tmztrockbox uses libmtp now
06:56:01tmztnot rockbox
06:56:24 Part BryanJacobs
06:57:36tmztrobin0800: I think we discussed this, it's a hal problem in ubuntu since rockbox shares usb vend/prod with OF
07:01:28robin0800tmzt: I you certain its right for the c200 because The OF works
07:02:28tmztthe OF works because MTP is enabled I assume, rockbox does not currently support MTP/PTP
07:02:49tmztare you saying e200 does not work in OF?
07:03:06robin0800tmzt: No msc
07:03:19tmztnot Auto?
07:03:46robin0800tmzt: see this
07:03:49tmztyou can check the hal logs but this is really a ubuntu problem
07:04:45tmztThe USB product ID used by Sansa e200 running in UMS/MSC mode, both by Rockbox as well as original firmware is 0x7421. The PID for MTP mode is 0x7420
07:05:06tmztwhich does lsusb report in rockbox?
07:06:55tmztand what does it report in OF
07:11:54robin0800tmzt: Yes I'm at work at present but now I know what might be wrong I'll check and report back later today I think when it dosn't connect lsusb can't read it
07:12:56robin0800tmzt: I have two now C240's
07:13:15tmzteven if libusb is using it should show in lsusb, if not that's probably a different problem
07:13:56tmztthe patch doesn't exactly make sense either, it doesn't seem like it should affect mass storage mode since that uses a different vend/prod
07:15:03robin0800I think in E200's case the usb id's are wrong in ubuntu
07:15:26 Join AndyIL [0] (i=AndyI@
07:15:35 Quit jmillikin (Read error: 60 (Operation timed out))
07:19:06 Join jeffdameth [0] (
07:22:09 Quit robin0800 (Remote closed the connection)
07:27:11amiconnUnhelpful: In what way does it look wrong? I didn't notice problems when viewing jpegs so far?
07:27:15 Quit AndyI (Read error: 110 (Connection timed out))
07:31:35Unhelpfulamiconn: i get a garbled mess instead of an image. see (i know the sample images are low-res, but they're quite clearly broken on my beast with r20608 and ipod video sim with fresh svn)
07:32:38 Join robin0800 [0] (
07:33:43 Quit jeffdameth1 (Read error: 110 (Connection timed out))
07:35:01 Join JdGordon| [0] (i=43a02c5a@gateway/web/ajax/
07:35:26robin0800tmzt: Usb id's are different in C200 0781 7450 Rockbox 0781 7451 OF
07:37:19amiconnUnhelpful: Hmm, never seen that with other images, although chroma subsampling is common afaik
07:40:02 Join theredone [0] (n=47b63257@gateway/web/cgi-irc/
07:40:41 Join Seed [0] (
07:40:58Unhelpfulamiconn: i thought it was supposed to be common, too. all of those are compressed with ijg cjpeg, and display correctly in a PC image viewer (feh). chroma subsampling is supposed to be a rather large portion of the compression gain for color images.
07:41:06theredonehey installing the new version will that delete my files
07:41:39Unhelpfuli have difficulty believing that either cjpeg is creating broken files or that nobody's noticed this before, so i'm kind of glad somebody else can see it. ;)
07:42:08amiconnThey're showing properly in FF3, windows image viewer and irfanview as well
07:42:45amiconnAn X5 sim only shows the non-subsampled image properly.
07:43:27amiconnWhat quality setting did you use in cjpeg?
07:43:55amiconnIirc there's a warning in the docs that very low quality settings would create coefficients not every decoder will understand
07:43:59Unhelpfuli would expect this problem bites grey and mono targets as well, since i'm pretty sure they use the same logic to identify chroma blocks (for skipping)
07:44:08robin0800Llorean: : Usb id's are different in C200 0781 7450 Rockbox 0781 7451 OF
07:44:26theredonehey installing the new version will that delete my files please help
07:44:30Lloreanrobin0800: You already mentioned that in the channel
07:44:38Unhelpfulcjpeg -quality 75 -sample 1x1,2x2,2x2 (or whatever else)
07:44:40Lloreantheredone: No, it won't, or we'd tell you it did.
07:44:51theredoneok thanks
07:45:32 Quit robin0800 ("No Ping reply in 90 seconds.")
07:45:57 Join robin0800 [0] (
07:46:45 Quit theredone ("CGI:IRC (EOF)")
07:48:02 Quit jhMikeS (Nick collision from services.)
07:48:08 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
07:49:27amiconnUnhelpful: Hmm, the warning applies to quality < 25 (see man page)
07:49:31 Join israr [0] (n=israr@
07:49:36Unhelpfulimagemagick, at least, does *not* subsample at default (or many typical) settings. i can't speak for any of the things that windows users might be using to batch scale+convert.
07:50:04amiconnWell digicams usually write subsampled files
07:50:24*amiconn would need a tool to check subsampling of existing pictures
07:50:39 Quit JdGordon| (" ajax IRC Client")
07:52:34Unhelpfulthey easiest way i've found is either the debug output in that patch, or djpeg -verbose $file >/dev/null
08:00:49amiconnUnhelpful: Your test images seem to have supersampled chroma, not subsampled, unless I'm interpreting the djepg output wrong in comparison with my digicam pics
08:01:40Unhelpfulamiconn: really? i was not aware that you even *could* have such a thing, but it would explain the problem...
08:02:30amiconnFrom an arbitrary digicam image: Component 1: 2hx1v q=0 \n Component 2: 1hx1v q=1 \n Component 3: 1hx1v q=1
08:03:28amiconnFrom one of your test images: Component 1: 1hx1v q=0 \n Component 2: 1hx2v q=1 \n Component 3: 1hx2v q=1
08:06:21amiconnSee /usr/share/doc/libjpeg-progs/wizard.doc.gz (or wherever your distro puts it)
08:07:47Unhelpfulso basically i was interpreting the sample data backwards... i'll go close my bug and feel stupid now. ;)
08:14:27 Quit robin0800 (Remote closed the connection)
08:15:55Unhelpfulall i need now, i think, is replacements for the bitstream functions, and to clone most of jpeg_decode into a function that does one row at a time. what did you think of the idea of a scaler callback that fetches from the in-memory yuv bitmap? the downside is that it would be scaling on draw, but that's probably not fixable if you want the decoded cache as yuv anyway... also, is the Y-only decoded image on greyscale/mono targets 8 bits?
08:15:55Unhelpful that would be *very* easy to feed to the scaler.
08:19:33 Quit z35 ("Leaving")
08:32:14 Join flydutch [0] (
08:34:30***Saving seen data "./dancer.seen"
08:35:36amiconnEach decoded component is a canonical 8-bit bitmap. The jpeg viewer feeds the Y component directly to grey_ub_gray_bitmap_part()
08:36:14amiconnFor YCbCr, the colour components may or may not have lower resolution than Y of course
08:37:22 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
08:37:44amiconnThe jpeg viewer keeps both the compressed data and several decoded zoom levels in memory, as YUV for space reasons (non-subsampled YUV would take the same amount of memory as RGB, subsampled YUV then saves)
08:38:31Unhelpfulfor greyscale targets, basically you need a context struct that knows what line is being read from, and fills a struct img_part in with the address of the start of the row, and the length.
08:39:00Unhelpful...and *a function that* fills etc :)
08:40:20amiconnOtoh, storing native bitmaps on colour targets (RGB565, RGB565swapped) would save space vs. YUV 4:4:4, and would be on par for YUV 4:2:2. It would need a little more vs. YUV 4:2:0
08:40:30 Join l403 [0] (n=l@
08:41:08Unhelpfulfor color targets, things are more complicated. the color scaler only understands rgb, so the callback needs to fill a buffer with rgb pixels and return a pointer to a struct img_part with the appropriate data. the buffer can be whatever size is convenient, provided the returned buffer never crosses a line end
08:41:17 Join Rob2223 [0] (
08:42:15amiconnI'm thinking about the jpeg viewer. The scaler output is a native bitmap, so it might in fact make sense to store these instead of YUV
08:43:06Unhelpfulin that case, it might make sense to use the fractional loader there, rather than having the entire YUV bitmap in memory at the same time as the native one.
08:44:05 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
08:47:32amiconnThe native bitmap isn't kept in memory atm, but I think that it could be kept instead of the YUV when the scaler will be added
08:53:14 Join petur [50] (n=petur@rockbox/developer/petur)
08:53:24 Join JdGordon [0] (
08:54:11Unhelpfulright, but what i'm saying is that if you want to use the scaler and keep a native bitmap, it might be best to use the fractional loader, and not have to have the whole YUV image in memory as well.
08:55:35amiconnYes of course. In this case the fractional loader needs to be able to read from ram though
08:56:32 Join midijunkie [0] (
08:56:56JdGordonB4gder: ping?
08:57:13B4gderyou called sir?
08:57:23JdGordongot a min to get my server going again?
08:57:37 Quit midijunkie (Client Quit)
08:57:52 Quit Rob2222 (Success)
08:58:03JdGordon2mbit up now so all builds should be sweet :)
08:58:44JdGordonas are all the paths... i tihnk :p
08:59:25 Join bertrik [0] (
08:59:58Unhelpfulright... that's a slightly tricky part, but i think that the getc/putc functions i wrote for reading in characters, and the rework i'm going to have to do for the bitstream stuff, could be set up so that they can be conditionally replaced with versions that just fetch the data from memory. most of the code could be shared, then, but the core jpeg loader wouldn't need to be extended for reading from a file in memory, and plugins that wan
08:59:59Unhelpfult to decode from memory would have to call a version built for pluginlib.
09:02:42Unhelpfulor, the core version could take function pointers for getc/putc and friends, but that seems like it would be bloating and slowing down the core version (via checking pointers before calling those functions, or having to call its versions via pointers) for something that only plugins would be using.
09:03:19amiconnAccessing single bytes doesn't sound very efficient
09:04:52Unhelpfulmost of the header data is byte-oriented, though. the getc/putc i wrote load 16B at a time into a buffer so that we're not constantly doing 1B calls to read.
09:08:03 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
09:11:02 Join ender` [0] (
09:16:37 Join Thundercloud [0] (
09:19:37 Quit evilnick (Read error: 104 (Connection reset by peer))
09:21:44 Quit JdGordon (Read error: 110 (Connection timed out))
09:21:57 Quit bertrik (Remote closed the connection)
09:32:03 Join ajb [0] (
09:33:20 Join hittudiv [0] (n=hittudiv@
09:35:01 Quit hittudiv (Remote closed the connection)
09:37:28 Join hittudiv [0] (n=hittudiv@
09:44:12 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
09:48:18 Join JdGordon| [0] (i=43a02c5a@gateway/web/ajax/
09:52:27 Quit hittudiv ("Leaving")
09:52:53 Join kugel [0] (n=kugel@rockbox/developer/kugel)
09:53:25 Quit Thundercloud (Remote closed the connection)
09:54:26 Quit JdGordon| (" ajax IRC Client")
10:01:23 Quit JdGordon (Read error: 110 (Connection timed out))
10:09:36 Quit BigBambi (Read error: 110 (Connection timed out))
10:12:29 Quit l403 (Read error: 104 (Connection reset by peer))
10:15:21 Join l403 [0] (n=l@
10:20:13 Join n1s [0] (n=n1s@rockbox/developer/n1s)
10:21:40 Join Lynx_ [0] (
10:25:17 Quit StealthyXIIGer (Read error: 110 (Connection timed out))
10:26:35 Join pyro_maniac [0] (
10:34:30 Join Grahack [0] (
10:34:33***Saving seen data "./dancer.seen"
10:42:23 Join barrywardell [0] (n=barrywar@
10:42:30 Join AndyI [0] (n=pasha_in@
10:44:01GodEaterwow - users list is a bit charged :-/
10:44:37GodEaterI don't subscribe to it anymore just to avoid all that, and yet I can't help reading the archives. It's like a car crash, you don't want to look, but you can't help it....
10:49:02*kugel doesn't know what happpens in the user list
10:49:42Bagderstay that way!
10:49:47Bagderdon't look!
10:49:53 Quit Grahack ("Leaving.")
10:50:02 Join BigBambi [0] (n=alex@rockbox/staff/BigBambi)
10:51:13GodEateryeah really. Don't.
10:53:13*linuxstb has to look now, after so many people saying not to...
10:53:16kugeltoo late :(
10:54:40GodEaterit's not even just the vitriol - it's the *dreadful* quoting styles people use too.
10:55:11ZagorI'm particularly fond of the sig-first style. it's somehow freshly innovative!
10:55:49Bagderit's the "top posting must mean I leave everything as-is and only fill in my blurb at the bottom" style
10:56:03Bagderor rather not top-posting
10:56:21Bagderthe art of stripping useless quotes is soon extinct
10:56:55GodEaterZagor: yeah, that's clearly a new form of art ;)
10:57:21*petur thinks they have a point, the unsubsribe field is a bit too much hidden on that page...
10:57:44GodEaterthey didn't need to bitch about it though
10:57:46Bagderit is
10:58:00Bagderit's the default mailman design though
10:58:09 Quit ajonat ()
10:58:11GodEaterZagor: so what happened to the new site going up last night then? :)
10:58:27Bagderand possibly we should have "Unsubscribe:" in the footer instead of "mail admin:"
10:58:30ZagorI got swamped by my tax returns :-(
10:58:42GodEaterBagder: that's a good idea
10:58:50peturso many retrns? Are you buying us a beer now?
10:58:51GodEaterZagor: oh dear :(
10:59:54Bagderfooter changed!
11:00:22Bagderit's interesting how that question only appears on the users list...
11:00:52Bagderthe dev list doesn't even have a footer
11:00:53 Quit israr (Read error: 104 (Connection reset by peer))
11:01:10 Join israr [0] (n=israr@
11:01:14kugelBagder: that's only because people don't want to unsubscribe from the other lists :)
11:01:15*Bagder tries to not make conclusions on people based on this...
11:01:17 Quit AndyIL (Read error: 110 (Connection timed out))
11:01:20GodEaterwe don't shout at people on the dev list, so they don't threaten to leave :)
11:01:24kugelI did for you :p
11:01:35n1sthe mails in the dev list usually don't make you want to hurt somebody though ;)
11:01:54*kugel disagrees
11:02:10Bagderwell, if we exclude jdgordon's mail? ;-P
11:03:22kugelhis mails are fine
11:03:50*Bagder doesn't want to spell out the ones he thinks kugel means...
11:03:53*amiconn wonders about all this setjmp and exit business
11:08:08 Quit ajb (Remote closed the connection)
11:08:38kugelBagder: seems like a good idea :)
11:08:39 Quit kugel ("ChatZilla 0.9.84-rdmsoft [XULRunner]")
11:13:51 Join codesquid [0] (n=quassel@kong.kawo1.RWTH-Aachen.DE)
11:14:33 Quit barrywardell ()
11:15:58 Quit gevaerts (Read error: 60 (Operation timed out))
11:16:22 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
11:27:03 Join EternalRains [0] (
11:33:24 Join PaulJam [0] (
11:37:27 Quit bs66_ (Read error: 60 (Operation timed out))
11:38:09 Join PyroBor [0] (
11:38:28 Join bs66_ [0] (n=sysuser@
11:39:56 Join kugel [0] (n=kugel@rockbox/developer/kugel)
11:41:52 Quit kugel (Read error: 60 (Operation timed out))
11:44:10PyroBorHellow. I installed rockbox on iriver H10 20GB MTP. I works great! but i have problem changing the Music directory its properties to not hidden. I am connected to the player over UMS mode. And i am using linux... how to change it to not be hidden?
11:44:14 Join EternalRains1 [0] (
11:45:07 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
11:45:40 Join tvelocity [0] (
11:45:48 Join kugel [0] (n=kugel@rockbox/developer/kugel)
11:45:53GodEaterPyroBor: do you have mtools installed ?
11:46:18GodEaterI think mattrib is what you want
11:47:17GodEatermattrib -h <your music directory>
11:47:46PyroBorthank you
11:49:15 Nick fxb__ is now known as fxb (
11:51:44 Join einhirn [0] (
11:52:20 Nick fxb is now known as fxb__ (
12:02:18 Quit EternalRains (Read error: 110 (Connection timed out))
12:02:27 Quit laburd ("leaving")
12:10:45 Join nibbler [0] (n=Nibbler@
12:19:33 Quit kugel (Read error: 110 (Connection timed out))
12:26:02 Join barrywardell [0] (
12:33:13PaulJamhi, it seems as if somehow a 160x128 theme (mipper) shows up for 220x176 targets on the theme site...
12:33:52 Join Grahack [0] (
12:34:37***Saving seen data "./dancer.seen"
12:38:05pixelmawell, it's a text only one and could work on (almost?) every target
12:40:23pixelmait could show up on more pages then though...
12:40:36 Quit nibbler (Read error: 110 (Connection timed out))
12:41:37 Quit EternalRains1 (Read error: 104 (Connection reset by peer))
12:42:36pixelmaand leads to the question once again how to deal with screenshots on themes that work in different resulutions (or colour depths etc.)
12:51:23 Part PyroBor ("Konversation terminated!")
12:56:21 Join kugel [0] (n=kugel@rockbox/developer/kugel)
13:00:30 Quit Riku (Read error: 54 (Connection reset by peer))
13:02:01 Join Lss [0] (
13:04:20 Join evilnick [0] (
13:06:13 Join Xerion_ [0] (
13:10:04 Quit Xerion (Read error: 60 (Operation timed out))
13:10:04 Nick Xerion_ is now known as Xerion (
13:16:29 Quit Seed ("cu, Andre")
13:17:58 Join MT [0] (n=chatzill@
13:19:08MTlinuxstb : there ?
13:20:51 Join robin0800 [0] (
13:21:48 Join gregzx [0] (n=chatzill@
13:22:03linuxstbMT: Yes
13:23:14MTlinuxstb : How can I play the raw audio frames with sox without converting them to wav first ?
13:23:26linuxstbWith "play".
13:23:49linuxstb"sox" converts audio, "play" plays it. ("play" is a utility that comes with sox)
13:24:22MTyes, I mean are there any specific options/flags I should pass to play ?
13:24:52linuxstbSomething like "play -t raw -r 44100 -c 2 -s file.bin" for signed 16-bit stereo at 44.1KHz - but check the man page (or "play −−help")
13:25:33MTYes that's what I wanted to know, thanks :)
13:27:42linuxstbYou may just simply want to add wav output to your code though - it's very easy. Some code you can steal is in apps/codecs/libffmpegFLAC/main.c
13:41:51MTlinuxstb : since bits/sample isn't given in rm header −− bits_per_sample = frame_size * 8 / samples_per_frame ?
13:43:58 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
13:44:17 Join roughL [0] (n=raphaels@
13:52:17 Quit pyro_maniac ("Leaving.")
13:52:32linuxstbMT: That flac program just hard-codes it to 16 bits per sample. I'm guessing that's what the Cook decoder will generate.
14:02:29 Join schrottplatz [0] (
14:08:01 Join LambdaCalculus37 [0] (i=44a0430d@rockbox/staff/LambdaCalculus37)
14:14:26 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
14:21:47 Join pyro_maniac [0] (
14:34:41***Saving seen data "./dancer.seen"
14:35:50 Quit Zarggg (Excess Flood)
14:36:24 Join Zarggg [0] (
14:36:45 Join ajonat [0] (n=ajonat@
14:38:27 Join PaulJam_ [0] (
14:42:15 Join PaulJam__ [0] (
14:49:06 Quit PaulJam_ (Read error: 60 (Operation timed out))
14:49:15 Join PaulJam_ [0] (
14:52:59 Quit bmbl (Read error: 110 (Connection timed out))
14:53:28robin0800In Rockbox are the USB Id's the same for the Sansa C240 and the C250?
14:55:11robin0800gevaerts: Are you aware that on my C240 They are different Between Rockbox and the OF
14:55:33gevaertsdoes it matter?
14:55:52LambdaCalculus37As long as they're valid.
14:55:58GodEatergevaerts: didn't we find out the other day they're using different vid/pid for MTP/MSC ?
14:56:34gevaertssomeone claimed that, yes. I haven't checked it personally
14:56:37robin0800gevaerts: Could This be the Linux Gnome Problem?
14:56:46 Quit B4gder ("It is time to say moo")
14:57:11gevaertsrobin0800: no. The gnome problem is a bug in the libgphoto2-2 package
14:57:16GodEatergevaerts: nor I - perhaps I should.
14:58:13robin0800gevaerts: I thikthats a E200 problem Not C200
14:59:11gevaertsrobin0800: if gnome decides to treat a device that advertises itself as MSC as MTP, gnome is *wrong*
14:59:27 Quit PaulJam (Read error: 113 (No route to host))
15:00:10robin0800Yes Its wrong for E200 but I 'm not sure it is for the C200
15:00:30GodEaterhow can it not be wrong for both ?
15:01:33LloreanGodEater: I think robin0800 is saying there's uncertainty as to whether it's treating the c200 as MTP.
15:01:40robin0800The E200 in MSC mode cant be seen in Rockbox or OF mode
15:02:04robin0800C200 works in OF
15:02:24GodEaterthat's not true
15:02:28GodEatermine certainly didn't
15:02:53 Quit tvelocity (Read error: 60 (Operation timed out))
15:03:09robin0800GodEater: What version of F/W
15:03:23Lloreanwhat's the point of all this discussion anyway?
15:03:42LloreanIt's something on the host end, so it's something you should be researching and reporting to the libgphoto2 guys if you think it's a different bug.
15:04:20*GodEater will have to wait for the "refreshing database" thing to go away before he can report the OF version
15:04:41robin0800Llorean: So its not the Rockbox Different USB I'ds Then?
15:05:26LloreanUSB ids are non-functionaly. They're just information for the host.
15:05:34 Quit PaulJam__ (Read error: 113 (No route to host))
15:05:38GodEater01.01.00F apparently
15:05:40LloreanIf it treats us differently based on them when we're claiming to be an MSC device, the host has broken behaviour
15:06:42GodEaterrobin0800: the problem is gnome is basing it's decision about whether the device is MSC/MTP on the USB ids - which it shouldn't - it should listen to what the device is saying it supports.
15:07:29robin0800Llorean: Have we the right USB Id's To tell it itts MSC then
15:07:47GodEaterrobin0800: you're not listening, it's NOTHING to do with the USB id
15:07:48Lloreanrobin0800: The USB IDs have nothing to do with telling it it's MSC.
15:08:51LloreanWe are telling it our device is MSC. It's ignoring this because it's using the USB IDs for something they're not meant to be used for, instead.
15:09:03LloreanThis is a bug _they_ need to fix.
15:09:22 Join ajonat_ [0] (n=ajonat@
15:09:42 Quit kugel (Read error: 110 (Connection timed out))
15:10:03 Quit ajonat (Read error: 60 (Operation timed out))
15:10:04robin0800Llorean: I'll add C200 as a comment to the bug
15:10:36LloreanIt's the base behaviour that's the problem. If they fix it in a way that affects the e200 but not the c200, they still aren't doing things properly anyway
15:11:54robin0800Llorean: Thanks for that.
15:14:39 Join tvelocity [0] (n=tony@
15:15:28 Quit robin0800 (Remote closed the connection)
15:19:38 Part LinusN
15:20:49 Join evilnick_7 [0] (i=0c140464@gateway/web/ajax/
15:21:15 Quit Grahack (Success)
15:22:32 Join Zoxc [0] (i=Zoxc@
15:23:58 Join AndyIL [0] (i=AndyI@
15:31:26 Quit AndyI (Read error: 60 (Operation timed out))
15:31:47 Join Grahack [0] (
15:37:27 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
15:37:30 Quit AndyIL ()
15:39:08 Quit codesquid (Remote closed the connection)
15:44:41 Quit shodanX_ (Read error: 113 (No route to host))
15:47:30 Join shodanX [0] (
15:51:24scorchesaratoga: you are now cloaked
15:56:33 Quit Zarggg (Read error: 104 (Connection reset by peer))
15:57:27 Join Zarggg [0] (
15:58:16 Quit Grahack ("Leaving.")
16:04:08MTlinuxstb : something is wrong with decoding :/
16:04:34linuxstbVery wrong?
16:05:05 Quit schrottplatz ("o.O")
16:06:03MTwell .. the duration is correct at least
16:06:20MTother than that .. it sounds like it's compsed solely of noise
16:07:41MTI still don't know where to start looking, but I'm pretty sure I'm sending the correct frames to the decoder.
16:08:29linuxstbI would add debugging statements to the ffmpeg decoder, and also to your version (printing out the values of some buffers), and compare the output.
16:09:06linuxstbThis is how I wrote my Monkey's Audio decoder - comparing the output of each stage of the processing in the original decoder, and mine.
16:15:47 Join robin0800 [0] (n=quassel@
16:16:40 Quit robin0800 (Remote closed the connection)
16:23:55 Join schrottplatz [0] (
16:23:59 Join evilnick1 [0] (
16:26:07 Join _lifeless [0] (n=lifeless@
16:31:16 Quit __lifeless (Read error: 110 (Connection timed out))
16:32:44 Quit antil33t (Read error: 110 (Connection timed out))
16:32:52 Join antil33t [0] (n=Mudkips@
16:34:43***Saving seen data "./dancer.seen"
16:36:04 Join toffe82 [0] (n=chatzill@
16:36:38 Quit Zarggg (Read error: 104 (Connection reset by peer))
16:37:12 Quit LambdaCalculus37 (" ajax IRC Client")
16:37:15 Join Zarggg [0] (
16:44:22 Join LambdaCalculus37 [0] (i=44a0430d@gateway/web/ajax/
16:46:17 Quit _lifeless (Read error: 110 (Connection timed out))
16:50:49 Quit evilnick (Read error: 113 (No route to host))
16:51:30 Join _lifeless [0] (n=lifeless@
16:52:03MTlinuxstb : what about converting the file to wav using signed 16-bit big-endian pcm, then extracting the produced frames with ffmpeg and comparing that with my output ?
16:52:17 Quit perrikwp (" ajax IRC Client")
16:52:43 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
16:53:46linuxstbMT: I don't think that will tell you anything - you want to compare the data at earlier stages, not just the final output.
16:54:04linuxstbSo the first thing would be the input into the decoder.
16:54:11linuxstbAnd then later stages...
16:55:45saratogaMT: I would output WAV when debugging WMA after it was mostly working (so that I could import into matlab easily for plotting)
16:56:15saratogainitially, I defined a lot of "dbprintf()" macros that could be turned on and off to print debug info
16:56:29saratogai had one that dumped the IMDCT data, another that dumped during quantization, etc
16:56:50saratogai put them in the same places in the integer and fp codecs so I could compare what was going on
16:59:29saratogasomething like this:
16:59:36 Join wincent [0] (
16:59:56saratogathough it may or may not be useful for you
17:00:29 Join wpyh [0] (n=william@
17:01:53MTThe input to the decoder has already been checked and is correct now.. but now when comparing the decoding data (output after transform, quantization .. etc.) what would be my reference ?
17:02:37saratogaI would compare it to the original ffmpeg codec at the same point in the code
17:03:05saratogaideally both codecs should give the same output (converting fixed to fp if needed)
17:03:24saratogaparticularly at this early stage where you have not optimized anything yet
17:05:25MTso I'd add debugging lines to cook decoder in ffmpeg and write a main() that uses this decoder with same file ?
17:05:56saratogathats one way to do it
17:06:07saratogayou could also recompile ffmpeg, though that might be a bit slow
17:06:22saratogait comes with its own main function for debugging
17:07:47 Quit Zagor ("Don't panic")
17:08:07MTrecompiling ffmpeg is the main reason I was trying to find another solution :/
17:08:28linuxstbMT: I don't think you can avoid that....
17:09:26saratogaMT: right now you're debugging the parser?
17:10:07 Quit kkurbjun (Remote closed the connection)
17:11:21MTsaratoga : The parser is working *almost* properly right now; it outputs properly constructed frames that should be sent to the decoder..
17:11:30MTso now I'm debugging the decoder.
17:11:47saratogawhat have you changed so far in the decoder?
17:13:30MTnot much .. it now uses rb's rand() instead of ffmpeg'srandom_seed() and it no longer needs AVCodecContext/AVFormatContext/AVPacket .. instead, it uses RMContext and RMPacket
17:14:27 Quit toffe82 (Read error: 104 (Connection reset by peer))
17:15:34saratogacould you keep an unmodified version of cook.c handy and compare to that? or do you need to change too much to get it to work with your parser?
17:16:10 Join toffe82 [0] (n=chatzill@
17:16:34saratogai should also ask if you got your parser to play an RM file before you started removing ffmpeg junk?
17:16:37MTI can't use an unmodified version with my parser currently if that's what you mean.
17:17:28saratogaah ok
17:17:35MTplay an RM file ?
17:18:07saratogai was wondering if you were able to take the output from your parser and feed it to the ffmpeg decoder and get correct PCM audio
17:18:18saratogabut it sounds like you're not there yet
17:18:39MTI did something similar ..
17:18:48MTnot playing audio though
17:19:34saratogasorry I mean just writing it to a file so you can compare it to ffmpeg
17:20:11MTffmpeg has a command which produces the descrambled cook audio frames that should be fed directly to the decode() function .. I made the parser output the same set of file, cat'd both outputs into 2 different files and diff'd them against each other
17:20:49saratogaah ok
17:21:03saratogabut you don't have cook.c compiling outside of ffmpeg then?
17:21:58MTI've isolated cook.c from ffmpeg before working with it.
17:22:15MTso it's compiling outside of ffmpeg
17:22:40MTis that what you mean ?
17:23:18saratogaMT: does the parser give the same output when you diff ?
17:23:41saratogabut you can't decode an RM file?
17:25:15 Quit bubsy ("Mrrrrreow!")
17:26:15MTI can *use* the decode function of cook.c and produce a decoded output.. the duration is correct, but it sounds horrible (mainly noise, no identifiable audio/music..) so it seems like there's something wrong with the decoding part.
17:26:26saratogaah ok
17:26:31saratogai misunderstood you before
17:26:44saratogai'll be back in a few minutes, I need to help someone with something
17:26:53MTtake your time
17:27:12 Nick gromit` is now known as jordan` (
17:31:26 Join PaulJam [0] (
17:31:38 Join StealthyXIIGer [0] (
17:38:54 Join PaulJam__ [0] (
17:38:56 Quit PaulJam_ (Read error: 104 (Connection reset by peer))
17:46:28 Join Guest_9526 [0] (n=Mudkips@
17:46:34 Quit antil33t (Nick collision from services.)
17:48:56 Quit wpyh (Read error: 110 (Connection timed out))
17:49:44 Join PaulJam_ [0] (
18:00:03 Join PaulJam___ [0] (
18:00:16 Quit PaulJam (Nick collision from services.)
18:00:21 Nick PaulJam___ is now known as PaulJam (
18:00:36 Join Lynx [0] (n=Lynx@
18:00:36 Quit PaulJam_ (Nick collision from services.)
18:00:41 Quit PaulJam__ (Nick collision from services.)
18:01:06 Nick Lynx is now known as Guest27896 (n=Lynx@
18:06:26 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:07:07 Join Lynx [0] (
18:07:32 Nick Lynx is now known as Guest16090 (
18:09:39 Quit perrikwp (" ajax IRC Client")
18:14:28 Join faemir [0] (
18:16:59 Quit Lynx_ (Read error: 110 (Connection timed out))
18:16:59 Nick Guest16090 is now known as Lynx_ (
18:17:43 Join kugel [0] (n=kugel@rockbox/developer/kugel)
18:18:21 Quit PaulJam (Read error: 113 (No route to host))
18:23:11 Quit petur ("work->home")
18:24:59 Join lymeca [0] (
18:25:42 Quit lymeca (SendQ exceeded)
18:26:20 Join lymeca [0] (
18:26:28MTlinuxstb : quick question because I have to go in a few minutes ; to use cook.c directly from ffmpeg means that I'll have to know how to initialize AV* stuff. Or can I just hard code COOKContext with the same values with which it's init'ed in my decoder, then I won't need any ffmpeg-specific stuff ?
18:26:36 Quit Guest27896 (Read error: 110 (Connection timed out))
18:27:19MTand for the input buffers, I'll just send it the output of either of the 2 parsers.
18:28:03linuxstbMT: Sorry, I'm not sure what you mean without looking at the code...
18:29:03MTah .. 1 minute.
18:34:35 Quit schrottplatz (Remote closed the connection)
18:34:44***Saving seen data "./dancer.seen"
18:34:56 Quit obo (Remote closed the connection)
18:35:34 Join obo [0] (
18:36:46 Join schrottplatz [0] (
18:39:14 Join jmillikin [0] (n=jmilliki@
18:39:20 Quit kugel (Read error: 110 (Connection timed out))
18:42:02MTlinuxstb : those are the two functions in cook.c that take in ffmpeg-specific data stuctures
18:43:03MTlinuxstb : (if you want to look at my code too) here's my code :
18:43:32 Join bertrik [0] (
18:43:33 Part pyro_maniac ("Leaving.")
18:46:20schrottplatzim not happy with the menu
18:46:27schrottplatzof my rockbox
18:46:54MTTo avoid using cook_decode_init() ( and dealing with ffmpeg's parser, stream reader, codec context .. etc) I could initialize COOKContext with the same values from my code , and for cook_decode_frame() I could just send the audio frames that are output by my parser or ffmpeg's .. would this be a correct approach ?
18:47:05evilnick_7schrottplatz: In what way, and how do you feel it could be improved?
18:47:24 Quit ajonat_ ()
18:48:10 Join LambdaCalculus3 [0] (i=44a0430d@gateway/web/ajax/
18:48:49 Quit LambdaCalculus37 (" ajax IRC Client")
18:49:01 Nick LambdaCalculus3 is now known as LambdaCalculus37 (i=44a0430d@gateway/web/ajax/
18:49:15schrottplatzi think we should have as the first entry
18:49:31schrottplatzinstead of files
18:49:49evilnick_7But that doesn't list ONLY music. It lists Files
18:50:00LambdaCalculus37It's a file browser, hence the name.
18:50:08schrottplatzand under music the first option resume playback and second database
18:50:09LambdaCalculus37People carry more than music on their players.
18:50:40schrottplatzso when i start my player
18:50:47schrottplatzi can navigate in my pocket
18:50:48bertrikI can see the point
18:50:56gevaertsschrottplatz: please try to keep thoughts on single lines
18:51:09*gevaerts mostly listens to non-music, and he doesn't use the database
18:51:12evilnick_7schrottplatz: That means that it'd be two button presses and also movement in the menu to resume music
18:52:17schrottplatzand things like system and settings should be one entry
18:52:33schrottplatzbtw playlist in the music list too
18:53:04linuxstbMT: Sorry, I don't have time to look at it in detail now. But your idea sounds sensible - to just initialise CookContext yourself.
18:53:37LambdaCalculus37schrottplatz: That doesn't sound very intuitive at all, to be honest.
18:53:56schrottplatzmaybe second entry in the main menu would be files (bc i cant browse trough the files in my pocket)
18:54:15gevaertsIt should at least be called "audio", *not* "music
18:54:44gevaertsschrottplatz: can you browse the database in your pocket?
18:54:47evilnick_7I do like the idea of having "Music" as an entry on the main menu. That's what most users would want to access most often (imho)
18:54:57MTlinuxstb : no need to look in detail, just wanted to know what you thought about this approach. Thanks.
18:55:07schrottplatzbut i want to have same things in the same menu
18:55:08gevaertsso why does that make Files different?
18:56:31schrottplatzgevaerts: understood?
18:57:13gevaertsI still don't see why (a) the pocket thing is relevant, and (b) how your proposal would make things easier
18:57:39schrottplatzits more quick to navigate
18:57:52gevaertsis it?
18:57:54 Quit MT (Read error: 104 (Connection reset by peer))
18:58:09schrottplatzin my case yes
18:58:20schrottplatzi dont know your flow
18:58:42gevaertsI don't use the database at all
18:59:01schrottplatzand where is your problem?
18:59:14gevaertsI don't have a problem. The current layout is fine
18:59:38*LambdaCalculus37 agrees with gevaerts
18:59:57LambdaCalculus37I don't use the database, either, except for PictureFlow.
19:00:54gevaertsthe database is the second entry. Surely that's not far to reach?
19:01:50 Join moos [0] (i=mustapha@rockbox/staff/moos)
19:01:52linuxstbschrottplatz: I don't think you're going to convince us to change the official menu structure. But if you can compile Rockbox yourself, it's easy enough to move menu items around and rename them.
19:02:10schrottplatzi think it would make more sense if there would be an entry with audio->resume playback; database; (later picture flow)
19:02:49gevaertsand why not files?
19:02:55schrottplatzlinuxstb: i just wanted to make suggestions...
19:03:22schrottplatzgevaerts: in this case then only music files
19:03:30evilnick_7schrottplatz: If there's a good enough reason to change (that will be beneficial for enough users) then we're open to suggestions. They would have to offer improvements to enough people though and not just for your use individually.
19:03:46schrottplatzfor example a folder /music/
19:03:53gevaertsschrottplatz: my personal opinion is that what you propose would make rockbox slower and harder to use
19:04:02gevaertsAlso a lot less flexible
19:05:02gevaertsWhy do you e.g. want to restrict where users can put audio files?
19:05:04bertrikschrottplatz, I think the problem with something like the main menu structure is that there's probably a lot of people having some kind of strong feeling about it, so changing it is a very time-consuming process.
19:05:40 Join Llor_Phone [0] (n=Llorean@
19:06:34schrottplatzi thought it would be easyer
19:07:25Llor_Phoneschrottplatz: the best way to get something like that changed is to write a patch.
19:07:53schrottplatzbut im to stupid to write a patch :)
19:07:55gevaertsThe big problem is that it's easy to assume that the way you're using it is the most natural way and that everyone will agree. In reality there is no best way, and everyone does something different. Everyone falls into that trap...
19:08:04Llor_PhoneThe current menu is there for a reason. You basically need to show your way is better with a patch people can use over time to compare
19:09:08bertrikDoes the database store anything other than tags for audio files? If not, couldn't it have a more specific name, like "Audio database"?
19:11:00Llor_Phonebertrik: In the future it may hold other media. RIght now, yes, audio only.
19:11:39gevaertsmaybe "media database" would work?
19:12:51evilnick_7Until anything else is supported then "Audio database" sounds better to me. Once other items are in the db then "Media database" would be more descriptive.
19:12:54Llor_PhoneCould, yes. I may have missed something. Was one of the problems a confusing name?
19:13:48bertrikLlor_Phone, sorry, no not really, the other discussion just got me thinking about some of the names used in the main menu
19:14:41Llor_PhoneMaybe something more like "Library"
19:14:50Llor_PhoneI do prefer shorter names.
19:18:38bertriklibrary sounds good
19:18:46 Quit jeffdameth (Read error: 60 (Operation timed out))
19:19:56 Join jeffdameth [0] (
19:21:00LambdaCalculus37Library works.
19:21:01 Quit schrottplatz ("o.O")
19:21:48 Quit Llor_Phone (Read error: 104 (Connection reset by peer))
19:22:23 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
19:24:24 Quit ender` (" Never say "Oooops" ... always say "Ahhh, interesting..."")
19:31:05 Join jgarvey [0] (
19:45:02 Join ender` [0] (
19:49:07 Join JdGordon| [0] (i=836b0070@gateway/web/ajax/
19:55:43 Join SirFunk [0] (
19:56:38 Join Thundercloud [0] (
20:04:14 Join petur [50] (n=petur@rockbox/developer/petur)
20:10:18 Quit SirFunk_ (Read error: 110 (Connection timed out))
20:18:21 Join Grahack [0] (
20:20:49amiconnlinuxstb: The cook decoder project will probably only support cook encoded realaudio files for a start? Are there plans to extend it to other realaudio codecs?
20:24:12 Join tomers [0] (
20:25:18linuxstbamiconn: Yes, assuming there are decoders. It depends how long the cook codec takes (the ffmpeg decoder is floating point).
20:25:50amiconnI was looking at this list:
20:26:33amiconnFor ac3 and aac we already have decoders, so it's "only" handling a different container
20:26:49linuxstbI think ffmpeg has decoders for RealAudio 1.0 and 2.0 as well.
20:27:02amiconnThat leaves VSELP, LD-CELP, Sipro and ATRAC as lossy codecs
20:27:14linuxstbAnd there's an ATRAC codec in ffmpeg too...
20:27:42linuxstbBut yes, MT's original proposal talked about writing a general RM demuxer, with cook as the first codec.
20:28:24linuxstbNot on that list is MP3, but the data is packed in a different way to normal mpeg audio streams.
20:28:58linuxstbMP3-in-Realaudio uses this -
20:33:13 Join tessarakt [0] (
20:34:47***Saving seen data "./dancer.seen"
20:41:20 Join itcheg [0] (i=41d59de2@gateway/web/ajax/
20:56:07 Join perrikwp [0] (i=982132ed@gateway/web/ajax/
20:57:50 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
21:05:50 Quit itcheg (" ajax IRC Client")
21:18:39 Quit SirFunk (Read error: 110 (Connection timed out))
21:19:46 Join SirFunk [0] (
21:34:46GrahackHi, is there a cheat sheet for the use of buttons on various players ? especially iPod ?
21:34:58LambdaCalculus37The manuals.
21:36:14GrahackI'm learning the various tricks in the manuals but would be glad to have an overall look on them... Maybe I could try to submit something if I end with something nice.
21:36:45LambdaCalculus37You're more than welcome to submit patches. :)
21:37:35Grahackmmmf, I wish I could submit C patches, but I'll submit wiki patches only
21:40:48evilnick_7The manual is not written in C, so if you feel there's a section that can be improved or added then the text itself would be okay. (although Tex[?] is preferred)
21:41:05n1sGrahack: maybe it could be done as an appendix for the manual. It's probably the only chance it has of staying up-too-date
21:42:59 Quit barrywardell (Remote closed the connection)
21:43:47LambdaCalculus37evilnick_7: LaTeX, to be exact.
21:44:06 Join darkham [0] (
21:44:46 Quit n1s ("Lmnar")
21:45:37evilnick_7LambdaCalculus37: Thanks. I keep forgetting the exact name. Is the casing important?
21:45:57GrahackI know LaTeX, but I guess it should be simple enough to translate to html with no problems, gonna look at TeX4ht
21:46:19bluebrotherthe manual uses TeX4ht to produce html output
21:46:25Grahackcasing helps not forgetting it commes from Tex
21:46:37bluebrotherit would be TeX to be exact :)
21:47:13evilnick_7Apologies, my memory is encoded with a lossy codec.
21:47:18Grahack(sorry, you're right, I was about to correct myself and now am punished for my lazyness)
21:50:57 Quit JdGordon| (" ajax IRC Client")
21:51:04 Join JdGordon| [0] (i=836b0072@gateway/web/ajax/
22:00:38saratogaBagder: could you disable femlab as a build server for a little while? i need to resolve issues with getting blacklisted on my network
22:02:06Grahackcould someone point me to the sources of this page please ?
22:03:18GrahackI'm quite lost n the svn repo... :)
22:03:27linuxstbIt's in "manual"
22:03:37linuxstbWhich is at the same level as apps, firmware, tools, etc
22:03:58peturstarts here:
22:04:08bluebrotheryou need other parts of the rockbox sources to build the manual
22:04:25LambdaCalculus37evilnick_7: That's the ASCII representation of the name.
22:04:53linuxstbGrahack: That actual page is manual/rockbox_interface/main.tex
22:05:15Grahackpetur & linuxstb: thanks
22:05:20*linuxstb thinks some "svn mv" would be helpful in the manual...
22:05:49bluebrotherGalois: manual/rockbox_interface/main.tex
22:06:03*bluebrother agrees with linuxstb and spots being too slow :o
22:06:19bluebrothergiving all those main.tex files better names would be a good thing too IMO
22:11:57 Quit petur (Remote closed the connection)
22:12:17Grahackone .tex for several players: NEAT guys, but the task becomes tricky for me now: my strategy so far: copy paste the source to grab the buttons of other players. Moreover I planned to make tables: adds a layer of complexity
22:14:50Grahackthere are \begin{btnmap} macros too that should be used...
22:14:54 Join webguest45 [0] (n=ad224b55@gateway/web/cgi-irc/
22:15:07 Quit webguest45 (Client Quit)
22:17:33 Quit bmbl ("Woah!")
22:21:49 Join kugel [0] (i=kugel@rockbox/developer/kugel)
22:22:34 Quit lymeca ("KVIrc Insomnia 4.0.0, revision: , sources date: 20090115, built on: 2009/03/07 00:45:02 UTC")
22:25:50Bagdersaratoga: will do!
22:27:23 Join Seed [0] (
22:29:48gevaertsJdGordon|: deleting the file you're playing is common enough for it to be supported from the WPS context menu :)
22:31:36*Llorean wonders if the WPS context menu should also have "Remove from Playlist"
22:32:21JdGordon|gevaerts: thats been an outstanding bug for a while, imo it shuold be removed
22:32:56JdGordon|remove from playlist makes sense... remove from the disk is just asking for trouble
22:33:49LloreanRemove from disk seems pretty safe too, at least we don't get people really complaining about it.
22:33:54evilnick_7Llorean: I'd like to see that personally
22:34:04LloreanBut it seems to me there should be "Remove from Playlist" and "Remove from Disk and Playlist"
22:34:27LloreanThe one option we do have is, I think, the least useful variant since it ought to be very uncommon that you want it gone from the disk, but still in the playlist simply getting skipped over
22:34:50***Saving seen data "./dancer.seen"
22:34:59JdGordon|Llorean: remove from disk is not safe at all... best case is its fully buffered so you can skip around inside the track without problems... thats lamost never the case though
22:35:21evilnick_7I find it useful to "preview" new albums and delete the songs I don't want to keep with me on the player when away from a computer.
22:35:45gevaertsI used to delete files from the WPS after listening to them. I suspected some problems at some point (like skipping to the next track being confused), but those were fixed
22:36:11LloreanJdGordon|: If it removed it from the playlist as well (which would include moving on to the next track) it wouldn't be a problem though
22:38:51 Join stripwax [0] (
22:39:07 Join EternalRains [0] (
22:39:23Grahackgood night guys, the cheat sheet thing seems quite huge to accomplish, I'll start for the iPod, then who knows...
22:39:25 Quit Grahack ("Leaving.")
22:43:44saratogaBagder: i spoke with my network admin, and he says that on the 18th the build script tried to initiate more then 200 connections in one hour, and was blacklisted again
22:44:13saratogado you have any idea how difficult it would be to make the system connect less frequently?
22:45:58 Quit JdGordon| (" ajax IRC Client")
22:46:51Bagderwell, we'd have to change the build system to keep the initial ssh connection up
22:47:26saratogahow often does it reconnect now? i've just started reading the code
22:47:31Bagderit's not really that hard, only a bit sensitive area since breaking that script breaks all builds
22:47:50Bagderfor each build, it does:
22:48:13Bagderbuild, copy, cleanup (3 connections)
22:48:40Bagderhm, or is it two copies
22:49:17Bagderyes: build, copy-log, copy-build (for some builds), cleanup
22:49:28Bagderand an initial svn up
22:49:48Bagderso if you build 7 zip-targets, you get 1+7*4 connections
22:50:27gevaertshm, doesn't modern ssh support multiplexing a connection?
22:50:53Bagderit could just re-use the same one anyway
22:50:59gevaertsMaybe just enabling that and starting a connection to each build server at the start would do the trick?
22:51:06Bagderexcept for the copy...
22:51:14gevaertshow is that done? scp?
22:56:34saratogahow would you reuse the connection? is there a way to actually control the shell remotely ?
22:56:53 Quit LambdaCalculus37 (" off I go!")
22:57:01 Quit perrikwp (" ajax IRC Client")
22:57:55gevaertsBagder: actually, scp seems to use the same connection just fine
22:59:20BagderI've never played with that feature
23:00:14 Quit moos ("Rockbox rules the DAP world")
23:00:38Bagderright, man ssh_config has the details too
23:00:52gevaertsyou could actually just keep one ssh connection open all the time
23:01:17Bagderone of the trickiest things in the build script is error handling
23:01:34Bagdera connection will not live fine "all the time" ...
23:01:45gevaertsit will if you put it in a loop :)
23:02:08gevaertsbut yes, it's not *that* simple
23:02:53Bagderand frankly, there are several things we could improve in the build script as has been mentioned many times
23:02:53gevaertswould this change error handling at all though? I'd expect individual sessions to look the same
23:03:40 Join JdGordon| [0] (i=836b0070@gateway/web/ajax/
23:04:31BagderI'm not sure, it depends on how to setup the master if the slaves finds it died
23:06:56 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
23:08:03gevaertsIf I understand things correctly, with the right (non-auto) setup the slaves can just use their own connection if the master has died
23:08:04 Quit Horscht ("Verlassend")
23:12:18 Quit StealthyXIIGer (Client Quit)
23:13:43 Quit Lss (Read error: 54 (Connection reset by peer))
23:24:56 Quit Zoxc ()
23:28:56 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
23:29:20 Join froggyman [0] (n=47ba40e2@gateway/web/cgi-irc/
23:36:15tomersgevaerts: In usb-drv-arc.c, init_endpoints(), does RX means also IN ? i.e. in the host point of view?
23:36:32*gevaerts looks
23:37:38 Join midijunkie [0] (
23:38:53darkhamhow goes the job on sansa fuze?
23:39:27darkhami know are "in progress", but some news?
23:39:45bluebrotherah sorry, mixed up fuze with view
23:39:56*bluebrother goes standing in the corner
23:41:29gevaertstomers: I suspect that RX means OUT. At least in the description for BURSTSIZE and TXFILLTUNING ( and that seems to be the case, so I'm assuming it's also true for the rest
23:41:56gevaertsdarkham: have you checked the wiki and forums? That's where the information lives
23:42:14 Join vedlith [0] (
23:42:27tomersThat means it is from the device point of view... Doesn't the manual apply both for d
23:42:32darkhamgevaerts: yes but not so much about that...
23:42:33tomersdevice and host?
23:43:03tomersThanks! That really helps.
23:43:37gevaertstomers: I think RX and TX are probably always from the chip point of view, whether it's host or device
23:44:56*kugel doesn't like bluebrother's proposal size
23:45:04kugeland it's not much different from the existing one
23:45:36gevaertsdarkham: is probably the most helpful. Note that "Yes" in the table may mean "it somewhat works, see forums and/or irc logs for details" though
23:46:04darkhamgevaerts: thanx
23:46:25bluebrotherwhy does it need to be much different?
23:47:26 Quit stripwax ("")
23:48:26 Join stripwax [0] (
23:49:13 Quit Horscht ("Verlassend")
23:52:50 Join iWant_to_hack [0] (n=47c3910b@gateway/web/cgi-irc/
23:53:07iWant_to_hackhello... i need some help
23:53:30 Join Zarggg_ [0] (
23:53:45iWant_to_hackhello... i need some help
23:53:46evilnick_7iWant_to_hack: Ask your question then!
23:54:12iWant_to_hackI want to hack my old i pod but dont know where to begin
23:54:26LloreanThe Rockbox manual is a good place to start.
23:54:44iWant_to_hackokay... ill check it
23:54:56bluebrotherreading the website might also help ...
23:56:14 Join itcheg [0] (i=62db4767@gateway/web/ajax/
23:57:29 Quit evilnick_7 (" ajax IRC Client")
23:57:30froggymaniWant_to_hack: what exactly are you trying to hack on your iPod?

Previous day | Next day