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 2006-12-05

00:00:19 Quit MarcoPolo (Read error: 104 (Connection reset by peer))
00:01:05TommyTottit just start over and over..
00:01:14TommyTottgets to the smileyface error icon
00:02:33TommyTottdid everyone die her??
00:03:07markunTommyTott: it's getting late in Europe..
00:03:15jhMikeSno, not yet :) just don't know what you're talking about
00:03:58TommyTottmy ipod didnt work, then it did, and now it just comes to that "smileyface error screen", and then it reboots, and do the same over and over
00:03:59*jhMikeS wonders what makes a smiley face on error
00:04:09preglowi think it's the ipod sad face
00:04:16TommyTottlike this smiley face :(
00:04:22preglowthat doesn't smile
00:04:22TommyTottyes, a sad smiley.. :P
00:04:37TommyTottyou call all those smileys..
00:04:38jhMikeSand I thought Apple was just sadistic
00:05:07TommyTotti dont think its apples icon.. rockbox i think
00:05:22TommyTotteitherway, i need help
00:05:48jhMikeSI don't have an iPod so not sure
00:06:10preglowit needs to be restorted
00:06:33TommyTottit worked 15mins ago..
00:06:40preglowthen i've no idea
00:06:45preglowand now i really need to sleep too
00:07:05TommyTottsoap: here???
00:07:24jhMikeSpreglow: gnight
00:08:55Kasperlethis one?
00:09:03thegeekis the rockbox bootloader faster than loader2?
00:09:20thegeekfrom video's I've seen of loader2+ipodlinux it takes fucking forever to boot
00:09:26thegeekwhich is really pretty stupid for a dap
00:09:43thegeekrockbox was lightning fast on my h120, and I loved it that way;P
00:09:46Kasperlerockbox boots quite fast with ipod loader 2
00:10:03thegeekthe video I saw was like 3 seconds just for loader2
00:10:29Kasperlehow long does the standard firmware display the apple logo after a hard reboot? ;)
00:10:32thegeekbut ipodlinux itself was horribly slow;P
00:10:55thegeekI dont know
00:11:21Kasperleif you're in loader 2, booting rockbox is fully usable after about 5 seconds
00:11:46TommyTottKasperle: yes, thats the one.. but it worked 20mins ago
00:11:49 Quit bluebrother ("sleep")
00:11:50Kasperlei reckon about 3 seconds of loader2 loading it into memory, max. 2 seconds for rockbox to boot
00:12:13Kasperlei haven't tried the rockbox loader yet, though
00:12:15 Quit charkins ("Ex-Chat")
00:12:25 Quit kubiix (Read error: 104 (Connection reset by peer))
00:12:35 Join kubiix [0] (
00:12:48*Soap is here
00:13:08Soapwhat icon do you see?
00:13:11TommyTottsoap: i now get the sad smiley icon..
00:13:14thegeekthat's not too bad Kasperle
00:13:24SoapTommyTott: that is apple's
00:13:40TommyTottoki.. anyhow.. what can i do??
00:13:48TommyTottit worked 20mins ago
00:13:54KasperleTommyTott: have you tried putting it into disk mode?
00:13:59Soaplet it charge more?
00:14:35 Quit kubiix (Client Quit)
00:14:36TommyTottyes.. it worked, but suddenly, i couldt access the disk in windows.. i took it out, and the icon showed
00:14:42 Join kubiix [0] (
00:15:18TommyTottsoap: i try to charge it, but it just keep turning on and off, and allways show that icon
00:15:49Soappress menu+select to reboot, and menu+play while rebooting to get it into disk mode.
00:16:40Soapand let it charge for well over an hour before attempting anything else.
00:17:04Soapit is easy for the ipod to consume more power than it gets over USB while low in battery in disk mode.
00:17:21Soaps/"disk mode"/emergency disk mode
00:18:14TommyTottwhen i charge in the wall, the sad smiley comes.. but when i charge in the computer, only the battery icon shows..
00:18:31TommyTottbut windows doesnt find it
00:18:58Soaplet's get it charged up some, boot it in to disk mode, then worry about Windows finding it.
00:19:09Soapipods behave very strangely when the battery is dead.
00:19:31 Quit kubiix (Client Quit)
00:19:37 Join kubiix [0] (
00:22:41 Join dau [0] (i=daurn@
00:22:43 Quit daurn|laptop (Read error: 104 (Connection reset by peer))
00:23:40 Quit linuxstb_ ("CGI:IRC")
00:24:26 Join qwertymkonji [0] (
00:25:06 Quit perpleXa ("Leaving")
00:26:59 Quit kubiix ("Miranda IM! Smaller, Faster, Easier.")
00:27:06 Part qwertymkonji
00:28:19 Part LinusN
00:39:24 Quit Arathis ("[rl_bot quit]")
00:45:51 Join MarcoPolo [0] (
00:46:14 Quit heyJulius ("( :: NoNameScript 3.81 :: )")
00:47:31pixelmajhMikeS: nice forum post... (that doesn't look any good in my browser) ;)
00:50:26unleetIm out.
00:50:35unleetcya guys tmr.
00:50:56 Quit unleet ("CGI:IRC (EOF)")
00:52:14 Quit Criamos (Read error: 110 (Connection timed out))
00:54:38Lars_GI wonder what patches I should apply
00:58:47 Join daurn|laptop [0] (i=daurn@
00:59:15 Quit dau (Read error: 104 (Connection reset by peer))
01:00:00 Quit daurn|laptop (Read error: 54 (Connection reset by peer))
01:00:14 Join daurn|laptop [0] (i=daurn@
01:04:10 Quit daurn|laptop (Read error: 104 (Connection reset by peer))
01:04:32 Join daurn|laptop [0] (i=daurn@
01:05:41***Saving seen data "./dancer.seen"
01:08:31 Quit PaulJam (".")
01:08:53 Quit BHSPitMonkey (Read error: 110 (Connection timed out))
01:11:33 Join habana [0] (i=58a10615@gateway/web/cgi-irc/
01:12:04habanahi all
01:12:16 Quit MarcoPolo (Remote closed the connection)
01:13:06jhMikeSpixelma: not in mine either so I edited it :)
01:13:12habanai read with pleasure that sansa can dual boot now ? in this case, i need to compile with Bootloaer option ?
01:13:31pixelmajhMikeS: thank you :)
01:15:56 Join safetydan [0] (i=cbca159f@gateway/web/cgi-irc/
01:16:12Lars_Grockbox compiles real reasy
01:16:33TommyTottpiece of crap ipod
01:16:38habanaplease, sansa lovers ! Tonight, i didn't forget my usb wire ^^
01:17:25Lars_GTommyTott: :P
01:18:23TommyTottLars_G: this piece of sh** wont work again
01:19:26habanaTommyTott: so bloated ? never tested ipood
01:19:53TommyTottwhat does that mean??
01:20:10 Quit excitatory (Read error: 145 (Connection timed out))
01:20:47habanaerr, let me look in a dictionnary :)in my bad englidh, i thought it was a bad ting or badly done
01:21:26TommyTotti might be.. im not so good in english myself
01:21:34 Join excitatory [0] (
01:21:58habanalol, it seems to have a different sense :/
01:24:36TommyTotti dont get it.. first, it went on and off with the "sad smiley" icon.. then, today, it didnt start at all.. then i charged it for a while, and pressed select+menu.. then it started normaly.. i pluged it into the computer, and everything was ok.. Then suddenly, it went back to switching on and off again..
01:25:40TommyTottnow, when charging in the wall, it goes on and off with the smiley, and when plugged into the computer, it only shows the battery icon
01:25:57TommyTottif i disconnect it, it starts the on and off thing again
01:27:23safetydanTommyTott, maybe just leave it charging for a few hours before you try again.
01:28:30*Lars_G bows
01:28:35Lars_Grocklbox rocks
01:28:43Lars_Gwhoever made the build system, you rule
01:28:49TommyTotti will try.. but when connected to the computer, it sometimes reboots and show the smiley, then go back to showing the battery icon
01:29:03 Join barrywardell [0] (n=barrywar@
01:33:35TommyTottpiece of sh**
01:35:47 Quit daurnimator ("Cyas later...")
01:36:01 Join daurnimator [0] (i=daurn@
01:36:27Lars_Gwhat was the patch number for the patch that adds the capability to possition menues in a theme?
01:37:32 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
01:38:20 Join |AhIoRoS| [0] (n=ahioros@
01:43:33 Quit Lars_G ("Leaving")
01:43:41 Join BHSPitMonkey [0] (
01:46:58 Quit |AhIoRoS| (Remote closed the connection)
01:59:25 Join t0dk0n [0] (
02:00:53 Part pixelma
02:02:47habanasansa ping ?
02:03:08 Quit daurn|laptop (Read error: 104 (Connection reset by peer))
02:03:08 Join dau [0] (i=daurn@
02:03:13rasherEh, what?
02:04:38habanajust to be sure to NOT brick my sansa when applying patch from 15.53.34 #
02:06:01barrywardellit works for me...
02:06:15 Join |AhIoRoS| [0] (n=ahioros@
02:06:18barrywardellbut you also need a e200.c for the bootloader
02:06:29habanarasher: Sorry , didnt want to wake you up "<rasher> Hvad"
02:06:48barrywardelland modify a couple of other files
02:07:01habanabarrywardell: hi and thanks
02:07:25habanabarrywardell: other patches ?
02:08:00barrywardellthat's my e200.c
02:09:22 Quit habana (Excess Flood)
02:09:41 Join habana [0] (i=58a10615@gateway/web/cgi-irc/
02:10:12barrywardelland that's a patch for the rest
02:10:21habanaok, stop
02:10:27barrywardellit will ONLY work with a bootloader build
02:10:40habanadont have to apply all these mods
02:10:52habanathought there was only one
02:11:05barrywardellthose 3 will allow you to build a dual boot bootloader on the sansa
02:11:26barrywardellyou could just wait a couple of days until it's added to cvs
02:12:15habanaThank you very much, i take these to play during lunch tomorrow.
02:13:28habanamust say im very impressed by the way you build free softwares and especially rockbox.
02:13:38habanaGreat model
02:18:38 Quit habana ("CGI:IRC (EOF)")
02:22:12 Part t0dk0n
02:25:50 Quit spiorf (Remote closed the connection)
02:25:53 Join HalloweenJack [0] (i=50cad4e0@gateway/web/cgi-irc/
02:26:03 Quit HalloweenJack (Client Quit)
02:29:21 Join Lars_G [0] (n=lars@unaffiliated/lars-g/x-000001)
02:29:54Lars_GQuestion is patch #5899 still useable or used at all? is/will this be included in current cvs?
02:32:49 Quit wooo (Read error: 113 (No route to host))
02:32:53safetydanLars_G, I've no idea if it's still usable. But it's unlikely to be included as a more generic solution (viewports) is what's wanted, or at least that was the talk a while back.
02:33:11*safetydan hasn't been keeping up
02:33:30Lars_Gstill it calls my atention, since the theme I have defined offsets
02:33:36Lars_Gwhich on the last nightly didn't apply
02:35:24 Quit dau (Read error: 54 (Connection reset by peer))
02:35:50 Join daurn|laptop [0] (i=daurn@
02:38:35 Join Marco [0] (
02:43:32MarcoI'm using an ipod 5g/video and the jClix theme
02:43:56Marcobut in order to use that theme, I need to use Julius' jbuild, which happens to be a little old
02:44:18Marcoare there newer versions of the jbuild available?
02:46:26safetydanMarco, if they're not on then there's no newer versions available.
02:46:42Marcois it recommened to patch the source yourself?
02:46:56MarcoI'm not sure if applying a few month old patch to a current CVS version is a good idea
02:47:39safetydanMarco, if you have the ability to fix any errors or patch failures that come up, you can do what you like.
02:48:00linuxstbI'm only guessing, but if there are no new versions of a custom build, that often means the included patches don't work any more with current CVS.
02:48:40linuxstbBut I'm sure the answer is somewhere in that 30-page forum thread about Julius's build...
02:50:13 Quit daurn|laptop (Read error: 54 (Connection reset by peer))
02:52:01barrywardellhabana: if you see this tomorrow, here's a single patch with all my previous changes combined.
02:52:25 Join daurn|laptop [0] (i=daurn@
02:55:31 Join habana [0] (i=58a10615@gateway/web/cgi-irc/
02:59:33barrywardellhabana: here's a single patch with all the changes:
03:00:09habanaoh tahnks, i was waiting in the other window
03:01:39habanaso i re "cvs update -dP" and "patch < diff" and can make nice dreams ?
03:03:08habanalatest tip so i could patch nicely ?
03:05:11 Quit rasher ("leaving")
03:05:44***Saving seen data "./dancer.seen"
03:09:28barrywardellmake sure you have a decrypted and header-stripped OF in
03:09:54barrywardell /.rockbox/OF.bin
03:11:44habanayou talk chinese to a french not triilingual and not speaking chinese.
03:12:09habanaCould you give me a single command line to apply for that ?
03:13:21barrywardellyou need to use mi4code to decrypt a sandisk PP5020.mi4
03:14:29habanaeasy ! Badger gave me the tip "wget && gcc -o mi4code mi4code.c -lgcrypt"
03:14:54barrywardellyeah, then './mi4code decrypt -s PP5020.mi4 OF.bin'
03:15:19barrywardellassuming you have a PP5020.mi4 from Sandisk in your current directory
03:16:04habanaok done.big tahnks
03:17:04 Join bro2baseball [0] (
03:17:13habanafor other reader : ./mi4code decrypt -s PP5022.mi4 OF.bin
03:18:09barrywardelloh yes, sorry
03:18:23habananope !
03:20:13 Join Kittt0s [0] (i=Kaa@
03:20:22 Quit Kitt0s (Read error: 145 (Connection timed out))
03:23:46 Quit Lars_G ("Leaving")
03:23:50habanamake in progress. big up for barrywardell !
03:24:04habanaFile encoded successfully
03:24:59barrywardellanybody with an ipod want to test that patch with the new crt0-pp.S?
03:27:31 Quit habana ("CGI:IRC (EOF)")
03:31:42bro2baseballwhat is the crt-pp.S exactl
03:34:56barrywardellit's the first bit of code that's run
03:35:30 Quit Rudy4Pez (Read error: 110 (Connection timed out))
03:35:55 Quit daurn|laptop (Read error: 54 (Connection reset by peer))
03:35:59 Join daurn|laptop [0] (i=daurn@
03:36:14 Quit funky ("leaving")
03:37:54 Part bro2baseball
03:45:40 Part safetydan
03:45:51TommyTottokey.. now i have charged my ipod for some time.. What shall i do?? the problem was that it started, but showed the "sad face" icon, then restartet, and did it over and over.. s what shall i do now??
03:59:10 Quit barrywardell ()
04:06:13 Nick Soap is now known as SoapIsOutOfTown (n=Soap@rockbox/staff/soap)
04:09:30 Quit |AhIoRoS| ("Abandonando, see you")
04:25:18 Quit niskel (Read error: 104 (Connection reset by peer))
04:27:09 Join dau [0] (i=daurn@
04:27:10 Quit daurn|laptop (Read error: 104 (Connection reset by peer))
04:27:38JdGordondoes anyone know if the player allows custom keyboards?
04:29:47 Join daurn|laptop [0] (i=daurn@
04:29:48 Quit dau (Read error: 104 (Connection reset by peer))
04:34:24 Quit jadakiss ("w00t!")
04:35:12*JdGordon thinks it doesnt
04:36:41 Join bender02 [0] (
04:37:47 Join webguest17 [0] (i=46b2299b@gateway/web/cgi-irc/
04:38:03webguest17can someone help me in here?
04:38:25 Quit webguest17 (Client Quit)
04:38:34JdGordonwe have a winnr!
04:38:51 Join webguest68 [0] (i=46b2299b@gateway/web/cgi-irc/
04:38:55webguest68i need some help
04:38:59webguest68can someone help me?
04:39:26JdGordonjust ask your damn question
04:39:26 Quit webguest68 (Client Quit)
04:44:06TommyTottokey.. now i have charged my ipod for some time.. What shall i do?? the problem was that it started, but showed the "sad face" icon, then restartet, and did it over and over.. s what shall i do now??
04:44:12TommyTottthats my question
04:47:26 Part bender02 ("[IRSSI]")
04:48:25 Quit daurn|laptop (Read error: 104 (Connection reset by peer))
04:48:27 Join daurn|laptop [0] (i=daurn@
04:50:34 Join jrockway [0] (
05:04:28 Join dau [0] (i=daurn@
05:04:29 Quit daurn|laptop (Read error: 104 (Connection reset by peer))
05:05:47***Saving seen data "./dancer.seen"
05:10:10 Nick myzar|away is now known as myzar (n=myzar@
05:19:38 Quit dau ("dau was using Daurnimator's Legendary Script")
05:47:45 Join dantheman [0] (n=danielma@
05:54:42 Join BHSPitLappy [0] (
05:57:27 Join nave7693 [0] (
06:00:44 Quit jaebird ("Leaving")
06:02:40 Join lostnihilist [0] (
06:03:34 Quit TommyTott ()
06:08:15 Join jaebird [0] (
06:16:28 Quit lostnihilist (Remote closed the connection)
06:21:19 Part nave7693
06:26:29 Join lostnihilist [0] (
06:59:22 Join webguest07 [0] (i=ccb977f2@gateway/web/cgi-irc/
06:59:38webguest07i need help
07:00:54webguest07serious help
07:01:13webguest07my 4g ipod only shows the apple on the screen and it doesn't mount or turn off
07:05:49***Saving seen data "./dancer.seen"
07:09:12 Quit webguest07 ("CGI:IRC")
07:11:10 Quit XavierGr (Nick collision from services.)
07:11:13 Join XavierGr [0] (
07:19:46 Join Rudy4Pez [0] (
07:24:14 Join petur [0] (n=petur@rockbox/developer/petur)
07:25:23 Quit Kasperle (Read error: 110 (Connection timed out))
07:26:33 Quit tychver ("Leaving")
07:26:38 Quit slarti (Connection timed out)
07:27:50 Quit dantheman (Success)
07:39:11 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
07:46:08SlasheriLinusN: morning, i got the unit back and now it's working great, thanks :)
07:46:16LinusNSlasheri: goodie
07:46:26Slasherihowever, it was flashed with the wrong h-115 firmware and no rockbox at all :D
07:46:42LinusNas you can see, I left the bdm connector in place, just in case
07:46:49Slasherihehe, yeah ;)
07:46:53Slasherinice :)
07:47:07LinusNand, yes, I didn't bother to put rockbox on it
07:47:07Slasheribut now i try not to brick it again
07:47:24LinusNh115? oops :-)
07:47:28Slasheriyep :D
07:47:44Slasheribut i managed to cross-patch it with a working h120 bootloader :)
07:48:11LinusNsorry about that
07:48:37Slasherinp :) probably i just would have need to send it back if flashing would failed :)
07:48:40LinusNwas in fact a long time since i had to unbrick a player
07:49:07Slasheriand now even my new bootloader seems to be working fine
07:49:38*LinusN hears toni braxton sing... "unbrick my daaaaap"
07:50:12 Quit jba (Read error: 110 (Connection timed out))
07:52:09LinusNok, so your bootloader works now?
07:53:54Slasheriyep :)
07:54:10Slasherinow supporting both ram and rom images, and reset cookie for the rom image also
07:54:29Slasheriand simple menu activated with the hold switch and when rockbox crash
07:54:47Slasheriusing that it's possible to select the default boot method when original fw is no longer present
07:57:05 Join Kasperle [0] (
07:57:23LinusNwith the hold switch? how?
07:59:14Slasheriactivating hold switch and then turning on player activates the bootloader menu (unless usb cable is present, then usb mode is activated). The menu has a 15s timeout after poweroff is issued
08:00:30Slasherior without hold switch activated and rockbox has previously crashed, menu is also activated but after the timeout default boot method is executed
08:00:32 Join MarcoPolo [0] (
08:01:44Slasherihowever, if original fw is still present, that menu is never activated and old rec-button combination works as before
08:02:01 Join ender` [0] (n=ender@
08:02:16LinusNehum, what if you try to use the hold switch to prevent it from turning on?
08:02:42Slasheriit always shuts off after 15s of inactivity
08:02:53Slasheriand there is no activity when hold switch is on
08:03:03LinusNah yes
08:04:47Slasheriwithout hold switch and cleanly shutdown and default boot method either rom or ram, all hardware is not initialized at all and rockbox is being booted as soon as possible
08:04:50LinusNdoes it still work on the h300?
08:05:06Slasherishould work fine
08:05:14Slasheribut i have no chances to test that :)
08:05:21LinusNi will
08:05:21Slasherii could provide a patch soon
08:05:24Slasherigood :)
08:05:41LinusNsince i have toni braxton by my side ;-)
08:06:23 Quit petur ("sssssssssss---------PLOP!")
08:07:03Slasheriwith the rom image, it's nice that you could just tap the play button twice to activate the resume :)
08:07:22Slasherithere is almost no need to wait, something like 500ms or less is enough between the presses
08:08:11LinusNooooh, nice
08:08:42LinusNhow much with the ram image?
08:09:58 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
08:10:38Slasherithat's not much either, should be almost same. i can do some measurements soon
08:13:21LinusNi can try it myself
08:13:40LinusNis all the other stuff (flashing code etc) in cvs?
08:16:20Slasheriyes :)
08:17:10Slasheriwell, of course the patch also changes and Makefile a bit to activate rom image generation for irivers
08:20:04SlasheriLinusN: btw, you could already try the ram image flashing. However, with the new bootloader loading should be faster
08:22:12LinusNspeaking of flashing, i have been thinking about an "emergency unpatch" feature in the loader
08:22:46LinusNthat is, remove the bootloader somehow
08:23:09LinusNi.e restore the reset vector
08:23:19Slasherihmm, i had that also in mind long time ago, but amiconn and others think flashing functions in the core are not a good idea
08:23:26Slasheribut maybe bootloader could be different
08:23:30LinusNi agree about that
08:24:21LinusNi don't want flashing code in the core either
08:24:57LinusNso i don't know how to solve this :-)
08:25:02Slasherihehe :D
08:25:23LinusNthere are of course ways to protect against accidentally running the flash code
08:25:34Slasherithat would be useful indeed if needing to send a unit to repair and remove the bootloader
08:25:57Slasheribut the could would need to be on flash because disk might be crashed
08:26:38LinusNwe could for example have the flashing code scrambled in flash and have it always run from ram, unscrambling it as we copy it to ram
08:26:50Slasherihmm, true
08:27:10 Join nave7693 [0] (
08:27:12LinusNbut making it that complicated will also increase the risk of bugs
08:27:50 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
08:28:25LinusNi like the cowon bootloader approach
08:28:41LinusNalmost unbrickable
08:35:09amiconnFlashing code can't be executed from rom anyway
08:36:56LinusNtrue, true
08:37:02 Part nave7693
08:37:20LinusNamiconn: but we could still make sure it can't be executed in-place
08:38:03LinusNwith several mechanisms, like the scramble-copy method and also with some global "key" variables
08:39:34amiconnAs soon as flashing code switches the rom to programming mode, it will crash
08:40:07LinusNyes, if you execute from rom
08:40:40LinusNso yes, we can probably put flashing code in flash quite safely
08:40:54LinusNi.e in the bootloader
08:41:34LinusNhowever, the h300 is a little nasty in that respect, since the flash reset pin doesn't seem to be connected to the main reset
08:43:06 Quit _Veseliq_ (Read error: 113 (No route to host))
08:43:35peturLinusN: if you manage to find some time, could you check/verify how VBUS and ID of the usbotg connector of the h3x0 are connected?
08:43:54LinusNok, i'll try to remember that
08:47:33JdGordonhey, whats the story with all the wps' on flysrpay?
08:49:20LinusNJdGordon: we once wanted people to submit wps's to flyspray for inclusion in cvs
08:49:49peturJdGordon: any idea on ?
08:50:48JdGordonLinusN: are the just goin to be left? or do we eventually dump/commit them?
08:50:59JdGordonpetur: i havnt noticed any problems... ill test now tho
08:52:43LinusNbtw, we could always protect the flash with the breakpoint logic
08:53:59JdGordonpetur: i just disabled DC, rebooted, re-enabled, rolod and didnt get the splash screen (tho i expected to!)
08:54:13peturdon't rolo
08:54:22peturalways powercycle
08:54:31JdGordonok, starting again
08:55:41peturjhMikeS: I think you got the file ok?
08:57:39LinusNJdGordon: yes, i think we could reject the wps'es in the tracker
08:57:41jhMikeSPetur: yup
08:58:07LinusNi'd like a default wps with more "rockbox" feeling to it
08:58:11JdGordonLinusN: I ask, because there are heaps of them, and removing them would make the open patches number much smaller :)
08:58:15 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
08:58:33peturduh.. that targa 30GB PDA I talked about earlier is next week available for 149 euro... runs linux, sources available... tempting
08:58:54LinusNi like icatcher a lot, but i'd still like a simpler one with less frills and more "rockboxish" whatever that is
08:59:24*JdGordon back soon, dinner
08:59:33petur (in Dutch but the specs are understandable - the screen spec is wrong imo)
09:01:00Slasherii like llamedbeta2 for H1xx
09:01:13Slasheriit's not even in cvs :/
09:05:53***Saving seen data "./dancer.seen"
09:06:04SlasheriLinusN: i will now generate a patch for you, assuming my latest testing works :)
09:06:59peturjhMikeS: strange post on taperssection, have copied it here: - don't have an h1x0 so I can't check that but it seems unbelievable
09:09:14amiconnpetur, jhMikeS: My first (and probably only) guess would be the wspll thing
09:09:54amiconnNot sure how that should result in hdd noise being recorded, but the UDA _is_ a strange beast
09:10:22jhMikeSpetur: saw that and I'm going wtf?
09:10:30peturI've never seen more than amplitude changes on hdd activation
09:11:26amiconnBtw, and cause I ranted about the noisy uda output earlier - seems this is a problem of the integrated headphone amp. Line out is ok
09:12:35jhMikeSI would think corrupting the data in the digital circuitry would make for something really nasty and loud but not a particular noise and why would the HD be louder? :\
09:13:05LinusNwell, the hdd is probably the biggest source of interference
09:14:28jhMikeSI take it CMC8s means using an external mic?
09:14:39jhMikeSright, of course
09:15:58peturmaybe the impedance of the mics is a lot different from the ones we use to test...
09:16:16jhMikeSI'm not hearing anything wacky myself though. Will test the mic that came with it.
09:16:17peturstill that couldn't explain the difference when running OF
09:17:41jhMikeSnot doing wspll for recording is an easy switch if that makes jet engine sounds
09:18:10peturI can always ask him to try something
09:18:27jhMikeSleave the airport?
09:19:56 Join spiorf [0] (
09:22:07 Join forehead [0] (i=54bd6183@gateway/web/cgi-irc/
09:22:18 Join miepchen^schlaf [0] (
09:22:32jhMikeSI'm gettin' nothing myself. everythings quite quiet
09:23:36peturI'll give hime the 'unable to reproduce' - maybe others in the forum will test too... maybe his h120 is faulty?
09:26:20jhMikeSI just turned up the gain so high the headphones got feedback (ouch!) and I could hear the neighbors whispering and put the mic right against the side of the unit and the spinup still only measured about halfway up the scale.
09:27:21peturwell nevermind then, I've asked for others to try and confirm
09:27:38 Quit forehead ("CGI:IRC (Ping timeout)")
09:29:24 Join kubiix [0] (
09:29:26jhMikeSI'll give him two builds, one with WSPLL enabled, the other without and ask him to say which one gives the noise and which doesn't. Wish I could double blind that and hand them out randomly.
09:29:57peturoh goodie
09:30:51jhMikeSWith just one trial he's got a 50% random hit so it not so good. Take one sans wspll and tell him it has it (placebo control)
09:32:00peturAnd iff I name them builds A and B >
09:32:27 Quit kubiix (Client Quit)
09:32:28LinusNbtw, could we try to detect buffer underruns and warn the user?
09:32:33 Join kubiix [0] (
09:33:12peturbuild-in quality control - I'd rather have *no* underruns :/
09:33:51LinusNof course, but it could help us find the root cause
09:34:06peturit would also make testing easier...
09:34:24 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
09:35:29JdGordonpetur: re 6408, adding a call to settins_save() when dircache finishes initiing wold probably fix the problem
09:36:01peturis there any good reason why shutting down from within the menu doesn't save settings?
09:37:42JdGordonbut thats not the issue...
09:37:51peturI know
09:37:54JdGordonadding settings_save() to then end of main() would work
09:38:01jhMikeScould just have him do a take in OF and RB starting with a reference signal to check for equal gain.
09:38:31jhMikeSunderruns? who's got the underruns?
09:40:34JdGordonpetur: hmm... crap... it didnt stop the scanning splash
09:42:12jhMikeSrecording can never go famine only flood except with spdif errors. a DMA error would terminate it on the spot.
09:43:20 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
09:43:34 Join Kapez [0] (
09:45:15peturcan't hold off work any longer ;) Later!
09:45:21 Quit petur ("worrrk")
09:48:00 Join miepchen^schlaf [0] (
09:48:45 Join austriancoder [0] (n=austrian@
09:48:56jhMikeSLinusN: were you talking about playback or recording re: underruns? *confused*
09:50:59LinusNi meant overruns of course
09:51:49JdGordonoh, is recording to mp3 safe with cvs? i have to record a lecture tomorow
09:51:50jhMikeSah...could loud enough vibration (concert hall) cause a drive to not write
09:52:36jhMikeSI just did a 2h 14m test and had no problems. mp3 should work all the same.
09:52:48SlasheriLinusN: ready :)
09:53:30Slasheriand you need also unless you change the crc values in iriver_flash plugin
09:55:22Slasheriplease try it on h120/140 also to see if it works
09:55:30jhMikeSI can only imagine a flood situation if the drive fails to write fast enough. Even the most demanding encoder (WavPack) at high samplerates can flood the output but the thread prio boost takes care of that almost instantly.
09:55:41markunLinusN: On my Gigabeat HDD spindown halts Starfield (just an example) for maybe half a second. What could be the problem? My H120 doesn't have this problem.
09:55:53 Join ProdigySim [0] (
09:55:58Slasherih110 is not yet supported, unless you try that too
09:56:46LinusNmarkun: that's odd - i wonder why it does that
09:57:36markunLinusN: if because of the font cache the hdd has to spin up again there is another short halt
09:58:11 Nick Kapez is now known as Xyh (
10:00:00jhMikeSDo most people have a decent experience with Cowon repair service? I fear they're gonna give me a runaround since they don't even let you talk to someone live but make you fill in a "ticket". I'm worried about being without the unit for too long.
10:00:49thegeekLinusN: what do you think about the 80gb 5.5g ipod?
10:02:55 Quit kubiix ("Miranda IM! Smaller, Faster, Easier.")
10:04:35LinusNthegeek: haven't had time yet to work on it
10:04:59thegeekah, I meant what do you think of it as a dap in general;P
10:05:25LinusNah, well, it doesn't look much different from the 5g
10:07:16thegeekbtw, I have a dead h120
10:07:29thegeekthe ata/usb bridge is fucked
10:07:32thegeekdo you need it?
10:07:45thegeekit's just laying around atm
10:07:45Slasherithegeek: hmm, how is it broken?
10:07:55thegeekwell, I tried to fix it;P
10:08:10thegeekone of the daughterboard connectors are bad
10:08:22thegeekcompletely b0rked now that I tried to fix it;P
10:08:45crwllbtw, are those "bricked-by-nokia" h100's always completely FUBAR?
10:08:52Slasherihmm, you borked solder inside it or something like that? =)
10:09:08 Join jba [0] (
10:09:26Slashericrwll: afaik, most are unless you try and replace the cpu
10:09:37thegeekSlasheri: it was broken already
10:09:43Slasheriyou need some special equipments however to unsolder the cpu
10:09:57 Nick crwll is now known as crwl (
10:10:11Slashericrwl: here in jyu we have some equipments to do that ;)
10:10:18thegeekafter my repair attempt it is certainly not better;P
10:10:18 Join forehead [0] (i=54bd6183@gateway/web/cgi-irc/
10:10:33crwlSlasheri, my h120 is fine but I know someone whose unit isn't :)
10:10:43LinusNcrwl: from my limited experience, the nokia-fried units could be salvaged unless you have tried to power it on with the charger attached
10:10:54 Quit Xyh ("PJIRC @")
10:13:17crwlLinusN, with the nokia charger or the correct charger?
10:13:37Slashericrwl: if you need a cpu replacement, i might be able to do that but it costs something like 20e for the new cpu and might fail
10:15:20 Quit forehead ("CGI:IRC (Ping timeout)")
10:18:47 Quit markun (Read error: 104 (Connection reset by peer))
10:20:41 Join markun [0] (n=markun@rockbox/developer/markun)
10:21:19LinusNSlasheri: soldering and unsoldering bga is surely tricky
10:22:12Slasheritrue, it requires some heat pistols at least
10:22:24LinusNyes, and patience
10:22:38scorchewhat is so tricky about it?
10:23:05LinusNthe pads are not visible, so you have to heat the chip with hot air
10:23:22scorcheoh fun
10:23:36jhMikeSjust send it to me and I'll talk to it :)
10:23:42LinusNand it can be a bitch to make all pads connect when you solder the new cpu back
10:23:55scorchewhy dont you just reflow?
10:24:35amiconnmarkun: Sounds like a threading issue. Can't think of anything else
10:24:50amiconnI wonder why the gigabeat code isn't in cvs yet
10:24:56markunamiconn: do you have some tips on how to debug?
10:25:16markunyes, should start to commit it
10:25:27LinusNscorche: well, that is basically what you have to do
10:25:38LinusNbut with a heat gun instead of an oven
10:25:59scorchei have had great results reflowing with a skillet
10:26:29 Join forehead [0] (i=54bd6183@gateway/web/cgi-irc/
10:27:24LinusNscorche: hehe, nice idea
10:27:50scorcheit works well =)
10:27:54LinusNthat would probably work if the pcb weren't already mounted on both sides
10:28:03scorcheand is quick and cheap
10:29:41crwloh, in fact...
10:29:44crwltipi^, listen up!
10:30:48 Join barrywardell [0] (n=barrywar@
10:31:35 Quit forehead ("CGI:IRC (Ping timeout)")
10:33:35 Quit spiorf (Remote closed the connection)
10:39:41JdGordoncan I force a settins_save before shutdown?
10:40:11JdGordonis there a reason we dont save before shutdown now?
10:43:03amiconnThe ata callback is called before shutdown, and that should save settings
10:43:13 Join pondlife [0] (n=Miranda@rockbox/developer/pondlife)
10:43:25JdGordononly if sometihng called settins_save before shutdown
10:43:54LinusNJdGordon: why call settings_save() if nothing has changed?
10:44:15JdGordonbecause something is being very wierd with the settings...
10:44:24LinusNno need to spin up the hard drive if it isn't necessary
10:44:30JdGordonre fs 6408
10:44:55LinusNthen i'd say that there is a missing settings_save() somewhere
10:45:04LinusNbut it should not be in the shutdown code
10:45:10amiconnI guess we should change the settings save handling a bit
10:45:17*jhMikeS thinks he'll design a pre-power supply for his H100 for safety (to protect against others). there's room to sneak another tiny board in there?
10:45:22JdGordonit does seem to be callin the settings flush callback, but the dircache settings doesnt always actually work
10:45:43amiconnWe used to only call settings_save when leaving the menu, but then we did so unconditionally
10:45:55JdGordonthat still happens
10:46:14LinusNJdGordon: sounds like a bug that shouldn't be solved by putting a settings_save() call in the shutdown code
10:46:15JdGordonthe change is moving the actuall writing into a callback, but tha seems t work, so i dont know
10:46:40amiconnWe could call it immediately after leaving an individual setting, and the advantage would be that we could do so only if the value actually changed
10:46:58LinusNwe want as few saves as possible
10:47:35LinusNjhMikeS: there's not much room in there
10:48:00LinusNand installing your device is not that easy, since you have to separate the two boards
10:48:07amiconnLinusN: settings_save wont actually write to disk until the next spinup, or until shutdown
10:48:34LinusNi know, but i don't want it to spin up when i shut down
10:48:59LinusNnot if there aren't any changes
10:49:54amiconnThat's why I suggest that method
10:50:03 Part austriancoder ("Kopete 0.12.3 :")
10:50:32amiconnIt both ensures that even a single change will get saved, but nothing gets written if there was no change
10:50:41amiconnWith the current method, we have 2 problems:
10:50:42LinusNJdGordon: logf is your friend :-)
10:51:04JdGordonno its not.... damn remote display is too small, so its too easy to miss the event
10:51:08amiconn(1) Nothing gets saved at all if someone enters the menu, changes some settings, and then powers down
10:51:16LinusN1) it doesn't get saved if you don't leave the menu before you shut down
10:51:18amiconn..without first leaving the menu
10:51:34LinusN2) it saves even if nothing has changed
10:51:38jhMikeSthere should be a global set of states and a dirty flag for settings, no? How bout one for UI busy too so we can display a wait at certain points and eat button presses?
10:52:03amiconn(2) There is a useless save if someone enters the menu, enters some settings and then leaves the menu without actually changing a settings value
10:52:15JdGordonwe could hash the global_settings mem and save if the hash changes..
10:52:19JdGordonbut thats overkill
10:52:23amiconnjhMikeS: You don't need an extra flag.
10:52:36amiconnJust registering the callback _is_ that flag
10:53:49jhMikeShmmm...and it remembers that for later?
10:53:55markunamiconn's suggestion looks nice and simple to me
10:55:06 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
10:55:26amiconnjhMikeS: The callback will be called at next spinup, or on shutdown/usb connect, whichever comes first
10:55:49amiconnIt will be unregistered after being called
10:57:06jhMikeSI'm trying to be lazy about it and allow a query from anywhere in a consistent manner. Of couse the callback could be a member of "the big board". NULL or non-NULL would be just as good.
10:57:25amiconnWhy do you need the ability to query?
10:59:23amiconnIf a setting actually changes, the callback for saving settings will be registered. This is completely self-contained.
10:59:39jhMikeSnot just for that, but a central place where the state in general is kept even for things like what screen is current, audio status, and more.
11:03:27jhMikeSIf you have that you can say mark the UI busy and buttons would repond accordingly and perhaps even a message/icon automatically displayed until it's available again.
11:05:18jhMikeSWhen voice reloads, things seem deadlocked until the drive spins up and it swaps the bits. If you push keys at that time you get an explosion of buffered presses. Just a whacked out idea I have. :\
11:05:54***Saving seen data "./dancer.seen"
11:06:29 Join miepchen^schlaf [0] (
11:08:04amiconnjhMikeS: I actually like the feature that rockbox buffers the keypresses
11:08:23amiconnThe archos firmware didn't, and was really cumbersome to operate because of this
11:09:18amiconnThe bitswap on swcodec should be gone soon, if I fiind the time to actually apply langv2 (to lang and voice) properly :/
11:10:29 Join bluebrother [0] (i=n99bOXfj@rockbox/staff/bluebrother)
11:12:07 Quit barrywardell ()
11:13:38jhMikeSThere's still the wait time involed for spinup and it appears locked during that. I wonder if some kind of auto detect is possible. Small viewports would be just the thing to have to implement it on the UI end. (I'll avoid using a tumbling hourglass :)
11:13:46Slasherihmm, maybe i will add backlight on/off at boot setting to the eeprom also. that way rockbox could fade in the backlight on boot when loading from rom :)
11:16:25jhMikeSHow bout having all the display settings available to the bootloader? Would also like to not initialize the pmu or lcd more than once either but rolo needs things reset right?
11:17:03amiconnThe bootloader shouldn't be unnecessarily complex
11:17:15LinusNamiconn: i agree
11:17:17amiconnAbove all it needs to be reliable
11:17:55LinusNi'd like it to work much like the cowon loader
11:18:29SlasheriLinusN: btw, did you try the patch yet? :)
11:18:45Slasheriamiconn: that would be only one if on boot
11:19:17jhMikeSmany of the lcd features should be disabled in a bootloader build then (esp x5 with the on off stuff). probably could trim a lot out of stuff.
11:21:34LinusNSlasheri: not yet, no time - at work
11:22:49Slasheriok :)
11:24:14jhMikeSCan't queue_wait_w_tmo take TIMEOUT_BLOCK? If it could many multiple function calls could be elimated. I'm I even being relevant on topic anymore? :) *tired and ready for bed*
11:26:08SlasherijhMikeS: then you want to use queue_wait
11:26:10Slasherithat always blocks
11:30:01jhMikeSI know but there's a lot of if (this) queue_wait else queue_wait_w_tmo, could just use one function
11:30:53jhMikeSmore like the way get_action works, just specify block, no block, or a timeout
11:31:10safetydanis __m68k__ defined for coldfire builds?
11:34:53 Join spiorf [0] (
11:35:23SlasherijhMikeS: hmm, sounds weird
11:40:02 Quit JdGordon ("Konversation terminated!")
11:40:07amiconnsafetydan: Why do you need that?
11:40:46safetydanamiconn, some existing code with m68k asm uses that define to decide whether to compile the C or asm version
11:41:46amiconnSimply check if it is, I'd say
11:41:55safetydanit is defined
11:42:09safetydanbah, now I remember why I got stuck
11:42:17safetydanError: invalid instruction for this architecture; needs 68000 or higher −− statement `addq.w #1,%d1' ignored
11:43:38amiconnYeah, coldfire doesn't support the whole m68k instruction set
11:44:10amiconnNamely, arithmetic instructions are only supported in the xxx.l form
11:44:29amiconnAt least that part is RISC
11:45:42 Join hickery [0] (
11:53:54PikccAside from better customisation of sound and being able to read MP3s from disk without a .db, what are the other reasons to use Rockbox on an iPod?
11:54:32Kasperleyou can have games and other plugins on the ipod, plus, you could write your own apps for it
11:55:04Pikcchm... I saw a feature mentioned on the forums, how would that work?
11:55:16amiconnBetter codec support (ogg, flac, wavpack to name a few)
11:56:38bluebrotherPikcc: already seen ?
11:58:04hickeryAnyone answer iRiver/H120 questions ?
11:58:21bluebrotherjust ask
11:58:38bluebrotherI don't know if I can answer it unless I know the question ;-)
11:58:42hickeryI just wanted to know if anyone had worked out the gain settings the iRiver firmware uses for recording levels, and thus what they might equate to in Rockbox
11:59:16 Quit jba (Read error: 110 (Connection timed out))
11:59:26hickeryAnd second, what gain iRiver uses when you select External Mic as the source
11:59:39bluebrotherno idea. I rarely use recording
12:09:06 Join _Veseliq_ [0] (n=veseliq@
12:12:00 Join thomasg_ [0] (
12:12:11 Quit Pikcc (Read error: 60 (Operation timed out))
12:17:10 Join funky [0] (n=repulse@unaffiliated/funky)
12:18:30 Part LinusN
12:28:17 Quit thomasg (Read error: 110 (Connection timed out))
12:31:31 Quit blue_lizard (Read error: 110 (Connection timed out))
12:37:16 Join blue_lizard [0] (
12:43:10Slasheriweird.. with rombox i still seem to have the problem that set_cpu_frequency(CPUFREQ_NORMAL); at very start crashes rockbox
12:43:17Slasheriwithout that call everything works fine
12:43:24Slasheriamiconn: any ideas of that?
12:52:35 Quit safetydan ("Ex-Chat")
12:54:19XavierGrI am amazed....
12:59:14markunXavierGr: ;-)
12:59:53markunDo you think this is an actual IRC log?
13:02:42 Join Jagan81 [0] (
13:05:31Jagan81hi. the french/EU volume-limit isn't in effect when running rockbox on iPods which has this, right?
13:05:56***Saving seen data "./dancer.seen"
13:06:33 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
13:16:04 Join lee-qid [0] (
13:16:34amiconnSlasheri: Not from just looking at the source.
13:16:35 Quit miepchen^schlaf (Read error: 113 (No route to host))
13:16:52amiconnWhere is that call that crashes?
13:17:01Slasheriamiconn: apps/main.c
13:17:35Slasherithat is really weird..
13:17:46Slasheriinside bootloader at the beginning it works ok
13:18:04Slasheribut not from rockbox running from flash
13:18:11Slasherihowever, running from ram it still works fine
13:18:36Slasheriand without that call everything works fine, cpu boosting also
13:18:43 Join anathema [0] (
13:18:56 Join Criamos [0] (
13:23:01 Join miepchen^schlaf [0] (
13:28:17Slasherihmm, and that cpu_boostid thing seems to write outside the buffer boundary..
13:28:27Slasherii would like to remove that whole thing
13:29:53Slasherior maybe not.. but still it seems suspicious
13:32:34markunhi hsaliak
13:33:15amiconnSlasheri: Speaking about boosting - I would prefer to know why rockbox is boosted after boot if voice ui is enabled
13:33:38Slasheriamiconn: hmm, that is happening on all targets?
13:33:45amiconnSlasheri: Does that call crash completely, or 'just' freeze?
13:33:53 Quit actionshrimp (Read error: 110 (Connection timed out))
13:34:09amiconnYes, I can observe it on all my swcodec targets
13:34:14Slasheriamiconn: i am not sure, it might just freeze in the while loop or something like that
13:34:50Slasheriweird.. i will try to rember look that later
13:35:08hsaliakhi markun
13:35:47Slasheriamiconn: hmm, might be due to lostlogic's improvements to the scheduler
13:35:54Slasheriso it no longer unboosts voice thread properly
13:36:12amiconnThe boost id just shows '?'
13:36:32 Part hsaliak
13:38:17petursearching the forum gives database error, is this known already?
13:38:54 Join ScoTTie [0] (i=skotty@unaffiliated/scottie)
13:39:06 Quit ScoTTie_ (Read error: 54 (Connection reset by peer))
13:39:48 Quit markun (
13:40:12Slasheriamiconn: correct, scheduler triggered boosts are not tracked
13:40:51Slasheriamiconn: the problem is probably that when scheduler blocks a task, boost state is kept. Only when sleeping, boost state is cleared
13:41:07Slasheriso if voice thread blocks while it's has boosted the cpu..
13:42:24Slasherinot sure why lostlogic implemented it that way
13:42:47NJoinmarkun [0] (n=markun@rockbox/developer/markun)
13:42:48 Quit markun (
13:42:48NJoinmarkun [0] (n=markun@rockbox/developer/markun)
13:42:50*amiconn can't say he likes that auto-boosting stuff
13:43:17preglowSlasheri: because blocking might end any second
13:43:29preglowSlasheri: and he doesn't want a short block to unboost, since that'll lead to a lot of boost/unboosting
13:43:50*preglow likes anything that leads to him not having to remember to unboost manualle
13:44:48Slasheripreglow: hmm, true. But what might cause short blocks?
13:45:18preglowSlasheri: can't think of anything
13:45:24preglowSlasheri: but i know that is the reason
13:45:35preglowi asked him myself a while ago
13:45:45Slasheriand those blocks need to be really short..
13:46:39Slasheriah, mutexes at least
13:46:55Slasherii liked more the idea cpu is boosted only after all threads are sleeping
13:47:13Slasherinow the boost is controller per thread
13:48:18amiconnEven unboosting while all threads are sleeping isn't always wise
13:48:32preglowbut it very often is
13:48:34amiconnOne or more threads might sleep for a very short time
13:48:45amiconn...needing full speed afterwards
13:48:55preglowno, i'd rather have it done automatically than mess around with unboost bugs
13:48:58preglowremember how many of those we had?
13:49:21Slasheriand code is a _lot_ simpler now
13:49:23preglowi think we had a couple that never went away
13:50:13Slasheriamiconn: such a task should manually cpu_boost/unboost
13:50:31preglowSlasheri: can that work without the scheduler interfering?
13:50:57Slasheripreglow: manually boostin/unboosting always works
13:51:17preglowbut in that case, all is good!
13:51:24Slasherischeduler knows also its own global boost state so it wont intefere with that
13:52:27preglowafter starting to use c++ properly, i can really see the value in having things done automatically
13:52:34preglowbut since c doesn't have destructors, this'll do nicely
13:53:38 Join JoeBorn [0] (i=rootmeis@
13:56:38 Quit JoeBorn (Client Quit)
14:00:48 Quit markun (Read error: 104 (Connection reset by peer))
14:00:54 Join markun_ [0] (n=markun@rockbox/developer/markun)
14:03:52 Nick markun_ is now known as markun (n=markun@rockbox/developer/markun)
14:07:09preglowme wantie
14:07:23tehsmocapacious. .I like that word
14:07:56BagderI bet it'll be available around 2012
14:08:10preglowthat would be quite ideal even for laptops
14:08:23preglowthen again, access times probably aren't extremely grand
14:08:46Bagderit seems laptops might skip the 1.8" step since there are now flash-only ones
14:09:48Bagderbut perhaps the flash prices won't be able to compete with the price
14:09:54*Bagder can't type
14:10:04amiconnThere were laptops with 1.8" HDDs years ago...
14:10:10Bagderyes, they exist
14:10:14Bagderbut are still rare
14:10:17amiconn(Toshiba libretto)
14:13:49preglow100 gigs of flash sounds like it's going to be expensive, though
14:14:15tehsmotechnology will improve, though
14:14:33tehsmoand currently HD's have the economy of scale advantage
14:14:37tehsmothat could change depending on the market
14:14:48 Quit pondlife (Read error: 104 (Connection reset by peer))
14:18:16 Join n1s [0] (
14:22:03 Join habana [0] (i=c2c7fca1@gateway/web/cgi-irc/
14:22:09habanahi all
14:29:09markunhi habana
14:31:05habanamarkun: wich player you own ?
14:32:08markuniriver h120 and gigabeat F40
14:33:05habanasansa and an old s1mp3 mpaler
14:33:29Jagan81I know it's off topic, but I was wondering if anyone knows if goPod will work on 5.5G iPods soon? (I e-mailed the developers, but no response...and I have an 80 gig'er so I can't use rockbox yet)
14:34:03BagdergoPod? what's that?
14:34:12habanawow, the f40 is pretty :)
14:34:27Jagan81Badger: a program that removes the french/EU volume limit
14:34:50lexbut not available for 5.5G :(
14:35:37daurnimatorwhats with the volume limit
14:35:41Jagan81no...but I hope rockbox will support the 80 GB soon, so I don't have to think about that
14:35:50markunhabana: I would have prefered a X30/60, but they are hard to get outside japan and australia
14:36:02daurnimatorif people are too stupid to turn it down, its their own fault
14:37:39Jagan81a french law says that a portable player should only be able to play approx. 100 dB max (_not_ loud enough for me..), and because of EU/EEC regulations all european iPods come with a volume "cap" at approx. 100 dB
14:37:59Jagan81goPod removes that restriction
14:39:42*amiconn wonders why people have problems with that volume limit
14:39:52 Quit hickery ("Leaving")
14:39:55habanaim french
14:39:57habanawe all know that up to 100 db, it damages the ear
14:40:09amiconnI don't even go near the possible maximum when listeing with earphones/headphones
14:40:14habanathis law is a good thing (at least for children)
14:41:04 Join barrywardell [0] (
14:41:10Jagan81habana: I could maybe agree with you before, but now iPods have a new feature where parents can set the maximum volume and protect it with a password
14:41:11peturthe only time I need/use full power output is when connecting to an external amplifier (line-out)
14:41:14habanaafter, if you know and accept, you can break your ears (as i actually do with the us sansa firmware to circumven (or near word) limitation
14:42:00markunhabana: why would you want to damage your ears?
14:42:11daurnimatorthats the worst thing i've ever heard
14:42:20habanaJagan81:imagine that some people dont use ipod in the world (incredible ;))
14:42:41daurnimatorparents are getting too much power over children
14:43:01Bagderwe can tell you're not a parent ;-)
14:43:01 Join Gnelik [0] (n=Miranda@
14:43:34daurnimatorproblem is with an issue like this: EVERYONE has a bias
14:43:48 Nick Gnelik is now known as Gnelik_very_happ (n=Miranda@
14:43:51habanamarkun: in the subway, classical music is nearly inaudible with limitation. and i need to listen to good vibes when leaving my job
14:44:25 Join TeaSea [0] (n=thunderc@
14:45:18habanaGnelik: dont do drugs ;)
14:45:21Jagan81lex: have you heard any news about 5.5G-support in goPod? those damn developers wont answer my email..
14:45:33n1shabana: get some sound insulating (or whatever you call them) headphones
14:45:42lexJagan81: nope :(
14:45:56lexmarkun: hey :)
14:46:04lexmarkun: i bought a psp a few weeks ago
14:46:19markunand now you are porting rockbox to it?
14:46:25lexnope :p
14:46:33Gnelik_very_happhabana: OK :)
14:46:41lexi guess i can do more with XMB than rockbox on it
14:47:35markunyes, rockbox is a bit limiting for a console/pmp
14:47:40daurnimatorwhy would you want rockbox on a psp?
14:47:57Bagderfor the fun of doing the port? ;-)
14:48:17daurnimatortheres much more worthy targets
14:48:28 Join TommyTott [0] (
14:48:40habanan1s: htanks but its too expensive
14:49:15markundaurnimator: which one are you thinking of?
14:49:29TommyTottanyone who helped me yesterday here now??
14:49:36markunworthy target
14:50:03*n1s drools
14:50:28daurnimatorn1s: they haven;t even released 40gb yet
14:50:39daurnimatorthat was annouch nigh on 1.5 years ago
14:51:10lexdaurnimator: yeah, there's a good target to port rockbox into: ipod 5.5G :)
14:51:13n1si know, i know but I can dream, can't I? :-D
14:51:18TommyTottn1s: my ipod crashed again yesterday.. but ow i fixed it.. i opened it an pushed on the HD cable.. and now it works.. :D
14:52:15n1syep, dodgy connections can make all sorts of fun things :-)
14:52:43 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
14:53:08TommyTottwhy didnt i think of that yesterday.. today i thought "i cant hear the HD starts spinning, so maybe thats the problem"
14:53:13TommyTottand it was.. :D
14:54:35TommyTottnow i have to backup my music berfore it happens again
14:54:38markundaurnimator: no 40gb??
14:54:57markunare you talking about the single platter one?
14:55:42n1sit was actually anounced before the dual platter 80 gigs but still hasn't been released afaik
14:55:49markunWell, it is used in the Gigabeat F41
14:56:00TommyTottanother q: are Anapod Explorer the best to transfer music to my ipod, or are there other good programs??
14:56:45n1smarkun: are those available somewhere?
14:57:21markunn1s: only in japan I think
14:58:16*n1s wants a gigabeat, but it is friggin' impossible to find them here exept the S models, they're just hard to find :-/
14:58:22barrywardellanybody here with an iPod that wants to test this patch: ?
14:58:40barrywardellit needs to be tested on bootloader builds as well as normal builds
14:58:40 Quit habana ("CGI:IRC (EOF)")
15:00:01 Join habana [0] (i=c2c7fca1@gateway/web/cgi-irc/
15:00:03markunn1s: they are not sold in my country either, but ebay has some F40 models
15:00:37habanabarrywardell: hi
15:00:58habanathanks for your support yesterday
15:01:46barrywardelldid it work ok?
15:02:22habanabooted on rockbox and writed error -1 on the screen
15:03:29habanadidnt have time to go further. but i lear a lot of things and hope ill manage to hack code and have fun with
15:03:40barrywardelldoes it load the sans
15:03:46barrywardellsansa firmware ok?
15:04:05habanacant tell you, it wass soooo late
15:04:14Gnelik_very_happare you speaking of sansa?
15:04:23habanai can retest tonight
15:04:32habanaGnelik_very_happ: yes
15:04:35Gnelik_very_happIf you like i can give you my graphic boootloader
15:04:43barrywardellerror -1 means it can't find rockbox.e200. you have to do a normal build, then do make and make zip then unzip to your sansa
15:05:30 Join Pikcc [0] (i=underwat@
15:05:41barrywardellmake && make zip
15:06:00***Saving seen data "./dancer.seen"
15:06:03 Join niskel [0] (
15:12:25 Join actionshrimp [0] (
15:14:00amiconnGnelik_very_happ: You're missing one letter ;)
15:14:59lexhmm :(
15:15:09lexsennheiser hd 201 or a 2GB memory stick pro duo
15:15:22lexi need new headphones because these are going to be replaced
15:15:30lexby a working one
15:20:07 Quit TommyTott ()
15:25:55markunlex: are the 201's so great?
15:27:10lexmarkun: dunno, but i need a new one
15:27:33lexfor my ipod and to the time when those are going to the guarantee
15:32:06 Quit midkay ("*poff*")
15:32:43 Join midkay [0] (n=midkay@rockbox/developer/midkay)
15:37:35 Join vcardenas [0] (i=c818044b@gateway/web/cgi-irc/
15:38:37 Quit daurnimator (Connection timed out)
15:38:38 Quit ender` (Read error: 110 (Connection timed out))
15:40:21 Join pondlife [0] (
15:41:43 Quit petur ("worrrk")
15:44:22Gnelik_very_happamiconn:Than my happynes in not so big :)
15:45:41habananot only size matters
15:45:52habanaquality first !
15:48:26 Join TommyTott [0] (
15:49:53 Join TeaSeaLancs [0] (n=thunderc@
15:51:14TommyTottcan anyone direct me to a site with rockbox ipod photo skins??
15:54:46 Join NickDe [0] (n=nicholas@
15:54:58 Quit NickDe (Read error: 104 (Connection reset by peer))
15:55:22 Join Yuki [0] (
15:57:36 Join NickDe [0] (n=nicholas@
15:57:49 Quit NickDe (Read error: 104 (Connection reset by peer))
15:57:51 Quit Pikcc (Read error: 110 (Connection timed out))
16:03:31 Join ender` [0] (n=ender@
16:04:02 Quit TeaSea (Read error: 110 (Connection timed out))
16:04:14TommyTottyey.. my ipod works again.. and with rockbox not bugging.. :D
16:17:49 Part Yuki
16:19:12 Join perpleXa [0] (n=perpleXa@unaffiliated/perplexa)
16:31:18 Join Kingstone [0] (i=GPool@
16:32:48TommyTotthmmm.. the scrolling are so slow.. are 15 the max speed?? (i cant notice any difference in the different settings)
16:32:54TommyTottanyone have any idea??
16:33:34markunThe font caching makes it slow
16:33:56TommyTottand the sulotion is?
16:34:02TommyTottno themes??
16:34:27markunTry the chicago font. Is scrolling still slow then?
16:34:59markunthe solution is to improve the font caching code
16:34:59 Quit funky (Read error: 104 (Connection reset by peer))
16:35:24 Join perplexity [0] (
16:35:24 Join funky [0] (n=repulse@unaffiliated/funky)
16:36:05TommyTottyes, its still slow
16:36:20TommyTottand now the playscreen theme is fucked up.. :P
16:37:22 Quit thomasg_ (Remote closed the connection)
16:37:47markunstill slow.. strange
16:38:19TommyTottits not extreemly slow..
16:38:46TommyTottbut, in apples fw, if you scroll slow, it scrolls slow, but if you do it faster, it speeds u
16:38:55TommyTottmaybe rockbox doesnt do that??
16:39:38 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
16:40:05TommyTottdo you know?
16:40:51 Join kubiix [0] (
16:40:55TommyTottit seems like it do now, but it has a lag when i scroll fast
16:41:10 Quit funky (Read error: 104 (Connection reset by peer))
16:41:42 Join funky [0] (n=repulse@unaffiliated/funky)
16:42:12 Join SaLoMoN [0] (
16:44:15 Quit funky (Read error: 104 (Connection reset by peer))
16:44:34TommyTottnooo.. it crashed.. what do i do???
16:45:01 Join PaulPosition [0] (
16:45:04TommyTotthehe, found it
16:46:41 Join funky [0] (n=repulse@unaffiliated/funky)
16:51:38 Join lowlight [0] (i=c730190a@gateway/web/cgi-irc/
16:54:09TommyTottbut the scrolling is still slow..
16:54:58lowlightI'm missing the obvious...where is MEM defined?
16:56:03PaulPositionBeing very interested in testing Barrywardell's patch for e200 and h10 bootloaders, I want to know if it's possible to cvs check rockbox-devel specific modules without overwriting those files in /rockbox/ that I've spent a day getting patches to sync with...
16:57:01barrywardellwhich modules do you want?
16:57:43Bagderlowlight: in the local Makefiles
16:57:51Bagder -DMEM=${MEMORYSIZE} usually
16:58:47PaulPositionBarrywardell - Hmm, I guess those that are needed to build the bootloader :p Is it only the bootloader module or would I need debug too?
16:59:12PaulPositionoh.. Just found out how to check separate module..
16:59:35barrywardellyou need some of the stuff in firmware too. and some stuff in apps
17:00:18PaulPositionbarrywardell - Well, the pseudo-mopule 'rockbox' is already there so...
17:01:13barrywardellyou need rockbox-devel
17:01:13 Quit funky (Read error: 104 (Connection reset by peer))
17:01:35barrywardellthat's rockbox + bootloader + manual iirc
17:01:54 Join funky [0] (n=repulse@unaffiliated/funky)
17:02:20lowlightahhh...why not in config-*.h?
17:02:26 Quit lowlight ("CGI:IRC (EOF)")
17:02:36 Join miepchen^schlaf [0] (
17:02:52PaulPositionahh okay then. Thanks :)
17:03:22 Join lowlight [0] (i=c730190a@gateway/web/cgi-irc/
17:04:06 Quit Rob2222 (Read error: 104 (Connection reset by peer))
17:05:33TommyTottanyone know anything about the scrolling problem??
17:06:02***Saving seen data "./dancer.seen"
17:07:03Kingstonewhy Apple did this to us?!? :\\
17:07:11Kingstonewe only wanted a different OS
17:07:17Bagderlowlight: the MEMORYSIZE is set by configure and would make more sense to be used directly
17:07:19Kingstonewhy did they encrypt the new nano
17:07:32Kingstonei guess they had enough of us
17:08:01lowlightTommyTott: you mean line scrolling or list scrolling?
17:08:01 Join freqmod [0] (
17:08:18TommyTottlist scrolling.. its so slow.. have an 4th gen photo..
17:09:17lowlightI believe it's know that list scrolling is slow on ipods. I think there's a scroll accelerator patch on the tracker.
17:10:11 Nick myzar is now known as myzar|away (n=myzar@
17:10:14TommyTotti only found one for 5th gen
17:10:18lowlightor maybe that was scroll wheel accelerator
17:10:35zylcheI thought you could change speeds...
17:10:49TommyTottscroll wheel accelerator yes, but is far as i know, its for the 5th gen video only
17:11:02zylcheahh, figures.
17:11:50lowlightmake it work for 4G's too :)
17:12:01lowlightthat's how things get done around here
17:12:01 Quit lowlight ("CGI:IRC (EOF)")
17:12:22 Quit ender` (" How many Dragon Ball Z characters does it take to screw in a lightbulb? Only one, but it takes ten episodes.")
17:12:43TommyTotti dont know how to do that..
17:12:48TommyTottthis is the one i found
17:13:01 Join lowlight [0] (i=c730180a@gateway/web/cgi-irc/
17:13:19TommyTottand it seems that it doesnt even work with new builds
17:13:37SaLoMoNdoes anyone knows how far the development of rockbox for sandisk players is?
17:14:50 Join My_Sic [0] (
17:16:13 Join ender` [0] (n=ender@
17:16:20lowlightcheck the sansa thread on the forum under "New Ports"
17:16:36PaulPositionBarrywardell - Okay, so expected behaviour : changing only the bootloader and not the rest of the build produces only a 'system files missing' error on my 5gb H10. Am compiling the rest now. −− one question, should the Original.mi4 file be _original_ or unencrypted?
17:17:28 Quit Gnelik_very_happ (Read error: 131 (Connection reset by peer))
17:17:33*Bagder created just now
17:18:06barrywardellPaulPosition: do you have a MTP H10?
17:18:30PaulPositionBarrywardell - yeah, MTP. why?
17:18:34*zylche made up
17:18:44zylcheah hah! Bagder!
17:18:54barrywardellthen the H10.mi4 you created won't work
17:19:02PaulPositionoh. duh.
17:19:17zylcheThe pages aren't updating their last modified date!
17:19:30PaulPositionlol What would I have overlooked? Or is that one wall you haven't broke down yet?
17:19:32barrywardelluse ../tools/scramble -mi4v3 bootloader/bootloader.bin H10.mi4
17:19:39 Join _FireFly_ [0] (
17:20:01PaulPositionbarrywardell - Ah, will have a look then, thanks :)
17:20:32 Join Paul_the_Nerd [0] (
17:21:46TommyTottArent there anyone who has a sulotion for the scroll wheel problem?? for the 4th gen ipod..
17:22:35zylcheBagder come back now!
17:22:48*zylche prods Bagder with a long pointy stick
17:22:57 Quit lowlight ("CGI:IRC (EOF)")
17:24:19Paul_the_NerdTommyTott: There's no reason to assume that someone has a solution to the problem. If there were a *proper* solution, don't you think it would get included?
17:24:29 Join lowlight [0] (i=c730180a@gateway/web/cgi-irc/
17:24:45 Join Rob2222 [0] (
17:25:24TommyTottPaul_the_Nerd: i found a solution for the 5th gen, so maybe thes is one for the 4th gen i havent found..
17:25:50Paul_the_NerdTommyTott: That patch works for all iPods, or can be made to work for all iPods, but it doesn't work with the new button methods, as you've pointed out.
17:26:45TommyTottso you say that all ipods are scrolling really slow, since that patch dont work with the new builds??
17:27:24TommyTottif thats i known problem, why arent it fixed, when there seems to be a know sulotion for it
17:27:32Paul_the_NerdThe 5G and 5.5G have the worst of it, since they have the most screen to update, but yes, they all scroll fairly slow
17:27:41Paul_the_NerdThere is not a known solution to it.
17:27:50Paul_the_NerdThe Scroll Acceleration patch doesn't really fix the whole problem.
17:28:05Paul_the_NerdAs I said, if there were a PROPER solution, it would most likely already be included.
17:30:01 Quit habana ("CGI:IRC (EOF)")
17:30:08TommyTottto bad..
17:30:46Paul_the_NerdWell, you're welcome to start searching for a solution as well.
17:33:05TommyTotti dont know any programing, so that would be difficult for me
17:34:38barrywardellPaulPosition: i forgot to mention that Original.mi4 needs to be decrypted and have its header stripped by mi4code.
17:35:00 Quit lowlight ("CGI:IRC (EOF)")
17:35:03PaulPositionbarrywardell - yeah, I guessed something similar. :)
17:36:11PaulPositionbarrywardell - Just the bootloader was enough to get into rockbox, but my build wasn't much old (two days..). Now to get into OF. :p
17:43:30PaulPositionbarrywardell - wow, the mi4code tool isn't the simplest thing around.. :o "mi4code decrypt -s Original.mi4 dummy" .. And then what? rebuilding would reencrypt, right? Do I just rename?
17:44:08PaulPositiongoffa - Woah.
17:44:46Paul_the_Nerdgoffa: I saw that. I wonder if it means 50gb single-platters.
17:44:48barrywardellmi4code decrypt -s H10.mi4 Original.mi4
17:44:54Paul_the_NerdI keep holding off on upgrading my H120.
17:45:03barrywardellwhere H10.mi4 is an iriver firmware
17:45:32barrywardellthen put the Original.mi4 file that's created in you System dir
17:45:56PaulPositionbarrywardell - Thanks, so renaming the 'dummy' file I produced should do the same..
17:46:20barrywardelli guees so
17:46:52goffadunno .. but i think that'd be pretty cool
17:47:01goffawould really open things up for those h120 people
17:47:14Paul_the_NerdWell, we already have 30 and 40 options.
17:47:33goffa10gb is a lot of music though
17:47:46goffai notice the difference between my x5 and my h140
17:48:22Paul_the_NerdIf you've got legal music, many people will have a hard time filling even 30
17:48:38Paul_the_NerdUnless you're encoding to lossless formats.
17:48:53PaulPositionI wish they didn't forget the 1" drives.. We want our bits standing up and not lying down, like in that 'get perpendicular' cartoon from Hitachi. :p
17:49:12Paul_the_NerdI'm sure the .85s will get some loving too, in time
17:49:22Paul_the_NerdBut Apple doesn't make one, so there's not the same economic pressure.
17:49:37goffafunny how they seem to control everything
17:50:16PaulPositionNonono, they're the rebel working against those that control everything? Didn't you get the memo?
17:50:21goffalike cars have ipod docks, planes are installing docks, hell i even saw a toilet paper doc yesterday
17:50:37 Join obo [0] (
17:50:55goffaoh yeah... ipod owners are terrorists and thieves
17:51:51 Quit lee-qid ("aufwiederbyebientotsayonara")
17:52:03 Join lost|X40 [0] (
17:54:07 Quit lostnihilist (Read error: 110 (Connection timed out))
17:54:11preglowwhat is the bootloader.img which is put in the ipod bootloader dir after compiling it?
17:54:23preglowbootloader image with no retailos?
17:55:32*preglow messes up his nano
17:55:54 Join TeaSea [0] (n=thunderc@
17:56:09preglowwhat the hell just happened
17:56:37Paul_the_NerdWhat'd you do?
17:57:03preglowi'm trying out barry's bootloader
17:57:13preglowand now the only thing i get is the apple logo
17:57:50preglowi'd actually forgotten how to manually enter disk mode and pressed the wrong keys...
17:57:58Paul_the_NerdBeen a while, 'eh.
17:58:16preglowapparently, i was so used to it just entering usb mode that at first i even didn't press any keys
17:58:37preglowbut yeah, know what bootloader.img is?
17:59:50 Quit Rob2222 ()
18:00:14Paul_the_NerdI have no clue. :)
18:03:02 Quit TeaSea (Read error: 104 (Connection reset by peer))
18:03:15 Join TeaSea [0] (n=thunderc@
18:04:57preglowwell, this certainly does not look so bright
18:05:01preglowthe patched bootloader hangs
18:05:05 Part vcardenas
18:06:36barrywardellwhat ipod model?
18:06:57barrywardellnvm. it's a nano
18:07:18preglowit just hangs at the apple logo
18:07:21preglowi've removed all local changes
18:07:36preglowthat is, the ones of consequence, mainly just some cache thing i think is redundant
18:08:08preglowthe hanging thing might suggest it's stuck waiting for the cop at some spot
18:08:59 Join bubi [0] (n=user@
18:09:01preglowyou're sure there's no sansa/h10 only stuff in here?
18:09:05bubihi there...
18:09:33barrywardellwhich cache thing did you remove?
18:09:53barrywardelli don't think there's any specific stuff there
18:09:54preglowline 672-3 in system.c
18:10:21preglowjust dummy reads from ram to clear the cache lines, i don't think it's needed
18:10:27preglowbut i put it back anyway
18:11:07bubii am trying to create the bootloader for my ipod video with firmware 1.2 but geht the error :
18:11:07bubiC:\Downloads>ipod_fw -g video -o rockboot.bin -i apple_os.bin bootloader-video.b
18:11:07DBUGEnqueued KICK bubi
18:11:07bubifread failed: No error
18:12:33 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
18:12:39preglowwell, do you have apple_os.bin and bootloader-video.bin in the current dir?
18:13:00 Join miepchen^schlaf [0] (
18:13:12bubibut. hmm apple_os.bin 0kb ? could not be ?!
18:13:20preglowthen there's your problem
18:13:51barrywardellpreglow: hmm, so it doesn't even start to show the rockbox bootloader screen?
18:14:26preglowbarrywardell: no, just the apple logo, which is under bios control
18:14:32preglowbarrywardell: eh, rom/flash
18:14:40preglowbarrywardell: just tried building my own with no patches, and it works
18:14:46bubiah know :)
18:14:52 Quit obo ("bye")
18:15:09preglowonly changes that affect me are the ones in crt0-pp.S
18:15:18 Join obo [0] (
18:15:35barrywardellyeah, they're the only ones that the patch affects for ipods
18:16:14preglowi'll have a look and see
18:16:18 Quit bluebrother ("leaving")
18:16:44 Join bluebrother [0] (i=Ftom1k5m@rockbox/staff/bluebrother)
18:19:19preglowgah, crt0-pp.S is a jumble of bootloader and non-bootloader code
18:19:27preglowprobably my fault
18:20:26barrywardelli have a version with just bootloader code:
18:20:39barrywardellthat should be equivalent to the version in my patch
18:21:50 Join Gnelik [0] (n=Miranda@
18:21:56 Quit TeaSeaLancs (Read error: 110 (Connection timed out))
18:23:15preglowit also had syntax errors :>
18:23:19preglowpad_skip is a label
18:24:10preglowbut no, no change in behaviour
18:27:44preglowtried removing the cop wait loop, no change
18:28:27GnelikWhy double values a allways rouded??
18:28:53barrywardellthere isn't a huge amount of change. it's mostly just rearranging
18:29:09preglowmyeah, i see
18:29:49 Quit bubi ()
18:29:52 Join shrift [0] (n=shrift@
18:32:28 Quit linuxstb (Read error: 110 (Connection timed out))
18:32:49 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
18:34:11preglowlinuxstb: there?
18:34:21preglowah, unattended join...
18:34:39shrifthi guys
18:34:42barrywardellthe current crt0-pp.S has two sections where it gets the "high part of our execute address"
18:34:52shriftcan anyone point me to where there is information about restoring an ipod?
18:36:03barrywardelland it uses 0xffffff00 in the current version (ie. the second section. the first one doesn't seem to be used at all)
18:36:20barrywardellmy version uses 0xff000000
18:37:03shriftI know there is the apple restore utility, but that is not working... If anyone knows where there is any more detailed documentation?
18:37:23preglowbarrywardell: i changed that to bic r4, pc, #0xff now, no change
18:37:52preglowin a related note, does anyone know where the apple bootloader loads us?
18:40:55Paul_the_Nerdshrift: There's not really much more documentation on it than the official documentation. Failures of the Apple Restore tool are a problem with the tool itself, anyway.
18:40:56Kasperleshrift: can you boot the ipod into disk mode?
18:41:11shriftI can get into disk mode fine
18:41:18shriftand I reformatted it to NTFS
18:41:23shriftthen ran the apple restore tool
18:41:27shriftit says it fails in windows
18:41:37shriftbut it does change the partitioning to the standard ipod setup
18:41:38Kasperleshrift: in my case, iTunes said "bleh, your ipod is broken, you can restore it to factory settings by pressing here ..."
18:41:48shriftit says that
18:42:02shrifthowever, it never boots into the ipod OS
18:42:05shrift: (
18:42:36shriftand I don't really care about that, but my understanding is that I need that working in order for rockbox to extract the files it needs?
18:42:47Kasperlecan you manually format the partition to fat32?
18:42:48barrywardellpreglow: using 0xffffff00 makes my H10 bootloader not work - same symptoms as your ipod bootloader
18:42:52 Join Arathis [0] (
18:43:35shriftwell, it is currently fat32
18:44:38shriftdo we have a pastebin type thing, or just paste into the chat?
18:44:51Paul_the_NerdWell, we have pastebin itself.
18:45:00preglowi don't get this
18:45:07preglowwhy does r8 get moved to r0 before main is called?
18:45:10preglowmain takes no parameters
18:45:22shrift Device Boot Start End Blocks Id System
18:45:22shrift/dev/sdc1 1 14 112423+ 0 Empty
18:45:22shrift/dev/sdc2 15 7296 58492665 b W95 FAT32
18:45:28shriftthat is what the ipod looks like now
18:45:35shriftso the partitions seem ok
18:45:36preglowi've got a worrying feeling i wrote that code too...
18:45:47shriftbut for some reason it's not putting the OS in...?
18:45:57Paul_the_NerdWhat type of iPod is it?
18:46:55shrift5.0 60gb
18:47:43barrywardelldan_a wrote the new bootloader crt0
18:47:49Paul_the_NerdWas it one of the 5.0s that Rockbox wouldn't install on?
18:49:28shriftI have had rockbox on the device before....
18:49:31shriftI kinda jacked it up
18:49:33shriftso I have to reinstall
18:49:48shriftit's not one of the new "5.5" gen ipods
18:50:14Paul_the_NerdWell, there were 5.0s that suffered the same problem as 5.5s.
18:50:25Paul_the_NerdHave you tried using a different version of the restore tool, by the way?
18:51:20shrifthmmmm, I have not yet tried a diff version of the tool
18:51:22shriftI will try that
18:51:45shriftthanks for the idea.... I still have an old one installed before they moved it all to the itunes integrated updater
18:52:00barrywardellpreglow: that mov r0, r8 looks unnecessary, but also harmless
18:52:42Paul_the_NerdGood luck with it, shrift
18:52:55shriftthanks, Paul_the_Nerd
18:53:13preglowbarrywardell: sure, r0 gets clobbered anyway, just wondering what was the rationale, heh
18:53:19 Quit Criamos (Read error: 104 (Connection reset by peer))
18:53:19oboPaul_the_Nerd: did you see ? Is it along the lines you were thinking of?
18:53:52preglowbut the assumption here seems to be that we're loaded at the start of ram or something
18:53:55preglowis this really so?
18:54:11barrywardellthat is so on the h10/sansa
18:54:28GnelikWhat is limitations working with double type
18:54:38GnelikOr some known bugs?
18:55:01preglowGnelik: what are you trying to do?
18:55:57Gneliki+=(double)(LCD_WIDTH/plot.x_mod) i -double xmod-double
18:56:15 Join Criamos [0] (
18:56:18Gnelikand return (int)x*x x -double
18:56:27Paul_the_Nerdobo: From the description, yeah.
18:56:54Paul_the_Nerdobo: Sad to admit, I haven't tried it, yet.
18:57:39preglowGnelik: those lines are seriously fucked up, are you sure you don't have a problem with your irc client?
18:57:57oboheh, no worries.
18:58:18 Join TeaSeaLancs [0] (n=thunderc@
18:59:10Gneliki+=(double)(LCD_WIDTH/plot.x_mod) where i is double and x_mod is double
18:59:12Paul_the_Nerdobo: But it's definitely what I had in mind (and would reduce instances of some of the more common iPod complaints/questions, such as "Why doesn't car adapter mode work?" and "Why do my settings keep clearing?")
18:59:41oboPaul_the_Nerd: car adapter mode is only enabled on the 5g at the mo...
19:00:18preglowGnelik: nothing wrong with that
19:01:23Paul_the_Nerdobo: Yup, but it shouldn't have worked anyway, because receiving power from the car adapter would cause it to boot into disk mode anyway, right?
19:01:44Paul_the_NerdNow, all someone needs to do is leave Hold on and Car Adapter mode should function fine (with that patch)?
19:01:46oboyes, unless you manually held menu
19:01:55oboAFAIK it should
19:02:15preglowbarrywardell: even if i calculate the pc at start: and use that for copying, it doesn't work
19:06:04***Saving seen data "./dancer.seen"
19:06:06amiconnGnelik: It seems that gcc sometimes doesn't understand what you want. I would simply rewrite the line as:
19:06:37 Quit XavierGr (Nick collision from services.)
19:06:40 Join XavierGr [0] (
19:06:41amiconni.e. first casting LCD_WIDTH into double, and then dividing
19:06:46Gnelikpreglow:Do you have a time? Look at
19:07:10Gnelikdraw_plot function
19:10:07Gnelik amiconn: Can't we DEBUGF double variables?
19:11:21preglowGnelik: not really, no
19:11:32Arathisbarrywardell: didn't watch the channel much the last days. was there any progress on dualboot for h10?
19:11:46preglowbarrywardell: ok, i now tried to just skip the iram copy code, then insert a straight reset sequency, and then i get the ipod support screen
19:11:55preglowbarrywardell: which at least is a difference
19:12:03preglowArathis: it's being worked on right now
19:12:22preglowArathis: it's pretty much working
19:12:27 Join God__Eater [0] (
19:12:34 Quit God_Eater (Read error: 110 (Connection timed out))
19:13:08barrywardellArathis: it works now :)
19:13:45Arathisbarrywardell: great! is the new bootloader up already?
19:13:53CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
19:13:53*Arathis want's to download :D
19:13:54preglowbarrywardell: deh, i can't replicate that, apparently...
19:13:58barrywardellit's not in cvs yet.
19:15:04Arathisbtw: are the bootloaders at build automatically out of cvs too?
19:15:17amiconnbarrywardell: Are there any features in the H10 retailos that make dualboot useful?
19:15:46barrywardellradio and recording are the big two
19:15:49barrywardelland wma
19:16:12preglowwell, this is annoying....
19:16:19Arathisuntil somebody implements them in RB *g*
19:16:28preglowi think i'll go get some dinner and wait for someone else to try too
19:17:04amiconnAh ok.
19:17:11barrywardellpreglow: me too
19:17:16*amiconn wonders what's up with linuxstb's recording code
19:17:35barrywardellamiconn: it works on H10 too, but with the same bugs
19:17:54amiconnWhat bugs?
19:18:04amiconnI cannot test, as my only PP target cannot record
19:18:14amiconn(DAC only - mini G2)
19:18:55barrywardellplayback is broken after recording
19:19:13barrywardellmp3 encoding doesn't work - probably endian issues
19:19:23 Quit TeaSea (Read error: 110 (Connection timed out))
19:19:48amiconnHMM :/
19:19:49Arathisbarrywardell: any idea when bootloader code will be in cvs?
19:19:56barrywardellArathis: look back on today's irc logs for a patch
19:20:06Arathisah, okay
19:20:47Arathisthanks for the work again to you and dan_a
19:20:48barrywardellArathis: it's not in cvs yet because the same code affects the ipods and seems to break the ipod bootloader
19:22:19preglowgod knows what's wrong, though
19:22:26preglowi now reset the cpu first thing
19:22:30preglowyet nothing new seems to happen
19:22:34preglowit just stays at the boot logo
19:22:38Arathisi'm off
19:22:40preglowi think i should see a blink or something
19:22:41 Quit Arathis ("[rl_bot quit]")
19:23:10barrywardellamiconn: there is a patch in the tracker
19:24:19barrywardellpreglow: very confusing. i'll try get a lend of an ipod color later and test on that
19:25:21 Join Cassandra [0] (
19:25:23preglowbut at least i can compile a working bootloader from the old code
19:25:27preglowso something has to be up
19:25:39*preglow wants jtag
19:25:51GnelikWhen i do temp.y = i*i; all works ok
19:26:01CassandraI want dancing badgers.
19:26:01*barrywardell has jtag on his H10
19:26:08preglowyou mean bagders
19:26:08CassandraNo-one ever gives me what I want.
19:26:14Gnelikbut when i do temp.y = count(i);
19:26:24Gnelikit dont work
19:26:25CassandraNah. I'm still mentally scarred by that picture on the forums.
19:26:37barrywardellnope, i soldered onto the jtag pins on my H10 and made a connector
19:26:52Gnelikstatic double count(double x) {return x*x;}
19:26:53PaulPositionBit weird... Went crazy on the EQ/crossfeed/etc. on my H10 and I get no skip but a *very* unresponsive system...
19:27:01preglowbarrywardell: well, does it work?
19:27:08barrywardellkind of
19:27:16GnelikHelp please!!
19:27:24barrywardellusing jtag tools i can detect two jtag devices
19:27:25CassandraSearch isn't working. Otherwise I'd provide a link.
19:27:28 Quit TeaSeaLancs (Remote closed the connection)
19:27:40 Join TeaSea [0] (n=thunderc@
19:27:50oboCassandra: check your email - there awaits a dancing badger :)
19:27:50Paul_the_NerdAugh, what keeps borking forum search. :(
19:28:15barrywardellbut it only recognises them as unknown
19:28:27preglowbarrywardell: which means you can't do anything with them, or?
19:28:34barrywardelli was recommended to try openocd but haven't got that to work yet
19:28:46barrywardellpreglow: not yet
19:29:38CassandraI would do, but Thunderbird ate my mail profile.
19:29:46CassandraStupid Thunderbird.
19:30:52*amiconn really wonders what could cause the hang on G5.5/80GB
19:30:56 Quit pondlife ("disconnected has pondlife")
19:31:06goffaCassandra: check your profile.ini file
19:31:14amiconnThis should be tracked down to the exact line of code that causes the hang
19:31:14goffayou should be able to switch to your old profile
19:31:28goffaif you are in win.. dunno about lin
19:32:11barrywardellpreglow: i gotta go now, but I'll check back in a few hours. and I'll check the logs to see if you found anything
19:32:22PaulPositionBarrywardell - Haven't got that dualboot to work on my H10 5gb MTP, by the way.
19:32:35barrywardellno? how far did you get?
19:32:59barrywardelldid you make a H10.mi4 that doesn't give System Files Missing?
19:33:12PaulPositionFirst try got me into the iriver's loading screen and hung..
19:33:27PaulPositionnext tries stop on rockbox saying 'loading iriver firmware'.
19:33:37 Quit God__Eater (Read error: 110 (Connection timed out))
19:33:46Cassandragoffa: All looks good - and the profile's there. F*** knows what it's on.
19:33:59PaulPositionNo 'system files missing' at all. Rockbox load real fine though.
19:34:04goffahmm... do you have more than one profile Cassandra?
19:34:10barrywardellPaulPosition: this is with my patch applied?
19:34:18goffaok.. that rules that out :(
19:34:18barrywardellsounds like how things worked before
19:34:30CassandraAnd it was working on Friday before I went away for the weekend.
19:34:47goffawhat if you made a new profile and copied your boxes to that?
19:34:50 Join God__Eater [0] (
19:34:56PaulPositionbarrywardell - Am thinking of something...
19:35:09goffaits probably the files themselves.. but that sometimes works
19:35:11PaulPositionbarrywardell - Please go on with your business, I'll catch you later :)
19:35:23barrywardellok, bye
19:35:23 Part barrywardell
19:35:30 Quit TeaSea (Read error: 104 (Connection reset by peer))
19:37:36GnelikPople help please with double types
19:37:46 Join Genre9mp3 [0] (n=MARVIN_T@rockbox/contributor/Genre9mp3)
19:38:32Gnelikint a = x*x; where x-double works well, int a = count(x); where count static double count(double x) { return x*x;}
19:38:57GnelikI a trying to make graph builder
19:39:48amiconnDo you declare count() before the statement?
19:42:35pixelmapreglow: Cassandra probably means this picture
19:43:35 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
19:44:40 Quit MarcoPolo ("Bye !")
19:47:32goffahmm.. i like the logo from alfred's post with rb_logo_mashup_blackbg.jpg
19:47:40goffawith the white box next to rockbox
19:47:49goffawith the downward facing note
19:48:30GnelikNo other thougths?
19:49:02markungoffa: which one?
19:49:43goffawith the white
19:49:53goffaer bluish white whatever
19:50:18Gnelikgrrrr.. :(
19:50:48goffajust looks clean
19:51:41 Quit Genre9mp3 ()
19:52:41goffamaybe i'm nuts.. but i like the upside down logo better :)
19:52:50goffabut i can see the case for the R
19:53:04CassandraI still want a cuddly badger a la Linux Penguin.
19:53:14CassandraPeople like cuddly animals.
19:53:17goffahe he he
19:53:33CassandraIt could have little headphones and everything.
19:53:40CassandraAnd we could sell stuffed toys.
19:53:54goffamove for a mascot?
19:54:31goffalike at rbcon you could have a guy in a badger costume walking around, etc
19:55:10CassandraYeah. People relate to mascots.
19:55:18*Cassandra is semi-serious.
19:55:52goffawould be a good promo thing
19:55:59goffato raise money for the project
19:56:28 Quit God__Eater (Read error: 110 (Connection timed out))
19:56:28 Quit TommyTott ()
19:56:30goffat-shirts, badger caps (w and w/o phones)
19:56:31 Join God__Eater [0] (
19:57:40goffawith the tagline "Badgers, we don't need no stinking Badgers!" (sorry had to throw that in somewhere, has to come up when badgers are mentioned)
19:57:42netmasta10bti hear that companies with mascots make 17% more proffit
19:59:55 Quit Siku ()
20:01:52Cassandragoffa - it's one of my favourite sayings.
20:02:16CassandraNow I have "We don't need no stinking badgers!" to the tune of The Wall going through my head.
20:02:19CassandraOh well.
20:02:58 Quit Gnelik (Read error: 104 (Connection reset by peer))
20:03:34oboThanks for sharing the misery :)
20:03:56 Join miepchen^schlaf [0] (
20:04:03 Join habana [0] (i=58a10615@gateway/web/cgi-irc/
20:04:56 Join dan_a [0] (n=dan-mirc@
20:05:44CassandraBesides, everyone loves badgers, right?
20:06:38CassandraBrilliant. So thanks to Dunderbird, I've lost my archived gmail posts and my Bayesian spam blocking database. Pfeh.
20:06:48preglowpixelma: hahahaha
20:07:02CassandraGood job my main account is on IMAP. Stupid Mozilla peeps.
20:07:34Cassandrapreglow, told you it was mentally scarring.
20:09:01preglowCassandra: as mental scars go, this one worked out nicely
20:09:44preglowi like those logos
20:09:54preglowvery much cleaner than the current design
20:11:55CassandraAs I said in the thread - I like them, but not enough for it to be worth changing our stationary.
20:12:24CassandraThey're missing some sort of personality or pizzazz or something beginning with P.
20:12:38preglowit's surely missing pizza
20:12:46preglowbut no, i think the current logo looks too crowded
20:12:57CassandraI'm no fan of the old logo, but replacing it with something else uninspiring doesn't strike me as a good move.
20:13:00preglowi'd love to see a new website design based on those
20:13:19Cassandrapreglow, I think most people are in agreement with that.
20:13:31CassandraI also think it's too orange, but I think that's a more personal thing.
20:13:48dan_aFWIW I like the logos in the thread more than the current one
20:14:03amiconnOh noes
20:14:07*bluebrother likes the current logo best
20:14:23amiconnCould someone please make this silly logo thread go away? grr
20:15:04preglowwhy silly?
20:15:10Paul_the_NerdI wouldn't mind booting to this one:;topic=6202.0;attach=1312;image but I'm kinda happy with the current website+logo combination.
20:15:23preglowCassandra: agree with the orangeness
20:15:44preglowPaul_the_Nerd: i think there's one colour too many in that one
20:16:21Paul_the_NerdI think the gray on the box should match the gray in the word, yes.
20:16:46dan_aIt's easier to argue about the logos than to work out how to make the bootloader work across iPods, Sansas and H10s...
20:17:11habanaPaul_the_Nerd: nice logo :)
20:17:33Paul_the_Nerdhabana: It's not mine.
20:17:52preglowtalking about logos also is a FUN break from looking at the apple logo straight for two hours
20:17:55preglowi like fun
20:18:01Paul_the_NerdI'm not wholly sure where the original attribution lies, The 'cube' is from somewhere on Misticriver.
20:18:01PaulPositionHehehe.. Well I can say that Barry's bootloader does dualboot on my H10 5gb. Yay!!
20:18:14 Join lowlight [0] (i=c730180a@gateway/web/cgi-irc/
20:18:45Slasherihmm, now iriver bootup looks quite "professional". After pressing power on, backlight almost immediately fades in and rockbox logo displays :)
20:19:21preglowSlasheri: ooh :)
20:19:43*amiconn is a bit surprised about linux media players
20:19:55amiconnNeither totem nor rhythmbox do gapless
20:21:29 Join Lear [0] (i=chatzill@rockbox/developer/lear)
20:21:39preglownothing does...
20:21:46tehsmomusicpd does :D
20:21:46preglowlinux media players bloody suck
20:21:48pregloweach one of them
20:21:58 Join wooo [0] (
20:22:00tehsmoeh, I disagree - musicpd is good
20:22:02lowlightmpd does gapless
20:22:17tehsmosupports gapless playback, lots of formats, searching through filename or id3 tags, etc.
20:22:29tehsmothat and it uses a daemon-client model, which is rather cool
20:23:08Slasheriamiconn: hmm, something seems to be wrong with the boosting code.. Now after undefined the cpu boost tracking, now iriver crashes at first boost also
20:23:14Slasheriand when loading from disk
20:23:37preglowtehsmo: does it require you to make a database?
20:23:58tehsmopreglow: yeah
20:24:05tehsmoyou give it a directory to search in and it generates one
20:24:12preglowthen it's out of the question
20:24:40tehsmoeh, I don't mind having to update the db every once in a while
20:24:41Cassandraamicon: if you'd like an alternative topic, I'd like to talk about how to automate downloading voicefonts.
20:24:42preglowalso, i'm not a fan of client-daemon stuff when it comes to media players
20:24:45amiconnNow I can understand why some people want the rockbox sim as a media player
20:24:50preglowtehsmo: i mind even having a database
20:25:05tehsmoit's worth it for the other features (keyword search, multiple clients including CLI ones, seemless switching from alsa playback to a shoutcast stream, etc.)
20:25:06preglowamiconn: it actually is the best thing i've got... now if only it used less cpu
20:25:10tehsmobut I can see where you're coming from
20:25:14*amiconn also doesn't want a database
20:25:18Slasherii think also rockbox would be superiour audio player engine even on linux
20:25:33amiconnpreglow: On linux I tend to believe this now. On windows there are better alternatives
20:25:44Slasheriof course it would have some limitations on pc.. but still
20:26:01Cassandraamiconn: sometimes I think you'd be happiest if rockbox ran on Stonehenge. ;)
20:26:17Slasheriamarok is quite good on linux, but it's database and ui is too laggy
20:26:26preglowamiconn: surely, yes
20:26:28Paul_the_NerdCassandra: The port's underway, but only getting to test code on solstices and equinoxes makes debugging a hassle.
20:26:30tehsmoeh ..amarok requires KDElibs iirc
20:26:31preglowamiconn: nothing beats foobar2000
20:26:37*tehsmo doesn't like his mp3 players to require gui libs
20:26:41lowlightthis one claims gapless too
20:26:45CassandraPrehistoric stone circle. I'm implying you like things no-frills.
20:26:46amiconnAnything that _requires_ to build a database is out of question for me
20:26:50preglowoh yes
20:26:53tehsmoalthough, I don't like them to require guis *at all*
20:27:03amiconnpreglow: Hehe, I never tried foobar, but there are numerous others
20:27:10Cassandraamiconn, but Rockbox doesn't require it. Just uses it if you want to.
20:27:45preglowamiconn: no wasted cpu/memory on a fancy gui, and it does almost everything you want
20:28:59 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
20:29:05amiconn_client/server can be useful if it doesn't make configuration too complex
20:29:07amiconn_Deliplayer is an example for client/server on windows
20:29:10Cassandrafoobar's not bad. And you can do replaygain tagging in it.
20:29:12 Quit amiconn (Nick collision from services.)
20:29:12 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
20:29:29*amiconn doesn't need or want replaygain either
20:29:36Slasheriargh.. does anybody mind if i rip off that boost tracking code?
20:29:49Slasheriit's even defined incorrectly..
20:29:57Slasheridisabling it crashes it
20:30:19*Cassandra likes replaygain.
20:30:20preglowSlasheri: eh, don't we need to put in a lot of manual unboosting again then?
20:30:36Slasheripreglow: no, it's used only for debugging
20:30:45CassandraMinimises faffing with the volume control without doing nasty intrusive things to your media files.
20:30:54Slasheripreglow: and there could be a better implementation..
20:30:59thegeekfoobar2000 is teh win
20:31:11amiconnSlasheri: pondlife might...
20:31:16Slasherieach boost/unboost could include the filename and line number from code if necessary
20:31:31preglowSlasheri: ahh, that, i don't care much about that, no
20:31:36Slasherihmm, but he said too it could be implemented better..
20:31:56amiconnI think tracking could perhaps be logf-only
20:31:57Paul_the_NerdScrap it, and if someone complains about it being gone, tell them it's their job to make it work right. :-P
20:32:04*Cassandra is overawed with the huge response to rbutil 0.2
20:32:11Slasheriamiconn: true, that sounds much better and simpler
20:32:19Slasheriwithout complex syntax and defines..
20:32:23Slashericould be fully automatic
20:32:30CassandraDid anyone try it and like it?
20:32:41oboDidn't have windows to try it :(
20:33:27Cassandra*nod* I'd try to knock up a Linux version, but my Linux machine is currently misbehaving and won't boot.
20:33:58CassandraThat after I gave it some TLC to fix some overheating issues it was having *sigh*
20:34:15CassandraSometimes I think technology just hates me. Why did I ever become a sysadmin?
20:34:22 Join MarcoPolo [0] (
20:34:50*obo has asked himself that as well
20:34:58oboDo you have a Makefile for it?
20:36:09CassandraNope, but Code::Blocks runs under Linux.
20:36:35Slasherii think i will just remove the boosting code for now, it seems quite buggy and write into outside memory addresses..
20:36:43Slasherino wonder if it might crash system
20:36:45 Quit anathema (Read error: 113 (No route to host))
20:36:52amiconnBoosting, or boost tracking?
20:36:57thegeekcode::blocks is great;P
20:36:58Slasheriah, boost tracking
20:37:45Slasheriand the implementation is overkill..
20:39:18bluebrotherCassandra, I tried it from wine but it crashed the moment when it tries to access the network
20:39:31bluebrotherbtw, is it really necessary to install that first?
20:39:35Cassandrabluebrother, I had one report of it working in Wine.
20:39:43CassandraInstall what?
20:39:49bluebrotherinstall the tool
20:40:04bluebrotherthe download itself is some windows installer
20:40:10CassandraOh, nope.
20:40:25CassandraJust copy the file and the ini file into the same place then run it.
20:40:35CassandraOf course you need to be able to get to them to do that.
20:40:55bluebrotherIMO an officiall installer should be a simple zip file and be able updating a build while also running from the player itself
20:41:08bluebrotheri.e. when stored on the player of course ;-)
20:41:31amiconnbluebrother: On the zune it could even do what you said first ;)
20:41:32CassandraI think it'll run from the player if you install it there.
20:42:13bluebrotherif it doesn't need that installation I guess so ;-)
20:42:43CassandraWont save your settings across machines.
20:42:47CassandraBut it'll work.
20:43:43bluebrotherwhy, doesn't it save its settings in the program folder?
20:46:40amiconnSaving settings in the program folder is about the worst thing to do
20:46:59bluebrotherin most cases, yes.
20:47:00oboCode::Blocks don't seem to have made any source code available in the last year, only windows binaries :(
20:47:03amiconn(unless the program folder is located on removable storage)
20:47:17*n1s gives up on installing "codeblocks" and waits for someone to make the utility work in linux
20:47:24n1sobo they have svn
20:47:47oboah, now where is that hiding? Could only find a page saying they weren't using CVS anymore
20:48:01bluebrotherfor a rockbox installer it would be good if it could run from the device, and it should also save its settings in that case
20:48:24obothanks n1s
20:48:58n1snp, if you can get it to compile on your system please tell me what you did
20:49:10Cassandrabluebrother, nope. Saves them in an OS dependent place.
20:49:29CassandraObviously on a Linux system, trying to save things in the program's own folder is a bad thing.
20:49:39amiconnSame on windows
20:50:22Cassandraobo: Follow the links from the main page to the nightly build.
20:50:37CassandraIt's a bit involved, but you do get to the forum links in the end.
20:50:55oboThey don't have a binary for my arch
20:51:04bluebrothersure, but the installer should be capable of running as "portable app"
20:51:12oboor at least I didn't see one - x86_64 linux
20:51:21Cassandrabluebrother: Suggest a good way of detecting how you're running from the device, and I'll implement it.
20:51:48Cassandraobo: Ah. Right. :(
20:51:52lowlighti wonder if list scrolling could be improved by using all scroll lines which save the text...then.each increment would just offseting the scroll lines up/down and add one new line (instead of fetching all the lines for each increment)
20:52:08 Quit MarcoPolo ("Bye !")
20:52:22obolowlight: have you seen TPs old patch?
20:52:33Cassandraalthough I'm not exactly sure you'd want the download cache on the device.
20:52:44bluebrotherCassandra, maybe a command line switch? And use a local configuration if there is one in the programs folder
20:52:58lowlightobo: no...but i thought it did something with the framebuffer
20:53:13 Join MarcoPolo [0] (
20:55:00obolowlight: FS #5591 just changes to list.[ch]
20:55:35 Join Rondom [0] (
20:57:10CassandraThe one thing I really wish I could stick in the installer is a firmware patcher.
20:57:27 Part Paul_the_Nerd
20:57:45CassandraBit of a major project, that one though.
20:57:47oboCassandra: I'm working on it... :)
20:57:57CassandraSince every bloody firmware's different.
20:58:03Cassandraobo: tell me more?
20:58:33bluebrotherthe *patcher's should be libraries ...
20:58:34lowlightobo: ok...I must have been thinking of something else then
20:59:06obobluebrother: to save on binary size?
20:59:07amiconnlowlight: Fetching the text is neglectible. All the graphical operations take way more time
20:59:27amiconnBut without actually seeing it, I don't know how bad it is
20:59:44amiconnLists are scrolling with decent speed on all my targets
20:59:57bluebrotherobo, no, to make it possible to add it to e.g. the installer as dynamic library
21:00:34Cassandraamiconn: btw - any idea why the font compiler doesn't seem to be paying much attention to the unicode codes in jackash-16.bdf
21:00:35oboI don't see why you'd need to do that?
21:00:46Cassandraall my international characters are screwy.
21:01:08CassandraAnd I'm pretty sure I went through by hand and gave 'em all the right unicode codes.
21:01:08bluebrotheryou don't need to, but I think it would make it easier. No chance of code duplication
21:01:47CassandraI'd love to have a bunch of libraries I could just link in to the installer.
21:02:14Learamiconn: If you want to fetch text from (i.e. search) the database, it does take some time, apparantly.
21:02:41*bluebrother wanted to give librarifying the patchers a shot so he can get some knowledge of that library thingy
21:02:59*bluebrother googles for a "donate time" button for his little webspace
21:03:37lowlightamiconn:you (and I) have small screens though
21:03:54amiconnThe H300 screen isn't small...
21:04:38obolowlight: currently my 5G scolls at 4.9 lines per second
21:04:47lowlightok then
21:04:56*goffa returns from lunch $.90 poorer.. damn high stakes gambling
21:05:39goffai read the convo about players.. i like mpd with gmpc as a front end
21:05:49goffaand sometimes no gmpc
21:05:58goffadepending on if i'm in the mood to see track names
21:06:01CassandraI still like badgers. It's a shame you can't play MP3s on them.
21:06:06***Saving seen data "./dancer.seen"
21:06:22*tehsmo likes the mpc frontend, since it works nicely over an ssh session for shoutcast stream control
21:06:37goffawell... before i switched to dapper i did :)
21:06:58CassandraYou know, I'm confused. I really don't think the 5g scrolls too slowly.
21:07:03goffanow i'm on edgy... so moving further from tunes on the badger
21:08:10Cassandragoffa - phear my geek cred. Colin Watson and Ian Jackson of Canonical are both personal friends. ;)
21:08:39CassandraOh, yeah, and I've met that Daniel Stenberg too - total loser, if you ask me.
21:09:16goffai met a unix user once
21:09:22preglowok, i'm tired of this fucking apple now
21:09:52Cassandrapreglow, maybe you're too young for it.
21:10:09goffathat's gonna leave a mark
21:11:48 Quit markun ("leaving")
21:11:58amiconnobo: H300 scrolls at about 15 lines per second. That's unboosted, and within a tagcache list
21:12:12 Join markun [0] (n=markun@rockbox/developer/markun)
21:12:37*bluebrother failed to find a time donation button / service
21:13:18*obo found one called redundancy
21:13:35 Part Jagan81
21:17:59preglowi've got a feeling i'm looking for the error in some completely different spot than where i'll fin dit
21:19:02 Join Rob2222 [0] (
21:22:54 Quit akaias (Read error: 110 (Connection timed out))
21:25:28 Join akaias [0] (n=akaias@
21:29:06amiconnHmm, someone should update the model list on the frontpage a bit
21:32:45amiconnBagder: If someone updates files in the www module, do those changes show up automagically?
21:33:07 Join Rob222241 [0] (
21:33:13markunamiconn: change the list in what way?
21:33:47amiconnG5.5 is supported now, except 80GB model
21:33:51Slasherilinuxstb: hmm, maybe you would like to answer to that..
21:34:04Slasheri(he needs probably the updated ipodpatcher utility)
21:34:21zylchedoing it manually on linux is funner <.<
21:34:43Slasheriexcept cvs bootloader for ipod does not support 5.5G yet
21:34:44Slashericore does
21:35:03lexamiconn: but
21:35:09lexamiconn: when can we boot into apple os too
21:35:10 Quit zylche (Remote closed the connection)
21:35:38 Join zylche [0] (
21:35:41amiconnThat's not a requirement for being counted as supported
21:36:03amiconnOn X5 and H10 we don't have (official) dual boot either
21:36:15lexbut like on other ipods you have
21:36:31lexand without apple os you can't take everything off from your ipod
21:36:57lexit's now like... 2 hours of music or 14 hours of music and videos
21:37:06lexbut not 14 hours of doom :D
21:37:28amiconnI get about 9 hours of music from my ipod... and I've scrapped the of
21:37:42 Quit perplexity (Read error: 110 (Connection timed out))
21:38:57amiconnSlasheri: What needs to be done in the bootloader? Just rebuild it with the new fat driver, or is there more?
21:39:18Slasheriamiconn: it needs change to the load address and some padding
21:39:27Slasherii haven't yet figured out how to auto-detect that
21:39:54Slasherior infact.. that was the ipod patcher
21:40:00Slasherinot the bootloader itself
21:40:26Slasheriyep, cvs bootloader works fine
21:40:52Slasheriand yes, i have already started adding a command line option for the 5.5g..
21:41:00Slasherididn't even remember that ;)
21:41:29 Quit Cassandra (" HydraIRC -> <- 100,000+ downloads can't be wrong")
21:41:54Slasherinow trying to finish the iriver flashing so i could do some commits and clean-up local repository :)
21:43:29Slasheriamiconn: however, that rom loading problem and crash on set_cpu_frequency seems to be a tough one :/
21:44:14amiconnI suspect something related to rom waitstates, but the code doesn't look suspicious...
21:44:37Slasheriamiconn: and all further cpu_boost:n works fine
21:44:58Slasheriso that's really weird.. even cpu_boost(false); should do the excact same thing as set_cpu_frequency
21:45:09Slasherisomehow it's only the first call that is causing the issue
21:45:44amiconnBut if that first call happens later, it's no longer an issue iiuc?
21:45:59Slasherithen it never crashes
21:46:44 Join [TEHb]_ [0] (
21:46:59preglowvery weird
21:46:59[TEHb]_linuxstb: Hi.
21:47:05[TEHb]_linuxstb: there?
21:47:34Slasheriamiconn: if you would like to try, i can send a patch and working bootloader
21:48:03 Join bluey- [0] (
21:51:38 Quit Rob2222 (Read error: 110 (Connection timed out))
21:52:50Slasheriamiconn: in fact i got some idea to try.. just a moment
21:52:52 Join bluestorm [0] (
21:53:23bluestormMy mom just bought a Ipod Nano
21:53:32bluestormi was considering putting rockbox on it
21:53:41bluestormhow can i know wich "generation" it is ?
21:54:00bluestorm(as it seems 2nd one isn't supported, whereas 1st is)
21:54:04Slasherii have no idea.. maybe somebody else here knows
21:54:10Slasheridid you search rockbox wiki?
21:54:45zylcheSlasheri, Quickly.
21:54:48scorchebluestorm: is the whole thing metal or is the front plastic and the back metal?
21:54:55bluestormgonna look
21:54:59[TEHb]_Guys, what's really mean -mlong-calls gcc option?
21:55:07*zylche tuts
21:55:10scorchestop please..
21:55:13zylcheah hah
21:55:20*zylche loops
21:55:23preglow[TEHb]_: it's for making gcc use long calls on arm
21:55:39[TEHb]_preglow: Already read description, but anyway, can't understand, what's really changed in the ASM?
21:55:42preglow[TEHb]_: ordinarily, arm uses short local jumps for all function calls, but we can't use that yet thanks to iram
21:56:01preglow[TEHb]_: what changes is that all function calls use 12 byts instead of 4
21:56:13preglowalso they execute around four cycles slower
21:56:14bluestormscorche: full metal
21:56:17bluestorm(execpt the roll)
21:56:23scorchethen it is a second gen
21:56:23preglowperhaps even more
21:56:32scorchethe roll?
21:56:38bluestormso there can be no rockbox for it ?
21:56:45scorcheatm, no
21:56:47bluestorm(the with circle used for selection)
21:56:58bluestormshould it be available soon ?
21:57:04scorchedoubt it
21:57:39[TEHb]_preglow: sorry, don't know, what's 'irams'. Just found this channel in the
21:57:54bluestormit's a shame ^^
21:57:57[TEHb]_preglow: BL used for doing short calls?
21:58:43*[TEHb]_ working with Intel Xscale CPU's
21:58:53scorchenot really..the 2nd gen is vastly will be a large amount of work
21:59:13preglow[TEHb]_: iram is ram that is internal to the cpu, the problem is it's address, which is 0x40000000
21:59:23preglow[TEHb]_: and that's out of bounds for arm's usual short relative jumps
22:00:39 Quit Kingstone ()
22:00:39[TEHb]_preglow: hmm, you mean, that we can't specify such big offset in the shift ?
22:00:57 Part bluestorm ("Konversation terminated!")
22:00:58preglow[TEHb]_: the branch instructions can only specify offset +/- 32 megs in size
22:01:14preglow[TEHb]_: ordinary ram is mapped at 0, iram is mapped at 0x40000000, which is way too much
22:01:35[TEHb]_I'm very interesting in switch method
22:01:42markunpreglow: and gcc doesn't choose from call to call what kind of jump it should use?
22:01:51preglowmarkun: sadly, no
22:01:57preglowmarkun: we haven't found a good way to make it do so yet
22:02:07preglowmarkun: it's really braindead, the only jumps we want to be long are those too and from iram
22:02:17preglowwhy it's so hard telling gcc that i don't know
22:02:29preglowseveral tens of kilobytes of rockbox code is wasted on those long valls
22:02:47amiconnThe current way to tell gcc when to use long jumps is (1) not sophisticated enough to tell what we really want (2) somewhat buggy
22:03:21amiconnGcc can be told to use long jumps whenever a function uses a section attribute
22:03:34Slasheriamiconn: ok, found the issue..
22:03:40Slasheriflash wait states seem to be wrong
22:03:45[TEHb]_but we can use longcall attribute for such functions?
22:03:51preglow[TEHb]_: nope
22:03:55preglowdoesn't work
22:03:58Slasheriwhen i changed it bigger, booted fine
22:04:05amiconn[TEHb]_: The longcall attribute is another such quirk
22:04:33amiconnThe necessity for long calls does neither depend solely on the caller nor solely on the callee, it depends on both
22:04:44amiconnIt seems gcc doesn't know that...
22:05:04Slasheriamiconn: but it's weird why it crahes only at the beginning..
22:05:17Slasheriand not later
22:05:22amiconnAn iram function could call another iram function with a short call, the same way as a dram function could call another dram function
22:06:03amiconnThere should be a way to tell gcc to use long call if, and only if, the destination lies in a different section
22:06:22[TEHb]_Therefore, the easiest way - just use -mlong-call flag ? :)
22:06:29amiconnEven better would be an option to specify section pairs for which to use long calls
22:07:35amiconn(and regarding the bugginess - you can't rely on the section attribute thing when not using -ffunction-sections. Sometimes it doesn't work)
22:07:37Slasheriamiconn: or could it be possible while PLL is locking, it could momentarily put cpu running at too high speed?
22:07:52Slasheriamiconn: so maybe after PLL has locked, we could restore the lower wait state
22:07:57amiconnThe cpu runs at plain xtal clock until the pll is locked
22:08:05amiconnThis is automatic
22:08:57Slasherithen probably 1 wait state is just not enough..
22:09:44Slasherii will change it to 2
22:09:59amiconnWhy would it only crash the first time, and only if this first time happens early?
22:10:10Slasherii have no idea..
22:10:58Slasheriat least CPUFREQ_MAX works ok with 4 wait states
22:11:19preglowgood thing we've got a code cache
22:11:46 Quit mirak (Operation timed out)
22:12:05Slasheripreglow: yep, i haven't even noticed a slowdown when running from rom
22:12:19Slasheribut need to do more measurements..
22:12:25amiconnRAM isn't exactly fast either
22:13:46amiconnOn coldfire targets the ROM is 16 bits wide, so no slowdown due to bus width
22:14:13amiconnBut even on archos where the ROM is only 8 bits wide, running from rom pays off
22:14:30amiconn..even though the slower execution speed is sometimes noticeable
22:14:30lowlightobo: had a look at that only optimizes the list when no text is changing & just the cursor is basically the beginning & end of the list display.
22:14:55Slasheri124/4 = 31, 45/1 = 45
22:14:57amiconnNamely when using tagcache^h^h^h^h^h^h^h^hdatabase
22:15:05Slasheriso it might be very possible 1 wait state is not enough..
22:15:21amiconnSlasheri: Very wrong calculation...
22:15:28Slasherihehe :D
22:15:29amiconnThe basic bus cycle is 3 clocks
22:15:38Slasheriah, i haven't read the spec yet
22:15:49amiconnSo 1 waitstate makes that 4 clocks, and 4 waitsates makes it 7 clocks
22:16:08*amiconn recommends MCF5249UM.pdf
22:16:16Slasheriyep, i have it :) checking now
22:17:13Slasheriamiconn: ok, then it should be more than enough..
22:17:38amiconnI cross-checked the timing with the eeprom datasheet
22:18:00amiconnBah, whatever_prom
22:18:35obolowlight: more useful with paged scrolling then?
22:20:40 Quit PaulPosition (Read error: 145 (Connection timed out))
22:21:15shriftbagder: who should I speak with about wiki editing privileges?
22:23:41shriftdoes anybody know to whom I should speak about wiki editing privileges?
22:24:05 Join mirak [0] (
22:24:06n1syes anyone who already has wiki editing privileges
22:24:22shriftwell, can anyone give "BrendanMartens" editing privileges, please?
22:24:36n1sif you promise not to spam :-)
22:24:44shriftI promise. : )
22:24:48Slasheriamiconn: hmm, now i don't understand.. looking at the spec it says that wait states are on bits 13-10
22:25:12Slasheribut looking at source, it seems bits 12-09 are used
22:25:36Slasheriand bit 9 is reserved, or has no defined function
22:25:48amiconnThe code uses bit 13..10
22:26:00Slasherireally? i will double-check
22:26:16amiconnYes, definitely
22:26:35amiconn0x0580 sets bit 10, 8 and 7
22:26:53n1sshrift: done
22:26:57shriftthank you, n1s
22:27:55Slasheriwas just so long bit string i readed it wrongly from the calculator
22:28:59*preglow does all his hex<->bit conversions in his head
22:29:00 Join Lars_G [0] (n=lars@unaffiliated/lars-g/x-000001)
22:29:12preglowproving i am a healthy young male
22:29:20*amiconn usually does the same
22:29:32Slasheripreglow: i trust more on the calculator as it's not so easy to make mistakes :)
22:29:42Slasheriand those mistakes are really hard to find
22:29:57preglowi do use a calc when the numbers get too long
22:30:08 Part Lars_G ("Leaving")
22:30:29Slasherihehe, sure. simple byte or nibble-byte calculations are also easy to do in head
22:35:52Slasheriamiconn: so PLLCR &= ~1; should immediately switch main clock to the xtal?
22:38:41 Quit Lear ("Chatzilla 0.9.75 [Firefox 2.0/2006101023]")
22:39:13markunI'm unable to enter the "View HW info" menu. Is there anything special I need to change for it to work?
22:40:00blue_lizardhallo all
22:40:05amiconnSlasheri: Yes
22:40:12blue_lizardis this archos player capable of using rockbox?
22:40:19amiconnmarkun: Yes, you need to implement it...
22:40:19Slasherimarkun: did you check debug_menu.c?
22:40:30 Quit bluey- ("Leaving")
22:40:39amiconnThe default function just returns imediately, so it appears as if you can't enter it
22:40:48 Join jba_ [0] (
22:41:07Bagderblue_lizard: nope
22:41:26markunamiconn: I thought I implemented everything.
22:41:33blue_lizardit would have been too nice
22:41:49markunI setup the right address in dbg_flash_id
22:42:28markunamiconn: ah, I see it now
22:42:44 Quit nudelin ("Hammer of Dawn is offline.")
22:46:27dan_aHmmm... does anybody know why on the PortalPlayer units we disable the I2S FIFO and then the interrupt when we stop playback, but disable the interrupt first when we pause it (this is in pcm_playback.c)? Does the order matter at all?
22:47:02 Join petur [0] (n=petur@rockbox/developer/petur)
22:47:28 Join forehead [0] (i=54bd7ef6@gateway/web/cgi-irc/
22:47:37peturoh please
22:47:48*petur slaps forehead
22:48:11Bagderdan_a: btw, I created today to have something to point users to
22:48:25 Join Soap [0] (
22:48:40*forehead pays back all the slapping
22:48:49 Quit forehead (Client Quit)
22:49:33dan_aBagder: Excellent. I'll update it with the obvious things ("No, Rockbox hasn't drained your battery that quickly") shortly
22:49:41amiconnBagder: Got my question?
22:49:45Bagderamiconn: nope
22:49:55*Bagder gets it now
22:50:05Bagderyes, the www module is supposed to update automaticly
22:50:38 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
22:50:44*petur is *very* happy: squeezebox2 returned from repairs and playing nice as ever
22:52:45linuxstb_Bagder: If you have a couple of minutes, can you put in and also extract ipodpatcher.exe from the zip file and put that in the same place (replacing the previous version).
22:53:14 Join nudelin [0] (
22:53:24Bagderwill do, in a few minutes
22:53:25spugthe font color in the picture menu is white instead of black like it is in the rest of the menus
22:53:35 Join CriamosAndy [0] (
22:53:41linuxstb_Bagder: Thanks.
22:53:44spughow come? is this color set somewhere else than the regular one?
22:53:55spugit'd be cool if i could read the picture menu text
22:53:56 Quit linuxstb (Nick collision from services.)
22:54:06 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
22:54:10n1sspug: in the jpeg viewer?
22:55:22 Join Davide-NYC [0] (
22:55:31 Quit Davide-NYC (Client Quit)
22:55:33amiconnlinuxstb: Does that one include G5.5 support?
22:55:45 Part lowlight
22:55:46linuxstbipod_fw.c still needs fixing though.
22:55:52spugn1s: i can't read any of the text in the JPEG viewer because of the color
22:56:10linuxstb(there's a patch, but it's not in CVS, and I don't know if the patch good...)
22:56:33n1sit used to have black background with white text but then someone changed so the backdrop is still there, I'd like it changed back as displaying a picture on top of another is not always a good idea
22:56:55linuxstbAFAIK, dual-boot is broken on the 5.5g - that's probably caused by ipod_fw.c not building the combined firmware image in the way the Rockbox bootloader expects.
22:57:22n1sspug: maybe I'll change it tomorrow if I find the time.
22:57:27amiconnThe background should be black while showing the jpeg, but the menu could continue using the user-selcted ui colours and backdrop
22:58:30n1samiconn: sounds reasonable, I use white text anyway so I didn't realize that was hardcoded before
22:58:42Kasperlelinuxstb: i think kalthare's original patch breaks ipod_fw for other ipods
22:58:45 Join Paul_the_Nerd [0] (
22:59:38 Join GliTch_ [0] (
23:00:42 Join Davide-NYC [0] (
23:01:00spugn1s: thanks a lot'
23:01:17linuxstbKasperle: Yes, I think it does. It should be possible to detect what kind of firmware image is being processed, and adjust automatically - i.e. all 5G users (old and new) would continue to just use the same "-g 5g" option.
23:01:20spugn1s: should i submit a bug report or something?
23:05:13 Join PaulPosition [0] (
23:06:08***Saving seen data "./dancer.seen"
23:07:51dan_aBagder: Any new info from MrH? It sounds like he's got a whole pile of things he's nearly ready to tell us about...
23:08:19Bagderno news since that last one, no
23:08:53 Join dantheman [0] (n=danielma@
23:08:58 Join linuxstb_ [0] (i=5343d4aa@gateway/web/cgi-irc/
23:09:51dan_aI've had a bit of a play with DMA on the flash, but not got it working reliably
23:10:21 Quit shrift (Remote closed the connection)
23:10:40Davide-NYCjhMikeS: what can I do to help with determining the validity of the recording bug?
23:10:55Bagderlinuxstb: those files are now in position on
23:11:15 Quit Criamos (Read error: 110 (Connection timed out))
23:12:29 Join merbanan [0] (
23:12:47 Quit dantheman (Read error: 104 (Connection reset by peer))
23:13:26linuxstb_Bagder: Thanks.
23:13:56 Quit Rondom ("Ex-Chat")
23:14:27amiconndan_a: Btw, I think ata-e200.c should be renamed
23:14:33amiconnIt's just not ata
23:14:58amiconnAll it has to do with ata is that it resembles the ata.c api
23:15:01 Join lee-qid [0] (
23:16:27preglowi don't really care, but people know how to use svn
23:17:09*amiconn is waiting for the svn move as well
23:17:28 Quit freqmod (Remote closed the connection)
23:17:43amiconnThere is a couple of files which should be renamed or moved as-is, but I don't want to lose the history
23:18:05preglowBagder: make a FinallySVN wiki page which lists what scripts need to be rewritten, then wait until they are :>
23:18:08Slasherihmm, one simple way to speed up scrolling on ipod seems to be limit the lcd refresh rate
23:18:19Slasheriand actually works quite well
23:18:37preglowSlasheri: speeding up scrolling on ipod should be done by rewriting the wheel handling, if you ask me
23:18:43 Quit MarcoPolo (Remote closed the connection)
23:18:48Slasheripreglow: yes, that of course too
23:18:55Slasheribut currently lcd is lagging too much
23:19:06Slasheri(unless pages scrolling is used)
23:19:09preglowand the best way would be to allow moving more entries in one go than just one
23:19:12Slasheriand even then
23:19:27 Join webguest91 [0] (i=5873b21c@gateway/web/cgi-irc/
23:19:28preglowbut of course, the 5g is slow as hell in just drawing too
23:19:37dan_aamiconn: I've got no problem with renaming it - although at the moment it's obvious what file ata_read_sectors will be in, even if there is no ATA in the code.
23:19:41Slasherihmm, indeed
23:19:44amiconnSlasheri: A good reason for optimisation...
23:19:49 Join MarcoPolo [0] (
23:19:59Slasheriamiconn :)
23:20:33amiconndan_a: Hmm, not really. The ather ata-<target>.c files only provide target specific bits & pieces for use in ata.c
23:20:55amiconnata-e200.c is self-contained, much like ata-mmc.c is for the Ondio
23:21:19amiconn(and ata_mmc.c is another candidate for renaming + moving)
23:21:48Slasheriamiconn: hmm, atm i couldn't find other solution than just using wait state 2..
23:21:50PaulPositionAny of you guys have any idea why I would get data abort on turning crossfeed on? Could it have to do with Barrywardell's bootloader which I tried (and love, dualboot!)?
23:22:00Slasherior boost cpu first to MAX frequency
23:22:43Slasherior maybe the pll is incorrectly initialized..
23:23:03Slasherithat would explain it
23:26:03habanaPaulPosition: how do you make the choice to boot one or another ? just finished to make the same for sansa
23:26:21dan_ahabana: Press left to boot to OF
23:26:39dan_a(hold it down until it says that it's loading Sandisk firmware)
23:27:19habanahas someone tested the graphical bootloader from Gnelink ?
23:28:45amiconnSlasheri: That doesn't make sense...
23:29:02dan_ahabana: Last time I checked it, it wasn't working properly (because the timer interrupt does not get enabled in the bootloader)
23:29:32amiconnSomething is running wild in an inconsistent way... at least there is no visible consistency
23:29:46 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
23:30:15Slasheriamiconn: hmm, true.. that might be also possible, then then it must be some interrupt
23:30:28 Join My_Sic [0] (
23:31:16amiconnHmm, now that you say it... interrupts aren't disabled within set_cpu_frequency()
23:31:30amiconn...and they shoudn't, because this function can take some time
23:31:45 Quit Rudy4Pez (Read error: 110 (Connection timed out))
23:31:49Slasherii will try that :)
23:31:56 Quit merbanan (Remote closed the connection)
23:32:04amiconnYou probably got that wrong...
23:32:14Slasheriah, hmm
23:32:18amiconnInterrupts should _not_ be disabled in set_cpu_frequency
23:32:43amiconnThis function can take several milliseconds, so disabling interrupts is out of question
23:32:44preglowdisabling interrupts for a whole pll relock would be very nasty
23:32:51Slasheribut disabling them temporarily would reveal if the problem is with interrupts
23:32:52amiconnYOu can do this as an experiment though
23:33:17Slasherihmm, but now need to go.. more testing tomorrow :)
23:33:27amiconnIf an interrupt happens after reprogramming the pll but before reprogramming the waitstates, ugly things might happenn
23:33:53amiconnHmm, but we could disable interrupts immediately before reprogramming the pll, and enable them after reprogramming the waitstates
23:34:09amiconn...before the wait-for-relock loop
23:36:38amiconnAnother option would be to track whether we're switching up or down (only relevant for CPUFREQ_NORMAL), and reprogram waitstates before the pll when switching up, and after the pll when switching down
23:39:34 Quit Davide-NYC ("Chatzilla 0.9.77 [Firefox 2.0/2006101023]")
23:40:13Slasheriamiconn: hmm, disabling interrupts didn't work either
23:40:32 Join TeaSea [0] (n=thunderc@
23:40:40Slasherieven after PLL has locked, if we change wait states back to 1, it will crash immediately
23:40:47pregloware you sure they are enabled?
23:40:56 Join webguest20 [0] (i=3e41833e@gateway/web/cgi-irc/
23:41:10Slasheripreglow: wait states? yes, those should be correct
23:41:18webguest20jhMikeS: ping
23:41:33amiconnSlasheri: Uh, but when using 1 waitstate later it does work, provided we don't switch to CPUFREQ_NORMAL during boot?
23:41:33 Quit PaulPosition ()
23:41:46amiconnThat doesn't make sense....
23:41:53Slasheribut if we first put cpu to CPUFREQ_MAX and after that -> CPUFREQ_NORMAL, no crashing with wait states 1
23:42:01Slasheriamiconn: correct
23:42:05amiconnI think this is a case for a logic analyzer run...
23:42:09Slasheriamiconn: cpu needs to be first at max frequency
23:42:21Slasherionly after that 1 wait state works..
23:42:44linuxstb_preglow: I just read your earlier question earlier about bootloader.img - did you figure it out?
23:42:47amiconnLemme check the datasheet again
23:42:49preglowlinuxstb_: no
23:43:02preglowi just ignored it
23:43:41linuxstb_It's a ready-to-dd boot partition image with the bootloader but no Apple OS - just dd if=bootloader.img of=/dev/sda1
23:44:15preglowah, like i guessed, then
23:44:35preglowdo you know where the apple bootloader loads us, btw?
23:44:35linuxstb_It's made by scramble.
23:44:39preglowstart of ram?
23:45:00linuxstb_Yes, 0x1000000.
23:45:00amiconnlinuxstb: On ipod it should be possible to build a rockbox.bin which can be put into the firmware partition, correct?
23:46:06linuxstb_Yes, I think I tried it successfully. A bit of a pain to upgrade though...
23:46:43amiconnI think that would be a somewhat convenient method for debugging the G5.5/80GB
23:47:02Kasperlelinuxstb_: why is that a pain to upgrade?
23:47:20 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
23:47:49 Quit MarcoPolo (Remote closed the connection)
23:47:52Paul_the_NerdKasperle: Having to dd or ipodpatcher every time you wanted to update your Rockbox version?
23:48:24Kasperleduh. i just confused the bootloader and rockbox :/
23:48:34Soapseperates the wheat from the chaff ;)
23:48:41 Join TeaSeaLancs [0] (n=thunderc@
23:48:58Kasperleyou could later integrate a feature into rockbox, that moves an upgrade file to the firmware partition though
23:49:10Paul_the_NerdMuch like the flashing for rombox, etc.
23:49:34Paul_the_NerdOne could click a .ipod file and have it Rolo, or choose open with, and "update.rock" it into the boot partition.
23:52:25 Quit blue_lizard (Read error: 110 (Connection timed out))
23:52:26 Part n1s
23:52:35 Part webguest20
23:52:47habanadan_a: little troube with OF into Rockbox bootloader (again for Sansa)
23:53:02dan_awhat's the problem?
23:53:05amiconnSlasheri: Hmm, I checked both datasheets again.
23:53:36 Quit _FireFly_ ("Leaving")
23:53:39habanai've got a "rockbox error: -4"
23:53:40amiconnThe slowest version of the flash eprom has 90ns /CS access time,
23:53:56habanai know for -1 ut not for this one
23:54:06preglowamiconn: there are several versions in use?
23:54:13habanais there a faq for these error messages ?
23:54:13amiconnand the coldfire read cycle fetches data (1.5+waitstates) bus clock cycles after enabling /CS
23:54:41amiconnbusclk == cpuclk/2, so one bus clock cycle is ~44ns at CPUFREQ_NORMAL
23:54:58amiconnpreglow: Probably not, but I don't know which version is used.
23:55:08preglowamiconn: ahh, like that
23:55:13Slasherioh, so that could be the issue
23:55:15dan_ahabana: I don't think there is a FAQ. Error -4 when you are trying to load the OF means that it couldn't read the whole file.
23:55:17amiconn2.5*44 == 110, so we're on the safe side
23:55:32amiconn...with 1 waitstate
23:55:41dan_ahabana: Have you definitely got a /.rockbox/OF.bin file?
23:56:03Slasheribut i will debug more tomorrow, now have to go :
23:56:04habanadan_a: i have a 6 GB version. let me verify for your answer
23:56:46dan_ahabana: Can you check how much free space you have, too?
23:57:01*jhMikeS sees he's been pinged a couple times but has been busy
23:57:14habanadan_a: not a free spaxe trouble
23:57:18My_Sichi every one
23:57:28habanaMy_Sic:hello salut
23:57:57My_Sicwhen the new version of rockbox will be released ?
23:57:58habanaMy_Sic: tu as quel lecteur ?
23:58:05My_Sicipod 5G 60Go
23:58:17My_Sicanciennement j'avai un H120
23:58:21Paul_the_NerdMy_Sic: What do you mean by "new version"? There are new builds frequently.
23:58:31dan_ahabana: I'm just wondering if you have used more than 4Gb already.
23:58:51habanadan_a: let me reboot

Previous day | Next day