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 2007-08-01

00:00:06 Quit obo ("bye")
00:00:33 Join hannesd [0] (
00:01:00Bagderrasher: yes, it should include "-I/path/to/sdl/headers"
00:03:00 Quit davina (Remote closed the connection)
00:03:03rasherThat didn't seem to work
00:06:33 Join webguest81 [0] (i=57030c0a@gateway/web/cgi-irc/
00:07:11 Quit jgarvey ("Leaving")
00:07:15webguest81sorry guys, do you know why all the wps for ipod video don't properly work on the ipod 80 gb???
00:07:19webguest81i'm going mad
00:07:29Soul-SlayerAre you using the official build
00:08:19webguest44Where can i find the folder where the .c files are on my computer? i'm trying to put matrix.c in my h10, but i cant find where to put them
00:08:35 Join skyfaller [0] (n=nelson@wikipedia/Skyfaller)
00:08:38Soul-SlayerYou have to compile it with the rest of the source code.
00:08:42rasherwebguest81, Have you installed the fonts? They can be found on the "current build" page, and are explained in the manual as well as on the extras page, which is also linked in the menu on
00:08:47DerPapsteven better: are you using wpses that are still up to date with the latest rockbox version?
00:08:56skyfalleryesss! just got my new iPod video
00:09:01DerPapsti don't have any issues with my wps on my 80GB iPod
00:09:15rasherwebguest44, You will need to compile the plugin, but first you will need to update it because it's horribly outdated and somewhat broken.
00:09:38rasherwebguest44, Considering how you phrased your question, I think the answer is that it won't work
00:10:17webguest81I'm gonna try right now
00:10:23Soul-SlayerHave fun
00:10:56DerPapstbut i had to fix minor parts of the wps because they were broken
00:11:05rasherBagder, any other ideas?
00:11:15webguest81DerPapst, have you installed the fonts?
00:11:22DerPapstrecrosscompile sdl?
00:11:25Soul-SlayerOf course.
00:11:39DerPapstSoul-Slayer == Derpapst?
00:11:54webguest44Can you apply a patch in windows?
00:11:57Soul-SlayerSoul-Slayer <= DerPapst :P
00:12:15Bagderrasher: can you try a somewhat older release and see if this problem was introduced with a rockbox change?
00:12:20rasherDerPapst, Hrm, I could try that, but it's been working for so long with no changes, so I'm rather of the opinion tht that's not the issue
00:12:28Bagderwell not release but rev
00:12:48DerPapstwebguest44: yes
00:12:52BagderI doubt think rebuilding sdl will do anything good for this
00:12:54Soul-Slayerwebguest44: Only if you're up for setting up a linux environment within windows, be it cygwin or vmware
00:13:02Bagderhm, "I don't"
00:13:11*Bagder types crappppppy
00:13:12webguest81ok, it does works!!!! thank you very much
00:13:19rasherBagder, do you remember off the top of your head how to checkout a specific date?
00:13:38rasherBagder, nevermind, I rtfm
00:13:39Bagderrasher: you can "svn up -r<rev>"
00:13:43DerPapstwebguest81: :)
00:13:59Bagderor "-r{date}" with a iso8601 date
00:17:26webguest44Soul-Slayer, What? I was trying to use microsofts Virtual PC 2007 thing but i dont have the right os. I have XP home, and it only works on the Professional version. Look it up on google. you can run linux in it if you want is the summary of thnis message, i don't want to do all this work to play Pong against my H10!!!
00:17:48 Quit ender` (" I worked for 22 years in the tourism industry in Niagara Falls. I have watched thousands of tourists cross the border in mid)
00:18:38 Quit webguest81 ("CGI:IRC (Ping timeout)")
00:19:08Bagderrasher: a test would be to change "SDL.h" to <SDL.h> in the files that use " instead of <
00:19:22Bagderfor #include
00:19:39Soul-Slayerwebguest44: Setting up the compiling environment in Windows is detailed in the wik
00:20:00rasherBagder, That sounds like an interesting experiment.. trying the old version first though
00:20:02 Join Febs [0] (
00:20:17DerPapstis the pong a.i. somewht decent?
00:20:29webguest44i think.
00:20:45DerPapstwhere is the patch?
00:20:47Soul-SlayerWell I wouldn't call it a 'revelation'... But it does what is advertised.
00:21:05webguest44on the traCKER
00:21:15Soul-SlayerI think he's like the number ^^
00:21:31*DerPapst too lazy to search
00:22:00*webguest44 is at peace.
00:23:25DerPapstwebguest44: is it FS #5855?
00:23:32DerPapstor another one?
00:23:40webguest44Ill look...
00:23:49Soul-SlayerYes it's that one
00:23:54 Quit idnar (Nick collision from services.)
00:23:57 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
00:24:28rasherBagder, nope, the old version doesn't work either. And neither mingw32, make or any other related package has been updated recently as far as I can tell. Trying to replace "" with <> now
00:24:47DerPapstmulti_pulse_patch? wth?
00:24:55webguest44DerPapst, yes.
00:26:04Soul-SlayerDerPapst: It's because if music is playing, the pads colours change to pulse with the music
00:26:14DerPapstlet's check it out ;)
00:26:49Soul-SlayerThere's only so much you can do with a game of pong...
00:26:54webguest44Yay!! i think it should be included with rockbox
00:27:07DerPapstwebguest44: what's the capacity of your h10?
00:27:17webguest4420 gb
00:27:36webguest44I have it right here
00:28:27*DerPapst builds it
00:28:34 Quit linuxstb ("CGI:IRC")
00:29:10*webguest44 is thinking of a funny and witty action.
00:31:11*DerPapst notices that cygwin is dogslow once again
00:31:27krazykitDerPapst, just install linux and be done with it :D
00:31:30DerPapsti'm doing 2 builds at a time :D
00:31:43DerPapsti have linux installed.
00:32:07amiconnHmm, one shouldn't have multiple targets connected and then copy rockbox over without first checking which target got what drive letter :/
00:32:13DerPapstbut my kernel is old and contains a bug which causes SDL not to build
00:32:21webguest44DerPapst, what player do you have
00:32:36DerPapstamiconn: hehe
00:32:36krazykitDerPapst, that makes no sense. the kernel doesn't affect sdl.
00:32:49krazykitamiconn, whoops, eh
00:32:51DerPapstwebguest44: 3G iPoD and 5.%G IPod
00:33:02amiconn5% iPod? ;)
00:33:19krazykitit's only a few of the cables he's building it for
00:33:31rasherBagder, still no luck, although I get a different error now
00:33:35Soul-SlayerHe stole a scrollwheel.
00:33:43DerPapstkrazykit: it does. SDL includes headerfiles and something is wrong with joystics
00:34:44krazykitDerPapst, −−without-joystick?
00:34:53DerPapstamiconn: when i build rockbox my computer laaaaaags. infach i type faster than my pc is spitting out the chars on the screen and then the shift key hangs from time to time too
00:35:01webguest44Does that even count as an ipod? Like a harddrive and the motherboard are 25%
00:35:31DerPapstkrazykit: i'm not a linux guru so i don't know if that works.i'
00:35:38DerPapstll try that tomorrow :)th
00:35:48krazykitDerPapst, it'd be in the ./configure −−help, probably
00:36:05 Quit barrywardell ()
00:36:07DerPapsti hit enter _before_ typing "thanks"
00:37:54 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
00:38:47*webguest44 waits for an update from DerPapst
00:38:54DerPapstthe h10 build is slower than the sim one.... odd. normaly the arm build is Faster ;)
00:39:19LloreanWhy is Fade on Stop/Pause enabled by default? As far as I'm concerned, it just makes the thing feel unresponsive.
00:40:08DerPapstere is the setting to kill that or at least shorten the fade out?
00:41:04DerPapst*indeed, where....
00:41:05Soul-SlayerLlorean: I think it's more to be aimed at the end user. One of the first things people I have shown Rockbox to have mentioned is 'how cool' a feature that is. Personally I don't find it much of a bonus either, but there we go.
00:41:48LloreanIt's the last thing I forget to turn off every time I end up accidentally clearing my settings
00:41:53DerPapsti think iT's a pita. i don't like it.
00:41:55LloreanDerPapst: Can't remember at the moment. Playback maybe.
00:42:05LloreanWe've gotten a bug report or two about it in the past.
00:42:21Soul-SlayerDerPapst: Settings -> General -> Playback -> Fade on stop/pause
00:42:23Soul-SlayerI believe.
00:42:38*Llorean goes to dinner
00:43:03Soul-SlayerLlorean: How on earth do you accidentally clear your settings often?
00:43:38Bagderit used to be very common
00:43:42DerPapstwow... finally building the h10 plugins :-/
00:43:45Bagderbefore we got the config in a text file
00:43:54Soul-SlayerWhere did it used to be stored?
00:44:03Bagderbinary in a sector
00:44:20Soul-SlayerSo every time you used a new build it wiped?
00:44:24Bagderand it was modified and thus made incompatible often, and then it was reset
00:44:25 Join flynux [0] (n=flynux@2a01:38:0:0:0:0:0:1)
00:44:50Bagderyou could still save/load text configs though, so it wasn't that annoting mostly
00:44:53DerPapstgood old times :D
00:45:05 Quit sharpe (Read error: 104 (Connection reset by peer))
00:45:21amiconnLlorean: 'Fade on stop/pause' is one of the first features I am disabling...
00:45:38*DerPapst is doing that too soon
00:45:42webguest44Derpapst, Yesssss!
00:47:02Soul-SlayerI wish I knew where all my disk space on my linux partition has run off to
00:47:05DerPapstgot builderrors...
00:47:54DerPapstbut it's only rockboy... so
00:48:06Soul-SlayerWhy did that error :S
00:48:41DerPapstmaybe i have messed with it and never reverted...
00:48:56DerPapstbut it didn't fail on any other builds so far
00:50:39webguest44is it done? is it done? is it done?
00:50:41DerPapstwebguest44: wait a sec...
00:51:07Soul-Slayerwebguest44: It's a game of pong with an AI element to it, calm yourself :p
00:51:13DerPapstwebguest44: do you care about rockboy?
00:52:05DerPapstif so you should backup rockboy.rock and the viewer config
00:52:56 Quit darkless (Read error: 101 (Network is unreachable))
00:54:22rasherBagder, ran out of good ideas?
00:54:51DerPapsthmmm.. svn up broke rockboy...
00:55:14rasherDerPapst, conflicts
00:55:24DerPapstdunno why though
00:57:32webguest44DerPapst, not really. I cant figure it out
00:57:55DerPapstor does it not download?
00:57:59webguest44o rockboy
00:58:23DerPapstcheck in ./rockboxviewers
00:58:35DerPapst.rockbox/viewers even
00:59:03webguest44what player does the .zip work with?
00:59:12Soul-SlayerH10 20gb
00:59:23DerPapstas you said...
00:59:40 Quit robin0800 (" HydraIRC -> <- Would you like to know more?")
01:00:07webguest44ohh. ok. im un zipping it now. will give feedback
01:00:13DerPapstcan i paly agains my ipod as well in this new pong? can't figure out how though
01:01:19DerPapstah.. figured out.
01:01:23DerPapstneat ;)
01:01:42webguest44Drum roll please...
01:02:01 Join muesch [0] (i=53b45170@gateway/web/cgi-irc/
01:02:03Soul-SlayerI'm finding it really hard to contain my excitement
01:02:29webguest441-0 im winning!
01:02:33DerPapstwebguest44: and you have the gbs codec now.. but it won't work well anyways... :P
01:03:13webguest44Thanks!!! will the plugin be put in the svn?
01:03:18 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
01:03:27mueschhey guys, a little question, do you know solutions to the problem that ipodpatcher doesn't find the g5.5 ipod?
01:03:32Soul-SlayerWho knows, it's been in the tracker for a long while
01:03:39 Join miepchen^schlaf [0] (
01:03:49Soul-Slayermuesch: Is it connected to your computer in Disk Mode?
01:04:00mueschit's in disk mode
01:04:02DerPapstI like the new pong :)
01:04:05mueschwith linux
01:04:15mueschi can mount it
01:04:23mueschbut ipodpatcher doesn't find it
01:04:29Soul-Slayerdoes fdisk -l find it?
01:04:30 Quit ShadowXP (Read error: 104 (Connection reset by peer))
01:04:44webguest44My backlight goes off too quickly. do i need to set new settings?
01:04:48 Join ShadowXP [0] (
01:05:11DerPapstwebguest44: the shouldn't have been overwritten
01:05:20mueschsoul-slayer: yes
01:05:26DerPapstmuesch: are you root?
01:05:40webguest44DerPapst, whats the gbs codec? sounds cool, but...?
01:06:13Soul-Slayerwebguest44: It's a codec (Like MP3, FLAC, OGG) which was starter implemented in a patch
01:06:30Soul-SlayerAnd I'm guessing DerPapst applied the patch and didn't revert it when he sent it to you
01:06:34DerPapstit's gameboy tunes
01:06:50muesch@soul-slayer: wheee :D it works. 1000 thanx
01:06:59mueschsomeone should write that into the wiki...
01:07:06Soul-SlayerEven though DerPapst said to be root, I was getting there :p
01:07:19Soul-SlayerIs it not there already?
01:07:24mueschi think it isn't
01:07:27mueschjust a sec...
01:07:34webguest44Ok. i guess i never needed my backlight on. i set it to 15s. enough for pong, eh?
01:07:37mueschi'm talking about the ipodlinux wiki...
01:07:54mueschhey, do you guys have pong on your ipods???
01:08:07Soul-Slayer"Warning: You may need to be the root user in order for ipodpatcher to have sufficient permission to perform raw disk access to your player."
01:08:11Soul-SlayerOh right
01:08:29Soul-SlayerPong comes with rockbox normally
01:08:47mueschthat's awesome :D
01:08:49DerPapstiPodlLinux coems with pong and an a.i. too
01:09:11mueschanyways, we'll install taht shit on my ipod tonight :D
01:09:20*webguest44 enjoys an ice cream cone.
01:09:23mueschi can't type anymore... :D
01:09:24*DerPapst made that wiki page...
01:10:04Soul-SlayerThen it's all your fault. Muahahah.
01:10:45mueschanyways, on this page it doesn't say the thing about the root.
01:10:52mueschbut oh well, never mind
01:10:55mueschthanx for the help
01:11:06 Quit bluebrother (Read error: 113 (No route to host))
01:11:06DerPapstit doesn't.
01:11:08Soul-SlayerHave fun
01:11:42 Quit rasher (Read error: 148 (No route to host))
01:11:52 Join hannesd_ [0] (
01:12:46webguest44i just can't beat the Ai
01:13:15***Saving seen data "./dancer.seen"
01:13:39DerPapstSoul-Slayer: does one have to be root to write the mbr with fdisk?
01:13:42DerPapstprobably yes...
01:13:49Soul-SlayerYou'd hope so
01:14:06*DerPapst adds wanrings to the wiki
01:14:22 Join darkless [0] (
01:14:24 Join gromit` [0] (
01:14:41*amiconn fetches the ata specs
01:16:01*webguest44 wishes the FM H10 radio would be availible as fast as the Pong AI
01:16:47Soul-Slayerwebguest44: Go ahead and write a patch for the radio then :P
01:18:46*webguest44 wishes to have more knowlege on the subject and all of the previous exursions to make the radio work.
01:18:55 Quit muesch ("CGI:IRC 0.5.7 (2005/06/19)")
01:19:27 Join a1titude [0] (
01:20:45Soul-SlayerLearn :)
01:20:52webguest44My player is really fast. like it responds well to the scrollpad and when when music is playing. More existing patches?
01:21:12Soul-SlayerThe CPU speed is boosted when music is playing
01:21:30amiconnNot always
01:21:54Soul-SlayerOne day I'll be able to sound confident and be right in sounding confident :p
01:22:02Soul-SlayerDoesn't work when there are professionals all over the place =$
01:22:22*webguest44 knows this, someone tried and could not even get power to the actual fm module
01:22:26 Join rasher [0] (
01:22:41Soul-SlayerWould have been barrywardell I imagine
01:23:36Soul-SlayerHe made the initial H10 port, and because the FM receiver is the same as in another iRiver, tried using it the same way but to no avail
01:23:55webguest44i assume so
01:24:23webguest44what about disassembling the OF
01:26:08Soul-SlayerI don't know, but I know it's not as easy as it sounds.
01:26:32webguest44on DOOM and Chessbox, I get incompatibilities issues.
01:27:03 Quit hannesd (Connection timed out)
01:27:04 Nick hannesd_ is now known as hannesd (
01:27:17rasherBagder, I just recompiled SDL, and everything works now. I have no idea what's going on - I haven't touch my crosscompiled SDL since the very first time I built it, many months ago. Weirdness.
01:28:49*webguest44 lets my roommate try out pong
01:29:45webguest44What was the purpose of 2 player pong? this would be diffucult to do.
01:30:26 Quit spiorf (Remote closed the connection)
01:30:54DerPapstwebguest44: simply make a backup of .rockbox/rocks/pong.rock before you update your build
01:31:18webguest44ok. good to see you again.
01:31:46DerPapstthen replace the one from the zip with the backuped one
01:31:57Soul-SlayerDerPapst: That wouldn't work though, if the build version is different to the .rock version it'll just complain about incompatabilities
01:32:05DerPapstuntil no one messes with the menu api this should work.
01:32:31DerPapstthen don't update :P
01:32:31Soul-SlayerI'm fairly sure
01:32:53webguest44Im fine with the build. i update every 3 weeks or so.
01:32:58DerPapstor bug the devs so they check it into svn
01:33:13webguest44the Who?
01:33:16*DerPapst would like to see that in svn too
01:33:55webguest44Are there any Devs watching?
01:34:03pixelmaDerPapst: guess you mean the plugin api though ;)
01:34:08DerPapstthere are a few here right now
01:34:19DerPapstpixelma: yes
01:34:42 Quit BHSPitMonkey (Connection timed out)
01:34:46webguest44Yournot a Dev?
01:35:10DerPapsti have contributed 2 small patches.. but that's all
01:35:41webguest44Why are there so many names but not many actualy talking
01:36:17DerPapstsome are only talking once in a month
01:36:22Soul-SlayerThey'll pounce when they see fit.
01:36:31*DerPapst thinks of miepchen^schlaf
01:36:44Soul-SlayerA strange species is the Rockbox Dev ;)
01:36:49 Part toffe82
01:37:10DerPapstthey are actually pretty active
01:37:57webguest44FIT! FIT! Will they talk now?
01:38:15 Join webguest89 [0] (i=52990f30@gateway/web/cgi-irc/
01:38:41*DerPapst counted 10 people with svn access right now.
01:38:48DerPapstbut i probably missed some ;)
01:38:59webguest44Cool DerPapst
01:39:19DerPapstmiepchen^schlaf: you aren't sleeping.... i knew that :P
01:39:23webguest89I'm having a problem installing rockbox onto my 5G Ipod, anybody willing to give some advice?
01:39:25webguest44webguest89, whats up
01:39:43Soul-SlayerDetails please webguest89
01:39:55webguest89I'm getting a data abort error whenever I play any MP3
01:39:58*amiconn counts 17
01:40:08Soul-SlayerAre you using the latest bootloader and SVN?
01:40:13webguest89it happens as soon as I hit play
01:40:13DerPapstsee? ;)
01:40:35DerPapsti probably don't even know all of them
01:40:38webguest89I'm thinking my Ipod may be on its last legs
01:40:38 Quit doc|home (Read error: 104 (Connection reset by peer))
01:40:53Soul-SlayerWell, does the apple firmware work? If so, I doubt there's anything wrong with it
01:41:30*webguest44 sugests PMing them!@
01:41:34webguest89the apple firmware does work
01:41:40Soul-SlayerWhen did you last install the bootloader?
01:41:45*amiconn wonders whether the nano boots with the CPU set to 80MHz
01:41:53DerPapstwebguest89: and you're using the latest build also?
01:42:02*Soul-Slayer wonders why it wouldn't
01:42:08webguest89last updated the bootloader today
01:42:32*Soul-Slayer checks recent commits to see what could have done this
01:42:42amiconnSoul-Slayer: All other ipods boot with the CPU set to 24MHz
01:42:50*webguest44 wonders about Ipodlinux
01:42:59Soul-Slayerwebguest89: What build last worked for you, or is this the first one you tried?
01:43:24webguest89this is the first one I have tried
01:43:29Soul-Slayeramiconn: I wish I knew how you worked all this out, lol
01:43:39Soul-Slayerwebguest89: Give me a few mins.
01:43:50 Join jhulst [0] (
01:43:51DerPapstwebguest89: there will be a new daily build soon. use that instead and when you get a dataabort as well we can easily check where it happens when you give us the address
01:43:58webguest89ok thanks
01:44:04amiconnSome info from the ipl wiki, disassembling a whole bunch of original firmwares, checking register dumps...
01:44:33Soul-SlayerYou make it sound so easy
01:44:47DerPapstwebguest89: but not the current one.. you might have to wait 24hours.
01:44:49*webguest44 adds a please-your-charming-and-wonderful-devs to any actions that involve the devs.
01:44:52amiconnFinding the boot speed is actually quite simple - just measure it (with a calibrated delay loop) without letting rockbox touch the setting before
01:44:57webguest89ok no problem
01:45:02DerPapsti don't know when a new dailybuild is triggert
01:45:12Soul-Slayerwebguest89: Do you have a regular iPod video or a 64MB one?
01:45:18amiconnDerPapst: 6am ce(s)t afaik
01:45:59DerPapstwebguest89: 30GB one?
01:46:10Soul-Slayer Install that build for me please and tell me if it works
01:46:40webguest89yeah 30 GB
01:47:06 Quit rasher ("leaving")
01:47:25Soul-SlayerIf that build works I imagine something from barrywardell's commit broke the 5g
01:47:38webguest89doing it now
01:47:40webguest44If no one gets my See FIT!! FIT!! joke, my life is nothing ;(
01:47:43Soul-SlayerOkay thanks.
01:47:53Soul-Slayerwebguest44: I don't get it.
01:48:40webguest89data abort at 0002fbe0 is the error I'm getting
01:49:02Soul-SlayerNot that then... Lets try a much older build then
01:49:02webguest44You said that they would pounce when they see 'fit'. FIT!! FIT!!
01:49:11Soul-SlayerAh, I see ^^
01:49:36 Join atsea- [0] (i=ariel@gateway/tor/x-64983e52a3cad138)
01:49:38Soul-Slayer Try that one webguest89
01:50:09 Part a1titude
01:51:34webguest89data abort at 00030110
01:51:36*webguest44 yawns.
01:51:49webguest89well, it's a different error this time
01:52:58Soul-SlayerLets get this right... You're using a 30GB iPod video, you installed the latest ipodpatcher today, and no builds seem to work?
01:53:35DerPapstthere are a vew reported cases about that.
01:53:35webguest44Why do you still want Rockbox?
01:53:44DerPapstuse a build about 2 weeks ago
01:54:03Soul-SlayerI just gave him a build from the 13th of July
01:54:10Soul-SlayerWait, was that a Friday?
01:54:19Soul-SlayerAck it was. :p
01:54:39webguest89I heard good things about the software, the main thing that attracted me is dragging and dropping MP3s onto the Ipod
01:54:47webguest44Try one from Feb. 21
01:55:10Soul-Slayer Earliest archived build on the site
01:55:12Soul-SlayerTry that
01:55:17webguest89I was thinking of getting a Iaudio or something soon anyway
01:55:31 Part skyfaller ("Leaving")
01:56:14Soul-SlayerHow do you identify between the regular build and the 64mb ram one?
01:57:05webguest44I had a good idea. What about putting a 20 Questions game as a plug in. Call it Qbox
01:57:21Soul-SlayerGet coding :)
01:58:12webguest89back in a tick sorry
01:58:12webguest44I dont think Radica will tell me those secrets
01:58:37DerPapstSoul-Slayer: there are 2 different build. so you don't have to identify them on runtime.
01:58:47 Join Robin0800 [0] (
01:59:10Soul-SlayerNo, but how would I as the end user identify it?
01:59:27Soul-SlayerI can't just look at my ipod and go 'Yeah, that looks like it's packing 64mb of ram' :P
01:59:29webguest44Qbox is a cool name, though. Right?
01:59:43webguest44The idea is cool, Right?
01:59:45Soul-SlayerWe already have xobox, zybox, pacbox
02:00:04webguest44But no Qbox
02:00:52 Nick pixelma is now known as pixelbox (i=pixelma@rockbox/staff/pixelma)
02:01:00 Nick pixelbox is now known as pixelma (i=pixelma@
02:01:04Soul-SlayerThe point I'm making is anything with the 'box' suffix is nothing new :p
02:01:09webguest89Sorry, I simply assumed that this was the 32MB model as I bought this Ipod pretty much as soon as they came out
02:01:29DerPapstSoul-Slayer: 30GB iPod video = 32MB ram 60/80GB iPod = 46MB ram
02:01:34DerPapst*64MB even
02:01:35Soul-SlayerI'm not saying you're wrong, I just don't know how you're meant to find out
02:01:38Soul-SlayerAhh ok.
02:01:43webguest89plus I don't have this search function that's available on the 5.5G Ipod
02:01:58Soul-Slayer5.5g = 60/80 or just 80?
02:02:05webguest44Soul-Slayer, but the idea is.
02:02:10webguest89no idea sorry
02:02:18Robin0800Soul-Slayer,what size is hard disk?
02:02:44webguest89i think 5.5G is just 30GB for the lower end and 80 GB for the higher end
02:03:01Soul-SlayerI don't have one, I'm just not sure how you go about working it out
02:03:12webguest44pixlema, WT heck
02:03:15Soul-SlayerAnyhow. webguest89, did you try the older build I said?
02:04:26webguest89data abort at 000305b0
02:04:31Robin0800webguest89,what size is hard disk?
02:04:54webguest89I think the problem lies within the Ipod, it's been performing very slowly recently
02:04:57Soul-SlayerI'm stumped.
02:04:58webguest89it's a 30GB
02:05:13webguest89thanks for your help anyway
02:05:19*webguest44 tries to get others to do his work, and that system mostly works.
02:05:32Robin0800webguest89, then 32mb
02:06:26DerPapstRobin0800: we told him/her already
02:06:29 Quit jhMikeS (Read error: 104 (Connection reset by peer))
02:06:41DerPapstok ;)
02:06:49*webguest44 is torn between the tv and this
02:06:50DerPapst(one can't be carefull enough)
02:07:15Soul-SlayerWhy does the 80gb have double the RAM?
02:07:34webguest89more oomph I guess
02:07:39 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
02:08:09Soul-SlayerUnless they just expect you to be using larger files, so it doesn't want to spin up the disk as often
02:08:27DerPapstSoul-Slayer: maybe the hdd drains more power while active. So you have double the size to preload more data
02:08:35Robin0800more videos perhaps
02:10:58 Part Robin0800
02:12:04DerPapstGood night everyone :)
02:12:06 Quit DerPapst ("So Long And Thanks For All The Fish!")
02:12:27 Quit Soul-Slayer ("Leaving.")
02:18:19 Quit webguest44 ("CGI:IRC")
02:19:21 Quit webguest89 ("CGI:IRC (EOF)")
02:21:50 Part pixelma
02:27:36 Quit jhMikeS (Nick collision from services.)
02:27:42 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
02:32:03 Quit Thundercloud (Remote closed the connection)
02:37:04 Quit ShadowXP (Read error: 104 (Connection reset by peer))
02:37:12 Join ShadowXP [0] (
02:52:29 Quit inmotion (Read error: 110 (Connection timed out))
02:53:14 Quit jhMikeS (Read error: 104 (Connection reset by peer))
02:54:05 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
03:13:17***Saving seen data "./dancer.seen"
03:40:38 Quit dandin1 ()
03:51:39 Join BHSPitMonkey [0] (
03:52:34 Join Soul-Slayer [0] (n=Administ@
04:03:45 Join chrisjs169 [0] (i=nuclears@unaffiliated/chrisjs169)
04:11:59 Quit jhulst (Remote closed the connection)
04:18:23 Quit chrisjs169 ("CGI:IRC")
04:18:59 Join chrisjs169 [0] (i=nuclears@unaffiliated/chrisjs169)
04:21:41 Join Soul-Slaye1 [0] (n=Administ@
04:23:21 Quit thegeek_ (Read error: 104 (Connection reset by peer))
04:23:30 Join thegeek [0] (
04:26:25 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
04:31:47 Quit chrisjs169 ("CGI:IRC")
04:35:04 Join linuxchuck [0] (
04:35:17 Quit Soul-Slayer (Read error: 110 (Connection timed out))
04:35:39 Nick Soul-Slaye1 is now known as Soul-Slayer (n=Administ@
04:36:35linuxchuckAlright... Who's got a moment to assist a striving c-coder-wannabe with intent for building a plugin? :-)
04:36:54*linuxchuck drops a pin and listens...
04:36:58 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:37:07linuxchuckyep... It's quiet now. :-P
04:37:12 Join miepchen^schlaf [0] (
04:39:02 Quit hcs (Read error: 104 (Connection reset by peer))
04:39:45linuxchuckreally quiet... :-/
04:39:53 Nick gtkspert_ is now known as gtkspert (
04:40:26Soul-SlayerUsually is at this time :p
04:43:12linuxchuckyeah. I've just hit a *tiny* speed-bump.
04:43:26linuxchuckI'll check in tomorrow, at a more reasonable hour.
04:47:01 Join inmotion [0] (n=tgang@
04:47:37TiMiD[FD]linuxchuck: I've got 8 minutes for you
04:48:09linuxchuck:-) Thanks! Here's my question:
04:48:40linuxchuckI'm building one of my first *official* c programs, and am working on a playlist builder as a plugin for rockbox.
04:49:07linuxchuckI'm looking for a way to enable d_type (found in dirent.h) for detecting file types
04:49:18linuxchucksuch as directory, file, and so-on.
04:49:29linuxchuckor, is there a better way to detect the file type?
04:49:58linuxchuckI'm basically currently working on the recursive directory scanner that hunts for mp3 files, and returns results.
04:50:01TiMiD[FD]let me check that ...
04:50:09linuxchuck'preciate it.
04:50:24TiMiD[FD]cannot you check the files by extension btw ?
04:51:05linuxchuckyes, but... there are a large number of "what if" scenarios I have to account for.
04:51:27linuxchuckOne of the first lessons I ever learned about writing programs, is to NEVER expect your users to do something smart.
04:51:38linuxchuckwhat if they create a directory with a .mp3 extension?
04:51:58linuxchuckthings like that. I'm doing my best to make the program "idiot resistant".
04:52:09TiMiD[FD]sorry I misenderstood ...
04:52:15LloreanYou shouldn't only expect music to be MP3s
04:52:21TiMiD[FD]so you wanna make the difference between files and dirs ...
04:52:21LloreanYou'll need to cover all audio types Rockbox does
04:52:22linuxchuckoh, it's no problem.
04:52:37LloreanWhat you'll really want to do is use the list of audio extensions Rockbox supports, then rule out directories.
04:52:43LloreanRockbox won't recognize music files with bad extensions anyway
04:52:47linuxchuckwhich is why I'll most likely set it up to take a command line argument for certain things.
04:52:55TiMiD[FD]isn't there a plugin that goes recursively through the directories already ?
04:53:10linuxchuckoh, wasn't answering your plugin question.
04:53:17linuxchuckis ther?
04:53:36LloreanThe configuration plugin for the random folder playback does, I believe
04:53:38linuxchuckperhaps I can look to that for inspiration.
04:53:43LloreanIt shows you a list of all folders and subfolders on your device
04:53:44TiMiD[FD]entry = rb->readdir(dir);if(entry->attribute & ATTR_DIRECTORY) ...
04:53:49linuxchuckexcellent! I'll hunt for it.
04:53:55midgey34i thought disktidy did as well
04:54:21linuxchuckum... I don't think the attribute is stored as a flag, so & won't work.
04:54:37linuxchuckit'd have to be == I think. If my research so far is correct.
04:54:55TiMiD[FD]it works like that
04:54:56linuxchuckand the types would be stored in a struct, and found under d_type
04:55:03linuxchuckah, ok. I'll give it a try.
04:55:12linuxchuckperhaps I've been barking up the wrong tree.
04:55:35linuxchuckI'll go through the two plugins mentioned earlier, and see if I can get my version working here.
04:56:02linuxchuckheh... then I'll have to learn how to cross-compile... but that's a whole different bridge to cross, and I've a ways to go.
04:56:42linuxchuckThank you both for your suggestions. I'll see what I can figure out with them.
04:59:28linuxchuckHere is another quick (and hopefully not too stupid) question:
04:59:32linuxchuckif you are still here...
05:00:13linuxchuckIs it reasonable for me to believe that I can build, compile, and test my program on my regular Linux system before attempting to cross-compile it for the Rockbox platform?
05:00:17DEBUGReceived signal 15 (SIGTERM), terminating (snapshot: fplrun.c line 385)
05:00:17***Saving seen data "./dancer.seen"
05:05:20***Started Dancer V4.16
05:05:20***Connected to on port 6667
05:05:20***Logfile for #rockbox started
05:05:24CtcpVersion from freenode-connect!freenode@freenode/bot/connect
05:05:25***Server message 501: 'logbot :Unknown MODE flag'
05:05:25Mode"logbot :+i" by logbot
05:05:25***Server message 477: 'logbot #rockbox :[freenode-info] help freenode weed out clonebots, please register your IRC nick and auto-identify:'
05:05:26 Join logbot [0] (i=bjst@gateway/web/cgi-irc/
05:05:26 Join pearldiver [0] (
05:05:26 Join inmotion [0] (n=tgang@
05:05:26 Join miepchen^schlaf [0] (
05:05:26 Join linuxchuck [0] (
05:05:26 Join thegeek [0] (
05:05:26 Join Soul-Slayer [0] (n=Administ@
05:05:26 Join BHSPitMonkey [0] (
05:05:26 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
05:05:26 Join ShadowXP [0] (
05:05:26 Join atsea- [0] (i=ariel@gateway/tor/x-64983e52a3cad138)
05:05:26 Join gromit` [0] (
05:05:26 Join darkless [0] (
05:05:26 Join hannesd [0] (
05:05:26 Join flynux [0] (n=flynux@2a01:38:0:0:0:0:0:1)
05:05:26 Join idnar [0] (i=mithrand@unaffiliated/idnar)
05:05:26 Join Febs [0] (n=chatzill@rockbox/administrator/Febs)
05:05:26 Join rotator [0] (n=e@rockbox/developer/rotator)
05:05:26 Join BigMac [0] (
05:05:26 Join bdgraue [0] (
05:05:26 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
05:05:26 Join Llorean [0] (n=llorean@rockbox/administrator/Llorean)
05:05:26 Join crashd_ [0] (
05:05:26 Join lostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
05:05:26 Join kubiix [0] (
05:05:26 Join hashier [0] (
05:05:26 Join ze [0] (
05:05:26 Join joshin [0] (n=joshin@unaffiliated/joshin)
05:05:26 Join daurnimator [0] (n=fake@unaffiliated/daurnimator)
05:05:26 Join BobShield [0] (
05:05:26 Join Siltaar [0] (
05:05:26 Join Rob2222 [0] (
05:05:26 Join maffe [0] (
05:05:26 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
05:05:26 Join GodEater_ [0] (n=bryan@rockbox/staff/GodEater)
05:05:26 Join tedr0ck [0] (
05:05:26 Join midgey34 [0] (
05:05:26 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
05:05:26 Join maddler [0] (
05:05:26 Join secleinteer [0] (
05:05:26 Join Paprica [0] (n=Paprica@rockbox/developer/paprica)
05:05:26 Join _jz [0] (
05:05:26 Join RaRe [0] (
05:05:26 Join FOAD [0] (
05:05:26 Join morrijr [0] (
05:05:26 Join gtkspert [0] (
05:05:26 Join luckz [0] (
05:05:26 Join Neovanglist [0] (
05:05:26 Join krazykit [0] (
05:05:26 Join Nimdae [0] (
05:05:26 Join annulus_ [0] (
05:05:26 Join alienbiker99 [0] (
05:05:26 Join midkay [0] (n=midkay@rockbox/developer/midkay)
05:05:26 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
05:05:26 Join Soap [0] (n=Soap@rockbox/staff/soap)
05:05:26 Join GodEater [0] (n=bryan@rockbox/staff/GodEater)
05:05:26 Join Galois [0] (
05:05:26 Join gxti [0] (n=gxti@foresight/developer/gxti)
05:05:26 Join Hadaka [0] (
05:05:26 Join gnakijefh [0] (i=0@
05:05:26 Join markun [0] (n=markun@rockbox/developer/markun)
05:05:26 Join crwl [0] (
05:05:26 Join kclaf [0] (
05:05:26 Join Ave [0] (
05:05:26 Join Slasheri [0] (i=miipekk@rockbox/developer/Slasheri)
05:05:26 Join sbeh [0] (
05:05:26 Join Rick [0] (i=rick@unaffiliated/rick)
05:05:26 Join Torin [0] (
05:05:26 Join sneakums [0] (
05:05:26 Join bagawk [0] (n=lee@unaffiliated/bagawk)
05:05:26 Join krylon [0] (
05:05:26 Join Seed [0] (
05:05:26 Join TiMiD[FD] [0] (n=TiMiD[
05:05:26 Join _pill [0] (
05:05:26 Join Bjoern-Erik [0] (
05:05:26 Join tris [0] (i=tristan@
05:05:26 Join preglow [0] (
05:05:26 Join gregj [0] (
05:05:26 Join Caliban [0] (
05:05:26 Join alberink [0] (
05:05:26 Join _w0rd54 [0] (
05:05:26 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
05:05:26 Join andrewg867 [0] (
05:05:26 Join jurrie_ [0] (
05:05:26 Join indro [0] (i=indro@
05:05:26 Join enyc [0] (
05:05:26 Join guyzmo [0] (
05:05:26 Join Overand [0] (i=overand@pdpc/supporter/active/Overand)
05:05:26 Join tumu [0] (
05:05:26 Join Presence [0] (
05:05:26 Join sslashes [0] (i=sslashes@
05:05:26 Join Battousai [0] (n=bryan@gentoo/developer/battousai)
05:05:26 Join SaLoMoN [0] (
05:05:26 Join Weiss [0] (
05:05:26 Join blithe [0] (
05:05:26 Join pabs [0] (
05:05:26 Join mtoups [0] (n=mtoups@HENSON.ISR.CS.CMU.EDU)
05:05:26 Join pixie [0] (
05:05:26 Join [mbm] [0] (i=mbm@openwrt/developer/mbm)
05:05:26 Join Kohlrabi [0] (
05:05:26 Join dionoea [0] (
05:05:26 Join feisar [0] (
05:05:26 Join DiDjCodt [0] (
05:05:28linuxchuckahh, gotcha.
05:05:59linuxchuckI understand. I'll get the simulator up and running first then.
05:18:47 Join webguest05 [0] (i=43baf8f7@gateway/web/cgi-irc/
05:19:23webguest05I was wondering if I could get some help with album art on rockbox
05:20:02webguest05i was wondering if i get some help with album art on rockbox
05:21:04 Quit webguest05 (Client Quit)
05:27:27 Quit midgey34 ("Leaving.")
05:36:58 Join bagawk_ [0] (n=lee@unaffiliated/bagawk)
05:47:21 Quit rotator ()
05:50:13 Quit bagawk (Read error: 110 (Connection timed out))
05:54:10TiMiD[FD]linuxchuck: if your code only touches files and spit a playlist file, then it's very probable that it'll work the same in the sim as on the device, but when you have a running code I recommend to test it on your player because sometimes there are surprises, like stack overflows or weird crashes
05:54:39 Join midgey34 [0] (
05:55:48linuxchuckI understand. Basically, I planned to develop it on my machine first because it is also a learning tool for me to get myself beyond the limits of what my C programming books can teach. I thought that if I was successful in making it work on a typical linux box, then I should consider porting it over to rockbox.
05:57:09TiMiD[FD]you can start with rockbox on the sim
05:57:13TiMiD[FD]it's not very difficult
05:57:24linuxchuckIt's technically my first "real" project outside of the canned ones provided in the books. It's the first one I've actually had a personal interest in making work. So hopefully, I'll stick with it, and make it into something functional and useful.
05:57:42TiMiD[FD]when you don't know how to do something just look how they do in the other plugins
05:58:23TiMiD[FD]are you new to programming too ?
05:59:04linuxchuckwell... that kind of depends. I'm new to c. I can make bash scripts sing and dance, and have a fair hand at PHP.
05:59:16linuxchuckSo programming concepts are not new to me, but I am a complete newbie to c.
05:59:44TiMiD[FD]so it shouldn't be that hard to get something working !
05:59:58TiMiD[FD]good luck
06:00:05linuxchuck:-) Thanks!
06:05:53 Part Soul-Slayer
06:22:51 Quit thegeek (Read error: 110 (Connection timed out))
06:22:51 Quit midgey34 (Read error: 104 (Connection reset by peer))
06:36:22 Quit bdgraue (Read error: 113 (No route to host))
06:42:33 Join thegeek [0] (
06:43:42 Quit linuxchuck ("g'nite")
06:52:49 Join midgey34 [0] (
06:55:31 Join ptw419 [0] (
06:56:53 Quit ShadowXP (Read error: 104 (Connection reset by peer))
06:56:57 Join ShadowXP [0] (
07:05:22***Saving seen data "./dancer.seen"
07:09:53 Quit Febs (Read error: 110 (Connection timed out))
07:10:51 Join daurn [0] (n=fake@unaffiliated/daurnimator)
07:19:54 Quit daurnimator (Connection timed out)
07:21:04 Join davina [0] (
07:22:06 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
07:27:14 Join scorche` [0] (n=scorche@rockbox/administrator/scorche)
07:32:02 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
07:32:23 Join miepchen^schlaf [0] (
07:46:37 Join BwS [0] (i=450ad212@gateway/web/cgi-irc/
07:47:38BwSHello, anyone here?
07:47:52 Quit BwS (Client Quit)
07:48:18scorche`gotta love those patient ones...
08:05:43 Quit kubiix (Read error: 110 (Connection timed out))
08:28:17GodEater_hit and run IRCers
08:30:19 Join Slowking_Man [0] (i=slow@wikipedia/Slowking-Man)
08:33:30 Join bluebrother [0] (
08:36:33 Join Rob222241 [0] (
08:38:54 Join Doomed_ [0] (
08:42:31 Quit midgey34 ("Leaving.")
08:45:48 Join midgey [0] (
08:49:49 Join ender` [0] (
08:51:23 Nick bagawk_ is now known as bagaw (n=lee@unaffiliated/bagawk)
08:51:29 Nick bagaw is now known as bagawk (
08:52:52 Join webguest85 [0] (i=c023111d@gateway/web/cgi-irc/
08:53:34 Quit Rob2222 (Read error: 110 (Connection timed out))
08:56:47 Quit alienbiker99 (Read error: 110 (Connection timed out))
09:00:43 Join MacGyver88 [0] (
09:00:49MacGyver88hello again!
09:01:48 Join petur [0] (n=petur@rockbox/developer/petur)
09:03:02MacGyver88do you know if there's an avi plugin for rockbox?
09:03:19peturthere is none
09:03:21scorche`no as in "rockbox cant"
09:03:24MacGyver88I mean, can I watch .avi movies on my iPod with Rockbox installed on it
09:05:24***Saving seen data "./dancer.seen"
09:05:25LloreanConvert them to mpeg
09:05:40LloreanThe iPod isn't powerful enough to scale them automatically anyway, so you'd have to convert them to at least the right size.
09:05:55MacGyver88ah that's the reason!
09:06:17LloreanThere's no sense doing a lot of work to support multiple formats if people *have* to convert anyway, right?
09:07:17*markun wonders if some people don't know that avi is a container format
09:07:30MacGyver88I didn't know that players aren't powerful enough to scale them.
09:07:39MacGyver88markun, i didn't know it :)
09:07:45scorche`markun: i would be willing to bet that most dont
09:07:49MacGyver88I'm not so expert in this kinda stuff
09:07:58markunThe Gigabeat S will probably be powerful enough
09:07:59LloreanYeah, AVI can hold many formats.
09:08:10LloreanThough most commonly these days it includes XViD or DiVX
09:08:15LloreanAt least, as far as I've seen
09:08:28peturmarkun: but mpegplayer doesn't read it, even if it contains an mpeg stream, right?
09:08:42markunbut mp3 is not the only popular audio codec used. I see a lot of ac3
09:08:54MacGyver88by format you mean kind of compression?
09:09:08LloreanMacGyver88: Yeah.
09:09:22MacGyver88I know that AVI can hold XviD or DivX
09:09:37MacGyver88so that's the main reason.
09:09:48MacGyver88pity :(
09:13:57LloreanAs I said, you'll have to convert 'em anyway, right?
09:14:02markunLlorean: more work for him :)
09:14:05 Join n1s [0] (
09:14:13MacGyver88right, markun... :)))
09:14:15Lloreanmarkun: It's only more work if he didn't need to convert.
09:14:39markunwell, his player has a chip which is powerful enough to do that, so it's a pity we can't make use of that
09:14:40LloreanThere's no different in the amount of work to convert, though it may take a little more (or possibly less) time depending on the speed of your encoder.
09:14:59LloreanI doubt the Gigabeat could decode MPEG-4 at 512x384 and scale to 320x240
09:15:13markunThe Gigaeat S?
09:15:16LloreanOh, the S
09:15:27LloreanIt *might* but it's iffy I think
09:15:38markunI believe the i.mx31 has special video hardware
09:15:45LloreanAh, well if you can access that, then sure
09:17:20GodEater_that would be nice - our first dap with h/w accelerated movie playback
09:17:54scorche`once again, assuming we can access it =)
09:18:19scorche`and someone writes the S-specific code
09:18:42GodEater_I was under the impression the h/w in the S was documented ?
09:19:28*scorche` doesnt know and is reading from his List O' Responses (TM)
09:23:18 Join bdgraue [0] (
09:23:34GodEater_is it possible to restrict a viewer to opening only one file, rather than <anything>.<extension the viewer is associated with> ?
09:24:40 Join B4gder [0] (
09:25:44*scorche` hands B4gder a cup of coffee and wags his tail
09:26:28B4gderwhat more can a man want?
09:26:41*GodEater_ has a long list
09:26:50B4gderyou're both crazy
09:26:50GodEater_which co-incidentally starts with women ;)
09:27:14*petur would be happy with more time (to begin with)
09:27:21 Quit MacGyver88 (" mIRC 6.21 in Italiano by TuttoIRC & ^DaNGeR^")
09:27:37GodEater_petur: that's second on my list
09:27:44scorche`well, technically, all i require is a hole and a few other interactable parts, but i think i am getting dangerous with the logs being saved...
09:27:56*scorche` issues an admin /clear
09:28:29GodEater_my eyes feel soiled
09:31:11 Quit lostlogic (Remote closed the connection)
09:31:27*scorche` apologizes for ending the conversation
09:31:27 Join lostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
09:32:53*scorche` wags his tail a bit more
09:33:33n1shmm, does the antiskip buffer thingy make sense for any flash targets?
09:33:57GodEater_none at all as far as I can see
09:34:17GodEater_if you're shaking your DAP so hard the electrons in it are jiggling, then you've got bigger problems than skipping music
09:34:47GodEater_like possibly the event horizon you're crossing...
09:34:48*scorche` throws a flash chip into a pool to see if it will "skip"
09:35:45*GodEater_ notes no-one answered his earlier question, and assumes therefore that the answer is "no"
09:36:03amiconnn1s: It does not, and iirc the Ondio didn't have it (for exactly that reason) before JdGordon's settings rework
09:36:41scorche`GodEater_: i cant see why you cant...(i was thinking of a way to say that soundly without a number of negatives)
09:37:07n1samiconn: the reason I'm asking is that apparently the langv2 cleanup patch removes it for ondio, but I think that should be a separate change, which I will do now
09:37:38amiconnpixelma had problems when trying that yesterday
09:38:01amiconn(_with_ the langv2 patch applied)
09:38:04n1samiconn: yes, I read her reply in the tracker
09:38:30n1sso sould I introduce a HAVE_FLASH_STORAGE or something?
09:38:57n1sor CONFIG_STORAGE = FLASH, maybe
09:38:58amiconnYes, probably
09:39:10amiconnBut checking for that might be difficult
09:39:21B4gderthe server list of "old" daily builds now also include map file zips
09:39:30amiconnThe Ondio sets HAVE_MMC, and the Sansa and iFP also have their own definitions
09:39:39B4gderwhich could be useful to track down addresses reported in bug reports etc
09:40:06amiconnBut the Nano, while being a flash target, uses a flash chip that's controlled by ATA commands and hence uses the standard ATA driver
09:40:23GodEater_scorche`: well how is that done in the viewers.config then ?
09:40:45n1samiconn: I meant put HAVE_FLASH_STORAGE in the config*.h files for those targets to make the check simpler
09:41:07scorche`GodEater_: oh...i was thinking code-wise
09:41:27Lloreanamiconn: Maybe something like NO_SPINUP ?
09:41:43GodEater_scorche`: I've already done that bit - what I want now is some way of stopping rockbox launching my plugin on files which don't match my explicit file name
09:42:07GodEater_although I guess I could just silently exit again
09:42:18amiconnn1s: Hmm... maybe that makes sense. Should be set for the sim then as well
09:42:36B4gderall salute scorche, now known as committer #55
09:42:49GodEater_hmm, no - that wouldn't work
09:42:57amiconnPerhaps the other definitons could be removed now with target tree in place... not sure though
09:43:13*GodEater_ waits for scorche` to change his nick
09:43:24*scorche` vows not to Screw Things Up (TM)
09:43:43scorche`GodEater_: this isnt my main box, so i dont use my main nick
09:43:47amiconnHmm, the hack in ata.c is another hint towards the recent nano problem
09:44:04peturB4gder: it would be nice to have a list of committers somewhere, I find myself using ohloh to look up if somebody already has access or not (I tend to forget a lot - bad memory)
09:44:10GodEater_scorche`: doesn't matter. Bagder has decreed you're now known as commiter#55
09:44:24amiconnB4gder: Do you know enough about ATA so you could answer questions about the init process, or do I have to wait for LinusN?
09:44:32*scorche` doesnt quite get GodEater_
09:44:41*amiconn also wonders where LinusN disappeared...
09:44:43B4gderamiconn: you'd have to bother linus
09:44:45GodEater_scorche`: never mind
09:44:45*scorche` "oh's"
09:44:54Lloreanpetur: 40 of the committers are registered as Developers in the forums
09:45:29B4gderpetur: the list is in the script, that is here:
09:45:31n1samiconn: apparently we have a HAVE_FLASH_DISK which is only used for the ifp (and only checked once) maybe that could be extended to other targets...
09:46:02peturB4gder: thanks!
09:46:08amiconnThat also defines the driver iirc (like HAVE_MMC on Ondio)
09:46:19amiconnSome code parts depend on HAVE_MMC
09:47:19amiconn(like the chunked voice file load, or the check whether to use the serial)
09:47:23 Join pondlife [0] (
09:47:49n1samiconn: oh, well, HAVE_FLASH_STORAGE it is then
09:47:57amiconnI think the best way to clean up this mess is to complete the target tree move for the old targets
09:47:57 Quit webguest85 ("CGI:IRC (EOF)")
09:47:59*amiconn sighs
09:49:42 Join christian_ [0] (
09:49:44 Quit darkless ("Leaving")
09:49:50 Join darkless [0] (
09:51:31christian_hi, I've got a question: How do I rate songs on my ipod running rockbox? And are there any programs to sync those ratings back to the pc?
09:52:24amiconnLlorean: around?
09:54:26 Quit hannesd (Read error: 104 (Connection reset by peer))
09:55:03*pondlife finally did the binary chopping required to work out what broke playback stopping on the sim, and it was
09:55:42pondlifeProbably just indicates taht the sim doesn't do queue_send() properly.
09:57:02pondlifeThis would explain several sim segfaults.
09:57:04christian_is anybody able to help me out on the problem above? I'd really appreciate it.
09:57:40peturchristian_: I don't know of any way to do what you asked
09:57:48pondlifechristian_: Ratings are available in the context menu, you need to be using the database and have collect runtime data enabled.
09:57:53pondlifeNo way to sync them to the PC AFAIK.
09:58:32 Join hannesd_ [0] (
09:58:45petur(my reply was only about the syncing to pc)
09:58:59Lloreanamiconn: I'm here
09:59:01pondlifeMaybe an FR for a plugin to copy the current ratings into track tags?
09:59:02*petur yawns and tries to wake up
09:59:11*pondlife slaps petur with a beer
09:59:27peturtwo triggers in one line!
09:59:27christian_whats an FR?
09:59:38pondlifeFeature Request
10:00:02pondlifeAlthough we're trying to clear them out this week, so maybe wait a while... ;)
10:00:06amiconnLlorean: Do you have new information about the problematic nanos?
10:00:09christian_ah alright! thanks a lot guys!
10:00:56 Join aliask [0] (
10:01:04amiconnpondlife: There was a tag editor plugin patch iirc
10:01:10 Quit preglow (
10:01:11 Nick hannesd_ is now known as hannesd (
10:01:25 Quit hannesd (Killed by (Nick collision))
10:01:28NJoinpreglow [0] (
10:01:28amiconnI think this could be extended to take its input from elsewhere, e.g. the db
10:01:52 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
10:01:52amiconnOf course it would also need to be extended for other tagging formats than id3
10:02:06JdGordonevening all
10:02:15pondlifehi jdGordon
10:03:05 Join webguest85 [0] (i=c023111d@gateway/web/cgi-irc/
10:04:29 Join BHSPitMonkey_ [0] (
10:05:04Lloreanamiconn: Unfortunately, no. No new posts or responses to the tracker entry
10:05:47pondlifeSo it seems to be related to the Apply flash loader version? Or is that just a theory?
10:05:54pondlifeApple, even..
10:06:33*petur read that as apple flashy loader
10:06:44pondlifeWell, there's a logo...
10:09:26amiconnThat teory has proven wrong
10:09:32amiconn*theory even
10:09:58pondlifeSo is it hardware revisions? :(
10:10:02amiconnBut the problem might be related to the factory installed flash version, as that may depend on the production batch
10:10:29amiconnI suspect the latest (?) batch uses a different flash disk that's more susceptible to timing issues
10:10:50amiconnI am trying to understand PP ata timing setup right now
10:12:40 Join hannesd_ [0] (
10:12:45 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
10:18:08 Quit BHSPitMonkey (Read error: 110 (Connection timed out))
10:18:38JdGordonare the addresses in the plugin .maps before being loaded? I'm trying to find 0xea00000 but cant see it in or
10:20:35amiconnThe addresses in the plugin .maps are where they are loaded. Plugins and codecs are linked to a fixed address
10:20:54JdGordonok, thats what I thought, but wanted to double check
10:23:23JdGordonhmm, the area around 0xea00000 isnt mentinoed in any of the maps... something is doing something very wrong :p
10:23:57amiconnDidn't you overlook a digit in that address? Addresses should have 8 digits...
10:23:58B4gderea sounds like an instruction
10:24:12 Join obo [0] (n=obo@rockbox/developer/obo)
10:24:59JdGordonthe panic is "Undefined instruction at EA000004", am I going about finding the cause the entirley wrong way?
10:25:27thegeekamiconn: keep up the great work on pp:)
10:25:38 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
10:26:19B4gderJdGordon: no, but it looks like it jumped out it the woods
10:26:36B4gdernow where is that typing school?
10:27:35*JdGordon not sure where that "in" fits to make the sentance make sense :p
10:28:08pondlifeThe woods are on the .map?
10:28:10B4gdersee it as a challange for the reader
10:28:25B4gderit matches my other lines: hardly any sense
10:28:55B4gderI should just drink coffee and not type
10:29:17*pondlife should drink tea, but has no mugs. Even his Rockbox mug has been packed away.
10:29:28pixelmaB4gder: beware of DerPapst's typo class ;)
10:29:39pondlifeteh best class ever
10:31:01amiconnJdGordon: That almost looks like it loaded an instruction as an address
10:31:27JdGordonamiconn: the microsecond timer doesnt work in sleep mode? is there any way to keep track of time while in sleep mode then?
10:31:40B4gderamiconn: exactly my feeling
10:31:54B4gderwhich could imply a stack overflow or similar
10:32:06B4gderhm, perhaps not
10:32:11JdGordonloading it onto my nano so I can try reproducing
10:32:20 Join gunluva [0] (i=467c11a3@gateway/web/cgi-irc/
10:32:29*JdGordon slaps slow download server... 165kb/s just isnt fast enough
10:32:53B4gderhaha, try again and you might get a faster one ;-)
10:32:58gunluvaHello, is rockbox being developed for Sansa c250?
10:33:10B4gdergunluva: not until you start it!
10:33:25gunluvaWhat would I need for that?
10:33:28B4gderand lowlight did get one he said, so there's hope
10:33:51B4gdergunluva: some skill, time and reading the NewPort wiki page
10:34:42pixelmasomeone started once
10:34:58B4gderyeah, e200tool etc works on it
10:35:14B4gderwe can in fact put a bootloader on it
10:35:34B4gderalthough we don't have code for its LCD so he never passed that part
10:35:38gunluvaOn a c200?
10:35:57B4gdersee that forum thread
10:36:01B4gderand the wiki page for the c200 series
10:37:28amiconnHrrrm, that 1st gen I bought (or thought to) turned out to be a 2nd gen :(
10:37:50*pondlife does the new port dance
10:38:14amiconnJust a single platter model...
10:38:22Lloreanamiconn: Misleading auction?
10:38:36B4gderargh, annoying
10:39:03amiconnMaybe the seller doesn't know the exact difference. I have to contact him...
10:40:26JdGordonhmm, cant reproduce the panic here... and the main thread is only at 46% after leaving mpegplayer so doesnt look like a stkov
10:40:34gunluvaSo what does e200tool do?
10:41:58JdGordone200tool fixes broken e200s (its a godsend :D )
10:42:26gunluvaSame for c200s?
10:42:38LloreanThough as a general rule the only people who break their e200s badly enough to need it are people who aren't able to follow the instructions to use it. ;)
10:42:49gunluvaGood. Mine is acting up.
10:43:09B4gderright, but people going for a port to c200 might need e200tool at times... :-)
10:43:18JdGordonLlorean: or just do something really stupid, usually accidently
10:43:37LloreanJdGordon: Just meant on average
10:43:42gunluvaAll I need is a firmware copy?
10:43:51B4gdergunluva: to do what?
10:44:01LloreanIt'll likely be very useful for porting the c200, and is almost certainly going to be essential to whoever ends up getting the e200R working
10:44:38gunluvaTo reset my c200.
10:45:11B4gderno, to "upgrade" to a new firmware you just use recovery mode
10:45:33B4gder"reset" isn't really anything but power off/on
10:46:03gunluvaMy player keeps powering itself on and off.
10:46:12gunluvaI haven't done anything to it yet.
10:46:14B4gderthat sounds like a bug
10:46:17B4gdersw or hw
10:46:31gunluvaI never had the prob before.
10:46:38gunluvaIt just started todat.
10:46:48B4gderthen it sounds like a hw bug
10:46:53LloreanI knew someone with a c200 with a similar problem, but they'd gotten it wet.
10:47:08gunluvaMines been dry.
10:47:43B4gdergunluva: you experienced in low level/embedded programming ?
10:47:45 Join ackbahr [0] (
10:48:57gunluvaThat's a negative, good buddy.
10:49:12B4gderoh, ok
10:49:34aliaskI have a C200 floating around at home, but I don't think my mum would appreciate me breaking her mp3 player :P
10:50:25gunluvaWell, it seems to be working now.
10:50:50gunluvaI went into recovery mode, and disconnected it after a while. Without changing anything.
10:50:56*n1s crosses fingers
10:51:02gunluvaWe'll see how long that lasts.
10:52:00gunluvaI'm currently adding my music back in.
10:52:05B4gdern1s: nice commit
10:52:47gunluvaNo spontaneous disconnects yet.
10:52:55n1sB4gder: thanks
10:53:16peturB4gder: does the current build page need manual tweaking to remove the gap?
10:53:41B4gderno, it'll go away when the older name no longer exists in the build logs
10:54:01peturis that log longer than the table?
10:54:06B4gderyes, much longer
10:54:17B4gderI think 7 days
10:54:49B4gdermainly to allow you to browse somewhat older build logs etc even when they run out from the table
10:55:01gunluva*sigh* I has oo many music.
10:55:07JdGordonamiconn: telling time while cpu is sleeping? do I need to use the rtc?
11:00:25gunluvaDisconnected itself.
11:00:30gunluvaAnd shut off.
11:00:51gunluvaPower button's not working.
11:02:04pixelman1s: that was beyond my knowledge... :)
11:02:56 Quit christian_ ("Leaving")
11:02:58pixelmabut crossing fingers didn't help
11:03:01n1spixelma: I will make a new patch when I have fixed the red for sansa
11:03:13n1snew lang cleanu patch*
11:03:24gunluvaWhere can I get a copy of the 1.00.04A firmware?
11:03:44B4gdergunluva: we talk rockbox here, you know
11:04:17gunluvaSorry. Am newb.
11:04:40scorche`gunluva: the guidelines of the channel are linked in the topic if you havent read them
11:05:25***Saving seen data "./dancer.seen"
11:06:00gunluvaWait, what'd I do wrong?
11:07:03B4gderwe try to keep the topics here rockbox-related
11:07:15gunluvaOh, I see.
11:07:20gunluvaNevermind, then.
11:08:10 Quit jhMikeS (Nick collision from services.)
11:08:15scorche`hrm...i dont remember the build table actually counting down live...
11:08:16 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
11:08:54pondlifescorche`: It's always over optimistic too - at least when I've watched it.
11:09:09scorche`better than pessimistic
11:09:28pondlifen1s: red part 2
11:09:30scorche`at least 30 seconds optimistic there
11:09:51n1spondlife: already committed :-)
11:11:20pondlifeI also committed a fix-ish thing for the sims... so further red might be my fault.
11:11:36 Quit tedr0ck (Client Quit)
11:12:04pondlifen1s: You may want to retest FS #6691 now too.
11:13:04pondlifeJust in case it was misdiagnosed.
11:13:28n1spondlife: will do
11:14:32n1spondlife: wow, one-word fix killing two bugs
11:14:41*pondlife would really like a brave soul to let the sim's use the "normal" threading code.
11:14:59pondlifeIt's a hack, really - the problem is fundamental though.
11:15:35B4gderyes, it would be good if it could use the real threading
11:15:52pondlifeStill, 2 less bugs on Flyspray...
11:16:06B4gderI've been pondering if setjmp/longjmp could be used for it
11:16:22pondlifeAh, does the build estimated time round up to the next minute?
11:16:52pondlifeWeird. I have noticed HH:MM:00 as an end time twice now.
11:17:15pondlifeThat could explain an over-optimistic-ness by an average of 30s.
11:17:18B4gdergreen and clean again
11:21:50*B4gder plays with adding the rev number for each build download
11:22:03B4gdersince they're not removed now prior to builds
11:23:12bluebrotherhmm, having a build-info file for the current builds would be a nice addon. That way rbutil could display the current svn revision number
11:23:58n1spixelma: new patch posted
11:24:30B4gderbluebrother: easily provided of course
11:24:40amiconnB4gder: The rev numbers might also be useful for the daily^H^H^H^H^Harchived builds
11:24:55B4gderyes, I'm planning to add them there as well
11:25:27 Join Thundercloud [0] (
11:27:02 Quit gunluva ("CGI:IRC (EOF)")
11:27:03 Join Febs [0] (
11:27:45peturB4gder: also on the front page / last 4 weeks page ?
11:27:54n1spondlife: FS #6691, still crashes
11:28:10B4gderpetur: you mean just telling the current rev number ?
11:28:19pondlifeI thought it would - that's probably the libmad optimisation then.
11:28:49peturB4gder: I mean in the table (maybe on top of the text that mentions the changes or the changed files)
11:29:13B4gderI'll work on it
11:29:59pondlifeMaybe change the word "Daily" to be "Archived" too.. to emphasise their out-of-date-ness?
11:30:02bluebrotherB4gder: what do you thing about using the svgs for the images in the table?
11:30:08pondlifeAnd the url:
11:30:20bluebrotheri.e. use a scaled-down version of them.
11:30:29bluebrothercould even make use of transparency ;-)
11:30:45B4gderbluebrother: sure, would be cool
11:31:07B4gderwouldn't it make sense to generate pics from them and use?
11:31:13bluebrotherdo you have a preferred size for the images? I can try making some later today.
11:31:35B4gdermost of the existing are roughly 60x80
11:31:37 Join obo_ [0] (n=obo@rockbox/developer/obo)
11:31:58bluebrotherok, then I'll do something similar sized.
11:31:59 Quit obo (Read error: 104 (Connection reset by peer))
11:32:00pondlifeB4gder: Those table cells are too tall...
11:32:07 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
11:32:11B4gderI'm experimenting a bit atm
11:32:19B4gderI took those high pics away again
11:32:27pondlifeOK, experiment on!
11:32:29B4gderI considered using them instead of the text
11:33:11pondlifeText is possibly better for blind people.
11:33:21B4gdernot with the proper alt texts etc
11:33:35amiconnB4gder: Imho the script on the build page checks too often if the builds are overdue
11:33:48B4gderyou mean the 10 second interval?
11:33:51amiconnIt checks every 10 seconds, I think 30 seconds should be enough
11:34:26amiconnThe 10 seconds might put quite some load on the server, especially if the firefox bug hits which makes it reload the whole page including images
11:34:30B4gderwill take away lots of load on the server when plenty people waits for an overdue round
11:34:34amiconn(get that here from time to time)
11:34:44B4gderI don't think it is a bug
11:34:48B4gderit is what the javascript does
11:35:25B4gderit reloads everything without considering cache
11:35:33 Join desowin [0] (
11:36:14pondlifeWhy does it need to reload before the build completes?
11:36:26B4gderto show the updated table
11:36:28amiconnB4gder: You could leave out the parameter
11:36:35pondlifeI mean now the old build is preserved, you don't see the table growing like you used to.
11:36:40B4gderah, no not before the build is done
11:36:42amiconn'true' forces a reload
11:37:05B4gderpondlife: they're no longer removed before the build so the old zips are still present until overwritten by the new one
11:37:06amiconnwindow.location.reload() is a non-forced reload
11:37:22B4gderok, I switch to that now
11:37:49pondlifeB4gder: Yes, so no need to do *any* reload until the build completes...?
11:37:52amiconnB4gder: Btw, I find it irritating that at the time the script does its last reload, the delta table is not yet updated
11:38:09B4gderpondlife: well, now there's a rev number added so you can in fact see them change when the new zips fly in...
11:38:15amiconnSo I have to update once more anyway
11:38:20pondlifeB4gder: Ah, gotcha
11:38:37bluebrotherhow about adding the release numbers to the size table as well?
11:38:39amiconnwindow.location.reload explained (in German):
11:41:18B4gderI read plenty explanations
11:42:08scorche`while we pick on tables on the site, is it reasonable to have (on /irc) links for or links to switch the table to link to viewraw without going to the .pl first?
11:42:11B4gderbluebrother: ah, yes I can do that now when I get the rev number... I'll start logging the rev number in the size dumps
11:42:18 Join Entasis [0] (
11:43:28jhMikeSamiconn: hello. around now.
11:47:41scorche`bluebrother: re rbutil.php, once i get off of vacation, i will hurry to fix a few things and implement a few back-end changes, then commit...would you like the php files now to start working on that though?
11:47:44jhMikeSabout 13h late :)
11:48:29bluebrotherscorche`: I could do that and implement it the same time in rbutil −− so if you have the files at hand it would be nice.
11:48:30B4gderany suggestion on how to provide the rev numbers for daily builds? I mean, should I make them available on the download mirrors as files somehow?
11:48:46bluebrotherotoh there is enough other stuff to do, so no hurry
11:49:00B4gderlike I could make simple files named "date-$date-is-rev-$rev" or whatever
11:49:15scorche`bluebrother: it is just transferring a few files atm, so i am fine with that =)
11:49:15bluebrotherB4gder: can you create symlinks on the mirrors?
11:49:26B4gderI'm not sure
11:49:30bluebrotherscorche`: great.
11:49:46B4gderstill, the order is date to a rev and we stored the builds on dates
11:49:49scorche`are you able to receive DDC?
11:50:03bluebrothernot sure −− I'm behind a NAT
11:50:19B4gderreceiving DCC is fine behind nats
11:50:36B4gder(the connect procedure is the reverse to what you'd expect)
11:50:39bluebrotherwell, I actually never tried with this setup
11:52:16scorche`bluebrother: well, i can give it to you now, but i will be changing a few things in it...(mainly a few vars and organization of)
11:52:52bluebrotheranything that would make it hard to merge later?
11:53:03scorche`not too hard
11:53:19bluebrotherok. Then let's just try and feel the pain later on ;-)
11:53:31scorche`i am not at home, so it may not like me even sending the file...
11:53:44linuxstbB4gder: Isn't there a "build-info" or similar file already available with details of the daily builds?
11:53:49bluebrotheryou could also email it to me
11:54:03scorche`can you PM me the address?
11:54:14amiconnjhMikeS: What would be needed to run the pp emu? And is it possible to log accesses to certain register ranges and where they came from?
11:54:14B4gderlinuxstb: there is, but that keeps the info only for the most recent build
11:54:28amiconnAlso, which PP model(s) can be emulated, and what other hardware?
11:54:51linuxstbB4gder: Could you just extend that?
11:55:12B4gderlemme see what I can do...
11:55:14jhMikeSamiconn: you need a partition dump from a sansa. it does log register access and is supposed to log the code that is running.
11:55:38amiconnHmm, I'd want to use it to run an ipod firmware...
11:55:44amiconn(or other PP target with ATA)
11:55:47jhMikeSI also suspect it could be adapted to run like another device but it's somewhat hard-coded atm.
11:56:31*amiconn has no experience with such emulators...
11:57:48 Join XavierGr [0] (
11:57:48jhMikeSyou'd have to probably setup some ata emulation. it just does the SD interface.
11:59:58jhMikeSamiconn: do you happen to know what the wrap period on USEC_TIMER is?
12:00:16amiconnYes, very simple...
12:00:19 Part maffe
12:00:26amiconnIt's a 32 bit counter
12:00:32 Join maffe [0] (
12:01:08B4gderbluebrother: will have the rev number after next commit, and is valid for the most recent commit build
12:01:09amiconnSo ~4294.96 seconds
12:01:52jhMikeShmmm...a bug reported with the e200 wheel suggests to me it could be sooner. hmmm.
12:02:01scorche`B4gder: did you see what i said about the irc page?
12:02:25scorche`while we pick on tables on the site, is it reasonable to have (on /irc) links for or links to switch the table to link to viewraw without going to the .pl first?
12:02:40pixelman1s: will try soon... just a note about the anti-skip setting: when I went through english.lang for the langV2 patch earlier this year, amiconn told me that anti-skip _might be_ usefull on swcodec flash players for crossfade. That's why I only excluded it for Ondio that time... not sure, someone should try (?). And as soon as there is a conclusion, we'd need to adapt the manuals btw.
12:03:06B4gderscorche`: ah, sure
12:03:37linuxstbscorche`: You also prefer the plain text files?
12:03:59scorche`many times, yes....the .pl is quite slow at times
12:04:15scorche`especially compared to the instantness of the text files
12:04:27JdGordonthere is most definatly a mepgplayer bug on at least the nano :(
12:04:37JdGordonand of course, its nice and random
12:05:00jhMikeSJdGordon: what?
12:05:01B4gderscorche`: there!
12:05:28JdGordonjhMikeS: undefined instruction at 0xEA000004
12:05:46scorche`B4gder: i meant for all of the archived ones as well
12:06:00scorche`not sure if adding a whole other table would be reasonable though
12:06:07B4gderthat would be just a bit too much
12:06:29 Quit ackbahr ("ChatZilla [Firefox]")
12:06:38B4gderscorche`: check the "list all" button in the bottom and download them all...
12:06:42jhMikeSI think the cache coherency isn't quite right still. I've gotten some things but after _restarting_ the plugin. not the first time I run it and that what the telltale sign of that.
12:07:09JdGordonI've now had it crash the first load, the 6th load and the 2nd...
12:07:19scorche`B4gder: that works, thanks =)
12:07:42B4gdernow lunch
12:07:50jhMikeShmmm...what was move out of iram there?
12:09:37*jhMikeS found it.
12:12:58linuxstbjhMikeS: ?
12:15:43jhMikeSyou took libmad code out of iram on pp502x?
12:16:04linuxstbWhy should that cause problems?
12:17:04preglowman, dircache is failing a lot for me these days
12:17:17jhMikeSI don't know if it is the problem but if there were some cache releated issue to it since the code is now is cached memory, it might.
12:17:25preglowevery other restart requires a foreground disk scan, and every once in a while, dircache doesn't even work
12:17:47 Join barrywardell [0] (
12:18:00linuxstbjhMikeS: I thought the only cache issues were with respect to data used by the video thread.
12:18:41 Quit scorche` (" to bed, i suppose...")
12:19:23jhMikeSI'm just making guesses atm. Those are the sorts of errors that came up before related to caching though.
12:35:46 Quit Siltaar ("Aurevoir -")
12:35:57preglowTiMiD[FD]: sounds like what i get, yeah
12:37:01TiMiD[FD]that's what I've got some month ago
12:37:07TiMiD[FD]now it isn't that systematic
12:37:23preglownah, it's not really consistent
12:37:30TiMiD[FD]same with me now
12:37:45TiMiD[FD]but at that time I could reproduce it one launch on 2
12:38:54TiMiD[FD]would be interresting to do some tegression tests to track the problem
12:39:22TiMiD[FD]because that used to work pretty well before
12:40:20TiMiD[FD]by the way would it be possible for me to get some rights on the tracker ?
12:40:37TiMiD[FD]to close bugs for example
12:42:49linuxstbTiMiD[FD]: You should ping Bagder
12:43:01preglowwhat he said
12:43:44linuxstbpreglow: Have you heard from saratoga recently?
12:44:19 Quit midkay ("Leaving")
12:44:33 Join midkay [0] (n=midkay@rockbox/developer/midkay)
12:45:51TiMiD[FD]Bagder: ping
12:47:13preglowlinuxstb: nope, i'm gonna mail him now
12:47:47amiconnBagder: ping... The build system doesn't seem to pick up my commit....
12:50:03preglowhow are the rest of the gsoc projects doing anyway?
12:51:36GodEater_preglow: well no-one understands how Nico_P is coming along except JdGordon ;)
12:51:42linuxstbpreglow: Silently...
12:51:53GodEater_preglow: and no-one's heard about TTS except markun, who says it's "coming along"
12:52:09*GodEater_ struggles to remember what GSOC project No. 4 is
12:52:21GodEater_oh yeah
12:52:25GodEater_no idea how that's going either
12:52:30linuxstbObviously the students are all too busy working on their projects to come to IRC to chat about them...
12:52:55aliaskUSB is the one I anticipate most of all. Mainly because the Gigabeat S port relies on it.
12:53:13linuxstbAnd that's the real iMX.31?
12:53:17GodEater_relies on it how ?
12:53:38linuxstbYou won't be able to dual-boot?
12:53:41aliaskWell, at the moment there isn't any way to get a rockbox.gigabeat on to the player
12:53:54markunGodEater_: I got an email from Ozgur today
12:53:57linuxstbAh, the original firmware doesn't give you a UMS mode?
12:54:08linuxstbWhat about some kind of "data" mode?
12:54:08GodEater_ewww - so how do you get code to run on it then ?
12:54:19aliaskThe OF is shocking, dual booting will NOT be a priority
12:54:29linuxstbI thought MTP allowed storage of datafiles/
12:54:41markunGodEater_: "So I have been working busily on those requests and I will continue until tomorrow evening. Then I will again have more time for GSOC until August 7th."
12:54:44aliaskPerhaps it does, but it's obfuscated
12:54:55aliaskI've never actually thought of that to be honest
12:55:09linuxstbBut that's also a pain if you're not using Windows...
12:55:17aliaskYes. Which I'm not.
12:55:29markunnot not using? :)
12:55:37aliaskShush :P
12:56:02GodEater_markun: does that email actually give any indication of status though ?
12:56:03aliaskWell, I have a windows box, (which I need to get code onto the S) but it's downstairs - so I may as well not have one :P
12:56:14GodEater_all it says to me is "I'm still working on it"
12:56:27 Join Theft [0] (i=58c98044@gateway/web/cgi-irc/
12:56:51*linuxstb wonders how many stairs aliask has
12:57:07*aliask hasn't ever counted
12:57:15aliaskBut I'm getting fit running up and down them testing things
12:57:29TheftHi all. Somebody know what is "ipodnano maps"?
12:57:31linuxstbGet a very long usb cable and vnc...
12:57:40*GodEater_ wonders if aliask has thought of moving the windows PC upstairs
12:57:55aliaskI can't - it's my parents. But the USB idea is good :)
12:58:07linuxstbTheft: They're information for developers - if you're not developing rockbox you can ignore them.
12:58:40linuxstbTheft: They list where in memory all the different bits of Rockbox are - so they help debugging.
12:59:59GodEater_aliask: what about a window vmware image ?
13:00:59aliaskNot enough space on my harddrive.
13:01:09Theft<linuxstb> Thanks
13:01:51linuxstbaliask: Then you need to hunt down austriancoder and crack the whip...
13:02:18 Quit Theft ("CGI:IRC")
13:02:18aliaskHe seems to be making good progress
13:02:22markunGodEater_: not really. I asked him to come here and/or update the TTS wiki page
13:03:06GodEater_markun: so he's not sent you any code or anything ?
13:05:26***Saving seen data "./dancer.seen"
13:05:46*pondlife whistles "Money for nothing"...
13:05:56GodEater_that's what I was thinking
13:06:21pondlifeHow come Google are paying out for such cases?
13:06:35GodEater_it's the mentor's call I believe
13:06:37linuxstbIsn't there still the end-of-project assessment?
13:06:47GodEater_they're supposed to get marked halfway through, and then at the end
13:07:53markunGodEater_: He's sent me some code, but I didn't take a close look at it yet
13:08:37GodEater_markun: is there a reason you've not posted it anywhere where others could look at it ?
13:08:42linuxstbWhat's he actually doing? Is the first step a plugin?
13:09:39*GodEater_ doesn't want to seem like he's picking on markun, he's just interested
13:09:56*linuxstb too!
13:10:15markunlinuxstb: yes, still working on the plugin
13:11:28markunI can host the rar file he sent me if you want
13:11:45GodEater_why not put it on the wiki ?
13:11:58pondlifeOr FS with a ref from the wiki
13:12:21GodEater_even better
13:12:34linuxstbmarkun: Please convert to something nicer than rar though...
13:12:56amiconnIs there a problem with rar?
13:13:09amiconn7zip is able unpack rar files
13:13:11B4gderrar is evil
13:13:25amiconnInsert 'to' where it fits...
13:13:48GodEater_rar == cthulu of compression formats
13:13:55amiconnB4gder: What's up with the build server?
13:14:05B4gderwhat why?
13:14:18amiconnIt didn't pick up my commit
13:14:24markunhttp:// (or .tar.gz)
13:14:47amiconns/build server/build master/
13:14:53B4gderI broke the perl...
13:15:13B4gderthere, now perhaps
13:15:16GodEater_JdGordon: awake ?
13:15:23markunlinuxstb, GodEater_: got the link?
13:15:46GodEater_got it
13:15:48*pondlife assumes that Nico_P must still be on holiday.
13:15:54B4gderthe build is on
13:16:17GodEater_pondlife: I think so
13:16:18pondlifeHmm, it started the countdown at ~4:40
13:16:41pondlifeYet builds are taking ~6:09 on average
13:16:42B4gdernot here
13:16:50B4gderare your clock off?
13:17:02B4gders/are/is whatever
13:17:27pondlifeNow I get it...
13:18:52*GodEater_ wishes JdGordon had finished his UsingTheMenuAPI page :(
13:20:24 Join Siltaar [0] (
13:20:44linuxstbB4gder: Is the build countdown reliant on the user's clock?
13:21:12B4gderyes, since it sets a target ETA and counts down to that
13:21:24B4gderthe target time being set by the server
13:21:39B4gderso if the client diff from that, the delta time is wrong
13:22:06peturI don't have a good feeling about the USB gsoc project atm, communicating with austriancoder is very troublesome, haven't heard of him in 12 days :(
13:22:39peturGoogle must have payed....
13:22:39linuxstbB4gder: Couldn't you just tell the client "x seconds from now" ?
13:23:05B4gderlinuxstb: I guess I could, but then I'd need to decipher and rewrite the javascript I got ;-)
13:23:14linuxstb;) If I'm ever bored...
13:23:31B4gderand btw, "from now" doesn't work
13:23:35B4gdersince the page can be reloaded
13:23:49B4gderyou don't want the countdown to restart then
13:23:49preglowpetur: how's the progress anyway?
13:24:00peturnot working yet
13:24:10preglowyeah, but is there potential?
13:24:23peturand imho we should have been further
13:24:33linuxstbB4gder: I don't understand... Why should it restart? Shouldn't it just correct itself?
13:24:48peturpreglow: yes, if he gets his act together and get active a bit more
13:24:51B4gderlinuxstb: not if the HTML would say "done in N seconds"
13:25:02B4gderlinuxstb: since that would be N seconds from... when?
13:25:14preglowpetur: btw, i've had tripel karmeliet now!
13:25:18preglowit's great
13:25:21XavierGrIMHO all GSOC projects (except WMA) are rather late in development :\
13:25:29*petur gets thirsty
13:25:33linuxstbB4gder: The html wouldn't say "N seconds", the javascript would display the time dynamically.
13:25:52preglowXavierGr: well, i think they should at least have commited something by now
13:26:01linuxstbSo when the page is loaded, it contains a variable with the countdown time. The javascript adds that to the user's current time to display the ETA
13:26:06preglowit's hard to tell if they're late or not if they don't
13:26:10B4gderlinuxstb: yes, but when you manually reload the page before the countdown is completed, it'd still say "N seconds"...
13:26:31peturwe should have put forward deadlines before the start of gsoc
13:26:39peturnow it's a bit late
13:26:58XavierGrlet's hope that we will have an end result anyway...
13:27:12preglowat least we have wma :P
13:27:33XavierGryeah at last there is no reason to have the OF on H100 series :D
13:27:37peturwell if they fail to deliver, I say we fail them (no money from google)
13:27:38linuxstbpetur: It's not too late to say something like "if you don't achieve X, Y and Z before the end of the SoC, you will get an unfavourable report from us."
13:28:03linuxstb(there's still a fair amount of time between now and the end of the summer...)
13:28:20peturlinuxstb: the problem is that google is black/white, you either pass or fail them
13:28:27B4gderyeah, failing to deliver at least partly working code is indeed not too "favourable" in my view
13:28:50linuxstbpetur: Then we make it black/white - if they don't deliver then we fail them...
13:29:01peturthat I would do anyway
13:29:21linuxstbOr at least don't make sufficient effort to deliver...
13:30:00B4gderexactly, we need to take our role seriously and responsibly so that we might be able to take on this in the future as well
13:30:50preglowit's a learning experience anyway, we'll have a better idea what to expect the next time around
13:31:09peturI've been meaning to discuss some things in private with the other mentors, but the committers ML is not ok because the students are on there too (or most of them)
13:31:14preglowand perhaps even have some guidelines in place
13:31:33preglowhow many of us are here now?
13:31:35preglowthree out of four?
13:31:47preglowthen why don't we just go #mentor-lol or something?
13:31:56peturoh you mean here
13:32:02preglowindeed :>
13:32:16*petur kicks markun awake
13:32:25linuxstbDoes anyone have any idea how many hours students are supposed to be devoting to their SoC work? Is it intended to be a full-time job?
13:32:49peturI see it as a payed job
13:33:52preglowi really just see it as open-source coding with pay and a requirement to deliver
13:34:04preglowwhich means you do what you have to do to get things done your own way
13:34:13preglowyou should communicate, of course
13:34:39B4gder now contains the rev number of the latest commit build
13:35:04peturpeglow: my student fails horribly at that point
13:35:09B4gderfor anyone wanting it automaticly I mean
13:35:17preglowwhich isn't good
13:35:19 Join Nick_Brackley [0] (i=cb57494b@gateway/web/cgi-irc/
13:36:18preglowwhich is one thing i think we should emphasise next time around
13:36:33preglowwe do quite a bit of communicating in rockbox, and gsoc students should be required to too
13:36:39peturI stressed that point very clear on his evaluation
13:36:50B4gderI find it particularly strange in autriancoder's case since he's been involved in rockbox for quite some time before this
13:36:57preglowthen he'd better deliver nice stuff to get the final payment, if you ask me
13:37:19 Quit pondlife ("disconnected has pondlife")
13:37:23 Quit ender` (Read error: 104 (Connection reset by peer))
13:37:47peturwe were having a debugging session, and he just quit without saying anyting and hasn't been around since then (12 days). No response on mails as well
13:38:10peturmaybe somebody bombed his home?
13:38:26XavierGrlet's hope that it isn't something serious
13:38:30aliaskBut austriancoder has already got some decent results (IIUC), and he's been keeping his code available - he just hasn't been keeping his information on official rockbox channels.
13:39:16peturaliask: I don't say he's not delivering anything, bu I get the feeling we just lost 2 weeks
13:39:34preglowmight just be off on vacation or something
13:39:37preglowbut should have said something anyway
13:39:41peturhe might have told
13:39:47 Quit secleinteer (Connection timed out)
13:39:57 Join alarmclok [0] (n=jjjs@
13:40:01peturI went off but I did mail him that
13:40:09*petur awakes
13:40:23preglowyeah, i've been very careful to at least let saratoga now about my numerous excursions
13:40:33preglownot that he needed me around anyway...
13:41:18GodEater_this was observed at DevConPub 2, where I mistook linuxstb as the mentor for wma ;)
13:41:40B4gderI'm sure he made you suffer for that! ;-)
13:41:57GodEater_no - he was most gracious
13:42:21linuxstbGodEater also wondered what I was going to spend my $500 on...
13:42:28GodEater_he said something about "that git preglow" :)
13:43:25GodEater_another PS3 wasn't it linuxstb ? After you've appropriated the money from the rockbox fund of course...
13:43:26 Quit Nick_Brackley ("CGI:IRC (EOF)")
13:44:39 Quit webguest85 ("CGI:IRC (Ping timeout)")
13:44:58linuxstbAt least I got a mug and a t-shirt.. ;)
13:45:07GodEater_without having to arm wrestle too
13:45:32preglowthe mentor cash goes to the org anyway, no?
13:45:36preglowwho will i have to blackmail for it?
13:45:39GodEater_yes - this was explained to me
13:45:56GodEater_I was clearly feeling a bit slow that evening - as I think I already knew that
13:46:06peturonly the org gets money
13:46:24*petur assumes B4gder filled in the tax form
13:46:33linuxstbGodEater_: At least you could tell my left earphone from the right...
13:46:42GodEater_hahaha - there is that!
13:46:55GodEater_have you written it on them in something with a bit more contrast now ?
13:47:19linuxstbNo, I still just wear them randomly - to avoid noticing Rockbox bugs.
13:47:40 Join Jon-Kha [0] (
13:47:59preglowso you don't count channel swapping as a bug, then
13:48:07GodEater_what channel swapping ?
13:48:17linuxstbHappens every time I put my headphones on...
13:48:47linuxstbBut I can't say I care which channel the music comes out of...
13:49:01GodEater_variety is the spice of life after all
13:49:36preglowalso chili
13:50:23 Quit aliask ("ChatZilla [Firefox]")
13:53:01 Join webguest85 [0] (i=c023111e@gateway/web/cgi-irc/
13:56:58linuxstbmarkun: Are you around?
13:57:40amiconnpreglow: Speaking about wma - do you think it's possible to emac'ify the mdct functions like you did with the libmad ones?
13:58:16amiconnIt seems to me that wma doesn't have much different cpu power requirements than mp3 - but on coldfire libmad is way faster than libwma
13:58:18markunlinuxstb: a bit
13:58:54linuxstbIs IRAM faked in the gigabeat .lds files?
13:59:26preglowamiconn: not in the same way, no
13:59:27amiconnlinuxstb: The Ixxx_ATTR macros are supposed to evaluate to nothing when iram shouldn't be used
13:59:37preglowamiconn: the mpeg mdct is fixed size, which makes it easier to asm optimise
13:59:39amiconnSo no need to fake something...
13:59:44linuxstbamiconn: I know...
13:59:55preglowamiconn: wma does multiple sizes. still possible, but a bigger job
14:00:07linuxstbamiconn: But my understanding of the .lds is that it _is_ faked for the gigabeat.
14:00:21amiconnpreglow: Iirc saratoga said that wma also uses fixed size mdct, just not only one
14:00:22preglowwhy would you fake that?
14:00:52preglowoh? my understanding is that ffmpeg uses pre-proc -> fft -> post-proc for mdcts
14:00:58linuxstbI'm guessing a quick hack at the start of the port which was never fixed...
14:00:58preglowand the fft is most certainly just one routine
14:01:23 Join ender` [0] (
14:01:24amiconn5 sizes iirc
14:01:30preglowyeah, something around that
14:02:24preglowsaratoga mentioned he might look into using the tremor mdct
14:02:40preglowhmm, or was that just merbanan giving me some advice
14:03:31linuxstbYes, I think merbanan was asking why we didn't use it.
14:03:57preglowyeah, that was it
14:04:13linuxstbSee the logs - 20070404
14:04:19linuxstbI mean 20070704
14:20:33alarmclokgzip doesn't work
14:20:51 Quit alarmclok ("gzip killed Tina Turner, the only daughter of jesus")
14:23:15GodEater_I wish just for once I'd get my pointer math right first time. <sigh>
14:25:35linuxstbGodEater: To answer your question from about 4 hours ago - why do you want to restrict your viewer to one .link file?
14:26:52linuxstbIt's not a viewer then - it's a plugin with a config file...
14:28:01GodEater_linuxstb: I wanted to make it easy to launch the plugin, so it's supposed to run when you select you file in the root of the player
14:28:16GodEater_linuxstb: but I also wanted to avoid having the possibility of self referential .link files
14:28:59GodEater_so I just wanted the plugin to launch into it's selection mode when you select the file
14:29:05 Quit webguest85 ("CGI:IRC (Ping timeout)")
14:29:20GodEater_and any other call to the plugin with a file argument is interpreted as the user wanting to add the file to the shortcuts list
14:29:43 Join webguest85 [0] (i=c0231115@gateway/web/cgi-irc/
14:29:45GodEater_unfortunately the side effect of this is clicking on any other .link file which are user created, means they get added into the file
14:30:11linuxstbCan't you just check the name of the file, and react appropriately?
14:30:33GodEater_what is appropriate action for another .link file though ?
14:30:54GodEater_add it to the "main" one, or use it as another list of shortcuts ?
14:32:20linuxstbYes, I would say that if you open a .link file, you're viewing that list of shortcuts. If you open any other kind of file, you're adding it to the standard .link file.
14:32:21GodEater_Llorean seemed to think it was a good idea to have all the shortcuts in only one place
14:32:58GodEater_that would seem to go against what Llorean advised however
14:33:00linuxstbThere's no harm in allowing a user to create multiple .link files if they so wish.
14:33:33linuxstbI can't see what problems it could cause.
14:33:54GodEater_well it causes me all sort of issues - since I'll pretty much have to rewrite the plugin =.
14:34:16GodEater_it was designed around only one place to store the shortcuts
14:35:53*linuxstb reads the source...
14:36:35GodEater_I can't think how to make it create anything other than a default link file either, not and keep it simple (i.e. no text editor functions)
14:37:40linuxstbI'm not suggesting you let users add files to other .link files - just view them.
14:38:03linuxstbI agree that would complicate things a lot and not be worthwhile.
14:38:10GodEater_well that would be fairly trivial then I suppose
14:38:28linuxstbBut if a user has (manually, or by renaming created another .link file, I can't see why they wouldn't be allowed to view it.
14:39:27GodEater_if Llorean shouts at me because of it I'm blaming you :)
14:40:28linuxstbLooking at your code, did you try to simply call "opendir" on the parameter? If that succeeds, then the parameter is a directory.
14:40:50GodEater_no - ripped the "is it a dir?" code straight from the properties plugin
14:42:06GodEater_I'll change it - your way is both simpler and more obvious
14:42:10linuxstbDo you even still need to care?
14:42:20GodEater_about what ?
14:42:31linuxstbWhether the parameter is a file or a dir.
14:42:44linuxstbI thought you could now add both to the list?
14:42:44GodEater_yes - I need to know whether to write a trailing "/" or not
14:42:47 Quit webguest85 ("CGI:IRC (Ping timeout)")
14:42:59GodEater_if you don't the filetree ends up "on" the directory, rather than "in" it
14:43:13linuxstbMaybe the user wants that...
14:43:26*GodEater_ looks for a gun to shoot at linuxstb
14:43:36*linuxstb should probably stop talking as he'll never use this plugin anyway ;)
14:43:44GodEater_it wasn't the way I interpreted the feature request
14:43:52linuxstbDoing what the customer asks for?
14:44:07linuxstbThat never works...
14:44:51GodEater_so I hear...
14:45:18GodEater_now why the bloody hell doesn't my delete function work ? :(
14:47:37 Join Fra [0] (
14:52:56GodEater_bugger this for a game of soldiers, I'm going to the gym
14:54:00FraSorry but I need 2 patches
14:54:06FraI click on 'em
14:54:23Fraand it returns a page that starts with
14:54:23FraOnly in rockbox_svn/apps: bookmark.c.orig
14:54:23FraOnly in rockbox_svn/apps: debug_menu.c.orig
14:54:23DBUGEnqueued KICK Fra
14:54:23Fradiff -u -r rockbox_svn.orig/apps/gui/backdrop.c rockbox_svn/apps/gui/backdrop.c
14:54:23Fra−−- rockbox_svn.orig/apps/gui/backdrop.c2007-06-24 21:29:10.343750000 -0500
14:54:23***Alert Mode level 1
14:54:23Fra+++ rockbox_svn/apps/gui/backdrop.c2007-06-24 23:14:40.281250000 -0500
14:54:32peturthat's a patch
14:54:34B4gderthat's a patch
14:56:22JdGordonGodEater_: still around?
14:58:09CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
14:58:09*JdGordon slaps GodEater_ for being stupid and excerising!
14:58:55amiconnJdGordon: Btw, I found a bug in the settings screen y'day
14:59:11amiconnTry changing the 'Status Bar' setting and see...
14:59:19*JdGordon tries
14:59:50JdGordonhehe, its not all that bad :p
15:00:09*JdGordon changes that to using a temp var
15:03:15Frais it possible for a patch to end with "enum {" ????
15:04:11idnardiff can do weird things ;)
15:04:24***Alert Mode OFF
15:04:35Fraok I trust you;
15:05:30***Saving seen data "./dancer.seen"
15:05:57B4gderdiff just shows diffs and with -u or -c it includes context
15:06:03B4gderso of course the context can contain whatever
15:06:34Franow I have just downloaded a patch
15:06:41Frahow can I apply it?
15:06:57FraI've installed DiffUtils and Patch too
15:07:22linuxstbFra: What development environment have you set up for Rockbox? (cygwin, vmware, linux, ....)
15:07:28B4gderFra: then you didn't read through the wiki url from above
15:07:37Frayeah I read
15:07:46B4gderyou apply it with 'patch'
15:07:50B4gderthen you rebuild
15:07:58FraI set up no development environment
15:08:13B4gderthen you need to do that
15:08:21Frahow can I?
15:08:23B4gderand that's one of the first things mentioned on that page you said you read
15:08:23linuxstbFra: That's the first thing you should do - they should come with the necessary patching tools.
15:08:51linuxstbFra: And you should try and build a clean version of Rockbox first, and then try to build versions with patches applied.
15:11:29B4gderbut be very careful or you might learn something in the process! ;-P
15:12:16FraI'm trying
15:12:49JdGordonB4gder: is the revision under the dowload link the revision of the avilable zip? or the one being built?
15:12:58B4gderif the available zip
15:13:03B4gderwhat you actually get by clicking
15:13:26JdGordonso they are all done with 90s+ still to go?
15:13:54B4gderyou mean 90s left of the countdown?
15:14:05B4gderthe countdown is just a guess
15:14:53B4gderbased on a moving average for the latest builds
15:15:03B4gderthe build time still varies a lot
15:16:13amiconnSure, it depends on how much has to be rebuilt (due to ccache)
15:16:36 Quit Fra (" mIRC 6.21 in Italiano by TuttoIRC & ^DaNGeR^")
15:16:36B4gderyes, and quite a lot on various local circumstances for each server
15:16:40amiconnIt takes rather long when a central header file was changed (worst case probably being config.h)
15:16:47JdGordonok, it just surprosed me that all the builds said the enw version already, might have been because I was just too slow
15:16:47B4gderand which server that gets to send the biggest files back
15:17:29amiconnBut why does the build page still show the counter when all builds are actually done?
15:17:44B4gderhow would it know that they are done?
15:17:44amiconnI mean when entering the page, or refreshing manually
15:17:59B4gderno, it doesn't
15:18:17amiconnSo what did JdGordon experience, then?
15:18:41JdGordonI guess it doesnt refresh untill the couter is finished
15:19:01B4gderit doesn't refresh until the counter reaches ETA, no
15:19:09amiconnJdGordon: Yes, it doesn't refresh by itself then, but then you should at least see _one_ old revision .zip
15:19:12B4gderand then it does it every 30 seconds (since a while back, used to be every 10)
15:19:40JdGordonamiconn: not nescacerrily... it still has to do the sim and bl's
15:19:56amiconnB4gder: Btw, the non-forced refresh is working fine
15:20:06amiconnAh, hmm, forgot the sims...
15:20:27B4gderamiconn: yeps, better indeed
15:20:40JdGordondo we really need a second sudoku solver? (
15:21:38linuxstbJdGordon: No, close it as out of date...
15:21:55JdGordonok, done
15:25:23JdGordoncan a codec person look at its a libdumb update (do we still use libdumb?)
15:26:49amiconnIt was removed from svn as nobody made a codec based on it
15:29:55 Join sharpe [0] (
15:34:38JdGordonis lars van de klomp a real name?
15:35:09petursounds like it yes
15:35:18peturDutch probably
15:35:19*JdGordon wonders if he has to add the name to credits for a simple patch which is 2 years old and wasnt even diff'ed properly
15:35:48peturI have added somebody in the past for a one-line diff
15:36:07JdGordonnames in credits already anyway :)
15:37:17barrywardellamiconn: i hadn't even noticed the missing 0. thanks for fixing that
15:38:01amiconnHehe, noticed it when adding another register to the ports debug. Wondered: "Huh? We want source #2 not source #3?"
15:38:06JdGordonhmm... /me remembers to put a space between the and ?
15:38:26*JdGordon 's auto-replace script needs some work it seems :p
15:39:32amiconnJdGordon: In the commit msg it's sufficient to just write 'FS#nnnn' or 'FS #nnnn'
15:39:44*JdGordon knows
15:39:47JdGordonI mean in irc...
15:40:02amiconnAh, that
15:40:09linuxstbJdGordon: I would say #2458 is out of date...
15:40:26amiconnWell, my client found the correct end of the link even without the space
15:40:39amiconn(as a trailing '?' isn't allowed in an url)
15:41:01JdGordonI tihnk we need to decide on a cutoff date for old patches where unless it still actually patches correctly it should be closed
15:41:33JdGordonseems Konversation is smart also and ignores the ?
15:41:35linuxstbJdGordon: I don't think we can make a strict rule - it depends how much that particular part of Rockbox has changed, or if similar features have been implemented since that patch.
15:42:00linuxstbe.g. a patch to a plugin could be valid for a long time...
15:43:16JdGordonspeaking of which... I forgot to check the builds page to see multiple revisions in the download table :p
15:44:34B4gderso commit more!
15:45:11petur939 and counting...
15:46:28*B4gder dreams of 800
15:46:37JdGordonnot a bad effort... what did it start at? 1000 something?
15:47:32B4gderjul 28 we had 199 bugs and 281 patches
15:47:41B4gderaccording to the automated mail
15:48:19JdGordon nice comment :p
15:48:33JdGordonby nice, i mean poiontless
15:48:43 Join Siltaar_ [0] (
15:49:17*bluebrother dreams of 500 open tasks left
15:49:34peturwhy not 0 ;)
15:49:38JdGordoneasy... blow away the flysrpay db and start again
15:50:18bluebrotherwell, might be easy. But going the crash route isn't a challenge.
15:50:58bluebrotherand we're doing this for the challenge, don't we? ;-)
15:51:00 Quit Siltaar (Read error: 110 (Connection timed out))
15:51:50JdGordona font person is needed for
15:52:22*bluebrother found another completely pointless comment
15:53:04JdGordonlostlogic: close 2958?
15:53:08barrywardellis this still valid: ?
15:53:19*amiconn found 'diskmode' and 'hotstuff' in the 2nd gen rom :)
15:53:23JdGordonI havnt seen that problem in ages
15:53:33*JdGordon wonders what 'hotstuff' is?
15:53:39JdGordonapple sanctionned pr0n?
15:53:45amiconnIt's a marker
15:54:04amiconnBoth markers must be set and then a device reset issued to make the flash loader go into diskmode
15:54:06bluebrotherstupid comments like "please fix this" in FS #6590 are quite tempting to reject a task ...
15:54:13bluebrotherbut that would be a wrong move.
15:55:41JdGordonsame user...
15:55:51JdGordoncan we send him a warning in FS?
15:57:05JdGordonrbutil can install the h300 bootloader now right?
15:57:20linuxstbamiconn: Isn't that the same as later ipods?
15:57:31amiconnIt's not exactly the same
15:57:51linuxstbAh, so the current code wouldn't work?
15:58:06 Quit B4gder ("It is time to say MOOO")
15:58:09amiconnLater ipods put diskmode\0\0hotstuff\0\0\1 at (IRAM_END - 256)
15:58:42amiconnFor 1st/2nd gen it looks like we need to put 'diskmodehotstuff' there
15:59:07amiconndiskmodehotstuff\1 even
15:59:43 Join pondlife [0] (
16:00:00linuxstbOK. Is the structure of the ROM the same as later ipods? i.e. can the diskmode and diagmode images be extracted via the flashsplit.c program here?
16:00:56amiconnUmm, didn't try
16:10:12 Join Febs_ [0] (
16:10:19JdGordonmidgey: ping?
16:13:02GodEater_JdGordon: back now - but abandoned my attempt to make sense of the menu api - so you're off the hook ;)
16:13:13 Join webguest79 [0] (i=c023110f@gateway/web/cgi-irc/
16:13:13JdGordonoh goody :)
16:13:25bluebrotherhmm. What do you guys think about moving the device selection in rbutil to its own tab?
16:13:25JdGordonit shouldnt be all that hard...
16:13:44pondlifeJdGordon: Might be good for you to ask GodEater what info he lacked...
16:13:51pondlifeA use case, if you will ;p
16:13:58bluebrotherthen the device path selection could go to the same so it wouldn't needed to have this in every installation window.
16:14:17GodEater_JdGordon: however, my question was around the statement on the wiki page you wrote where you say "callback is the function to call before displaying the menu blah waffle (explained later)"
16:14:18JdGordonGodEater_: pondlife rekons I shouidl ask you what info is lacking so it can be added
16:14:22GodEater_and then it wasn't
16:14:38JdGordonhehe, umm... hmm :p
16:14:52GodEater_at least - if it was, it was so subtle I missed it
16:15:02GodEater_your example passes NULL for it too - so that didn't help
16:15:12pondlifeGodEater_: Subtle stuff indeed
16:15:22JdGordonif I had net access in my lectures tomorow I'd fix it then, too tired now
16:15:24GodEater_pondlife: you agree it's missing then ?
16:15:26JdGordonremind me tomormow
16:15:30GodEater_JdGordon: no rush
16:15:40pondlifeGodEater_: No idea, never tried using it.
16:15:45GodEater_pondlife: wimp
16:16:07pondlifearm wrestle?
16:16:09*JdGordon having fun with the "rejected" button :)
16:16:12GodEater_bring it!
16:16:38pondlifeyou win
16:16:47*GodEater_ 's legs are the only aching bit after the gym today, so he feels up to the challenge
16:16:56*pondlife has never been to a gym
16:17:07GodEater_you must have won last time - otherwise how do you have a mug and t-shirt ? ;)
16:17:25pondlife(or they'll wonder where the other 20 went)
16:17:47*GodEater_ lost to a lot of people in the pub that night it seems
16:17:57GodEater_complete strangers most of them too
16:18:08pondlifeBut now they are walking adverts
16:18:08 Quit webguest79 (Client Quit)
16:18:14pondlifeWith no URLs on :(
16:18:14GodEater_this is true
16:18:26GodEater_assuming they're wearing them and not cleaning the car with them
16:18:40pondlifeThey wouldn't dare...,
16:18:56GodEater_well they took the liberty of defeating me
16:18:58lostlogicJdGordon: looking
16:18:58bluebrotherJdGordon: SDHC cards use a different addressing, so the current code won't work with them
16:19:00GodEater_I reckon they would dare
16:19:20bluebrotherunless someone has added detection of SDHC (but I haven't noticed something like that)
16:19:38JdGordonbluebrother: right, but no devs own a sdhc, and it being in the top votedd tasks annoyed me
16:20:02pondlifeHmm, looks like NYC will be the first DevConBar? Or will it be a DevConDeli?
16:20:08 Quit tumu (Read error: 104 (Connection reset by peer))
16:20:13 Quit BigMac ("Leaving")
16:20:14JdGordonpondlife: a quote from a comment on FS you added... "I'll commit this one soon unless anyone objects."
16:20:26pondlifeJdGordon: Did I not commit it?
16:20:29JdGordon... dated...Monday, 13 November 2006, 17:47 GMT+10
16:20:36bluebrothermaybe this votes stuff should be hidden in FS?
16:21:00GodEater_pondlife: DevConStarbucks I thought
16:21:01amiconndiskmodehotstuff\1 works :)
16:21:12pondlifeJdGordon: Ah, another of my orphans
16:21:18pondlifeWill resurrect
16:21:35amiconnlinuxstb: Do you think we should have separate firewire handling (even though it will currently just do the same as usb)?
16:21:46amiconnOr should I just hack it into the usb handling?
16:21:58pondlifeHack! Hack! Hack!
16:23:11linuxstbHopefully we'll need to split them quite soon - when our own usb mode code starts to get committed...
16:24:18amiconnYeah, but 1st/2nd gen has no usb
16:24:25bluebrotherdoes anybody know the state of that btw?
16:24:31 Join Jens [0] (i=Jens@pdpc/supporter/active/Jens)
16:24:43amiconnAnd 3rd gen is probably different, as our current detection doesn't seem to work according to DerPapst
16:24:45linuxstbBut I'm not really sure at what point it needs to be split - e.g. at the apps level , we don't want to handle separate SYS_USB_CONNECTED and SYS_FIREWIRE_CONNECTED messages...
16:25:10amiconnlinuxstb: hmm
16:25:14linuxstbYes, I'm sure I remember dan_a saying it wasn't implemented.
16:25:25amiconnThat means we should perhaps rename it
16:25:53amiconnIt's definitely compiled in - see my commect at the end of pp5002.h about the hack
16:26:24amiconnFirewire detection is dead simple - just checking a port pin
16:26:31JdGordonwhat do we do with a 6month old translation update?
16:26:48bluebrotheris the lang file in svn even older?
16:26:51amiconnBut I would like to see these checked on other ipods than those I have
16:27:04*barrywardell directs pondlife to as well
16:27:11 Quit Febs (Read error: 110 (Connection timed out))
16:27:35amiconnWhat I know so far: Mini G1: C1 = firewire detect, C2 = firewire power
16:27:36JdGordonbluebrother: no idea...
16:27:40amiconnMini G2: same
16:27:49pondlifebarrywardell: Do you reckon I got the pointer maths correct there?
16:27:53amiconn4th gen (bw/color): same
16:28:08amiconn5th gen: L3 = firewire power (only)
16:28:09*pondlife wants to trigger a queue overflow...
16:28:29amiconn2nd gen: C7 = firewire detect, C4 = firewire power
16:28:47amiconnThat leaves 1st gen, 3rd gen, and nano
16:28:56barrywardellpondlife: i looks right to me. don't know what removing the ifdef is for though
16:29:28amiconnI _guess_ that nano is same as video (or does the nano support firewire?)
16:29:37pondlifeBecause I like removing #ifdefs, and cpu_boost() is #defined out of existence anyway
16:32:56barrywardellpondlife: looking at playlist.c now, I don't think the patch is necessary any more. is that right?
16:33:52 Quit JdGordon ("Konversation terminated!")
16:33:59 Nick Febs_ is now known as Febs (
16:34:00pondlifebarrywardell: Correct
16:34:17barrywardellok, i'll close the task
16:35:44 Join toffe82 [0] (
16:37:54 Join DerPapst [0] (
16:38:02pondlifeHmm it's a bit easy to overflow a queue.
16:38:05DerPapstamiconn: nano has no firewire
16:38:14pondlifeMainly holding down a key when nothing is looking
16:38:28amiconnhi DerPapst
16:38:35DerPapsthi amiconn ;)
16:38:39amiconnDo you have a firewire cable for your 3rd gen?
16:38:47 Quit Entasis (Read error: 104 (Connection reset by peer))
16:38:55DerPapst(how could i recharge it otherwise?)
16:39:02amiconnAnd does your PC have a firewire port as well?
16:39:08DerPapstyes again
16:39:20amiconnCould you test something for me?
16:39:27DerPapstbut i'm at work still...
16:39:37*pondlife wonders why thread->name is always blank.
16:39:40amiconnah, so later perhaps?
16:39:43DerPapsti can in ~2h30'
16:39:48amiconnnice :)
16:40:00DerPapstyou can prepare everythin already :P
16:40:19amiconnThe first test just requires svn
16:40:44amiconnJust 'View IO Ports', your fw charger, cable, and the PC
16:41:22barrywardellis there anyone that can test I can't test it.
16:41:26amiconnHopefully I will have working firewire detection (with reboot into diskmode) in rockbox tonight
16:42:37*amiconn wonders when dan_a will be around again :\
16:42:52pondlifeWe scared him off with our DevConPubs
16:44:06pondlifeSeriously, am I wrong to expect a string from q->thread->name ?
16:44:19amiconnThere should be one
16:45:56pondlifeI can't see anything that sets q->thread (aside from to NULL).
16:46:58pondlifeAh, wakeup_thread
16:48:23amiconnMaybe it's only set when some additional debug stuff is enabled?
16:48:34*amiconn remembers some #ifdef stuff in there
16:50:45 Quit sharpe ("Leaving")
16:51:32 Quit ender` (Read error: 104 (Connection reset by peer))
16:55:50amiconnpixelma: ping...
16:58:58 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
16:59:59linuxstb_amiconn: I've just updated the flashsplit.c utility on the IpodFlash wiki page to support the 1g/2g. It extracts the application images to ".ipod" files suitable for use with RoLo (if it worked...). Your image contains three - diskmode, diagmode and diskscan. (I haven't seen diskscan before...)
17:00:56DerPapsti remeber hearing of diskscan
17:01:09DerPapstsomewhere on the iPL wiki i guess
17:02:22DerPapsthere we go...
17:03:35DerPapstit even says that one can extract the bootloader...
17:03:54linuxstb_Well, the bootloader is everything else...
17:04:49linuxstb_There is a simple directory at the end of flash listing the available images - with offsets, lengths and checksums, similar to the content of the firmware partition directory on disk.
17:05:30linuxstb_So it seems the main bootloader does what it does, and then either loads the firmware from disk, or loads one of the "apps" in flash.
17:05:32***Saving seen data "./dancer.seen"
17:05:37linuxstb_(depending on detected keypresses)
17:05:39DerPapstcan your tool extract the bootloader too?
17:06:08 Nick hannesd_ is now known as hannesd (
17:06:19linuxstb_No, but that's just from the start of the file onwards. All it would do is truncate the file.
17:06:38DerPapstbecause certain persons might be interested in findong out what it writs to ata reated... thingys
17:06:40 Join dandin1 [0] (
17:06:58DerPapstah ok.
17:07:37DerPapstfindong <−− wth ;)
17:07:42linuxstb_But the whole thing is really the bootloader - you could think of those apps as being inside the bootloader.
17:07:47 Join spiorf [0] (
17:07:54DerPapstahh ok.
17:08:12DerPapstwell then nvm ;)
17:08:23*DerPapst shuts up again
17:11:01 Join spiorf_ [0] (
17:14:54bluebrothern1s: is FS #6691 still valid=
17:17:58 Quit spiorf_ ("Read error: 110 (Connection timed out)")
17:18:17 Join spiorf_ [0] (
17:19:53 Part maffe
17:22:55barrywardelllinuxstb_: would it be hard to add an 'update OF version' option to sansapatcher?
17:24:47amiconnlinuxstb: Ah, nice :)
17:25:01 Quit desowin ("use linux")
17:25:04amiconnI wondered why some routines seem to be present in the flash multiple times
17:25:18amiconnI guess there's one for each application (which needs it)
17:25:56*GodEater_ thinks that barring the feature of using mulitple .link files, which he's still not sure about, his plugin is done.
17:27:03linuxstb_GodEater_: As I said before, ignore me - I'm not a potential user of your plugin...
17:27:17linuxstb_barrywardell: I've forgotton how sansapatcher works...
17:27:29pixelmabluebrother: n1s said it still isn't solved (today's log at 11:27)
17:28:17barrywardelllinuxstb_: ok, i'll just have to have a look myself
17:28:33 Join freqmod [0] (
17:29:04*amiconn thinks one key to battery savings on PP is figuring out the unknwon bits of the DEV_EN register
17:29:09linuxstb_barrywardell: Why would you want to do that anyway? Isn't it easier to just update the OF in whatever way Sansa suggest, and then just re-run sansapatcher to install the RB bootloader again?
17:29:10amiconnBoth on PP5002 and PP502x
17:29:26jhMikeSamiconn: maybe is of interest and could explain the behavior differences in core patches when constantly messing with cpsr
17:29:37barrywardelllinuxstb_: yeah, but that takes longer and I'm impatient
17:29:44 Quit spiorf (Read error: 110 (Connection timed out))
17:30:07linuxstb_barrywardell: Do you often update the OF?
17:30:33GodEater_having never written so much code for rockbox before, what is the criteria for acceptance of my work? Is there testing it must undergo? Or just a general code review ?
17:31:01barrywardelllinuxstb_: I'm looking at FS7207. If I could easily update the OF, I could implement it for all firmware versions quickly
17:32:10linuxstb_GodEater_: There's nothing formal... Basically it needs to be generally accepted (i.e. no-one shouts at you in IRC), and a dev needs to take an interest.
17:32:38pondlifebluebrother: Yes, n1s tested #6691 earlier today and it's still a problem.
17:32:39*GodEater_ braces for being shouted at
17:34:03pondlifebluebrother: I've put a note on the bug.
17:35:49pondlifeGoing for the really old ones, would deserve an Out Of Date....
17:36:12pondlifeOver 3 years between notes - good going!
17:36:48*GodEater_ dons a placard to drum up interest too
17:37:33amiconnjhMikeS: Problem #1 doesn't apply to us afaik. We don't have such dual-use functions
17:38:10amiconnAnd messing with the fiq shouldn't stop rockbox from booting without freeze, as there are no fiqs fired in that stage
17:38:52amiconnqueue_post does checked whether it was called in interrupt context?
17:38:54jhMikeSwe also don't seem to allow fiq during irq
17:39:13jhMikeSyes, because it saves and restores cpsr
17:39:54jhMikeScpsr_irq has the F bit set in crt0(-pp).S
17:40:20amiconnI'm not sure whether we need to allow this
17:40:21pondlifeWhat's holding up RockCalendar ( - there's been hours of work put into it and it looks good.
17:40:43amiconnpondlife: Last time I checked it wasn't really usable on small screens
17:40:57pondlifeCan a plugin not be used on limited targets?
17:41:06pondlifei.e. included only on some targets.
17:41:06linuxstb_It's also full of nasty custom menus... (or at least, used to be)
17:41:15jhMikeSI'm allowing it now. I think I want to use interrupts for task switching.
17:41:30amiconnI wouldn't want to have two plugins for the same purpose, split by target
17:41:43amiconnjhMikeS: hmm?
17:42:07jhMikeSit would simplify the scheduler and be much better for dual core setup
17:42:09amiconnWe don't want preemptive multithreading
17:42:18jhMikeSthat's another subject
17:42:29jhMikeSthough I don't understand the object to it.
17:42:52amiconnIt requires _a lot_ more precautions than cooperative scheduling
17:43:25amiconnCurrently a thread can rely on not being interrupted at an arbitrary position
17:43:27jhMikeSa few but this dual core will basically intro
17:43:43amiconnYes, but there are other targets than just PP
17:43:57jhMikeSbut it can't. it really simplifies nothing at all unless you can count on the code you call never yielding.
17:44:10amiconnyes, exactly
17:44:24jhMikeSrockbox is too complex to count on that
17:44:52amiconnWith premptive threading, you can't even rely on not eing interrupted when not calling any subroutine at all...
17:45:16jhMikeSmost code needn't deal with it anyway
17:46:04*amiconn wonders how interrupts can be used for thread switching while staying cooperative
17:46:05 Join Febs_ [0] (
17:46:07jhMikeSmost of the code that runs on multiple threads now is full of holes.
17:47:37jhMikeSexactly the same way it would if it were preemptive but instead of a timer doing it, the call to a task switching function causes the interrupt
17:48:04amiconnAnd why is that better than just letting yield() do the work?
17:48:13amiconnSounds like more overhead for nothing to me...
17:48:18 Join Xerion [0] (
17:49:18jhMikeSit would except the interrupt interaction with the scheduler would be serialized. this would simplify dual core interaction especially since wakes invoked by one core would be handled on the waking threads core, during interrupt context.
17:50:08jhMikeSI can't see where threre's really any overhead at all since the return from the handler would load the new context...not much different than now.
17:50:41amiconnYou cause extra interrupts, which involves additonal stack accesses
17:50:51 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
17:50:54 Quit Thundercloud (Remote closed the connection)
17:50:54amiconnAnd I wonder how this should work on other targets than arm
17:51:02 Join Thundercloud [0] (
17:51:11 Quit davina (Remote closed the connection)
17:51:21*amiconn doesn't really understand what interaction is there, and why it needs to be serialized
17:51:42jhMikeSI'm wondering a bit about that. Does the CPU on say coldfire store the context or does the service routine?
17:51:50amiconnThere were times when the scheduler was simple, so that an average rockbox coder could understand it, and it worked reliably...
17:52:22jhMikeSI think it need to catch up to more advanced devices.
17:53:08 Join davina [0] (
17:53:16jhMikeSthings can't run completely isolated in a dual core environment. there will be points were scheduling must be synchronized between cores.
17:53:20*amiconn doesn't see the need for that
17:53:36amiconnYes, dual core might be a problem
17:53:56amiconnColdfire saves the status register and the pc on the stack on interrupt
17:53:57jhMikeSthen dual core support will be hard
17:54:11 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
17:54:44jhMikeSwhat about the data and address registers? the routine saves those?
17:55:31 Quit Febs (Read error: 110 (Connection timed out))
17:56:02jhMikeSit's not even safe to just start a new thread on the COP from the CPU since it messes with data the COP is changing.
17:56:41amiconnYes, on cf the routine has to save/restore additional registers
17:58:32amiconnI still don't understand the need for sw ints...
17:58:34jhMikeSoverhead is really minor then since the prologue stores the context from the current task, and the epilogue loads for the next.
17:59:05amiconnISRs have to save *all* registers, including scratch
17:59:19amiconn(d0/d1/a0/a1 on cf)
17:59:32jhMikeSyes, not too much more
18:00:09jhMikeSfrankly I just loathe the cooperative stuff anyway
18:00:13amiconnThat is if it uses those of course
18:00:31 Join My_Sic [0] (
18:01:27*amiconn probably has to stop doing more complex stuff for rockbox if we'd go preemptive
18:02:34jhMikeSthe biggest concern is device access really. plugins, UI code and such probably wouldn't be much different if at all.
18:03:40amiconnAnd I quite like being able to do low level stuff without locking each and every tiny step
18:03:53amiconnI wouldn't be able to follow it, I think :\
18:05:06jhMikeSthere's a crutch aspect to cooperative that when something that didn't yield suddenly starts to, the code falls apart. at least it would have to be secure from the start.
18:05:25amiconnYes, with way more overhead than now
18:06:11jhMikeSthat's the thing, a mutex should only force a task switch on block otherwise it's negligable really.
18:06:39amiconnYou have to watch for obscure deadlock situations etc...
18:06:48amiconnThe equivalent of arm's swi instruction on coldfire would be trap, on SH it would be trapa
18:06:52jhMikeSblah, those can still happen
18:07:02amiconnA trapa on SH is quite costly - 15 cycles
18:08:23amiconnEh, wrong, just 6
18:08:25 Quit petur ("work->home")
18:08:46jhMikeSthe same deadlocks are possible in a cooperative system. it really behaves exacly the same way.
18:08:48amiconnIt also saves sr and pc, like on coldfire
18:08:49 Quit ze (Read error: 110 (Connection timed out))
18:09:15amiconnIn a cooperative system, mutexes don't need to be atomic
18:09:35jhMikeSactually, they are already for CF and ARM anyway
18:09:49 Join ze [0] (
18:10:24jhMikeSdoes SH have a similar test and set instruction (I have read the ds too carfully)
18:10:54amiconnIt has tas
18:11:34jhMikeSon cf I used bset.l and sne.b and bset.l sets the status depending on the dest value
18:11:37amiconntas reads a byte from memory, and sets the t flag if that byte was zero. It also sets the MSB of the memory byte in the same operation
18:12:01amiconn(tas == test-and-set)
18:12:06jhMikeSjust like bset.l
18:12:27amiconnIt can only set one bit (the msb)
18:13:08jhMikeSas long as mutex->locked is set to a nonzero value atomically
18:14:13amiconnIt doesn't read the value back into a register directly but into the t flag, but that's no problem I think
18:14:18jhMikeSthe sim runs preemptive threads, doesn't it?
18:14:41amiconnIt simulates rockbox threading by interlocking the threads
18:16:09jhMikeSthere's code specifically to interlock them?
18:16:26pondlifeI'm not conviced it simulates rockbox threading that well...
18:16:43 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
18:16:51pondlifePlenty of playback race conditions don't happen in the sim.
18:18:03amiconnCheck thread-sdl.c
18:18:17amiconnIt does simple round robin scheduling by this interlocking
18:18:27amiconnYou see: simpler scheduler->less problems
18:18:37*jhMikeS is trying to find where is locking code is
18:18:52amiconnAnd preemptive scheduling will make things far more complex...
18:19:12amiconnSDL_mutex *m;
18:21:00 Join jgarvey [0] (
18:21:16jhMikeSwell, that's a mutex. I guess I'm familiar with what SDL is up to.
18:21:51 Quit ze (Remote closed the connection)
18:21:55 Join ze [0] (
18:21:55amiconnAfaiu this mutex makes sure that only one of the thread can run at all times
18:23:03jhMikeSI think on whole, it would make things easier.
18:24:48 Quit pondlife ("disconnected has pondlife")
18:24:58amiconnMe doesn't believe this
18:25:05*amiconn even
18:25:26jhMikeSyes, I see it emulates serialized threads with preemptive ones.
18:26:44*jhMikeS always enjoyed preemptive multithreading programming
18:27:15amiconnIt'd cause me headaches... that's why I stayed away from it so far
18:27:44jhMikeSit's more like an ice cream passes
18:28:47 Join ze_ [0] (
18:29:11 Quit ze (Read error: 104 (Connection reset by peer))
18:29:16 Nick ze_ is now known as ze (
18:29:27jhMikeSdealing with interrupts in the pcm and elsewhere is already like it not to mention seeing so many things open already. More mutexing is needed currently.
18:31:35jhMikeSa good deal of the stuff I did for recording allows multithread access to things that yield and just allowing that really makes things simpler than they would be if I went through pains to avoid it.
18:34:47 Quit obo_ ("bye")
18:35:21 Quit HellDragon (Read error: 104 (Connection reset by peer))
18:35:22amiconnIs it certain that the 2nd argument of test_and_set is always 1 ?
18:35:36 Join HellDragon [0] (i=Nocebo@unaffiliated/helldragon)
18:37:55 Join petur [0] (n=petur@rockbox/developer/petur)
18:38:13jhMikeSit should be
18:39:10amiconncoldfire test_and_set seems to rely on that as well
18:39:52jhMikeSat least the arg has to be 8 bits
18:40:07 Join lostnihilist [0] (
18:40:17amiconnHmm, it doesn't...
18:41:57jhMikeSFor general code, it should use 1. For something CPU-specific it could use whatever is available.
18:42:54 Quit DerPapst ("work->home")
18:43:23 Join Arathis [0] (
18:45:01 Quit petur ("real life (tm) - back tonight late")
18:52:18jhMikeSon dual core, locking both schedulers because of a task switch on one core won't allow the maximum potential. just the short situation of waking a thread on the other core or creating/terminating a thread on it needs syncing but mixing this up makes things more complicated that just using interrupt context for the switching. whether every other CPU must follow too is another story.
18:52:38 Quit Febs_ ("ChatZilla [Firefox]")
18:56:09 Join bluey- [0] (
19:00:44*amiconn does an experiment
19:02:08 Join bluey^ [0] (
19:03:08jhMikeSbasically, threads should be woken/started/removed by the core they belong to
19:05:35***Saving seen data "./dancer.seen"
19:11:16 Join webguest46 [0] (i=485af9fd@gateway/web/cgi-irc/
19:11:40webguest46does rockbox work on 2nd gen ipod nano
19:12:00 Quit XavierGr (Nick collision from services.)
19:12:01*jhMikeS at least wants to commit to allow FIQ during IRQ
19:12:04 Join XavierGr [0] (
19:12:07bluebrothersee the frontpage.
19:12:21webguest46what if i put the first gen on my 2nd gen
19:12:29bluebrotherit won't work
19:12:32webguest46will that mess it work
19:12:58bluebrotherI don't know how the Ipod will react to this. But it _won't_ work
19:13:12webguest46are you guys planning on making rockbox for 2nd gen
19:13:12bluebrotherbesides ipodpatcher won't install the bootloader
19:13:23bluebrotherno. Ports are not planned.
19:13:33bluebrotherports are done by people starting to work on a port
19:13:52 Join obo [0] (n=obo@rockbox/developer/obo)
19:14:02Doomed_the 2nd gen has encryption or something, theres a thread in the forum
19:14:28bluebrotherthere are basically 2 issues with the 2nd gen nano:
19:14:37bluebrother(1) it has a _completely_ different hardware
19:14:51amiconnToo bad that the return value of an asm block can't just be a flag -> 2 insns wasted
19:15:01bluebrother(2) the system is completely encrypted, thus it is not possible running our bootloader the same way we do as on other Ipods.
19:16:26*GodEater thanks heaven that linuxstb was foresighted enough to prevent users breaking their 2nd gen nanos with ipodpatcher
19:17:04jhMikeSamiconn: the sne.b on cf is a wasted instruction. the C code could use the status from bset.l but doesn't. :\
19:17:10bluebrotherwell, I'm pretty sure Itunes restore would be able fixing that
19:17:16amiconnjhMikeS: Yes, same on SH
19:17:23bluebrother(at the expense of loosing all music on the player of course ;-)
19:17:32GodEaterbluebrother: indeed
19:17:49amiconnI have to move the T flag to a register (with movt), then the subsequent code tests that register, setting the T flag...
19:18:08 Quit bluey- (Read error: 110 (Connection timed out))
19:18:25amiconnThe atomic mutex doesn't have an influence on bin size
19:18:27jhMikeSlol...figures they wouldn't make the assembler more analyitical.
19:18:40jhMikeScompiler too
19:18:51*amiconn will test whether it actually works
19:19:21jhMikeSwhat did it generate from the C?
19:19:33 Quit webguest46 ("CGI:IRC (EOF)")
19:19:46amiconnProbably something similar, just not atomic
19:22:34 Quit bluey^ ("Leaving")
19:23:47jhMikeSamiconn: the addresses of CPU/COP_INT_CLR are definitely correct for PP5002? I just want to make sure all sources are masked before clearing the F bit in crt0-pp.S when setting up irq mode.
19:23:56 Join donutman25 [0] (n=chatzill@
19:24:20donutman25hi everyone
19:24:31amiconnjhMikeS: Yes, they should. Why do you want to do that?
19:24:35 Quit XavierGr (Nick collision from services.)
19:24:38 Join XavierGr [0] (
19:24:51donutman25how do i use the dd command to retrieve data from the hidden partion on sansa?
19:25:45GodEaterdonutman25: what OS you on ?
19:26:09GodEaterdo you the device name of your sansa ?
19:26:28 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
19:27:12jhMikeSamiconn: to allow FIFO refills to be prompt always
19:27:21*GodEater tries the question again, this time in english
19:27:32GodEaterdonutman25: do you know the device name of your sansa ?
19:27:38donutman25what is the device name?
19:27:43 Join Cripps [0] (
19:27:44linuxstb_donutman25: I wouldn't call it hidden - it's very visible. Try running "fdisk -l" to see a list of all the disks attached to your computer, and their partitions.
19:27:45amiconnjhMikeS: Yes, but do you expect any interrupts to be pending at this point?
19:27:47GodEaterthat'll be a know then
19:27:53GodEateror a no
19:28:00GodEaterlord - brain has short circuited
19:28:06amiconnHehe, I made a mistake in my test_and_set for SH
19:28:18amiconnThe return value is the wrong way round...
19:29:39jhMikeSamiconn: paranoid ... we're left with whatever the bl left.
19:30:20GodEaterlinuxstb: I think my plugin is exposing a bug in dircache
19:30:21linuxstb_donutman25: Are you still with us?
19:30:28linuxstb_GodEater: Impossible!
19:30:39donutman25yes i am
19:30:51donutman25where can i find the device name?
19:30:57 Part Jens
19:31:04GodEaterlinuxstb: if you choose a shortcut before the cache has finished updating, the plugin tells you it's not on the disk anymore
19:31:05linuxstb_donutman25: Did you read my reply to you?
19:31:09 Join BigMac [0] (
19:31:10GodEaterwhich is a blatant lie!
19:31:18 Join Arathis_ [0] (
19:31:21donutman25yes i did
19:31:32linuxstb_Did you run "fdisk -l" ?
19:31:36 Quit Arathis (Nick collision from services.)
19:31:38CrippsI would just like to say thank you to everyone that is working hard on Rockbox. I just bought a Cowan iAudio X5 recently, and popped rockbox on 'er ... I love it. I'm hoping to also do some contribution in my spare time ;)
19:31:40 Nick Arathis_ is now known as Arathis (
19:31:53linuxstb_donutman25: Can you paste the output to ?
19:32:02donutman25ok hold please
19:32:11jhMikeSamiconn: system_init in the bl doesn't change whatever OF setup so imho safe side is best.
19:32:59amiconnHmm, but you need to restore it afterwards, otherwise the OF might not like it
19:33:27amiconnWhy is system_init in the bl important, btw?
19:33:36amiconnShouldn't we just do it for main rockbox?
19:33:45GodEaterdonutman25: do you even have your sansa plugged in ?
19:33:45jhMikeSI am doing it for main only
19:34:11linuxstb_donutman25: OK, that says that the disk in your Sansa is /dev/sde and /dev/sde2 is the second partition (the "hidden" one).
19:34:48GodEaterlinuxstb: where's his system drive though ?!?
19:35:19linuxstb_So the command you need is "dd if=/dev/sde2 of=mysansa.bin"
19:35:36donutman25ok thank you so much
19:35:39GodEaterdonutman25: what are you planning on doing with this file ?
19:35:40linuxstb_GodEater: Maybe he doesn't have permissions to read them, or he just pasted the relevant part...
19:35:40jhMikeSwe can also clean the fiq vector out of there since it's set in the core pcm anyway...just reserve the vector space
19:36:22donutman25GodEater: I just wanna poke around in it to see if i can learn something
19:36:34GodEaterdonutman25: good luck then!
19:37:14GodEaterlinuxstb: what did you think of my bug? Plausible? Or is my own code to blame do you think ?
19:37:36linuxstb_GodEater: I don't know - dircache is one of those many black holes in Rockbox that I don't understand.
19:38:00GodEaterit's Slasheri's baby is it not ?
19:38:08linuxstb_It is.
19:38:15*GodEater prods at Slasheri
19:39:10GodEaterdonutman25: btw - have you read this page ?
19:40:28*jhMikeS tries to write white holes but probably only manages Hawking Radiation
19:45:56SlasheriGodEater: currently dircache can't handle live updating the cache while it's still building, however in that case all file operations affecting the cache should just block
19:45:57Crippswow ... power management for the iAudio X5 is sparse ...
19:46:51linuxstb_Cripps: What do you mean?
19:47:52 Quit barrywardell ("Konversation terminated!")
19:47:58 Quit BigMac ("Leaving")
19:48:17 Join BigMac [0] (
19:48:22Crippslinuxstb_: there are a few functions in power-x5.c that are bool function(yadda){return false}
19:48:35Cripps*more or less
19:48:38 Quit Arathis (Read error: 104 (Connection reset by peer))
19:49:58linuxstb_Cripps: Most of those are for the simulator (the version of Rockbox that compiles and runs on your PC).
19:50:50 Quit pearldiver (Read error: 110 (Connection timed out))
19:50:58Crippslinuxstb_: ahh, understood ... although, that doesn't explain : staticbool powered = false; ;)
19:51:11Crippsoops: static bool even
19:51:43linuxstb_That's set by the tuner_power() function.
19:52:24linuxstb_But yes, it doesn't appear to do much to the hardware...
19:52:27Crippslinuxstb_: ah. I see; it's just being set to false as an initiation value. Wow. My C is *rusty*
19:53:09Crippslinuxstb_: if I were to have a desire to improve power management on my model, where would you suggest I start reading?
19:54:06linuxstb_I don't know - neither the iaudio hardware or Rockbox power management are things I know much about. Hopefully someone who does know is reading this...
19:54:22Crippsthat would just be too perfect ;)
19:54:45jhMikeSthere's no HAVE_TUNER_PWR_CTL for the radio. I don't know why it needs the function really.
19:57:33amiconnjhMikeS: It's needed for the archos FM Recorder and (old) Ondio FM
19:57:46jhMikeSyes, and e200
19:58:27jhMikeSfor everything else, no reason for even a stub
20:00:42GodEaterSlasheri: would that include a simple open() ?
20:02:16jhMikeSwhat about an issue with dircache and rolo firmware change detection now that dircache is transparent? it used to avoid using it.
20:03:50linuxstb_The non-cached function should still be accessible - I don't know if it uses it though...
20:04:47jhMikeSnot last I checked. I don't think it got updated.
20:07:19 Join DerPapst [0] (
20:07:37*DerPapst trys out his maybe new keyboard...
20:07:56DerPapstamiconn: i can do test now if you want
20:08:05amiconnjhMikeS: Dircache has to rescan after usb anyways
20:08:44amiconnDerPapst: Nice :)
20:09:00amiconnFirst take your 3rd gen, start rockbox, and go into 'View I/O Ports'
20:09:26amiconnCheck the value of GPIO_C
20:09:46amiconnThen plug it into your charger, and watch the value
20:11:01DerPapstfrom B6 -> A6 on charger...
20:11:33amiconnAh, so that's the same as on 2nd gen like I expected
20:11:44amiconnNow plug it into your pc and check the value again
20:11:51jhMikeSamiconn: will dircache rescan before the rolo check though?
20:11:53DerPapstfrom B6 -> a& -> 26 on firewire
20:12:14amiconnYOur computer has a 6-pin firewire socket?
20:12:31amiconnNice, then it's all the same :)
20:12:37DerPapstgood ;)
20:12:46DerPapstshall i do this with usb too?
20:12:58amiconnBit 7 is firewire detect, bit 4 is power detect
20:13:27linuxstb_DerPapst: Do you have a usb charger as well?
20:13:38DerPapstonly firewire.
20:13:56linuxstb_The problem we had with USB on the later ipods was distinguishing between a PC connection and a charger connection...
20:14:48amiconnNow all that's left is the nano...
20:14:50DerPapstnone of the values changed... i mean GPIO_B and D are changing all the time....
20:15:15DerPapstamiconn: i think it will behave like the video
20:15:33amiconnI think so too, but I'd like to see that verified
20:16:32DerPapstbest would be to go to the firums and make a topic with subject "NEED TESTERS" you'll have 10 pages filed in a second :D
20:16:47amiconnlinuxstb: I think we can distinguish USB and USB power without enabling the usb controller...
20:16:59linuxstb_On the later ipods?
20:17:23linuxstb_That would be nice.
20:17:26amiconnOn my mini, both B1 and D4 change when plugging it into the computer
20:18:40*Llorean sighs
20:18:51LloreanStill no responses on the Nano thread, or on the flyspray task where I've asked for more information
20:19:04LloreanThey were responding at least once a day until I asked them to do something.
20:19:06amiconnHmm, same on an usb charger...
20:19:18DerPapsttheir ipod magically worked again... ;)
20:19:22amiconnLlorean: Do you have a firewire cable & charger?
20:19:53LloreanNo, sorry
20:20:01LloreanOnly USB
20:20:09amiconnSorry I mean B0 and D3
20:21:32amiconnThe 3rd gen should detect USB power on D4 according to ipl
20:22:18 Join EbErT [0] (
20:23:08 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
20:23:24 Join miepchen^schlaf [0] (
20:25:56GodEaterLlorean: I guess I'll extend the plugin to handle mutliple link files then if you think it's a good idea.
20:26:48amiconnDerPapst: Could you check port D with an USB charger again?
20:26:48 Join desowin [0] (
20:27:01amiconnKeep an eye on the first hex digit
20:27:30LloreanGodEater: If you just make it a viewer for .link files, that should cover nearly all the functionality I described anyway, right?
20:27:39amiconnIf it changes constantly both with and without the charger, the cycle might be different
20:27:55LloreanThe only thing that doesn't cover is "automatically executing single-line link files"
20:28:26GodEaterLlorean: it's a bit more involved than that, but it shouldn't be *too* hard
20:28:46LloreanI thought you were earlier having trouble *not* having it works as a viewer for link files.
20:29:01GodEaterLlorean: it's registered as a viewer for all .link files
20:29:22GodEaterbut that simply means rockbox launches the plugin with the link file name as a parameter
20:29:51GodEaterwhich is effectively the same as launching it with the "Add to shortcuts" option in the context menu
20:30:19GodEaterso I'll need to add some logic for testing if the file given as an argument has the .link extension
20:30:21LloreanSo it needs to check if the parameter is a link filename, and load it as a link file rather than adding.
20:31:03GodEaterwhich is why one of my testers ended up with "other" .link files just being added into the "main" one
20:31:33GodEaterto be honest, I'm not sure how much more I want to do to this until I get a feel for if it's wanted in svn or not
20:31:44LloreanStill, I think it'd be good to have it works as a generic viewer for .link files, since there's no real harm in that.
20:31:44LloreanThough by no means do you need to implement that.
20:31:44DBUGEnqueued KICK Llorean
20:31:50GodEaterI'm dreading the "Don't add more stuff to the context menu" crowd already
20:33:31GodEaterLlorean: dare I ask if you've actually tried it out yet ?
20:34:28LloreanI have not. It's not really something I can even imagine an edge case situation where I'd have a use for it, in all honesty.
20:35:18GodEaterI thought that when I started writing it
20:35:24GodEaterbut I've actually started using it
20:35:27GodEaterwhich surprised me!
20:36:57 Join Guile`` [0] (n=Guile@
20:37:53amiconnlinuxstb: Could you verify whether your color behaves the same as the mini g2?
20:38:37linuxstb_amiconn: Are you asking about firewire or usb, or both? I don't have a firewire charger.
20:38:41amiconnI found C1 = firewire detect (low active), C2 = firewire power (low active), D3 = USB power (high active)
20:39:16amiconnDo you have a firewire socket + cable?
20:39:36amiconn6-pin or 4-pin?
20:39:54linuxstb_6 I think - give me a few minutes to find it...
20:40:07 Join XavierGr_ [0] (
20:40:08 Quit XavierGr (Nick collision from services.)
20:40:09 Join GodEater__ [0] (n=vircuser@rockbox/staff/GodEater)
20:42:01CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
20:42:01*jhMikeS recommends for a look
20:43:34DerPapstamiconn: sorry was afk...
20:43:38DerPapsti'll read up
20:44:56DerPapstamiconn: i don't have a usb charger. only firewire... but maybe i can build something ;)
20:45:22amiconnUSB power can be checked with any USB socket except on a passive hub
20:45:41*DerPapst as an active one...
20:46:30linuxstbamiconn: It's a 6-pin cable.
20:47:01 Join Soul-Slayer [0] (n=Administ@
20:47:26amiconnlinuxstb: Plug it into your pc & ipod color and watch GPIO C values...
20:47:29 Quit donutman25 (Read error: 104 (Connection reset by peer))
20:48:26DerPapstfrom C/8 -> D/9 (letter and number being both the first value)
20:48:46linuxstbamiconn: It changes from 0x36 (normal) to 0x30 (firewire attached to PC)
20:49:16DerPapstlast value stays at 4
20:49:39amiconnDerPapst: Okay, so D4 it is, high active
20:49:48amiconnIt also means the G3 can be charged from USB
20:50:02DerPapstthe of doesn't allow that
20:50:17amiconnHmm, then why does it detect power? weird...
20:50:23DerPapstand nither does diskmode-
20:50:37amiconnlinuxstb: The value doesn't toggle between 30 and 31 when plugged?
20:52:20 Quit ptw419 ()
20:52:37amiconnjhMikeS: I tested atomic mutexes on SH. No binsize penalty
20:53:10 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
20:54:19 Join maffe [0] (
20:54:20jhMikeStoo bad there's no reduction from it
20:54:49 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
20:55:02amiconn4 instructions with svn are replaced by 4 insns with atomic mutexes
20:55:10amiconnlinuxstb: [20:50:46] <amiconn> linuxstb: The value doesn't toggle between 30 and 31 when plugged?
20:55:38linuxstb_Not that I could see.
20:55:49 Quit linuxstb (Nick collision from services.)
20:55:55amiconnOkay, so perhaps the controller isn't enabled
20:55:56*jhMikeS totally munged the term "spinlock" here
20:56:05 Nick linuxstb_ is now known as linuxstb (i=5343d4aa@rockbox/developer/linuxstb)
20:56:27amiconnI should commit DEV_EN display in View I/O Ports
20:57:23jhMikeSView I/O Ports doesn't scroll does it?
20:57:33DerPapstwow... one can access the G3's hhd from a pc while in diagnostic mode ^^
20:58:28amiconnjhMikeS: No it doesn't
20:58:47amiconnBut imo the GPIO values can be displayed in a more compact way
20:58:58jhMikeShow so?
20:59:03DerPapstthere is an option in diag mode called USB DISK
20:59:16amiconnGPIO: A:xx B:xx C:xx D:xx et
20:59:22DerPapstif i enter that and attach my usb cable i can access the iPod.
20:59:23jhMikeScould be done with a list
20:59:52jhMikeSyou can printf whatever you want to a line in the list
21:00:06DerPapstif i press >> it says FW Disk...
21:00:08amiconnThen you can't watch the effect of button gpio
21:00:12DerPapstjust cool :D
21:00:33DerPapstand goes in to diskmode now... that is odd
21:00:45amiconnThe ports debug screen only has a single button to leave it, and that triggers on release, on purpose
21:01:16jhMikeSsome small tweaks would allow it I think
21:01:16 Join SliMM [0] (n=chatzill@
21:01:23SliMMhey guys
21:01:50amiconnjhMikeS: Nah, imo it's not worth the hassle. Some of the clock register stuff can probably go away now
21:02:18jhMikeSI was thinking cycling the hold switch or something could lock/unlock it
21:03:48amiconnThe hold switch? ANd what about targets without hold switch?
21:04:31jhMikeSwell, they must have something to click that doesn't involve a scoll action
21:04:42amiconnDerPapst: Wanna test a small patch
21:05:33*DerPapst fires up cygwin
21:05:39***Saving seen data "./dancer.seen"
21:06:07 Quit My_Sic (Read error: 113 (No route to host))
21:06:12DerPapstif it works ;)
21:06:33 Quit spiorf_ (Connection timed out)
21:07:07DerPapstok.. recived
21:07:22 Join spiorf_ [0] (
21:07:50*DerPapst wonders how to dcc files himself...
21:08:11amiconnAfter installing the build, connect the ipod to the PC via firewire. Then enter 'View HW Info'. It should reboot into diskmode
21:08:29amiconnIf you enter 'View HW info' without firewire connected, it should display as normal
21:09:42amiconnIt should also not reboot when connected to the charger
21:10:39jhMikeSthose FIQ stacks are giant
21:10:59amiconnAre they? Just 256 words...
21:11:29jhMikeS1KB? I don't think we use 1/100th of that.
21:11:36DerPapstamiconn: ok
21:11:58amiconnDerPapst: It only tries to detect firewire when entering that screen
21:12:54*DerPapst sees some fancy memcpy action :)
21:15:01*DerPapst woders where )0x40017f00 is....
21:17:14*linuxstb curses Llorean for always replying to forum threads first...
21:17:22*Llorean looks innocent
21:19:11linuxstbDerPapst: The end of IRAM
21:23:10amiconnDerPapst: It may be that the memcpy action isn't correct for the G3. Then it will reboot, but not go into diskmode
21:23:14DerPapstLlorean has nearly half of the total forum posts...
21:23:28DerPapstamiconn: ok
21:23:52*DerPapst installs
21:23:56amiconnIf that's the case, it would be useful to have a dump of the G3 ROM
21:24:09LloreanDerPapst: No, about 10%
21:24:12DerPapstthe aupd image isn't decrypted
21:24:24linuxstbLlorean: That's OK then ;)
21:24:35DerPapstLlorean: i sad nearly... ;)
21:25:26DerPapst10% is nearly 50% ;P
21:27:01DerPapsti bet one day the reply button is going to take revenge on you becuase you've clicked it too often
21:29:06DerPapstamiconn: worked
21:29:07 Quit toffe82 (Read error: 104 (Connection reset by peer))
21:29:11DerPapstand it's in diskmode :)
21:29:21amiconnYay! :)
21:29:52amiconnSo I can start adding that to the correct places...
21:30:19linuxstbSo I assume the same IRAM magic works on both 2g and 3g?
21:30:46*DerPapst wonders if it would be possible to carge a g§ iPod in rockbox using USB
21:31:04amiconnWell, ipod charging itself is hardware
21:31:09amiconnSo just try it
21:31:22DerPapstok then i know it doesn't wirk
21:31:28DerPapst*work even
21:31:32 Join toffe82 [0] (
21:31:33amiconnStart ipod, go to 'View battery', plug in USB, and let it sit for a while
21:32:04amiconnIt might be that the charger needs to be enabled somehow
21:32:13GodEater__Llorean: I have it handling all .link files now, and it does the "single entry" one as you suggested too
21:32:25amiconnI don't need to do that on 2nd gen (for firewire charging of course)
21:32:27GodEater__all I need to do now is stop the user from attempting to delete things from their user generated one
21:32:41linuxstbGodEater__: Why?
21:32:48*linuxstb waits to be told to shut up
21:33:16 Join rasher [0] (n=rasher@rockbox/developer/rasher)
21:33:33preglowshut up!
21:34:01DerPapstdoes rockbox do idle poweroff in debugmenu?
21:34:22amiconnSo disable poweroff before
21:34:35DerPapstgood idea ;)
21:34:46*amiconn has it disabled on his G2 because it would just reboot otherwise
21:35:21GodEater__linuxstb: cos Llorean changed his mind!
21:35:50linuxstbGodEater__: I mean why is it a problem if a user wants to delete items in their own .link files?
21:36:15GodEater__because the delete function is hard wired to delete from the main shortcuts file
21:36:37rasherAny perl/shell gurus that can explain this: (why the third execution of gives "1" rather than "test" as output)
21:37:00GodEater__I figure the user has already gone to the trouble of manually creating a .link file with either the rockbox text editor (yuk), or on their PC
21:37:04rasher(genlang has this problem)
21:37:14GodEater__they aren't going to object to editing it again to remove a favourite they don't want any more
21:37:20linuxstbGodEater__: I assumed your plugin would load the shortcuts file into RAM, allow the user to either select one, or delete some, and then if the user has deleted some, write it back to disk (using the same filename)
21:38:16*GodEater__ fumbles for that gun again
21:38:32linuxstbIt's OK, I'll take preglow's advice now.
21:39:36GodEater__I'll be honest, it wasn't as much work getting it *loading* user files as I thought
21:39:38preglowyou might not want to mistake it for advice!
21:39:47GodEater__but the delete routine made me cry this afternoon
21:39:47linuxstbrasher: I don't know what the -s option does, but when you run "perl" it's not using the -s option, but ./ is.
21:39:55GodEater__and I don't want to go near it again unless I have to
21:40:24rasherlinuxstb, well there you go
21:40:44 Join ender` [0] (
21:41:04GodEater__doesn't calling "perl something.perl" ignore the shebang line ? even
21:41:36rasherYes. The -s is exactly the problem (it enables some argument parsing)
21:41:37linuxstbYes, that's what I just said ;)
21:41:39GodEater__whilst "./" runs the interpreter specified in the shebang, along with arguments
21:41:54GodEater__I'll shut up and go back to my C then shall I
21:42:11rasherI think I was mistaking -s for being strict mode, and not caring about it.
21:42:16amiconnlinuxstb: I wonder whether I should rename the USB_* definitions to something more generic, and what that could be...
21:43:16 Nick gnakijefh is now known as Anakin (i=0@
21:44:07 Join petur [0] (n=petur@rockbox/developer/petur)
21:46:55rasherGodEater__, Thanks, but I found this already. Problem solved long ago
21:49:34DerPapstGodEater__: yet anouter idea for your plugin... somethin like a create option in its menu when you started it from the plugin browser. You select create new shortcut and then the browser opens, then you browse for the file/dir you want to link to and then you hold down select for a fiew seconds and the link is made :)
21:49:43*DerPapst runs far far away
21:50:36*DerPapst made some nice typos again (<−− shouting from far far away)
21:50:52GodEater__DerPapst: I think the way I came up with for creating / adding is better than that
21:51:01peturDerPapst: KISS please
21:51:05linuxstbamiconn: Which USB_* definitions are you talking about?
21:51:16linuxstbSYS_USB_* ?
21:51:17pixelman1s: you around?
21:51:29amiconnSYS_USB_CONNECTED, USB_IPODSTYLE, ... the whole thing
21:51:32DerPapstpetur: isn't that kiss (at least for the user to use it...)
21:51:55amiconnMaybe I should just pretend that firewire is a kind of USB....
21:52:16GodEater__DerPapst: that seems a long winded way to create a shortcut to me
21:52:16linuxstbI don't think it's a big issue - perhaps just leave it as it is for now...
21:52:35DerPapstGodEater__: hoe do you create them?
21:52:40GodEater__Long select on the file/dir and then choose "Add to shortcuts" seems easier to me
21:52:53GodEater__much quicker
21:53:12DerPapsti didn't know that is possibe ;)
21:53:13*GodEater__ thinks bluebrother may shoot him though for another entry in the context menu
21:53:29GodEater__DerPapst: you live and learn ;)
21:53:35DerPapsthe can't shoot that far :P
21:53:40DerPapstso no worries.
21:53:41linuxstbGodEater__: Does your plugin need to be a plugin? Would it be simpler to just keep it as a viewer only?
21:53:49DerPapstadd what ever you want to the menu :D
21:54:22GodEater__linuxstb: you know what our users are like for "where do I download X" when we've made it a viewer ;)
21:54:40GodEater__as far as I can see, it loses nothing for being a plugin too
21:54:55DerPapsti wondered that to not long ago
21:54:56GodEater__if someone *wants* to navigate to it in the plugins menu, who am I to stop them ?
21:55:03DerPapstwith the midi plugin ;)
21:55:10*petur feels it's about time to store plugins in categories (folders) like demos, games, utilities
21:55:26*krazykit agrees
21:55:36GodEater__the plugin browser would need some work
21:55:37DerPapstand there is already a FS entry with a patch about that
21:55:42GodEater__it can't see folders at the moment
21:56:00amiconnWell, not filtering folders would be simple, but not a proper solution
21:56:04*GodEater__ thinks he's probably trying to teach petur to suck eggs though
21:56:09DerPapstbut there are some issues that i don't understand
21:56:25*petur looks at opened can or worms
21:56:32amiconnI'd want to browse 'Spiele', 'Anwendungen' etc, not 'Games', 'Applications'
21:56:49peturamiconn: yes
21:57:22GodEater__ah of course
21:57:35DerPapstnow i understand that issue again :)
21:58:29linuxstbWhy not just have subdirs in plugins, and have new menu options which open those subdirs instead of the main plugins directory?
21:58:44linuxstbThen those menu items can be localised.
21:59:09amiconnIt adds a few language strings though (unavoidable)
21:59:22linuxstbAlthough that doesn't solve the problem of the plugins themselves having English names...
21:59:42GodEater__if someone wants to put a definitive list of categories together, I might attempt that too - I think I could manage it
21:59:46amiconnThis random_folder_advance_config.rock should perhaps be a direct-load plugin then
22:00:07amiconn(like credits.rock already is)
22:00:08linuxstbI've thought about that - you mean a menu option which just runs that plugin?
22:00:37DerPapstGodEater__: there is one in that patch already
22:00:57amiconnlinuxstb: Yes, somewhere near the option to set folder advance to 'Random'
22:01:08amiconn(not that I would ever use it....)
22:01:59DerPapstcan't one do that in the plugin? e.g. in podzilla2 the module has a path where in the menu it shows up. and with a lanuage file you can chnage that.
22:02:00GodEater__DerPapst: FS# ?
22:02:12jhMikeSamiconn: pp5002 allows ldrb access to PROCESSOR_ID?
22:04:19GodEater__DerPapst: that's not the approach just discussed, although it might be a nice place to start I suppose
22:04:39amiconnjhMikeS: I would think so, although the OF doesn't do it
22:05:02amiconnBut there are other places where the OF does long access and we do byte access
22:05:09amiconn(ata registers for instance)
22:05:39DerPapstGodEater__: at least it contains the list you wanted to see ;)
22:05:40amiconnI didn't even try dualcore so I can't say for sure
22:05:53amiconnI want to get firewire handling in place, then sleep
22:06:27amiconnI mean CPU sleep (well, perhaps amiconn->sleep() ) before
22:07:00*amiconn counts 22 open windows on his desktop
22:07:02jhMikeScrt0-pp.S currently does ldr then and ,#0xff but just seems unneeded
22:07:21amiconnAnd 38 firefox tabs
22:07:41jhMikeSno multimonitors?
22:07:48 Join webguest89 [0] (i=52994c2e@gateway/web/cgi-irc/
22:08:00amiconnAt least it's 1400x1050
22:08:13webguest89I was on yesterday under the same name
22:08:36webguest89just thought that I'd mention that I have fixed the Ipod
22:08:42jhMikeSthat's a weird res
22:09:14GodEater__no it's not
22:09:41webguest89I had to do a factory reinstall of the Ipod software, then put Rockbox on again. Works like a charm
22:09:43krazykitamiconn, sounds like you need virtual desktops :D
22:10:04amiconnkrazykit: So I'm always searching where my windows are? Nah, surely not
22:10:05jhMikeSGodEater__: just for laptops? I usually stay clear of them.
22:10:07 Quit inmotion (Read error: 110 (Connection timed out))
22:10:43amiconnThat's what a taskbar is for. I have it in 2-row mode atm
22:10:58 Quit desowin (Remote closed the connection)
22:11:05amiconn(and no nasty grouping, which would have the same effect as virtual desktops)
22:11:49*GodEater__ likes virtual desktops
22:12:17DerPapstwebguest89: good to hear :)
22:12:26DerPapstthough i can'T remember the proble :D
22:12:35DerPapstproblem even
22:12:41amiconnjhMikeS: I saw desktop lcd monitors with the same resolution
22:14:03*GodEater__ misses his 1900x1200 laptop
22:14:14jhMikeShmm...must be an lcd thing then. I'm partial to CRTs but maybe I'll be forced to get one when I can't keep these running any longer.
22:14:25*amiconn hates CRTs meanwhile
22:14:41GodEater__they didn't make many widescreen format CRT monitors
22:14:50GodEater__so no wonder it looks like an odd res
22:14:57jhMikeSkeep 'em stocked with fresh caps since those dry out
22:15:01amiconn1400x1050 is 4:3
22:15:11amiconnAnd I don't like widescreen for the computer
22:15:31amiconnFor TV it's nice though
22:17:14 Part maffe
22:17:16GodEater__widescreen on a PC is something I've grown used to
22:17:20GodEater__I like it now
22:17:22*DerPapst agrees
22:17:23 Quit webguest89 ("CGI:IRC (EOF)")
22:17:35DerPapst... to amiconn's statement
22:17:47GodEater__2 text editor sessions sit nicer side by side
22:17:52 Join maffe [0] (
22:18:06DerPapstGodEater__: yes. the reason why i have 2 screens :P
22:18:18*DerPapst can't live without them anymore
22:19:28GodEater__I have the luxury of two screens at work
22:19:33DerPapstbrb ~15-20minutes
22:19:35GodEater__but still wish they were both 16:9
22:23:25 Quit GodEater__ ("User pushed the X - because it's Xtra, baby")
22:28:33Bagderoops, the build rev numbers were wrong
22:28:41Bagderthey were all showing the m5 one ;-)
22:28:43 Quit ender` (" Good programmers treat Microsoft products as damage and route around them.")
22:29:20*linuxstb fines Bagder a week's Rockbox wages
22:29:48Bagderoh man, that hurts
22:30:23 Quit davina ("xchat on Ubuntu 7.04")
22:31:09jhMikeShmmm...the COP never seems to be switched back to supervisor mode after setting up undefined mode :\ (What am I missing here?)
22:32:01 Join ender` [0] (
22:34:46 Join barrywardell [0] (
22:36:52amiconnjhMikeS: !
22:37:02jhMikeSam I right or hallucinating?
22:37:13amiconnIt looks like you're right
22:37:14*petur returns to his 1920x1200 laptop :)
22:37:30jhMikeSguess I'd better fixur uppur
22:37:57amiconnSo the cop runs in undefined mode....
22:38:07*amiconn wonders what that might imply
22:38:11jhMikeSheh...that seems appropriate
22:38:33amiconn...especially if an interrupt occurs
22:39:23jhMikeSsansa boots ok with it fixed
22:40:47 Quit Xerion (" ")
22:42:06jhMikeSundefined instruction is the lowest priority exception mode
22:42:14DerPapstbäck :)
22:42:45Bagderhey 7zip has 614 open bug reports, so we're not #1 ;-)
22:43:06amiconnjhMikeS: What happens if it runs in undefined mode and an undefined instruction is encountered?
22:43:08Bagderbut they only have 75 patches
22:43:36Bagderthey won the "best project" in sourceforge "Community Choice Awards"
22:44:51amiconnjhMikeS: Quick check on 2nd gen shows no problems with your commit
22:44:54DerPapstwell they have slightly more users.. otherwise the rockbox project would win for sure ;)
22:45:04amiconn(but then we're not running dualcore on PP5002)
22:45:25BagderDerPapst: yeah they probably have magnitudes of more users
22:45:47amiconnAnd I did a demanding test: running mandelbrot.rock while playing music
22:45:57BagderI was just surprise to see the "best project" have that many open bugs etc
22:46:38jhMikeSamiconn: I
22:46:48jhMikeS'm not sure it can enter UD mode again
22:46:58 Join tedrock [0] (
22:47:00jhMikeSat least the commit is good :)
22:47:20jhMikeSprobably should just change the mode there now too
22:47:29 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:47:43 Quit amiconn (Nick collision from services.)
22:47:45 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:47:50 Quit tedrock (Client Quit)
22:52:13 Join McEnroe [0] (
22:52:43McEnroeis there any info which themes are cjk "compatible"?
22:53:34linuxstbI don't know about themes, but there is a page for fonts -
22:54:01 Join tedrock [0] (
23:00:35 Quit tedrock (Client Quit)
23:04:23 Quit spiorf_ (Remote closed the connection)
23:04:29*midgey cheers after fixing the sim on his mac
23:04:34 Quit ender` (" The sentence 'On the fish and chips-sign he wanted to have a hyphen between fish and and and and and chips.' would be a lot )
23:05:20 Join tedrock [0] (
23:05:42***Saving seen data "./dancer.seen"
23:06:05*barrywardell cheers with midgey
23:06:41midgeywell, for some reason SDL refused to read files on the disc :-/
23:07:24linuxstbWhat has the file handling got to do with SDL?
23:07:53midgeyeh it probably didnt
23:08:20linuxstbHow did you fix it?
23:08:40midgeyinstalling a new version of sdl
23:09:02midgeywell, technically an old version
23:09:05 Join ender` [0] (
23:09:22linuxstbHave you tried installing qt on your Mac? (for rbutilqt)
23:09:39midgeyno, only on my linux computer
23:09:45midgeydoes it not work?
23:10:26linuxstbQt didn't compile for me. The make failed with errors about unknown object types (IIRC)
23:10:48*midgey tries
23:10:59*amiconn has his mini g2 auto-rebooting into diskmode when connecting firewire :)
23:12:11 Join alienbiker99 [0] (i=484c2e88@gateway/web/cgi-irc/
23:13:24DerPapstto firewire power + data or power only or data only?
23:13:38amiconnData only of course
23:13:47DerPapstnice thing :)
23:14:12amiconnThis will work for mini g1, mini g2, color and g4
23:14:34amiconnNow to write usb-pp5002.c which will do the job for g1..g3
23:14:49amiconn(except that it shouldn't be called "usb"...)
23:16:00amiconnWell, for g3 it will also contain usb code, but that has to be written by someone else
23:16:31 Quit alienbiker99 (Client Quit)
23:16:44 Quit ender` (Read error: 104 (Connection reset by peer))
23:18:30DerPapstbut then it has to be usb-fw-502X.c too
23:18:30*jhMikeS is going to try a whacky experiment
23:19:04*linuxstb puts on his tin hat
23:19:10 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
23:19:18 Join miepchen^schlaf [0] (
23:19:41jhMikeScodec on two cores at once
23:19:48 Join ender` [0] (
23:20:00DerPapstgood luck :)
23:20:31linuxstbWhat do you expect?
23:20:40*DerPapst wonders how to do a svn diff so new files are included in the diff file and old are removed...
23:21:00DerPapstlinuxstb: maybe no crashes... ;)
23:21:01jhMikeSthe spc codec to run the emulator on cop and the sample output to main
23:21:30jhMikeSor cpu rather
23:21:37jhMikeSthey'll run in parallel then
23:21:43linuxstbAh, you mean making a specific codec multi-threaded?
23:22:20jhMikeSas an initial tester
23:22:45 Join Rondom [0] (
23:23:03linuxstbI understood "codec on two cores at once" to mean the same code running on both cores, which did indeed sound whacky...
23:23:56jhMikeSwell, if caching weren't an issue I probably would try that too
23:24:20linuxstbBut that's how I imagined Rockbox would make use of dual-core - specific CPU-intensive parts would be made to use the COP, rather than making Rockbox in general become multi-core friendly...
23:24:42jhMikeSsome parts should be
23:25:24jhMikeSlike pcm output should be started on the initiating core and such. perhaps allow disk I/O and some other limited access by the COP.
23:26:12*jhMikeS was never thinking full SMP
23:28:42bluebrotherDerPapst: svn add the files first or append the new files by diffing to /dev/null
23:29:23 Quit midgey ("Leaving.")
23:30:47DerPapstbluebrother: thanks. will try
23:34:44 Quit EbErT ()
23:35:47DerPapstafter adding the file with svn add it didn't show up in the diff file and doing "diff file.c /dev/null > burrito.diff" has another syntax
23:37:03obodiff -u file.c ....
23:37:06amiconnyeah, svn diff does a unified diff by default
23:37:29DerPapstah thanks
23:42:57linuxstbsvn add should work though...
23:43:58*linuxstb wonders what DerPapst is writing
23:44:02DerPapstit is svn add file.c or am i wrong?
23:44:42linuxstbThat's right.
23:45:39DerPapstodd... now it's there :D
23:46:38 Quit McEnroe ("leaving")
23:48:05DerPapsteh? a t the end of the diff: Property changes on: bin2c.c <new line> ___________________________________________________________________ <new line> Name: svn:executable <new line> + *
23:49:28barrywardellare there plans to commit toni's arm disassembler to svn?
23:49:32DerPapstlinuxstb: have to dissapoint you. i'm just updateing ipodpatcher with Loader 2 :P
23:49:47 Quit linuxstb ("CGI:IRC (Ping timeout)")
23:50:07DerPapsthe quit because og that now i guess....
23:50:10 Quit jgarvey ("Leaving")
23:52:16 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
23:52:53 Nick linuxstb_ is now known as linuxstb (i=5343d4aa@rockbox/developer/linuxstb)
23:53:44 Join rotator [0] (n=e@rockbox/developer/rotator)
23:53:58linuxstbbarrywardell: It's been mentioned before, and everyone is in favour. I guess Toni has just never got around to it (IIRC he has svn access)
23:54:24barrywardellit's very useful. I definitely think it should be committed
23:55:03linuxstbI think there was a short mailing list thread about committing it...
23:55:06 Quit freqmod (Read error: 104 (Connection reset by peer))
23:58:05linuxstbYou could just commit it if you wanted - the only question is where in svn does it go? Either a new "utils" directory (gets my vote), or in tools.
23:58:51amiconnDerPapst: Would you prefer the firewire logo or usb logo one the G3? USB isn't detected atm anyway...

Previous day | Next day