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 2005-11-13

00:02:01XShocKnot expensive at all.
00:02:23XShocKactually to be exact, making wakeup will be much harder.
00:02:49lokkianyhow, a calendar and a clock
00:02:55lokkithats just amazing
00:03:11XShocKjust regular clock+calendar is just an RTC chip, capacitor(you can actually ommit it too), resistor. actually even send free samples of their RTCs.
00:04:49XShocKto make wakeup you will also need progmrammable counter, another chip.
00:05:10lokkiwhich will probably not fit into the case
00:05:43XShocKnot really. rtc is a 4 pin, 2 by 2 mm chip. counter is also pretty mauch the same size.
00:05:51XShocKrtc is a 8 pin ...
00:06:31XShocKyou also need a 32.768khz quartz
00:09:30amiconnXShocK: Many rtc chips have an alarm function
00:11:07XShocKtrue. ijust used the one that don't :)
00:11:58XShocKsince i don't really need alarm,but will probably alter it to have it anyway, will change chip..
00:16:50 Quit lokki ("CGI:IRC")
00:16:53XavierGrand how the clock stores the time, when the unit shuts down?
00:17:15XavierGrDoesn't it need a battery?
00:18:04XavierGror it will draw power for its need from the main battery?
00:18:29XavierGrand what current consumption does it have, if it draws power from mains?
00:18:31XShocKyes. it draws power from main battery
00:19:15XShocKwhen main 3.3 volts are disconnected, it switch directly to battery and turns into lower consumption mode.
00:19:20XavierGrI mean, will it have a major difference in battery consumption?
00:19:31XavierGrwhich is?
00:19:59XShocKi actually couldn't measure power, my ,multimeter somehow measured 0. :) while i crearly saw a since wave on the quartz which meant it was working.
00:20:15XavierGrit must be lower than 1-2mA to say that it will not make a difference
00:20:27XavierGrI guess I am not sure
00:20:36XavierGranyway have to eat now
00:21:03XavierGrPlaese make a wiki tutorial on how to do it, I am very interested into this. Please.....
00:21:14XShocKit is 100microampere
00:21:53XavierGrthen great
00:22:27XShocKhmm.. it is actually even less. 800-1200nA
00:22:44 Join arkascha_ [0] (
00:23:06XShocK100microA is a standby current. 800-1200nA is on BACKUP battery
00:27:27XShocK0.0008 mA looks really scary. :)
00:28:44amiconnThe RTC chip used in the archos recorders has a backup battery current of < 500nA
00:31:53XShocKI liked that Tektronix TDS2024 osciloscope so much...
00:32:53amiconnlinuxstb: This rrockbox devkit uses a stone-age sh-elf-gcc ...
00:33:01 Quit _arkascha (Read error: 110 (Connection timed out))
00:36:05linuxstb_amiconn: Yes, but the ipod_fw.c errors are caused by the native compiler.
00:36:39linuxstb_I've emailed bluechip. I don't know if he actively maintains his devkit.
00:37:25amiconnCurrent native gcc on cygwin is 3.4.4. Currently recommended sh-elf-gcc is 3.3.6
00:39:08 Quit ashridah ("gone")
00:39:08amiconnWhat's the recommended arm-elf-gcc?
00:39:11 Join ghode|afk [0] (
00:40:15linuxstbI'm using 3.4.4
00:41:43linuxstbIPL still use 2.95 for the kernel, so I have no idea if 3.4.4 is the best choice. I just decided to try the newest version (ignoring gcc4)
00:42:07amiconnYeah, rockbox code isn't really up to be build under gcc 4
00:42:18amiconnLoads of signedness warnings...
00:44:54 Quit ender` (Read error: 113 (No route to host))
00:54:51 Quit arkascha_ (Remote closed the connection)
00:57:53***Saving seen data "./dancer.seen"
01:14:40 Join tucoz [0] (
01:16:18 Quit tucoz (Client Quit)
01:16:20 Join tucoz [0] (
01:16:59tucozHi, I had not updated rockbox for some time and just notice that the shut down splash is now not gray scale anymore? is this intended?
01:18:43 Quit tucoz (Client Quit)
01:20:42 Join DJDD [0] (
01:21:22 Join tucoz [0] (
01:23:53tucozI just thought that the amount of commits made recently may have caused this.
01:29:14tucozAnyway, it is all the splashes. I just thought I mention it in case it is an unwanted behaviour.
01:32:28linuxstb_tucoz: Yes, I think I caused that with one of my commits. It was not intended.
01:34:54tucozok, seems like the ipod port is moving in the right directon though
01:35:37tucozNice work. Btw, I had a look at a friends nano the other day. Have to say that it is pretty mp3-player
01:36:04tucozgood night
01:36:06 Part tucoz ("Leaving")
01:40:57 Quit paugh ("whoops")
01:43:11amiconnlinuxstb: Turns out this splash() background problem was indeed introduced by one of your commits - but due to my wrong hint :-(
01:43:46linuxstb_I realised that.
01:43:49amiconnOnly the old splashes are affected
01:44:28amiconnObviously I didn't look closely enough. LCD_MAX_LEVEL-1 != LCD_WHITE ...
01:44:32linuxstb_But the old code - using LCD_MAX_RED-1 for example wasn't very intuitive.
01:44:52linuxstb_I mistakenly read that as being the maximum...
01:45:05amiconnYes, even the -1 isn't intuitive. It only works nicely for 2bit greyscale
01:45:35amiconnFor higher depths -1 would only give a very faint effect
01:46:02amiconn(2*LCD_MAX_LEVEL/3) would be more appropriate
01:46:31linuxstb_The plugins use that in some places.
01:46:42amiconn...but for colour lcd, maybe a coloured background would be even better?
01:47:29linuxstb_For the colour LCDs, I think we'll need some kind of (user-configurable) colour scheme. But I'm happy to ignore that for now.
01:48:45amiconnSupport for custom colours (and greyscales) needs some work in the scrolling functions
01:54:59linuxstb_So should I define LCD_LIGHTGRAY and LCD_DARKGRAY in lcd.h and just use those for now?
02:08:29 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
02:09:27 Quit Moos ("Glory to Rockbox")
02:22:30 Nick Rick is now known as Gibbed (i=rick@unaffiliated/Rick)
02:22:43 Nick Gibbed is now known as Rick (
02:22:58 Quit gromit` (Remote closed the connection)
02:35:42 Join gromit` [0] (n=gromit`
02:35:49 Quit gromit` (Remote closed the connection)
02:36:04 Join gromit` [0] (n=gromit`
02:41:59 Part linuxstb_
02:42:44 Quit solexx_ (Read error: 110 (Connection timed out))
02:57:54***Saving seen data "./dancer.seen"
03:07:13 Join _user_ [0] (
03:14:49 Join solexx [0] (
03:15:59 Quit _user_ ("CGI:IRC (EOF)")
03:17:32 Quit dpassen1 ()
03:27:36 Join Zezayer [0] (
03:31:24 Join actionshrimp [0] (
03:39:31 Quit Zezayer ("CGI:IRC")
03:57:18 Quit actionshrimp ("a bird in the bush is worth two in your house")
04:01:57 Part XavierGr
04:26:51 Join whatboutbob [0] (
04:32:57whatboutbobhow do you guys (rb devs) feel about press coverage?
04:35:07whatboutbobto clarify, i can prolly get an article written about rockbox for some aussie newspapers...just wanna know if that'd be ok with y'all?
04:47:26 Join Paul_The_Nerd [0] (
04:54:13thegeek_I know there have been articles about rockbox before
04:54:21thegeek_so I think it would be ok
04:54:33thegeek_but I am in no way offically part of the rockbox team
04:57:56***Saving seen data "./dancer.seen"
05:01:21XShocKbtw, is it normal that ihp120 target does not build with DEBUG option?
05:02:00XShocK/home/rinat/rockbox/apps/codecs/libmad/global.h:36:4: error: #error "cannot define both DEBUG and NDEBUG"
05:02:47Paul_The_NerdI dunno, but it sounds kinda unlikely
05:04:02 Quit DJDD ("Trillian (")
06:11:43 Join DJDD [0] (
06:33:38 Quit whatboutbob ("CGI:IRC (EOF)")
06:44:26 Quit ehntoo (Remote closed the connection)
06:57:58***Saving seen data "./dancer.seen"
07:00:56 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
07:09:24 Quit RotAtoR ()
07:46:17 Quit XShocK (Remote closed the connection)
07:52:16 Join Jungti1234 [0] (n=jungti12@
08:00:18Jungti1234How do you become if apply H100 rockbox firmware to H300?
08:14:36 Join ashridah [0] (
08:16:56Jungti1234Is no there a person to answer on my question?
08:18:18ashridahJungti1234: what was the question?
08:18:27Jungti1234How do you become if apply H100 rockbox firmware to H300?
08:20:08ashridahi doubt the H300 would even accept the H100 firmware
08:20:23ashridahbut the LCD and probably other components are different
08:21:28ashridahso the unit would fail to restart at best, and would probably be unfixable.
08:23:38Jungti1234Thank answer.
08:26:20Slasherihi :)
08:26:54Slasheriamiconn: now the player works fine again, it took something over 10h with hdd spinning do completely discharge the battery
08:27:32ashridahSlasheri: rofl.
08:57:59***Saving seen data "./dancer.seen"
09:16:51 Join itsjohnc [0] (
09:17:04 Part itsjohnc
09:33:07amiconnSlasheri: Nice to hear :-)
09:36:34amiconnI think there are 2 reasons why the battery lasted so long: The hdd was spinning but not accessed, it auto-parked the heads and that lowers power consumption. And the cpu ran at 11 MHz.
09:43:54 Join Mxm`Pas`Bien [0] (
09:57:18 Quit Maxime (Read error: 110 (Connection timed out))
10:09:16 Quit Mxm`Pas`Bien (Read error: 110 (Connection timed out))
10:42:10Bagderhaving mat holton set to moderated on the rockbox list was a good idea
10:42:34Bagderwe get viruses with his From: address sent to the list several times a week
10:53:08 Quit Jungti1234 ("Bye Bye~")
10:57:35amiconnBagder: What is bad (for him) is that these viruses are very likely sent from elsewhere
10:57:45amiconnIsn't there a better method to filter?
10:57:51Bagderyes there is
10:58:00Bagderwe should run a proper viruscheck on all incoming mails
10:58:02***Saving seen data "./dancer.seen"
10:58:11Bagderwe that is still on the todo
10:58:14Bagderbut that
11:12:23 Join amiconn_ [0] (
11:14:52 Quit amiconn (Nick collision from services.)
11:14:52 Nick amiconn_ is now known as amiconn (
11:16:29 Join mashalla [0] (
11:18:03 Join Mxm`Pas`Bien [0] (
11:20:55 Join ender` [0] (i=ychat@
11:42:42 Quit Mxm`Pas`Bien (Read error: 104 (Connection reset by peer))
11:42:44 Join Mxm`Pas`Bien [0] (
11:46:38 Join len0x [0] (
11:46:48 Quit mashalla (Read error: 110 (Connection timed out))
11:48:28linuxstbMorning all. How do people feel about splitting crt0.S into CPU specific files? i.e. crt0-sh.S, crt0-m68k.S, crt0-arm.S and crt0-calmrisc16.S
11:49:25Bagderif you think that improves readability and not hinders maintainability, then sure
11:52:18 Part len0x
11:54:21 Join linuxstb_ [0] (
12:04:54 Join LinusN [0] (
12:04:56linuxstb_Bagder: I don't feel that strongly about it, so unless others say "yes, good idea", I'll just leave it.
12:05:31BagderLinusN might have an opinion?
12:05:37LinusNnot really
12:05:48Bagderthen lets leave it as-is for now
12:06:01linuxstb_OK, no problem.
12:06:02LinusNthe linker cmd files might have to be changed though
12:06:26linuxstb_But is that simply changing the INPUT()?
12:06:50LinusNhmm, looks like they are prepared for that already
12:08:33LinusNlen0x is planning to let the playback status icon reflect the radio status too
12:08:51LinusNso that the Play icon is shown when playing radio
12:09:07 Quit linuxstb (Read error: 110 (Connection timed out))
12:09:58LinusNi don't really mind, but i'm not sure i like the way it is implemented
12:10:54linuxstb_On a different subject, the iPod builds are now clean enough to add to the automatic builds. But I'm not sure how many we need to add - currently there is bootloader/normal/simulator for Nano and Color/Photo.
12:11:10linuxstb_Shouldn't we have a radio icon instead of re-using the play icon?
12:11:26LinusNi think i'd like that better
12:12:08Bagderlinuxstb_: we could start with a few
12:12:13Bagderand expand later if need be
12:13:07Bagderwhat's the binary name of rockbox on ipod?
12:13:11Bagderrockbox.ipod ?
12:13:33 Join Jungti1234 [0] (n=jungti12@
12:13:44linuxstb_I'm not sure if we should have different extensions for the different builds though.
12:14:02Bagderon iriver we don't
12:14:08Bagderwe just store the model name in the file
12:14:12Bagderand have the bootloader check it
12:14:20linuxstb_That's the approach I've adopted so far.
12:16:10Hadakawhat?? rockbox for ipod in development? wow
12:16:32Hadakacan a nano be made to play ogg?
12:17:23 Join linuxstb [0] (
12:17:23BagderI believe it already can with ipl
12:18:48Hadaka(I didn't even know ipod linux ran on nano yet)
12:19:08Bagderit does
12:19:23Hadakabut that's way nice, seems like I have to get one of those then
12:19:29 Join Maxime` [0] (
12:20:41Bagdernow commit something!
12:21:41 Join _FireFly_ [0] (
12:21:52HadakaI think I might commit myself to installing rockbox 2.5 on my archos today ;)
12:21:58 Join DangerousDan [0] (
12:26:17Bagderipod builds added
12:26:40Hadakaon that note, I seem to have trouble locating where are the instructions for rockbox 2.5 in flash
12:26:43BagderI'm sure I messed up something, but that'll show soon :-)
12:27:43Hadakaah, bootbax, found it
12:27:46 Quit Kohlrabi ("Leaving")
12:28:20 Quit Mxm`Pas`Bien (Read error: 110 (Connection timed out))
12:28:29 Join Kohlrabi [0] (
12:28:39 Quit Kohlrabi (Client Quit)
12:28:44 Join Kohlrabi [0] (
12:28:54 Quit Kohlrabi (Client Quit)
12:29:08linuxstbA change has been committed....
12:29:33 Join Kohlrabi [0] (
12:32:59 Quit linuxstb_ (Read error: 110 (Connection timed out))
12:33:09 Join mashalla [0] (
12:38:38 Join Moos [0] (
12:45:58Bagdernow that is a table
12:46:10Bagdercheck the cvs build table
12:47:04_FireFly_wow who was the elefant which had sit down on the table;)
12:47:56linuxstbNice :)
12:48:20linuxstbI'm thinking of disabling the building of the plugins for now - that's what is causing all the errors in the main build.
12:48:55Bagdersure, we did that in the iriver beginning too
12:49:13linuxstbOK, another commit on the way.
12:49:42BagderLinusN: seeing we miss a h300 boot build there, you think it is premature to add one?
12:50:55 Join actionshrimp [0] (
12:51:29linuxstbIf I make two commits in quick succession, will only the first commit be used in the build? Or do I have a few seconds to do the second one?
12:51:47Bagderit checks for new changes every 60 seconds
12:52:06Bagderso you have 0-60 seconds for a second commit ;-)
12:52:31linuxstbOK, I'll just separate them by a semi-colon on the same line.
12:53:27linuxstbOK, the next builds should be yellow for the main targets.
12:53:28BagderI'm leaning towards a system that will require N seconds of no commits before it actually starts the rebuild
12:53:39Bagderto allow a series of commits better
12:53:43 Join andy_ [0] (
12:54:05_FireFly_this shouldn't be so hard to implement
12:54:21 Join ep0ch| [0] (n=ep0ch|@
12:54:22Bagdernot really, it just conflicts somewhat with how the scripts work atm
12:54:33linuxstbMaybe just increase the gap from 60 seconds to 2 or 3 minutes.
12:54:43andy_is it only me who would like line-selector to default to inverse bar instead of pointer?
12:54:57ep0ch|the new memcpy doesnt seem to affect the codecs boost ratio at all :(
12:55:11linuxstbep0ch|: The codecs probably don't use it yet.
12:55:27ep0ch|they have to be written to use it?
12:55:29_FireFly_Bagder: what have you done to the wps-code ;) i have to update my wps-widget again ;)
12:55:47Bagdermy change was very small
12:55:56ep0ch|i thought the new memcpy would override the standard one
12:56:27linuxstbep0ch|: I was wrong, looking at the code, I think they do use the new memcpy. But obviously not enough for it to be noticable.
12:56:47linuxstbIt's only memmove that's implemented locally for the codecs.
12:57:02ep0ch|oh ok
12:57:23ep0ch|how's the ipod coming along? :)
12:58:05***Saving seen data "./dancer.seen"
12:58:16ep0ch|i see lots of red
12:58:17linuxstbRockbox now compiles (almost) cleanly for the iPod. The work is now to slowly fill in the gaps.
12:58:28linuxstbep0ch|: Those are just the plugins - see the results of the next build.
12:58:50linuxstbIt's just because there are no button mappings defined for each plugins.
12:58:57linuxstbSo I've disabled the plugin building for now.
12:59:11ep0ch|so how far does booting get?
12:59:49linuxstbThe bootloader works (booting either ipodlinux or the Apple firmware). But there is not yet a functional Rockbox for it to load.
13:00:28 Quit ashridah ("Leaving")
13:02:56ep0ch|ipod have two cpus right? do you think rockbox will need much work to utilise both cpus?
13:04:29linuxstbI think we would just use one CPU for the codecs, and one for everything else. So it shouldn't be too hard - the codecs already run in their own thread.
13:04:56LinusNdo we know how to communicate between the cpu:s?
13:05:46linuxstbI don't, but the IPL people do - they've got both CPUs working. I would also expect it to be documented in standard ARM documentation as well."> <- How about this mp3 player?
13:06:35 Join ender1 [0] (i=ychat@
13:07:24ep0ch|i have a feeling the hardware in that one is similar to ihp
13:12:56_FireFly_looks nice
13:14:55Jungti1234And this? -
13:14:56_FireFly_but to much windows related pics for my taste ;)
13:15:30_FireFly_and i can't understand korean or what that language is
13:16:07andy_Jungti1234: cute menu graphics.. we should have a cute menu option in rockbox :)
13:17:40Jungti1234How about mp3 player kit?
13:18:05Jungti1234Popularity is good very in the Korea.
13:18:30Jungti1234But, there is no many a person who buy because of expensive price.
13:20:03Jungti1234Use translation.
13:23:44 Quit ender` (Read error: 110 (Connection timed out))
13:26:09linuxstbI mentioned this yesterday, but the last two warnings in the iPod builds are actually bugs. Someone familiar with databox should try and fix them (values in the range -1 to 130 being stored in a char). Only the arm compiler gives a warning because chars are unsigned by default on that platform, but it's a problem with signed chars as well.
13:31:11 Quit andy_ ()
13:31:37MoosHCl is your man ;)
13:37:25Jungti1234Found interesting image.
13:39:27_FireFly_has someone of the devs found some time to look at my wps-widget (
13:39:35Jungti1234Does iriver hate ipod? :)
13:40:11LinusN_FireFly_: zip?
13:40:18_FireFly_it looks like or it's only a marketing gag
13:40:57 Join len0x [0] (
13:40:59_FireFly_LinusN: yepp
13:41:08LinusN_FireFly_: i'd prefer a single patch file
13:41:11 Quit Maxime` (Read error: 104 (Connection reset by peer))
13:42:37_FireFly_the problem was i had made the diff with "cvs diff" but due the anonymouse checkout of the tree i can't add my files to the tree so i had made a zip
13:43:04_FireFly_the current filename-sheme of my patch might be not the best
13:43:31LinusNthe "g" in gwps.c might not be necessary
13:45:07_FireFly_i will change it
13:45:26_FireFly_any other suggestions ??
13:46:26LinusNnot yet, will have a look
13:46:58LinusNbasically, you've moved the wps code to a "widget", so it can have two instances?
13:47:33_FireFly_with a few changes yes i think
13:48:22_FireFly_currently some functions uses the gui_sync* function of other widgets
13:50:13LinusNnitpick: if(!wps_data || !buf) return false;
13:50:22LinusNbreak that into two lines
13:50:38_FireFly_why ??
13:50:55LinusNeasier to read imho
13:51:45thegeek_LinusN : thanks for the connector, it works good, but I found out that the actual problem was not the connector, it was the two small resistors on the ground channel
13:51:54thegeek_L5 and L6
13:52:40LinusNaha, so it works now?
13:52:52thegeek_it worked great last night
13:53:00thegeek_however, I think my harddrive is failing
13:53:10thegeek_I've been getting ata -35 errors for some time now
13:53:25thegeek_and it seems my little operation has made it worse
13:53:41_FireFly_LinusN: if i change gwps.h to wps.h then which header(this file or the file in apps/ ) is matched first is wps.h is included
13:53:41LinusNit might not be the hard drive after all
13:53:49thegeek_very true
13:54:03LinusN_FireFly_: aaaah, didn't think of that
13:54:43LinusNthegeek_: does it work better if you push the ata daughterboard together with the mainboard?
13:54:55thegeek_but when I hear the hardrive spin up, when it fails I have a "click"
13:54:57thegeek_let me check
13:55:33thegeek_it worked
13:55:51LinusNthen you have a problem with the connector between the boards
13:55:59thegeek_I just squeezed right over the connector to the harddrive
13:56:07LinusNa common issue with these players
13:56:39thegeek_good news:)
13:56:41thegeek_I guess
13:56:45thegeek_is it fixable?
13:56:55LinusNyes, but it takes quite some soldering skills
13:57:22thegeek_as in
13:57:33thegeek__very_ good soldering skills?
13:58:25 Quit Sandking ()
13:58:41 Join Mxm`Pas`Bien [0] (
13:58:42LinusNwell, i've done that repair a few times now, and it requires SMD soldering experience
13:59:02thegeek_it seems to work great if I just keep a little pressure on the connector
13:59:29_FireFly_tape it ;)
13:59:34thegeek_perhaps I could just add a little object there that would exert pressure when the case is screwed together
14:00:32LinusNyes, that might make it work for a little longer
14:00:51LinusNi have heard of other people doing that
14:04:07LinusN_FireFly_: at first glance, your widget looks good
14:04:54LinusN_FireFly_: what's the idea with gwps-common?
14:05:12 Quit gromit` (Read error: 104 (Connection reset by peer))
14:05:19thegeek_when the hardrive spins down, is there supposed to be a click
14:05:31thegeek_guess my fix is no 100%
14:05:43_FireFly_in the first place i had the idea to spilt of these function which are used from both displays
14:06:06SlasheriHmm, now my secondary bootloader and rombox works fine :)
14:06:14LinusNSlasheri: wee!
14:06:25 Join gromit` [0] (n=gromit`
14:06:30_FireFly_but now i think these file can be renamed
14:06:31SlasheriNext step would be to change the initial reset vector to replace the primary boot loader
14:06:55SlasheriBut i think i don't want to try that yet.. :)
14:07:07_FireFly_LinusN: now you have your single patch-file ;)
14:07:13LinusNsaw that, gr8
14:07:34LinusNSlasheri: so you have a flashing plugin?
14:07:39SlasheriLinusN: would you like to test my bootloader in few days before i try booting it directly?
14:07:49SlasheriLinusN: yes, it works great :)
14:08:02MoosSlasheri: Hi, have you got a BDM?
14:08:07SlasheriMoos: no
14:08:14LinusNSlasheri: submit a patch
14:08:43SlasheriLinusN: ok, thanks. I will do that soon :)
14:09:02MoosSlasher: you've got courage to do those things without BDM :)
14:09:17LinusNMoos: finland is not that far from sweden :-)
14:09:41LinusNhe can row over to stockholm and unbrick it :-)
14:09:56SlasheriLinusN: now the flash plugin erases section 0x8 - 0x1effff and places my secondary boot loader at 0x8. Then it places rockbox at 0x10000
14:10:07Slasherihehe :D
14:10:14 Quit gromit` (Read error: 104 (Connection reset by peer))
14:10:55 Join webguest65 [0] (
14:11:04webguest65hi everybody
14:11:07 Join gromit` [0] (n=gromit`
14:11:21webguest65is somebody working on aac?
14:11:42thegeek_seems to work ok now
14:11:59thegeek_if it just last a little longer I'm happy
14:12:04thegeek_thanks for all the help LinusN:)
14:12:26webguest65what does it need to play realtime?
14:12:45LinusNSlasheri: i had no idea that you could erase from 0x08, i thought you had to erase the entire sector
14:13:20SlasheriLinusN: yes, i had to erase entire section but i have a safe guard in the code that immediately after the erase replaces the reset vectors
14:14:13 Nick ender1 is now known as ender` (i=ychat@
14:14:20SlasheriLinusN: when the only bootloader would be at 0x8, user could in future safely upgrade the rockbox because flashing wouldn't touch any cricital sectors
14:15:34LinusNso you have two boot loaders, one primary and one secondary?
14:15:43Slasheriyes :)
14:15:56Slasheriand i haven't touched the primary bootloader at all
14:16:09Slasherii just press rec+play when i want to test my boot loader
14:16:27Slasheri(the secondary bootloader is almost same as primary, only a few little modifications)
14:17:46SlasheriLinusN: now the secondary boot loader seems to work well but i am afraid to change the reset address to that bootloader.. If it fails, i could no longer boot at all
14:20:14SlasheriLinusN: btw, do you think it would be a very bad idea to save settings to the last sector in flash? (that would require erase cycle after 8 save cycles). I am going to try if that could give any decrease on boot time
14:21:52LinusNSlasheri: i wouldn't want to write to the flash more than necessary
14:23:26ep0ch|you know those 128 bytes that are available to store settings with...
14:23:27Slasheriok, that would then probably be a bad thing to do
14:23:45Slasheriep0ch|: those are not enough :/
14:23:50ep0ch|could it be possible just to save the frequently changed settings in there
14:24:03ep0ch|like volume, etc
14:24:25ep0ch|that would save writing to the disk on shutdown on quite a few occasions
14:24:55LinusNthe settings are saved whenever the disk spins up
14:26:07ep0ch|i thought the hardisk always spun up on shutdown to write the settings
14:29:08LinusNnope, only if the settings have changed since the last spinup
14:30:16ep0ch|well, i just checked, i started up rockbox waited for the cache check to finish and spin down. Then switched off, the HD spun up and did something.
14:30:17 Quit gromit` (Read error: 104 (Connection reset by peer))
14:30:53 Join gromit` [0] (n=gromit`
14:31:04Slasheribtw, with flashed rockbox we can safely disable that cache check also when it's not necessary :)
14:34:17 Quit Kohlrabi ("Leaving")
14:34:18 Quit Mxm`Pas`Bien (Read error: 104 (Connection reset by peer))
14:34:41 Join Maxime` [0] (
14:35:37webguest65i read somewhere it's because of the compiler (gcc 3.x.x instead of gcc 4.0) true?
14:36:05webguest65i mean aac playback
14:38:01ep0ch|i believe aac gained near realtime playback under gcc4. however other codecs suffered, i.e. less effecient.
14:39:34webguest65so, either the other codecs have to be rewritten for gcc4 or aac has to be further optimized with gcc3 ?
14:40:20ep0ch|aac has to further optimized whatever
14:40:46webguest65i thought somebody allready worked on that quite hard
14:50:57 Quit webguest65 ("CGI:IRC")
14:51:20 Join Maxime`` [0] (
14:55:07 Join Maxou` [0] (
14:55:52 Quit Maxime`` (Read error: 104 (Connection reset by peer))
14:58:06***Saving seen data "./dancer.seen"
14:58:22LinusNhow about making the default iriver font a little bigger?
14:59:13Maxou`snap font is great :p
14:59:20ep0ch|well i'm using nimbus-12, wish there was a nimbus-10...
14:59:52len0xnimbus is a great font indeed
15:00:14crwli personally think fonts as small as snap are almost useless, and my vision is ok
15:00:56crwl(and it's sad that almost every WPS in WpsGallery is based on some ridiculously small font)
15:01:11LinusNcrwl: in the iriver, yes, but not on the archos
15:01:26crwlLinusN, indeed
15:01:52LinusNi personally use nimbus-12
15:02:07crwlme too
15:02:18crwlit's a good font
15:02:30LinusNa chicago clone
15:02:41 Quit Maxime` (Read error: 110 (Connection timed out))
15:02:54LinusNgives that nice ipod look :-)
15:03:13ep0ch|which file specifies the default font?
15:05:07LinusNit used fonts/rockbox_default.bdf
15:05:10linuxstbwebguest65 (if you read this message in the logs): the status of aac is "in progress". It's being worked on a little, but it isn't anyone's top priority. It did gain significantly from gcc4, but as you said, the other codecs seemed to suffer a little.
15:05:42 Quit Jungti1234 ("Bye Bye~")
15:06:24SlasheriLinusN: with default settings iriver boot time from flash: 4s
15:08:17ep0ch|how are you timing this? from pressing 'play' to when?
15:08:37Slasheriep0ch|: from play until directory tree is shown
15:08:53 Join webguest24 [0] (
15:09:17ep0ch|its about 5s normally then
15:09:30Slasheriyes, i got 6s
15:10:01webguest24hi there, does nayone have time for a noob question about his iriver?
15:10:03ep0ch|i think the realtest should include having resume playback enabled
15:10:45ep0ch|well more realistic
15:10:48webguest24ive got a file in my .rockbox directory that i cant access/delete. its called like B something, with no extension and is about 1MB, how do i get rid of it?
15:11:13webguest24i think its buggering up my WPS, as they cant load the bmps
15:11:35webguest24do i have to format or defrag or what?
15:12:07ep0ch|how have you tried deleteing it so far?
15:12:25webguest24shift delete
15:12:34ep0ch|via usb?
15:12:40linuxstbwebguest24: Are you using a very new version of Rockbox? The way that the WPS handles bitmaps has changed.
15:12:41ep0ch|what about rockbox delete
15:12:42webguest24i also tried renaimg it but it doesnt work
15:12:46webguest24yes, via USB
15:13:08webguest24well rockbox doesnt allow you to navigate to .rockbox folder
15:13:19ep0ch|change navigation view
15:13:33ep0ch|Show Files: All, should do
15:14:18Moospress lon a/b and down joistick
15:15:13webguest24it wont delete
15:15:22webguest24but it will rename
15:16:02webguest24maybe the file is like a bad sector or something? a dodgy file?
15:16:10Maxou`u tried a chkdsk ?
15:16:41webguest24no, how would i do that?
15:16:50webguest24and might it wear out the disk?
15:17:33Maxou`chkdsk youririverDriveLetter:
15:17:39_FireFly_I think Maxou` mean scandisk ;) under windows
15:17:41webguest24in MS DOS?
15:17:44 Join tucoz [0] (
15:17:51Maxou`in a command window yes
15:18:00Maxou`Start > Run > Cmd
15:18:01Maxou`here ^^
15:18:17 Join ehntoo [0] (
15:18:48tucozWhat exactly does the crt.S file do, and what does it mean?
15:19:25tucozI mean, what does crt stand for?
15:19:44webguest24so when im in the cmd window what do i type? chkdsk F: or whatever?
15:20:38LinusNtucoz: it performs the initializations for the C runtime code
15:21:00LinusNand i believe it stands for C RunTime
15:21:19tucozLinusN, ah, ok. Then I was on the right track.
15:22:00webguest24ah i got it
15:22:35webguest24its truncating the file, whatever that means!:)
15:22:59Slasheriep0ch|: 8s to wps from disk, 6s from flash
15:23:25tucozhmm, was the lineselector not inverted by default? I seem to recall that it was, but when my settings was reset I got the little tiny arrow instead.
15:24:07webguest24the file is still there, what the hell do i do?
15:24:25ep0ch|tucoz: it was for a brief period iirc
15:24:25_FireFly_backup your files and format the drive
15:24:45Maxou`webguest24: what's the exact name of the file
15:25:09webguest24it is called "p"
15:25:12tucozep0ch|, oh. Then it should be inverted by default imho
15:25:21webguest24and in the .rockbox directory
15:26:31webguest24chkdsk says /p first allocation unit is not valid
15:27:18LinusNi also think the default should be inverted
15:27:54Maxou`chkdsk /F F:
15:28:21webguest24have done, 97% complet
15:28:28tucozwebguest24, is it possible to delete the file from within rockbox?
15:28:35webguest24no, tried that
15:28:46Maxou`with the /F ?
15:28:54webguest24do i want to convert lost chains into files?
15:29:00webguest24yes with the /f
15:29:20LinusNwebguest24: yes, it you want to see what the file contained
15:29:39webguest24right lets see if it worked:)
15:30:06webguest24wahoo! p is gone
15:30:21Maxou`whatever we say, chkdsk is great sometimes lol
15:30:28webguest24thanks a bunch guys
15:30:33webguest24and thanks chkdsk
15:34:33 Quit webguest24 ("CGI:IRC")
15:35:06tucozregarding crt.S, is the S meaning subroutines?
15:35:31LinusNS == assembler source
15:35:37*tucoz writes really weird sentences today
15:36:13tucozLinusN, ok. You were correct about the c-run-time
15:36:42tucozBut I do know to little on compilers to understand what it does ;)
15:42:52tucozLinusN, you better watch out. A wolf is on the loose in Stockholm. :D
15:43:17*LinusN locks the door
15:48:43 Quit Lynx_awy (Read error: 104 (Connection reset by peer))
15:49:31 Join Lynx_awy [0] (
15:51:51 Part len0x
15:55:33SlasheriLinusN: i am implementing a failsafe menu to the new bootloader with options: "A) Boot from disk B) Boot from rom C) Enter USB mode D) Reset settings E) Boot from address"
15:55:56Slasherithat would then pop up if user presses rec
15:56:43Moosgreat job Slasheri and this without BDM :D
15:57:01_FireFly_tucoz: about the wolf here is a short tale which i had made :):
15:57:03_FireFly_The Wolf will come to bring death to everybody
15:57:04_FireFly_nobody have seen the wolf with his own eyes
15:57:04_FireFly_they found only some paw print of a wolf near the death bodys
15:57:04DBUGEnqueued KICK _FireFly_
15:57:04_FireFly_but the mysteric thing is, no wounds was found on the death bodys
15:57:04_FireFly_so nobody knows if the killer is a wolf oder not
15:57:07_FireFly_and nobody found the answer for the mysterium
15:57:28MoosSlasheri: good luck for your futur progress
15:57:40solexxis the information about the iriver remote on the iIriverStatus page still accurate?
15:57:43solexx( Currently the file browser and partial menus are supported)
15:57:47SlasheriMoos: thanks, i try not to brick my player :)
15:58:11_FireFly_i have made a wps-widget for multi-screen
15:58:18_FireFly_it can be found on the tracker
15:58:23Moosis it finished?
15:58:36Moosfull I mean :)
15:58:37_FireFly_what the wps-widget ??
15:58:52tucoz_FireFly_, nice rhyme
15:58:59_FireFly_tucoz: thanks
15:59:25CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:59:25*tucoz likes the a wolf oder not
15:59:41Moos_FireFly_: have you speak with TiMiD about your new patch?
16:00:08LinusNSlasheri: "enter usb mode"? why?
16:00:09_FireFly_yepp he has given some suggestions about it :)
16:00:26_FireFly_TiMiD are u here ??
16:00:31Mooshe like your patch?
16:00:53SlasheriLinusN: that is not probably necessary, i will figure it out later :) of course it would enter usb mode automatically also if it detects usb
16:01:03Slasheribut it could give user an option not the enter the mode
16:01:10Slasheriwith a little timeout for example
16:01:14LinusNand what would B) be?
16:01:23_FireFly_i don't know becuase i didn't said something about the current version of my patch but LinusN seams to like it
16:01:31Slasheriit would do the normal flash boot
16:01:32_FireFly_he didn*t
16:01:49Moos_FireFly_: mk
16:01:49SlasheriE) would be for debugging and developers
16:01:57Slasheriit would allow to run code from any address
16:02:02LinusNSlasheri: i think it should be on anither button
16:02:19SlasheriLinusN: Hmm, the failsafe menu?
16:02:33LinusNit would be so cumbersome to boot the original for those who still do that
16:02:35Slashericurrently i have put it to hold switch
16:02:56SlasheriLinusN: that bootloader would be available only when rockbox is flashed..
16:03:11Slasheriso it wouldn't be possible to boot original firmware with that bootloader at all
16:03:21Slasheriunless rolo could do that in future
16:03:27Mooswe don't need it :)
16:03:35LinusNok, so rombox needs a different primary boot loader
16:03:56LinusNthe i don't mind the rec button
16:04:09Slasherithat is mainly patched original bootloader with ifdefs
16:04:18Slasheriah, ok :)
16:05:08LinusNi guess two options would be enough: "boot from disk" and "boot from address"
16:05:20Slasherihmm, true
16:05:27Slasheriand maybe also the reset settings
16:05:29Moossounds very good
16:11:56LinusNi think the settings reset should be in rockbox itself and not in the boot loader
16:12:47LinusNthe boot loader should be as dumb as possible
16:13:31SlasheriHmm, yes. But somehow it would be good if user could reset the settings before rockbox is fully booted. Either from bootloader or rockbox itself
16:14:21LinusNyes i fully agree with that, but not in the boot loader
16:15:35SlasheriWhat about giving bootloader a number or something like that, which would be passed to rockbox when it's booting?
16:15:47Slasherithen it could do some things depending on that value from bootloader
16:17:24LinusNsounds complicated
16:17:43LinusNthen you have to update the bootloader if you add new startup features
16:18:05Slasheriuser could enter any value from bootloader, and rockbox would interpret it
16:18:14Slasherisomething like a linux bootloader :)
16:18:24LinusNwhy from the boot loader? i don't get it
16:18:25Slasheriwhere kernel interprets the boot options
16:19:13SlasheriHmm.. if we have a menu there, it would be easy to read the value in the failsafe mode. Of course that would not be the best way for user
16:19:33LinusNespecially since the non-rom bootloader doesn't have that menu
16:19:57LinusNi wast to keep the bootloader functionality to a minimum
16:21:12Slasherihmm, then rockbox could check some button values or implement own menu there. I really don't know what would be the best choice :)
16:26:32LinusNthe best choice is obviously to do it in rockbox
16:26:49LinusNso you can extend that menu without having to reflash the boot loader
16:27:00Slasheriyes, that would be good
16:30:55 Quit HCl (Read error: 110 (Connection timed out))
16:32:31 Join preglow [0] (
16:32:48preglowdamn, you guys have been busy
16:37:44linuxstbWelcome back preglow.
16:39:27preglowyou're right, the ipodloader button driver is glitchy
16:39:37preglowi did a hack to make it load firmware according to keypress
16:39:42preglowbut it doesn't always work
16:40:05linuxstbI've been playing as well, and I think the trick is to press the key as quickly as possible - before the apple logo appears.
16:40:20preglowwell, not exactly an elegant solution
16:40:36preglowi wonder if we should set up the full interrupt driven driver in the bootloader, already
16:40:46linuxstbThat's why I wanted to write a "proper" driver - based on the IPL kernel driver.
16:40:46preglowbut i also suspect the best thing is to touch as little as possible :/
16:41:02preglowbut yes
16:41:25preglowwhat should go first? making rockbox work?
16:42:03linuxstbIt would be nice to get Rockbox running. That's just waiting for crt0.S to be written now.
16:42:21linuxstbAs you can see from CVS, rockbox.ipod is compiling now.
16:42:42preglowwe also need threading, no?
16:44:02linuxstbYes - it obviously won't do anything useful without threading.
16:44:30preglowthat shouldn't be too hard, if we forget about the second core for the moment
16:46:03preglowi'll have a look at it now, as a matter of fact
16:52:12 Join _DangerousDan [0] (
16:55:18Maxou`hum, what do yo umean about running "rockbox fro mflash"? you mean 'copy' the .rockbox folder into the flash memory of the player?
16:55:54preglowwell, primarily rockbox.iriver
16:56:13Maxou`and, what's the advantages? (other than 2s less @ boot?)
16:56:28preglowfaster boot, more memory for file buffering
16:56:35Maxou`ok I see
16:56:48LinusNi don't buy the memory argument though
16:58:10***Saving seen data "./dancer.seen"
16:58:19 Quit DJDD (Read error: 110 (Connection timed out))
16:59:30preglowLinusN: well, me neither
16:59:47preglowand don't really the faster boot either
16:59:48LinusNwhat's the gain? 200-300k?
16:59:57preglowbut then again, i'm not the one coding this
17:00:27Moosless spinning disc, right?
17:00:44preglownot by any amount that will matter
17:01:22linuxstbAlmost a screendump feature:
17:01:50preglowi'm doing a big pscp here, and suddenly nothing but putty and pscp works
17:02:06tucozcool, you are really going somewhere with this it seems.
17:02:18linuxstbI'm just trying to implement the screen_dump function for 16-bit displays. Get rid of a few warnings.
17:02:22ep0ch|is the read speed from flash and ram similar on ihp?
17:04:02preglowep0ch|: perhaps similar, but not equal
17:04:45preglowwhy the hell doesn't web browsing work
17:05:03preglowopera has its bugs, obviously
17:05:15LinusNnorwegian crap :-)
17:05:30ep0ch|wow an opera user
17:05:50 Quit DangerousDan (Read error: 110 (Connection timed out))
17:05:53preglowi just recently started to use it
17:06:01preglowused to use firefox
17:06:06preglowwhich isn't without its annoying parts either
17:06:48ep0ch|i always found opera couldnt render quite a few pages well, but that was a good few years back
17:06:48linuxstbAh, that's more like it:
17:07:02tucozpreglow, use ie for a wonderful browsing experience ;)
17:09:51preglowep0ch|: only problem i have with opera is connectivity issues
17:09:55preglowsometimes it just stalls in loading
17:09:57preglowwhen there's no reason
17:10:27Maxou`i quite like opera
17:10:39Maxou`except some things but..
17:28:35 Join chopped_pork [0] (
17:41:34tucozOT, does static linking mean that I 'bundle' the executable with the library? If used with e.g. boost, will this give me faster executables?
17:41:54preglowyou bundle the library _in_ the executable
17:41:57preglowfaster, perhaps
17:42:06_FireFly_maybe faster but bigger executables
17:42:09preglowdepends on whether any other programs use boost as well, at the same time
17:42:22 Part ep0ch|
17:42:24preglowmost of boost is header based anyway, so you don't have a choice
17:42:46tucozpreglow, _FireFly_ : ok, thanks. Hmm, what does that mean? that I can not link it statically?
17:43:03preglowtucoz: there is nothing to link against in the majority of boost. all the functionality is in .h files
17:43:14preglowtucoz: the code is inlined
17:43:17tucozpreglow, oh, I see.
17:43:27preglowthere are some expecptions
17:43:31preglowline filesystem and python wrapper
17:43:45*tucoz understood anyway
17:44:30linuxstbAnyone know if a linux kernel has a built-in checksum, or anything else I can use to check that it's valid?
17:44:36tucozboost is really good I think. The little I have used it has helped me lots
17:44:45preglowyes, boost is a must for c++ coding, if you ask me
17:44:49preglowi use it all the time
17:45:21tucozpreglow, cool. I used stl all the time, but there were some parts missing there, so boost is a nice addition to that.
17:45:36preglowi use stl all the time too
17:45:44preglowbut alone it's a bit lacking
17:45:49tucozI mean, stl exclusively.
17:46:07tucozBut, then I had to reinvent the wheel all the time
17:47:32preglowwhat's up with the 'started' variable in thread.c ?
17:47:49preglowseems like some archs does started thread detection in a different way without any very good reason
17:48:36LinusNit depends how the thread is started
17:49:52LinusNyou either start it by "manually" putting a valid stack frame on the thread stack and do a context switch, or you use the "started" flag to make the scheduler start it explicitly
17:51:04preglowthe why doesn't all archs have 'started'?
17:51:40preglowforget me
17:52:04LinusNforget who? :-)
17:52:11preglowthere you go
17:52:21preglowbah, who puts the stack pointer in a non-last register!
17:52:22preglowarm does
17:55:04 Join amiconn_ [0] (
17:56:05amiconn_LinusN, preglow: All rockbox targets use a 'started' indication in the scheduler, the forged context method is long gone
17:56:39amiconn_The difference is that the gmini uses a separate 'started' flag, while the others use the start address field for that
17:56:58preglowwhy does it do that? just hasn't been fixed yet?
17:57:06amiconn_If the start address is non-NULL, the thread needs to be started. NULL indicates already started
17:57:35amiconn_The gmini port is stalled, and I didn't want to change code that I can neither compile nor test
17:59:11preglowok, thanks for clearing up
17:59:22preglowi'll delete that in the ipod target, then
17:59:45amiconn_The SH and coldfire targets both use the start address technique
18:00:02amiconn_I'd suggest to do the same on arm
18:02:07 Join RotAtoR [0] (
18:02:35preglowi'll just assume members in a struct that are 4 bytes big will never be padded here
18:02:40preglowsound assumption, yes?
18:02:53preglowthe members are 4 bytes big, not the struct iself ;)
18:04:48amiconn_I think this assumption is safe, at least on a <= 32bit architecture
18:05:07 Quit mashalla (Read error: 110 (Connection timed out))
18:06:17 Join DJDD [0] (
18:13:21preglowr0-r3 are both scratch and used for parameter passing on arm
18:14:09preglowso i still need to save them, then
18:14:36amiconn_SH1 is similar; r0-r7 are scratch, r4-r4 are used for parameter passing
18:14:43amiconn_r4-r7 I mean
18:15:24preglowof course
18:15:29preglowyield takes no parameters
18:18:33 Quit amiconn_ ("CGI:IRC (EOF)")
18:20:51preglowanyone know about arm and alignment needs?
18:21:06preglowthink i'll align stack by four to be sure
18:23:23 Join ashridah [0] (
18:25:16TiMiD_FireFly_: !
18:25:31TiMiDshow me your wps :)
18:25:53_FireFly_look at the tracker :)
18:26:37ashridah_FireFly_: that's right, you wrote a remote wps patch or something didn't you?
18:27:04_FireFly_not a remote-wps patch a multi-screen-wps-"widget"
18:27:39TiMiDI look
18:27:57TiMiD(sry, but I havn't a lot of time, lot of things to do those days ...)
18:30:26 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
18:31:02preglowlinuxstb: do you know how wide the ram bus is on the ipod?
18:31:20preglowin some cases they say it actually pays to compile to thumb code
18:31:52 Join Gibbed [0] (
18:31:53 Quit Rick (Nick collision from services.)
18:32:02 Nick Gibbed is now known as Rick (
18:32:06ashridah_FireFly_: it seems to have a compilation error in gui/gwps.c
18:32:16ashridahgui/gwps.c: In function 'wps_data_load':
18:32:16ashridahgui/gwps.c:647: error: syntax error before 'if'
18:33:02ashridahmissing semicolon
18:33:21ashridahworks after adding it to gwps.c:646
18:33:38_FireFly_i see :)
18:33:44_FireFly_sorry for that
18:35:44_FireFly_fixed on tracker
18:36:13ashridahso what does the patch do anyway? :)
18:36:30ashridahwait. hang on, i've copied the wrong :)
18:36:36_FireFly_adds multi-screen support
18:37:11_FireFly_and as in my unofficial remote-patch it adds also loading of *.rwps (remote-wps) files
18:37:30linuxstbpreglow: No, I don't know.
18:39:15ashridah_FireFly_: got it working. hot.
18:39:37linuxstbpreglow: Interesting thread:
18:40:45_FireFly_ashridah: thanks
18:41:00_FireFly_i hope it will get in cvs soon
18:43:20preglowwhat, there are numbers there that suggests gcc 2.95 is faster than 3.4...
18:43:24preglowgenerates faster code, that is
18:44:09_FireFly_i think that depends on the target arch
18:47:00preglowwhat is the procedure register on sh? same as link register on arm?
18:49:45 Quit DJDD ("Trillian (")
18:58:11***Saving seen data "./dancer.seen"
19:02:49 Join Kohlrabi [0] (
19:03:10 Quit ashridah ("paintball :)")
19:05:15LinusNthink so
19:05:27LinusNit contains the return address in a subroutine call
19:06:25 Join amiconn_ [0] (
19:07:07amiconn_preglow: The SH procedure register stores the return address from a subroutine.
19:07:24amiconn_bsr/jsr don't push it on the stack
19:08:04amiconn_This means the lowest-level subroutine calls don't need to access memory
19:12:33preglowamiconn_: same as on arm, thebn
19:19:50 Quit lostlogic (Read error: 110 (Connection timed out))
19:22:39 Join Lear [0] (
19:24:09 Join lostlogic [0] (
19:24:22preglowi love the fact that you need to bloody buy protective cases for ipods in multipacks
19:25:07thegeek_preglow : when I found out, I was upset
19:25:24preglowwhat the hell is up with it anyway?
19:25:29preglowwhy would i want _FIVE_ cases?
19:25:31linuxstbpreglow: Obviously you want them in different colours to match what you are wearing...
19:25:36thegeek_it is extremely annoying
19:25:42preglowlinuxstb: i will kill then maim steve jobs
19:27:04preglowyou need to buy the arm instruction reference
19:27:06preglowi love people
19:27:24_FireFly_moep :)
19:27:46thegeek_I feel like some carmageddon
19:27:51thegeek_anyone remember that game?
19:27:56thegeek_wonderful stress reliever
19:28:04preglowplayed it a ton of years ago
19:28:06thegeek_nr 2 was the best
19:29:40preglowcan only remember the first one
19:30:25thegeek_it seems they are making a nr.4
19:42:53 Quit amiconn_ ("CGI:IRC (EOF)")
19:47:26preglownow, if there was some way to test threading...
19:53:59linuxstbYou can probably just create some threads manually in the bootloader that display a counter on the screen and then yield.
19:54:43linuxstbIn fact, the bootloader at least starts the scroll thread.
19:55:06chopped_porkhm... how can i convert a .ttf font to .fnt? i'd like to check if verdana 8 looks as good on rockbox as it does in x11 :)
19:56:57tucozchopped_pork, maybe fontforge can help you with that
19:58:08tucozor maybe this:
19:59:29tucozand run the convbdf tool found in the rockbox/tools dir
19:59:31chopped_porkgreat, thanks, i'll look into that
19:59:46tucozwhich will convert a .bdf font to fnt
20:00:29chopped_porkoooh ttf2bdf is for windows, evil :>
20:00:41*preglow suddenly realises he's in windows
20:00:54chopped_porkah much beter now ^^
20:01:03tucozhehe, what a nice suprice
20:06:15chopped_porkconvbdf says Error: Too wide characters (>16 pixels) :/
20:08:36chopped_porkoh it worked with a 6px font size
20:08:45chopped_porkand 7px too
20:09:20_FireFly_chopped_pork: remember you don't have a big resolution on the device :)
20:09:43_FireFly_so 6 or 7px fonts-size is big enough
20:10:26_FireFly_or i'm wrong
20:14:40chopped_porkalright so either i've done something wrong or the default font actually *is* verdana
20:29:01 Quit _DangerousDan (Read error: 104 (Connection reset by peer))
20:32:39 Join DangerousDan [0] (
20:40:51 Join Sandking [0] (
20:46:48 Join len0x [0] (n=len0x@
20:46:55 Join markun [0] (
20:53:14preglowamiconn: how do we know that gcc always allocates a scratch register for the attr parameter in the store/load_context functions?
20:58:13***Saving seen data "./dancer.seen"
20:59:45 Quit BirdFish ("reboot")
20:59:47linuxstbpreglow: I've been working on a bootloader a little - I'm about to commit a new version with the ipodloader button code which lets you triple-boot depending on keypress.
21:01:30preglowwhich workS?
21:03:13 Join XShocK [0] (
21:03:33linuxstbpreglow: If you press the key very very early, then yes. So it's not perfect, but it's good enough for now.
21:03:36preglowthe bootloader code or the proper interupt driven code?
21:03:49linuxstbYes, just the bootloader code. We can replace it later.
21:03:51preglowi've got one which does that as well
21:03:54XShocKnow my iriver shows time and date.. :))
21:04:04preglowbut it doesn't look like shit, so go ahead and commit yours :)
21:04:17 Join yngwi [0] (
21:04:29LinusNXShocK: nice!
21:05:25preglowplease tell me you've implemented a low power sleep mode :>
21:05:47XShocK:) no. i added RTC chip to it.
21:06:59Maxou`it's possible? o_o
21:07:09 Join linuxstb_ [0] (
21:11:37linuxstb_preglow: I load the linux kernel and rockbox into a temporary buffer first. So if there is a problem, we can still fall through to the apple firmware.
21:12:05linuxstb_I'm just reducing the amount of screen clutter (so it will be fine on the Nano), then I'll commit.
21:12:57preglowthink i'll boot linus now, so i can test this code
21:13:12*LinusN is reading the "seconds" register of the h300 rtc :-)
21:13:35preglowlinuxstb_: ipods _do_ have an rtc, yes?
21:13:51linuxstb_preglow: Yes.
21:20:52 Quit linuxstb (Read error: 110 (Connection timed out))
21:21:38 Join BirdFish [0] (n=bradbox8@
21:26:19 Quit CoCoLUS (Remote closed the connection)
21:26:40 Join petur [0] (
21:27:36 Join CoCoLUS [0] (
21:30:26preglowthe gfx card fan adjustmant prog doesn't work after i updated my ubuntu
21:31:56 Quit linuxstb_ (Read error: 110 (Connection timed out))
21:33:53 Join Paul_The_Nerd [0] (
21:34:35 Join linuxstb_ [0] (
21:38:41preglowlinuxstb_: going to commit soon?
21:40:58 Quit petur (" Want to be different? HydraIRC -> <-")
21:41:08preglowhah, my bootloader now panics :-)
21:43:27Paul_The_NerdAnd that deserves a smile?
21:44:09Paul_The_NerdIsn't a panic generally a bad thing?
21:44:16preglowoh yes
21:44:24linuxstb_preglow: Soon. Do you know how many lines of text your Nano can display in the bootloader?
21:45:55preglowlinuxstb_: nine
21:46:14linuxstb_My bootloader displays 10 in the worse case... I'll remove one :)
21:46:34preglowbut i wonder where this stokov comes from
21:46:38preglowthe scrolling thread doesn't work in any case
21:47:26linuxstb_Your ipod build has the stack overflow detection working?
21:47:53preglowi removed the threading #ifdef
21:47:59preglowso it's got thread code built in
21:48:10LinusNdoes your crt0.S fill the stack with 0xdeadbeef?
21:48:17preglowLinusN: :-)
21:48:22LinusNthere you go
21:50:05 Join mashalla [0] (
21:50:12 Quit Sandking ()
21:58:53 Join muesli_- [0] (
21:59:41linuxstb_preglow: OK, I've committed my updated bootloader.
22:00:13linuxstb_When you actually want to run rockbox, you'll need to remove the #ifdef
22:00:46 Quit Kohlrabi (Nick collision from services.)
22:00:51 Join Kohlriba [0] (
22:04:56preglowlinuxstb_: okies
22:08:41 Quit BirdFish ("studies")
22:09:07 Join ]RowaN[ [0] (
22:09:54]RowaN[whats the general feeling on moving the main rockbox exe to inside the .rockbox directory, to make the root look a bit cleaner. i know its been disucssed before
22:10:21LinusNi'm not against it
22:11:45muesli_-would be cool (just my 0.0002cents...)
22:11:49Paul_The_NerdI think it'd be kinda nice too.
22:12:11linuxstb_It means that if you don't have "show all files", then you can't browse to it to reboot. But maybe that's not an issue.
22:12:39Learfor developers it might be...
22:13:12linuxstb_Personally I would prefer a reboot option in the menu - maybe the top option in Debug
22:14:06Paul_The_NerdI always kinda wondered why there wasn't one.
22:14:20muesli_-would be cool, too (just my 0.0002cents again...)
22:14:35]RowaN[well maybe the bootloader could execute /.rockbox/rockbox.iriver if it doesnt find a copy of the file on the root.. but maybe thats a needless complication of a bootloader which is supposd to be kept simple
22:15:08preglowbut not much, i suppose
22:16:00linuxstb_Another option is to make .rockbox just hold configuration files, and put the binaries (rockbox.iriver, the plugins and the codecs) somewhere else.
22:16:01Paul_The_NerdYou could just change the display supported files so it ignored rockbox.iriver. I know it's kinda counterintuitive since it's technically supported, but that way you wouldn't have to rewrite the bootloader.
22:16:55LinusNthere is another reason to put in in .rockbox:
22:17:10]RowaN[wow just seen ipod builds on the daily page =]
22:17:29LinusNso many lamers extract only the rockbox.iriver file to the player, this way they *have* to extract .rockbox
22:17:30]RowaN[is there a public bootloader for ipods?
22:18:02preglowlinuxstb_: what the hell? does arm stacks grow upward?
22:18:16]RowaN[oh no this channel is going to end up full of ipod owners! ='[
22:18:25preglowlinuxstb_: or is you putting _stackbegin in sp an error?
22:21:11tucoz]RowaN[, join #ipodlinux for a taste of how that will be ;)
22:22:10]RowaN[hmm i guess i have slightly less reason to laugh at ipod owners now though
22:22:23tucozlinuxstb_, have the ipodlinux community shown any interest in the rockbox port?
22:22:48linuxstb_preglow: It could be a mistake.
22:22:58]RowaN[the sound quality from my ipod sucked (tinny, no bass, EQ made everything distort), but everyone else seems to think its ok.. maybe mine was faulty.. doesnt seem likely though
22:23:14linuxstb_preglow: Compare my code with startup.s in the ipodloader directory.
22:23:41linuxstb_tucoz: No, none at all.
22:23:44tucozI just wonder if the sound quality of the bundled head phones is any good, or if people just use them as a fasion statement
22:24:40tucozlinuxstb_, that is too bad. Well, maybe there will be interest in a while.
22:25:24tucoz(I meant fashion :)
22:25:26linuxstb_I don't think many people there know what Rockbox is. Or maybe they are just very happy with ipodlinux.
22:26:48tucozThat is probably true. Just curious, does it take long time to boot ipl?
22:26:54]RowaN[tucoz: it was dire with my 3rd gen ipod, original headfones or my own sony ones (which sound great with my iriver, and with my 50 cd player)
22:27:04]RowaN[it was nice of apple to refund me so promptly though =]
22:28:04preglowlinuxstb_: it could be that they've swapped the stackbeing and stackend symbols
22:28:05tucozhehe, ok. (dire means crap right?)
22:28:12preglowlinuxstb_: in rockbox, i believe stackend is the one with the higher address
22:28:27linuxstb_That could explain the corruptions then.
22:28:39preglownot really
22:29:29preglowadding another sprintf should just overwrite previously overwritten stack
22:37:28 Join arkascha [0] (
22:44:45amiconnpreglow: Regarding your register question - we can't make that bullet-proof. Gcc seems to use a scratch register by convention.
22:45:04preglowamiconn: that might become a very interesting bug one day
22:45:28amiconnIf we want to make it bulletproof, we would have to write the context switching in assembler
22:47:50Paul_The_NerdIs the directory navigation menu item related to the ability to change directories with the remote while in WPS?
22:49:59 Quit arkascha (Client Quit)
22:51:24 Quit len0x (Read error: 145 (Connection timed out))
22:52:41 Join len0x [0] (
22:54:02_FireFly_Paul_The_Nerd: not only also with the main (PLAY+LEFT/RIGHT or short+long press LEFT/RIGHT
22:54:36Paul_The_NerdI was just curious.
22:54:41Paul_The_NerdI like the "move to another folder" bit of it.
22:54:49Paul_The_NerdBut not the button press aspect which seems to be new. =/
22:55:27*Paul_The_Nerd shrugs.
22:57:06len0xwho doesn't like what? :)
22:57:23Paul_The_NerdFor example, if you're playing and reach the end of a directory, with it enabled it goes on to the next directory, but if it's disabled it doesn't seem to.
22:57:56len0xwhat's wrong with that behaviour ?
22:57:59Paul_The_NerdThat part I like
22:58:17***Saving seen data "./dancer.seen"
22:58:18Paul_The_NerdThe part I dislike is that the addition of the "Folder Skip" button combination is enabled/disabled by the same menu choice, it seems.
22:58:46len0xoption doesn't have anything to do with skip
22:59:03len0xskip to te next folder will alwasy work
22:59:35_FireFly_just tested
22:59:38Paul_The_NerdEarlier, I'd turned it off, and then Source/Bitrate didn't skip folders on me.
22:59:45Paul_The_NerdLemme double check, maybe somehow I just didn't press it right.
22:59:57len0xthere was bug at some point
23:00:12len0xI know because I coded it :)
23:00:43_FireFly_radio does now also work on the remote :)
23:01:01len0xI know :)
23:01:22_FireFly_i have a quick-hack to show the keyboard om the remote
23:01:23 Quit dpassen1 ()
23:02:01len0xI'm not completely done with radio.c yet...
23:02:07_FireFly_it was/is a part of my unofficial remote-patch
23:02:11len0xkeyboard on remote ?
23:02:26len0xdoes it fit ?
23:02:50_FireFly_yepp i have revert it to the old 4 page version for the remote
23:02:54 Join Paul_The_Nerd_ [0] (
23:03:07Paul_The_Nerd_Sorry 'bout that
23:03:17Paul_The_Nerd_My laptop had apparently come unplugged and hybernated.
23:03:51Paul_The_Nerd_Okay, so apparently I was mistaken
23:04:05Paul_The_Nerd_Does Source/Bitrate do nothing if you're playing a playlist then?
23:04:17_FireFly_len0x: look at
23:04:30DBUGEnqueued KICK len0x
23:04:30len0xPaul_The_Nerd_: obviously
23:04:44Paul_The_Nerd_Then there's some sort of bug in it
23:04:44len0xworks only for folders
23:04:55Paul_The_Nerd_I thought I was accidentally hitting the Source/Bitrate button while listening to my rock playlist
23:05:17_FireFly_Paul_The_Nerd: i have just test it it doesn't work when you play a playlist
23:05:23Paul_The_Nerd_For some reason in the middle of the music playing, a song will end, and instead of moving to the next playlist entry, it moves into either the recording folder, or the patchset folder for the old midi plugin thing
23:05:43Paul_The_Nerd_I'd thought I was accidentally triggering the folder navigation, but apparently it's something else.
23:07:18LinusNthere is something fishy with playlist playback
23:07:26LinusNsuddenly it just freaks out
23:07:32Paul_The_Nerd_Excellent! I'm not crazy!
23:07:44Paul_The_Nerd_I'd assumed it was error on my part rather than a bug.
23:07:57Paul_The_Nerd_Carelessly moving the wrong thing on the remote.
23:08:39len0xwhich build are you running ?
23:09:41 Quit Lear ("Chatzilla [Firefox 1.5/undefined]")
23:09:52Paul_The_Nerd_And choosing Info->Version gets me... the credits?
23:10:21len0xlol indeed
23:11:05_FireFly_it shows shortly the version and then the credits
23:11:13Paul_The_Nerd_I didn't know that.
23:11:21Paul_The_Nerd_I'd always just run the credits plugin.
23:11:54amiconn_FireFly_: <_FireFly_> yepp i have revert it to the old 4 page version for the remote <= it was 3-page before...
23:12:36_FireFly_amiconn : yepp it is 3 page
23:12:39 Quit mashalla ()
23:13:03_FireFly_for the remote only in my patch
23:13:28_FireFly_which is a split-off of my old unofficial remote-patch
23:13:29amiconnLinusN: Not only playlist playback is fishy...
23:14:11amiconnToday I used my iriver for real a bit. The Stop button didn't always react as expected
23:14:30amiconnGot that two times
23:14:49LinusNempty wps screen?
23:15:06preglowscroll thread still isn't working
23:15:07amiconnNo, wps was still shown, and music still played
23:15:09len0xI know how to do empty wps scree :)
23:15:40amiconnJust as if I didn't press the button, but lock was not enabled. The volume control worked
23:16:14amiconnThe first time I got it to react after a couple of tries. The second time the only working method was the 8-second shutdown
23:16:39Paul_The_Nerd_I've had that happen too amiconn, but only once so far.
23:16:39len0xa bug: play something -> to go the radio -> exit radio via stop -> press play - blanks screen
23:16:42amiconnSo I couldn't have been the button itself, otherwise the 8-second shutdown wouldn't have worked either
23:17:09len0xbecause playlist is not saved
23:17:45len0xactually no
23:18:47_FireFly_len0x: i can*t reproduce this bug
23:19:01len0xI was sure that how I got it... (may be I pressed stop instead of play indeed)
23:20:10len0xbut I just tried and couldn't either...
23:20:12_FireFly_if i press stop i see shortly the wps and then the filetree
23:20:26 Quit Paul_The_Nerd (Read error: 110 (Connection timed out))
23:20:38 Nick Paul_The_Nerd_ is now known as Paul_The_Nerd (
23:21:12len0xfor some reason playback is resumed properly now (previously when entering radio current playback position was not saved)
23:21:34len0xi.e. when I press play
23:22:14preglowi do some longs lcd_puts calls, then call yield in a loop for a long time, this should be sufficient to see some scrolling, yes?
23:22:55amiconnIs the timer interrupt already implemented?
23:23:07preglowshould that matter?
23:23:12amiconnOf course
23:23:21preglowof course it will
23:23:22amiconnThe scrolling thread uses current_tick...
23:23:43linuxstb_current_tick is a macro pointing to the RTC
23:23:56preglowoh, so it should work
23:23:59linuxstb_The ATA driver (and other places) already use it.
23:24:58_FireFly_good night everybody
23:25:40 Quit _FireFly_ ("Leaving")
23:26:24 Join MacDancer [0] (
23:29:16preglowwonderful, now my ipod just refuses to boot
23:29:25preglowafter displaying a battery symbol
23:29:38preglowi guess i'll have to assume it doesn't charge via usb in linux, then
23:29:46markunyou can't have bricked it, can you?
23:30:00preglowmarkun: that would have been, very, very hard
23:30:18preglowit looks to me like it's pretty close to impossible
23:30:27markungood to hear
23:30:38preglowit's got a bootloader in flash that can always boot in disk mode
23:30:56preglowbut still, this is somewhat strange
23:32:27preglowit plain refuses to boot
23:32:30linuxstb_preglow: Hold switch?
23:32:39preglowlinuxstb_: no
23:32:46preglowlinuxstb_: i've never used that yet
23:32:48linuxstb_It's only me that does that then :)
23:33:01preglowit displayed a battery symbol on screen when i tried to boot it
23:33:14preglowand now it doesn't do anything when i try to turn it on
23:33:21linuxstb_Mine charges fine from Linux.
23:33:29preglowi'm pretty damned sure mine did as well
23:33:43linuxstb_It does just sound like the battery is empty though.
23:33:54preglowi'm also pretty sure the battery wasn't completely emptyu
23:34:12preglowthis was just normal development, i can't have bricked it doing this
23:34:32linuxstb_So the reboot combination works, but it goes straight to a battery icon?
23:34:37preglowi'll just boot windows and see what happens
23:34:39preglownothing happens
23:34:42preglowit never turns on
23:34:45preglowdisplays nothing
23:35:25linuxstb_I would just try charging it for a while. I too would be extremely surprised if you've bricked it.
23:36:11preglowi've plugged it straigt into my box now, perhaps the lcd panel usb hub crashed or something
23:36:32preglowno, i refuse to believe it's possible to brick it doing something like thsi
23:36:38preglowi've done way crazier things
23:36:46preglowlike the time cygwin overwrote the partition table and mbr
23:37:15amiconnAll lcd monitor usb hubs I've seen so far are passive. Unlikely that it is sufficient for recharging your ipod...
23:37:35preglowamiconn: worked fine so far...
23:37:42 Join _DangerousDan [0] (
23:38:13preglowi got it going again
23:38:14amiconnHow do you know it charged for real?
23:38:26preglowamiconn: well, i assume it does charge for real when it claims to be charging for real
23:38:31preglowamiconn: it also reported full battery
23:38:35preglowbut i've got it going again
23:38:44preglowthere was some strange lockup in the apple flash
23:38:56preglowi must have not held the reset combo for long enough
23:39:04CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:39:04*preglow wipes away sweat
23:39:23*linuxstb_ passes preglow a cold beer
23:41:01preglownow there's something i could have used right about now
23:41:15preglowoh well, i'll continue tickling the threading
23:41:32preglowfun part is the battery is almost full
23:41:56]RowaN[err whaddu i have to do to get my bmp showing in the wps now
23:42:41amiconnpreglow: There's some thread and timer tick test code in firmware/test/kernel
23:42:58amiconnIt seems that it needs a gdb stub though...
23:43:00 Quit ehntoo ("Leaving")
23:43:08Paul_The_Nerd]RowaN[: I think it needs to be in a folder with the same name as the .wps file
23:43:14preglowi'll just spawn a couple of threads that does simple stuff
23:43:22Paul_The_Nerdin .rockbox/wps/ you have bob.wps and a folder named bob with all the .bmp files
23:43:26amiconnDoes the ipod have an rs232 interface?
23:44:08linuxstb_Or at least the 3G and earlier models did. I think the 4G models do, but it may not have a linux driver.
23:44:35linuxstb_I think it's used to connect to the various ipod peripherals
23:45:36]RowaN[Paul: still not working
23:45:55]RowaN[do i have to put the image pathname in the wps now?
23:49:58preglowi created two threads that just incremented variables
23:50:05]RowaN[the info on is out of date
23:50:09preglowbootloader now creashes, runs, crashes, runs, crashes, runs
23:51:09 Join HCl [0] (
23:51:25preglowi am a fool
23:51:42]RowaN[can anyone help me with wps?
23:52:09 Quit muesli_- (Read error: 110 (Connection timed out))
23:52:31preglowok, threading works
23:52:50preglownote to self: rockbox does not handle threads exiting very well
23:52:54 Quit DangerousDan (Read error: 110 (Connection timed out))
23:52:55linuxstb_preglow: Nice.
23:53:27Paul_The_Nerd]RowaN[: I'm not 100% certain. I moved my .wps file into the WPS folder, moved my images into a subfolder and everything worked fine
23:53:37linuxstb_Does that mean you can tick the first box here?
23:53:37 Join ehntoo [0] (
23:53:43 Quit ehntoo (Client Quit)
23:53:49preglowyeah, but i'll give it a few more tests first
23:54:06preglowscrolling still doesn't work for some raso
23:54:06Paul_The_NerdThe .wps file still simply calls it backround.bmp and it's in .rockbox/wps/paul/background.bmp with the .wps being .rockbox/wps/paul.wps
23:54:34linuxstb_Are the necessary LCD functions implemented?
23:54:40 Join ehntoo [0] (
23:54:59preglowperhaps, perhaps not :)
23:55:14]RowaN[hmm thats how i have it, but image not showing
23:55:27linuxstb_preglow: I would guess not - I only did the very minimum to get the text displaying.
23:55:36Paul_The_NerdWhat's the line in your WPS that loads the image?
23:56:55]RowaN[ah whoops
23:57:03]RowaN[i renamed the image to rowan.bmp
23:57:06]RowaN[forgot to change wps
23:57:26]RowaN[a shoolboy error
23:57:37Paul_The_NerdIt happens
23:58:15preglowlinuxstb_: looks like it's all implemented to me

Previous day | Next day