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 2007-07-24

00:00:26 Join safetydan [0] (i=3b6408cc@rockbox/developer/safetydan)
00:00:31PaulJamWidget: what database settings do you use? i havn't had duplicate entrys in my database for a long time.
00:01:14LloreanThe simplest solution is just to play the duplicate song, then choose ID3 Info in the context menu and scroll to the bottom to see where the real file is located.
00:01:34WidgetPaulJam: loadtoRAM, AutoUpdate, GatherRuntimeData are the ones turned on
00:01:56Widgetit's just rebuilding the DB atm, but i'll check that in a second
00:02:30 Quit safetydan (Client Quit)
00:02:43Widgetcan the iriver run as a music player with USB power? it just displays a USB symbol here, which I assume means it can't
00:02:46 Quit robin0800 (Read error: 110 (Connection timed out))
00:02:51PaulJamhmm, i use the same settings
00:03:27linuxstbWidget: Which iriver?
00:03:33 Join safetydan [0] (i=3b6408cc@rockbox/developer/safetydan)
00:03:34Widgetlinuxstb: h320
00:03:36 Quit safetydan (Client Quit)
00:03:47linuxstbWidget: Then yes - hold MENU as you insert the USB cable (IIRC)
00:03:50 Quit Rob2222 ()
00:04:08Widgetlinuxstb: menu being the a.b thing
00:04:29pixelmabtw... does someone know if "gather runtime data" depends on another setting (like "Auto-Update" or the like)? The wiki doesn't state one but simply enabling gave me no result on Ondio (wanted to test it)
00:04:48Widgetlinuxstb: hmm, didn't do anything different here, but no matter
00:06:31 Join safetydan [0] (i=3b6408cc@rockbox/developer/safetydan)
00:07:55PaulJamhmm, my h300 just froze after disconnecting USB. it shows the USB logo and below it the text "Lese Festplatte...". disk is still spinning.
00:08:24linuxstbWidget: Try REC..
00:08:49Widgetlinuxstb: w00t!
00:10:36*jhMikeS looks in the IRC history and wonders about the meaning of "restructured". Just say it. It's ok: "Completely screwed it up so what it's supposed to do is completely unrecognizeable." :P
00:11:05 Quit robin_0800 (Read error: 110 (Connection timed out))
00:11:19 Quit Lear ("ChatZilla [Firefox]")
00:12:01 Quit safetydan ("CGI:IRC (Ping timeout)")
00:12:49davinaniht night
00:12:52 Quit davina ("xchat on Ubuntu 7.04")
00:13:49Widgetwell, reinitialising the database fixed it
00:14:17PaulJamWidget: do you use dircache?
00:14:35WidgetPaulJam: turned it on before initialising just now
00:14:54linuxstbjhMikeS: :) Talking about mpegplayer, have you experienced the stack overflow issue people have been reporting? It seems to only occur without a mpegplayer.cfg file.
00:15:14linuxstb(and only on Coldfire afaik)
00:16:05jhMikeSlinuxstb: no, I only ever really use it on targets that can actually play video well which is basically sansa and gigabeat
00:16:14PaulJamif you had it off before then this could have caused the duplicates. if dircache is off then the database only picks up new songs, but doesn't delete songs that aren't there anymore, so if you move files on the harddrive you get duplicates.
00:16:16 Quit saratoga ("CGI:IRC (Ping timeout)")
00:17:49PaulJam(in combination with autoupdate)
00:18:12jhMikeSlinuxstb: which stack? the main stack?
00:18:21linuxstbYes, the main stack.
00:18:36WidgetPaulJam: that'll be it then
00:20:03linuxstbIt seems to happen when saving the .cfg file, and it only seems to be a tiny overflow. So I'm not sure how best to fix it - if we remove that large libmad buffer from the stack, it will probably significantly slow down playback on Coldfire, but I don't have a colour Coldfire target to properly test.
00:21:01 Join robin0800 [0] (
00:21:05Widgeti really like the idea of mpegplayer, but it doesn't fast enough on my h320, and without being able to seek, its use is limited
00:21:10jhMikeSThe next move with that is to have a UI thread and a plan to not require adding another thread slot in the core.
00:21:15WidgetI should really learn how to use the iriver firmware
00:22:09*jhMikeS is working it toward seeking...the reliable thread control is part of that and needed so something was hacked together to get around the current multicore problems.
00:22:39thomasgrockbox rocks :)
00:23:05linuxstbWhat's the cost of adding an extra thread slot? How can it be avoided?
00:23:23linuxstbSo the intention is four threads? video, audio, buffering and UI?
00:23:39WidgetjhMikeS: cool!
00:23:41jhMikeSlinuxstb: I have a way to make scrolling work on both displays with one thread and no compromise.
00:24:23jhMikeSlinuxstb: I think that's the best arrangement.
00:24:59Widgetbedtime for me, cya all
00:25:13 Quit Widget ("Leaving")
00:25:23linuxstbHmm, the H300 builds fine with that libmad buffer in normal IRAM, rather than on the stack...
00:27:08jhMikeSgtg. mom's b'day and we're all going for sushi. bbl
00:27:15linuxstbOK, have fun...
00:27:27linuxstbBut the ipods don't build...
00:31:59 Quit jhulst ("Konversation terminated!")
00:36:18 Quit PaulJam (".")
00:38:14*pixelma still shakes head about "Knopf Beleuchtung"...
00:41:17 Quit [1]robin0800 (Read error: 110 (Connection timed out))
00:41:27 Join [1]robin0800 [0] (
00:43:42 Join robin_0800 [0] (
00:46:41Nico_Pbluebrother: trying to compile your rbutilQt
00:47:01Nico_Pwhat am I supposed to do? qmake-qt4 then make doesn't work
00:47:32 Quit sarixe ("Peace")
00:48:53oboNico_P: in the root, or in src?
00:49:02Nico_Pobo: that was in the root
00:49:03 Quit krp (Read error: 60 (Operation timed out))
00:49:19oboshould work in src
00:49:20Nico_Pin src it seems to work better :)
00:49:46 Join krp [0] (
00:51:47Nico_Pwow nice :)
00:53:21Nico_Pthis might actually motivate me to start fillding with C++ and work on RBUtil
00:54:56 Join HellDragon_ [0] (n=Nocebo@unaffiliated/helldragon)
00:55:12 Quit HellDragon (Nick collision from services.)
00:55:20 Nick HellDragon_ is now known as HellDragon (n=Nocebo@unaffiliated/helldragon)
00:58:00 Quit adiamas ("ChatZilla [Firefox]")
01:03:05 Quit Entasis (Read error: 104 (Connection reset by peer))
01:03:18 Quit obo ("bye")
01:03:35 Quit XavierGr ()
01:06:47 Quit MournBlade ("CGI:IRC (EOF)")
01:07:51 Quit robin0800 (Read error: 110 (Connection timed out))
01:08:00 Quit [1]robin0800 (Read error: 110 (Connection timed out))
01:08:55***Saving seen data "./dancer.seen"
01:09:59 Join XavierGr [0] (
01:10:55*linuxstb spots a green build table and goes to sleep
01:10:57 Quit linuxstb ("CGI:IRC")
01:16:04 Quit robin_0800 (Read error: 110 (Connection timed out))
01:17:55 Quit Arathis ("Bye, bye")
01:19:45 Join [1]robin0800 [0] (
01:24:01 Join robin0800 [0] (
01:24:02 Part toffe82
01:26:14 Join robin_0800 [0] (
01:29:40 Quit Nico_P (Remote closed the connection)
01:30:35 Join [2]robin0800 [0] (
01:30:35***Alert Mode level 1
01:30:35DBUGEnqueued KICK [1]robin0800
01:30:35DBUGEnqueued KICK robin0800
01:30:35***Alert Mode level 2
01:30:35DBUGEnqueued KICK robin_0800
01:30:35DBUGEnqueued KICK [2]robin0800
01:30:35***Alert Mode level 3
01:31:53 Join [3]robin0800 [0] (
01:31:53***Alert Mode level 4
01:31:53***Alert Mode level 5
01:31:53DBUGEnqueued KICK [3]robin0800
01:31:53***Alert Mode level 6
01:31:53***Alert Mode level 7
01:31:53***Alert Mode level 8
01:31:53***Alert Mode level 9
01:32:47 Join helpme401 [0] (i=4122e228@gateway/web/cgi-irc/
01:32:59helpme401hello can someone help me out with something?
01:33:28 Join [4]robin0800 [0] (
01:33:28***Alert Mode level 10
01:33:28***Alert Mode level 11
01:33:28DBUGEnqueued KICK [4]robin0800
01:33:28***Alert Mode level 12
01:33:28***Alert Mode level 13
01:33:28***Alert Mode level 14
01:33:28***Alert Mode level 15
01:33:28***Alert Mode level 16
01:33:34helpme401I recently installed rockbox on my 2nd generation ipod mini
01:33:59helpme401I put one song on there and it worked fine but then I deleted it and added other songs
01:34:23helpme401now on the roxkbox explorer it shows that song I deleted being there but none of the other songs I recently put on
01:34:35helpme401however the song I deleted shows up under the name but will not play
01:34:53iambenright, the database probably has not been updated
01:34:59helpme401how do you do that?
01:35:31 Join kkurbjun [0] (
01:35:35iambengo into your settings, general settings, database settings
01:35:44iambenyou should probably turn autoupdate on
01:36:02iambenand also update it now
01:36:24 Join MournBlade [0] (i=4401597c@gateway/web/cgi-irc/
01:36:32iambenit does it in the background, i dont really know how long it takes tbh
01:37:32helpme401how do you get out of search
01:38:10 Quit relaxed ("")
01:38:34iambenwhat search?
01:38:38helpme401i got it
01:38:45helpme401thank you for your help
01:43:29***Alert Mode OFF
01:43:39 Join [5]robin0800 [0] (
01:43:39***Alert Mode level 1
01:43:39***Alert Mode level 2
01:43:39DBUGEnqueued KICK [5]robin0800
01:43:39***Alert Mode level 3
01:43:39***Alert Mode level 4
01:43:39***Alert Mode level 5
01:43:39***Alert Mode level 6
01:48:54helpme401Is there a way to make the Database directory the first selection on the list?
01:50:07 Quit [1]robin0800 (Read error: 110 (Connection timed out))
01:51:15 Quit [3]robin0800 (Read error: 110 (Connection timed out))
01:52:08 Quit robin0800 (Read error: 110 (Connection timed out))
01:53:23 Quit [2]robin0800 (Read error: 110 (Connection timed out))
01:53:40***Alert Mode OFF
01:55:32Soul-Slayerhelpme401: If you change it at source and recompile, yes. But there isn't a feature in Rockbox to do it
01:55:45pixelmayou can't reorder the menu but you could set "database" as startup screen
01:56:02helpme401oh cool thanks
01:56:07helpme401this program roxors
01:56:16 Quit robin_0800 (Read error: 110 (Connection timed out))
01:58:52helpme401anyone know how to increase the font size?
02:00:54 Join jhulst [0] (
02:02:22 Quit Soul-Slayer (Read error: 104 (Connection reset by peer))
02:05:52 Quit [4]robin0800 (Read error: 110 (Connection timed out))
02:07:08 Quit [5]robin0800 (Read error: 110 (Connection timed out))
02:08:40 Join kubiixaka [0] (
02:21:29 Quit jhulst (Read error: 110 (Connection timed out))
02:25:31 Quit kubiix (Read error: 110 (Connection timed out))
02:34:21 Join toffe82 [0] (n=chatzill@
02:37:29 Part pixelma
02:38:22 Quit kkurbjun (Read error: 104 (Connection reset by peer))
02:44:42 Join PenguinMasta [0] (i=47cde811@gateway/web/cgi-irc/
02:45:59PenguinMastahello all
02:47:03PenguinMastahow can i tell what Gen my ipod nano is
02:47:33iambeni've only held an ipod in my hands once in my life
02:47:39PenguinMastai was looking on the wiki but didn't find anything... then googled it and didn't see anything :(
02:47:57PenguinMastai want to rockbox but noticed that one gen 1 was supported
02:49:46 Quit spiorf ("Read error: 110 (Connection timed out)")
02:50:40LloreanPenguinMasta: Is the front White or Black, with a silver back?
02:51:52PenguinMastait's white with green
02:52:08iambeni dunno anything about ipods but i can google like a sumbitch:
02:52:10LloreanIt's a 2nd generation.
02:52:59PenguinMastadamn i must of not worded right
02:53:46PenguinMastai tried this "what gen is my ipod nano"
02:55:17PenguinMastathanks for the help :)
02:55:26iambeni try
03:05:13 Quit PenguinMasta ("CGI:IRC (EOF)")
03:08:56***Saving seen data "./dancer.seen"
03:11:38 Join saratoga [0] (i=980398fe@gateway/web/cgi-irc/
03:12:11 Quit helpme401 ("CGI:IRC (EOF)")
03:13:32 Join jhulst [0] (
03:14:25saratogapreglow: do you know what the hold up on FS #6705 (MPA optimization for ARM) was?
03:17:02saratogaI'd be more then willing to take a look at it if someone could fill me in on what needs to be done
03:17:08 Join wishes [0] (n=wishes@
03:18:35 Join aliask [0] (
03:19:27 Join Rob2222 [0] (
03:19:56Lloreansaratoga: If I recall, it broke mpegplayer because of the way it used IRAM? (Or maybe I'm thinking of a different patch)
03:20:14 Join Daishi [0] (
03:21:50saratogaLlorean: thats what I thought, but looking at the comments, linuxstb mentioned getting mpegplayer working back in may
03:22:36saratogasince it seems to considerably reduce the boost ratio on the PP targets, I'd be interested to see what it does for battery life
03:27:57 Join krylon [0] (
03:28:15krylonJust made a post about my ipod that repeatedly gets an undefined instruction lock up :(
03:32:29saratogakrylon: this sound like it :
03:34:53krylonat least im not alone. i was using up tissues very quickly. from reading all of those replies, there is no fix, or am i mistaken?
03:35:37krylonwhere do i download an updated bootloader? only way i can find a download is in the manual that i posted a link to
03:35:42saratogakrylon: i have no idea, this is the first i've heard of it
03:35:54saratogaupdating a bootloader is basically just repeating the rockbox install procedure
03:35:59krylonive tried searching the forums each time i see a new error code, but i come up with nothing
03:36:24saratogayeah the error code is an address, so it'll change slightly for each compile
03:37:54 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
03:38:53 Quit Thundercloud_ (Read error: 104 (Connection reset by peer))
03:39:34saratogaLlorean: for what its worth, the MPA patch and mpegplayer work fine for me
03:40:22TiMiD[FD]around how much percent of changes should I consider changing the file copyright ?
03:42:45krazykitTiMiD[FD], sorry, what?
03:44:24TiMiD[FD]there are copyrights in rockbox source files right ?
03:44:56TiMiD[FD]but if I change the content of one file so that 95% of the code is mine, not the original
03:45:06TiMiD[FD]what do I do ?
03:45:12krazykitliscense it gpl
03:45:26TiMiD[FD]I know
03:45:35krazykitrockbox is liscensed as gpl. any derivative works must be liscensed gpl
03:46:19TiMiD[FD]I'm not talking about the liscence though
03:46:35hcslicense, by the way
03:46:59TiMiD[FD]I'm talking about copyright
03:47:05hcsare you adding code and wondering if it is substantial enough to add your name to the copyright notice?
03:47:15saratogahe means the copyright date on the GPL notice
03:47:21saratogawhen should that be updated
03:47:24TiMiD[FD]I'm changing code completely
03:47:25 Quit Seed (Nick collision from services.)
03:47:33 Join Seed [0] (
03:47:40TiMiD[FD]the only thing remaining is ... some inline pictures
03:48:00 Join kubiix [0] (
03:50:12 Join safetydan [0] (i=3b6408cc@rockbox/developer/safetydan)
03:50:53 Quit saratoga ("CGI:IRC (EOF)")
03:51:25safetydanTiMiD[FD]: you could always do "Copyright <your name> based on work by <Previous name>"
03:51:59 Join Soader03 [0] (i=18caf221@gateway/web/cgi-irc/
03:52:41TiMiD[FD]I think it would be nice to have an history in the files
03:53:05TiMiD[FD]people who made significant changes to the code could add their names
03:54:49 Quit Soader03 (Client Quit)
03:55:45 Join Soader03 [0] (
03:55:55safetydanTiMiD[FD]: you could always do "Copyright <your name> based on work by <Previous name>"
03:58:32 Part Soader03
04:00:35 Join relaxed [0] (n=relaxed@unaffiliated/relaxed)
04:01:45 Quit atsea- (Read error: 104 (Connection reset by peer))
04:03:04 Quit kubiixaka (Read error: 110 (Connection timed out))
04:03:44 Quit safetydan ("CGI:IRC (Ping timeout)")
04:19:00 Join kubiixaka [0] (
04:25:51 Join atsea- [0] (i=ariel@gateway/tor/x-31889521755e0562)
04:25:52 Part toffe82
04:27:24 Join ptw419 [0] (
04:28:11 Join Caliban_ [0] (
04:28:26 Join Bossier [0] (
04:28:38 Nick Bossier is now known as tdtooke (
04:30:46tdtookequick question: I just noticed that a scrollwheel acceleration change just got commited for the sansa e200. I'm wanting to use this for my iPod so will I be able to use the same WHEEL_ACCELERATION_FACTOR he has defined for the e200, or would I need to figure out what it should be for the iPod?
04:41:02 Quit Snake__ (Remote closed the connection)
04:41:14 Quit kubiix (Read error: 110 (Connection timed out))
04:43:58 Join toffe82 [0] (
04:44:03 Quit Caliban (Read error: 110 (Connection timed out))
04:44:53krazykittdtooke, i recall the commit comment saying it was a good general accel for any scrollwheel target
04:49:22 Quit jhulst (Read error: 110 (Connection timed out))
04:51:49tdtookeOk, I think what I'll do is just do a diff adding those defines to config-ipodwhatever.h and throw that up on the ipod scrollacceleration patch page and see what happens.
04:52:24LloreanWhy add it to a patch entry that's a different method?
04:52:31LloreanAnd why not make sure it works first?
04:52:43tdtookeI was going to do that first, just didn't say it here
04:53:00tdtookeIf this works well it'll make that patch not necessary
04:53:10LloreanThen that task can be closed.
04:53:13iambendid those e200 accel changes actually enable scroll accel, or did it just lay the framework to make it possible?
04:53:22iambenforgive me for asking a "user" question =)
04:53:24LloreanBut please don't add .diffs to a flyspray entry that aren't the same patch.
04:53:47tdtookeOk, just thought the guys that prowl that page would definitely be interested
04:53:58LloreanThey might be, but that's a Rockbox page, not their page.
04:54:04LloreanAnd we like the tracker to stay organized.
04:54:45LloreanI'm pretty sure that the e200 acceleration makes use of the way the wheel driver is written, and would require changes to the iPod wheel driver for it to work the same.
04:58:11tdtookeHmm.. on the ipod acceleration patch they just basically change the offset parameter select_at_offset uses based on whatever acceleration factor you have set
05:01:14tdtookeI think you're right on having to modify the driver, I see that he changed that on the e200 so that means I'd have to do something similar for the iPod and that's definitely out of my league at this time.
05:02:09 Nick lostnihilist_ is now known as lostnihilist (
05:03:12iambenso the e200's accel is already implemented, thanks guys!
05:08:59***Saving seen data "./dancer.seen"
05:11:34 Join RaRe [0] (
05:17:18 Join mike__ [0] (n=mike@
05:24:25 Nick tdtooke is now known as Bossier (
05:24:28 Quit Bossier ("( :: NoNameScript 4.02 :: )")
05:31:52jhMikeSthe acceleration just sets something up so it can be used but e200 is the only wheel driver that outputs the right data for it atm. sure ipods and perhaps H10 could utilize it but the scroll drivers need modifying for it.
05:32:18 Quit BigMac (Read error: 110 (Connection timed out))
05:41:32 Part Llorean
05:48:05 Join saratoga [0] (i=41becb3b@gateway/web/cgi-irc/
05:54:31 Quit rotator ()
06:17:39 Quit Nibbier (Read error: 110 (Connection timed out))
06:18:05 Join Nibbier [0] (
06:21:28 Join TheCreationist [0] (
06:21:51TheCreationistCould someone help me figure out a glitch I'm having with Rockbox on a Sansa e260?
06:22:55 Quit TheCreationist (Remote closed the connection)
06:25:14 Quit saratoga ("CGI:IRC (EOF)")
06:26:25 Quit Daishi ("Client exiting...")
06:26:30 Quit lazka (Remote closed the connection)
06:32:40 Part MournBlade
06:41:00 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
07:08:05 Quit Siltaar ("CGI:IRC (Ping timeout)")
07:09:00***Saving seen data "./dancer.seen"
07:12:52 Quit hcs ("Leaving.")
08:06:07 Join entheh [0] (
08:06:08 Part toffe82
08:07:46 Join pmhaddad [0] (
08:07:50 Part pmhaddad
08:20:03 Quit goffa (Read error: 113 (No route to host))
08:32:56 Quit idnar (Nick collision from services.)
08:32:58 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
08:44:58GodEaterdamn - no dan_a tonight
08:45:21 Join Rob222241 [0] (
08:49:04 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
08:51:14 Quit aliask ("ChatZilla [Firefox]")
08:52:30jhMikeSJdGordon: care to do tester really quick? I can't make a panic now no matter what I do. I suspect that delay might not be needed now either.
08:52:54JdGordonyeah ok
08:56:05 Join kaaloo [0] (
08:56:17 Part kaaloo
08:56:47jhMikeSI hope it doesn't need that sleep anymore
08:58:48JdGordonjust got a panic
08:58:59JdGordonnot 100% sure i roloed to it though
08:59:22JdGordonyep, panic 14
08:59:40jhMikeSduring insert?
09:00:24JdGordonno, after instert, going to files from the main menu
09:00:58JdGordonvery reproducable
09:01:15JdGordonthere is no disk activity between the insert and the going into files
09:02:01 Quit Rob2222 (Read error: 110 (Connection timed out))
09:09:02***Saving seen data "./dancer.seen"
09:09:05jhMikeSyesterday it was just during swaps?
09:10:06JdGordoncant rewmember how i tested yesterday, but yeah probably
09:10:20JdGordoni tihnk i stayed in the browser the whole time
09:10:26JdGordonwhereas now i was changing in the menu
09:10:38jhMikeSand now after booting, just going to "File" does it?
09:11:33jhMikeSWhat about "View Disk Info"?
09:12:16JdGordonhmm... wait a sec, i tihnk the wrong version is being loaded
09:13:23JdGordonok, definatly running your version now
09:13:27JdGordonand still oanic
09:14:04jhMikeSwhich version _were_ you running?
09:14:08JdGordonhmm... panic on removal
09:14:17JdGordonmy dodgey one from yesterday
09:14:34jhMikeSpanic on removal and browse?
09:14:45JdGordonthat last panic was when i was in the tree, then inserted and it didnt come uo, then remove, then panic
09:15:51CtcpIgnored 7 channel CTCP requests in 13 minutes and 4 seconds at the last flood
09:15:51*JdGordon back in a while
09:16:12jhMikeSnot too long...I just need to see if it's the delay removal? :)
09:16:19 Join TTSbox_ [0] (
09:21:55 Join linuxstb [0] (i=5343d4aa@rockbox/developer/linuxstb)
09:24:51 Join davina [0] (
09:25:31 Quit entheh ("^~")
09:30:39 Join lost|X40 [0] (
09:40:02 Quit lostnihilist (Read error: 110 (Connection timed out))
09:46:48 Join Nick_Brackley [0] (i=7dff0332@gateway/web/cgi-irc/
09:50:19 Quit TTSbox__ (Read error: 110 (Connection timed out))
09:58:11JdGordonjhMikeS: got another mi4 to test?
09:58:22daurnimatorhey JdGordon
10:00:08jhMikeSJdGordon: In a second. All panics disabled too.
10:03:01jhMikeSstill no delay but I wonder what will happen
10:05:36JdGordonjhMikeS: yeah, seems to be good
10:06:22jhMikeSit's probably sending multiple interrupts and ripping the device out while it's trying to read but it should actually be trying multiple times now
10:07:10 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
10:08:50linuxstbjhMikeS: Is there a #define to indicate the available playback samplerates for the different devices?
10:09:10jhMikeS#define HW_SAMPR_MASK in the config
10:09:39jhMikeSHW_SAMPR_CAPS rather
10:10:12jhMikeSsee pcm_sampr.h and pcm_sampr.c as well for what gets #defined as a result of it
10:10:34linuxstbOK, thanks.
10:12:46jhMikeSJdGordon: then this panicing thing just has to go as it's too much trouble and not a reliable measure of failure.
10:13:57jhMikeSthings really need proper mutexing too
10:18:49 Join pondlife [0] (
10:20:35 Join Siltaar [0] (i=c134d0e9@gateway/web/cgi-irc/
10:26:31 Join obo_ [0] (n=obo@rockbox/developer/obo)
10:30:57 Nick obo_ is now known as obo (n=obo@rockbox/developer/obo)
10:32:03 Quit linuxstb ("CGI:IRC")
10:42:16 Part maffe
11:03:10JdGordonjhMikeS: panic wouldnt be so bad if it wasnt such a pain to reset it
11:06:53 Quit bluebrother (Read error: 113 (No route to host))
11:09:05***Saving seen data "./dancer.seen"
11:11:03jhMikeSJdGordon: and if it actually meant anything important :-) lots of stuff can occur during insert which basically means it should just process the next message and try again.
11:21:43jhMikeSthe problem is that the card gets detected long before it stops moving in the slot. the controller will even send interrupts with an unchanged line level which I'm guessing means the card should be reinitialized.
11:21:48 Join desowin [0] (
11:31:18 Join ganny [0] (i=9088689c@gateway/web/cgi-irc/
11:31:47 Join Entasis [0] (
11:33:25 Join maffe [0] (
11:33:58gannyAnyone here have thougts about the sound quality with Rockbox on the Sansa E200 series?
11:34:01amiconnjhMikeS: I don't know how the sansa handles it, but on the Ondio the mmc detection is debounced and delayed about 30 ticks
11:36:22jhMikeSwe're using interrupts instead of wasting CPU with tick tasks. the hardware controller seems to inform it of what needs doing.
11:37:06amiconnYes, and then you run into problems because the card is still moving...
11:37:48jhMikeSso a mount attempt might fail but then the controller just says "try again" and that seems to work just fine
11:43:35 Join spiorf [0] (
11:45:59 Quit ganny ("CGI:IRC")
12:00:51 Part maffe
12:01:00 Join maffe [0] (
12:04:13 Part maffe
12:11:33 Join maffe [0] (
12:13:41 Join bluebrother [0] (
12:13:43 Join thomasg_ [0] (
12:15:00 Join linuxstb [0] (i=d556da1b@rockbox/developer/linuxstb)
12:26:58 Quit desowin (Read error: 104 (Connection reset by peer))
12:29:15 Join desowin [0] (
12:31:05 Quit thomasg (Read error: 110 (Connection timed out))
12:32:13 Part maffe
12:32:36 Join maffe [0] (
12:32:38 Quit maffe (Remote closed the connection)
12:32:51 Join maffe [0] (
12:34:58 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
12:36:49 Join gast42 [0] (i=864c039a@gateway/web/cgi-irc/
12:37:46 Join wrobbie [0] (
12:38:07 Join borisyeltsin [0] (
12:38:21 Join Rob2222 [0] (
12:38:32 Join RaRe` [0] (
12:38:51 Quit billytwowilly (Read error: 104 (Connection reset by peer))
12:40:52 Join moos [0] (
12:47:24gast42hi, i just downloaded the bleeding edge build for H300 and the version in rockbox-info.txt is "r13969M-070724". A while ago someone said that the "M" means that the source was modified. how can that be with the official build?
12:48:23linuxstbIt's normal, but I can't remember the reason why it happens.
12:49:12GodEaterM is for Magic ;)
12:49:42linuxstbor Magners...
12:49:50*linuxstb is thirsty
12:50:56 Join jgarvey [0] (
12:51:00 Quit Nick_Brackley ("CGI:IRC (EOF)")
12:51:10GodEaterlinuxstb: saving yourself for the pub this evening I hope
12:51:24 Join Nick_Brackley [0] (i=7dff0332@gateway/web/cgi-irc/
12:51:26pixelmaI believe it's because of some changes to solve dash/bash issue on the ubuntu servers
12:51:45GodEaterpixelma: that sounds extremely plausible
12:53:04gast42so no magic build :(
12:53:08 Quit RaRe (Read error: 110 (Connection timed out))
12:53:44 Join RaRe [0] (
12:57:07amiconnM is modified, and iirc the ubuntu build problems were solved in svn a while ago
12:57:27amiconnSo build servers delivering modified builds need to be checked and fixed
12:57:45 Quit Rob222241 (Read error: 113 (No route to host))
13:01:52linuxstbCan you remember exactly what was happening?
13:02:07 Quit RaRe` (Read error: 110 (Connection timed out))
13:02:10amiconnIt was a dash vs. bash syntax problem iirc
13:02:33GodEaterdoes rockbox-info.txt tell you which build server generated the .zip ?
13:02:35linuxstbYes, but why does that cause svn to report the code as modified?
13:02:52amiconnlinuxstb: Hmm, wasn't it you who pointed out the solution (using printf)?
13:03:14amiconnErm, because the build script are also in svn?
13:03:16 Quit Nibbier ("bei uns hat noch niemals nicht koana koan durscht ned ghabt")
13:03:19linuxstbYes, but I thought the problem I solved was just that manual builds were failing.
13:03:28 Join Nibbier [0] (
13:03:44 Join voltagex [0] (
13:03:49linuxstbAh, so the build server admins needed to modify the build scripts?
13:04:15 Part maffe
13:04:24*amiconn thinks that build servers delivering modified builds should be excluded from the build process, after retrieving 'svn diff' output for investigation
13:04:47*linuxstb checks what his server is doing
13:05:15voltagexI'm trying to build RBUtil on Mac OS X
13:05:44voltagexand it's failing in the last steps: lipo: can't figure out the architecture type of: /var/tmp//cc507TKy.out
13:05:47 Join maffe [0] (
13:05:48pixelmaGodEater: the H300 build was made by debussy...
13:06:26*GodEater doesn't think his own build server should be a problem - it's never even seen dash
13:06:34linuxstbvoltagex: Did you build wx with universal binary support?
13:07:08linuxstbAnd I assume you're using (modified for your own paths) the Makefile.macosx file?
13:07:31voltagexyes I modified the makefile
13:07:39voltagexchecking what I did with wx now
13:09:08***Saving seen data "./dancer.seen"
13:09:46voltagexyes, it was universal
13:13:44linuxstbThen I'm not sure what the issue is - I don't recall ever getting that error message.
13:16:22voltagexI think I used a newer version of wx
13:16:30voltagexyep...2.8.4 instead of 2.8.0
13:17:00GodEaterI'd be surprised if it was that causing the problem
13:17:21pondlifeHmm, what determines a "modified" build? Even if I use straight SVN source, I get an M in there...
13:19:47amiconnsvnversion −−help confirms that 'M' means modified
13:20:11pixelmapondlife: do you use tortoisesvn - or something that needs different line endings, maybe?
13:20:19pondlifeCygwin svn
13:20:26pondlifesvn diff gives no output.
13:20:38bluebrothersvn should handle line endings itself
13:20:42amiconnMy own local builds are always marked 'M' because I have some changes in config-*.h files
13:21:09bluebrotherdid you try svn revert? Maybe that gives a clue ...
13:21:40pondlifeNo, I don't think I've ever used svn revert. I tend to patch/unpatch my way around.
13:21:53bluebrothermaybe it's a timestamp issue?
13:22:12*pondlife rebuilds, just in case.
13:22:30amiconnjust issue 'svnversion' in your svn root
13:22:34bluebrotherI could imagine that reverting a patch gives a different timestamp than reverting
13:22:36voltagexI'll have a look at rbutil again in the morning
13:22:47amiconn^ quick check whether the source is modified or not
13:23:54pondlifeHmm, will delete and re-svn up I think.
13:24:15amiconnAnd 'svn diff' tells you nothing?
13:24:27amiconnMaybe some files' properties are different
13:24:37bluebrotherdoes svn revert . -R tell something?
13:25:43pondlifeLine ending differences would show up, right?
13:26:12bluebrotherafaik they should.
13:26:14linuxstbvoltagex: Maybe the Makefile needs tweaking for newer wx versions - I'm still using 2.8.0 on my Mac.
13:26:39voltagexlinuxstb: thanks for that, gotta go, bye!
13:26:41 Quit voltagex ("leaving")
13:26:56amiconnIn linux I get no M for a clean working copy
13:27:21*linuxstb wonders what "mixed revision" means - I get "13967:13969" for my checkout...
13:27:42bluebrotheryou have files for each version
13:27:55linuxstbAn svn update fixed it...
13:28:00bluebrotherI got that a couple of times when I ran svn up in the manual folder and forgot the rest.
13:28:01amiconnlinuxstb: This is because your working copy is at 13967, while the main repo already is at 13969
13:28:27Nick_Brackleyi just build a clean version and got r13969-070724
13:28:33linuxstbMust be something to do with me having committed the last two changes without doing an update...
13:28:41pondlifeAnyone else using Cygwin?
13:28:43amiconnThen the versions are technically mixed, because some (i.e. most) files didn't change between 13967 and 13969
13:29:01amiconnpondlife: Yes, me
13:29:08pixelmame too
13:29:22amiconnBut I always have changes against svn in my main copy
13:29:56pondlifeOK, I have a clean copy and svnversion no longer gives me the M.
13:30:08pondlifeNow for some binary file comparisons.
13:30:49amiconnWhen binary files differ, 'svn diff' will show them (just printing "binary file versions differ")
13:31:18pondlifeI mean comparing text files for line-end differences.
13:31:50pondlifeBTW is the iPod 4G build supposed to be stable at the moment? (re:
13:31:56 Quit relaxed ("")
13:32:40amiconneurgh - so it actually has the same problem as the H10
13:32:43pondlifeOr is the 30Gb actually a 5G...?
13:33:14amiconnThis smells like PP5020...
13:33:51*pondlife gets a little sick of duplicate PortalPlayer failure reports...
13:36:03 Quit gast42 ("CGI:IRC (Ping timeout)")
13:36:43pondlifeHmm, arm-wresting.. guess I won't get a T-shirt then.
13:36:50pondlifenor mug
13:37:04 Join PaulJam [0] (
13:39:32*pondlife lols at LinusN's "...sadly unusable"
13:40:41GodEaterif it's any consolation, I lost (or rather gave up) at the last arm wrestle I had....
13:40:51GodEaterI'd rather play pool for them - but there's no pool table at the pub
13:41:56pondlifeamiconn: While you're collecting bug reports ;) .. did you also see the Nano report on the dev ML?
13:45:27amiconnNo. Just checked mail...
13:46:04amiconnNo bugreport appeared
13:46:20 Quit linuxstb ("CGI:IRC (Ping timeout)")
13:46:29 Join BHSPitMonkey_ [0] (
13:53:20 Join unknown [0] (
13:55:02 Quit BHSPitMonkey (Read error: 104 (Connection reset by peer))
13:55:58unknowndoes the sd driver for the e200 target support 2gb sd cards?
13:56:46bluebrotherif it's not SDHC then yes.
13:56:57amiconn2GB shouldn't be SDHC
13:57:00bluebrotherhigh capacity cards.
13:57:16bluebrotherit shouldn't be, but I don't know if there are SDHC cards around with 2GB
13:57:25amiconn(unless the boundaries are different than for MMC)
13:59:01 Quit PaulJam (Read error: 113 (No route to host))
13:59:13amiconnThe mailing lists are slow...
13:59:26 Join Thundercloud [0] (
14:00:50amiconnAnyone around with a nano who could make a dump of the original firmware?
14:00:54GodEaterBagder took the mail server on holiday with him ;)
14:01:25GodEateramiconn: I assume you mean the content of flash ?
14:01:37GodEaterjust the 1st partition ?
14:01:49amiconnI mean the original firmware binary
14:02:15*GodEater is lost
14:02:32unknowni can't find the technical specs for the e200
14:02:37amiconnNot the whole partition, just the firmware binary
14:02:57amiconnipodpatcher -rb firmware.bin iirc
14:03:03scorcheamiconn: hold a sec
14:03:23amiconn(unless the original firmware is completely removed) took 10 minutes to redistribute the mail
14:06:19scorche-rfb it seems
14:07:46 Quit maffe (Remote closed the connection)
14:07:59scorcheamiconn: shall i try DCC?
14:08:21unknownok, i found that it should support 2gb microsd cards
14:08:26unknownso i'm off to buy one :)
14:08:32 Part unknown
14:09:22 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
14:11:46pixelmapondlife: I had the same (svnversion giving me the "M"odified) in an old tree - the diff showed nothing but I've found a few ".mine" and other files from conflicting updates. Removing those and issueing an "svn resolved" helped
14:12:38amiconnAh, yes, -rfb sounds better
14:12:55amiconnscorche: Try it - receiving should work at my end
14:13:17amiconn(sending doesn't atm, because of a bug in Hydra 0.3.160 - I should downgrade until fixed)
14:13:36pixelmaI thought that was the different way around...
14:13:51scorche..i am using 0.3.160 and it seems to be sending just fine...
14:14:02amiconnAre you behind a nat router?
14:15:17scorcheshould be
14:15:19pondlifepixelma: I bet that's it.. I have plenty of other files there (.origs and ".mines")
14:15:24 Join linuxstb [0] (i=d556da1b@rockbox/developer/linuxstb)
14:15:39amiconnre linuxstb
14:16:06amiconnlinuxstb: Do you remember the (slightly) colourised sudoku numbers?
14:16:23amiconnI wonder whether they should be commited or ditched
14:17:10scorcheamiconn: would you like me to checksum it just to be sure?
14:17:45amiconnMight be a good idea
14:18:01linuxstbamiconn: I don't think I was a big fan of them. I'm used to Sudoku being black and white..
14:18:25 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
14:18:59amiconnbd5564b769645af0fae76269c667bf00 */cygdrive/d/#Daten/RB/apple_os_nano/nanofirmware.bin
14:19:00wrobbiepondlife: i'm here too (nano problems). might be faster :)
14:19:52pondlifewrobbie: I don't have a Nano myself, just wanted to make sure you and amiconn got together ;)
14:20:01scorcheamiconn: yup...same here
14:20:37pondlifeanyways... lunchtime.
14:24:06 Join DerPapst [0] (
14:25:51DerPapstamiconn: You can download all current firmwares from here and extract the firmware with ipod_fw for all non 5.5G iPods. This will result in the same that ipodpatcher -rfb does. but you wont have to wait for someone to dump the firmware for you ;)
14:26:04JdGordonamiconn: is there any way we can pass the bootloader version to rockbox? really getting sick of the comments "have you got the newest bootloader"
14:26:45 Quit jgarvey ("Leaving")
14:27:18 Join jgarvey [0] (
14:30:16 Join maffe [0] (
14:36:48amiconnwrobbie: Could you check Info->Debug->View HW info and tell me the PP version?
14:38:32 Join relaxed [0] (n=relaxed@unaffiliated/relaxed)
14:38:48 Quit hcs ("Leaving.")
14:45:14wrobbieamiconn: Sorry, didn't see your message. PP version PP5022C
14:45:25DerPapstIs there somewhere a chart which iPod have which PP chips inside? i know that 1G-2G have PP5002 and 4G b&w, Mini 1G have PP5020
14:45:45amiconnI just wanted to make sure there aren't some nanos with PP5020
14:47:03amiconnDerPapst: G1..G3 have PP5002, G4 (bw & color/photo), Mini G1 have PP5020), Nano and G5/G5.5 have PP5021 (displayed as PP5022C in software), and Mini G2 has PP5022 (displayed as PP5022B in software)
14:47:53jhMikeSamiconn: I find it a little odd that caches are invalidated in the code that is loaded (in crt0-pp.S) and not by the code that did the loading before actually branching into it iiuc.
14:48:00wrobbieamiconn: you have no problems on your nano?
14:48:12DerPapstwasn't sure whether the 4G photo/Color is PP5020 for sure.
14:48:16amiconnIriver H10 has PP5020, Sansa E2x0 has PP5024 (which is a PP5022 with tacked-on AS3514, and is displayed as PP5022 in sw)
14:48:23GodEaterwrobbie: don't think amiconn owns a nano
14:49:57pixelmaDerPapst: ;)
14:50:56*amiconn can't find anything special in the nano's clock setup routine
14:51:06DerPapstpixelma: i'm currently editing that page to fix some issues :P
14:51:12jhMikeSperhaps I'm just not following this correctly right now :\
14:51:21amiconnIt's the same as the G4 color uses, and the only difference to the G5 routine is the lack of PP5026 handling
14:52:22 Join kaaloo [0] (
14:52:24DerPapstpixelma: because the CPU row for Color iPods says "Portal Player PP5020. (chip revision may differ)" and i'm unsure that is meant with "chip revision may differ"
14:52:27 Part kaaloo
14:52:59amiconnThe revision is the letter after the PP502x
14:53:18amiconnSO far I've only seen PP5020E, PP5022B and PP5022C
14:53:18DerPapstAh ok.
14:53:52wrobbieamiconn: I can try tweaks and patches if you know of any :) I could also try moving to HEAD and backing out those two revs, but I don't know if they have further dependencies.
14:53:53DerPapstPP5020E is used in 1G mini 4G (b&w/Photo/Color)?
14:54:14DerPapsts/1G mini/mini 1G,
14:54:16amiconnI don't know whether it's all of them
14:54:37amiconnI just know that the Mini G1 I had for testing and my H10 6GB have PP5020E
14:55:06DerPapstthanks for the info :)
15:09:09***Saving seen data "./dancer.seen"
15:11:36JdGordonzzzzzzzzzzz...... boooorred
15:12:35linuxstbJdGordon: How's the buffering/playback stuff coming along?
15:12:50JdGordonstalled for the moment
15:13:13Nico_PJdGordon: no progress on merge.patch ?
15:13:28*Nico_P hasn't taken the time to take a good look at it
15:13:37Nico_Pand no progress on the rest either
15:13:40 Join Jens [0] (i=Jens@pdpc/supporter/active/Jens)
15:13:52Nico_Ptomorrow I'll be bakc home so hoepfully I'll have some time to wrk on it
15:16:17*amiconn spots some typing accomodation ;)
15:16:46*jhMikeS is too tired to say that there's no free buffering lunch
15:17:00 Quit Siltaar ("CGI:IRC (Ping timeout)")
15:20:46jhMikeSJdGordon: no further sansa panics?
15:21:12JdGordonhaccnt been using my sansa since that testing...
15:21:14jhMikeSof course not, but you know what I mean
15:21:24JdGordoncommit the fix :)
15:21:39JdGordonI need to update my bootloader again so i can boot from sd
15:22:06jhMikeSjust wanted to know if you performed the swap abuse protocols
15:23:21JdGordonsoon ill try again... acidently deleted the whole .rockbox folder without backing up my config and themes
15:23:28jhMikeSI can push the card in (but not enough to lock it in so that it springs back out) as fast as I can do it and there's no problems
15:23:30JdGordonreloading it all now so i have music tomorow
15:23:54 Quit jgarvey ("Leaving")
15:24:37*jhMikeS will introduce an occasional write corruption bug as a bonus just to frustrate and require reformatting :p
15:25:42JdGordoni tihnk both my cards are already full of corruption
15:27:21 Join thegeek [0] (
15:27:25JdGordonok nice.... i have almost worse than a panic...
15:27:29JdGordonan empty filebrowser
15:28:00*JdGordon slaps self... view mode was supported :p
15:28:09JdGordonno it wasnt....
15:28:13*JdGordon unslaps self
15:28:35JdGordoni have a very empty root folder~
15:28:39 Quit crwl (Read error: 110 (Connection timed out))
15:29:07jhMikeSnot my fault
15:29:25*JdGordon isnt so sure....
15:30:52jhMikeShuh? no way would it goof sd0
15:31:14JdGordonwierd, i dunno, reboot fixed it
15:34:23jhMikeSI think that happened to be once but not with patched code
15:35:52 Join unknown [0] (
15:36:56unknownwhere can i see the free discspace?
15:36:59 Quit mike__ (Remote closed the connection)
15:39:52pondlifeNico_P: You seem to have done some Rockboxing whilst on holiday :)
15:41:14 Quit ptw419 ()
15:41:47unknownok, found it :)
15:41:56unknownthe sd card stuff sometimes gives an error
15:42:01unknownbut works otherwise
15:43:12jhMikeSlinuxstb: I really intended HW_HAVE_xx/HW_HAVE_xx_() to be used and not bitmask testing outside of the header.
15:43:41 Quit thegeek_ (Read error: 110 (Connection timed out))
15:43:55Nico_Ppondlife: yes a bit, but I didn't have enough time to do some big work
15:44:07pondlifeAn understanding gf, then ...
15:44:13Nico_Pyes :)
15:44:28pondlifeI wasn't pestering you to finish MoB or anything, honest.
15:44:49*pondlife doesn't intend that to sound sarcastic
15:44:58 Part maffe
15:45:02JdGordoncourse not
15:45:10 Join maffe [0] (
15:45:18Nico_Phehe :)
15:45:32Nico_Ppestering me is Llorean's role ;)
15:46:23 Join Arathis [0] (
15:47:05Nico_PI'm still a bit unclear on how I'm going to make things work though... the whole point of RB in RB was to avoid having to move the metadata around and it turns out it's not helping much
15:47:29pondlifeWell, always KISS your code.
15:47:32Nico_Pso I'm wondering whether I should go back to before I implemented that or still use it
15:48:02pondlifeIf it doesn't help your strategy, remove it.
15:48:24Nico_PI have to do some serious thinking about all this
15:48:46*amiconn always has to think twice about "rb in rb" (rockbox in rockbox?)
15:49:01Nico_Palso about how to handle the metadata... completely separate handles or else ?
15:49:12JdGordonseperate handles
15:49:35pondlifeWhen do you need to keep metadata handy, but not audio (or vice versa)?
15:50:03amiconnE.g. if a track is longer than the buffer, and we want to support next track info all the time
15:50:23amiconn(a feature that could be made ramsize dependent)
15:50:33pondlifeThere are probably more use cases - maybe make a list?
15:51:15*JdGordon slaps pondlife
15:51:22JdGordoni finished with bloody use cases last semster
15:51:25JdGordonnever again :'(
15:51:35pondlifeYou can never have enough use cases..
15:51:56 Part unknown
15:52:34JdGordonalrighty, bed time, ttyl
15:52:42 Quit JdGordon ("Konversation terminated!")
15:54:01Nico_Ppondlife: you scared him off :p
15:54:18linuxstbjhMikeS: Why not test the bitmask? (seems like one abstraction too many...)
15:54:46Nico_Paudio needs metadata, not the opposite... so the need for metadata disappears at the same time as audio
15:54:58linuxstbjhMikeS: But if you want to change it, please do.
15:55:08Nico_Pfor the next track we need the metdadata before the actual audio
15:56:34amiconnlinuxstb: Hmm, maybe sudoku could have an option to select either mono or coloured numbers?
15:57:09pondlifeSo we're only likely to have the next track info fragmenting the buffer at any one time?
15:57:11amiconn(does it have any option atm?)
15:57:27linuxstbThere's a menu, but I can't recall any options..
15:59:00 Quit Nick_Brackley ("CGI:IRC (EOF)")
16:00:00pondlifeNico_P: ^^^ ?
16:00:11jhMikeSlinuxstb: no, it's just meant as a shortcut really. the other HW_HAVE_xx_() is also just cleaner for defining arrays.
16:00:27 Quit Entasis ("Leaving")
16:00:33 Quit moos ("Glory to Rockbox")
16:00:41Nico_Ppondlife: what was suggested was to use a static buffer to have only the essential info about the next track
16:00:51Nico_PI think that's quite KISS
16:00:55pondlifeThat's what I was thinking.
16:01:06pondlifeNo need for "next track album art" is there?
16:01:20Nico_PI don't think so
16:01:40*jhMikeS just likes to hide all the complication way back in things so it's easy to be lazy later :)
16:01:43amiconnNext track album art is a contradiction in itself...
16:02:15pondlifeUnless you have "next folder" enabled.
16:02:16bluebrotherbut users will scream for support for it :(
16:02:57pondlifeSo just hang onto the next track metadata as supported by current WPS tags - not too big a block is it?
16:03:14pondlifeThen, the buffer can just roll around freely.
16:04:00amiconnI'm still not convinced it's a good idea to put current & next track metadata into the main buffer
16:04:24Nico_Pamiconn: you mean the ID3 data or ID3+AA+cue ?
16:04:44amiconnI mean all metadata
16:05:51amiconnImho all the various suggestions sound too complicated for what it's worth
16:06:23pondlifeSo how would AA from metadata be achieved? Or would it not?
16:06:38amiconnWell, that's what I am not sure about
16:07:13amiconnSupporting AA requires a rather big metadata buffer (that's why buffering subsequent metadata on buffer is a good idea)
16:07:48amiconnIf it was just for me, I would never start thinking about supporting AA...
16:08:31pondlifeSame here, but it seems to be the #1 thing users want - hence the custom builds, in the main.
16:08:42amiconnBut with current (+next) metadata on buffer and long tracks, you either have to move metadata around all the time, or handle a buffer with multiple gaps in it
16:09:01Nico_Pthe static buffer for AA could be of the appropriate size for the current WPS
16:09:06amiconnAnd what about wrapping metadata? That should never happen imo
16:09:18Nico_Pand maybe we could do bitmap resizing before copying the AA to it
16:09:21pondlifeAs long as the gappiness and/or moving and/or wrapping is abstracted, it doesn't matter to me.
16:09:24Nico_Pamiconn: I agree
16:09:58amiconnpondlife: That is the point where complexity starts. You cannot fully abstract gappiness
16:10:31amiconnThere are codecs which _require_ contiguous data blocks of a certain size (ranging from a few KB to the full file size)
16:10:52pondlifeWhich ones need the whole file?
16:11:09amiconnAnd that's an swcodec only issue, btw. The MAS doesn't care about it as it does its own buffering
16:11:11pondlifeI mean, at the moment..
16:11:26amiconnAtm: none. But all the mod players are like that...
16:11:41*pondlife cares about mod even less than AA.
16:11:52pondlifeNot that that's a good argument... :)
16:12:02 Quit maffe (Remote closed the connection)
16:12:04amiconnThe sid player just gets around it by copying everything to the remainder of the codec ram - works for sid as those files are small, but won't work for many mod formats
16:12:21amiconnI do care about mod formats - unlike AA
16:12:27*Nico_P is off for a bit
16:12:28 Join maffe [0] (
16:12:31*amiconn even wants to bring mod playback to archos
16:13:09*amiconn has about 21000 mods (that includes sids - which are the least interesting of the mods imho)
16:13:29*pondlife never had a C64,Amiga or Atari, so that's probably why he's missed out on mods.
16:13:54amiconnI would also like to see midi support in the main playback engine - but this will be _very_ complicated
16:14:39*DerPapst <3 midi + mod :)
16:14:52pondlifeMIDI + video via codecs !!
16:15:01amiconnOne possible solution would be to make the loader also buffer the instruments - but then you'll end up with buffering the same instruments over and over when playing several midis in sequence
16:15:30amiconnpondlife: Hehe, and that's what I would not like in the main pb engine: video
16:15:44amiconnVideo is clearly an add-on for a dap firmware, not a core function
16:15:44pondlifeWhy not?
16:16:05amiconnAnd the requirements of video and audio playback are rather different
16:16:38pixelmaI wouldn't want a video to start while I have the dap in my bag and accidentally inserted a video file in the playlist
16:17:09 Join n1s [0] (
16:17:19amiconnWhat the main pb engine could support though would be playing the audio part of a video (iiuc the wma codec already allows this)
16:17:24DerPapsti bet that issue could be solved with a setting :P
16:17:33amiconnDerPapst: How?
16:17:46DerPapstpixelma's issue.
16:17:48amiconnVideo doesn't use a wps, audio does. How would you handle this?
16:18:07DerPapstonly play back the audio stream until you explicitly start a video
16:18:26bluebrotherin that case one could extract the audio on the pc
16:18:40bluebrotherless file size, less disk spinups, longer battery life ;-)
16:19:24*amiconn got an undefined instruction in the clock setup even though set_cpu_frequency() is in IRAM
16:19:26pixelmaDerPapst: then don't have it in the main playback engine - case closed...
16:19:38DerPapste.g. if you have misic videos on a 5G iPod and you play back music in shuffle mode it also plays the aoudio stream of these videos
16:19:49amiconn(the grrr wasn't relaxed to bluebrother's remark)
16:20:32DerPapsti don't care if video playback is in the video engine or not.
16:20:50*bluebrother doesn't care about video on a dap
16:20:51DerPapsts/video engine/playback engine
16:21:19pixelmareally, I don't see why it has to be there - if I want to watch a video, I take the device out and explicitely start it
16:22:59DerPapsttrue... i thought more about something that is done in the wma codec. it can play beack the audio stream of wmv files as well. so if you have a big collection of misic videos and yo don't have the mp3s or whatever you can listen to the audio without watching the video.
16:23:34DerPapstwhile being in the WPS
16:23:55 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
16:24:37amiconnDerPapst: Yes, although bluebrother also has a point here
16:25:01amiconnDemuxing the video on a PC and then playing the audio part saves quite some spinups
16:25:29DerPapsti don't have any music videos so i don't care about that too much.
16:25:52linuxstbDerPapst: That's more by accident than design though...
16:25:55linuxstb(wmv playback)
16:26:08DerPapsti know ;)
16:26:47bluebrotherthen extend the wma codec to reject every file that contains a video stream ;-)
16:30:53*linuxstb wonders if saratoga has made any progress with wma seeking.
16:31:57amiconnSeeking and lsp are the 2 things still missing, correct?
16:32:52 Join Pro90 [0] (n=mongo@
16:34:14Pro90how can i change the etc/profile of cygwin with nano
16:34:48GodEater"nano /etc/profile" ought to open it in the editor
16:37:17Pro90but there is no PATH=/usr/local/bin:/usr/bin:/bin:/usr/X11R6/bin...
16:39:11GodEaterso ?
16:39:53Pro90i want to do the step 4 in the cygwin HowTo
16:40:18Pro90Edit using nano or vi (not Notepad or Wordpad in Windows) /etc/profile and add ":/opt/sh/bin:/opt/m68k/bin:/opt/arm/bin" (without the quotation marks) to the end of the line that says:
16:40:18Pro90 PATH=/usr/local/bin:/usr/bin:/bin:/usr/X11R6/bin...
16:40:18Pro90so that it reads: PATH=/usr/local/bin:/usr/bin:/bin:/usr/X11R6/bin:/opt/sh/bin:/opt/m68k/bin:/opt/arm/bin:$PATH
16:40:18DBUGEnqueued KICK Pro90
16:40:18Pro90Save the changes, exit cygwin and restart cygwin.
16:40:31Pro90this is what i want to do
16:42:09linuxstbamiconn: Yes, I think so.
16:42:40linuxstbamiconn: Although there are still a few bugs in the parser with unusually structured files (my code...)
16:43:40amiconnlinuxstb: Oh, I forgot other missing things: optimisation, optimisation and optimisation
16:44:02CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
16:44:02*amiconn wonders when our emac wizard will return
16:44:10 Quit Bagder ("It is time to say moo")
16:44:38Pro90plz help
16:44:44linuxstbamiconn: Yes, we're missing Coldfire gurus...
16:45:13linuxstbPro90: Your /etc/profile file doesn't contain any lines with "PATH" in them?
16:45:27GodEaterPro90: then just add a line which reads "PATH=$PATH::/opt/sh/bin:/opt/m68k/bin:/opt/arm/bin"
16:45:38GodEaterand put it right at the bottom of the file
16:45:56linuxstbDid you want two colons after $PATH?
16:46:00 Part maffe
16:46:17 Join maffe [0] (
16:46:17GodEaterlinuxstb: no - well spotted - you win a free beer tonight :)
16:46:38 Nick Caliban_ is now known as Caliban (
16:46:38linuxstbHurray... What about the t-shirts? I guess it's no fun now that there are 4 t-shirts and a maximum of 4 people coming...
16:47:01GodEaternot really - you'll get one of those by default ;)
16:47:19DerPapstPro90: open the file in any text editor and check if the line is there. Don't save the file though
16:48:09GodEaterlinuxstb: what time are you aiming to get there tonight ?
16:48:13DerPapstor you can use a text editor that is able to work with unix newlines for editin that file instead of vi or nano
16:48:35DerPapste.g. proton (<−− google)
16:48:53GodEaterooh a new text editor to look at ;)
16:50:09Pro90when i open it with word pad there's the line i need...
16:51:05DerPapstwell then either nano is stupid or... you don't know how to use nano :P
16:51:07 Join chimera78 [0] (i=cf728bfe@gateway/web/cgi-irc/
16:51:15Pro90okay it is also there when i open it with nano
16:51:18DerPapstbetter use a gui editor for that task
16:51:38 Quit relaxed ("")
16:51:49chimera78I just downloaded the latest build, but the zip seems to be password protected. What is the password? And why does that happen periodically? I could not find any information on the webpage
16:51:56DerPapstanyways.. get a text editor that can handle unix newlines
16:52:07bluebrotherchimera78: the zip is not password protected.
16:52:07Pro90how can i save the file with nano
16:52:22DerPapstyou'll need it once you start messing with the rockbox source code
16:52:38bluebrotherPro90: seen the line at the bottom of nano?
16:52:45bluebrother^ means the Ctrl-Key.
16:53:54GodEaterso Ctrl+X to exit, and it'll ask if you wish to save
16:54:29linuxstbGodEater: I should be able to get there for around 6pm. What about yourself? pondlife is due at 7pm?
16:54:35 Quit Nibbier (Remote closed the connection)
16:55:00GodEaterI need to pop home first - I've had a shirt related accident
16:55:03GodEaterso need to chane
16:55:11GodEaterbut I could still make 6-ish I think
16:55:19 Quit lost|X40 ("Leaving")
16:57:15pondlifeSee you there
16:57:57GodEaterindeed ;)
16:58:03 Part chimera78
16:59:22Pro90compiling still doesn't work
16:59:36GodEaterPro90: "source /etc/profile"]
16:59:40linuxstbHave you closed down cygwin, re-opened it, and re-run configure?
17:00:25bluebrotheror re-started bash?
17:01:16linuxstbPro90: You should type "echo $PATH" at the cygwin prompt to check if the PATH is now correct.
17:01:27*GodEater prints out the obligatory Rockbox banner
17:01:28Pro90thanks a lot Guys
17:03:51n1slinuxstb: good news! your patch fixed the lng file generation :-)
17:05:10 Join lostnihilist [0] (
17:05:16 Join Nibbier [0] (n=sven@
17:06:00linuxstbn1s: Good. The only thing I don't like is that "make zip" doesn't check if apps/features exists. But then it doesn't check if Rockbox has been built anyway...
17:06:30*amiconn made a tiny change to the clock setup. No crash so far
17:06:54amiconnBut that doesn't mean much - hafta run it for extended times (just 35 minutes now)
17:07:18linuxstbn1s: Did you read my comment about button.h ?
17:07:47linuxstbI assume you've looked at an apps/features file and seen the mess it causes?
17:08:20n1slinuxstb: yes, I agree it should be in the target config*.h files for targets that has it
17:08:23markunjhMikeS: Do you have an answer for this guy?
17:08:55GodEaterFebs mentioned that at the first DevConPub
17:08:59 Quit Nico_P (Remote closed the connection)
17:09:00GodEaterI've never come across it myself
17:09:11***Saving seen data "./dancer.seen"
17:09:14GodEateror not noticed it at least
17:09:21linuxstbn1s: Will you move it? That could be done as a separate commit at any time. I probably won't have chance for a couple of days.
17:09:57DerPapsti often enough swap my ear plugs without noticeing it :P
17:10:10n1slinuxstb: yes, I'll do it now, then resync the lang cleanup patch with your changes included and ask around for people to test/comment
17:10:19linuxstbMy earplugs don't even have L+R marked on them, so I have no idea what I'm listening to...
17:10:57linuxstbn1s: Maybe it should be changed from HAS to HAVE as well - to be consistent with everything else?
17:11:15 Join kaaloo [0] (
17:11:25n1slinuxstb: yes, I thought about that, is there a reason it is different?
17:11:26 Part kaaloo
17:13:43linuxstbn1s: I can't think of a reason.
17:14:25linuxstbBTW, in case anyone is wondering what we're talking about, it's the HAS_BUTTON_HOLD #define which is in button-target.h. We want to put it in the config-$TARGET.h file instead.
17:14:49pondlifeSounds like it should be in the config headers...
17:15:43amiconnI think that the config-*.h files should only contain stuff that can be selected. Stuff which is defined by the hardware should be defined in the target tree
17:15:56*pondlife longs to try Rockbox on a scrollwheel (with old and new acceleration)
17:16:39linuxstbamiconn: The problem is the file apps/features.txt being used for langv2. This needs to test HAS_BUTTON_HOLD, but if features.txt #includes button.h, it ends up containing C function prototypes etc...
17:16:58amiconnThe drawback is of course that apps code which needs to know about features has to include a shitload of headers
17:17:15amiconn...unless there's some clever inclusion in a central header
17:19:12amiconnThen we need either some clever processing for features.txt (a perl script?), or split the headers into #define stuff and prototypes
17:21:18*amiconn would think that processing a text file with perl so that only the lines with # remain shouldn't be too difficult
17:21:43amiconnEven though I am no perl wizard...
17:22:00linuxstbNot even perl, a simple grep should do it...
17:22:19*amiconn wrote very few perl scripts for rockbox being one of them
17:24:23*bluebrother remembers the FS cleanup
17:24:43bluebrotherok, so does anyone have objections against my announcement for the ML?
17:24:57linuxstbdelete * from flyspray;
17:25:56GodEatertruncate is more efficient ;)
17:26:10bluebrotherclose all entries?
17:26:13amiconnOne whole album played w/o crash ... let's see what happens with the next
17:26:13linuxstbbluebrother: Can you remind me what it said?
17:28:19linuxstb30 July isn't quite a week away any more.. I find mentioning the day of the week helpful - i.e. Monday 30 July to Sunday 5 August.
17:28:40bluebrotherI already changed that to "starting next week" ;-)
17:29:47bluebrotherbut if you think it's better we can postpone it another week.
17:30:00amiconnArathis: around?
17:30:12bluebrothermy major concern was if it's a good idea to ask for help that way on the MLs.
17:30:34amiconnI *think* I have a solution for the occasuíonal crashes on PP502x
17:30:39linuxstbMy only comment is that I think that "this patch needs to be synced" can be useful (when phrased like that, rather than as an order for someone to sync it). But apart from that it seems fine to me - we should try and encourage users to help out with flyspray.
17:31:29amiconnbluebrother: Some rather old issues are still unsolved. Perhaps they should get some special mark in oder to avoid tracking them down over and over?
17:31:53amiconn(one issue that comes to mind is the "norwegian language course" issue, which will be solved with MoB)
17:32:08bluebrotherold issues as in not present in the tracker?
17:32:14bluebrothernorwegian language course?
17:32:17linuxstbbluebrother: Maybe also worth saying that people are welcome to continue after the 5 August - we don't want to give people the impression it's just one week when this would be useful.
17:32:34amiconnOld issues as present in the tracker, but not being clear enough that they are still valid
17:33:01bluebrotherlinuxstb: seen the remark at the end of the text? (line 21 in pastebin) ;-)
17:33:13Arathisamiconn: yes
17:33:27amiconnbluebrother: The issue with very small tracks so that the number of track entries is exhausted (30 on swcodec, 16 on hwcodec) before the buffer is full
17:33:52linuxstbbluebrother: You just added that! :)
17:34:16amiconnHwcodec does something relatively harmless - it plays the 16th track over and over until you skip manually once - then the issue might happen again after the next 16 tracks
17:34:28amiconnI don't know what swcodec does in this case
17:35:26amiconn"Norwegian language course issue" refers to how this issue was first observed
17:36:00amiconnLanguage courses often use lots of tiny audio snippets. So tiny that they even manage to overflow the 16 track entries with a <2MB buffer
17:36:18Arathisamiconn: something to test for me? :)
17:36:27amiconnWant a build?
17:36:43amiconn(20GB iirc)
17:36:47Arathiswhat does it do?
17:37:05amiconnIt should fix the crashes, although I'm not sure yet
17:37:11amiconnHence it requires testing
17:37:30Arathiswell than of course YES! :D
17:37:31amiconnI got no crash for >1 hour now, but of course that doesn't mean much
17:37:50Slasheriamiconn: swcodec should stop buffering when that limit is reached (at least it did that before)
17:37:50amiconnGimme a few minutes (cygwin is slow, you know)
17:38:39Arathisamiconn: I'll do a battery-test with it. wanted to do that anyway
17:39:22bluebrotherok, updated the text a bit: (lines 11 and 19)
17:39:32pixelmaSlasheri: have a question related to "gather runtime data" - does it rely or depend on other database settings (couldn't find something in the wiki but it looks like it doesn't work here "out of the box" on my Ondio)
17:39:53bluebrotherI think that "language" issue could just get a comment in the task.
17:39:57Slasheripixelma: nope, it doesn't depend on any other setting
17:40:47Slasheripixelma: so it should always work if db is initialized
17:40:56pixelmaso it would be enabling that option with initialized database
17:41:04amiconnWhere is the runtime data gathered nowadays?
17:41:18Slasheripixelma: correct
17:41:57pixelmaSlasheri: should the played songs appear in "recently played" (using svb tagnavi.config)
17:42:03Slasheriamiconn: in the tagcache db's main index entries for tracks as 4 bytes numerical data per tag type
17:42:49Slasheripixelma: yes, but those doesn't appear immediately. you may have to play a few songs or stop playback to force an update
17:43:35Slasherionly when playback code generates unbuffer events (when loading new tracks), old played tracks are added there
17:44:06Slasherihowever, modifying manual rating immediately commits the update to the db
17:44:07pixelmaI played maybe 8...9 tracks (from internal memory this time) and stopped, there is nothing in "recently played"
17:44:24Slasheripixelma: interesting.. which player do you have?
17:44:37pixelmaas I said, that was a test on Ondio
17:44:59amiconnMaybe the recently played db stuff suffers from the same problem as the scrobbler stuff on hwcodec?
17:45:03Slasheripixelma: ah, then it's possibly an issue with the hwcodec playback engine not generating those unbuffered events correctly
17:45:10*amiconn wants to see engines unified
17:45:30Slasheriand if that doesn't work, probably no statistics works at all
17:45:40pixelmamaybe that's the same as with the non-working last-fm log...
17:45:53amiconnSlasheri: The hwcodec engine does generate unbuffer events, but maybe someone changed the api for swcodec and forgot to adapt the hwcodec engine, or tell someone with a hwcodec target to do it?
17:46:27Slasherihmm.. i will check the code
17:47:10pixelmathe weird thing with last-fm is that once in a long while, you'll get a log file with a folder or two
17:47:15 Quit gregj (Read error: 110 (Connection timed out))
17:47:25linuxstbWhat exactly is an "unbuffer" event? Is that different to "track change" ?
17:47:28pixelma(on hwcodec I mean)
17:47:39linuxstb(is there such a thing as a "track change" event?)
17:47:49amiconnlinuxstb: It's when the track gets removed from the on-buffer tracklist
17:47:52Slasherilinuxstb: yes, unbuffer event is generated only when disk spins up
17:48:04amiconnIt was implemented this way in order to save extra spinups
17:49:26 Quit wrobbie (Read error: 110 (Connection timed out))
17:49:46*amiconn now has in fact more swcodec targets than hwcodec targets
17:49:55amiconnThe ratio is 5:4 atm
17:49:56pondlifeIs there a "disk spinup" event?
17:50:13pondlifei.e. so lazy writers can avoid spinup themselves.
17:50:18 Join spiorf_ [0] (
17:50:47amiconnpondlife: You can register for an ata spinup callback
17:50:50 Join jgarvey [0] (
17:50:56amiconnThis is what settings do
17:51:22linuxstbWhat was the original use for "unbuffer" ?
17:51:40pondlifeOr older?
17:51:57linuxstbI'm assuming if it exists in hwcodec it's older...
17:52:11amiconnOlder - gather runtime data
17:52:16amiconn(with the old db)
17:52:29amiconnBut not as old as swcodec - it was also backported
17:53:33 Join PaulJam [0] (
17:53:34Slasherihmm, at least it seems that unbuffer events are registered correctly. The next issue could be non-working on-buffer-event
17:54:18 Join My_Sic [0] (
17:54:44Slasherinothing can be updated in the database unless a valid entry id from tagcache is received in buffer event
17:54:58bluebrotherok, any last comments on my fs mail? Otherwise I'll send it
17:55:07amiconnSlasheri: Is this stuff logf'ed?
17:55:17Slasheriamiconn: yes, almost everything should be
17:55:41n1swhat do we call the micro sd card volume on sansas? "mSD"?
17:55:46amiconnHmm, so comparing a logf log swcodec vs. hwcodec should help figuring out what's going wrong?
17:56:14Slasheriamiconn: tagtree_unbuffer_event() in tagtree.c is the code updating that data and logf should tell if it fails and why
17:56:34Slasheriamiconn: indeed
17:57:00Slasheriamiconn: and almost every possible failure in tagcache engine is logf'd too
17:57:19amiconnIsn't that called database nowadays? ;)
17:57:43Slasherinot internally in the code, hehe :)
17:57:51*bluebrother preferred tagcache ;-)
17:57:53linuxstbOr in Slasheri's brain...
17:59:23pondlifeThe "Bring Back TagCache" campaign starts here :)
17:59:24Slasheriand when speaking about tagcache, one can be almost sure what code it refers to :)
17:59:37Arathisamiconn: I'm getting an error unpacking the *.zip
17:59:37*amiconn wonders whether we'll ever see an official H1x0 bootloader that supports rockbox in flash :(
18:00:07Slasheriamiconn: hmm, that just needs to be released. Did you try the 7-pre3?
18:00:30Slasheriamiconn: one problem is that bootloader no longer compiles due to the binary size..
18:00:34amiconnI won't try flashing before we have an official bootloader
18:00:35*linuxstb needs to update his rockbox-in-flash on his h140 but it always makes him nervous
18:00:40bluebrotherthere is this issue that you can't start the player with the remote
18:00:42 Join Domonoky [0] (
18:00:54Slasheriamiconn: well, imho 7-pre3 is stable
18:00:54pondlifebluebrother: I fixed that for H300 with non-LCD remofe
18:00:57pondliferemote, even
18:00:59Slasheriamiconn: and well tested
18:01:08bluebrotherpondlife: in the bootloader?
18:01:12 Join low_light [0] (i=c730180b@gateway/web/cgi-irc/
18:01:28linuxstbSlasheri: Can the plugin reflash an official iriver .hex firmware file?
18:01:28amiconnArathis: Hmm, strange, the .zip is fine here
18:01:41bluebrotheroh, nice. Then someone needs to test and update the bootloader ;-)
18:01:43amiconnBut I'll upload it to my provider's webspace - maybe that works better
18:01:47pondlifebluebrother: It's in SVN now, not sure if it fixes H100 though.
18:01:59Slasherilinuxstb: not directly, it needs to be decrypted on computer first to a .bin file
18:02:04*n1s wants to see a h300bootloader that doesn't start the OF when power is plugged in while off...
18:02:22Slasherilinuxstb: but it should be possible to add that feature to the plugin directly as well..
18:02:22linuxstbSlasheri: But in theory it could - if we added the decryption code from descramble?
18:02:40pondlifeThe SVN H300 bootloader has working USB too.
18:02:48pondlifeOn all players, not just some.
18:03:19pondlifen1s: That would be nice, but LinusN had a good reason for not enabling it.
18:03:27*pondlife can't recall said reason.
18:03:29linuxstbSlasheri: Also, I was wondering if we could distribute a "rockbox.hex" file that a user could install using the iriver firmware? So a user who just wants to install Rockbox on a h1x0 doesn't have to worry about patching an official firmware first.
18:03:39Slasheriamiconn: and btw, what do you consider as being official?
18:03:39 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:04:08Slasherilinuxstb: that is very possible
18:04:17n1spondlife: I think it was to be able to boot the of even if rockbox button driver were broken in the early days, can't imagine why it should remain that way
18:04:51pondlifeI think it was low-battery related.
18:04:56amiconnn1s: Charging is still not 100% clean on pcf targets
18:05:02pondlifeSomething about the bootloader's charging
18:05:08Slasherilinuxstb: and then it would be good to have direct support to flash back from .hex because no ROM dump can be taken in that case
18:05:33n1sI always charge in rockbox and have no problems, just have to remember to power on before plugging in...
18:05:40linuxstbSlasheri: That's why I asked ;)
18:06:42 Quit spiorf (Read error: 110 (Connection timed out))
18:07:26 Join kubiix [0] (
18:08:34*DerPapst knows that "rrrr" isn't a good sign
18:08:51 Quit spiorf_ (Remote closed the connection)
18:09:13amiconnMy provider's upload interface doesn't like me
18:09:51amiconnAh, finally :|
18:10:23pondliferrrr is better than grrr though.
18:10:48bluebrotherand better than brrr :)
18:11:33*pixelma waits for a "(grrr)" from Jens again ;)
18:12:43 Quit jgarvey (Read error: 110 (Connection timed out))
18:13:07 Join jgarvey [0] (
18:13:13DerPapstand an even longer grr[...]rr will happen when his h100 is frozen again
18:13:35amiconnThe H100??
18:13:38pixelmait's nor an H100...
18:14:16bluebrotherxor? ;-)
18:15:40*linuxstb hugs grep and his local copy of the IRC logs ;)
18:15:49*bluebrother just sent a mail to the lists
18:16:45bluebrothernow you only need to filter out those hilights and | wc it ;-)
18:17:05*pixelma likes the presence of Llorean's statement in that list the most :)
18:17:42amiconnThe errror one?
18:18:03pixelmayep, just funny that it was caught there too
18:18:11linuxstbbluebrother: I'll leave that to someone that comes up with a better regexp for amiconn's noises.
18:18:26bluebrothermaybe "hmpf" is worth grepping to? ;-)
18:19:14amiconnSecond album w/o crash
18:20:05*amiconn queues 2 more
18:20:59pixelmalinuxstb, bluebrother: maybe that would be something for the IRC stats ;)
18:21:45Arathisamiconn: started test. I'm off in about an hour or less so I can report tomorrow or so
18:25:12amiconnI think if it doesn't crash in the first ~40 minutes that should be a good sign
18:25:15bluebrotherDomonoky: new tarball uploaded ...
18:25:29Domonokybluebrother: nice..
18:26:44Domonokybluebrother: can you give me the link to the tar ball again, please ?
18:26:50pondlifelinuxstb: "grrrrrrrrrrrrrrrrrr!!!!!!" sounds hard to beat.
18:27:04bluebrotherDomonoky: sure ;-)
18:27:37bluebrotherhmm, can git pull a repo that's published through http?
18:29:16*amiconn wonders what the pp chip is doing
18:31:32DerPapstsome black magic
18:32:02bluebrotherwell, the chip is black at least, isn't it? ;-)
18:32:08amiconnLooks like some bits in 0x600060a0 might be set after switching the clock source to 24MHz which aren't set before
18:32:32 Join spiorf [0] (
18:32:36*amiconn should try to capture them
18:34:25Domonokybluebrother: i can compile your new version withou problems, but installing failed. ( wrong device, i selected h120, and it trys to download sansa, and it downloading also silently failed, and the the zip failed, because the zip is 1 k file )
18:34:50*Domonoky will take a look a the code..
18:36:26 Quit maffe (Remote closed the connection)
18:36:36n1sany gcc gurus around? if so: would it be possible to give the compiler a header file and some defines and get back a list of wht was defined in the header file? and nothing else?
18:37:03bluebrotherDomonoky: strange, it worked fine here.
18:37:15Domonokybluebrother: i recompile in the src dir, and now downloading works, but its the wrong target..
18:37:33bluebrotherhave you re-run qmake?
18:38:38bluebrothern1s: you could let the compiler preprocess the header file and grep for #define in the output
18:38:51*linuxstb leaves for DevConPub2
18:38:56 Quit linuxstb ("CGI:IRC")
18:39:12*pondlife heads to DCP2 too
18:39:28 Quit pondlife ("beer, beer, beeer")
18:39:31*bluebrother wants too but is too far away
18:42:18n1sbluebrother: I can't get that to work, the output doesn't contain any defines at all...
18:43:36 Join RaRe` [0] (
18:43:52amiconnThe #defines will be gone after preprocessing...
18:44:04amiconnJens: ?
18:44:41Jens[18:10] <amiconn> Arathis:
18:44:48amiconnYeah, and?
18:45:03amiconnSorry that I have such a rare first name...
18:45:20JensIt just triggered me, that's all.
18:45:24JensAwesome name, sir.
18:45:52JensActually, "Jens" is currently the most common name in Denmark.
18:47:11*amiconn also gets triggered
18:47:26*amiconn being German, but still, this name is rather common here
18:47:49n1sbluebrother: aha the -dD option does almost what I want, a little healthy grepping on that output and I should be fine :-)
18:47:53JensJens strikes me as a very danish name.
18:48:08JensBut it's biblical, so I guess it goes anywhere.
18:48:37amiconnArathis: Any uncouth behaviour so far?
18:50:03Arathisnot so far :)
18:52:02bluebrotherDomonoky: what build environment are you using? A simple mingw installation?
18:52:28Domonokybluebrother: yes , actually the mingw you can get with qt..
18:52:43*Domonoky is now compiling qt in debug mode.. :-)
18:52:56bluebrotherhmm, I tried that yesterday but it seems that broke with path delimiters
18:52:57*amiconn is looking forward to receiving his ipod G1 and G2...
18:53:10 Join Siltaar [0] (
18:53:20bluebrotherso you're using a simple cmd shell? Or anything else?
18:54:02Domonokyyes, command shell.. couldnt get a nice IDE for Qt running, so just cmd shell :-)
18:54:34*bluebrother fires up the virtual windows and wonders why building didn't work yesterday
18:55:26 Quit obo ("bye")
18:56:38 Quit low_light ("CGI:IRC (EOF)")
18:58:01*DerPapst is scared and hides
18:58:39amiconnundefined instruction at 0x00058BE0
18:58:59amiconn...after almost 2h 30 min
18:59:37 Join pmhaddad [0] (
18:59:52 Quit RaRe (Read error: 110 (Connection timed out))
18:59:58bluebrotherargh. Stupid backslash path delimiter :(
19:00:01DerPapstbummer :(
19:00:15 Nick UNKNOWNFILE is now known as UNKNOWNIDLE (i=c@unaffiliated/unknownfile)
19:00:27 Quit Arathis ("Bye, bye")
19:01:12pmhaddadso i noticed that rockbox is supported by the e200 series of sandisk sansas. I was looking into purchasing a C140. Is the firmware/hardware on those so radically different that porting rockbox to a C140 would be a major undertaking?
19:02:52PaulJampixelma: i saw that you translated Latin1 with "Westeuropäisch", but isn't cp1252 the codepage that is called Westeuropäisch (Western European)?
19:03:33 Join low_light [0] (i=c730190b@gateway/web/cgi-irc/
19:03:44pixelmaisn't cp1252 cyrillic?
19:04:00amiconnNo, cp1251 is cyrillic
19:04:04dionoeano, it's the default codepage on windows in France at least
19:05:09amiconnCP1252 is a Microsoft-extended ISO8859-1
19:06:17pixelmapmhaddad: there was someone starting with a C200 port (see new ports forum ). Don't know how similar they are but you can only find out if you try to gather more hardware info
19:06:47 Join lazka [0] (
19:08:01PaulJamamiconn: thanks for the link. i was only looking at (and didn't even read the whole page. sorry)
19:08:11bluebrotherd'oh! This stupid windows uses the wrong make.exe :(
19:09:09pmhaddadpixelma, ok i'll look into it more depth. the c140 is on $45, but i really don't want to use the built in firmware, if i think it's feasible for me to port rockbox to it i will
19:09:14***Saving seen data "./dancer.seen"
19:09:50 Join lxx [0] (n=lexx@
19:10:01pixelmaPaulJam: I must admit that translating isn't easy... so suggestions are welcome
19:10:46lxxi have one album in one flac file and cue file
19:11:06lxxrockbox support playlist in cue files?
19:11:11 Join gregj [0] (
19:12:28 Join Xerion [0] (
19:13:05lxxwhat i need to do
19:13:50n1slinuxstb: for the logs: using -imacros to include the button.h file in the makefile got rid of the function prototypes, no extra parseing needed :-)
19:13:53lxxwhen i try open cue file in rickbox its just opened but music not start playing
19:14:27n1slxx: IIRC you need to enable cue sheet support in the settings somewhere
19:14:48lxxi did it
19:14:54lxxbut its no help
19:15:16lxxmay be couse flac file and cue file in diffrent folders?
19:15:21PaulJampixelma: what do you think of translating "Now Playing" with something like "Aktuelle Wiedergabe". for me "Spielt gerade" somehow sounds strange. but maybe thats only me.
19:16:15pixelmathe flac and the cue have to be named the same (without the extension) - and you need to enable cuesheet support in the settings). Read a bit in the "General settings" chapter in your player's manual about it
19:16:28pixelma^@ lxx
19:18:08lxxthey need to be in the same folder or its no matter?
19:18:13*DerPapst agrees with PaulJam
19:18:34n1shmm, the page should probably be deleted now that the patch has been committed a while ago, I guess this will go to the logs, LinusN Bagder Zagor
19:18:38pixelmaPaulJam: I don't like "Spielt gerade" much myself... only thing is that it's a bit shorter
19:18:50pixelmathan "Aktuelle Wiedergabe"
19:19:10pixelmalxx: I think they have to be in the same folder, yes
19:19:25*amiconn now tries with the special ipod color trickery disabled
19:19:57amiconnStrange thing is that all PP5020+ ipods (except the minis) are doing this trickery
19:20:34lxxthank for the help
19:20:45Domonokybluebrother: in the updateDevice function you set "defaults/platform" in the user settings, but when you download rockbox you use "defaults/curplatform" , that cant work ? :-)
19:21:36 Quit thomasg_ (Read error: 110 (Connection timed out))
19:21:55 Quit lxx ()
19:21:57 Join miepchen^schlaf [0] (
19:21:59pixelmaPaulJam: I first wanted to get rid of the worst things, like "Knopf Beleuchtung" *seems like I can't get over it* ;)
19:22:07 Join thomasg_ [0] (
19:22:08amiconnIt almost seems that the PP5020 gets out of control of its memory interface - but only after thousands of clock switches...
19:22:58*amiconn would like to know what 'undefined instruction' it reads when this exception is triggered
19:25:45 Join maffe [0] (
19:26:16Domonokybluebrother: if i change it, so it uses "defaults/platform" when installing, it works..
19:26:33DerPapstamiconn: is there no way to dump taht?
19:26:46Domonokyan i have seen you have built the logfile function, and the configure screen is nice.. :-)
19:26:52bluebrotherdamn, seems that was a line I missed when changing it.
19:26:58amiconnIt should be possible to re-read from memory, but perhaps it already fixed itself by then...
19:27:04 Join McEnroe [0] (
19:27:11bluebrotherI guess I didn't notice because my config still contains old lines.
19:27:49McEnroehow do i fast forward in mpegplayer (ipod)?
19:27:54bluebrotherbtw, what do you think about including the license in the about window?
19:28:22Domonokyand i think the updatedevice function is on the wrong signal, it gets triggerd when you highlight a string, so if you abort a selection, you hav the wrong device..
19:28:36DerPapstMcEnroe: you can't
19:28:43pixelmaMcEnroe: currently not at all (it's not implemented yet)
19:28:48DerPapstit isn't implemented at all
19:28:50Domonokybluebrother: the licence screen is nice.. :-)
19:33:49McEnroeDerPapst: pixelma: it guess it's already planned, isn't it?
19:34:58bluebrothergrrr. Something is really strange with kdevelop
19:35:17bluebrotherI *did* change that lines. But for some reason kdevelop overwrote it again.
19:35:32bluebrotherthat's also the reason why the log file is messed up :(
19:35:42*n1s posts an IMHO commit-ready lang v2 cleanup patch to and hopes people can test/comment
19:35:47Domonokybad kdevelop
19:36:28bluebrothermaybe I should give eclipse another try
19:36:56Domonokybluebrother: about the log file, i think it would be better to use a ini file for this, like the wx version does it..
19:37:26bluebrotherI haven't checked the wx version −− do you assign a key for each file?
19:38:18Domonokyyes, the key is the path and the value is the number you installed it..
19:38:26 Quit Xerion (Read error: 104 (Connection reset by peer))
19:38:40*amiconn is poking in the dark, it seems :(
19:38:52bluebrotherthe number I installed? You mean version number=
19:39:01pixelmaMcEnroe: Rockbox and estimates when something will be done don't go well together... I think I've read somewhere that one developer wants to look into it (not even sure though). And it's not a priority as Rockbox is mainly about music...
19:39:22pixelma(playing music)
19:39:45XavierGramiconn: bootloader version 7 for H120/H140 is very stable. I've been using rockbox on flash on my H140 for some months without any problem at all
19:39:47Domonokyno, the number of installs, so it gets incemented everytime you install that file again.
19:40:20bluebrotherhmm. What't the purpose of this?
19:40:41XavierGrSlasheri: do you recall if H100 models are flashable with the current version 7 bootloader? I have an H115 and it is a shame not to flash rockbox on it, OF is wasteful
19:41:05Domonokythe only purpose is, that you dont have to truncate the file, ( so you keep history) and dont have dublicates
19:41:37bluebrotherah, so you append each new install with a new section to that file?
19:41:46Domonokybecause you want to use this log, also when installing themes, or any thing else..
19:41:54bluebrotherwell, my intention was to remove an old install before installing anyway.
19:42:18bluebrotherI was wondering if it makes sense to use a different file for themes
19:42:42Domonokybluebrother: you can only remove if you have the log.. otherwise you would delete user config..
19:43:06Domonokyperhaps sektions for the different parts of installations ?
19:43:34Domonokyso a section for rockbox-base, one for fonts, one for themes, one for doom.. etc..
19:43:48Domonokythen you could provide uninstallation for all of this..
19:43:59bluebrotherwell, you could just skip removal if the log file isn't present.
19:44:03low_lightamiconn: what are have you been trying on the pp5020? maybe I can look through the m:robe bootloader.
19:44:41Domonokybluebrother: sure
19:45:02pixelmaMcEnroe: but I guess that's a feature that is desired to have (to have a properly working mpegplayer). Only no-one knows when it'll be done ;)
19:45:20Domonokyi think sections in one install log would be nice, otherwise you would have many many different install logs :-)
19:45:21amiconnlow_light: I am trying to get clock changing stable
19:45:50amiconnMaybe the m:robe original firmware could give some clues... the H10 firmware isn't very helpful
19:46:05SlasheriXavierGr: probably it doesn't work and could brick the player because of the wrong dram parameters
19:46:25Slasheribut it should be possible to do by recompiling the bootloader for H115
19:46:28amiconnIt seems to use lists which are built at boot and then referenced, unlike the ipod firmwares which use literal constants most of the time
19:46:58 Join obo [0] (n=obo@rockbox/developer/obo)
19:47:04Domonokybluebrother: you should get this code into a svn, the others could also work on it :-)
19:47:06amiconnMaybe the emulator would help understanding what's going on - if I knew how to use it...
19:47:48bluebrotherin what way? Creating a new module?
19:48:14bluebrotheror as a subfolder for rbutil so we have both the same time?
19:48:57bluebrother(btw, feel free to send me patches until it's in svn ;-)
19:52:16 Join crwl [0] (
19:54:13 Join Rondom [0] (
19:57:03oboor svn copy the current rbutil folder to branches and work on it there?
19:59:11 Quit iamben (Read error: 110 (Connection timed out))
19:59:37*bluebrother is out for the next 2 or so hours ... but a branch might be an idea.
20:02:01XavierGrSlasheri: anyway that someone could make it work? (or is it only linus that can do this without bricking concerns)
20:02:07 Quit krp (
20:02:07 Quit Hadaka (
20:02:07 Quit Jens (
20:02:07 Quit preglow (
20:02:31Domonokybluebrother: about the code for installing: you should check all userinput values at the begin of accept, and put it in the userconfig, and afterwards use the config, and not the values from the ui directly.. :-)
20:02:39NJoinJens [0] (i=Jens@pdpc/supporter/active/Jens)
20:02:39NJoinkrp [0] (
20:02:39NJoinpreglow [0] (
20:02:39NJoinHadaka [0] (
20:03:03low_lightamiconn: maybe this helps?
20:03:33low_lightlooks like some cop sleeping going on during the freq change
20:05:00SlasheriXavierGr: if everything goes well, just recompiling _should_ be enough. But nobody has tried that..
20:05:22Slasheriand the current svn bootloader does not even compile because of binary size and isn't safe to try
20:08:24amiconnlow_light: Hmm, this looks somewhat different from both ipod and H10 code, but this Cxx_CTL handling _other_ than bit 31 is also present in H10 (and Sansa) code
20:09:13amiconnI wonder what bit 30, 27..24 and 7..0 are doing (those are touched in all PP firmwares more or less)
20:10:05amiconnThe H10 doesn't seem to sleep the cop during clock changes, and neither do the ipod firmwares
20:10:52low_lightamiconn: that is from the bootloader not the OF. I'm not sure to what degree (if at all) the bootloader used freq scaling
20:11:55amiconnThat's why I prefer looking at of disasseblies, not bootloaders
20:13:10low_lightwell, the OF was too big for me to tackle
20:14:13amiconnSearch for constants which occur in clock setup, like 0x60006020, 0x60006034 etc
20:14:41*amiconn guesses the m:robe OF is quite small compared to e.g. ipod video
20:15:26 Quit Jens (
20:15:26 Quit preglow (
20:16:16NJoinJens [0] (i=Jens@pdpc/supporter/active/Jens)
20:16:16NJoinpreglow [0] (
20:19:27low_lightIt's 1.8 MB
20:19:48 Join Llorean [0] (
20:20:25amiconnThe video firmware is 7.1 MB
20:21:12amiconnI am doing this with a 2-way strategy. I have a raw disassembly created with arm-elf-objdump, and have the binary loaded in ida
20:21:12 Quit low_light ("CGI:IRC (EOF)")
20:22:05 Join low_light [0] (i=c730190b@gateway/web/cgi-irc/
20:23:45amiconnI load the raw disassembly into an editor, and search for magic constants. Then I go back to the start of the subroutine (not difficult to recognise), go to this address in ida, and press 'c'
20:32:09 Quit pmhaddad ()
20:37:20 Quit desowin ("use linux")
20:38:25 Quit Pro90 (Read error: 110 (Connection timed out))
20:39:41 Join slie [0] (
20:41:15slieI've got a few questions
20:42:13 Nick slie is now known as valiant (
20:43:37*valiant shrugs
20:43:45 Quit valiant (Client Quit)
20:46:16 Part maffe
20:46:28 Join maffe [0] (
20:47:11 Part maffe
20:47:30 Join maffe [0] (
20:51:11 Join entheh [0] (
20:53:01 Join web-taz [0] (
20:56:18*low_light doesn't understand arm 32bit assembly
20:57:05 Nick web-taz is now known as webtaz (
20:57:10 Join salty-horse [0] (n=ori@pdpc/supporter/active/salty-horse)
20:57:39 Nick webtaz is now known as web-taz (
20:58:21salty-horsehi. any sansa e200 devs/users around?
21:02:57 Join barrywardell [0] (
21:04:46 Join Xerion [0] (
21:05:24 Nick UNKNOWNIDLE is now known as UNKNOWNFILE (i=c@unaffiliated/unknownfile)
21:09:15***Saving seen data "./dancer.seen"
21:09:46 Part maffe
21:14:37joshinsalty-horse: ask your question. If someone knows, they might answer.
21:15:54salty-horsei'd like to purchase a sansa, now that rockbox supports it. does it support all E2* (non R) models?
21:16:02 Quit McEnroe ("leaving")
21:16:22salty-horsethe wiki says power handling is "Not yet optimal" - does it mean the battery has a shorter life on rockbox?
21:16:54DerPapstyes and yes
21:16:56 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
21:17:04salty-horsedoes ogg playback require more power than mp3? (if the mp3 playback is somehow hardware accelerated)
21:17:17 Join miepchen^schlaf [0] (
21:17:24DerPapstit's not hardware accelerated
21:17:36DerPapston the sansa
21:17:40n1ssalty-horse: they are both decoded by the cpu and should be fairly equal in speed
21:17:43LloreanBut ogg/vorbis is a more complex codec.
21:17:49LloreanIt's a bit slower
21:18:01salty-horseslower how? is it noticable in playback?
21:18:06 Join maffe [0] (
21:18:19LloreanNo, "slower" just means "the CPU has to spend more time boosted, so battery life might be a little bit worse."
21:18:25DerPapstthere are charts comparing decoding speed
21:18:34n1ssalty-horse: if you have very hight bitrate files or use a lot of dsp stuff it might be
21:19:20salty-horsei encode oggs at 44,100hz
21:19:35n1ssalty-horse: yes that is the _sample_ rate
21:19:44n1sbut at what bitrate?
21:19:48LloreanActually, interestingly enough it looks like Vorbis performs better at higher bitrates on ARM, while MP3 performs better at lower bitrates
21:20:17salty-horseLlorean, and the sansa is ARM-based?
21:20:25n1ssalty-horse: if you do not use very much dsp like eq crossfeed etc you should be fine
21:20:34salty-horseare there arm optimizations in the ogg source code, or is it all compiler magic?
21:20:37LloreanAt 160kbps, the performance table suggests Vorbis is better than MP3 for the moment
21:20:50LloreanThough odds are MP3 can be optimized to outdo Vorbis, as it does on Coldfire targets.
21:21:03n1ssalty-horse: yes Tremor has some ARM asm code in it
21:21:06Lloreansalty-horse: There's a wide range of optimizations in most of our codecs.
21:21:49n1salso Tremor was written from the ground up to be used on an ARM based platform
21:22:22salty-horseLlorean, "our"? don't you use the codecs from the big libraries? (xiph, mad, etc)
21:22:48Lloreansalty-horse: Do you assume that they were incorporated completely unmodified, and never optimized for our hardware?
21:22:59salty-horseabout movie playback, is mpeg the only supported codec?
21:23:09LloreanMPEG 1/2, yes
21:23:37salty-horseLlorean, i would assume that any optimizations on your part are integrated back to the upstream projects :)
21:24:54DerPapstit is for some parts
21:26:27 Part maffe
21:28:07 Quit lazka (Read error: 110 (Connection timed out))
21:30:38 Join maffe [0] (
21:31:01salty-horsehow does file transfer currently work on the sansa? AFAICT there is no usb support
21:31:24n1ssalty-horse: you have to use the original firmware for that
21:31:50n1sIIRC the simplest way is to just plug in usvb while the sansa is off
21:33:23 Quit barrywardell ()
21:33:47salty-horsen1s, how does that work? the usb powers the flash drive and my pc does the mounting?
21:34:59n1ssalty-horse: I would guess that the sansa bootloader powers on, detects usb and boots...
21:35:15n1salthought I have never used one, why don't you just try it?
21:35:35DerPapsthe doesn't have one :P
21:35:37salty-horsen1s, no sansa yet - or any other portable player for that matter. doing market research :)
21:35:52salty-horsea friend has one, but he won't let me "experiment" (meaning: b0rk his device)
21:35:57 Quit web-taz ("( :: NoNameScript 4.02 :: )")
21:44:36PaulJamn1s: i noticed that since you reworked the set time & date screen the statusbar flickers everytime you change a value or move the cursor.
21:44:50amiconnLlorean: I'd rather think that Vorbis, aac and wma on coldfire can be optimised to get nearly as fast as mpeg audio
21:45:21amiconnThe thing is that mpeg audio on coldfire uses the emac very efficiently, unlike tremor & co
21:45:54n1sPaulJam: hmm, ok, I'll look into it, probably tomorrow, on which target, I din't notice anything on my h300, or maybe I'm just blind...
21:46:26n1sah, /me is probably blind then :-)
21:46:50amiconnsalty-horse: Btw, the "big" libraries are most often not very usable on embedded systems (the exception being mad here)
21:46:58PaulJamwell, it's not a big issue, but i thought i should mention it.
21:47:18n1sPaulJam: no bug is too small :-)
21:48:55amiconnn1s: I noticed that too on H10, but wasn't sure whether it's a new effect
21:49:24amiconnI thought maybe it was like this all the time and just shows up on H10 because of the slowish lcd driver
21:49:40 Join kaaloo [0] (
21:50:03n1samiconn: the other bug which you told me about a couple of days ago should be fixed now at least, have you tested?
21:51:03 Part kaaloo
21:51:24amiconnI'll check..
21:51:35*amiconn doesn't use the set time & date screen often
21:51:36PaulJamfor me it is fixed.
21:51:52amiconnI just noticed it because the H10 was new, and I had to set time & date
21:53:49amiconnLooks like it's fixed. Checked both on Recorder (german, rockfont-8) and H300 (russian, nimbus-19)
21:55:18n1samiconn: ah, good, that screen does still not handle very big fonts but that is not a new issue...
21:58:53amiconnAnyone with an ipod color/photo around?
22:00:08DerPapstto get a firmware dump?
22:00:09jhMikeSmarkun: yes, I had a very long answer for that guy and yet another suggestion to devs to get the other PP targets to do what e200 does (which sould require minimal poking).
22:00:20amiconnDerPapst: No, that I have already
22:00:29DerPapstok :)
22:00:48amiconnI would like someone with an ipod color to perform a test with bit 30 and 31 of 0x600060a0 reset permanently
22:01:36amiconnI wonder what these bits are good for - the ipod color is the only target which has them set normally, and handling them seems to be the cause of unstable clock switching
22:02:12amiconnRight now I am running a test without this handling on H10, and it's running stable for ~3 hours now - the longest time so far
22:02:49amiconnThe ipl wiki says bits 30 and 31 of 0x600060a0 might have to do with lcd handling
22:02:54amiconn...on ipod color
22:06:46jhMikeSamiconn: PP5020 troubles? I can't believe that. ;) What does "getting out of control" mean?
22:07:14 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
22:08:06 Join hannesd [0] (
22:12:24 Join web-taz [0] (
22:12:34 Join DefineByte [0] (i=5751c305@gateway/web/cgi-irc/
22:12:40 Quit web-taz (Client Quit)
22:13:17n1sPaulJam: want to try a fix for the flicker bug?
22:14:02 Quit DefineByte (Client Quit)
22:14:20PaulJamI'm currently on windows. so i can't compile. but if you send my a binary i could test that.
22:14:27 Join DefineByte [0] (i=5751c305@gateway/web/cgi-irc/
22:14:32DefineByteoops disconnected
22:14:36jhMikeSn1s: the flicker in the time screen statusbar?
22:14:59n1sjhMikeS: yes
22:15:25 Quit chrisjs169 ("Konversation terminated!")
22:15:46jhMikeSI saw that on e200 just yesterday
22:16:26DefineByteis diff -u -r olddir newdir > patch the best way to make patches? for some reason the formatting looks different to some of the patches I've downloaded on flyspray
22:17:48DefineByteis it just the difference between svn and diff?
22:17:52jhMikeSDefineByte: why not "svn diff > <patchfile>" ?
22:18:24 Join webguest16 [0] (i=4f421b5b@gateway/web/cgi-irc/
22:19:08DefineBytehmm if i could understand that command i might try it >.<
22:19:35DefineBytei'm trying out cygwin for the 1st time
22:19:45*amiconn wonders what idea jhMikeS has about packed L-R samples
22:20:01 Quit low_light ("CGI:IRC (EOF)")
22:20:09n1sDefineByte: just type "svn diff > my_patch.diff" in you svn root (if you checked out the source with svn
22:20:18jhMikeSamiconn: sansa already uses them so why not get them working on the other pp targets?
22:20:26amiconnYes, but how?
22:20:27PaulJamn1s: that seems to fix it.
22:20:28DefineBytei didn't check it out
22:20:35DefineBytei just downloaded the source
22:20:45amiconnI thought you already know how to do it
22:20:50n1sPaulJam: thanks, will commit then, was a simple one-liner
22:20:59 Quit XavierGr (Nick collision from services.)
22:21:03 Join XavierGr [0] (
22:21:34jhMikeSamiconn: It already worked partially on the H10 and it could record. Perhaps just the audio chip setup needs some changing or something.
22:22:27jhMikeSsomeone tested this for me. whatever pp target I get ahold of myself I'll make it work.
22:22:59amiconnI could test on several, but atm I am still busy with this clock stuff :|
22:23:04jhMikeSThe only change I made was to use the register values from the sansa and leave IIS in slave mode instead of Master.
22:23:05n1sDefineByte: well, then the svn diff command wont work :-/
22:23:29n1sbut yeah the one you suggested should work
22:23:45jhMikeSyeah, np. that's important work.
22:23:48amiconnThere must be a reason why the other targets use IIS master
22:23:57amiconnDid you try master mode on sansa?
22:24:09DefineBytehmm, guess I better check out svn >.>
22:24:10*amiconn isn't even 100% sure what that means
22:24:29PaulJamn1s: thanks for the fast fix :)
22:24:39jhMikeSamiconn: sansa is master mode since the AS3514 can only run is slave mode. others have the audio codec as master.
22:25:19jhMikeSthe master provides the LRCK and BCLK, the slave follows
22:25:23n1sif (buildtable.color==green)nls->sleep();
22:27:58 Join My_Sic [0] (
22:30:02n1swell, good night then
22:30:06 Part n1s
22:32:15 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:32:29 Quit amiconn (Nick collision from services.)
22:32:31 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:34:50 Quit XavierGr (Nick collision from services.)
22:34:54 Join XavierGr [0] (
22:34:59 Join chrisjs169 [0] (
22:36:23 Quit XavierGr (Nick collision from services.)
22:36:27 Join XavierGr [0] (
22:38:54 Quit My_Sic (Read error: 104 (Connection reset by peer))
22:40:28 Quit Jens ()
22:42:47*amiconn has an idea...
22:44:49jhMikeSamiconn: about?
22:46:09amiconnThe WM* format setup
22:46:28amiconnDo you still have your L-R packed samples patch for non-AS3514 around
22:47:05 Part DefineByte
22:47:15jhMikeSwell, it _was_ part of a huge pcm update but I can provide it. it had some locking problems on H10 when trying to play after recording.
22:47:47amiconnplayback after recording is broken on anything pp afaik
22:48:02 Part webguest16
22:48:03 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
22:48:34jhMikeSfrankly I'm sure sure why there should have been any lockups but what's new with that.
22:49:23 Join daurn [0] (n=fake@unaffiliated/daurnimator)
22:52:05 Quit tris (Connection reset by peer)
22:52:14 Join tris [0] (i=tristan@
22:53:02jhMikeSamiconn: at the top of pcm-pp.c just make sure SAMPLE_SIZE_16BIT is defined.;filename=big-pcm-update.diff
22:53:10 Quit chrisjs169 (Remote closed the connection)
22:53:57 Join chrisjs169 [0] (
22:54:35Bagderbluebrother: good call on the cleanup week!
22:55:23amiconnjhMikeS: That's a huge patch.....
22:55:48*amiconn thought packed L-R samples are already in svn for sansa?
22:58:07 Quit daurnimator (Read error: 110 (Connection timed out))
22:59:28jhMikeSyes, but some stuff to use them for other targets isn't
22:59:53 Quit Domonoky (Read error: 104 (Connection reset by peer))
23:00:56jhMikeSit's huge but it's a big cleanup of the pcm interface that even removes code duplication in the sim
23:01:00 Join lukasm [0] (
23:01:12amiconnAh, hmm...
23:01:28*amiconn isn't sure whether he can make this huge patch running with packed samples
23:01:37amiconnShould it run as-is on other pp targets?
23:02:10 Join My_Sic [0] (
23:02:35jhMikeSit's supposed to but there was just that thing with going back to playback.
23:02:52jhMikeSOn H10 anyway
23:02:55*amiconn never tried recording on a PP target yet
23:03:51 Quit XavierGr (Nick collision from services.)
23:03:54 Join XavierGr [0] (
23:03:57jhMikeSe200 has no trouble at all nor do cf and s3c2440
23:04:24amiconncf meaning X5 in this case?
23:04:32jhMikeSx5, H120
23:05:20jhMikeSoh and i2s-pp.c needs IIS_FIFO_FORMAT_LE16 instead of IIS_FIFO_FORMAT_LE32 set
23:05:48amiconnHmm. Is there a reason not to commit this patch? More testing? Anything else?
23:05:58amiconnI mean without the packed sample thing
23:06:04 Quit XavierGr (Client Quit)
23:06:15 Join XavierGr [0] (
23:07:18jhMikeSamiconn: H10 did the lock even without packed samples. wouldn't play with them, but would record.
23:07:49amiconnHmm, does it play without packed samples or not?
23:08:11amiconnSo there shouldn't be a problem...
23:08:33amiconnPlayback after recording is broken on all PP targets in svn afaik (except maybe sansa)
23:09:03amiconnOn all PP target which can record of course
23:09:16***Saving seen data "./dancer.seen"
23:09:20 Quit davina ("xchat on Ubuntu 7.04")
23:09:31jhMikeSOne problem with recording currently was that r12 is not preserved in the c fiq handler and calling the c functions changes it. banked r12 was relied upon to keep its value.
23:10:31jhMikeSI moved the asm FIQ banked regs to r10 and r11 to be relied upon since those are supposed to be callee preserved
23:12:00jhMikeSamiconn: but in effect, no new bug would be present even if the reason is different?
23:12:38amiconnDoes any target use the C fiq? I thought all targets use asm and the C version is just there for reference?
23:13:03amiconnI'm not sure about H10 recording
23:13:10*bluebrother returns
23:13:17jhMikeSall the recording is the C fiq_record
23:13:38jhMikeSfor playback they all use the asm
23:15:24amiconnShouldn't the __attribute__ ((interrupt ("FIQ"))) handle the register stuff for us?
23:15:27jhMikeSr12 corruption after recording was getting sansa at random too when I worked on this.
23:15:50 Quit tris (No route to host)
23:16:10 Quit GodEater (Read error: 110 (Connection timed out))
23:16:12*amiconn wonders why the sansa uses __attribute__((naked)) and hand-crafted register saving
23:16:20bluebrotherBagder: thanks :)
23:16:23jhMikeSthis removes that too
23:17:07jhMikeSGCC FIQ handling is buggy and using the stack within the handler itself will trash things
23:17:26amiconnThe fiq has a separate stack
23:18:31amiconn(or at least that's what I understand)
23:18:32jhMikeStrue, but something is wrong with GCC regarding the banked sp and registers in FIQ.
23:19:11*amiconn decides that this is too complex to figure out in parallel with clock handling :\
23:19:14jhMikeSIf you use a variable that compells stack use (like an output pointer param), things die rather fast
23:20:16jhMikeSI'll gladly commit it if this transistion bug is nothing new
23:21:40amiconnPlayback on H10 running for 4 hours now w/o crash...
23:21:45lukasmhi there... i've a problem loading original h10 firmware with rockboxloader... it's an mtp device. can anyone help me ?
23:22:11bluebrotherlukasm: only if you dare to ask the question ;-)
23:22:48lukasmi think i do so
23:22:55jhMikeSamiconn: another thing I'm not certain about is the mods to pp5002. I was going to wait until after getting what I need to check that out on 3g.
23:22:56*amiconn wishes linuxstb would be around - damn DevConPub2 ;(
23:23:21amiconnjhMikeS: I'll receive 2 PP5002 targets, hopefully soon
23:23:35*DerPapst has two too
23:23:35amiconnIn fact I'll pick up the G2 in less than 2 days
23:23:51amiconnDerPapst: Do you have a G1 and/or G2? ;)
23:24:01*jhMikeS is just ticked atm that he has one right here on his desk he can't connect :\
23:24:01 Join BrandoNT [0] (
23:24:03DerPapstnope... :P 3 3Gs
23:24:22lukasmbluebrother: i've searched forum wiki etc pp, no hint. the player stops whith "ROLO executing" message
23:24:42amiconnjhMikeS: Those cables should be available from numerous places...
23:24:53bluebrotherI don't know if rolo works reliably on the h10 yet. What are you trying to achive?
23:24:55amiconnG1 and G2 use standard firewire cables
23:25:08amiconnG3 uses the same cable as the later ipods afaik
23:25:17jhMikeSamiconn: what about the charging. I mean, I can rig something up or find a cheap firewire card I guess (if such a thing exists).
23:25:17DerPapstcorrect ;)
23:25:25DerPapsteither usb or firewire
23:25:41amiconnThe G3 is said to only charge from firewire?
23:26:06amiconnAnyway, there are nuermous firewire chargers around as well
23:26:08DerPapstunfortunately you cannot charge it from usb :(
23:26:09lukasmi'm trying to get the original firmware running for radio-testing purposes
23:26:18BrandoNTWhere can i find RockBoy
23:26:36DerPapstit's already installed
23:26:49BrandoNTisi t?
23:26:52BrandoNTYOU SURE?
23:26:56DerPapstyou only have to select the rom from the file browser
23:27:02DerPapstI AM
23:27:12BagderBrandoNT: the manual is a good read
23:27:19 Join tris [0] (i=tristan@
23:27:22bluebrotherlukasm: to boot the OF hold Prev while booting
23:27:27*amiconn fetches a beer
23:27:28bluebrotheryou can't boot it from a file
23:27:37DerPapstamiconn: prost :D
23:28:10*jhMikeS doesn't wanna hear about alcohol after so much sake yesterday :P
23:28:12BrandoNTTHANK YOU
23:29:21lukasmoh... looks like i need to get my glasses cleaned ^^
23:29:42*DerPapst hugs his 2 displays
23:31:00BrandoNTMy battery!
23:31:07BrandoNTrockbox drains my battery!
23:31:11BrandoNTlike never before!
23:31:26BrandoNTI fully charge, use for 5 min
23:31:33BrandoNTbattery says 80percent
23:32:00 Join Jess [0] (i=46ef580c@gateway/web/cgi-irc/
23:32:05DerPapstthat sounds impossible.
23:32:08LloreanBrandoNT: 1) Please read the channel guidelines.
23:32:13jhMikeSamiconn: one thing that stikes me odd, is the headphone socket config pcm_rec_dma_init doesn't change it back to output ever again :\
23:32:29Llorean2) Rockbox is known to use more battery on PortalPlayer based targets right now. We haven't hidden this. Perhaps you should research software *before* installing it.
23:32:42DerPapstor your battery is already dead and needs to be replaced
23:33:06BrandoNTI was aware of 'more battery usage'. but not 'total' battery usage
23:33:10JessHi all. I've got an old Archos Recorder Jukebox 20. The archos docs list a 999 file limit per directory. does RB get around this?
23:33:11bluebrother2a) the FAQ is usually a good thing to read
23:33:23BagderJess: yes
23:33:24LloreanBrandoNT: "Total" means "All", which of course depends on how you use it.
23:33:26BrandoNTyesyes.. why read that when i can bother you guys?
23:33:40krazykitBrandoNT, because those are the rules
23:33:42*amiconn decided that >4 hours is enough testing for stability and puts a build onto his H10 with reinstated but modified 0x600060a0 handling
23:33:43BagderJess: but we don't advice using such horrible dir structure
23:33:48LloreanBrandoNT: Because if you refuse to read documentation, you'll eventually be kicked from the channel for blatantly ignoring the guidelines.
23:33:52bluebrotherBrandoNT: because it's *really* annoying to tell things again and again and again?
23:34:12JessBagder, huh? such horrible dir structure? I'm confused.
23:34:15BrandoNTALL: i apologize, i very young
23:34:28BagderJess: having more than 999 files in one dir is plain silly
23:34:31bluebrotherbecause devs should better develop instead of answering questions that are documented for exact that purpose?
23:34:39Bagderbecause it makes a mess
23:34:48Bagderwhy not use multiple directories?
23:35:24BagderJess: you don't use multiple folders on your computer for anything?
23:35:28Jessi keep all my albums sorted by artist/album. but i have a huge collection of single songs that I like to shuffle. These all go in the big dir.
23:35:48bluebrotheryou don't need all files in a single folder to shuffle
23:36:09BagderJess: going techy, allowing that many files will require a huge buffer that you could otherwise use for better things
23:36:13bluebrotherRockbox is heavily based on playlists, thus you shuffle a playlist.
23:36:17Jessthe single songs are one hit wonders and such. folks I dont want and entire album of.
23:36:34amiconnApart from being quite silly, rockbox does support bigger dirs than 999 files
23:36:50amiconnThe default is less though
23:36:54Bagder... as I already said ;-)
23:37:00*bluebrother points to the manual once more ;-)
23:37:09Lloreanbluebrother: Or you could just have \Music\Singles\A\, \Music\Singles\B, etc, then insert "Singles" into a playlist when you're ready to start playing.
23:37:15Jessin the orig archos OS, i could only shuffle a single dir. I'm thinkin of switchin to RB.
23:37:36bluebrotherLlorean: that was something I thought about ;-)
23:37:38 Join TOuf [0] (i=9052c506@gateway/web/cgi-irc/
23:37:48BagderJess: rockbox is way way waaay better than OF on archos
23:37:51bluebrotheror make a folder Singles with subfolders for every decade
23:37:54 Join haemmy [0] (n=stefan@
23:37:58Jessin RB, is it possible to create a playlist of the whole drive and shuffle that?
23:38:03bluebrotheror ... there are quite a couple of options.
23:38:15*bluebrother points to
23:38:25BagderJess: and rockbox will load that playlist in a few seconds
23:38:28amiconnjhMikeS: what headphone socket config?
23:38:35Jesswill such a huge playlist bog the machine?
23:38:43Bagdercompared to the OF in archos, loading playlists is amazingly much faster
23:38:45*amiconn is a bit slow today obviously :\
23:38:50TOufhello, i have a problem with the installation
23:38:52BagderJess: no
23:39:09bluebrotherJess: you can remove Rockbox without any problems. I suggest you just try it.
23:39:11TOufi am using ipodpatcher.exe but it says it cannot find any ipods even though my ipod is connected on disk mode
23:39:28bluebrotherdo you have administrator rights?
23:39:33Jessdo many of you use the archos to make bootlegs? do you get decent quality?
23:39:42lukasmbluebrother: thanks for that hint, but according to the graphic in the quickstartguide it's not the prev-button to press but the cancel button. but thanks anyway ; )
23:39:57BagderJess: most of us no longer use the archos series at all ;-)
23:40:04bluebrotherwell, maybe that changed and the manual was forgotten. Will take a note.
23:40:31Jessbag, err... umm... I'm skint! the archos still works, so I'm stuck with it.
23:40:55amiconnYes, and the archos still is a fine machine imo
23:41:04amiconnWith rockbox, that is
23:41:08Bagderit is
23:41:12amiconnWithout it, it's almost unusable
23:41:13Jessdoe it make a decent bootleg with a good mic?
23:41:30*amiconn never used it for recording with an external mic
23:41:41amiconnAny of my targets, in fact
23:41:43wishesrockbox ftw, thanks guys for making it cool :D
23:41:59Jessit wouldl be nice if I could plug into the board, but most shows won't let you.
23:42:05wisheseveryone looks at the ipod and goes 'wtf? that looks cool! what did you do?'
23:42:07 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
23:42:29bluebrotherTOuf: have you installed any other firmware like IPL to the ipod?
23:42:35bluebrotheror repartitioned?
23:42:49TOuf[ERR] No ipods found, aborting
23:42:56DerPapstipodpatcher detects the ipod regardless of iPL
23:43:09TOufno i just formatted the ipod
23:43:25bluebrotherhow did you reformat the Ipod? Does AppleOS still work?
23:43:30DerPapstmaybe you've killed the firmware partition?
23:43:30Jesswell, thx for the info, guys! I'll give it a go. Should I use RB 2.5, or the newest available?
23:43:33TOuflet me check
23:44:12bluebrotherJess: using a current build is suggested.
23:44:13amiconnJess: Use the latest build. 2.5 is hopelessly outdated by now
23:44:30Jessall good. tyvm!
23:45:16 Nick lymeca_ is now known as lymeca (n=lymeca@unaffiliated/lymeca)
23:46:44amiconnJess: With recent rockbox on archos recorder, you can even record as .wav (although it drains battery and is a bit cumbersome to use atm)
23:47:21TOufbooted into appleos, rebooted into disk mode and same error now :(
23:47:28DerPapstdoes it record to mp3 as default?
23:47:48amiconnYes, because that's what the hwcodec supports
23:48:03DerPapstTOuf: and the apple os didn't complain?
23:48:37TOufand the folder inside the ipod are calendars contacts ipod_control and notes
23:49:09DerPapstTOuf: what iPod do you have?
23:49:14amiconnBut the hwcodec can be reprogrammed, and we use the loadable pcm codec (developed by Micronas and Archos) with permission
23:49:30TOuf5th gen 30gb
23:49:36*jhMikeS is having a multithreaded heart attack peeking into the file system. cooperative multitasking doesn't have rules any different from preemtive once a thread "preempts" itself by calling yield which the ata drive does alot of.
23:49:41DerPapstamiconn: cool
23:49:57DerPapstamiconn: i didn't know that archos knows about this project :D
23:50:05TOufi had rockbox installed, wanted to revert from senabs to normal but i messed up with the themes
23:50:16TOufand decided to start everything from scratch
23:50:24amiconnIt does - and at some point even considered distributing rockbox with its products
23:50:39DerPapstwow o.O
23:51:31DerPapstTOuf: and you have resored your iPod with itunes?
23:51:32 Quit haemmy ()
23:51:37amiconnjhMikeS: ?
23:51:50TOufshould i?
23:51:51DerPapsthow have you uninstalled the bootloader then?
23:52:08TOufwith ipodpatcher.exe
23:52:12TOufpressed U
23:52:32 Join kubiixaka [0] (
23:52:52 Join jack|Argon [0] (
23:52:53DerPapstTOuf: unteresting.
23:52:54jhMikeSamiconn: I'm worried about simultaneous write access to files...not even on the same file even.
23:53:20 Quit Jess ("CGI:IRC (EOF)")
23:53:35amiconnjhMikeS: Why is that a problem? The file system is working fine for several years with cooperative multitasking
23:53:42DerPapsthmmm.. unsintallatioin worked so far for me.
23:53:50jack|Argonhi, does rockbox for e200 have FM tuner support?
23:53:53DerPapstthat might be interesting for linuxstb.
23:53:57TOufwtf just downloaded ipodwizard and it says no ipod found
23:54:10DerPapstbut he's drinking lots of beer right now :P
23:54:16TOufshould i reinstall itunes and restore the ipod from there?
23:54:35DerPapstTOuf: i think that would be the easiest + fastest way.
23:54:53TOufoh god i hate itunes :P
23:54:56TOufon it
23:55:04DerPapstdunno what's wrong with your ipod.
23:55:34DerPapsti would say your firmware partition is somehow messed up but the OF still works ^^
23:56:09jhMikeSwriting is rare compared to other activity though but two threads writing might come back with the same free sector for two different files unless cache_fat_sector always "gives" the sector to the thread that called the function. I'm just getting up to speed with this aspect.
23:56:18TOufcrap i got to go
23:56:29TOufi will do what you said and probably gonna come back later
23:56:30jack|ArgonDoes any one know if rockbox for e200 has FM tuner/video playing support?
23:56:31TOufthanks a lot
23:56:43 Quit TOuf ("CGI:IRC")
23:56:58amiconnjhMikeS: How could that happen?
23:57:00DerPapstafaik yes.
23:57:05DerPapstbut i can be wrong
23:57:09amiconnThe only situation where the ata driver yields is spinup
23:58:04jhMikeSnot really. it yields very often, so often recording doesn't even do yield itself when writing.
23:58:47jack|ArgonDerPapst: is there anywhere I can look so I can know for sure, also, does it support all e2XX models?
23:58:59jack|ArgonDerPapst: I'm doing research for a friend

Previous day | Next day