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 2006-01-12

00:00:08tomalI am new here. I am the one that is working on the ifp port.
00:00:30Midgey34nice work so far
00:00:33BHSPitMonkeywhat's an ifp?
00:00:41Midgey34iriver flash player
00:01:12Bagdertomal: I guess you know that linuxstb and preglow are mr ipodport
00:01:15linuxstb_Welcome - I'm one of the ipod devs.
00:01:32tomalBagder: Yes I know them
00:01:47linuxstb_You've just missed preglow though.
00:02:20tomalIs there anything you would like to ask me about the patch I sent recently?
00:03:21linuxstb_No - as I said in my email to rockbox-dev, I'm in favour of it being committed to CVS.
00:04:00BagderI think one of you arm guys should do that
00:04:00 Quit webguest60 ("CGI:IRC (EOF)")
00:04:50 Join webguest68 [0] (
00:04:51linuxstb_I'll be happy to - if there's no objections from anyone else.
00:05:04tomalI'll be happy too :)
00:05:12Bagdergo go go ;-)
00:07:37linuxstb_OK, I should be able to do it in an hour or so.
00:07:47 Quit xmixahlx ("blah blah blah")
00:07:57BHSPitMonkeydo what?
00:08:00 Join fobbo [0] (
00:08:11linuxstb_tomal: Is the patch on Sourceforge still the latest, or do you have more?
00:08:25ashridahBHSPitLappy2: commit a the ifp patch to cvs
00:09:02saa[b_r]iderMidgey: that makes sense! now I know what ifp and ihp mean!
00:09:03ashridahhmm. correct me if i'm wrong, but aren't there drastic differences in capabilities across the iFP range?
00:09:10linuxstb_BHSPitLappy2: The ifp also has an arm processor - so there is code we can share between the ipod and ifp projects.
00:09:16tomallinuxstb_: It is the latest. I haven't had much time for rockbox in last few days.
00:10:09linuxstb_tomal: OK - it's a very impressive solo effort so far.
00:10:17peturCan somebody tell me how to create one patch for multiple files in multiple directories?
00:10:17Bagderyes indeed
00:10:38Bagderpetur: make single ones and append them to a single file
00:10:55Bagderas long as you keep the paths "sane" it works brilliantly
00:11:05peturah ok...
00:11:19Bagderand by "sane" I mean having the same root
00:11:39peturI still don't know what's wrong with the ones TortoiseCVS generates :( will do it the manual way...
00:12:06petur(iRiver recording gain patch underway...)
00:12:38tomallinuxstb_: Thanks. I suppose I am not too far from playing audio, but it may take few weeks.
00:12:38linuxstb_tomal: What's the next step for the ifp port?
00:13:16Bagdertomal: you'll get all my thumbs up if you beat the ipod guys to music playback ;-)
00:13:24 Quit menollo ("CGI:IRC (EOF)")
00:13:28tomallinuxstb_: Currently I am working on a USB driver that will make the player behave like a serial port
00:13:44*Bagder turns it into a competition
00:13:44tomallinuxstb_: I'd like to write a GDB target for the player
00:14:18tomallinuxstb_: That would make experiments with it easier.
00:14:39linuxstb_I'm sure that would be useful. We should do the same for the ipod - there's a serial interface in the dock connector which is used to communicate with peripherals.
00:15:13linuxstb_But I'm not sure if it's working under ipodlinux for the latest ipods.
00:15:25tomallinuxstb_: I have also identified parts of the original firmware responsible for audio output and probably for setting bus and cpu clocks.
00:16:19 Quit uwe_ (Read error: 110 (Connection timed out))
00:16:45 Quit ender` (Read error: 113 (No route to host))
00:16:52linuxstb_What do you use for disassembling? IDA Pro?
00:18:01tomallinuxstb_: I used IDA Pro. Now I am using my own code analyzer (yes, I know, some of you already think I am insane)
00:18:16Bagderoh yes we do
00:18:17 Join Blipus [0] (
00:18:28Bagderbut good insane
00:18:36 Join uwe_ [0] (
00:18:37 Join wizisi2k [0] (
00:18:42 Part wizisi2k
00:19:13Blipusi hear people here are disassembling ipod firmware?
00:19:31tomallinuxstb_: The analyzer is written in Lisp. It can, for example, find out that at a specific point in code there is a constant in a specific register
00:19:48linuxstb_Blipus: Nope. But needing to.
00:20:03BHSPitMonkeyBlipus: :P sorry
00:20:09BHSPitMonkeyand that's not what i told you
00:20:20Bagderipod, ifp, iriver, gigabeat, iaudio. its a general dissassemblefest
00:20:21tomallinuxstb_: Very useful when the compiler decided to put a constant to a register at the beginning of a long procedure.
00:21:58tomalI haven't released the analyzer, because it's quite user unfriendly. But if you would like to see it, or its results, I may send something.
00:22:12 Quit lodesi ("Leaving")
00:22:21XavierGrtomal: I am jealous of your work. It should be very fun to do such work. I am thinking to disect my old crappy napa flash player!!
00:22:29linuxstb_Is it very specific to the ifp, or could it help other arm players?
00:22:52linuxstb_(Toshiba gigabeat and ipod)
00:23:13tomalI guess it could be used for others, unless the C compiler used made a strange code.
00:23:31darklessI've been searching on, here (the irc logs) and head-fi... Noone seems to know if the iriver is capable of producing bit-perfect output via optical out
00:24:08darklessI can't test it myself since I don't have anything w. bit-perfect digital in
00:24:28tomalIts use requires some Lisp knowledge (there is no user interface, you just call Lisp functions), but I can write a little howto.
00:24:41linuxstb_I believe people have outputted DTS audio via the optical out - implying that it is bit-perfect.
00:24:48adamdullentywhich iRivers have optical out?
00:24:54XavierGrdarkless: In the original firmware it fails to do so.
00:25:00adamdullentycool, didnt know that
00:25:13darklessadamdullenty: the entire H1X0 series
00:25:26XavierGrBut I think that this is the case in Rockbox too I am not sure.
00:26:54peturwhat's wrong with 'cvs diff -u apps/settings.c > gainpatch1.txt' ? it says 'no such directory'
00:27:11peturexecuted from rockbox dir
00:27:37darklessthanks linuxstb_, it does seem like it works
00:27:58 Join Jungti1234 [0] (n=jungti12@
00:31:54 Join Kyl3 [0] (
00:32:34 Join ts-x [0] (
00:35:48 Join einhirn [0] (
00:36:35ashridahgreat. something's changed. now i'm getting stock spam all of a sudden
00:38:40 Quit Kyl3 ("( :: NoNameScript 3.81 :: )")
00:38:59Jungti1234Bootloader V5!!
00:39:40Jungti1234It's no problem.
00:41:16 Join einhirn_ [0] (
00:42:02peturIf anybody cares to try the new iRiver recording gain control, the patch is out:
00:42:34 Quit webguest89 ("CGI:IRC (EOF)")
00:43:44darklessalright, so now that we've established that the iriver does seem to have bit-perfect optical out, what settings affect the bitstream? I know that crossfeed and replaygain does, anything else?
00:45:29XavierGrpetur what does it do exactly?
00:46:28peturanalog and digital gain are merged into one big range, alows to select both channels to change them together.
00:46:55XavierGrin the main record screen?
00:47:10XavierGrnice indeeed
00:47:15peturso no more gain settings in recording settings
00:47:21XavierGrI was always annoyed with the 2 gain settings
00:47:41XavierGrIt should be commited....
00:47:42darklesspetur: very nice
00:47:42peturthere's an indicator to tell you if the gain change you do is analog or digital
00:48:02peturand it should preserve best quality
00:48:12petur(first do analog gain, then digital)
00:48:21 Join akaidiot [0] (
00:48:23peturlike a digital camera ;)
00:49:08peturjust: I'm not sure I created the patch file correctly...
00:49:29Paul_The_Nerd"Digital Zoom" always seemed like the kind of thing you never wanted to actually use.
00:50:11Paul_The_NerdI mean, can't you just use magnification in whatever program later?
00:50:29peturwe had this discussion some days ago
00:50:54Paul_The_NerdWell, regardless, I do like the idea behind your patch.
00:51:03peturthe decimator in the UDA 1380 acts on the 24bits of the ADC output which is then reduced to 16bit
00:51:23peturso it's better than what you'd do to the wav file
00:51:50*petur gives credits to amiconn for the idea
00:52:20webguest68since when does pressing left in the root dir bring you to the WPS? or was it always like that?
00:52:27 Part LinusN
00:52:32linuxstb_tomal: Are you using the same gcc/binutils as we're using for the ipod? i.e. gcc 3.4.x and binutils 2.16 compiled for arm-elf?
00:53:34tomallinuxstb_: No. gcc 3.3.6 and binutils 2.15.
00:54:06linuxstb_Is there a reason for 3.3.6?
00:54:18Midgey34huh? −−> zip warning: name not matched: .rockbox/rocks/Copy
00:54:31Bagderspace in the file name
00:54:43Bagderbuildzip not like
00:54:45tomalNo. I have chosen it almost randomly
00:54:45Midgey34oh alright
00:55:03Midgey34I accidently copied a rock...
00:55:20 Quit einhirn_ ("Miranda IM! Smaller, Faster, Easier.")
00:55:31tomallinuxstb_: Is there any significant difference?
00:55:35linuxstb_Also, which iFP models will your port support?
00:55:48 Join einhirn_ [0] (
00:55:58linuxstb_tomal: I don't know - I just chose the latest available versions, and they seem to work fine.
00:57:48 Nick BHSPitLappy2 is now known as BHSPitLappy (
00:57:57tomallinuxstb_: I own iFP 790. It will most probably support iFP 795 and iFP 799 if I know what exactly flash chips are inside them.
00:58:04 Quit einhirn (Read error: 110 (Connection timed out))
00:58:33tomallinuxstb_: iFP 8xx are probably the same as 7xx inside
00:58:49linuxstb_I've added a "#define IRIVER_IFP7XX_SERIES 1" to config-ifp7xx.h
00:58:49markunyes, they are
00:59:09linuxstb_But maybe ifp7 isn't the best prefix?
01:00:13tomallinuxstb_: I think it's ok. I can't think of anything better.
01:00:40linuxstb_We need a define like that - for example the CONFIG_CPU definition isn't defined in the simulators, so we'll need to change some #ifdefs to use IRIVER_IFP7XX_SERIES instead.
01:00:55linuxstb_OK, I'll keep it as IFP7XX
01:02:22 Quit Kohlrabi (Read error: 104 (Connection reset by peer))
01:02:27tomalAFAIK iFP780 unlike the others uses internal PNX0101 USB controller, so I may be unable to write USB driver for it in the future.
01:03:18Paul_The_NerdSo, it's more of iFP79x maybe?
01:06:22tomalThe only difference between 780 and 7xx is a different USB controller and less amount of flash, so it's rather 7xx
01:08:08tomalI own only iFP790. I only know about the other players what I read on the Internet.
01:11:50 Join thegeek [0] (
01:13:33 Quit webguest68 ("CGI:IRC (Ping timeout)")
01:13:39 Quit Janco ("CGI:IRC (Ping timeout)")
01:14:15 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
01:14:58 Quit adamdullenty ("CGI:IRC (Ping timeout)")
01:14:58 Quit ts-x ("CGI:IRC (Ping timeout)")
01:16:20 Quit petur ("here today, gone tomorrow")
01:17:14 Part tomal
01:22:37 Join linuxstb_ [0] (
01:28:12 Quit linuxstb_ ("CGI:IRC")
01:29:30Jungti1234Anyone has H300?
01:30:05 Quit einhirn_ ("Miranda IM! Smaller, Faster, Easier.")
01:32:19 Quit Moos (Read error: 104 (Connection reset by peer))
01:36:01linuxstbRockbox target #16 now in CVS thanks to Tomasz.
01:37:30linuxstbtomal: If you read this in the logs, I've just made a few minor cosmetic changes to your patch - adding the IRIVER_IFP7XX_SERIES define, and using CPU_ARM in more places instead of checking for either PP5020 or PNX0101.
01:38:32linuxstbI've also made metronome compile for the ifp by adding some (random) button definitions - metronome now works on all targets. You may want to check the button definitions and change them.
01:38:43lostlogicthis thing is getting big... one of these days it'll start needing code documentation and policies or something.
01:39:00linuxstbWe've got those already.
01:39:02 Quit muesli- ("ich will Khe!!!")
01:41:56 Quit fobbo ()
01:48:33Midgey34is a new forum going to be opened up for the ifp series now?
01:54:15***Saving seen data "./dancer.seen"
01:55:43 Quit Jungti1234 ()
01:57:38markunMidgey34: ask jeff
01:57:38 Quit thegeek (Read error: 104 (Connection reset by peer))
01:59:30Midgey34hm, anyone know what z80 core is used in rockboy?
02:00:00linuxstbrockboy is just a port of gnuboy
02:00:11linuxstbSo it's whatever gnuboy uses.
02:01:20 Join DJDD_ [0] (
02:01:35 Quit Mulzi-SAW (Read error: 104 (Connection reset by peer))
02:04:41 Join alphakiller [0] (n=bohas@
02:05:02alphakillerheheh, I still couldn't boot rockbox on ipod
02:05:03linuxstbGood morning.
02:05:05alphakillerjeez :P
02:05:14linuxstbI thought you succeeded last night?
02:05:26linuxstbOr was that just using the bootloader to load the Apple firmware?
02:05:35alphakillerI installed bootloader ... but I couldn't boot up the kernel
02:06:59 Quit akaidiot (Connection timed out)
02:07:32alphakillerI compiled the rockbox kernel
02:07:41alphakillerand extracted into my iPod
02:07:45alphakillerfailure :(
02:08:07alphakillerRockbox error -2
02:08:08linuxstbAh yes - you got the "-4" error?
02:08:33alphakillernow, -2
02:09:05linuxstbHow big is the rockbox.ipod file on your ipod?
02:09:35linuxstbAnd another question - what's wrong (in C) with: *(x++)=f(*x)
02:09:48linuxstbgcc tells me "warning: operation on `x' may be undefined"
02:12:41lostlogic*ponders* I think we need to get the mregparm option supported on our targets. would make some inlines unnecessary, and improve performance overall, methinks.
02:14:53alphakillerhmm strange
02:14:59alphakillerseems my files on my iPod are messed up
02:15:09linuxstbCould be a dodgy fat32 partition
02:15:18linuxstbHave you restored it recently?
02:15:24linuxstb(using the Apple restorer)
02:19:39alphakillerI'm formatting with the windows now :P
02:20:26XavierGryou cant with windows I think
02:20:27linuxstbPaprica: You probably already know this, but it's possible to get stuck in brickmania with the ball bouncing from one corner to the opposite corner. I don't know if that's avoidable.
02:20:40linuxstbIt's only a 30GB ipod.
02:20:52XavierGrexceot with a third party program. (partition magic or swiss knife)
02:21:00XavierGrah okay then
02:21:00alphakillerI'm doing it..
02:24:35 Join aliask [0] (
02:24:51alphakilleroh rockbox loaded :D
02:25:18linuxstbA painless process... :)
02:25:22alphakillernow check sum errror
02:25:33alphakillerwhen rolo loads ..
02:25:40alphakillerI'm on a menu with an X and rockbox aside
02:25:51linuxstbThat's not a menu, that's the file browser.
02:25:59alphakilleroh :P
02:26:02linuxstbBut your hard disk is now empty....
02:26:05alphakillerhow do I run apps :P
02:26:12alphakillercan I load apps from it ?
02:26:12linuxstbHold down menu until the main menu appears.
02:26:31linuxstbThere are plugins - choose "Browse Plugins" from the menu.
02:27:40alphakillerI do select
02:27:45alphakillerit hangs on this menu
02:28:01linuxstbYes, hanging isn't uncommon at the moment.
02:28:18linuxstbYou can try disabling the CPU cache - remove the call to ipod_init_cache() in firmware/system.c
02:28:38linuxstbThat obviously slows it down, but seems to greatly improve stability.
02:30:01alphakillerit doesn't really hangs ..
02:30:13alphakillerit just don't go to the Browse Plug in
02:30:24linuxstbWhat do you mean?
02:30:33Paul_The_NerdSometimes when you choose Browse Plugins you get an empty screen instead of a list
02:30:43Paul_The_NerdOr it just holds for a few seconds, then does nothing leaving you in the menus still
02:31:00alphakillerhmm, oh I just figured out that my select button isn't workin on this menu
02:31:37Paul_The_NerdHit >>|
02:32:04alphakillerohh :P
02:32:09linuxstbThat's just to confuse the tourists...
02:32:16Paul_The_NerdStill doesn't always work though
02:32:55alphakillerhmm nice :P
02:33:04Paul_The_NerdBut, Bejeweled is PRETTY
02:33:12alphakillernice OS :P
02:33:24alphakillerwhat do I really need to work on it ?
02:33:41XavierGrtime I guess :)
02:33:57alphakillerhow do i guit sokoban
02:34:10Paul_The_NerdYeah, I couldn't quit some of the plugins either.
02:34:11linuxstbMENU is generally mapped to "cancel"
02:34:19alphakillermenu + select
02:34:29linuxstbBut yes, the key mappings have sometimes been chosen too quickly.
02:35:44XavierGrwhat is left for an iriver realese? h1xx
02:36:18linuxstbBug fixing I guess.
02:36:38linuxstbThe remote needs more work as well.
02:36:44XavierGrand maybe optical recording
02:36:57linuxstbMaybe, but that's not a show-stopper.
02:37:10alphakillerlinuxstb: should I start to code the multi-core driver ?
02:37:14alphakilleron iPod ?
02:37:17alphakilleror hack, dunno ?
02:37:33alphakillerow, bejeweled delays to start ?
02:37:41XavierGris there any intesion to integrate the wav2wv and the mp3 encoder to the recording options?
02:37:51linuxstbXavierGr: I don't like the way it's possible to select options using the remote that are not supported on the remote - i.e. being forced to get my h140 from my pocket to quit a screen.
02:38:14XavierGrlinuxstb: yes that true
02:38:39alphakillerbejeweled doesn't start here :(
02:38:45alphakillerI get a blue screen
02:38:51linuxstbSometimes plugins will start the first time, but not the second.
02:38:58linuxstbIf you reboot, it may work.
02:40:03Paul_The_NerdYeah, try rebooting
02:40:12linuxstbXavierGr: Yes, I'm sure there's interest in recording codecs - but I would say they are not vital for a first official release.
02:40:15alphakillersame shit
02:40:35linuxstbHave you tried disabling the cache yet?
02:40:39alphakillersokoban always starts
02:41:07alphakillerI dunno if you may get interested
02:41:27alphakillerI have a Java OS bases
02:41:42alphakillerI may "boot up" from rockbox
02:41:45alphakilleras plug in
02:41:51alphakillerdunno if you may get interested
02:41:53linuxstbMy personal interest is in getting the ipod to play music - I don't think java will help that.
02:41:55 Quit Sparticus (Read error: 104 (Connection reset by peer))
02:42:00alphakilleron running java on iPod :D
02:42:41linuxstbBut everyone does what they want to do - there's no dictator in Rockbox.
02:43:58 Join linuxstb_ [0] (
02:44:04alphakillerhow do I boot on the Apple firmware ?
02:44:30linuxstb_alphakiller: Hold down MENU as the ipod is booting.
02:44:34alphakillerhmm :D
02:44:46linuxstb_You need to press MENU very early in the boot cycle and keep it pressed
02:45:17linuxstb_The bootloader will then say "Loading original firmware..." if it detects the keypress - which it doesn't always do.
02:45:34alphakillersuccess :P
02:45:55alphakillerbut do I have stuff like "printf" on your lib ?
02:46:08linuxstb_We have snprintf
02:46:14linuxstb_And then lcd_puts()
02:46:22alphakillerhmm ..
02:46:34XavierGror splash()
02:46:40alphakillerit would be neat if you had a wrapper that controls the lines
02:46:41linuxstb_If you look at the plugins in apps/plugins/ you will see what sort of functions are available.
02:46:51alphakillerlcd_puts() you need to update the "line"
02:47:08linuxstb_Yes - but that's just passing "line++"
02:47:19linuxstb_Or "line" if you don't want to update the line.
02:47:50alphakillerand how do I erase a line ?
02:48:48linuxstb_And there's no malloc() in Rockbox.
02:48:57lostlogicsay I had a m68k gcc that enables mregparm and needed a HXXX user to test out a rockbox built with it... any takers?
02:49:16Paul_The_NerdI'll test on H120
02:49:28lostlogicw00t! I'll post a zip shortly!
02:49:50XavierGrwhat's a mregparm?
02:49:53lostlogicwont' effect the bootloader so if worse comes to worse, you can get to USB mode and put a normal build back ;)
02:50:09lostlogicpasses X function arguments in registers instead of on the stack
02:50:35XavierGrand that will make things faster?
02:50:36alphakillerI'll just use rockbox to boot my OS ...
02:50:36Paul_The_NerdYup, I'll try anything as long as it doesn't involve flashing. :)
02:50:40linuxstb_Does Rockbox compile OK with that?
02:50:55lostlogiclinuxstb_: not sure yet, I'll let you know momentarily
02:51:45 Join Jungti1234 [0] (n=jungti12@
02:52:21linuxstb_alphakiller: As long as you release it under the GPL, we'll steal the best bits for Rockbox.
02:53:11Jungti1234Frame Skip addition
02:53:42XavierGrI think this is commited Jugti
02:53:56Midgey34how many more did you add? I went up to 6
02:54:00lostlogicbah, gcc tells me invalid option when I just compiled it with the option enabled for m68k!
02:54:03linuxstb_Hurray, the build table is still green after the ifp7xx commit.
02:54:47*Paul_The_Nerd cheers for green build tables.
02:54:49lostlogic</idiot> I had reversed the gcc patch before the last build.
02:55:42linuxstb_alphakiller: So what does your java-based OS do? Why would someone want to use it?
02:55:50Jungti1234XavierGr: It added more some.
02:56:26Jungti12340, 2, 3, 5, 7, 9, 11
02:58:53XavierGrJungti: Maybe if you could try to make it run faster, then it could get commited... ;)
02:58:59BHSPitMonkeyhey people
02:59:24alphakillerlinuxstb: it's for enthusiats
02:59:35Jungti1234XavierGr: :)
02:59:42alphakillerthe same way why one would change from the Apple's Firmware for Rockbox
02:59:48BHSPitMonkeyknow those little bars in forum signatures that say "<something> user"
02:59:57BHSPitMonkeyin case there wasn't one for rockbox, i just made one
02:59:59lostlogiclinuxstb_: large portions of rockbox do compile with it, first problem came in the debug_menu... I'll attribute away places that break for now.
03:00:40 Quit linuxstb_ ("CGI:IRC")
03:01:02BHSPitMonkeypoor linuxstb_, using a web client
03:01:51Paul_The_Nerdlostlogic: Are you just going to want me to see if it runs, or is there any specific tests you'd like me to try once I have it?
03:02:08linuxstbBHSPitLappy: I'm back on my normal PC nwo.
03:02:10lostlogicPaul_The_Nerd: first... does it run, then some kind of performance measure... whatever you can think of
03:02:27lostlogicbut first I have to get it to compile which will take some trickery around our inline asm.
03:02:30XavierGrPaul the nerd you can run the wav2wv benchmark twice and see
03:03:06alphakillerlinuxstb: It can run Java :P
03:03:10alphakillerand be very portable
03:03:15alphakillerdoes multi-threading, ipc hehehe
03:03:18alphakillerfor Java ..
03:03:28alphakillerit's just for hobbist :D
03:03:40linuxstbSounds very cool - but I've yet to be attracted to java.
03:04:41lostlogichmph, the compilation failure was due to a bad assembly constraint in the first place.
03:04:41alphakillerheheh :P
03:04:53alphakillerlinuxstb: I might put it as linux.bin :p
03:04:56alphakillerand load it hehehe
03:05:12 Join Membrillo [0] (n=sam_kill@
03:05:20linuxstbThe aim of Rockbox on the ipod is to be a complete replacement for the Apple firmware - with the aim of playing music better than the original firmware.
03:05:38alphakillerbut how ??
03:05:40lostlogiclinuxstb: system.h:set_irq_level the constraints should be =d and +d not =r and +r.
03:05:46alphakillerplay music is just play musinc ..
03:05:53BHSPitMonkeyand just as soon as you figure that out, iPL will be most grateful :)
03:06:11lostlogicthat was the only compile failure.
03:06:25linuxstbGapless playback for a start, plus support for a large variety of audio codecs, plus independence from itunes...
03:07:03linuxstbPlus the choice of either playing files directly from the filesystem or from a "tag database"
03:07:07BHSPitMonkeyi love that rockbox has audio channel options
03:07:32BHSPitMonkeywhen i suggested a mono mode to the iPL devs, they saw it as useless :P
03:07:48*BHSPitMonkey needs to learn to program for himself though... so he shouldn't be complaining
03:08:12linuxstblostlogic: I think that's part of the ifp commit...
03:08:13Membrillomore developped codecs, giving greater sound quality without the restrictions of lower level audio decoding?
03:08:31lostlogiclinuxstb: shouldn't be, it's in a CPU_COLDFIRE ifdef
03:08:43linuxstbHah, a different set_irq_level then :)
03:09:12alphakillerbut you guys are mostly OS hackers too :P
03:09:12lostlogicshould I submit a patch, or just harrass people since it's a 1liner?
03:09:28 Join YouCeyE [0] (
03:09:29lostlogicPaul_The_Nerd: sorry I'm taking so long, I built it for H3XX the first time, forgetting what you would be running it on.
03:09:35linuxstbalphakiller: We are, but our motivation is better audio playback on our DAPs.
03:09:49linuxstbWell, for most of us anyway I think.
03:10:38alphakillerwell, I liked that Plug In browser ..
03:10:44alphakilleryou can expand your toys on iPod
03:10:47Membrilloalphakiller: Rockbox over the original firmware just fills in the gaps. Gives us DAP enthusiasts what we want rather than what the big companies want to put in
03:11:04Membrillobloody trans-national corportaions :P
03:11:42 Join gromit` [0] (n=gromit`
03:12:43Jungti1234BHSPitMonkey: What's that?
03:12:44Paul_The_NerdCopying over now
03:12:58BHSPitMonkeyJungti1234: just one of those forum signature things
03:13:00XavierGrlost logic I am interested too. Any patch to test for myself. I have an H140
03:13:03BHSPitMonkeyfor rockbox users
03:13:07BHSPitMonkeyme = bored
03:13:15MembrilloWould someone mind changing the version string on the H300 bootloader? it still says version 3
03:13:19lostlogicXavierGr: same link... 120 and 140 are the same except for harddrive size
03:13:49Jungti1234BHSPitMonkey: ah :) good
03:13:50XavierGrAH I thought you said it was for the H3XX
03:14:03Paul_The_NerdMusic playback doesn't work
03:14:14Paul_The_NerdI04: IllInstr at 00000002
03:14:22BHSPitMonkeyjust bored, and i had just made some iPL ones
03:14:23XavierGrmake a wav conversion
03:14:28BHSPitMonkeyand thanks :)
03:14:33lostlogicbah, none of the plugins will work :( sorry guys
03:14:41Paul_The_NerdAh, and it requires a pin to reset!
03:14:52XavierGrthen how can we test it?
03:14:58lostlogicyou can't yet
03:15:01XavierGrno playback no plugins then?
03:15:13Membrillobeg your pardon, the bootloader does say version 5
03:15:15lostlogicI'll have to work on it more. I'm sure it has to do with how the codecs are created.
03:15:27Jungti1234What is different that?
03:15:38XavierGrdo you think that it can greatly imrpove speed times?
03:15:53lostlogicXavierGr: honestly not sure, it's a thought though
03:16:26 Nick bagawk_ is now known as bagawk (
03:16:39Paul_The_NerdI'll be happy to test it next time too
03:16:53Paul_The_NerdI keep several implements suitable for resetting on hand for just such occasions
03:17:12lostlogicthanks, I'll ping again after reading the assembly for the codec and plugin APIs to figure out why they break.
03:17:37XavierGrlostlogic: but the CPU resgister is more fast than the iram right?
03:17:52alphakillerforever :P
03:17:53Paul_The_NerdSure thing, I'll be here for hours yet, and back tomorrow. Heh
03:18:10alphakilleralways register is faster
03:18:36XavierGrI thought so
03:18:43linuxstbBut won't the functions themselves be slower if they have less registers available? I'm assuming that's the implication.
03:19:01XavierGrthat's a valid point
03:19:34alphakillerwhat about stack ?
03:20:03linuxstbwhat about it?
03:20:05alphakillerthat's why modern cpus aims to have lots of registers
03:20:17alphakillerwhen you fullfil regs, you use stack
03:21:09lostlogicthis does not make less registers available to functions −− there really is very little cost to it, because anything you pass in on the stack is most likely copied into a register first thing inthe function _any way_ so thi sjust skips that save and load that are wasted 90% of the time.
03:21:50lostlogiclooks like the loose coupling of the API functions means that any functions exported to the API are going to have to be attributed to regparm=0 though :(
03:22:26linuxstbBut it could be useful internally in the codecs.
03:22:43lostlogiclinuxstb: yeah, and useful internally in the kernel...
03:23:14alphakillermy opinion is that such things would be faster if written on asm
03:23:23lostlogicwrite the whole thing in asm, eh?
03:24:00XavierGrwell it "could" improve speed greatly!
03:24:16XavierGr[11:02] markun: yes, install cvs tools. Then: cvsdo add 'file'
03:24:27XavierGrinstall what cvs tools?
03:24:31lostlogicthe places where it would make a big difference in rockbox are already inline assemblified
03:24:44lostlogicor have had crazy people like me go over them to make sure that GCC is doing the Right Thing (TM)
03:24:55XavierGrI use the DevKit do i need something more than that?
03:25:25alphakillerwell at least on mips and x86, I won gcc :P
03:26:23lostlogicalphakiller: you beat GCC on IA32!? that's hard to believe, it's pretty hardcore there.
03:26:30alphakillerheheh yeah ..
03:26:36alphakillera hand made asm OS :P
03:26:50alphakillerand I even counted cpu cycles
03:26:59alphakillerehehe parallelizing all :D
03:27:16 Quit Paul_The_Nerd ("Leaving.")
03:29:59lostlogicbut the question is: why do the calls _to_ the API functions try to use register params, but the exported versions of said functions do _not_
03:34:05linuxstbtime for bed. g'night.
03:35:45XavierGris there a staight forward method to include a new file to cvs diff?
03:36:00ashridahXavierGr: 'cvsdo' if you've got it
03:36:01lostlogicthat's wha tthe cvs tools are for cvsdo etc.
03:36:15ashridahyou can use 'cvsdo add' and it won't attempt to contact the repository
03:36:24ashridahor something to that effect
03:36:37XavierGrwhere can I get the cvsdo?
03:37:05ashridahi would have expected cygwin's cvs package to already have it
03:38:22XavierGrdo you have it?
03:39:32 Quit darkless (Read error: 104 (Connection reset by peer))
03:39:40ashridahi don't have cygwin handy, unfortunately. i assume the command 'cvsdo' doesn't exist in your cygwin environment?
03:39:48 Quit Membrillo ()
03:39:52ashridahyou might be able to use the cygwin setup program to add cvstools
03:39:54 Join darkless [0] (
03:40:00XavierGrwell I use the devkit and it seems not
03:40:14alphakillerI installed everything on my own here
03:40:17alphakillerI use cygwin
03:40:28alphakilleryou will only need a linux livecd to install your bootloader
03:40:52ashridahalphakiller: what's that got to do with his cygwin environment not having cvstools? :)
03:41:19ashridahXavierGr: does the devkit include a copy of the setup.exe for cygwin?
03:42:35XavierGrlet me check
03:42:47alphakillerIf I were he, I would download and install cygwin with binutils/gcc binaries and sources and cvs + automake + autoconf and gnumake and file utilz
03:42:57ashridahalphakiller: why?
03:43:07ashridahjust adding cvs tools should do the trick, with luck.
03:43:31alphakillerhmm ... I prefer to install things manually :D
03:43:36alphakillerI know what to do :D
03:43:47alphakillerit's not tricky building a cross compiler
03:44:20ashridahyeah, but it'd be a waste of his time and bandwidth, since he already has a working cross compiler, just not a copy of cvsdo.
03:45:13XavierGrthat's right I maybe only need to find an exe file.
03:45:51ashridahwell, one tarball
03:45:58ashridahactually, gimme a sec.
03:46:27XavierGrthere isn't the setup.exe file that cygwin comes with.
03:46:31ashridahcan probably just download the tarball for the tools
03:47:29ashridahcan you run 'cvs -v' in your cygwin environment and tell me what version of cvs you're running?
03:47:44lostlogicagh, the memcpy and memset assembly routines are what's hurting. I can fix this, but it'll be a pain in my butt.
03:48:30XavierGrConcurrent Versions System (CVS) 1.11.6 (client/server)
03:48:32 Join akaidiot [0] (
03:48:52ashridahokay. i don't think it matters either way, so download
03:49:21ashridahi'll check, but i'm pretty sure you'll be able to untar that in the root of your cygwin tree, and it'll install what you need
03:49:22XavierGrthanks ashridah, I will try straight away.
03:49:49ashridahthat's the one
03:49:53ashridahcd /
03:50:04ashridahtar xjvf /path/to/cvsutils-blah.tar.bz2
03:50:16ashridahwill unpack it, hopefully the copy of 'cvsdo' will end up in the path in the right place.
03:50:31 Join DaLToN [0] (n=dsfasd@
03:53:07DaLToNare you died?
03:53:34DaLToNwhat it means "ang"
03:53:56Jungti1234 úų?
03:54:07DaLToNpls in english
03:54:19***Saving seen data "./dancer.seen"
03:54:21DaLToNi cant unnderstand them
03:54:39 Quit mikearthur (Remote closed the connection)
03:54:44Jungti1234What is meaning of 'Are you died?'?
03:55:50ashridahgotta say, he looks pretty lively to me
03:56:09DaLToNcos all of them a little silent
03:56:20Jungti1234ah :)
03:56:26ashridahyou just joined. you missed all the action
03:56:36Jungti1234I misunderstood meaning.
03:57:47DaLToNwhat is your nationality
03:59:14DaLToNi a turk
03:59:20DaLToNdo you know turkey
03:59:28Jungti1234yes :)
03:59:32BHSPitLappyyes, we all know what turkey is.
03:59:41ashridahmmm. turkey
03:59:44DaLToNhow to you?
03:59:54BHSPitLappyhow to you too.
04:00:06Jungti1234Tasty turkey
04:00:24XavierGrashridah: Thanks for the help but unfortunately it didn't worked out. I exctracted the tools my self because the tar commnad didn't worked.
04:00:48ashridahhmm. cygwin's tar may or may not have had bzip2 available or something
04:00:55Jungti1234They are our brothers.
04:00:58 Quit Jungti1234 ()
04:00:59XavierGrI will try to find the file in a *.exe format. cvs is in a *.exe format too.
04:01:28ashridahaah, wait, those should have been .exe's.
04:01:39ashridahi'll poke harder, hang on
04:01:55XavierGrI renamed them to exes but it crashed :)
04:01:58DaLToNnow in my country and all moslem country celebrate muslim festival of sacrifices
04:03:04 Part DaLToN
04:03:21XavierGrashridah: If you have a cygwin enviroment you can give me your exes...
04:03:46ashridahXavierGr: well, you could build the cvsutils yourself, i suppose
04:03:56ashridahXavierGr: that's the problem, i don't right now
04:03:59 Quit DreamTactix291 (Read error: 110 (Connection timed out))
04:04:02XavierGrah sorry you said that you don't have one right now
04:04:57XavierGrhmm I think I knwo how to do it.
04:06:15ashridahXavierGr: you do realise that the 'cvsdo' script is a perl script, not an executable, right? :)
04:06:39ashridahproblem is, i don't know if the devkit has perl :(
04:07:08XavierGrwell I tried just in case. But yeah even when I try to compile it fails
04:07:14XavierGrI think I can find it.
04:07:27ashridahwell, compiling it shouldn't be necessary, really, they're perl scripts after all.
04:08:16 Quit alphakiller ()
04:09:28ashridahextract the 'cvsdo' script from the original .tar.bz2, and make it executable, if it's going to work, it should run okay
04:09:34ashridahif perl isn't there tho, can't do much
04:09:36XavierGrashridah: I will open cygwin setup get the cvsutils tarball and skip the installation
04:10:00ashridahyeah, the cygwin setup.exe should detect the cygwin install already, and just install the necessary parts, with luck
04:10:39BHSPitLappythat's what it does all right.
04:11:31 Quit Rob2222 (Read error: 104 (Connection reset by peer))
04:13:36 Join Rob2222 [0] (
04:23:26DJDD_Heya, how goes the Rockbox H3xx firmware? Does it really have all the features of the H1xx firmware?
04:24:00 Quit XavierGr ()
04:26:52 Join Membrillo [0] (n=sam_kill@
04:27:37MembrilloDid anyone here read the post on MR about a possible instrument tuner? do you think its possible?
04:30:00Membrilloon any platforms with a mic or line in
04:30:34 Join mymomthelush [0] (
04:30:41 Part mymomthelush
04:31:05DJDD_i really doubt it
04:31:16 Quit Nibbler (Read error: 110 (Connection timed out))
04:31:40Membrillowhy so? too hard to analyse the signals pitch?
04:31:44Membrilloneed extra hardware?
04:31:55 Join Nibbler [0] (
04:31:58 Join XavierGr [0] (
04:33:03DJDD_I just don't think anyone has the will to right the code for it
04:33:26DJDD_could be wrong though. Don't take anything i say as fact :P
04:35:22Membrillolol yeah of course
04:36:16Membrillowell im sure a couple of people out there play instruments. I mean you could even line in guitars etc
04:36:35DJDD_Though, i can say with fact that the iRiver H1xx series would probably be the most capable hardcore. It has line in and out on it
04:36:53DJDD_hardcore=hardware :P
04:37:11Membrillolol yeah
04:37:14Membrillowell so does H300
04:37:55DJDD_Ah yes, but the H1xx has DIGITAL line in and out ;)
04:38:04DJDD_or optical rather
04:38:26Membrilloi reckon its doable
04:38:33Membrilloand useful
04:39:10DJDD_Well, if i were you i'd create a thread on it and try to raise some awareness. If its popular enough the devs might move it up the priority list
04:39:14Membrilloif it could be done, you could have like a monotone composer like on the old phones :P
04:39:31Membrillotheres already a thread
04:56:20 Join thegeek [0] (
05:03:36 Join Rob2222_ [0] (
05:06:03 Quit akaidiot (Connection timed out)
05:07:40 Join mymomthelush [0] (
05:07:51 Part mymomthelush
05:11:15 Quit Nibbler ("can't stop the signal!")
05:12:23 Join Nibbler [0] (
05:20:23lostlogicanyone who knows m68k asm around?
05:20:26 Quit Rob2222 (Read error: 110 (Connection timed out))
05:20:26lostlogicmovm.l (%sp),#0x4fc
05:20:32lostlogicwhadat do?
05:23:51Membrilloim not minding :P
05:23:56Membrilloi have no idea what your on about
05:24:08lostlogicwhat DAP you have?
05:24:32MembrilloH300 and iPod vid
05:24:36MembrilloH320 that is
05:24:42MembrilloiPod vid 30g that is
05:24:53Membrillo30 gig not 30 generation haha
05:24:56lostlogicwant to test something for me on the H3x0 in a lil' while (because I don't have one, but I'm working on something slightly flakey)?
05:25:02Membrillosure thing
05:25:24lostlogicoh no, I'm working on breaking the whole shebang by using a patched GCC and -mregparm=3
05:25:35lostlogicbut I've had to rehack the memcpy and memset functions first to prevent mass stupidity.
05:26:00Membrilloim not sure wether its wise to say this, but im confused :-)
05:26:39lostlogicwhy wouldn't it be wise!?!?!?
05:26:52lostlogicI'll transfer to you a new −− hopefully it will run faster
05:27:08Membrillowhats different about it?
05:27:13lostlogicif you can time or measure speed somehow with what you're running now and then do the same with what I send you that'd be good.
05:27:20lostlogicinstead of passing function arguments on the stack
05:27:25lostlogicit will pass up to 3 in registers
05:27:42Membrilloso in peasant language...
05:27:58Membrilloit does good stuff?
05:28:16Membrillowhatever. It cant brick my player so ill try it
05:28:20lostlogichopefully it will be faster
05:28:27lostlogicright, it can't, because I'm not touching the bootloader yet.
05:28:36Membrilloi got that far
05:28:42Membrilloill see how rockboy runs :)
05:28:48Membrillowill that work?
05:28:50Membrilloas a test
05:29:04lostlogicsure −− should have it ready in a few... a wav2wv test would also be good if you have that capability.
05:29:24Membrillodo I? im not sure
05:36:42lostlogicignore that
05:37:39 Quit jelle-k (Read error: 110 (Connection timed out))
05:38:20XavierGrlost logic I can test too if you need testers.
05:38:29lostlogicI'll break everyone! :):)
05:38:31XavierGrBut I would need a patch or an h100 build
05:38:45lostlogicXavierGr: can't patch unfortunately because I have to use the GCC I built here
05:39:21 Join DreamTactix291 [0] (
05:39:45Membrillohaha, whats this breaking your on about :P
05:39:53XavierGrah ok then. I could use a build if you have the time to compile.
05:40:06lostlogicyeah, I'll try, after I figure out why it's behaving badly in one case.
05:40:10XavierGrlostlogic: what DAP do you have?
05:40:15lostlogicnone atm
05:40:23lostlogicI broke my H340 this past weekend, and am awaiting it's replacement
05:40:44XavierGrargh my condolences!
05:40:55XavierGrdid iriver rma'ed it?
05:41:22lostlogicwas my own dumb fault for futzing with it's charge input voltage, no, I bought a used one on Ebay
05:42:04XavierGrwhat did you do exactly? put a different charger?
05:42:05MembrilloXavierGr: that sounded slightly like a yokel... "did iriver rma'ed it?"
05:42:22Membrillosomething Cletus off the simpsons would say
05:42:40Membrillohillbilly. I dunno whatever word
05:42:59MembrilloI dont know what words used in your neck of the woods
05:43:25XavierGrwhat I said is probably wrong because there is no need for an 'ed when It is a question
05:43:52Membrillohaha just forget it man
05:44:02Membrilloim not making sense and i think im confusing you
05:44:08lostlogicMembrillo: don't harrass the non-native speakers!
05:44:45XavierGrhehe, anyway lostlogic: how much did you paid for your new H340?
05:44:56XavierGrargh again I did the same mistake....
05:44:58lostlogic275USD (used)
05:45:12lostlogicXavierGr: don't worry about it, we can understand you.
05:45:32XavierGrI am thinking of upgrading but I am not sure...
05:46:04Membrillolostlogic: do you still want me to test something?
05:46:18lostlogicMembrillo: yeah, finally copying over the file for real
05:46:29lostlogicsame link as before, ready now.
05:47:14Membrillojust getting it now
05:47:54XavierGrMembrillo: to test you can use the wav2wv plugin twice and compare the timings.
05:48:07XavierGryou need a wav file for that.
05:48:17Membrillowhere can i get the wav2wv plugin
05:48:21lostlogiclet's hope it boots even
05:48:23Membrilloill just record something to wav
05:48:27 Join atudar [0] (
05:48:36lostlogicthanks guys!
05:48:51lostlogicXavierGr: wait, not done copying yet, sorry
05:48:53XavierGrMembrillo: hold down the navi button and open the wav with the wav2wv plugin.
05:49:08lostlogicready now
05:50:01Membrillojust recorded somthing to convert
05:50:53Membrillodoes it give me a time at the end?
05:51:05Membrilloencoding time or something?
05:51:15lostlogicI think you need to time it, but I'm not sure.
05:51:16atudarHiya - has anyone had problems booting into the default iriver firmware after installing rockbox on an h320?
05:51:16XavierGrI am not sure. I will try right now
05:51:21lostlogicneve rused the plugin myself
05:51:33Membrilloatudar: nope its very safe
05:51:47lostlogicare we still at a zero brick ratio?
05:51:55Membrilloit times
05:51:57Membrillofor you
05:52:04Membrillo6 secs for first test
05:52:20atudarUnderstandable - but I've been unable to get it to boot into the iriver firmware without the thing being plugged in to charge
05:52:40lostlogicatudar: using which bootloader versoin?
05:52:42atudarbut it's not much of a problem because my only reason to boot into the iriver firmware is to watch videos right now
05:52:45Membrilloreflash, i think your using an old verstion of the bootloader
05:52:46XavierGrhmm 22 seconds internal timer and 25.26 my timer.
05:53:01XavierGrfirst test todays build now lets try lostlogics.
05:53:05atudarIt's version 3
05:53:12lostlogicatudar: what normally happens when you try to get into the iRiver firmware?
05:53:15atudarread the weird things going on with 4
05:53:27lostlogic5 is OK but doesn't change the issue you are having
05:53:29 Join Lost-ash [0] (
05:54:08atudarIt will come up with the usual rockbox bootloader and say that it's going to boot into the original firmware and then it turns off
05:54:10Membrillolostlogic. Both took 6 seconds
05:54:14atudarbut rockbox boots up fine
05:54:18Membrilloill try a longer recording maybe
05:54:21lostlogicatudar: you just need to keep holding play
05:54:22***Saving seen data "./dancer.seen"
05:54:23lostlogichold record
05:54:30lostlogicpress and hold play until you see the iRiver firmware booting
05:54:44atudarI love you
05:54:46atudarcan I have your babies
05:54:58lostlogicMembrillo: it might not be a noticeable difference −− it was just a theory of mine which I decided to spend time trying out ;)
05:55:06lostlogicatudar: uhh ... ... I'll ask my gf :-P
05:55:25XavierGrnope sorry lostlogic but my unit can't even boot.
05:55:31atudargood idea - thanks though
05:55:41lostlogicXavierGr: eek :(
05:55:49Membrillomine boots but the encoding time is the same
05:56:03atudarnow is anyone else working on any nifty wps for the h300?
05:56:18Membrilloyou can make your own? its very easy
05:56:20Membrillobut yes
05:56:24lostlogicMembrillo: no prob, thanks for testing
05:56:24Membrilloa lot of people are
05:56:33lostlogicXavierGr: lemme make anothe rone for you
05:56:46lostlogicXavierGr: I didn't make clean before building yours, might have botched it.
05:56:50XavierGrMembrillo: Did you tried on a bigger wave file? 6 seconds is a very short wav to test.
05:57:06Membrillono, 6 seconds was the encoding time
05:57:11XavierGrMake on with a minute and time it yourself don't belive the internal timer.
05:57:12Membrillobut i will try a longer one
05:57:16atudarI'm trying to make 2 pictures switch in one place like an animated gif but I can't seem to get it working
05:58:31XavierGrlostlogic: only the rockbox.iriver file is neede right? (though I tried both with the full folder)
05:58:47lostlogicneed everything to test any plugins
05:58:51lostlogicbut to boot jsut the rockbox.iriver
05:58:57Membrilloim trying a 5minute wav now
05:59:11XavierGrMembrillo: nice
06:00:45Membrillojust cutting pasting a few things to desktop
06:00:53Membrillocant fit two 5min wavs :P out of space
06:01:08XavierGryou have yours full?
06:01:24Membrillomy 20gb is full
06:01:30atudarwhy would you leave empty space on it?
06:01:39XavierGrI feel lucky that I have a 60GB one
06:01:40Membrilloits so depressing having to delete music off a full DAP
06:01:48Membrilloim going to buy a new HDD someday
06:01:53 Quit RotAtoR ("zzzzzz")
06:01:55atudar!! 60gig...
06:01:59atudaryou modded it?
06:01:59XavierGratudar for file managing
06:02:26Membrillohow much did that put you back?
06:02:38atudarI need about 60 gigs to contain my music in one place
06:03:48XavierGrsorry lostlogic it will not boot with this either.
06:04:26MembrilloXavierGr: how much did a 60gb cost?
06:04:39lostlogicXavierGr: that is sad. I guess it's not a big enough difference from the IRAM stack.
06:04:57lostlogicMembrillo: XavierGr: thanks very much for being guinea pigs for my crazy ideas, too bad it didn't pan out.
06:05:04XavierGruhm I dont clearly remember maybe 170euros I am not sure
06:05:10Membrilloim testing now
06:05:14Membrillo5min wavs
06:05:29XavierGrlostlogic: wait for Membrillo outcomes
06:05:30Membrillothe internal timer seems to be accurate
06:05:38XavierGrmaybe we can get some staff out of it.
06:05:52Membrillofar out!!!! how do i light the lcd without aborting the operation
06:06:06XavierGrpress up?
06:06:13Membrilloup aborts it half the time
06:06:31atudaralright - I'm back to work on my wps, thanks for the quick help
06:06:44XavierGrthen swithch hold?
06:06:47 Quit atudar ("CGI:IRC")
06:06:57XavierGror have a flashligth with you?
06:06:58Membrillothats an idea
06:07:15lostlogicXavierGr: flashlight no help on H3x0
06:07:19XavierGror make the lcd light all the time from the settings.
06:07:29Membrillocan you do that?
06:07:44XavierGrit is better to have the screen up the same time for both tests to do that?
06:07:54Membrillowhat setting is it
06:08:08 Quit ashridah (Read error: 110 (Connection timed out))
06:08:10Membrillofound it
06:08:20XavierGrI don't know about H300 settings. But it should be there.
06:09:50Membrillotest is running now
06:09:54XavierGroh you can test it with the new mp3_encoder plugin too.
06:10:26Membrilloowchie. Got E2cs this morning and my ears havent adjusted yet
06:10:39Membrilloi think my ear lobes are getting stretche
06:11:12lostlogicMembrillo: which tips are you using?
06:11:22Membrilloits not the tips
06:11:26Membrilloits the back bit
06:11:32Membrilloi think i have them adjusted badly
06:11:34XavierGrah I use e2 too. Great phones
06:11:46Membrillogot them off ebay. arrived this morning
06:11:54XavierGrI use the plastic flanges NOT the foams
06:11:55Membrilloim using the small black plastic ones
06:12:31lostlogicMembrillo: you are wearing them around over the tops of your ears right?
06:13:20Membrilloi think i have left in right though
06:13:25Membrillowhich could be causing problems
06:13:32lostlogicthey must be shpaed different than E3, because I dont' feel anything on my ear lobes
06:13:42Membrillonot my lobes
06:13:49Membrillothe inner bit of my ear
06:13:52Membrillobadly worded haha
06:13:58Membrillonot the inner bit
06:14:04Membrillobut the... gar... i dunno
06:14:06lostlogicahh, gotcha yeah, I know what you mean
06:14:22Membrillothe big hole lol
06:14:44Membrillojust need to get used to them
06:14:49Membrilloive never used earphones before
06:14:53Membrilloalways used chunkys
06:15:23lostlogicearbuds are actually worse on that area than inears, in my experienc, so you chose the right ones.
06:15:40XavierGrMembrillo: what about the tests?
06:15:52Membrilloalmost finished
06:16:21XavierGroriginal or new?
06:16:51Membrillojust starting new now
06:17:40Membrillowas gta or gta2 released for gameboy?
06:17:45 Quit Nibbler (Read error: 110 (Connection timed out))
06:17:50 Join Nibbler [0] (
06:18:04XavierGrhow much did it took for original?
06:19:17Membrillo179 seconds
06:20:51XavierGr5 minutes into 3 that is fast.
06:21:50Membrillolittle technical difficulty
06:21:52Membrillojust a sec
06:22:33Membrillowait its ok
06:22:35Membrillomy bad
06:22:37Membrillogoing now
06:22:42Membrillolooks like your build is kicking but
06:22:50lostlogicthat would be happy making
06:22:57lostlogicafter this test, sleep
06:23:01lostlogicwill play more after work tomorrow.
06:23:08XavierGrhow much?
06:23:36XavierGrMembrillo: It seems faster?
06:24:33Membrilloit looks like it will be dead even
06:24:48lostlogichehe, thus is life
06:24:48Membrillo20% in 34 seconds so far
06:25:02Membrillo30% 52s
06:25:19Membrillo40% 70s
06:25:37Membrillo50% 88s
06:25:41Membrillowill be slightly quicker
06:25:56XavierGrha it got 1.5 second ahaed
06:26:22Membrillo75% 133s
06:26:23XavierGrI mean 4s
06:27:03XavierGrand final?
06:27:21XavierGrthat very good!
06:27:24lostlogicso .5% fasterish
06:27:28XavierGroops now
06:27:29Membrilloi had the charger in for the second one. wil that make a difference?
06:27:33Membrilloi was almost flat
06:27:34XavierGrs/ now
06:27:44XavierGryeah no good
06:27:45Membrillodidnt think so
06:28:00 Part Midgey34
06:28:18XavierGronly 1 second? this is dissappointing.
06:28:21Membrillohmmm do you know if Gta 1 or 2 was released for gameboy?
06:28:33lostlogicsorry guys, my idea didn't pan out. thanks for testing
06:28:37XavierGrthe original gta?
06:28:41lostlogicI'm sure I'll have more bogus and/or good ideas in the future
06:29:00XavierGrnight keep up
06:29:01Membrillobadluck man
06:29:14MembrilloXavierGr: original or gta2
06:29:26XavierGrMembrillo: You mean Grand Theft Auto?
06:29:31lostlogicMembrillo: it's the way it goes in trying to optimize such things, I'm used to it, just not used to having to have other people test ;)
06:29:43XavierGrI don't think that they port games like the original
06:30:06XavierGrmaybe a modified version that is called gta
06:30:11XavierGrbut I don't know
06:30:22Membrillofair enough lostlogic. Keep up the good work though
06:30:32MembrilloGta = Grand theft auto
06:30:33XavierGris rockboy realtime with latest additions, I think not, right?
06:31:16 Join Paul_The_Nerd [0] (
06:31:24lostlogic(talking in my sleep: see if rockboy is any different on my build)
06:32:13XavierGrI was very disappointed with the outcome of the test. I was like: "Oh my god this is gonna speed up my player 20%!!" :(
06:32:30XavierGrBut thanks for trying lostlogic!
06:32:32Paul_The_NerdIs there a new lostlogic build?
06:32:35lostlogicXavierGr: win some, lose some, and rarely does one win as big as 20% in optimizations.
06:32:57XavierGrI know I was just dreaming/
06:33:34XavierGrdid you commited your ogg vorbis optimizations in the end?
06:33:41Paul_The_NerdSorry I vanished, people showed up with birthday wishes.
06:33:56XavierGryou have your birthday today?
06:34:13Paul_The_NerdTomorrow really
06:34:17Paul_The_NerdWhich, here, is 30 minutes away.
06:34:19lostlogicXavierGr: yeah, all of the worthwhile stuff in my vorbis optimizations got put in by Lear... he's working on some mor stuff there, I think.
06:34:23XavierGrHappy Birthday then!! :)
06:35:01lostlogichappy birthday Paul
06:35:12XavierGrPaul_The_Nerd: Where is the bouffee? (table with foods) ;) lol
06:35:31Paul_The_NerdI am the ripe old age of 23. Hehehe
06:35:31lostlogicif you also want to be disappointed in the lack of better performance, you can check out my build for H1x0 at
06:35:45lostlogicPaul_The_Nerd: welcome to the 23 club :)
06:35:49Paul_The_NerdIs it worse, or just equal?
06:35:51XavierGris it now fixed?
06:36:11lostlogicXavierGr: did you retry the second time?
06:36:14XavierGrI am 22. I will join next year.
06:36:25lostlogicPaul_The_Nerd: equal as far as anyone can tell
06:36:33 Quit kkurbjun ("Leaving")
06:36:46lostlogicoh, shit, nevermind, apparently it works on H3x0 and not on H1x0
06:36:48XavierGrdid you upload a new one?
06:37:10lostlogicXavierGr: I uploaded a second copy of that link right there after doing a clean build of it
06:37:18XavierGrI will try once more maybe it was the cache of my browser
06:38:08lostlogicsigh, I'm surprised it didn't improve more... we must already have all the functions that are called enough to matter inlined.
06:38:41XavierGrnah it is the same. But anyway 1 song per 179 seconds is too little, but maybe in other tests it will get better.
06:38:52Paul_The_NerdI just realized, last time I tested it used the wrong rockbox.iriver
06:39:01Paul_The_NerdI forgot to place the new one in the .rockbox folder to overwrite the old
06:39:24lostlogicugh, apparently it's broken on H1x0, and I'm too tired to fight with it tonight, so I'm done
06:39:40lostlogicmaybe tomorrow I'll have my own player again to test my code on ;)
06:39:44lostlogicg'night for real :)
06:39:47XavierGrMembrillo: I hope you did a good test, Imagine if by mistake you did the same test twice :shock
06:40:54 Quit Membrillo ()
06:41:08Paul_The_NerdG'night lostlogic
06:41:20Paul_The_NerdAnd, nothing even shows on my LCD with it. It's very VERY not working.
06:41:48XavierGrsame with me
06:47:34 Quit Lost-ash (Read error: 110 (Connection timed out))
06:48:32 Join akaidiot [0] (
06:52:17 Part Paul_The_Nerd
06:59:38 Quit thegeek (Read error: 113 (No route to host))
07:23:44 Join menollo [0] (
07:27:59 Quit menollo (Client Quit)
07:28:07 Join menollo [0] (
07:33:48 Quit pabs (Read error: 110 (Connection timed out))
07:35:49 Quit menollo ("CGI:IRC (EOF)")
07:36:55 Quit saa[b_r]ider (Read error: 110 (Connection timed out))
07:39:14 Join CrookedRain [0] (
07:41:52 Join B4gder [0] (
07:42:29 Join jelle-k [0] (
07:42:36B4gdermorrn morrn!
07:45:22 Join pabs [0] (
07:54:10 Join gtkspert [0] (
07:54:23***Saving seen data "./dancer.seen"
08:00:17 Join ender` [0] (i=ychat@
08:05:48 Quit akaidiot (Read error: 110 (Connection timed out))
08:09:42 Join gtkspert_ [0] (
08:11:44amiconnGood mornin'
08:15:07 Quit darkless (Client Quit)
08:19:12 Quit Coke (Nick collision from services.)
08:19:31 Join Coke [0] (
08:19:51 Quit Coke (Nick collision from services.)
08:20:07 Join hiro^ [0] (
08:22:56 Quit gtkspert (Read error: 110 (Connection timed out))
08:25:54 Join ashridah [0] (
08:38:14Bger mornin!
08:38:19BgerXavierGr :)
08:53:33 Join _FireFly_ [0] (
08:57:28 Join Kohlrabi [0] (
09:02:21 Quit uwe_ ("Leaving")
09:04:46 Join TravisH [0] (
09:06:46 Quit DreamTactix291 (Read error: 110 (Connection timed out))
09:14:52 Join saa[b_r]ider [0] (n=saab_rid@
09:18:13XavierGrOh Morning Bger!
09:18:25 Join DangerousDan [0] (
09:18:35XavierGrMorning amiconn and Bagder
09:27:00XavierGrOh Bger one of these days I will start to make the changes to battery_bench that we were talking about bfore Christmas.
09:28:34 Join Membrillo [0] (n=sam_kill@
09:29:41B4gderhey Membrillo, did you notice the bleeding time stamp thing works now?
09:29:55preglowlostlogic: i don't think m68k-gcc supports register parameter passing unless you patch it first
09:30:01Membrillosay what?
09:30:51Membrillocool. So is that the time of the last bleeding build?
09:31:46Membrillohaha, is it pushing to display the CVS description too? haha, if you cant be bothered thats cool
09:31:55 Quit Kohlrabi ("Leaving")
09:32:07Membrillojust means ill be able to display a changelog as well
09:32:12BgerXavierGr oki :)
09:32:27B4gderMembrillo: you can get that by using the timestamp
09:32:38B4gderfor example:
09:32:57Membrilloof course
09:32:59Membrillothank you
09:33:15B4gdermy pleasure
09:33:36Membrilloso i will be able to display the last change now, terrific
09:33:43 Join LinusN [0] (
09:33:55Membrilloso does that page display the last change, or the last couple of changes
09:34:03Bgermorning , LinusN
09:34:26B4gderMembrillo: the CVS html contains what seems to be the changes that went into that bleeding build
09:34:31B4gderit could be one, could be many
09:34:58XavierGrMembrillo: Wacky proposal, what if it can see which version the user has installed and then copmare with the latest. Then log all the changes from this build.
09:35:20amiconnBagder: Shouldn't that read ?
09:35:32XavierGrMembrillo: You are talking about your program that downloads and installs rockbox right?
09:35:55MembrilloXavierGr: yeah... im just thinking it over
09:36:01B4gderamiconn: yes, you're right
09:36:08Membrilloim wondering how you would make a comparison
09:36:32XavierGrnow that I think of it only if you make a file on installation.
09:37:05XavierGrthe binary doesn't hold any sort of that info.
09:41:41MembrilloXavierGr: what..?
09:43:19XavierGrIf you want to output the whole progress between 2 installations you could create a small hidden file on target that has the date of the latest build. Then upon update your program will check the 2 dates, and output all commited changes.
09:44:57Membrilloyeah, thats a good idea. I could do that. but the problem I have is getting the commited changes. I can get the latest change thanks to b4dger but I'm still working out how to get more than that
09:45:28XavierGrhmm yes that could be a problem.
09:45:36MembrilloB4gder: you couldnt make your script, add the timestamp to the page rather than delete the old timestamp and replace it?
09:48:25 Join akaidiot [0] (
09:50:07B4gderand when would it stop?
09:50:24B4gderI suggest you solve that problem in another way
09:50:43Membrillotrue lol
09:51:03Membrillogar... im not a very skilled programmer lol. This is getting over my head
09:52:31Membrillowell ill put in that you can see the last change, then i can think about more than tht
09:53:11B4gderthe CVS log is so crude anyway so most users won't appreciate it
09:53:45Membrilloyou think? i mean i look at it, but you dont think many people do?
09:53:50XavierGrWell it is the only source of monitoring progress for me.
09:54:27***Saving seen data "./dancer.seen"
09:55:43B4gderI bet many people look at it, I believe fewer people actually can make any sense from it
09:56:40Membrillohaha well ill use it. It will be a good experience for me as a learning coder
09:59:44Membrillohmmm if i download the page as html. then get rid of all the html up to the first <table> tag. Then add a <html> tag at the start. Will that just be the CVS changes table?
10:00:36B4gdersounds error-prone
10:01:18Membrillobecause I can do that easily enough
10:02:01B4gderI can offer a direct URL to the latest-cvs only
10:02:45Membrilloyep thats cool
10:02:49Membrilloi think ill just use that
10:03:14Membrillowow, thats perfect!
10:03:21Membrillothats exactly what i needed
10:03:23B4gderin spite of its name, that is the last N entries as shown on the front page
10:04:06Membrilloand thats automatically updated?
10:04:26Membrillothat makes it very easy on my part
10:12:53 Join amiconn_ [0] (
10:26:42 Join Zagor [0] (
10:31:05 Quit amiconn (Read error: 110 (Connection timed out))
10:31:05 Nick amiconn_ is now known as amiconn (
10:37:34 Join Jungti1234 [0] (n=jungti12@
10:39:01 Join DreamTactix291 [0] (
10:42:48 Join MrShlee [0] (
10:44:01Membrillohowdy. thanks B4gder btw. This is actually working
10:44:06Membrillonot coming along badly
10:44:07MrShleeHey the H3XX bootloader V5 doesn't seem to support the new H3 firmware :|
10:44:49LinusNMrShlee: ?
10:45:48Membrilloit works fine
10:49:19 Quit gtkspert_ ()
10:51:23MrShleeMembrillo: test it yourself, goto download the eu 1.29 firmware (labled 1.30). Try and patch it using the fwpatcher loader.
10:51:28MrShleeSecond opinion
10:51:47_FireFly_MrShlee fwpatcher isn't the bootloader ;)
10:52:11MrShleeI know technically it isnt.. but the patcher doesn't like the checksum
10:52:23_FireFly_why do you blame the bootloader ??
10:53:29MrShleeI'm not blaming anyone. ha just informing the general collective about the changed crc hash
10:53:47B4gderthat's a fwpatcher thing, not a bootloader flaw
10:54:19MrShleeand I shall now kill myself for having such a grave mistake :|
10:54:22_FireFly_but you had said that the bootloader doesn't support it ;)
10:54:25B4gderso what is the correct md5 we should use?
10:54:33LinusNhmm, do you think i can just update the md5 sum and still call it 1.29?
10:54:34MrShleeI explained it to LinusN the currect way
10:54:47LinusNin fwpatcher i mean
10:54:58MrShleethey released 1.29 twice.. so it uses two checksums
10:55:00LinusNnever mind, i'll add 1.30 too
10:55:06MrShleeIm currently using the original first release 1.29
10:55:13MrShleeand that patched fine.
10:55:45MrShleemust have been something wrong with it. for them to pull it and re-release as 1.30
10:55:49MrShleebut nothing is offically documented.
10:55:55LinusNas usual
10:57:33_FireFly_MrShlee i can*t find an 1.30 fw on only the 1.29
10:57:50 Join linuxstb_ [0] (
10:57:57 Join thegeek [0] (
10:58:26MrShlee_FireFly_ look at the 27/12/05 release. it IS labled 1.29 on the website
10:58:29LinusN_FireFly_: it's called 1.29on the web page , but the file says 1.30
10:58:39MrShleebut the file links to 1.30
10:59:14_FireFly_there is no fw published on 27.12.05 only 1.12.05
10:59:24*MrShlee looks
10:59:39 Quit linuxstb (Read error: 113 (No route to host))
11:00:16MrShleenumber 12 | H300 Firmware V1.29 - EU | Dec 27, 2005
11:00:29MrShleelinks to H300_V130(eu).zip (2.1 MBytes)
11:01:07_FireFly_then they didn't published it yet on the sites ;)
11:01:13 Quit DreamTactix291 (Read error: 110 (Connection timed out))
11:01:30MrShleeI know you guys must deal with alot of trolls.. but you shouldn't get so flamey so fast
11:01:57B4gderwe're not flamey, we're just picky on the details
11:01:57MrShleeI am here to help you fools.. shouldn't be so fast to call me a lier.
11:02:03MrShleeohh all good
11:02:31LinusNMrShlee: fwpatcher updated
11:02:36MrShleedetails have never been my forte
11:02:40MrShleeThanks LinusN
11:03:30LinusNsneaky, those iriver guys
11:04:31Jungti1234What are different 1.29EU and 1.30EU?
11:04:42LinusNJungti1234: nobody knows except iriver
11:04:51Jungti1234ah.. :) ok..
11:05:10B4gderyou _sure_ they know? :-P
11:05:25Jungti12341.30EU doesn't support video.
11:05:47Jungti1234But, 1.29EU supports video.
11:05:56MrShleeo_0 you sure
11:06:02Jungti1234I consider it strangely.
11:06:04MrShleethats a pretty major fix
11:06:05B4gderso then perhaps iriver doesn't know either ;-)
11:06:08 Join Polo_o [0] (
11:07:41Jungti1234I want this really. -
11:07:43 Quit akaidiot (Read error: 110 (Connection timed out))
11:09:10Jungti1234It should like to be possible at any cost.
11:10:42MrShleeJungti1234.. you weren't joking they removed video support.. crafty bastards!
11:11:13Jungti1234They are bad.
11:11:56Jungti1234Iriver has big problem.
11:14:13LinusNJungti1234: - wow! someone has been busy
11:14:41LinusNbut as usual, it's always easy to draw a gui example, but someone has to code it...
11:14:45B4gderLinusN: and what candy it is
11:14:54B4gdertripple amen to that
11:15:06_FireFly_yeah and i think to support this in rb many changes might be needed
11:15:07Jungti1234um.. I can't do it. :(
11:15:07B4gderand consider font sizes, lcd size, color depths
11:15:16LinusNand languages
11:15:45LinusNdamn good looking though
11:15:46Jungti1234it can change much.
11:21:20MrShleerockbox feels alot faster with new firmware/new bleeding edge - Might be a placebo but good work :D
11:21:41B4gderplacebo is good for you!
11:26:01 Quit MrShlee (Read error: 104 (Connection reset by peer))
11:29:04Jungti1234Can I do C really?
11:30:17LinusNdidn't tomek malesinski create a simulator for ifp790?
11:30:34LinusNinstruction sim that it
11:30:52preglowhe's created an entire simulator for it
11:31:15preglowthis alone proves he's quite mad
11:31:20preglowbut i've always liked mad people
11:31:28LinusNi hope he's willing to share it with us in the cvs
11:31:30B4gderrockbox material indeed
11:32:17linuxstb_LinusN: He mentioned he probably would last night. It's in lisp though...
11:32:35LinusNhe *is* mad
11:32:41aliaskHaha. Oldschool.
11:32:51preglowlisp :////
11:32:53linuxstb_Sorry, he has a code analyzer/disassembler - that's in lisp.
11:33:07linuxstb_I'm sure the emulator is in C
11:33:12preglowit is
11:33:14preglowi saw the source code
11:33:34LinusNi saw a film about that, i believe it's called Dead Language Society ;-)
11:34:46 Join gtkspert [0] (
11:35:57 Join B4gd3r [0] (
11:36:25 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
11:36:42Jungti1234What is 'B4gd3r'?
11:37:10 Quit DJDD_ (Read error: 110 (Connection timed out))
11:37:13B4gd3ra bot!
11:38:10 Quit B4gder (Nick collision from services.)
11:38:17 Nick B4gd3r is now known as B4gder (
11:41:30MembrilloLinusN: you werent possibly trying to make a joke about Dead Poets Society were you.
11:41:54LinusNMembrillo: took you a while to get it, didn't it? :-)
11:42:02 Join NicoFR [0] (
11:44:02Membrilloyep haha
11:44:06Jungti1234Rockboy is interesting...
11:44:12Membrilloim like... what the hell was he making a crack at
11:45:50ashridahi'm actually half surprised there isn't a dead language society.
11:45:58Membrillothere is
11:46:01Membrilloits called America
11:46:10ashridahprogramming languages, i mean.
11:46:15ashridahbut yes
11:46:40ashridahamerica takes the prize as being the only society where the language was shot by its native speakers :)
11:49:14aliaskDoes anyone have eli's bmp viewer's source? I lost mine in a format, and is down.
11:49:51Jungti1234Don't connect.
11:50:03aliaskDo you have the source code Jungti?
11:50:12NicoFRi have it too
11:50:13Jungti1234um wait
11:50:37Jungti1234He has it.
11:51:41 Quit `3nergy ("changing servers")
11:51:58 Join `3nergy [0] (
11:52:08aliaskCheers Nico.
11:54:23 Join TCK [0] (
11:54:29***Saving seen data "./dancer.seen"
11:55:15aliaskAh wait, I was actually looking for the one which displays a BMP file. It was in the extras folder, called plugbmp.c
11:55:27 Join DrMoos [0] (
11:56:02Jungti1234I was playing xobox, suddenly power off.
11:56:33 Quit TravisH ()
11:56:51NicoFRaliask: are you trying to modify/optimise it ?
11:57:21aliaskI just want to play around with the idea of using it to display bmp's for a user interface.
11:57:43aliaskThose pics of Jungti's made me want to do it. :)
11:58:03NicoFRwhat pics ?
11:58:13NicoFRlink ?
11:58:35LinusNis anybody up to the task of working seriously with the gui appearance?
11:58:51aliaskDon't look at me, I'm doing it to learn C.
11:58:52Membrilloyou would get great praise from thousands of people!
11:58:54LinusNnot just drawing pictures, but doing the actual work
11:59:36LinusNthis person will get a lot of heat from old rockbox users and devs
11:59:49Membrilloit wont be perfect?
11:59:54aliask"Wouldn't it be good if..."
12:00:11aliask"No I like it better this way..."
12:00:26NicoFRJungti: this is impressive
12:00:42LinusNbecause many oldtimers, especially those with old archos units, don't like changes, and particularly not changes that waste screen estate
12:00:50NicoFRJungti: did you do it ?
12:00:53Membrillostuff them
12:01:01Membrillorockbox needs sexing up
12:01:11Jungti1234NicoFR: Any Korean
12:01:20LinusNi am one of those oldtimers, but i'm not *that* anal about it
12:01:58MembrilloLinusN: what daps do you have?
12:02:03Membrilloarchos' i assume?
12:02:08LinusNalso, this person needs to have an understanding about the needs of the blind community
12:02:20aliaskWell, wouldn't it be only for targets with larger screens?
12:03:01 Quit `3nergy ("changing servers")
12:03:23 Join `3nergy [0] (
12:03:32LinusNi have archos player, rec v1, rec fm, iriver h110, h120, h140, h320, h340 and iaudio x5 :-)
12:03:43aliaskHoly cow.
12:03:54Bgerreally holy cow ... :)
12:03:59LinusNaliask: it would be a pain to have different ui's for different platforms
12:04:06Jungti1234LinusN: You are rich person?
12:04:08Membrillowhat do most of them do....?
12:04:09NicoFRLinusN: WOW :o
12:04:12Membrillosit around your house?
12:04:21Membrilloor can you not afford a house now
12:04:23LinusNno, i'm not rich, but we have a rockbox donation fund
12:04:31Membrilloyou have so many DAPS you live in a box
12:04:32aliaskWell, I was thinking more along the lines of just having some #ifdef's. Is that practical?
12:04:36LinusNi need those for developing and teting
12:04:42Membrilloah fair enough
12:05:04LinusN(and i didn't say how many of each i have :-)
12:05:11Membrilloso why do you have h320 and h340 etc?
12:05:15NicoFRor both your H3xx US versions ?
12:05:21BgerLinusN u should take a picture of all them (one picture with all these players on a table for example)
12:05:22NicoFR* or -> are
12:05:44aliaskAll displaying the rockbox logo (if possible).
12:05:49LinusNi have 1 320 with bdm attached, 1 int 320 and 1 us 340
12:06:16Jungti1234I have h120, h320
12:06:17aliaskDoes the BDM require the player to be naked?
12:06:20LinusNthe us 340 is for upcoming drm experiments
12:06:27LinusNaliask: yes
12:06:40aliaskYou must be good friend's with your local DAP store.
12:06:53LinusNi'm a regular ebay citizen :-)
12:07:13aliaskAh, I'm sure that eases the funds a bit.
12:07:30LinusNthe us340 is not mine, i am borrowing it from a dedicated rockbox fan
12:09:04Jungti1234Didn't you feel H300 LCD's problem?
12:09:18LinusNwhich one?
12:09:37LinusNboth problems?
12:09:44LinusNwhat problems?
12:10:14Jungti1234viewing angle?
12:10:27Jungti1234Angle of vision
12:10:35LinusNit changes color when viewing from an angle, yes
12:11:16Jungti1234It makes people's eye fast tiredly.
12:11:40LinusNi think the h300 lcd is damn good, except for the angle problem
12:11:44Jungti1234There is H300's LCD two.
12:12:05Bgernew LCD ?
12:12:18Jungti1234old LCD, new LCD
12:12:49LinusNi don't think i've seen a difference on my players
12:12:52Jungti1234We call new LCD that is 'Horizontal LCD'.
12:13:59Jungti1234and we call old LCD is 'Vertical LCD'.
12:14:24 Join muesli__ [0] (i=muesli_t@
12:14:49Jungti1234It makes eye as isn't good.
12:15:25Jungti1234This problem is issue in the Korea now.
12:16:48 Join mawe [0] (
12:20:30 Join thames [0] (i=XiRc_V26@
12:20:50thamesHello :)
12:21:18mawehi all. i can't find any info about developing plugins for rockbox. are there docs for the plugin api somewhere?
12:22:37linuxstb_Only the source code AFAIK - see plugin_struct in apps/plugin.h and all the existing plugins
12:22:46DrMoosLinusN: hi, is the LCD is good in the X5 compare to h3XX?
12:22:49linuxstb_There is a GraphicsAPI page on the Wiki which will probably be helpful as well.
12:23:21 Quit TCK (Read error: 104 (Connection reset by peer))
12:23:29LinusNDrMoos: the x5 lcd is crisp and bright, but it is rasterized, i.e you see the pixels
12:23:42linuxstb_mawe: apps/plugins/helloworld.c is a good place to start.
12:24:08DrMoosLinusN: k, thanks for info
12:24:15maweok, thanks!
12:24:24Jungti1234DrMoos: X5 LCD is better than H300.
12:24:58DrMoosdid you try it both?
12:25:22Jungti1234I witnessed it.
12:26:50LinusNi like the x5 a lot
12:26:55DrMoosLinusN: I'm wondering if a human brain could work on H3xx bootloader and Iaudio one in same time? :D
12:27:11DrMoosLinusN: I planed to buy a 60 go one
12:27:16LinusNthe x5 will probably be my player of choice when i have rockbox running on it
12:27:29maweoh, one more thing: i tried rockbox on my iriver yesterday for the first time, very impressed!! is there some way to get support for powered mics like in the iriver fw?
12:27:35DrMoosX5 60 gb yes
12:27:50LinusNmawe: there is support already
12:28:06LinusNmawe: use Line In as the source and adjust the ADC Gain
12:28:08DrMoosmawe: you can forgot the original fw ;)
12:28:26 Join mikearthur [0] (
12:28:32maweah, fanstastic. what does adc stand for?
12:28:50LinusNAnalog to Digital Converter
12:29:07mawehm... should have guessed that... ;)
12:29:30mawerecording to mp3 doesn't work yet, though. or does it?
12:29:53preglowbut you can record, then encode to mp3 :-)
12:29:55Jungti1234DrMoos: There is X5 to 30 GB in the Korea.
12:31:25mawenow all we need is a chromatic tuner plugin and an abc file viewer and the iriver is the perfect musicians companion! :) will try to tackle the instrument tuner
12:31:32DrMoosJungti: are'nt the 60gb one?, strange Iaudio is Korean company at start no? :)
12:32:26Jungti1234yes. but, there are no 60 GB.
12:33:30DrMoosJungti1234: here in France is worth, but internet is our friend :)
12:33:53DrMoosvery rare french internet shiper in France
12:34:05DrMoosand 0 market
12:34:51Jungti1234ah hahaha
12:34:56 Join amiconn_ [0] (
12:35:02mawewhat is the difference between the adc gain and the gain shown on the recording screen? will i get clipping or loose resolution if i set the adc gain to max and level-in the recording with the gain slider?
12:35:03Jungti1234what is 0 market?
12:35:08 Join lavi [0] (i=XiRc_V26@
12:36:00DrMoosJungti1234: here in france isn't stores for buy it( and test it)
12:36:05laviHello ~
12:36:57Jungti1234I understood.
12:37:13DrMoosJungti1234: I planed to buy a 60 one in Germany ;)
12:37:35 Quit Shani}{Coder (
12:37:35 Quit markun (
12:37:35 Quit thames (
12:37:35 Quit NicoFR (
12:37:35 Quit amiconn (
12:37:35 Quit _FireFly_ (
12:37:35 Quit XavierGr (
12:37:35 Quit edx (
12:37:35 Quit godzirra (
12:37:35 Quit Ismo (
12:37:37 Nick amiconn_ is now known as amiconn (
12:37:46Membrilloserver split
12:37:49Jungti1234DrMoos: How much?
12:37:55NJointhames [0] (i=XiRc_V26@
12:37:55NJoinNicoFR [0] (
12:37:55NJoin_FireFly_ [0] (
12:37:55NJoinXavierGr [0] (
12:37:55NJoinedx [0] (
12:37:55NJoinShani}{Coder [0] (i=Coder@Is.A.TclCoder.Org)
12:37:55NJoinmarkun [0] (
12:37:55NJoingodzirra [0] (n=shawn@
12:37:55NJoinIsmo [0] (i=laitinei@
12:38:38aliaskmawe: I wrote crappy wiki page about how to make plugins for rockbox, I'll see if I can find it
12:39:08aliaskAh here it is:
12:39:37 Quit lavi (Client Quit)
12:39:43 Join DreamTactix291 [0] (
12:40:11Jungti1234I read it, but programming is difficult. :(
12:40:25mawealiask: thats great, thanks!
12:40:31DrMoosJungti1234: here or in Switzerland for 370 euros
12:40:32aliaskHope it helps.
12:41:12aliaskOh, also, if anyone disagrees with anything I wrote there, then I'm probably wrong. Go ahead and update the wiki so we can all learn from it :)
12:41:13LinusNi have a 20gb x5, gotta get hold of a 40gb single platter hdd
12:41:30muesli__LinusN need a trader?
12:41:55DrMoosLinusN: I don't know yet if the 60gb one is mono or dual plater
12:42:14Bgervery probably dual
12:42:22Jungti1234DrMoos: I don't know German.
12:42:24DrMoosBger: I hope so
12:42:30muesli__i would guess 99,9%
12:42:35DrMoosJungti1234: me too ;)
12:42:44Jungti1234How much?
12:42:46DrMoosHi muesli btw
12:42:54Bgerit it's single plater that means they are ready for a 120GB dual ...
12:42:55muesli__hola moooos
12:43:14LinusNmuesli__: i need a place to buy it , and money :-)
12:43:16muesli__60gb single??
12:43:29markunLinusN: I also couldn't find anymore info on the Gigabeat G's CPU
12:43:40muesli__LinusN ;)
12:43:52markunThe DAC is a Cirrus Logic CS43L42
12:44:33LinusNmarkun: i emailed them
12:44:33LinusNasking for the data sheets
12:44:33LinusNthe chip looks very simple, no custom dsp stuff
12:44:45Jungti1234DrMoos: Don't you know how much it is?
12:45:08DrMoosJungti1234: 370 euros in Switzerland for 60gb one
12:45:43Jungti1234438000 won..
12:45:51markun8kB cache and 80kB IRAM
12:46:01LinusNaliask: i don't think you should have lessons in C on that page
12:46:35LinusNthat could be a separate page, this page should only be about plugin specifics
12:46:44DrMoosLinusN: did you already have time to take a look at harware things in x5?
12:46:48 Join Rob2222 [0] (
12:47:06aliaskWell, when I was learning to code plugins, I was really learning C at the same time, so when I was writing it I found it hard to differentiate between what was needed and what wasn't.
12:47:24Membrilloon the volume bars for listening to music (ie on the iamp300 WPS) can you raise the minimum value? my bars always hang up around the top and it doesnt give them much freedom to move
12:47:28LinusNDrMoos: yes
12:48:07DrMoosLinusN: how looks the beast? :)
12:48:46LinusNDrMoos: it's pretty similar to the h300 in many respects
12:48:52 Join akaidiot [0] (
12:48:59LinusNsame pcf50606 chip for instance
12:48:59LinusNthe lcd interface is a disappointment
12:49:58LinusNthe lcd controller is 16-bit, and emulates 18-bit color by dithering
12:50:20DrMoosI read in the wiki are using coldfire 5250, are they real differences with 5249 one?
12:50:31DrMoosLinusN: O.O ok
12:50:35LinusNit seems cowon decided to use the fake 18-bit mode, and this uses the 9-bit lcd bus interface
12:51:04LinusNwhich forces us to do two writes per pixel in lcd_update()
12:51:13aliaskWhy would they do that?
12:51:26 Quit _FireFly_ (Success)
12:51:35LinusNthey probably wanted to do 18-bit color
12:51:42 Join Maxime` [0] (n=flemmard@
12:51:57aliaskBut if the LCD hardware is only 16 bit, it won't matter will it?
12:52:09 Join Febs [0] (
12:52:53LinusNand the coldfire bus is 16-bit only, so it can't do 18-bit writes
12:52:53LinusNso it is split into dual 9-bit writes
12:53:02LinusNthe 5250 has more IRAM (128k vs 96k) and more dma channels
12:53:10 Quit thames (Connection timed out)
12:53:16aliaskSo it needs two writes to update the LCD?
12:53:59 Join petur [0] (
12:55:45 Quit LinusN (Read error: 104 (Connection reset by peer))
12:56:14peturmawe: did you get an answer about your recording question already?
12:56:49DrMoosgood to know :)
12:57:29DrMoosLinusN: already planed something for this IRAM plus and dma chanels?
12:58:42aliaskThe lcd framebuffer might fit in, but it wouldn't leave much space for other stuff.
13:01:11B4gderwell, the x5 only has 160x128
13:01:16preglowcan anyone think of a more flexible way to define what code goes in iram?
13:01:25preglowwe're going to need to differentiate between platforms there
13:01:42preglowipod code is larger than m68k code, iaudio has same size code, but more iram, etc
13:01:53 Quit ehntoo ("Leaving")
13:01:58 Quit Rob2222_ (Read error: 110 (Connection timed out))
13:02:39B4gderpreglow: its a tricky nut to crack...
13:02:42preglowB4gder: indeed
13:02:49preglowB4gder: i'd hate to see #ifdefs around every single ICODE_ATTR
13:02:57preglowbut i fear it's the only solution
13:02:58B4gderthat'd be terrible
13:03:48B4gderor some oddly contructed ICODE_ATTTR_LESS_IMPORTANT defines
13:03:52 Join thames [0] (i=XiRc_V26@
13:03:56preglowproblem is, we'd need many of them
13:04:07preglowthere are several cases that need differentiation
13:04:15preglowlike same arch, but different iram sizes
13:04:20preglowthen again, different arch, same iram size
13:04:25preglowand combinations :-)
13:04:34DrMooshehe :)
13:04:47B4gderwell, perhaps we could perhaps have a prio num and have config-*.h define what prio nums that go into iram
13:05:01B4gderor something in that style
13:05:14preglowyeah, that's what i thought about first
13:05:39preglowproblem is, once we bump into a problem on one particular arch/iram size combination, we'll need to do an adjustment that will also possibly affect other archs
13:06:20B4gderyes, we need to be able to enable/disable specific things to go into iram
13:06:27preglowand iram is sufficiently important to performance that what we really want is a completely tailored setup for each arch
13:06:37preglowbut i don't know how that can be done without ifdef hell
13:07:14B4gderI'd consider ICODE_IATTR_[NAME] to be better
13:07:27B4gderwith NAME being lots of different ones
13:07:31preglowahh, yes
13:07:36B4gderand config-*.h setting up which ones to go into iram
13:08:30Jungti1234I will come again
13:08:32preglowthat would pretty much be the complete solution i was talking about :-)
13:08:35 Join gtkspert_ [0] (
13:08:35 Part Jungti1234
13:08:46 Quit Rob2222 (Read error: 110 (Connection timed out))
13:08:57preglowwe'll probably need this in the core and with some select plugins/codecs
13:09:12 Quit Maxime (Read error: 113 (No route to host))
13:11:00 Quit NicoFR ()
13:14:05 Quit muesli__ (Read error: 110 (Connection timed out))
13:14:36 Join DangerousDan [0] (
13:14:45thamesI have a shooting pain in my eyes because of H300 LCD problem. :<
13:15:13aliaskWha?! Really? With the new LCD?
13:15:41thamesnew LCD?
13:16:06 Join muesli__ [0] (i=muesli_t@
13:16:09aliaskApparantly there are two versions of the LCD, one which has the problem and one that doesn't.
13:16:27aliaskDoes it start hurting after prolonged viewing?
13:17:09aliaskHrm, maybe this is the motivation for removing video in the 1.30EU firmware.
13:18:13peturI bet you were playing games on it :D
13:19:15Membrilloor looking at optical illusions.
13:19:19 Join Jungti1234 [0] (n=jungti12@
13:19:49Jungti1234I came back!
13:21:15 Quit gtkspert (Read error: 110 (Connection timed out))
13:23:09 Join LinusN [0] (
13:23:25thameshello :)
13:24:02aliaskwb Jungti1234, LinusN.
13:26:46thamesaliask : sorry, i would like further information new LCD.
13:27:18aliaskI'm just regurgitating what I heard earlier tonight from Jungti1234.
13:27:44aliaskLet me look up in the logs.
13:29:04thamesum... do you have same LCD problem?
13:29:05 Join adamd [0] (
13:29:32adamdhey, just wondering... is the v5 bootloader now safe? [v4 was unsafe wasn't it?]
13:29:46aliaskYes, I've personally tested it, and it works great. :)
13:29:54adamdcool :) what was the problem with v4?
13:30:05adamdand what's new with v5
13:30:17aliaskToo complicated for me to understand. Basically a chip was turning itself off after waiting too long.
13:30:31aliaskNothings new with v5, its just an unbroken v4.
13:30:35B4gder... since the bss had grown larger
13:30:46adamdhaha, well what was new with v4
13:30:55B4gder.. due to the unicode support in the fat code
13:30:58adamdi'd look for a changelog but my lunch break is about to end
13:31:09adamdin fact there it goes
13:31:26aliaskPlug usb in while the player is off, and it boots rb instead of iriver firmware.
13:31:36adamdahh that'll be better
13:32:15aliaskthames: Apparently the old LCD is a "vertical lcd" and the new one is horizontal. It causes eye strain apparantly.
13:32:20adamdneed to run
13:34:39Membrilloyou mean, it refreshes vertically\horizontally?
13:34:53Membrilloisnt it just a matrix of pixels?
13:35:38preglowbut you can refresh the screen in tons of ways
13:35:38LinusNthe original h300 lcd is actually 176x220, rotated 90 degrees
13:35:56preglowis that so
13:35:58peturat least it makes no difference for the software or we would have seen complaints...
13:36:47LinusNthe rockbox lcd driver sets the lcd controller to rotate the GRAM matrix
13:37:17Jungti1234heh #iriver
13:38:40 Nick thames is now known as lavi (i=XiRc_V26@
13:39:39 Quit Bger ("[BX] A friend with weed is a friend indeed")
13:39:47 Part Jungti1234
13:39:51 Quit muesli__ (Read error: 110 (Connection timed out))
13:40:21Membrilloi get what you mean
13:40:32aliaskOh, I thought it was to do with the viewing angles being greater in one direction than the other.
13:40:41aliaskI guess I'm not always right :)
13:41:09Membrilloso the LCD is naturally vertical? but its been rotated for installation in the iriver?
13:41:14Membrilloor the other way round
13:41:17 Quit saa[b_r]ider ()
13:41:57B4gderit seems most of the 220x176 lcds are targetted for/used in phones, and thus they are higher than wide
13:42:20B4gdernot that it matters when you can rotate them
13:42:25aliaskIt's true: my phone is 176x220.
13:42:46B4gderif you google for those dimensions, you'll find a 100 phones with it
13:42:50 Join muesli__ [0] (i=muesli_t@
13:43:16aliaskYeah, it wouldn't supprise me. It's a decent size for a phone.
13:43:30Membrillobut the phones dont have the processing power to use it like the iRiver does
13:43:37B4gderthey don't?
13:43:42Membrillowhats the res of the ipod 4g?
13:43:44B4gderI'd say most certainly have
13:43:55aliaskSome phones have 500mhz+ ARM processors.
13:44:08B4gderthere are phones with win CE in them
13:44:12B4gderinagine the power they need
13:44:28aliask(those phones run windows mobile)
13:44:31Membrilloi guess i just have the image in my head that phones are shitty
13:44:54B4gderand most phones these days run java apps
13:44:59B4gderthey EAT cpu
13:45:06Membrilloyou know they have windows 95 emulated on PSPs
13:45:31Membrilloyeah Java is very common on phones these days
13:45:56petur320x240 30fps DivX is no problem on a 300MHz WM2003 device...
13:46:16Membrillowhats the iriver 120mhz huh
13:46:24B4gderpetur: you mean ARM?
13:46:33aliaskBecause it won't have to resize the video at those dimensions for most PDAs.
13:46:46peturXScale is arm, yes
13:47:14 Quit lavi ("Ping timeout for]")
13:47:25B4gderpetur: is that really only done in pure sw?
13:47:37peturI think they have some graphic acceleration - not sure
13:47:57B4gderI'd say it is likely they have some sort
13:48:17B4gderI worked with a 300Mhz scale
13:48:20peturbut betaplayer runs on most pda's, so I guess it works without it.
13:48:39preglow320x240 divx isn't impossible on 300mhz xscale
13:48:52preglowdepends on whether interpolation is done in hardware, i guess
13:49:20B4gderanyway, we're drifting... :-)
13:49:52aliaskSo who wants to work on 220x176 divx on a 124mhz coldfire?
13:50:09B4gderaliask: there's work started, check the patches
13:50:17peturI'll just promote my recording gain patch before I go :D
13:50:43LinusNpetur: auto-scaling, or two settings?
13:50:45aliaskI should really check the patches more often.
13:51:05LinusNthe devs should check the patches more often :-)
13:51:05peturanalog+digital combined
13:51:23peturand left+right together trick
13:51:44LinusNleft+right together?
13:52:14peturyes. i you move down to gain, it first select L+R together, then L, then R
13:52:26aliaskI still think that my temporary fix for the yesno screen should be committed until we get the updated lang system.
13:52:29petureasy to change bothe channels at the same time
13:52:44LinusNpetur: will it work on the archos?
13:53:09peturonly implemented for UDA1380
13:53:22LinusNthe l+r thing too?
13:53:32peturI'm affreaid, yes
13:54:20peturwant it for the other ones too, heh?
13:54:30***Saving seen data "./dancer.seen"
13:54:43LinusNthe line in gain is l+r on the archos as well
13:55:03peturis it? haven't checked that
13:55:28Membrillohas anyone got auto gain adjusting for recording on their minds?
13:55:46LinusNnot me
13:55:55LinusNi'm far too busy to think of that
13:55:57Membrillopetur? your mr record :P
13:56:01B4gderI have a cup of coffee onb my mind
13:56:22B4gderand in my hand
13:56:36Membrilloits too late for coffee here
13:56:39Membrilloim on to tea
13:56:50peturI have tea and work in front of me...
13:57:08Membrillook cool
13:57:23*B4gder does the track-the-linux-trail-driver-update thing today
14:04:15lostlogicpreglow: I patched it... and then fixed the memcpy and memset asm routines to not assume stack args, but sadly there was no performance gain at least on wav2wv (regparm=3)
14:04:33preglowmeasuring core code would be better
14:04:37preglowsince that doesn't have iram stacks
14:04:44preglowthat is, the core thread probably does, but not the rest
14:05:25lostlogicpreglow: do you have an idea on how to measure something like that? when I have a DAP again, I'll revisit.
14:06:03 Quit akaidiot (Read error: 110 (Connection timed out))
14:07:22preglowno idea, you're the profiler dude ;)
14:07:59peturgot to go..
14:08:03 Quit petur ("CGI:IRC 0.5.7 (2005/06/19)")
14:08:07preglowit's really bloody weird that d0, d1, a0 and a1 aren't used for parameters
14:08:10preglowthey're scratch regs anyway
14:08:14preglowhow could it possibly hurt
14:08:28lostlogicpreglow: yeah, like why did it take GCC this long to learn to use regs for parms.
14:08:29 Join Jungti1234 [0] (n=jungti12@
14:08:38preglowit's a platform thing
14:08:50preglowthe motorola compilers probably never did, and gcc continued for portability
14:08:56preglowarm, for example, does register passing
14:09:04preglowand x86-64 might be
14:09:25 Quit Maxime` (Read error: 104 (Connection reset by peer))
14:09:33lostlogicI'm not sure if profiling will 1) work or 2) show any difference evenif there should be one with regparm due to the extra function calls at the start and end of every function.
14:09:46preglowthat's probably true
14:09:47preglowbut anyway
14:09:54preglowthere will be a performance gain
14:09:56preglowno question about it
14:10:02preglowi can't possibly imagine there not being any impact
14:10:36preglowso if regparm functionality were to become standard in gcc for m68k, then i'd go for enabling it
14:11:06markunLinusN: can you help me identify the last 2 components?
14:11:10lostlogicthe wav2wv test that Membrillo did for me last night showed no statistically significant change...
14:11:37lostlogicI wonder why it hasn't been patched in −− the patch I found on the MLs is pretty straightforward for enabling it.
14:12:02 Join linuxstb [0] (
14:12:33 Join Maxime [0] (n=flemmard@
14:12:33preglowbbl, meeting
14:13:30linuxstbI would expect that frequently used functions would have already been optimised to be inline anyway - so it doesn't really surprise me that the gain isn't significant.
14:14:29Membrilloi did a wav2wv test on a 5 and a bit minute wav and the difference was half a second i think
14:14:39 Join DJDD_ [0] (
14:14:56Membrillobetween the original and lostlogics optimization
14:15:08lostlogiclinuxstb: yeah, that's what I concluded −− might be able to save some code cache space by de-inlining without taking a big performance hit with regparm, but that's mroe than I'm likely to do anytime in the very near future.
14:16:15B4gdermarkun: I get lots of hits for "508a2" - but they all seem to assume we know what it is...
14:16:47 Quit Maxime (Read error: 104 (Connection reset by peer))
14:16:53 Join Maxime [0] (n=flemmard@
14:17:39markunB4gder: on the X series this is writen on the IC: P003 526AA
14:18:13markunThere are 3 of them, located near the USB ATA bridge
14:19:50adamdare there any plans to have a more accurate radio scanner... ie smaller steps than 0.1MHz
14:19:55adamdon the H300
14:19:56linuxstbmarkun: Are you going to try requesting more source from Toshiba, or are you happy to leave it as it is?
14:20:34linuxstbDid they even include the .config file for the kernel?
14:20:46markunno, they didn't
14:20:59linuxstbBad Toshiba....
14:24:51linuxstbI would reply to them, thanking them for the source, but pointing out to them that they the GPL requires them to also provide "the scripts used to control compilation and installation of the executable", referring them to section 3 of the GPL.
14:25:52B4gderyou could also make the sources they sent available
14:26:04B4gderat least as a patch against whatever kernel release they used
14:27:26markunok, I will
14:29:47Jungti1234LinusN, Are you there?
14:29:58 Quit muesli__ (Read error: 110 (Connection timed out))
14:30:23Jungti1234I have question
14:33:39Jungti1234LinusN: Do you know H300's time initialization?
14:34:31LinusNthe iriver firmware does it wrong, we do it right
14:34:47jelle-kwhat do they do wrong?
14:35:03B4gderthey set the year with a wrong base
14:35:07LinusNthey use the "year" field in the RTC in the wrong way
14:35:42LinusNi'm wondering if we should adapt on the h300 or not
14:36:03LinusNit's quite inconvenient for the h300 users when they run the original firmware
14:36:43linuxstbWhat base does the h300 firmware use?
14:36:59LinusNthey set it based on 1970
14:37:08linuxstbhaha. So leap years are wrong?
14:37:11LinusNwe use 2000
14:37:32B4gderyes, leap years are wrong
14:37:40jelle-kmy pc uses 1970 too?
14:38:13B4gderyou mean your motherboard's RTC?
14:38:34jelle-kasrock crap :P
14:38:41B4gderI'd guess it uses whatever that RTC says it should use
14:39:02B4gder1970 is just the regular time() base
14:39:08B4gderit doesn't mean the RTCs have to use it
14:40:21linuxstbCould we do an "if rtc_year > 30 then year=1970+rtc_year else year=2000+rtc_year " hack when reading the RTC?
14:41:25B4gderbut that screws up the original fw's view
14:41:34lostlogicproblem is that if it's set _right_ the original firmware reads it stupid and resets it
14:42:29LinusNwe could just do it in the same wah as the original, but that screws up the leap year detection
14:43:04linuxstbAnyone fancy disassembling and fixing the iriver firmware? :)
14:43:04B4gderand that feels... wrong
14:43:10 Quit Febs (Read error: 110 (Connection timed out))
14:43:15LinusNbut that shouldn't matter, right?
14:43:23 Quit ashridah (Read error: 110 (Connection timed out))
14:43:28LinusNwhen is the next leap year quirk?
14:43:46linuxstbthis year
14:43:55linuxstbYear 36 according to the iriver firmware.
14:44:00Nibblernah? really?
14:44:12B4gderamiconn did several leap year tests the other day
14:44:26linuxstbSo h300 users will have a 29 February this year :)
14:46:03Jungti1234Rockbox likes 2040!
14:46:08linuxstbBut I think h300 owners would be able to live with resetting the time once every 2 years - so my vote goes for copying the iriver bug.
14:46:23*lostlogic shudders
14:49:34*Membrillo realises he's never even realised that the time is on rockbox now
14:50:26markunB4gder: Is there a way to compare 2 diffs?
14:50:45LinusNmarkun: not without getting a severe headache :-)
14:51:02lostlogicamiconn: I'm guessing you'll hate this idea, but what do you think about getting the assembly versions of memset and memcpy into gcc inline assembly instead of straight assembly so that they aren't assuming a particular parameter passing model? (doesn't matter in practice until/unless the assembly targets get regparm support)
14:51:12B4gdernot that I know of, no
14:51:14linuxstbmarkun: I would apply the diffs to two copies of the source files, and then do a diff on those two sets of files.
14:51:17*Jungti1234 ah
14:51:44B4gderyes, I would do the linuxstb approach too
14:51:51markunlinuxstb: yes, I also wanted to do that, but the rmk6 patch didn't cleanly apply to the kernel source..
14:52:46markunWell, a diff of the 2 diffs also shows what was added by Toshiba, but it doesn't look very nice :)
14:53:33 Quit B4gder ("time to say moo")
14:53:57linuxstbI can imagine that would be confusing...
14:54:52markunHere's what that looks like:
14:54:59markunNot unreadable
14:56:41 Join Kohlrabi [0] (
14:57:39maweis there a device independent way to get a ringbuffer of audio samples from mic input?
14:58:13maweor even just the raw samples, could build my own ringbuffer....
14:58:44LinusNmawe: not yet, at least not in a plugin
14:59:03mawei see the pcm_record.c does creates a ringbuffer.... but it lookds device dependent
14:59:19maweLinusN: ok... shame
14:59:25LinusNwe only have one platform that can record pcm data
15:00:07LinusNonce we have several, a device independent way will probably be created
15:00:21LinusNmawe: tuner plugin?
15:00:38LinusNkind of guessed that
15:00:57maweso there's no way to get to that data from a plugin at the mo, correct?
15:01:17LinusNyou are much welcome to tinker with the pcm recording functions to implement that feature
15:01:47LinusN(getting pcm data in a plugin, that is)
15:02:14mawehm. i think i will try that. pcm only works on iriver?
15:02:21LinusNso far, yes
15:02:53maweok, thanks! back to the drawing board ;)
15:03:01LinusNhave fun
15:03:29 Join webguest23 [0] (
15:03:56 Quit webguest23 (Client Quit)
15:04:02 Join webguest62 [0] (
15:05:03webguest62I really fail to see a difference between "Insert" and "Insert Last", anyone care to enlighten me
15:05:31 Join muesli__ [0] (i=muesli_t@
15:05:34Membrillowell insert is insert next
15:05:46LinusNwebguest62: "insert" inserts after the last inserted file, "insert last" inserts last in the playlist
15:05:46Membrillothere is another that i didnt know
15:06:01Membrillooh really?
15:06:02amiconnlinuxstb: I vote for doing the date handling correctly in rockbox on H300
15:06:04Membrillothats cool
15:06:07Membrillodidnt know that one
15:06:12LinusNi think so at least
15:06:38MembrilloLinusN: so whats the difference between insert and queue?
15:07:02webguest62with queue the songs are removed after being played
15:07:05LinusNinserted files stay in the playlist after being played
15:07:20Membrillothats why they always disapeared
15:07:28LinusNprobably :-)
15:07:33amiconnlostlogic: Making the mem* functions inline would be a performance loss and code bloat
15:07:43amiconnToday, memset() and memcpy(9 are in iram
15:07:46LinusNa cache killer
15:08:05markunShouldn't patch also create dirs?
15:08:16amiconnAlso, memcpy() is almost 1KB
15:08:32webguest62now, when one creates a new playlist there is only a "Insert" menu no queue menu, so if shuffle is selected down the line, one hears that first album again
15:09:22Membrilloyeah... why cant you insert shuffled first off?
15:10:12webguest62or allow first added to be "queued" as well
15:10:53webguest62so that if shufle is selected down the road, first added won't be heard again
15:11:16Membrillogood question
15:11:48MembrilloLinusN: has this just not been put in yet, or left out for a reason?
15:12:04 Quit Vlad0man (Read error: 110 (Connection timed out))
15:12:10lostlogicamiconn: not _inline_ put them in a regular function as inline ASM instead of in a .S file...
15:12:27 Join Vlad0man [0] (
15:12:42LinusNMembrillo: i think it's just a glitch
15:13:11LinusNthe reason you can't queue first is probably because you need a playlist to queue to, and that doesn't exist
15:13:13Membrillook good good
15:13:28lostlogicamiconn: I've done it for the coldfire versions in order to test out regparm, works, the only difference in the assembly generated is that the registers are saved to the stack at the start of the function, which I assume would be done in the caller if it was calling the assembly version.
15:13:33LinusNbut rockbox could just as well create an empty list and queue to it
15:14:01webguest62so how to prevent first "added" from being heard again without manually removing them ?
15:14:11MembrilloLinusN: that would be good. so you could insert shuffled as well?
15:14:31LinusNsounds reasonable
15:14:39LinusNi don't have time to do it though
15:15:10webguest62Membrillo: after insertng the first album you can always "reshuffle the playilist
15:15:11MembrilloLinusN: yeah thats cool. as long as its a bug though, and not been left out intentionally
15:15:21Membrillohow do you reshuffle?
15:15:29LinusNi don't think it's intentional
15:15:31webguest62under the wps quick menu
15:16:04Membrilloholding A-B you mean
15:16:10Membrillois that the quick menu?
15:16:19Membrilloyou cant reshuffle the playlist there
15:16:24webguest62wps, long play press > playlist
15:16:45LinusNthat's the context menu, not the quick menu
15:17:32amiconnlostlogic: I don't see the problem. The m68k abi does specify the parameter passing
15:17:34Membrilloso whats the quick menu
15:17:51webguest62long press a-b and see
15:17:57LinusNMembrillo: hold A-B
15:18:18amiconnHaving the function completely in asm has lower overhead for calls, because the coder can decide which registers to save at which point
15:18:26Membrillothought so
15:18:36LinusNlong-navi -> context menu, long-a/b -> quick screen
15:18:46Membrilloi know its there, just didnt know it was named the quick screen
15:18:46 Join Febs [0] (
15:18:55 Join ^Guest37784 [0] (
15:19:07amiconnC functions don't allow to share code, and I want memmove() and memcpy() in one
15:19:18amiconnIt's already prepared, just not finished yet
15:19:25 Join zagor_testing [0] (
15:19:45 Part zagor_testing
15:21:09Membrillofrom wps going to the context menu. I see open with. Is that opening the now playing song with something?
15:21:50webguest62select playlist
15:23:13 Quit DJDD_ (Read error: 110 (Connection timed out))
15:23:39 Quit aliask ("G'night")
15:23:41lostlogicamiconn: the problem isn't a problem until/unless we want to enable regparm. but your point about code sharing is well taken :(
15:23:47 Part webguest62
15:24:24 Quit Membrillo ()
15:26:02 Quit Kohlrabi ("Fast alle Menschen sind Regenwürmer")
15:27:15^Guest37784LinusN: if i want to save with
15:27:24^Guest37784more than 1 save
15:27:28^Guest37784i need another file?
15:27:54 Quit Paprica (Nick collision from services.)
15:28:03 Nick ^Guest37784 is now known as Paprica (
15:28:13LinusNPaprica: please elaborate
15:28:43Papricajust check: you want that i ask again?
15:29:14LinusNwhat do you want to accomplish?
15:29:26 Quit DrMoos ("CGI:IRC (Ping timeout)")
15:29:29Papricai want to save with configfile_save
15:29:56Papricamore that one variable
15:30:01Papricai need to files?
15:30:14Papricamm how can i do that
15:30:17LinusNjust add another struct to the configfile array
15:30:50Papricasomthing like this?
15:30:50Papricastatic struct configdata config[2] = {
15:30:50Paprica{ TYPE_INT, 0, 1, &first_week_day, "first_week_day", NULL, NULL }
15:30:50DBUGEnqueued KICK Paprica
15:30:50Paprica{ TYPE_INT, 0, 1, &iriver_fw, "iriver_fw", NULL, NULL }
15:31:04Papricaok 10q
15:31:18LinusNyou forgot a comma though ;-)
15:31:38Papricaah right
15:31:48Papricamm and when i save or load
15:31:52Papricai do configfile_load("cal.cfg", config, 2, 0);
15:32:06LinusN*bing*! 10 points!
15:32:06Paprica2 is the num of item, isnt it?
15:32:25 Join DrMoos [0] (
15:32:44CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:32:44*Paprica take his points and back to work
15:41:40 Join Membrillo [0] (n=sam_kill@
15:45:57linuxstbamiconn: IMO, the biggest problem regarding the date is not the fact that the iriver firmware displays it incorrectly, but that it will apparently automatically reset dates it thinks are wrong - i.e. any date set by Rockbox.
15:46:14preglowwe should set it like the iriver firmware does, really
15:46:14linuxstbResulting in the "bug" that the date is reset every time a Rockbox user boots into the iriver firmware.
15:46:31LinusNlinuxstb: yes, and a nuisance for the user
15:46:43Membrillowhats negative about doing it the way the iriver fw does?
15:46:56LinusNMembrillo: leap years will be wrong
15:47:36Membrillobetter changing it every 4 years than changing it every time you accidently press record when booting
15:47:42markunHere's the patch of the Gigabeat's linux source:
15:47:50 Join Bger [0] (n=Bager@
15:47:55linuxstbMembillo: It's every 2 years, but I agree.
15:48:12Membrilloleap years are every four years aren't they?
15:48:20 Join webguest81 [0] (
15:48:21 Join akaidiot [0] (
15:48:32markunLinusN: There is also some NMS7210 stuff in the patch
15:48:35Membrillowhen the olympics are on, those years are leap years
15:48:48linuxstbYes, but the iriver firmware will incorrectly think 2006 s leap year, and incorrectly think 2008 is not etc
15:49:00webguest81h1x0 is not building with logf enabled :-\
15:49:09LinusNwebguest81: dsp.c?
15:49:15Membrillo2 years is a long time
15:49:27Membrillostick to that
15:49:33webguest81LinusN: yup
15:49:38Membrilloit wont offend anyone
15:50:44LinusNwebguest81: line 469, change to #if 0
15:51:47 Quit Membrillo ()
15:51:54webguest81Builds now, yes
15:52:09webguest81at least, it continues
15:52:43webguest81ah, another error
15:52:47amiconnlinuxstb: Using the wrong iriver approach in rockbox breaks leap year handling. I wouldn't care about iriver fw
15:53:42webguest81amiconn: annoying if you use the iriver fw for recording though (if you want mp3 recording)
15:53:45linuxstbamiconn: Personally I agree with you, but in this case, I think we should think of the users. I'm sure lots of h300 users still use the original firmware for either video or WMA playback. You also need to use it for flashing.
15:54:22FebsI just uploaded the iAmp300 wps and theme to the patch tracker. If anyone with commit access would commit it to CVS, I would appreciate it. This WPS should probably replace iAmp in H300 builds.
15:54:33***Saving seen data "./dancer.seen"
15:54:45LinusNwebguest81: same there, #if 0
15:54:58FebsThe .zip file contains everything: wps, rwps, .cfg file for the theme, and all necessary bmps.
15:55:17preglowstill register allocator thing?
15:55:27LinusNlinuxstb: ans also because you use the iriver firmware for charging
15:55:32preglowi thought someone said we used fomit-frame-pointer in debug builds now
15:55:47LinusNi don't think we do
15:55:59LinusNnobody changed it i guess
15:56:16Papricahaha i found a temp (i hope) solution for the years problem in the rtc
15:56:24Papricai use it in rockcalendar
15:56:35webguest81that was the last one
15:56:38amiconn_if_ we adjust for the buggy iriver fw, I'd adjust it so that the values are in the valid range for iriver, but also leap years are correct
15:56:52amiconnThis will result in a 1-year difference between rockbox and iriver
15:57:10amiconnStill I think someone should point out the leap year bug to iriver
15:57:13Papricai will release a new version in a while
15:57:24linuxstbamiconn: I'm sure they will in about 6 weeks time...
15:58:30linuxstbamiconn: Do you know what the valid range is for the iriver?
15:58:35 Quit DreamTactix291 (Read error: 110 (Connection timed out))
15:59:06linuxstbWould you just use a base of 1972?
16:00:29Xerionthe correct solution would be to make rockbox have every feature of the iriver firmware ;)
16:04:36markunthe continuous HDD spinning was also a great feature..
16:04:59markunIs it fixed in newer firmwares btw?
16:05:54Xerionno idea, i don't use them :)
16:06:10markunno, me neither
16:06:53BHSPitLappygreetings, #rockbox
16:07:06 Join darkless [0] (
16:08:08preglowit was fixed somewhat
16:08:34preglowi don't think they've got a threaded architecture, so is probably somewhat difficult, and requires checks bloody everywhere
16:09:24webguest81Anyone interested in a fun aac problem? I hit metadata.c:1150 with the chunk being "iods"
16:09:30markunamiconn: you'll like this. The new Gigabeat has even more buttons on the side ;)
16:10:11linuxstbwebguest81: Is that with a Nero-encoded file?
16:10:52preglowdidn't lear write a new mpeg4 parser?
16:10:56webguest81linuxstb: iTunes, I think
16:11:14linuxstbwebguest81: Is there anything unusual about it?
16:11:17webguest81Don't know for sure though
16:11:29webguest81Not that I know of
16:12:12linuxstbpreglow: I don't know. The one currently in Rockbox is one I wrote based on the code in the alac decoder. It needs a lot more work.
16:12:57preglowi think he started coding a new one
16:13:04preglowhe mentioned he was playing a nero encoded file at one point
16:13:27linuxstbThat's cool. It's not worth investingating bugs in the existing parser then.
16:13:38linuxstbAlthough the code in metadata.c is different.
16:14:02webguest81okay, I'll try again once a new parser appears
16:14:07preglowi'm not sure, though, may be delirium
16:14:07amiconnImho the correct solution would be to remove the iriver fw
16:14:27linuxstbwebguest81: I would still be interested in a copy of that file though - it sounds like a good test file.
16:14:34amiconnmarkun: Gah, ugly
16:14:45webguest81linuxstb: I'll put it somewhere
16:15:27preglowwhat's up with all these clever socket positionings?
16:15:32preglowac adapter on the _side_ ?
16:17:01markunamiconn: what do you think of this one?
16:17:27LinusNpreglow: not half as bad as having the headphone connector on the side (iaudio x5)
16:17:51markunLinusN: yes, don't understand why they did that..
16:19:22 Nick webguest81 is now known as nicknick (
16:20:00 Nick nicknick is now known as nnick (
16:20:16amiconnlinuxstb: The valid date range should be retrievable from the irc logs. I let a H300 owner carry out some tests some days ago
16:20:49preglowLinusN: indeed not
16:23:30amiconn starting at 23:07, and continuing
16:25:15amiconnI'd choose an offset of 36 years instead. That means iriver will show 2005 if rockbox is set to 2006
16:25:33amiconn...but the leap years will be correct
16:27:03nnicklinuxstb: did you get my pm?
16:27:39 Quit Febs ("CGI:IRC (EOF)")
16:28:35markunWow, I think I've found an even more ugly player:
16:29:52nnickthat looks like it could be a phone
16:31:03lostlogicw00t, my replacement iRiver is at UPS for me today.
16:31:03markunyes, it looks like one, but they call is "Digital Audio Player
16:31:06markunmaycom XT-21"
16:31:51*lostlogic writes "I will not fry this player by being a moron" on the chalk board repeatedly
16:33:20preglowtry being a fool next time
16:33:21preglowworks for me
16:33:35 Join Borg_Number_One [0] (
16:33:49*preglow prepares to be assimilated
16:34:26 Join menollo [0] (
16:35:12DBUGEnqueued KICK Borg_Number_One
16:35:12Borg_Number_One(google: bnobtc or Borg Number One or: "atasx")... this will help you to understand my question: Where can I get ,1ArchosUnlock.exe ?
16:36:42LinusNi don't know
16:37:26Borg_Number_OneDo you know, what I mean with ,1ArchosUnlock.exe at all? ,1:)
16:39:31 Quit mikearthur (Remote closed the connection)
16:39:41***Alert Mode level 1
16:39:41Borg_Number_OneReffering to this: ,1 there already was someone who asked for the tool.
16:40:14 Join mikearthur [0] (
16:41:07muesli__quick usb-charging already possible?
16:42:31lostlogicmuesli__: on what player?
16:43:33lostlogicIf my experiments with the power supply chip are correct, you can make it charge from USB on rockbox by starting a plugin that ignores the USB connected event and plugging in USB. Don't quote me on it, and it's not officially supported yet.
16:44:03muesli__but sounds like a good approach
16:44:53lostlogicI believe the concern is that USB does not provide enough power to both charge and operate the player, so you shouldn't try that in a plugin that boosts or uses the disk either.
16:45:11lostlogicso I take no responsibility for any bad effects of listening to me :-D
16:45:45Papricah3xx users?
16:46:18 Quit Shani}{Coder (No route to host)
16:46:24LinusNlostlogic: that may be true
16:46:25muesli__lostlogic i wouldnt use charging via usb but some have a broken charging plug
16:46:53markunBorg_Number_One: you might want to email the guy from
16:48:22lostlogicLinusN: 500mA is <1/2 of the current that would be needed to do a full speed fast charge, which would explain the 2A or 1.2A spec on the AC adapter
16:49:42***Alert Mode OFF
16:49:58 Join |Beowulf| [0] (
16:52:03Borg_Number_One,1<markun> Borg_Number_One: you might want to email the guy from Yes... it would be the best solution.
16:53:01Borg_Number_OneOr I ask to figure out the number.
16:54:19 Quit hiro^ (Read error: 104 (Connection reset by peer))
16:58:17 Join _FireFly_ [0] (
16:58:50 Join webguest33 [0] (
17:05:54 Quit akaidiot (Read error: 110 (Connection timed out))
17:05:59 Quit Zagor ("Client exiting")
17:13:14 Join Rob2222 [0] (
17:13:29Borg_Number_OneMhhh... might also help to find archosunlock.exe ,1:)
17:17:02Papricammm linusn
17:17:08Papricahighscore_save work fine?
17:20:23Borg_Number_OneDoes someone know "Alex Mina's" current homepage?
17:22:37 Quit menollo ("CGI:IRC (EOF)")
17:23:54 Quit Borg_Number_One ()
17:26:19lostlogicamiconn: you want to reuse code for memmov and memcpy to reduce rockbox's footprint?
17:30:50 Quit edx (Read error: 110 (Connection timed out))
17:31:37LinusNlostlogic: who said anything about fast charge?
17:34:39lostlogicLinusN: the 'normal' LiPoly / LiIon charge cycle is a 1A constant current charge up to cell nominal voltage, 400mA at 5V will not supply this.
17:35:48LinusNyes, but a lower current only means a slower charge, right?
17:37:20lostlogicright, but based on that observation it would make sense that it's insufficient current to power the player and charge?
17:37:55LinusNthat depends on how much current the player draws
17:38:03 Join DreamTactix291 [0] (
17:38:16 Join mirak [0] (
17:38:21lostlogicany idea how much that is during disk spin up and/or boost?
17:38:37LinusNnope, i haven't measured this
17:39:15amiconnlostlogic: Yes, especially since these fns are in IRAM
17:39:31amiconnmemmove is basically a doubled memcpy with 2 branches
17:39:52amiconnOne branch does copy forward, the other does copy backwards
17:40:03amiconnmemcpy() is just the forward branch
17:40:58lostlogicyah, too bad there's no way to share the code using C.
17:41:28amiconnlostlogic. LinusN: The LiIon charge cycle is _not_ constant current, but constant voltage (usually with current limitation)
17:41:34lostlogicwhen you do a jcc in assembly, what does GNU-as do with it? turn it into a bcc of the minimum length that will handle the offset?
17:41:55lostlogicamiconn: it is constant current for the first hour, then constant voltage once the threshold voltage is reached.
17:42:37amiconnIt's not time dependent at all
17:43:01amiconnA lower max. current will just increase the charging time with no other adverse effect
17:44:16LinusNstill, we need to select the proper charging mode when connected to usb
17:44:47lostlogicI wonder how the PCF50606 would handle the case where the current draw on it exceeds that from the charger, and it has a charged battery connected.
17:45:24LinusNhow could it draw more current than the charger delivers?
17:45:46lostlogicLinusN: when the demand from the player exceeds the supply from the charger.
17:45:52 Join Febs [0] (
17:47:03LinusNwell, that depends on the adapter, i'd say
17:47:36lostlogichow the PCF50606 will behave in that condition depends on the adapter?
17:47:37Slasheriwhile connected to usb and transferring files, the disk can take easily over 500 mA alone
17:47:55LinusNlostlogic: yes, if the adapter drops the voltage or not
17:48:07Paprica[::] <Paprica> highscore_save work fine?
17:48:26LinusNPaprica: nobody has ever used it
17:48:33lostlogicSlasheri: you know this? So when connected, what mode is the power supply in? USB powered, or battery powered, or can it draw on both as needed?
17:48:35LinusNfeel free to try it
17:48:58LinusNlostlogic: the hard drive always takes its power from the battery
17:50:08Slasherilostlogic: i don't know that, but at least H100 series player draws about 500 mA from battery when the disk is transferring data (or even more while playback is active and buffering)
17:50:16lostlogicI see... which would be why the charger always operates in trickle charge mode after completing charge in the iRiver firmware.
17:50:50lostlogicso then people who have had damage to their batteries from leaving them constantly plugged in are probably leaving them plugged in and _idle_ as opposed to plugged in and running.
17:54:38***Saving seen data "./dancer.seen"
17:55:51Papricawhat it means?
17:55:52Papricabrickmania.c:2535: warning: braces around scalar initializer
17:55:52Papricabrickmania.c:2535: warning: (near initialization for `')
17:55:52Papricabrickmania.c:2535: warning: initialization from incompatible pointer type
17:55:52***Alert Mode level 1
17:55:52Papricabrickmania.c:2535: warning: excess elements in scalar initializer
17:55:52***Alert Mode level 2
17:55:52Papricabrickmania.c:2535: warning: (near initialization for `')
17:55:54Papricabrickmania.c:2535: warning: excess elements in scalar initializer
17:55:56Papricabrickmania.c:2535: warning: (near initialization for `')
17:56:17 Quit YouCeyE ("Leaving")
17:56:52 Join ghode|afk [0] (
17:57:29 Quit _FireFly_ ("Leaving")
17:57:35lostlogiclaptop battery protecting systems have features where to save your battery if you leave it plugged in all the time, it will only activate the charger when battery reachees 9X percent rather than constantly trickling. That would be a nice feature.
17:57:35Papricaits fine now
17:58:13 Quit Jungti1234 (Read error: 104 (Connection reset by peer))
17:58:57 Quit mikearthur (Read error: 104 (Connection reset by peer))
17:59:13 Join mikearthur [0] (
18:01:56amiconnlostlogic: A LiIon charger must have battery protection, otherwise using it would be really dangerous
18:02:51lostlogicamiconn: sure, but trickling and holding it at 4.2V is not healthy compared to letting it fall to ~4.05 and then topping up to 4.2.
18:02:57amiconnSince LiIon charging is constant-voltage, battery protection is achieved by precise enough voltage regulation
18:03:24 Quit mikearthur (Read error: 104 (Connection reset by peer))
18:03:29 Join egotrippen [0] (
18:03:34 Join mikearthur [0] (
18:05:53***Alert Mode OFF
18:06:17 Join gtkspert [0] (
18:06:27 Join einhirn [0] (
18:07:45mirakanyone uses ubuntu ?
18:08:39nnickI do
18:11:03 Quit markun (Read error: 113 (No route to host))
18:11:46 Join markun [0] (
18:14:53 Join NicoFR [0] (
18:16:02 Join DT291 [0] (
18:16:02 Quit DreamTactix291 (Read error: 104 (Connection reset by peer))
18:16:28miraknnick: do you manage to compiler the simulator ?
18:17:34nnickDon't have any compilers installed
18:17:41mirakdamn you
18:17:47nnickDo all that on my other box, which is Debian
18:17:55nnickThat works nicely though
18:17:58miraknnick: would you try on ubuntu ?
18:18:09miraknnick: just the simulator
18:18:13mirakwith standart gcc
18:18:18miraknot the cross compiler
18:18:21mirakI have a problem
18:18:25mirakwith time.h
18:18:41nnickAh, I do have gcc.. hang on
18:18:51mirakI got builds error with timespec type
18:19:35linuxstbamiconn: I've been reading those IRC logs, the important facts seem to be that a) the iriver firmware adds 1965 to the value of the RTC year field to get the actual year; and b) valid years according to Iriver are 1980-2064 - i.e. 2015 to 2099 in Rockbox time.
18:20:05linuxstbSo I agree with your idea - if Rockbox uses 1964 as the base year, then leap years should be OK, and the iriver firmware will be happy.
18:20:56nnickmirak: which target?
18:21:48 Quit DrMoos ("CGI:IRC")
18:22:16linuxstbmirak: Did you say that the simulator worked for you in the past?
18:26:10CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
18:26:10*amiconn wonders what lead the iriver engineer to choose *1965* as the base (!!)
18:26:38linuxstbMaybe his/her birthday? :)
18:26:43amiconn(add an 's' where appropriate)
18:27:16egotrippenhey all
18:27:48miraknnick: H300
18:27:57miraklinuxstb: yes it worked.
18:28:08linuxstbWhat changed? Your installation or Rockbox?
18:28:17 Quit gtkspert_ (Read error: 110 (Connection timed out))
18:28:31mirakI think none
18:28:50mirakthe update to breezy happened before I started rockbox I think
18:28:51 Join adamdullenty [0] (
18:29:26egotrippenin a line of code like 'extern void __________,' does the 'extern' mean that it's accessing from another file? getting the value from another file?
18:29:55lostlogicegotrippen: yes, that means that the symbol will be linked from another object.
18:31:12nnickmirak: x11 or sdl?
18:31:56adamdullentyI asked this earlier but didn't see a reply - any clue whether there's anyone focusing on improving the resolution of the radio scanner?
18:32:07adamdullentyie. changing it from 0.1MHz to 0.05MHz like the iRiver firmware
18:32:34adamdullentyor whether it's pegged to be improved at any point
18:33:10nnickargh, I lack a ton of libraries
18:33:50adamdullentyjust out of curiosity... the only thing stopping me getting rid of using the original firmware entirely is the radio is better [currently], has recording , and the battery life is much better. but this should change, just wondering about the radio in rockbox
18:34:43lostlogicadamdullenty: battery life is still better in iRiver firmware? I thought we were about on par now.
18:35:18adamdullentymuch better... mind you I suppose I listen to a lot more FLACs in Rockbox. that's probably it, didn't think of that
18:35:36miraknnick: any
18:35:49adamdullentymy favourite albums, live shows, etc
18:35:57miraknnick: intersting, I maybe miss one lib
18:35:58lostlogicyeah, you've gotta compare same codec to same codec.
18:36:08lostlogicflac will involve much more disk buffering than lossy codecs.
18:36:14adamdullentyyep yep
18:36:21adamdullentythis is true
18:36:48lostlogica lot of people make attribution errors with rockbox because it does so much more, and if you use a lot of those additional features, you will of course pay the cost in battery life.
18:38:03adamdullentyso for mp3 playback and radio listening, on the iriver, rockbox is almost on par with iRiver? how tiny is the difference
18:38:56lostlogicadamdullenty: in my experience, with ogg q7, they are close enough to be hard to tell apart in battery life.
18:38:57adamdullentybtw I am totally in love with rockbox, it turns already the best MP3 player on the market [in my opinion] into... something godlike. it really does.
18:39:00lostlogicnever listened to the radio on either
18:39:15adamdullentyAh great... didnt realise
18:39:57adamdullentyI just know that when I'm using the player with rockbox at work, I'll be getting worried about the battery at about lunch, when it's at about 40% left. with the original firmware I'd not even be worried about it at the end of the day, still like 50% left
18:40:23adamdullentyyet, yeah, i'm not listening to FLACs and as many WAVs on the original firmware
18:40:25 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:40:40lostlogicwow, that's quite a difference... are you on the latest rockbox?
18:40:48amiconnWhy would you want to tune the radio in 0.05 MHz steps?
18:40:59linuxstbadamdullenty: You could try editing apps/recorder/radio.c to change the definition of FREQ_STEP (near the top of the file) from 100000 to 50000. I have no idea if it would work though.
18:41:30adamdullentyI can get much better reception tuning by 0.05 steps in the iRiver firmware
18:41:48amiconnAfaik, there are no terrestrial stations at xx.x5 MHz frequencies in europe (cable is different, but the iriver has no socket for that)
18:42:07 Join eEye [0] (
18:42:15adamdullentyit really helps, when reception is bad [I work pretty much in the metal roof of a warehouse, in an area with crap reception anyway. it seems to help to be able to tune more exactly]
18:42:26adamdullentyradio 1 and 2 have ranges
18:42:31adamdullentylike radio 1 is 97-99MHz
18:42:33linuxstbadamdullenty: I wouldn't believe percentages - the real measure is how long the battery lasts from fully charged to no charge.
18:42:42nnickmirak: what's your error?
18:42:50linuxstbYou should be able to get about 12-13 hours of constant FLAC playback - if you don't do anything else apart from listen.
18:43:19adamdullentylostlogic: I am on the latest rockbox, I think... compiled it from source last night with the colour changer plugin and a couple of patches
18:43:33amiconnlinuxstb: Would be an interesting test on H300
18:43:41adamdullentyhmm really? wouldn't have thought that. i'll maybe do a test sometime
18:43:48amiconnOf course it depends on the bitrate
18:44:13linuxstbMy tests were with very high bitrate FLACs - I think my album averages about 950kbit/s
18:44:18amiconnlinuxstb: What average bitrate were your test files?
18:44:22miraknnick: it complains about variable with unkown type
18:44:29amiconnSame as mine, then
18:44:37*amiconn hopes his H340 arrives soon :)
18:44:52miraknnick: /usr/include/bits/stat.h:71: error: field ‘st_atim’ has incomplete type
18:45:03linuxstbThey were also "-q 8" files which are the most computationally difficult to decode.
18:45:03adamdullentyI'm confused by what you mean by your albums are high bitrate... why would the bitrate be changable, it's all lossless
18:45:15miraknnick: it starts with an error into lcd-x11.c
18:45:27adamdullentyunless... it needs higher bitrate for.. more complicated music? or something? :S
18:45:31linuxstbAll lossless codecs are VBR - the bitrate varies depending on the content
18:45:45nnickmirak: I don't get that now.. I've installed libc6-dev x-dev libx11-dev libxt-dev
18:46:03nnickI do however get "/usr/bin/ld: cannot find -lXmu"
18:46:23mirakI have this packages
18:46:29mirakyou use breezy or dapper ?
18:47:09egotrippeni'm trying to change the 'invert bar' selector to a 'reverse scroll' selector, but i'm running on pretty much no experience
18:47:18egotrippeni found three places i think it could be
18:47:27nnickmirak: dapper
18:47:55linuxstbWhat is a "reverse scroll" ?
18:48:13 Join akaidiot [0] (
18:48:31nnickmirak: I'll try on a breezy box now
18:48:43egotrippeninstead of using the inverse of the colors, it just swaps the FG and BG
18:49:09egotrippeni've got blakc BG and orange-y FG, so the inverse is white and light blue
18:50:23adamdullentyegotrippen: firstly, great that you're a flips fan :) [hope that's it... or that'll make no sense, lol] and secondly: d'you mean when it highlights anything it insists on inverting the colours instead of maybe reversing them? yeah... i use blue on black and it inverts to yellow on white. impossible to read... happy to hear you're trying to fix it
18:51:04adamdullentycool... that'd be excellent
18:51:24egotrippeni pretty much have to do what i can and then implore the gods, though
18:52:00egotrippenin lcd-16bit.c, there's a bit about STYLE_INVERT, but it looks like that's referring to inverting the screen
18:52:24 Join petur [0] (
18:52:53miraknnick: I have a dapper in dual boot, but it was awfull
18:53:09nnickworking fine here
18:53:21egotrippenhm... none of these look quite right
18:54:19 Join SteveL__ [0] (
18:54:57egotrippenin list.c, there's a line 'display->invertscroll(0, selected_line);'
18:55:02egotrippenbut changing it didn't seem to do anything
18:56:50lostlogicamiconn: do you avoid using a jrs and rts in your memcpy / memmove code sharing?
18:56:51adamdullentysomeone must know where this is, i know two or three other people at least have been looking for it
18:57:07adamdullentythe invert code I mean
18:57:09egotrippenyeah? that's good
18:57:27egotrippeni haven't looked through the IRC in awhile, but i asked on the RB forums
18:57:37egotrippenfound enough to start
18:58:35adamdullentyI've just seen people on about it on forums
18:58:39adamdullentyonly found this place last night :)
18:59:44egotrippeni found this: void (*invertscroll) (int x, int y);
18:59:55egotrippenin screen_acces.h
19:00:14egotrippenbut it still doesn't look like a real definition
19:00:31nnickmirak: getting the exact same error on breezy
19:00:58 Quit nnick ("CGI:IRC")
19:01:03 Join webguest81 [0] (
19:01:07 Nick webguest81 is now known as nnick (
19:01:08adamdullentyanybody know where the selection inverter code is? who wrote that part... they might know where it is
19:01:24nnickcan't you just grep for it?
19:01:44egotrippenit should have to include LCD_DEFAULT_BG and LCD_DEFAULT_FG, i think, in order to know what those values are
19:02:31nnicksearch for it
19:02:55nnickit's defined in the firmware/lcd-* fies
19:03:42egotrippenoh, man
19:03:52egotrippeni didn't look far enough down lcd-16bit
19:06:21nnickmirak: working now.. apt-get install xaw3dg-dev libc6-dev x-dev libx11-dev libxt-dev
19:06:26nnickand it should work
19:06:37nnicknot sure x-dev is necessary
19:08:26 Quit linuxstb ("CGI:IRC")
19:10:10 Join edx [0] (
19:11:35egotrippenso in a line like 'drawmode = (DRMODE_SOLID|DRMODE_INVERSEVID);'
19:11:56egotrippenDRMODE_INVERSEVID would be the inverse scroll
19:12:22 Join Kohlrabi [0] (
19:13:01peturquestion: has somebody else already noticed that sometimes the brightness is set incorrectly? The parameter is ok and brightness is ok once you reset it to that value. w_e_i_r_d
19:13:20peturcan't reproduce now, but saw it earlier tonight...
19:13:55peturbingo - just did it again :(
19:13:55 Quit webguest33 ("CGI:IRC (Ping timeout)")
19:15:04 Quit ghode|afk (Read error: 104 (Connection reset by peer))
19:16:07miraknnick: I try
19:16:13egotrippendamn, this is over my head
19:16:32egotrippenback to the forums with this problem
19:16:50nnickegotrippen: you need to not change the drawmode, instead swap fg and bg
19:17:18miraknnick: seems better with this libs ^^
19:17:48egotrippeni see where it defines fg and bg as colors
19:17:58egotrippenbut not where it applies them
19:18:31 Join vektek [0] (
19:18:41 Join ghode|afk [0] (
19:19:33 Quit vektek (Client Quit)
19:19:35egotrippenthere's a bit "fg_pattern = fg_color; bg_pattern = bg_color;"
19:19:48egotrippenshould i change fg_color and bg_color?
19:19:55 Join vektek [0] (
19:20:08nnickswapping them should do it
19:20:15egotrippenok, i'll try it
19:20:15nnickand not using that drawmode up there
19:22:02vektekhas anyone made a wps with a power metre going up instead of sideways?
19:23:21 Part Polo_o
19:23:33egotrippenwon't that just swap the white and light blue with each other, though? i'm not seeing where it takes the inverse of the colors
19:27:26 Join _DangerousDan [0] (
19:28:57egotrippennope, no change at all
19:29:03egotrippenwhite background, light blue text
19:29:27nnickthe swap is in setting the drawmode to inverse
19:29:56nnickI don't really know how that code works
19:30:04 Quit vektek ("CGI:IRC (EOF)")
19:30:55amiconnlostlogic: What do you mean?
19:31:01egotrippendrawmode inverse doesn't seem to be defined anywhere
19:32:02egotrippenhm.... if that change had made white text on a blue background, i would just take out the inverse
19:32:11 Join Philip_0729 [0] (
19:32:28amiconnAnd yes, DRMODE_INVERSEVID should do. No need to swap colours
19:33:15egotrippenah, i think i see
19:33:30 Join vektek [0] (
19:34:36egotrippeni need SOLID+INVERSEVID
19:37:26egotrippentrying it....
19:38:29adamdullentygood luck
19:38:39adamdullentyfingers crossed
19:38:47egotrippennope, got an error compiling
19:39:35egotrippenit said 'inversevid undeclared
19:40:12egotrippenmm, maybe after the first one
19:41:02 Join manhattan [0] (
19:42:17egotrippenif this doesn't work, i gotta go anyway. i'll work on it more tonight
19:43:41egotrippenchanging the drawmode freaks it out
19:43:48 Quit DangerousDan (Read error: 110 (Connection timed out))
19:43:50manhattanI'm thinking of getting an mp3 player. I'm guessing I would want something in the neighborhood of 20+ gigs. I would like the player to be able to play mp3s, FLAC, m4a, ogg... and whatever else popular codecs are used. I have been looking at And I'm curious to know what is a decent player that the rockbox drivers are fairly mature and stable for. Help?
19:44:53 Join linuxstb__ [0] (
19:45:09egotrippeni've got another couple of ideas to try. i think i can get working what i want tonight
19:45:49 Quit linuxstb_ (Read error: 110 (Connection timed out))
19:45:58vektekis there a way of making the power meter go up instead of accross on the wps?
19:46:23 Quit egotrippen ("CGI:IRC")
19:46:49preglowlostlogic: yo, did the profiler work with faad?
19:47:09adamdullentymanhattan: i'm completely in love with my iRiver H340
19:47:20adamdullentywith all its features off the shelf it's completely amazing... with Rockbox it's godlike
19:47:36adamdullentyit'd do any of those codecs I think... not too sure about m4a but the rest definitely
19:47:49preglowif you don't care about colour display, you might get a h1x0 too, which i think is a better player
19:47:50adamdullentyH340 is 40gigs, H320 is 20gigs
19:47:57 Quit mawe (Read error: 110 (Connection timed out))
19:48:01adamdullentyall the other iRivers are meant to be impressive too
19:48:02 Join eagle730 [0] (
19:48:10preglowgot tasty spdif
19:48:11adamdullentyyep... h1x0 have optical out as far as I know?
19:48:14preglowand in
19:48:20eagle730hi, all
19:48:26adamdullentyah fun
19:48:27vektekI have the h120 and it's fantastic
19:48:54manhattanadamdullenty but the 340 is something in the neighborhood of four hundred USD.
19:49:01preglowlinuxstb__: you got any idea what might be the reason this thing refuses to boot retailos if i don't skip the button check? :/
19:49:02miraknnick: I have another error
19:49:06mirakstubs.c: In function ‘rtc_read_datetime’:
19:49:06mirakstubs.c:181: error: ‘time_t’ undeclared (first use in this function)
19:49:27adamdullentymanhattan: tis worth it!!
19:50:08mirak time_t now = time(NULL); is the faulty line
19:50:16manhattanadamdullenty not to say that it isn't... But I'm not sure I can afford it. Beyond that, for four hundred I MUST have m4a support on account of I have a lot of m4a files
19:50:31nnickhrm, I'd have thought time_t was defined in one of the packages I pasted
19:50:32nnickhang on
19:51:19manhattanadamdullenty So you say you see godlike level improvement form the default firmware to rockbox?
19:51:41 Join muesli- [0] (n=muesli_t@
19:51:43 Quit gtkspert ()
19:54:39***Saving seen data "./dancer.seen"
19:54:47peturCan I assume there are several threads started already by the time settings_apply() is called?
19:55:20peturI think there is a race condition somewhere causing some updates to the hardware not to be done
19:55:47petur(maybe it's just the brightness setting itself but I'd doubt that...)
19:57:21nnickmirak: are you sure you have libc6-dev ?
19:57:28eagle730does anybody knows things about the hardware of the iFP Iriver-models?
19:58:11eagle730i need more things than these in the wiki...
19:58:29 Quit muesli__ (Read error: 110 (Connection timed out))
19:59:41 Quit Seed (Nick collision from services.)
19:59:47 Join Seed [0] (
20:00:19eagle730i actually have a "philips hdd050 (gogear)" which contains same cpu, display controller and power-management-chip as the iFP series... and there are a few more similarities...
20:00:55markuneagle730: what else do you need to know?
20:00:57eagle730so i would like to ask, how the chances are, that the ifpport runs on this player...
20:01:49markuneagle730: It there a firmware update available for the device?
20:02:14eagle730it is stored in a hidden folder on the media (HDD)
20:02:40 Quit Rob- (
20:02:48eagle730the saa7750 fetches the update automatically out of this folder after unplugging
20:03:06eagle730as far as i know, not
20:03:34markuntomal is the ifp guy. You should talk to him when he's here.
20:04:00eagle730the firmware is named "Encrypted_FWImage_HDD070_5.2.ebn" which should be encrypted ;)
20:04:39eagle730but the update-program decrypts it before transferring and stores it in another folder on the pc's hdd
20:04:52adamdullentymanhattan: the four features I thought the iRiver lacked out of the box have been added by Rockbox. On-the-fly playlist creation [got addicted to this on my friend's iPod], gapless playback [VERY few players have this], recording direct to WAV from the line in [orig firmware records to MP3 instead, useless for live shows] and playback of FLAC
20:04:59 Quit Seed (Killed by (Nick collision))
20:05:00 Join Seed [0] (
20:05:00NJoinRob- [0] (
20:05:04adamdullentywith those it's perfect. for me anyway.
20:05:05miraknnick: sure
20:05:10 Quit Seed (Nick collision from services.)
20:05:13eagle730so in this temporary folder there's a file "fwimage.ebn"
20:05:16 Join Seed [0] (
20:05:21markuneagle730: what else did you want to know about the hardware that is not in the wiki?
20:05:46eagle730its about the little hardware differences...
20:06:20markunAre all the components on this page the same?
20:06:30adamdullentyand the colour screen, with light blue text on black background, is goooooooooorgeous. goes really well with the player's black casing. ah wow. it's god in mp3 player form. :)
20:07:38eagle730not all...
20:07:50eagle730i looked at the page a few hours ago...
20:08:00markuneagle730: well, just wait for tomal
20:08:06eagle730and compared them with my player's hardware
20:08:29adamdullentymanhattan: i'd guess the H320 [20gig, as you said you'd be interested in] would be about 170 USD less tahn the H340. it's 100 GBP less anyway... £160 from
20:09:25eagle730the differences are the flash-memory (mine has a HDD), the SDram (other manufacturer) and the flash-rom for firmware, but they shouldn't matter, cause the sdram and rom are similar types...
20:09:44eagle730when does he usually comes online?
20:11:01markunHe was here 20 hours ago
20:11:10eagle730hm ^^ k
20:11:30lostlogicamiconn: do you use JSR in the code, or only standard bra jccs?
20:11:40lostlogicpreglow: haven't tried it myself with any codec other than Tremor
20:11:44eagle730ty for the help...
20:11:49 Join uwe_ [0] (
20:12:11amiconnlostlogic: No jsr in memset() / memcpy(), that would be slow. Only bra and bcc
20:12:24amiconn(and rts of course)
20:12:27lostlogicamiconn: and even in merging memmove into memcpy and sharing code, still no jsr?
20:14:16eagle730er ^^ last question... can i use the iFPport under windows? the flasher is a linux c-file... can i simply compile it under win32-c?
20:14:37lostlogicok, I give up, I can't do that with C inlined assembly... Your other objections to moving it into a .c file proper constraints and separated asm volatiles can work around... but not two functions sharing codespace w/o JSR.
20:14:55markuneagle730: the ifpport is very new. I wouldn't try anything yes.
20:15:34markuneagle730: are you a programmer?
20:16:28 Join vger_ [0] (
20:16:48amiconnlostlogic: jsr implies additional rts insn(s), and one rts takes 5 cpu cycles on coldfire
20:17:57 Quit SteveL__ (Remote closed the connection)
20:18:01mirakin int rtc_read_datetime(unsigned char* buf) the line "time_t now = time(NULL);" time and time_t can't be found, though I include time.h. anyone know what's the missing include or package ?
20:18:43 Join SteveL__ [0] (
20:19:28nnickmirak: are you sure you have libc6-dev installed?
20:20:46lostlogicamiconn: I know. Which is why I'm giving up on trying to move those functions into C inline asms... although it may be worth revisiting if gcc ever gets regparm support on m68k.
20:20:46 Quit nnick ("CGI:IRC (EOF)")
20:20:54mirakyes I am sure
20:22:02adamdullentyEditing the definition of FREQ_STEP to 50000 did help, it now steps in 0.05MHz steps
20:22:03mirakI have a k7 kernel, maybe that's a problem ?
20:22:04adamdullentywhich is great
20:22:11mirakI don't see a libc6-dev for K7
20:22:18adamdullentybut it still rounds up to the nearest .1 in the radio screen... anyone know where this might be changable?
20:23:08eagle730markun: i can program a bit... im in the "learning process" ;)
20:23:21 Quit linuxstb__ (Read error: 110 (Connection timed out))
20:23:29 Join linuxstb__ [0] (
20:23:40eagle730markun: why do you ask?
20:23:48markunjust interested
20:24:10manhattandoes the rockbox work on any creative players?
20:25:04adamdullentyhmm dont know
20:25:14eagle730markun: im currently searching an decompiler/disassembler for the arm 720t, the cpu of my player... perhaps the disassembled code could offer some news :P
20:25:45eagle730markun: ida has problems with this type of arm-processors...
20:25:47 Join stamppot [0] (
20:25:56markundoes it?
20:26:14markuncan I have the binary?
20:28:25eagle730yes... it only supports standard arm, arm710a, armb and xscaleB/L
20:28:52eagle730binary of the firmware? do you think its of use for you? ;)
20:29:00linuxstb__preglow: No, I have no ideas about the bootloader problems. The button code in our bootloader should be the same as from ipodloader, which has no problems loading the apple firmware.
20:29:36linuxstb__manhattan: No, Rockbox doesn't work on any creative players.
20:29:39eagle730which files do you need?
20:29:43 Nick linuxstb__ is now known as linuxstb (
20:30:06manhattanont he rockbox site is the iRiver H10 (20GB) anH1x0 or are those differnet models?
20:30:27RickH1x0 = H100 (I think), H120, H140
20:30:32markuneagle730: fwimage.ebn
20:30:46eagle730how shall i send them to you?
20:30:53eagle730do you have icq or sth like this?
20:31:20eagle730i've the webclient for irc ;) so over this its impossible
20:32:06markunyou can email it to marcoen at gmail dot com
20:33:15eagle730do you also need the encrypted ver?
20:34:24eagle730when the fw is on the mp3 player, it has the ending .oke
20:34:33eagle730ill rar it and send it to you
20:34:47 Quit muesli- (Read error: 110 (Connection timed out))
20:35:12linuxstbeagle730: You need to wait until "tomal" is around in IRC - he's the only Rockbox person who knows the details of the ifp7xx, and the sole ifp developer.
20:35:59linuxstbAre you able to work on a port yourself?
20:36:44eagle730i don't know...
20:36:51eagle730i thought about it already...
20:37:03markunlinuxstb: "20:08 < markun> eagle730: well, just wait for tomal" :)
20:37:22linuxstbmarkun: :) I lost my connection then...
20:37:53markunlinuxstb: the linux source was not so useless afterall btw
20:38:03linuxstbThat's good.
20:38:21linuxstbBut have you managed to run your own code yet?
20:38:26markuncontains all we need for a ata driver
20:38:34markunno, still not
20:38:55eagle730the hardware is quite similar and the sourcecode of the ifp code is avaiable... i'll try to work on something like this and publish it for the community, when i'm succesful ;)
20:39:10mirakI have time.h in my includes time_t now = time(NULL);
20:39:16linuxstbAdapting the rockbox ATA driver for the ipod was simply a matter of defining the register addresses and fixing a few endian issues. So it seems very generic.
20:39:18mirakthough this fails
20:39:20markungtkspert now managed to get samba and his usb nic to work. next he'll compile busybox with telnetd
20:39:30eagle730only the flashing methods differ...
20:40:10linuxstbmarkun: Are you still lacking a cross-compiler?
20:40:17linuxstb(for Linux I mean)
20:40:21 Join SteveL___ [0] (
20:41:07eagle730afaik, the hdd050 also bases on i2c and ata... it should (theoretically) work
20:41:11markunlinuxstb: yes, still. gtkspert just built one, I will use the same guide he used.
20:41:23markunAnd now I can use the toshiba linux source.
20:41:39eagle730the email is sended...
20:41:58eagle730what do you want to do with the flashrom-image?
20:43:06eagle730if you success to disassemble it, let me know :P
20:43:16manhattanwhat i "seeking" when talking about codecs?
20:43:21eagle730you should have my email now...
20:43:27markunyes, I have it
20:43:51linuxstbmanhattan: skipping from one part of a file to another
20:45:22manhattanrockbox doesn't list m4a on the codecs list. Does that mean that it's not supported or is AAC (MP4) the same?
20:45:51linuxstbm4a isn't a codec - it's a "container format" that can contain a variety of different codecs.
20:46:18linuxstbBut Rockbox supports both AAC and Apple Lossless (ALAC) in an m4a container.
20:46:22manhattanoh I see... so if I have an m4a file, it may or may not be played?
20:46:34manhattanohhh I understand.
20:47:01manhattanwith regards to codecs realtime means, what; recording?
20:47:04linuxstbBut AAC support in Rockbox is not perfect yet - it doesn't decode fast enough to replay in realtime, so it pauses every couple of seconds.
20:47:22manhattanso relatime means playback?
20:47:47eagle730realtime is the speed, in which it should be heard...
20:48:32manhattanoh okay... So mp4 aren't ready to be played using RockBox yet?
20:48:47linuxstbAAC mp4 isn't, no.
20:48:51mirak the error is at the end ...
20:49:01linuxstbBut ALAC (Apple lossless) works well.
20:49:23linuxstbAnd of course, Rockbox doesn't support DRM.
20:49:56manhattangot it... I have a discography in mp4... What's DRM? :-\
20:50:18eagle730Digital Rights Management...
20:50:19linuxstbDigital Rights Management - i.e. encrypted audio files.
20:50:29eagle730i was faster ;)
20:51:35mirakwhen I do printf '#include <time_t>\ntime_t x;\n' > foo.c; make foo.o this works. That's a small test with a time_t type and a <time.h> include
20:51:37linuxstbeagle730: I'm going to ask you my now standard question - any idea if your phillips player can play back mp2 files (MPEG layer 2) ?
20:51:46eagle730its a method, to encrypt audio files, that you need a "key" for them.... if you buy .wma musik, for example, its the copy-protection... so noone can listen to it on another pc
20:51:47manhattanit would seem one of the sounder options to do, if I plan to use rockbox and I want a 20 gig player, would be to get an iriver h120
20:52:20eagle730why your standard question? :P
20:52:39linuxstbAt the moment yes. Hopefully in the near future, the iaudio x5 and ipods will work with Rockbox as well as the h1x0 and h3x0 do now.
20:52:59eagle730linuxstb: i have no idea....
20:53:15linuxstbeagle730: Because I'm trying to find out if any players support mp2 - so I'm asking everyone that comes here. :)
20:53:27eagle730linuxstb: i can only assume....
20:53:44eagle730do you mean mp2 audio or video layer? or both?
20:53:52linuxstbmp2 audio.
20:54:02manhattanboy those h120s are pretty rare to find.
20:54:02linuxstbAs used in digital tv/radio broadcasts in Europe.
20:54:28eagle730i live in europe ^^ but i never watched such a broadcast :P
20:54:42eagle730i test it.... thats the only way ;)
20:55:02 Quit SteveL__ (Read error: 110 (Connection timed out))
20:55:32manhattanholy smokes they're $300+
20:55:38eagle730but the hardware-decoder (saa7750) supports OFFICIALLY only wma (DRM) and mpg3...
20:56:36 Join tomal [0] (
20:56:52tomalHi! I see you are talking about me.
20:57:33eagle730hi, tomal ^^
20:58:42eagle730apparantly, mine doesn't support mpeg2-audio...
20:58:43preglowyou mean digital satelite uses mpeg2 audio?
20:58:46linuxstbI've uploaded a short (1MB) sample MP2 file here: I would be interested to know if it plays on your philips.
20:58:58linuxstbpreglow: Yes. Generally at 256kbps
20:59:13preglowsmall wonder the picture is so poor
20:59:15preglowmpeg2 everything
20:59:24linuxstbNo, sorry - it's MPEG-1, Layer II
20:59:30preglowahh, yeah, same shit
20:59:44preglowthere is no such thing as mpeg2 layer2 anyway, i believe
20:59:46preglowtomal: sure we are
20:59:46eagle730so it's mpeg-1 Layer II audio?
20:59:47linuxstbWhy do you say the picture is poor?
20:59:53preglowlinuxstb: because it is?
20:59:57linuxstbThat's up to the broadcaster.
21:00:02preglowi can see encoding artifacts everywhere
21:00:10preglowon most channels
21:00:15preglowmy favourite is sports channels
21:00:21preglowwith lots of motion, that is
21:00:28linuxstbThat's because MPEG-2 is optimised for 6Mbit/s, and most channels use around 3Mbit/s-4Mbit/s.
21:00:28adamdullentydigital TV is crap
21:00:49eagle730good ol' analog ^^
21:00:53linuxstbBut it's exactly the same technology as DVDs.
21:01:05adamdullentywhich they're switching off by... 2008, is it? or longer
21:01:07preglowlinuxstb: know, i'm too fond of dvds either
21:01:13preglowlinuxstb: since i usually can see the encoding artifacts...
21:01:22preglowi'm _not_ too fond of dvds...
21:01:30adamdullentybring on blu-ray!
21:01:37eagle730yes :P
21:01:43preglowthey'll mess that up as well
21:01:44preglowbelieve me
21:01:57linuxstbThat will just be high-definition, but also with insufficient bitrate
21:02:02adamdullentylike only using 10gigs and the rest for .... advertising :P
21:02:07preglowlinuxstb: exactly
21:02:19eagle730but could it be, that your decoder produces the artifacts?
21:02:21linuxstbI've seen some HDTV streams from the USA - 1920x1080i @ 12Mbit/s
21:02:24preglowand somehow they managed to include interlaced video in the spec
21:02:34eagle730interlaced? -.-
21:02:36preglowinterlacing needs to be abolished
21:03:02adamdullentyI've watched music videos captured from HDTV
21:03:05adamdullentylooks pretty good to me
21:03:33linuxstbAgain, it depends if the broadcaster has used a high-enough bitrate.
21:03:38manhattanis the iRiver h120 remote essential?
21:03:52linuxstbmanhattan: No.
21:03:58eagle730^^ interlaced is one of the quality problems between analog and digital video... and they want to include this in hdtv?
21:03:59linuxstbBut it's very nice.
21:04:28manhattansomeone is selling an iriver h120 for 190usd but w/o control...
21:04:30adamdullentylinuxtsb: what do the 1920x1080i @ 12Mbit/s look like? [what's the little "i" for?] - i realise this is highly
21:05:00eagle730tomal: i was told, that you are the one here, who develops for the Iriver iFP?
21:05:15tomaleagle730: Yes, I am.
21:05:35linuxstb1080i is 1080 lines interlaced - 60 fields/sec for NTSC countries, 50 fields/sec for PAL. The other HDTV standard is 720p - 60 frames per second.
21:05:56linuxstb(well the 59.97 fps nonsense...)
21:06:32linuxstbadamdullenty: They look good, but you can see definite artifacts due to insufficient bitrate.
21:07:12eagle730tomal: i have the "philips hdd050" and noticed, that the hardware is quite similar to the iFP series until 5xx ... so i thought about the iFP port to work on such a philips player... or with slight changes...
21:07:31manhattanwhy ar the iRive h120 so impossible to find?!?!?
21:07:46linuxstbBecause iriver stopped making them about a year ago
21:07:50adamdullentywould anyone know where the code for the FM radio screen is kept? specifically, the part which rounds the current frequency up to the nearest 0.01MHz?
21:08:05manhattanAnd the h10 is its replacement?
21:08:08tomaleagle730: Yes, I read the logs. However, I work on ifp790 port. It uses PNX0101 chip.
21:08:22linuxstbSame with the h3x0, but they were discontinued more recently.
21:08:30adamdullentyLinusN: radio.c holds the display stuff too?
21:08:45tomaleagle730: SAA7750 is similar in terms of features, but has different harware registers.
21:08:54eagle730tomal: hm :/
21:08:54adamdullentyI just edited the minimum step to 50000 instead of 100000
21:09:10manhattanAnd Rockbox doesn't work on the h10, right?
21:09:22adamdullentywhich works fine [wanted more accurate scanning] but the display rounds the current frequency up to the nearest 0.1MHz
21:09:38eagle730tomal: i'm thinking about working on a port/reverse-engeneering of the firmware
21:09:59eagle730tomal: how do you try to reverse-engineer the 790's firmware?
21:10:20tomaleagle730: LinusN has got SAA7750 datasheets, that should make it possible to create the port
21:10:26eagle730i've problems with IDA, it seems that it doesn't support arm 720t's
21:10:46LinusNadamdullenty: line 781
21:11:09tomaleagle730: I used IDA and it worked quite well, but now I am using my own disassembler/code analyer
21:11:17eagle730LinusN: can you mail me the saa7750 sheets
21:12:27LinusNeagle730: i /msg'd you
21:12:54tomaleagle730: what are your problems with IDA? Can you see any code that makes sense?
21:14:10manhattanI give up!
21:14:34adamdullentywith what?
21:14:37preglowtomal: isn't arm720t a standard armv4 core?
21:14:38 Quit eagle730 ("CGI:IRC (EOF)")
21:14:44 Join eagle730 [0] (
21:14:49adamdullentyactually yeah somebody pay attention to this poor chap :)
21:14:50tomalpreglow: It is.
21:15:04preglowthen i don't understand why ida pro shouldn't support it completely
21:15:05eagle730tomal: had a problem with my connection...
21:15:14LinusNmanhattan: no it doesn't worh on the h10
21:15:14adamdullentyalthough the info you're after is probably on the website
21:15:22manhattani can't seem to find a decent 20gig + player that will let me have all the formats I want the closest thing would seem to be the iRiver h10 but can't get rockbox on that.
21:15:39tomalpreglow: maybe there are endianess problems?
21:15:40LinusNmanhattan: iaudio x5 is a pretty safe bet
21:15:46eagle730tomal: as which filetype do you open the firmware and with which processor type?
21:15:58 Join Moos [0] (
21:16:14adamdullentymanhattan: what's wrong with the H320, for your needs?
21:16:30manhattanit's a little bit out of my price range.
21:16:33manhattanand rare to find
21:16:38adamdullentyI know I'm biased... but I can't see a single thing wrong with it. it's 100 GBP less than the H340
21:16:57adamdullentyhmm, not too rare... I think they still make them..
21:17:01tomaleagle730: I would have to reboot to Windows to check it. I just loaded it as a raw binary file and chose an ARM cpu
21:17:08manhattanno they've stoped making them
21:17:10lostlogicmanhattan: Linus is right.
21:17:11 Join ts-x [0] (
21:17:18adamdullentyhave they??
21:17:28adamdullenty........ why? are they absolutely nuts?
21:17:37manhattanyeah I read somewhere that the h10 20gb verisonis it's replacement
21:17:44tomaleagle730: is the firmware available on the Internet? I could take a look at it
21:17:47lostlogicthey will make U10 based players or something
21:17:58eagle730this would be great...
21:18:10eagle730i would try to make a port for this type of player...
21:18:19eagle730the firmware isn't avaiable...
21:18:28manhattanI didn't see the X5 listed on the rockbox site... That' why it hadn't occurred to me to reasearch it.
21:18:37LinusNmanhattan: i'm working on it
21:18:38manhattanbut it doesn't seem ungodly expenseive
21:18:44eagle730in the internet it's only in an encrypted form with the update-program...
21:18:46lostlogicLinusN: how's progress on that, BTW?
21:18:52ts-xmanhattan: Where are you located? If you're in the US, there are many H320 refurbs on ebay right now for ~$225
21:18:58LinusNlostlogic: slow, but steady ;-)
21:19:08mirakpreglow can you check that error at the end of the file ? I can successfully build a test.c with #include <time.h> time_t x; though
21:19:14lostlogicLinusN: awesome −− can't wait to see what you do with 4 DMAs in your hands :-P
21:19:17eagle730i could email it to you... on the player it's uncrypted in an folder...
21:19:24LinusNlostlogic: haha
21:19:39manhattandamn x5 is pretty slick lQQking
21:19:44eagle730tomal: i could email it to you... on the player it's uncrypted in an folder...
21:19:49lostlogicmanhattan: the only downside really is the low res screen
21:20:04manhattanI wans't planning on watching movies on it, or anything...
21:20:26tomaleagle730: If it's not bigger than 5 MB, then mail it to
21:20:27manhattanI only care about it's music playing capabs.
21:21:10manhattanand to be able to drag and drop files -so i can use it as an external drive form time to time.
21:21:16lostlogic*shrug* I'm just screen picky, I couldn't bring myself to go from the H340 screen to the iAudio screen, despite other nice features of the iAudio.
21:21:25mirakor LinusN ? (is probably hiding :D )
21:21:43peturLinusN: any idea why *sometimes* the brightness doesn't get set on RB load. The settings are OK but it looks like it wasn't passed to hardware.
21:21:48manhattanbut the h3*0 are outrageously priced... and pretty hard/rare to find.
21:22:17lostlogicmanhattan: I just picked up my replacement H340 US on Ebay for $275... used, but *shrug*
21:22:37linuxstbmirak: The error is self-explanatory - 'time_t' is undeclared in stubs.c, line 181
21:22:38peturLinusN: this also happens when using rolo...
21:23:44 Join simprix [0] (
21:24:03miraklinuxstb: yes, but how to fix that is less evident
21:24:30eagle730tomal: the mail is on the way :)
21:24:33mirakLinusN: I include <time.h> . When I do a test file like #include <time.h> time_t x; this works
21:24:46ts-xThe H320's on US ebay right now with remote for $200-$225 are a bad deal either
21:24:50manhattanholy smokes the iaudio plays FLAC out of the box!
21:24:51mirakLinusN: hem sorry, I was speaking linuxstb
21:25:14eagle730tomal: what was this "tool" of yours of which you spoke?
21:25:16adamdullentyice :)
21:25:40 Join muesli__ [0] (n=muesli_t@
21:26:10tomaleagle730: it is a code analyzer, written in Lisp and just for the task of reverse engineering the iFP7xx firmware
21:26:40tomaleagle730: as a result it is quite user unfriendly, a little buggy, but very useful
21:27:00manhattandamn... some of the people in here are some bad asses, huh?
21:27:18linuxstbmirak: It's hard to debug your problems. I would add "#warning here" messages to various include files (both Rockbox and Linux includes) to try and understand what's going on.
21:27:27 Join DangerousDan [0] (
21:27:30eagle730tomal: it could work for the saa7750-based players, too, or not? the code shouldn't differ much, hm?
21:27:56LinusNmirak: try #include "time.h" instead of #include <time.h>
21:28:05miraklinuxstb: I have run a gcc -E to see if the include is happening but it happens.
21:28:10mirakI tried that
21:28:17tomaleagle730: it should work, unless the used C compiler generated a strange code
21:28:30miraksame error
21:28:38linuxstbmirak: Which time.h is it including - the Rockbox one, or /usr/include/time.h ?
21:28:50 Part simprix ("Leaving")
21:28:54tomaleagle730: If I find some more time I will publish the analyzer with a short manual on how to use it
21:28:55miraklinuxstb: I tried #include "/usr/include/time.h"
21:29:00mirakand it doesn't work as well
21:29:09miraklinuxstb: <time.h> include /usr/include/time.h
21:29:13eagle730tomal: can you mail the compiled tool to my mail-address? you should have it from the mail i sent...
21:29:29peturLinusN: any idea? (^^)
21:29:40LinusNnot really
21:30:13eagle730tomal: i wouldn't need a manual, only a list of the commands / commandline-variables...
21:30:38adamdullentywhat's a good code editor which shows you the line number...
21:30:54 Part ts-x
21:31:42peturLinusN: could it be the HDD activity messes up the I2C timing?
21:31:52manhattandoes this look like a safe/smart bet?
21:32:13tomaleagle730: I got your email. However, I suppose the firmware is still encrypted.
21:32:43tomaleagle730: I cannot see any ascii strings.
21:33:00lostlogicpetur: that sounds reasonable. I had some weird I2C quirks when audio was buffering when I was repeatedly reading the I2C to watch teh PCF50606
21:33:00 Quit Xerion (" ")
21:33:32eagle730me neither... this is why i thought, ida or sth else has problems with the arm720t.... hm....
21:33:40tomaleagle730: I cannot also see ARM code. Most of the ARM opcodes has e5, e3 as their most significant byte
21:33:51eagle730tomal: :/
21:34:02eagle730tomal: thats bad... hm....
21:34:09preglowtomal: you've broken the ifp encryption totally, i take it?
21:34:58tomaltomal: Yes I did it. That's why I know specific properties of the ARM opcodes.
21:35:11eagle730tomal: the firmware-update-tool works as follows:
21:35:31tomaleagle730: could you give me URL of the firmware upgrade?
21:35:39eagle730tomal: it reads a file beginning with "encrypted-" from the disk (fwimage)
21:36:11eagle730tomal: it makes sth with it (i thought decryption) and saves it in his temporary folder...
21:36:35eagle730tomal: then it copies this file to the players "_system/firmware/" folder
21:36:58eagle730if the player is disconnected from usb now, it checks the folder for new FW and flashes...
21:37:06miraklinuxstb: when doing gcc -E foo.c | grep __time_t I obtain
21:37:07mirak__extension__ typedef long int __time_t;
21:37:07miraktypedef __time_t time_t;
21:37:25eagle730tomal : until now, i thought it would only flash an encrypted file to his rom :/
21:37:34mirakthis doesn't gappen when I do a gcc -E on stubs.c . "__extension__ typedef long int __time_t;" doesn't exist
21:37:37eagle730tomal: I'll search a link
21:37:59mirakdoes rockbox have some redefine or undef that could break that ?
21:40:40eagle730this should be the link...
21:41:01eagle730tomal: how did you break the encryption, and which tools did you use?
21:43:14 Quit _DangerousDan (Read error: 110 (Connection timed out))
21:43:17tomaleagle730: I looked at differences between similar versions of the firmware and it gave me some hints about the encryption method.
21:43:26linuxstbmirak: Yes, I'm sure the Rockbox Makefiles will change the include path
21:43:46LinusNmirak: include stdio.h after time.h
21:45:00mirakLinusN: no succes
21:45:02eagle730tomal: could the encryption method be similar on saa/pnx based players?
21:45:20mirakwonder why I didn't gave up. I must be sick or something :)
21:45:41adamdullentyLinusN: most of this code around line 781 seems to be about recording from the radio, rather than displaying the current frequency on the screen
21:45:55tomaleagle730: I would rather connect an encryption method with a manufacturer.
21:46:02miraklinuxstb: if I do include "/usr/include/time.h" shouldn't I be sure I use the correct file ?
21:46:13LinusNmirak: check if _TIME_T_DEFINED is defined
21:46:17tomaleagle730: There is no encryption modules in saa/pnx afaik.
21:46:33LinusNadamdullenty: really?
21:47:12LinusNadamdullenty: in my source, line 781 is:
21:47:15LinusN freq = curr_freq / 100000;
21:47:19 Join akaidiota [0] (
21:47:20adamdullentyLinusN: yes although I had to count the lines myself :P could you recommend a decent code editor which shows line numbers?
21:47:29LinusNadamdullenty: emacs
21:47:33adamdullentyLinusN: ah, mine isn't... I could just search for that - thanks
21:47:37adamdullentyokay, cheers
21:47:41linuxstbadamdullenty: Search for LANG_FM_STATION
21:47:54eagle730tomal: so possibly the firmware does the de/encryption?
21:48:22tomaleagle730: That was the case with the ifp790.
21:48:24mirakLinusN: I added a #ifdef _TIME_T_DEFINED #warning #endif and seems it's not defined
21:48:45adamdullentyoh, that's where I was, I just got confused as to where it says how many of the decimal places to show
21:48:48linuxstbadamdullenty: You will probably also want to edit the definition of LANG_FM_STATION in apps/lang/english.lang to be %d.%02dMHz
21:48:49eagle730tomal: than it'll be the same thing with the hdd050, i think...
21:49:08adamdullentyLinusN: Cheers, thats the kind of thing I was looking for :)
21:49:17tomaleagle730: I also think so.
21:49:56linuxstbadamdullenty: Change line 781 to be freq = curr_freq / 10000; and then change the two instances of "10" in the next line to be "100"
21:50:11linuxstb(I think...)
21:50:25eagle730tomal: in the fw-updater folder exists a dll called "firmware.dll"... do you think it could have sth to do with the crypting?
21:50:28adamdullentylinuxstb: not quite sure what that's doing but okay, ta
21:51:06adamdullentylinuxstb: ... i think it makes sense
21:52:43mirakok I have put a warning inside /usr/include/time.h, and it doesn't happear on rockbox make, however it happen on the little test programm
21:52:51 Quit Febs ("CGI:IRC")
21:52:53miraklinuxstb: the path are changed
21:53:14tomaleagle730: if Philips wanted to keep the code secret, then they would rather not put the decrypting code somewhere outside the player
21:53:44tomaleagle730: so I suppose there is no decryption code in dll.
21:53:58 Join Mmmm [0] (
21:54:43***Saving seen data "./dancer.seen"
21:55:32eagle730tomal: hm.....
21:55:52 Join Xerion [0] (
21:55:55adamdullentyIn the /* Tune in and delay */ section, the sleep(1), is 1 pausing 1 second every time the frequency is changed? and would reducing this value make it scan through frequencies quicker?
21:56:22eagle730tomal: if sth like this exist, we would need a program to read an flash-rom over usb... ;)
21:57:04linuxstbadamdullenty: "1" is 1/100th of a second. That's the minimum period Rockbox can sleep for.
21:57:48adamdullentylinuxtsb: ah okay. what's it there for?
21:57:59linuxstbI guess it's just to give the tuner time to settle
21:58:12adamdullentyah suppose so. how would I make it scan through frequencies a lot quicker?
21:58:24adamdullentyor is this one of those "Go and find out yourself" questions :P
21:58:35peturtake bigger steps :D
21:58:38tomaleagle730: reading a flash rom over USB may require cracking into a player
21:58:39linuxstbYou can't - unless you scan in larger steps
21:58:44adamdullentyah ok
21:59:21adamdullentyi'm thinking of a scan-in-bigger-steps-whilst-holding-down-another-button kind of thing maybe...
21:59:44tomaleagle730: you know, it is like with a computer connected to the Internet, you can't read some information unless you crack to it
22:00:21 Join solexx [0] (
22:00:53tomaleagle730: however, it may be possible to read it by connecting directly to the chips
22:01:12 Quit adamdullenty ("CGI:IRC (EOF)")
22:01:17eagle730tomal: it was a joke ;) if it would be that simple, like reading a pc bios, all the guys out there wouldn't sit for hours over decryption...
22:01:41 Quit akaidiot (Connection timed out)
22:01:54eagle730tomal: theoretically it has to be possible to connect to them... but with which software?
22:02:00miraklinuxstb: when I include #include "/usr/include/time.h" it include the right file but after that the error I have is stubs.c:187: error: dereferencing pointer to incomplete type
22:02:14mirakprobably the include went wrong
22:02:25mirakI don't see why I am have this problem and nobody else
22:02:25eagle730tomal: with this software you must be able to send and receive single custom commands...
22:02:54eagle730tomal: or we have to think about coding an own tool for this ;)
22:03:22peturlostlogic, LinusN: did a simple hack of readback to verify and rewrite if different. guess what: it works now. Something must be messing up the I2C :(
22:03:49 Join adamdullenty [0] (
22:03:52tomaleagle730: I thought about connecting some electronic equipment directly to the chips and reading their contents.
22:04:02peturI guess this isn't accepted as good solution eh?
22:04:21tomaleagle730: I do not know about it though.
22:04:53lostlogicLinusN: think there might be an issue with interrupts during I2C r/w truncating a write or read cycle badly?
22:05:07linuxstbmirak: So "struct tm" is not defined now?
22:05:55LinusNlostlogic: i2c operations to the pcf50606 should be interrupt protected
22:06:02eagle730tomal: if you would unsolder them, you could put them in an chip-flasher/reader ;) but it would be hard to do this without damaging the player ;)
22:06:20peturLinusN: what about DMA, can that be it?
22:06:29 Join menollo [0] (
22:06:41LinusNlostlogic: you need a set_irq_level(MAX_IRQ_LEVEL) before the i2c operation
22:06:44eagle730tomal: but in the datasheet are the commands for writing/reading/erasing the rom...
22:07:10lostlogicLinusN: I assume the brightness setting code does that?
22:07:15lostlogicor is that the problem?
22:07:22eagle730tomal: if you could send them over usb, the decryption problem would be gone ;)
22:07:26LinusNthe brightness setting should do it
22:07:35peturno irq calls in pcf50606.c
22:08:22peturwill try... and what do I set it afterwards?
22:08:37eagle730tomal: i've to leave soon... we could write an email to the other, if there is progress...
22:08:46LinusNint oldlevel = set_irq_level(MAX_IRQ_LEVEL);
22:08:54LinusNbla bla;
22:08:59peturyep, ok
22:09:14peturyes yes
22:09:31miraklinuxstb: I don't know, I don't know what means the error
22:09:32peturdidn't know set returned the old value
22:09:33tomaleagle730: I will have not much time for Rockbox in the following days
22:09:52tomaleagle730: I will rather work on an iFP prt
22:10:38tomaleagle730: in the meantime you can try to compare similar versions of the firmware
22:10:52eagle730tomal: ok...
22:11:00tomaleagle730: as with iFP790 it may give some hints
22:11:28tomaleagle730: read the wiki pages about ifp and ihp cryptanalysis
22:11:55eagle730tomal: i'll read the saa7750 datasheets etc... eventually there are useful hints for the reading of fw files...
22:12:12mirakLinusN: and I still have warning: implicit declaration of function `time'
22:12:25linuxstbeagle730: Maybe some other projects have decrypted philips firmware for other kinds of devices. It could be worth a quick google.
22:12:37 Quit solexx_ (Read error: 110 (Connection timed out))
22:13:21tomaleagle730: I guess there are no commands for reading an unencrypted firmware over USB, but maybe it would be possible to overflow some buffers and execute your own code :)
22:13:35eagle730linuxstb: i searched for sth like this... but apparently nobody (or at least i couldn't find somebody) did ever an dcryption of this or modified the firmware...
22:13:42eagle730tomal: :)
22:14:08LinusNmirak: that sucks
22:14:19eagle730tomal: possibly ^^ the code has then to be in arm720t format?
22:14:43mirakLinusN: a lot
22:15:11LinusNmirak: this was ubuntu, right?
22:15:14mirakLinusN: in thoery what should happen ? the time.h used should be the one of rockbox or the one of linux ?
22:15:19eagle730tomal: but how to overflow a buffer of an mp3 player? with gigantic fragmented mp3s? ;)
22:15:33LinusNmirak: i believe it should be linux
22:15:38mirakLinusN: yes, and nnick tried on the same ubuntu breezy and it works for him
22:15:52LinusNeagle730: either that, or bad id3 tags
22:15:57mirakLinusN: the make file default to rockbox time.h
22:16:31mirakLinusN: I mean I have put a warning inside /usr/include/time.h and it's not called by make
22:17:18peturLinusN, lostlogic: that seems to work out OK, will post a patch. btw, what a funny define (HIGHEST_IRQ_LEVEL) :D
22:17:41eagle730tomal: hm... the saa7750 has an input for "remote control"
22:17:45LinusNmirak: my sim also includes rockbox time.h
22:19:03XavierGrMy cousin is considering buying an H300. What are the differences between the H300 port and the H100. (I mean what is lacking from the H300 compared to H100)
22:19:10eagle730tomal: if we knew the syntax of this, it could possibly execute commands in the chip, like firmware reading.... with a bit programm code, it could be possible to let the player save the firmware as a file on hdd
22:19:15lostlogicpetur: cool −− when I'm doing pcf50606 work tonight or tomorrow I'll keep this in mind.
22:20:27tomaleagle730: isn't it just for a remote control like in case of a TV set?
22:21:10adamdullentyso eng: "Station: %d.%02dMHz" is right, for wanting it to show 0.05MHz increments in the frequency... ?
22:21:13 Join ashridah [0] (
22:21:16adamdullentyin the lang file
22:21:20eagle730tomal: thats the question...
22:22:11eagle730tomal: if its only a simple electronic which sends predefined commands (next track, play, pause) this is nonsense...
22:22:35eagle730tomal: but when it can send commands to the chip it would be an interesting point ^^
22:22:42linuxstbadamdullenty: Yes - plus the other changes I mentioned (100000 changes to 10000 and 10 changes to 100)
22:23:21adamdullentychanged that too... what I ended up with was that it just went 91.01, 91.02, ..., 91.09, then 92.00, 92.01, etc
22:25:26linuxstbMmm. 91.05MHz is 91050000Hz. So freq (curr_freq / 10000)=9105, so freq/100 is 91, and freq % 100 is 5 - so it should display 91.05
22:25:52eagle730tomal: i'll leave in 5 minutes....
22:27:17adamdullentyoops wrong pc
22:27:21lostlogicLinusN: what do you think of having defines for what type of power supply chip a unit has instead of the HAVE_CHARGING and HAVE_CHARGE_CTRL defines?
22:28:25peturLinusN: I just gave you some work :D
22:28:42eagle730tomal: :P trying to mail philips technical support would be funny ^^ i wonder what they'll say ;)
22:28:48MmmmLinus: I've been mucking about with metronome.c a bit and made a small bugfix (you could go above 400bpm)
22:29:18 Join Mark_ [0] (
22:29:22eagle730tomal: they won't give the decrypted version, because all the programm code of this is used in the new philips products as well ;)
22:30:13 Join YouCeyE [0] (
22:30:38eagle730tomal: see you ;) i leave now
22:30:39 Quit eagle730 ("CGI:IRC")
22:30:45tomaleagle730: see you
22:31:45LinusNMmmm: i think the BUTTON_REPEAT cases should go 1 step
22:32:16LinusNlostlogic: sure, why not?
22:32:21 Join alphakiller [0] (n=bohas@
22:32:29alphakillerjust a dumb question
22:33:03lostlogicLinusN: cool −− that's my mission then, to in some relatively clean way work LiPoly charge control into things, including settings for topping charge threshhold and stuff.
22:33:04alphakillerwhich function is loaded at the rockbox kickstart
22:33:15MmmmLinus: 1 step all the way? yeah I suppose that would be ok, although from 0 to 400 is a long way! :)
22:33:42linuxstbalphakiller: main() in apps/main.c I believe
22:34:17linuxstbYou may also want to look at firmware/crt0.S but it's a bit of a #if hell.
22:34:36peturcrt0 is normally the start of everything ;)
22:34:41MmmmLinus: also the MET_PLAY_STOP on line 895 doesnt work! (try playing a song and then starting metronome) if you replace this line with rb->pcm_play_stop() it works fine but I get the feeling i'd be told off for hacking if I did that :) Must be something to do with the way MET_PLAY_STOP is defined or something!
22:35:22mirakLinusN: linuxstb it works !!
22:35:36MmmmLinus:Would it be ok to put a few #ifs in to fix this or definitely not?
22:36:13linuxstbmirak: Congratulations. What did you have to do?
22:36:57preglowalphakiller: as always, start: in crt0.S is called first
22:37:23preglowwhich in turn calls main(), yes
22:38:03miraklinuxstb: I redownloaded the sim module and remove some code in stat.h
22:38:20amiconnHmm, I'm thinking about what the little plugin binary header should contain
22:38:59amiconnAt least some magic, start address, end address and entry point like the rockboy overlay on archos does
22:39:26linuxstbAnd a version...
22:39:29amiconnWould it make sense to include the target id & plugin api version?
22:39:36linuxstbAh, "magic"
22:39:57amiconnI'm also undecided about a good magic number
22:40:01linuxstbYou can use the model name (or number) that's used in the bootloader
22:40:10linuxstbAt least for iriver and ipod targets
22:40:26amiconnToday, _most_ plugins are different per target, but not all
22:41:10amiconnEspecially for archos recorder v1, v2 and fm recorder most plugins are binary identical
22:41:41amiconnIntroducing the target id would make them unique, making the windows installer bigger
22:41:58amiconnA propos windows installer... LinusN?...
22:42:55XavierGramiconn: Did you recieved your H300?
22:43:00amiconnNo, not yet
22:43:17LinusNamiconn: what about it?
22:43:31lostlogicamiconn: dude, you ordered that thing way before I ordered my replacement, and mine is waiting for me at UPS today.
22:43:47amiconnQuite empty...
22:43:57XavierGrLinus: You that have both. Does the upgrade (from H100 to H300) worths the money?
22:44:03LinusNamiconn: ouch
22:44:13 Part Philip_0729
22:44:21 Part tomal
22:44:24amiconnThey are still not working...
22:44:27LinusNXavierGr: i wouldn't "upgrade" to the h300 from h100
22:44:37lostlogicXavierGr: is a cross-grade, not an upgrade... lose spdif out, get larger size, get USBOTG, get color screen...
22:44:44preglowh1x0 is better than h3x0, if you ask me
22:45:02LinusNit all depends on what you want
22:45:04XavierGrwow preglow, all that for the spdif?
22:45:10preglowall what?
22:45:25preglowi prefer the h1x0 screen to the bad h3x0 one
22:45:27XavierGrH100 better than H300 I mean.
22:45:37preglowand then there's spdif
22:45:40LinusNi also prefer h100
22:45:46preglowplus, its slightly better looking ;)
22:45:53XavierGrIs this true? Bad Colour LCD?
22:46:05LinusNthe color lcd on the h300 rocks imho
22:46:06XavierGrI thought that H100 had a terrible screen too.
22:46:08lostlogicshrug, I like it fine, it's better than X5, right Linus?
22:46:11amiconnlostlogic: Last but not least: get front buttons :)
22:46:24preglowoh? the one i saw wasn't too clear
22:46:27preglowperhaps it was the condition
22:46:32preglowlighting condition, that is
22:46:40lostlogicamiconn: the button layouts each have their up and downsides... so I can't say one is better or worse.
22:46:45preglowscreen on my nano seems tons better
22:46:45XavierGramiconn: That's bad isn't it?
22:46:53*amiconn would like h300 button layout with h100 lcd
22:47:09preglowgo ipod! all front buttons!
22:47:24preglowi'd make an excellent propaganda minister
22:47:25XavierGrOSBOTG is a temtive feature...
22:47:30LinusNimho, the h300 buttons are far too hard to push, too much resistance
22:47:32*lostlogic returns to pondering another DAP IPV or gigabeat
22:47:35XavierGrfanboy ;)
22:47:43XavierGr(that was for preglow)
22:47:52lostlogicXavierGr: remember usbotg only works right on the international hardware, unless you mod the player
22:47:57darklessbit-perfect optical out... H1x0 beats them all... at least I haven't seen anything else sporting that feature
22:48:08amiconnpreglow: iPod != front buttons
22:48:13LinusNi like having a joystick, but the h100 joystick is too slippery for me
22:48:27preglowyeah, me too
22:48:32LinusNthe x5 joy is much nicer imho
22:48:41XavierGrLinusN: You can mod it ;)
22:48:44preglowit happens quite often that my thumb just slides off
22:48:55LinusNi have put hot glue on mine
22:48:58amiconnBtw, is it just me, or does h1x0 rockbox sometimes not react on the first press of Stop in wps?
22:49:14XavierGrSo you did eventually mod it.
22:49:16amiconnToday I had to press 3 or 4 times until playback stopped...
22:49:20LinusNamiconn: i haven't noticed
22:49:24preglowme neither
22:49:36lostlogicamiconn: is it a slightly long press, but not long enough to be a LONG press?
22:49:57amiconnHmm, maybe. Did you notice the same?
22:50:03 Quit ghode|afk ()
22:50:27XavierGr this is the insternationla version except from the cradle right?
22:50:39amiconnThe buttons have bad 'feedback' whether they are actually pressed
22:50:41lostlogicamiconn: on H3x0, I noticed that a just-too-long stop press did nothing.
22:52:00LinusNmaybe BUTTON_STOP|BUTTON_REPEAT is incorrectly handled?
22:52:23amiconnLinusN: I tried the X5 in a shop. The joystick is at the right side of the front plate iirc?
22:52:32LinusNi believe the music stops when you release the button
22:52:34mirakwith gdb, how do I put a break point in a particular line of a file.c ?
22:52:52LinusNand maybe STOP_PRE isn't correct
22:52:57Bagderb file.c:1234
22:52:58LinusNamiconn: yes
22:53:07peturWTF? Did a CVS update and now the build is using 4.0.2 again??? Must still be hiding somewhere :(
22:53:11XavierGrso lostlogic what happend to your yesterday's patch? Is it a no go?
22:53:23markunlostlogic: what's DAP IPV?
22:53:23preglowbah, i've completely forgotten about the eq
22:53:25lostlogicXavierGr: yeah, not worth it
22:53:30amiconnImho this is a strange design. I would expect the joystick in the center or at the left side of the front plate. The way it is, the joystick is cumbersome to handle if I hold the X5 in the right hand
22:53:33preglowsomeone want to finish it for me? :/
22:53:38lostlogicDigital Audio Player iPod Video
22:54:13LinusNamiconn: i agree
22:54:41 Join leftright [0] (
22:56:31leftrightamicconn: I have also noted that the stop button has to be pressed just right or it doesn't respond
22:57:13XavierGrDoes anyone has the cvsdo binarys?
22:57:18leftrightamiconn even
22:57:25XavierGrI use the devkit and I cant install it.
22:57:31 Join fobbo [0] (
22:57:48Papricawhat is the max value of integer?
22:58:07Bagder2^32-1 if unsigned
22:58:09preglowwhich integer?
22:58:32XavierGrDoes anyone here use the CYGWIN enviroment?
22:58:48amiconnIidos WPS_EXIT_PRE is unnecessary for iriver. We don't have an event associated with long press of Stop
22:58:57Paprica//say $calc(2^32-1)
22:59:12preglowjust ordinary int?
22:59:24lostlogic /exec -o echo "2^31-1"|bc -l (signed integer)
22:59:28preglowdoesn't work :P
22:59:32LinusNamiconn: no, but someone could perhaps exit a menu with STOP, right?
22:59:45amiconnYes, and?
22:59:45Paprica//say $calc(2^32-1)
22:59:51amiconnPaprica: The max. value of int is INT_MAX
23:00:07amiconnIt's _not_ a fixed value, it depends on the target architecture
23:00:09preglowPaprica: i use irssi
23:00:14amiconn(16, 32 or 64 bit)
23:00:17 Quit YouCeyE ("Leaving")
23:00:18Papricahuh ok
23:00:20lostlogicpreglow: then use mine ;)
23:00:29Bagder(not many ints are 64 bit...)
23:00:29 Join YouCeyE [0] (
23:00:29LinusNthen a stray BUTTON_OFF release event may exit the wps
23:00:39preglowBagder: sparc and mips, i think
23:00:44preglowthose are the ones i think have got a 64 bit int
23:00:50amiconnThe BUTTON_OFF is taken by the menu exit
23:00:53preglowsparc does, no?
23:00:56Bagderpreglow: no
23:01:01Bagderthey all use 32 bit ints
23:01:02amiconnwps will only see BUTTON_OFF|BUTTON_REL
23:01:17LinusNexactly my point
23:01:21amiconnThat's no problem if WPS_EXIT is triggered on a press
23:01:23preglowthen what did the commotion about x86-64 having a 32 bit int come from?
23:01:54amiconnLinusN: Who says that this has to stay? On recorder we have WPS_EXIT BUTTON_OFF
23:02:03amiconn...and no _PRE
23:02:32BHSPitLappypreglow: what's up
23:02:33LinusNhmmm, that's strange
23:02:39preglowBHSPitLappy: what what?
23:02:43preglownot much of anything
23:03:08amiconnpreglow: No idea, but I didn't say the int width maps directly to the machine bit width
23:03:14BHSPitLappyanything interesting in the ipod-rockbox world today
23:03:35LinusNamiconn: you're right, we should change WPS_EXIT to BUTTON_OFF and remove WPS_EXIT_PRE
23:03:46amiconnAfaik, for _some_ 64 bit archs int is 64 bit (IA64?), and for 8 bit archs int is 16 bit afaik
23:04:03alphakillerheheh ... on mips int is 32 bits
23:04:07alphakilleronly long is 64
23:04:11Bagderwell I said "(not many ints are 64 bit...)"
23:04:12alphakiller(mips 64 bits)
23:04:24 Quit akaidiota (Connection timed out)
23:04:26Bagderit is?
23:04:33lostlogicLinusN: did you end up soldering the BDM connector onto the X5 board somehow?
23:04:47amiconnalphakiller: Yeah, but mips was extended from 32 to 64 bit as was x86-64, right?
23:04:56Bagderchecking size of time_t... 4
23:05:00Bagderon MIPS 64bit
23:05:02amiconnOtoh, IA64 was developed as 64 bit from the beginning
23:05:18Bagderie it is 32bit
23:05:21LinusNlostlogic: i haven't done that yet
23:05:41lostlogicLinusN: not going to be necessary?
23:05:42LinusNlostlogic: i think i'll wait until i really need to
23:05:54LinusNthe x5 has a nice brick-protection
23:06:06LinusNthe bootloader is always there
23:06:11lostlogicah, nie!
23:06:12preglowso, like ipod
23:06:24preglowgod knows how many times that's saved me, already
23:06:27lostlogicpreglow: darned fanboy
23:06:59preglowLinusN: bootloader can go usb?
23:07:39amiconnHmm, it seems the irivers are the only targets which have no brick protection
23:07:41LinusNpetur: you're a windows fanboy
23:08:07 Quit godzirra (Read error: 104 (Connection reset by peer))
23:08:16LinusNpreglow: yes, and it does the flash update as well
23:08:27preglowvery nice thing to have, that
23:08:45preglowit allows fools like me to do bootloader work :>
23:09:01preglowsecondary bootloader work, at least
23:09:20LinusNunfortunately, this means that the boot will be quite long on the x5
23:09:25lostlogicmarkun: does gigabeat have brick protection like this?
23:09:37LinusNpetur: (1) - your patch contains ^M
23:09:55peturforgot to dos2unix it
23:10:00LinusN(1) the variable is called iOldIrqLevel
23:10:09*Bagder considers offering his
23:10:22preglowuse a proper editor
23:10:30 Join mirak_ [0] (
23:10:33*lostlogic waves his vim around
23:10:35Bagderit doesn't always help
23:10:37LinusNpetur: read docs/CONTRIBUTING
23:10:43Bagderpeople use C:\blabla\ a times
23:11:13*preglow hugs his vim
23:11:38mirak_ I have an error at line 46 of the Makefile : Makefile:46: *** missing separator. Stop.
23:12:37MmmmLinusN: Here's a patch that fixes all the bugs! Hope that's ok.. Im going snoozing now, Night...:)
23:13:12lostlogicman... one patch... fixes _all_ the bugs! ... ... no more rockbox work :(
23:13:16mirak_I see no error in that makefile though
23:13:19 Quit mirak (Connection timed out) the metronome.... heh heh...
23:14:42LinusNMmmm: what do you think about going in steps of 1 instead of 10?
23:14:43*petur studies the contributing doc while he rebuilds his build environment he just screwed up
23:15:24XavierGrahh notepad2 is a nice little editor. Highlight capabilities, tab to space line markers, whitespace-tab highlighting and even those linux carriage returns!
23:15:48XavierGrNow all my patches will be rockbox contribute-compatible
23:16:44peturlucky you...
23:16:51leftrightamiconn, How is the wakeup feature progressing
23:16:53MmmmLinusN: not sure, I think it may be a bit of a pain going 400 in steps of 1! have to try it to know though eh?
23:18:04markunlostlogic: I don't think there is a seperate bootrom
23:18:32lostlogicmarkun: but you've avoided bricking yours so far how?
23:19:30alphakillernext week I'll see if I can start working on the COP stuff
23:19:31markunyes, I've not writen anything to flash so far
23:19:38mirak_what's wrong in that Makefile at line 46 ?
23:19:46mirak_ Makefile:46: *** missing separator. Stop.
23:19:50alphakillerI'm very interested on getting all this stuff working nicely under iPod :D
23:19:54mirak_I see nothing schoking
23:19:55alphakillerfor now I'm just playing with my demos
23:20:00alphakillerand trying to understand the interface
23:20:10LinusNmirak_: is it really that makefile
23:20:15Bagderwrong makefile
23:21:03Bagderuse make −−print-directory
23:21:23Bagderto see in which dir it fails
23:21:45mirak_LinusN: I have realised that probably not
23:24:11mirak_I will go through all the silly errors
23:24:23 Join RotAtoR [0] (
23:25:47 Quit Mmmm ()
23:26:38 Quit menollo ("CGI:IRC (EOF)")
23:26:41 Join webguest02 [0] (
23:31:27mirak_make[3]: dlltool: Command not found
23:31:30mirak_what is that ?
23:31:34mirak_dlltool ?
23:31:45Bagderpart of libtool iirc
23:31:45 Part leftright
23:31:49Bagderfor cygwin
23:32:11Bagderhm no
23:32:28BagderI have a i586-mingw32msvc-dlltool in my win32 cross-compiler
23:33:07mirak_well I just try to build the sim for X11 on linux
23:33:27Bagderand which makefile said that error?
23:34:24mirak_the one I did for xvid codec
23:34:33mirak_It's copied on rockboy makefile though
23:34:45 Quit vger_ ("[BX] Get your free warez from!")
23:35:17mirak_maybe it have been updated on cvs
23:35:32Bagderwe've used dlltool for ages
23:35:36Bagderfor win
23:36:19Bagderlike in rockboy
23:36:31Bagderso you're trying to use a windows-only section on linux
23:37:17 Quit SteveL___ (Connection reset by peer)
23:37:20 Join Gibbed [0] (
23:37:23 Quit Rick (Nick collision from services.)
23:37:32 Nick Gibbed is now known as Rick (
23:37:50mirak_ok I don't have sdl make code
23:38:31Bagderthis is not related to sdl
23:40:02 Join _DangerousDan [0] (
23:41:55 Join Midgey34 [0] (
23:44:26mirak_Bagder: the make file for my plugins didn't contained sdl build code
23:44:47 Quit _DangerousDan (Read error: 104 (Connection reset by peer))
23:44:57Bagderok, so when not x11 it assumed win32
23:45:13 Join linuxstb_ [0] (
23:45:52linuxstb_mirak: I found that Makefile problem yesterday - do you want my fix?
23:46:30 Quit DangerousDan (Read error: 110 (Connection timed out))
23:47:10XavierGrmirak what are you up to?
23:47:21XavierGrI mean what are you working now?
23:49:53mirak_Bagder: break points doesn't work for the libs
23:50:00mirak_at least I don't know how to do
23:50:09 Join Janco [0] (
23:51:50mirak_how to use gdb break points with dynamically loaded libraries ?
23:53:04linuxstb_I don't know, but you can use DEBUGF in your plugin now it works in the sim
23:53:22linuxstb_(it works like printf and prints to the console)
23:54:44***Saving seen data "./dancer.seen"
23:54:46mirak_this time it worked
23:54:53mirak_it failed for plugin_main
23:55:01 Quit NicoFR ()
23:55:01mirak_and sample4 the enter point
23:55:56*preglow vanishes
23:58:35mirak_the sim strictly respect the memory layout of rockbox on the target hardware ?
23:58:53Bagderoh no

Previous day | Next day