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 2008-12-30

00:00:51 Quit ender` (" If the art world was ran by the RIAA, you would have to lease special glasses if you wanted to see a painting. -- Rodney Ca")
00:06:29 Quit solexx_ (Read error: 110 (Connection timed out))
00:14:10mcuelenaeregevaerts: I get several USB GET_DESC requests (1, 2 and 3) on the Onda; does that imply that the host receives stuff I send through EP0?
00:17:07 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
00:17:56 Part consquigulator
00:18:25 Join BigBambi_ [0] (
00:18:26 Quit BigBambi (Read error: 110 (Connection timed out))
00:20:10 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
00:24:47 Join casainho [0] (n=chatzill@
00:25:02casainhohello again :-)
00:25:14casainhoI am getting this warning: warning: implicit declaration of function ‘disable_irq_save’. Why? :-)
00:26:05mcuelenaereit means you don't have disable_irq_save() defined in system-target.h or you haven't included that file
00:27:48 Quit tvelocity (Remote closed the connection)
00:30:05casainhomcuelenaere: But looking at other targets, they don't disable_irq_save() defined on system-target.h??
00:30:44mcuelenaerehmm they have disable_interrupt_save() defined, so my mistake
00:30:51mcuelenaeredid you include system.h?
00:31:41casainhothe error starts here: CC firmware/kernel.c
00:31:43casainho/home/cas/Documentos/rockbox_player/rockbox/firmware/kernel.c: In function ‘tick_add_task’:
00:31:45casainho/home/cas/Documentos/rockbox_player/rockbox/firmware/kernel.c:100: warning: implicit declaration of function ‘disable_irq_save’
00:32:02casainhoso, system.h on kernel.c file?
00:32:45mcuelenaeredisable_irq_save() seems to be a #define in system-target.h:
00:33:21 Join einhirn [0] (
00:37:16 Join itcheg [0] (i=62db4767@gateway/web/ajax/
00:39:07 Quit Zambezi (Remote closed the connection)
00:39:37casainhowell, it's on target/arm-arm-system.h
00:40:10casainhomy processor is an ARM9, so, I don't know why I get that warning...
00:42:15casainhoanother warning: firmware/panic.c:51: warning: implicit declaration of function ‘disable_fiq’
00:42:23casainhoshould I be worried with them?
00:42:25 Quit Zagor ("Leaving")
00:46:18***Saving seen data "./dancer.seen"
00:48:48 Quit casainho ("ChatZilla 0.9.84 [Firefox 3.0.5/2008121622]")
00:48:53 Join Zambezi [0] (
00:51:43 Quit Hillshum ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
00:56:11 Quit bmbl ("Woah!")
01:05:43 Quit culture (Read error: 110 (Connection timed out))
01:15:09 Quit scorche (Nick collision from services.)
01:16:02 Join scorche [0] (i=Blah@rockbox/administrator/scorche)
01:24:25gevaertsmcuelenaere: you mean in the bDescriptorType field? If so, then yes, that would mean that the host gets something back
01:24:58mcuelenaerethat's good news :)
01:26:00gevaertsindeed :)
01:26:47mcuelenaereonly problem is that I just did something which breaks USB and I can't figure out what :/
01:27:43 Join offset [0] (
01:29:16mcuelenaereok, got it to work back. I get STRING 0 & STRING 3; what does this mean?
01:31:00gevaertsIt means that the OS asked what language your strings are in and what the serial number is
01:31:39mcuelenaereit stops communicating after that
01:31:45mcuelenaereI'll post a log
01:32:17gevaertsI think the next step would be a SET_ADDRESS
01:32:33 Quit jhMikeS (Nick collision from services.)
01:32:39 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
01:33:31 Quit nplus (Remote closed the connection)
01:34:27mcuelenaereI'm not sure at what step I disconnected and the process started over
01:34:32 Join BigBambi [0] (
01:34:47mcuelenaere(and the GET_DESC x messages are prefixed with usb_core: of course)
01:36:39mcuelenaereoh and could the absence of a call to usb_core_init() do anything harmfull?
01:36:58 Quit BigBambi_ (Read error: 110 (Connection timed out))
01:37:24*gevaerts checks
01:39:54gevaertsI think it won't matter, but to be sure maybe start with only USB_CHARGING_ONLY defined in firmware/export/usb_core.h (so no USB_STORAGE or USB_SERIAL). That uses no non-0 endpoints at all, so it sgould simplify things
01:40:06mcuelenaerehmm ok
01:40:30mcuelenaereand the absence of the usb_core_init() call is due to usb_enable(), which is called from the USB thread AFAICS
01:40:36mcuelenaereso there's something peculiar there
01:41:00gevaertshm, indeed.
01:41:12 Join iPlay [0] (n=569b26f8@gateway/web/cgi-irc/
01:41:29mcuelenaereUSE_CHARGING_ONLY is already defined in BOOTLOADER
01:41:42gevaertsIf you run with only USB_CHARGING_ONLY, maybe check firmware/usb.c to see if the calls to usb_core_enable_driver() are correct
01:42:30iPlayanyone help me out. ipods running rockbox np. but i cant get the utility to install at all
01:42:37mcuelenaereah USB_DRIVER_MASS_STORAGE isn't #ifdef'ed
01:42:39iPlaycan't find a .dll when i launch the exe
01:43:52gevaertsiPlay: have a look at
01:43:58 Quit MethoS- (Remote closed the connection)
01:44:09mcuelenaeregevaerts: is this correct?
01:44:29iPlaynice one
01:45:04mcuelenaerehmm I seem to have the same behaviour with it
01:46:15gevaertsmcuelenaere: it's correct, although I think that USB_CHARGING_ONLY needs the same treatment, and from reading the code I don't think it really matters (although just doing it for correctness is a good idea anyway)
01:48:59 Quit tessarakt (Read error: 110 (Connection timed out))
01:49:09gevaertsmcuelenaere: does your code call usb_start_monitoring() somewhere?
01:49:30 Join tessarakt [0] (
01:49:56mcuelenaeregevaerts: you mean like in the bootloader? no
01:50:22gevaertsOK. That explains why you don't get the correct events in the usb thread
01:50:30gevaerts(unless you get them some other way)
01:51:11mcuelenaereah ok, but it seems like the USB tick driver doesn't get them either
01:51:34mcuelenaereor there's something wrong with logf()
01:52:13gevaertsThere's this if(usb_monitor_enabled) in there
01:52:59gevaertsSo it needs usb_start_monitoring() somewhere to get it to actually do something
01:53:34mcuelenaereah you're right, I missed that
01:54:56gevaertsMaybe copy the usb logic from bootloader/gigabeat-s.c
01:56:02mcuelenaeredoes the main thread has to acknowledge the USB request too?
01:58:22gevaertsYes and no. SYS_USB_CONNECTED actually means that the UMS driver wants the disk for itself. As long as you're not far enough along to get near that it doesn't matter much
01:58:58mcuelenaereah ok. It seems that usb_core_init() gets called now
01:59:04mcuelenaerebut it doesn't help me any further though
02:00:52gevaertsdo you handle USB_REQ_SET_ADDRESS correctly? i.e. do you tell the hardware what its new address is, or does it handle that itself?
02:01:19gevaertsThat's actually usb_drv_set_address() to be precise
02:01:43mcuelenaereyes I set the REG_USB_REG_FADDR register
02:01:50mcuelenaerein that function
02:01:59mcuelenaerethat should be enough according to the datasheet
02:02:58mcuelenaerethat register does have this described at its seventh bit: Set when FAddr is written. Cleared when the new address takes effect (at the end of the current transfer).
02:03:25mcuelenaerebut I don't presume that's of any importance?
02:04:34gevaertsI don't think so, but I don't actually know...
02:04:47 Quit HellDragon (Read error: 104 (Connection reset by peer))
02:05:09mcuelenaereit never seems to get handled in any of the code from the manufacturer..
02:05:27 Join HellDragon [0] (
02:05:34gevaertshm, sounds a bit like the tcc controller then
02:05:36*mcuelenaere thinks it's something with ACK'ing USB requests
02:05:57gevaertsThat one handles SET_ADDRESS internally, so you don't actually need to do anything
02:06:12mcuelenaereI think the ZVM's chip has the same
02:06:28gevaertsIt's not simply ACKing. If that was the case you'd only ever see GET_DESC 1
02:07:21mcuelenaereperhaps it's not sending the whole message?
02:07:30mcuelenaereie it doesn't send every last packet or so
02:07:35 Quit tessarakt (Connection timed out)
02:07:54 Join tessarakt [0] (
02:07:54gevaertsAll these should fit in a single packet
02:08:26mcuelenaerehmm true
02:08:28gevaertsThe thing is that your log stops at SET_ADR, and repeats after that (so I suspect a reset there).
02:08:44mcuelenaereit doesn't really stop at SET_ADR
02:09:16mcuelenaereI didn't give all information, because I need to copy it manually :)
02:09:33mcuelenaere(type every character on the screen into pastebin)
02:09:52mcuelenaereit seems to give up on a GET_DESC 2
02:09:53gevaertsIs there usb activity after it?
02:10:17mcuelenaereyes, a GET_DESC 1,2,3,3 and 2
02:10:23gevaertsok. that's probably the "complete" USB_DT_CONFIG then
02:10:57*gevaerts tries to remember if high-speed control packets are 64 bytes
02:11:07mcuelenaereehm I forced the lowest speed
02:11:17*mcuelenaere doesn't know the correct term
02:11:23mcuelenaereso they're all 64 bytes
02:11:26*gevaerts tries to remember if full-speed control packets are 64 bytes :)
02:11:39mcuelenaereyes they are:)
02:11:42mcuelenaereI messed up with bulk packets
02:11:49mcuelenaerethose can be either 64 or 512
02:11:53mcuelenaeredepending on the speed
02:12:01gevaertsor less :)
02:12:03mcuelenaereand control packets are always 64bytes AFAIK
02:12:13mcuelenaereyes or less :)
02:12:58 Quit MarcGuay (" ajax IRC Client")
02:13:14mcuelenaerewhy is usb_core_ack_control() sometimes an usb_drv_send() call and sometimes an usb_drv_recv() call?
02:13:30 Quit bertrik ("Leaving")
02:13:31mcuelenaeredoesn't the USB chip _sends_ an ACK?
02:13:39gevaertsTry logging config_descriptor.wTotalLength at the end of case USB_REQ_GET_DESCRIPTOR:
02:14:52gevaertsusb_core_ack_control() may be a bit of a bad name. This is about acking the transaction, not the individual packet
02:15:03mcuelenaerewTotalLength = 9
02:15:14mcuelenaereah you meant the contents probably?
02:15:40mcuelenaerehmm sorry I thought I was logging the length of it
02:16:42gevaertsI basically want to find out if this descriptor is more than 64 bytes or not
02:17:06mcuelenaereand the writeFIFO() also gets called with length=9 so that's correct
02:17:09gevaertsIf it is, throw out everything you can so you have only the charging only thing. That should fit
02:17:48gevaertsIs it also 9 for the second GET_DESC 2? That one should be larger
02:18:34mcuelenaereit seems to be 9 for all GET_DESC :/
02:18:36*gevaerts is confused...
02:19:04mcuelenaereI added 'logf("G: %d", config_descriptor.wTotalLength);' at line 618 in usb_core.c
02:20:39 Quit iPlay ("CGI:IRC (EOF)")
02:20:42 Join crackerizer [0] (n=7d1a8eb8@gateway/web/cgi-irc/
02:21:34gevaertsI think that means that no driver is enabled (i.e. the appropriate usb_core_enable_driver()s didn't get called). I think you get a non-compliant device then...
02:22:04mcuelenaerehmm that would mean there's nothing wrong with the USB driver itself :)
02:22:10mcuelenaerelet's hope you're right :)
02:22:28crackerizeris somebody here own an iaudio m3??
02:22:46crackerizeror please tell me what is the build target for it??
02:23:18gevaertsI'd recommend double-checking usb.c for the right usb_core_enable_driver(xxx,true) calls
02:23:39mcuelenaereI get the Rockbox Media Driver windows thingy
02:23:47gevaertscrackerizer: I don't understand the question...
02:23:57mcuelenaereyou were right, just simply enabling the CHARGING_ONLY did the trick :)
02:24:11gevaertsOK. Now go back to storage :)
02:24:23mcuelenaerenow I just need to implement EP1 support :)
02:25:40 Join CaptainKewl [0] (
02:26:09crackerizergevaerts: I'd like to know what is the compiler's target for building rockbox for m3.
02:26:17 Join Barahir_ [0] (
02:26:40 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
02:27:02gevaertscrackerizer: you mean in tools/configure?
02:27:05*mcuelenaere thinks he'll take the shortcut and directly implement DMA support
02:28:19crackerizergevaerts: i mean what cpu which m3 uses..
02:28:37gevaertsah, ok
02:29:00gevaerts knows all :)
02:29:09crackerizerI dont see this model on the wiki page.
02:29:30gevaertsI do...
02:30:02crackerizerThank you gevaerts :)
02:30:39 Quit matsl ("Leaving")
02:31:31gevaertsNo problem :)
02:33:18 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
02:34:47 Quit BigBambi (Read error: 110 (Connection timed out))
02:35:46 Join BigBambi [0] (
02:35:52 Join Thundercloud [0] (
02:37:26 Quit crackerizer ("CGI:IRC (EOF)")
02:40:10 Quit Barahir (Read error: 113 (No route to host))
02:42:35 Quit jhMikeS (Nick collision from services.)
02:42:41 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
02:44:02 Quit moos ("Rockbox rules the DAP world")
02:46:22***Saving seen data "./dancer.seen"
02:48:48 Quit evilnick (" ajax IRC Client")
02:49:52 Join Barahir [0] (
02:53:10 Quit einhirn (Read error: 110 (Connection timed out))
02:55:57gartral1no, the processor arch type, i think
02:56:07gartral1oops, wrong tab
02:58:56 Quit tessarakt ("Client exiting")
03:04:20 Quit Barahir_ (Read error: 113 (No route to host))
03:10:51 Quit mcuelenaere ()
03:11:59 Quit itcheg (" ajax IRC Client")
03:16:17 Quit JdGordon (Remote closed the connection)
03:17:41 Join JdGordon [0] (
03:20:20 Quit Slack__ ("Ex-Chat")
03:23:32 Quit Bensawsome (K-lined)
03:26:24 Join Strife89 [0] (n=michael@
03:28:49 Quit perrikwp (" ajax IRC Client")
03:29:45 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
03:31:23 Join nosa-J [0] (
03:35:14 Quit DerDome (Nick collision from services.)
03:35:17 Join DerDome1 [0] (
03:35:29 Nick DerDome1 is now known as DerDome (
03:46:33 Join gromit`` [0] (
03:46:41 Quit gromit` (Connection timed out)
03:52:18 Nick fxb is now known as fxb__ (
04:04:55 Quit Thundercloud (Remote closed the connection)
04:06:07 Join tyfoo2 [0] (
04:06:38 Quit Horscht ("User was distributing pornography on server; system seized by FBI")
04:20:37 Quit tyfoo (Connection timed out)
04:20:41 Quit faemir ("Leaving")
04:22:13 Quit Strife89 ("Bed.")
04:27:34 Join Slack_ [0] (
04:29:29Unhelpfulamiconn: i can do it that way. i'm already going to have to put plugin-related stuff in bmp.c is it's going to be shared with pluginlib, so adding an init function doesn't *really* upset me... but it also makes one more difference in the plugin between using the core loader and the pluginlib loader. it would be nice if the only difference was essentially the name you call, ie bmp_load_file vs rb->etcetcetc
04:43:49 Join Seed [0] (
04:46:24***Saving seen data "./dancer.seen"
04:47:04 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:51:57 Part gartral1
04:55:32 Join Barahir_ [0] (
05:00:46 Join anewuser [0] (i=anewuser@unaffiliated/anewuser)
05:12:01 Quit Lss (Read error: 54 (Connection reset by peer))
05:12:23 Quit Barahir (Read error: 110 (Connection timed out))
05:12:37 Quit idshark (Read error: 104 (Connection reset by peer))
05:30:00 Join toffe82 [0] (n=chatzill@
05:39:48 Join itcheg [0] (i=62db4767@gateway/web/ajax/
05:46:47 Quit tyfoo2 ("Carpe diem")
05:55:44 Join Bensawsome [0] (
05:56:07 Nick Bensawsome is now known as Bensawesome (n=Bensawso@unaffiliated/bensawsome)
06:00:11 Quit itcheg (" ajax IRC Client")
06:01:58 Join theredone [0] (n=18837ab0@gateway/web/cgi-irc/
06:02:53theredonei understand you are the answer guys
06:04:12theredonei have an i pod g5 with rockbox and a dlo docking station i cant get video with it and the remote doesent work any way to fix it
06:04:57 Quit at0m ("CGI:IRC (EOF)")
06:05:10JdGordontheredone: rockbox doesnt support tv out
06:05:33scorcherockbox does not support the video out at the for the remote part, you can check the tracker for a patch for the apple accessory protocol
06:05:45scorcheactually, i take that back...i think that was recently committed..
06:05:51JdGordonthe IAP patch was commited...
06:06:14 Join Actium [0] (
06:07:00Actiumhiya, I have a sansa c200 that I just flashed to firmware v01.01.00P and rockbox utility won't detect it
06:07:20ActiumI thought maybe it was some of that msc/mtp mode nonsense because before I had a v01.05.00P so I reflashed it to this version
06:07:26Actiumbut rb util still doesn't find it
06:07:35Actium"No Sansa Detected!"
06:08:03 Quit amiconn (Nick collision from services.)
06:08:05 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
06:08:07theredonewerre can i find that
06:08:26theredonewere can i find that
06:08:48 Join pixelma_ [0] (n=pixelma@rockbox/staff/pixelma)
06:09:22theredoneyes please
06:10:03JdGordonfind what?
06:10:27 Quit pixelma (Read error: 110 (Connection timed out))
06:10:28theredoneiap patch
06:10:52scorcheit was committed...meaning it should be in the latest builds
06:11:38theredonehmm ok well it doesnt work any leads i can follow to git it fixed
06:13:20scorcheare you running the latest build?
06:13:48theredoneya downloaded it yester day
06:14:33theredoneunless something happened overnight
06:14:57 Join itcheg [0] (i=62db4767@gateway/web/ajax/
06:15:47theredonethis is the one i got
06:24:36theredonewell ill ceep lokin and check back later
06:25:14 Quit itcheg (" ajax IRC Client")
06:37:01Actiumsansapatcher can't find it either!
06:37:22LloreanActium: It's in MSC mode, and the tool is being run with admin privileges?
06:37:41LloreanWhat OS are you on?
06:37:44Actiumwindows XP
06:37:50ActiumI'm trying it on a different machine right now to rule that out
06:38:36Actiumalso you forgot to package mingwm10.dll with rbutil
06:38:42Actiumor whoever does that
06:38:46LloreanNo, it wasn't supposed to be packaged with it.
06:38:48LloreanIt was compiled wrong.
06:38:59Actiumwell luckliy I've got that dll
06:39:08LloreanThe utility shouldn't require any external dlls.
06:39:38Actiumyeah, stuff compiled w/ mingw shouldn't require mingwm10, I thought that was strange
06:39:39LloreanBut due to the holidays, it's a bit slow in getting a fixed version arranged and posted.
06:41:32Actiumnope, other comp says no sansa detected
06:41:35ActiumI shall try with my old sansa
06:42:55Actiumnew sansa fails
06:43:34Actiumold one has firmware 01.00.03A new has 01.01.00P
06:45:14LloreanI don't have my c200 on hand, but I'm pretty sure I installed it with 01.05.00P. That's the one where you need to hold left while hold is on to manually enter MSC mode during connection, right?
06:46:05Actiummy new one used to have that firmware, then I flashed to 01.01.00P, because it stil couldn't find it
06:46:12Actiumalso, I didn't have to do the hold left nonsense
06:46:23LloreanIn 1.01 or 1.05?
06:46:24Actiumprobably because this laptop has mtp drivers removed
06:46:28***Saving seen data "./dancer.seen"
06:46:42LloreanIt had an actual drive letter, and didn't simply say "Sansa c200"?
06:48:35Actiumand it had one for the microsd slot
06:48:38Actiumtried this on 2 different comps
06:49:12LloreanAnd you simply get a message saying the player wasn't found?
06:50:10advcomp2019v1.01.00 has a menu option for msc mode that is the why they went back to it from v1.06.00(if i remember the numbers right)
06:50:34 Quit agaffney (Read error: 104 (Connection reset by peer))
06:50:36ActiumI did the menu option for msc mode, it's in msc mode
06:50:44Actiumalso this is not my screenshot but it's exactly the error I get
06:51:41LloreanWell, unfortunately there's probably not much more we can do for you unless you can figure out what makes your player different from all the others it does work with.
06:52:05Actiumis there anywhere else I can obtain alternate firmware to try flashing it with?
06:52:33LloreanThere's the SansaLinux project, but they simply use our tools so it won't help.
06:52:37Actiumis there a method of extracting the firmware off of my other player that does work so that I may load it onto this one?
06:53:02LloreanIt's probably not the firmware that's the problem.
06:53:07ActiumAlso, what does the letter mean after the firmware? If I try to flash mine with a 01.01.00A (instead of P) will it brick?
06:54:36LloreanThe letter represents the region, I believe. It shouldn't brick, they're really quite difficult to brick. The best you can do, really, is disable them until you go through the recovery procedure.
06:55:18 Join agaffney [0] (n=agaffney@gentoo/developer/agaffney)
06:57:07ActiumDisable what?
06:57:12ActiumAlso, where's the sansapatcher source code
06:57:26LloreanActium: disable the player. As in, it'll be unusable until you recover it.
06:57:36LloreanAnd the sansapatcher source code is in SVN along with all the other project source code.
06:59:47 Join thegeek_ [0] (
07:00:33 Quit Aurix_Lexico (Read error: 110 (Connection timed out))
07:01:42 Quit _lifeless (Remote closed the connection)
07:01:45 Join __lifeless [0] (n=lifeless@
07:05:18 Quit nosa-J (Read error: 110 (Connection timed out))
07:05:24 Join nosa-J [0] (
07:09:31 Quit JdGordon (Read error: 104 (Connection reset by peer))
07:12:28 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
07:15:13Actiumstrange, I flashed it twice with 01.01.00A firmware and it stil says 01.01.00P
07:15:55 Quit XavierGr ()
07:16:01LloreanYou aren't really flashing it, the firmware is just stored in a different hidden partition in the storage.
07:17:10LloreanActium: As I said, it's probably not the firmware that's the problem. You may wish to investigate elsewhere. Make sure the USB PID/VID are what the patcher expects, etc.
07:17:12Actiumstorage is flash memory right
07:17:36 Quit thegeek (Read error: 110 (Connection timed out))
07:17:36Llorean"flashing" as a verb almost always refers to flashing ROMs when discussing computers, not writing to flash memory.
07:19:19LloreanIt's a more important distinction here because for several of the players there is a way to actually flash Rockbox to the ROM for other benefits than on-disk installation.
07:22:43 Part toffe82
07:27:27Actiumthe device manager pages look almost identical
07:28:02Actiumhow can I check the ids in windows? no lsusb :|
07:29:06 Join Darksair [0] (n=user@
07:30:33LloreanThere's no simple way, as far as I know.
07:50:52JdGordonyou can get it with a bit of pain in device manager (i tihnk)
07:56:18 Join ender` [0] (
07:57:16 Quit theredone ("CGI:IRC (Ping timeout)")
08:01:48 Quit nosa-J (Read error: 110 (Connection timed out))
08:06:08 Quit Darksair (Read error: 104 (Connection reset by peer))
08:09:11JdGordonwhy the heck does zxbox seem to check the buttonbar height but not use it?
08:12:56 Join Darksair [0] (n=user@
08:14:00 Join Rob2222 [0] (
08:16:20JdGordonamiconn_: do you think there is any problem with fiddling keeping an assumption in the buttonbar code that it will never be used with targets with a remote lcd?
08:32:22 Quit Rob2223 (Read error: 110 (Connection timed out))
08:33:03 Quit _Auron_ ("Infinity repeatedly denies rumours of plotting with zero to bring down the Universe.")
08:35:09 Join _Auron_ [0] (
08:37:29 Quit perrikwp (" ajax IRC Client")
08:39:08 Join tophe [0] (
08:39:24 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
08:46:29***Saving seen data "./dancer.seen"
09:03:30 Quit amiconn (Remote closed the connection)
09:03:30 Quit pixelma_ (Remote closed the connection)
09:04:31 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
09:04:31 Join pixelma [0] (n=pixelma@rockbox/staff/pixelma)
09:11:20 Quit Actium (Read error: 110 (Connection timed out))
09:18:03 Join miepchen^schlaf [0] (
09:19:05 Join Actium [0] (
09:29:14 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
09:30:03amiconnJdGordon: Imho this assumption would be bad. The main idea behind the apps/gui/ stuff is multiple display support
09:40:33JdGordonyeah.. but the buttonbar is pretty target specific...
09:41:45 Join anarky97 [0] (n=kpr@
09:41:59anarky97Alguien aqui?
09:42:42JdGordonenglish only here
09:42:51anarky97No comprende,senor
09:43:03anarky97Tengo una pregunta
09:43:09 Quit _Auron_ (Read error: 110 (Connection timed out))
09:43:12anarky97Lol, I'm screwin w/ ya, I speak english
09:43:34anarky97This isn't necessarily only for rockbox
09:43:39JdGordonyou know you just lost at least potential answereed for being a dick
09:43:39anarky97But it happens in #rockbox a lot
09:44:06anarky97Bein a dick? Wtf
09:44:13anarky97I'm sorry. Did I call you a faggot?
09:44:27anarky97Or do you throw tantrums like that when someone makes a joke
09:44:34anarky97Fuck you.
09:44:41anarky97I don't need your answer then, you little pussy bitch.
09:44:47anarky97Cocksucking faggot. There, happy?
09:44:48 Part anarky97
09:59:29JdGordonkugel.. if you want 9603 (?) commited... join in the next 15min....
10:00:43 Join moos [0] (i=Mustapha@rockbox/staff/moos)
10:02:41 Join einhirn [0] (
10:07:40 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
10:22:09 Join bertrik [0] (
10:32:21 Join fredddy [0] (
10:40:50 Quit bmbl ("Woah!")
10:43:58 Join culture [0] (
10:46:33***Saving seen data "./dancer.seen"
10:49:34 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
11:20:19 Join einhirn_ [0] (
11:22:05JdGordon... the iap patch had heaps of tabs :/
11:23:06 Join tvelocity [0] (
11:24:00 Join Torne [0] (
11:24:23Torneis there anything in a rockbox binary that actually identifies it as being a rockbox binary?
11:24:39Tornea string at a known offset or something?
11:25:19 Join w1ll14m [0] (
11:25:47w1ll14mi was wondering.... is there a way to improve the gigabeat S antenna ?
11:26:32Tornei'm looking at ways to support having rockbox in the ipod OSOS image and still be able to boot the retail firmware, and considering patching the bootloader to detect that this has been done automatically
11:26:48 Join kugel [0] (n=chatzill@unaffiliated/kugel)
11:26:58kugelJdGordon: here I am!
11:27:02Tornebut there's not much to identify rockbox.bin as being rockbox..
11:27:11*kugel was busy trying to get his driver license
11:27:32w1ll14mKugel: and did you ?
11:27:45kugeluhm, no
11:27:49w1ll14mto bad....
11:28:02JdGordonkugel: too slow :D
11:28:10w1ll14mi'm going for take #4 at 7 january
11:28:28JdGordonna, I looked at the patch and I remembered why i didnt like it... I'm generally annoyed with how the bars and lists [dont] work together
11:28:29w1ll14mi hope i will get it right then.....
11:29:12JdGordongrr... why isnt my SYS_BLAA event getting into default_event_handler() ?
11:29:35kugelJdGordon: yea, but it still has nothing to do with my patch
11:29:45JdGordonyes and no...
11:29:46Tornehm, i guess i could modify crt0-pp to have some identifying text :)
11:29:54JdGordonadding that hide_bars param is a bad idea...
11:30:00kugelif anything, I actually move the bar setup into viewport management, and away from list management
11:30:05Tornethere's that padding gap for the exception vectors..
11:30:59kugelI think it's better to tell it to hide the bar once at init, instead of constantly re-init'ing lists because of taht
11:31:05w1ll14mTorne, what are you trying to achieve
11:31:17JdGordonkugel: what I want to do is make the buttonbar a "wgdet" for screens so they have to setup the rest of the screen viewport for it, and statusbars should be considered "outside" the usable screen which has to be forcibly removed if thats wants
11:31:40Tornew1ll14m: i'd like to stuff rockbox into the ipod's OSOS image, so that the ipod rom loads it into ram, but still use the rockbox bootloader so it can support loading the the retail os from disk
11:32:05w1ll14mhmm you mean using rockbox fully from ram
11:32:13Tornewell no, it's just avoiding loadign two OS images
11:32:18Tornethe ipod rom loads whatever is in OSOS
11:32:26kugelJdGordon: then the bar code needs to know about all viewports, sounds like unneeded complexity to me
11:32:30w1ll14mactualy, only the bootloader is loaded in the beginning
11:32:32Torneby default that's the retailos, with the rockbox bootloader pasted on the end
11:32:39Torneno, it loads teh whole thing
11:32:44kugeland I don't think I break your plans
11:32:51Tornethen the bootloader just branches back to the start if it wants to run retailos
11:33:20Torneit means startig rockbox involves loading the whole retailos then immediately overwriting it with rockbox
11:33:25w1ll14mi've always had my bootloader with both
11:33:32JdGordonkugel: no it wont...
11:34:00Tornew1ll14m: all ipodpatcher does is change the entry point address of the OSOS image
11:34:10w1ll14mit does
11:34:17kugelthen I don't see what's wrong with it
11:34:20Torneit still loads the retailos into ram
11:34:26w1ll14mnot fully
11:34:40w1ll14mafter you boot retail you'll see it takes ages to load it
11:34:47Tornethat's its startup procedure
11:34:49Torneit's jst slow :)
11:34:55kugelbetter to have 1 function which is able to setup viewports as you need them, instead of messing with several parts of the code to finally reach your goal
11:34:57Tornethe whole of the OSOS image is in ram, is my oint
11:35:23Torneit hasn't moved itself around into iram or any of the other things it has to do on start
11:35:27kugelparticulary if it saves multiple inits
11:35:40Tornebut it has loaded up a big blob of stuff that i rarely want to use.
11:35:47w1ll14mhmmm ....
11:35:48Tornewhen it could've instead loaded rockbox :)
11:35:52w1ll14mnever had that kind of problem
11:35:56Torneit's not a problem
11:35:57w1ll14mi use a ipod video
11:36:02Torneit's just a slight boot time improvement
11:36:09w1ll14mhmmm ...
11:36:20Torneif rockbox is in the OSOS image, isntead of retailos, it boots a little faster
11:36:24Tornebut then you can't start retailos
11:36:39Torneso, i'm looking at fixing hte bootloader to allow that
11:36:50w1ll14mthat will be a hell of a job
11:36:53Tornenot really
11:37:01Torneyou can do it as a five line bootloader hack
11:37:09Tornebut that's strictly a hack
11:37:12w1ll14mnot familiar with that
11:37:14Torneit can't then load rockbox from disk any more
11:37:16kugelJdGordon: look, given the buttonbar was a widget, the buttonbar code would just do viewport_set_defaults with the settings->buttonbar as the hide and pass it to the lists
11:37:26kugelthat will work nicely with your widget idea
11:37:27Torneand thus that's never going to go into svn
11:37:33kugelI imagine at least
11:37:41Tornethis guy did it :)
11:38:01Tornei've tried his way and it works, but is a hack
11:38:10w1ll14mit also makes upgrading rockbox a lot more complicated, so i don't think they will do that
11:38:13Torneso, seeing abotu inventing a better way that doesn't break any existing functionality for people who do it normally
11:38:16Tornew1ll14m: it wouldn't be mandatory
11:38:23 Quit einhirn (Read error: 110 (Connection timed out))
11:38:27Tornethe idea is that it will work exactly the same as it does now
11:38:34 Join bluebrother [0] (n=bluebrot@
11:38:38Tornebut if you switch the OS images over with ipodpatcher it will then work my way
11:38:43Tornewithout requiring a hacked/recompiled bootloader
11:38:45kugelJdGordon: I think your idea is rather unrelated to my patch, and it's still a idea and there's no implementation (i.e. possible future code), and thus no reason for rejection, IMHO
11:39:06*JdGordon shouldnt have brought up the patch....
11:39:08Torneit means the bootlaoder needs to be able to tell if the image in ram is rockbox, hough
11:40:01Torneit can identify the retail image because it says "portalplayer" at offset 0x20
11:40:21Tornebut there's no comparable method for rockbox.bin
11:40:24kugellucent: ping
11:40:29w1ll14mthat would mean you have to write some way of detecting the image at the place of the OSOS image
11:40:41Torneyes. it already does for retailos
11:40:58Torneit doesn't just blindly branch back to 0
11:41:02Torneit checks first :)
11:41:21Torneso atm if you replace the OSOS image but keep the rockbox bootloader it will just fail if you have hold switched on
11:41:23kugelJdGordon: if it's a tough decision for you we can always mail to the -dev list and ask other people
11:41:41JdGordonoh its not...
11:41:45Torneunless you saved retailos as apple_os.ipod
11:41:49Tornewhich it tries first.
11:42:01JdGordonthe changes to the bars bit is the only reason i wont it out...
11:42:04w1ll14mhaven't gone that deep into the bootloader
11:42:07kugelapart from that, my patch is pretty much a requirement for a future customlist implementation which does it properly
11:42:12Tornes'in bootloader/ipod.c
11:42:15Torneit's pretty straightforward
11:42:20w1ll14mhold on
11:42:44kugelJdGordon: in the current code list/menu code handles bar setup, which is even worse I suppose
11:42:48 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
11:42:57JdGordonyes it is.. which is why i want to do something about it
11:43:06JdGordoncommiting that patch as it is will cause more work...
11:43:20JdGordonI'm working on a statusbar manager now which should clear everything up
11:43:20kugeldo you think so? explain please
11:44:01w1ll14mTorne: i see
11:44:15kugelless work for your patch I'd think
11:44:22Tornew1ll14m: there's a nice gap at the beginning of rockbox.bin where it's later going to write the exception vectors
11:44:27Torneit's just filled with zeros and is never read
11:44:41Tornei could put a string in there that would identify it, then have the bootloader check for that
11:44:57w1ll14mthat would be the 'nice' way
11:45:08linuxstb_Torne: Unless I'm misunderstanding you, what you want can already be done - use ipodpatcher to extract the OSOS image to a file called appleos.bin, and then you can dual-boot it.
11:45:17Tornelinuxstb_: no, you can't
11:45:36Tornelinuxstb_: you can't RoLo retailos
11:45:38kugelJdGordon: I don't really change so much, I just move the bar setup into viewport management, I actually fail to see how it touches your work
11:45:38Torneit doesn't work
11:46:06linuxstb_i'm not talking rolo. i'm talking loading it from the bootloader
11:46:09JdGordonkugel: because I would have to undo your entire patch and change it all.. instead of removing 4 or 5 lines in total...
11:46:16Tornethe bootloader can only load rockbox from disk
11:46:25Torneit doesn't know how to detect that rockbox is already in ram
11:46:28Tornethat's what i'm talking about fixing
11:46:38w1ll14mTorne: you want both rockbox and appleos in osos image ?
11:46:42Torneno, just rockbox
11:46:47Torneappleos will be on disk
11:46:50Tornewhich is already supported
11:47:31kugelso it adds more work for your not-yet-working-possibly-future-code and you don't want to commit it for that?
11:47:53Tornethere's no point having both in there, that would take just as long to load
11:48:04w1ll14mthan that patch should be sufficient to boot rockbox.ipod from osos image instead of appleos
11:48:17w1ll14mif holdswitch is on it loads image from disk
11:48:26Tornew1ll14m: that patch is, yes, ut tht patch breaks the old way
11:48:30Torneso it will never get applied to svn
11:48:38Torneso i will have to have a custom bootloader forever
11:49:04w1ll14mi always build bustom bootloaders + svn builds my self
11:49:05kugelthat's not reasonable to me
11:49:21linuxstb_Torne: I still don't understand what you can't do..
11:49:30*linuxstb_ thinks he gets it now...
11:49:49w1ll14mloading rockbox.ipod from ram to improve boot time
11:49:58Tornelinuxstb_: if the OSOS image consists of rockbox followed by the rockbox bootloader, then when it goes to load rockbox it will load it up again from disk
11:50:07Torneexactly the same as if OSOS was retailos+bootloader
11:50:11Torneit can't tell the difference
11:50:37Tornethere isn't anything *I* can't do here :)
11:50:48Tornebecause i have a hacked bootloader that always does it this way
11:50:52Torneas per that patch
11:51:12 Quit miepchen^schlaf ()
11:51:18Tornei'm talkiing about people with standard builds being able to do it if they want, by having a non-hacky solution that can actually go into svn
11:51:39linuxstb_is that really worth the inconvenience? i mean is there a noticable speedup compared to loading rockbox from the fat partition?
11:52:04Torneit's noticable, unless the dircache needs updating
11:52:10Tornein which case that kinda dwarfs the actual boot time
11:52:13linuxstb_(the inconvenience when upgrading)
11:52:34Tornewell, that's up to the user, is my point :)
11:52:46Tornei'm not recommending that this be considered the normal way
11:52:56Tornejut an option for people who wanat to fiddle :)
11:53:15Torneall it will cost to support is a few bytes increase in bootloader size
11:53:15w1ll14mi still think it's saver to never touch the osos image again after installing the rockbox bootloader
11:53:27w1ll14mbecause it can damage while updating the new rockbox binary
11:53:38w1ll14mthat will make things verry hard to fix
11:53:44Tornew1ll14m: you can always recover that trivially
11:53:49Tornejust boot into disk mode
11:53:53Tornedoesn't touch OSOS
11:53:56w1ll14mNo problem for me
11:54:08w1ll14mbut think of people who don't understand
11:54:20w1ll14mthis way you can almost always recover rockbox
11:54:22Tornepeople who don't understand aren't going to do this in the first place
11:54:37Tornethey'll install with rbutil and it will work like it does now
11:54:44w1ll14mbecause the functionality is not there
11:55:01w1ll14mbut if it was a simple checkbox, then most likely it will be used
11:55:04Torneis it any different to people opting to use the ipodlinux loader?
11:55:08bluebrotherpeople who don't understand tend to do all sort of things they don't understand ;-)
11:55:26Tornew1ll14m: it's not going to be, though
11:55:33Torneyou'd just have to do it yourself with ipodpatcher
11:55:57JdGordonHow do people feel about the statusbar being forcibly redrawn every 250ms by the default event handler?
11:56:12JdGordonunless it was explictly disabled for the curent screen
11:56:15w1ll14mthat would cause extra cpu usage i guess
11:56:27Tornew1ll14m: i dunno. i'm talking about supporting it as in "the bootloader understands it", not supporting it as in offering any help for people to do it :)
11:56:56linuxstb_Torne: I wouldn't object to adding it to SVN.
11:57:23Tornelinuxstb_: even though it means changing the crt0 for portalplayer? :)
11:57:42kugelJdGordon: how often is it redrawn now?
11:58:09w1ll14mTorne: if it's stable it shouldn't be a problem i gues
11:58:21*kugel rather thought of a statusbar tick-task which would be disabled altogether if there's no active statusbar, but didn't code anything yet
11:58:34w1ll14mof course it will be tested before it's added to svn
11:59:09Tornei guess the other question is what should the identifying info *be*
11:59:18Tornejust the string "rockbox"?
11:59:34w1ll14mmaybe "rockbox 3.1"
11:59:49Torneit should probably have the ipod model identifier in it, actually
11:59:57Tornethe way rockbox.ipod does
12:00:38w1ll14mi'd say rockbox + version string
12:01:04linuxstb_Torne: Why add identifying info? Can't you just see if the Apple firmware is there, and if not, assume rockbox?
12:01:14JdGordonkugel: as often as screens want it... actually I tihnk alot more than 4x/s
12:01:17Tornelinuxstb_: you could, but that opens you up to exciting new failure modes
12:01:28linuxstb_Ew.g. ?
12:01:35linuxstb_i mean e.g.?
12:01:50Tornelinuxstb_: if the image in OSOS is useless/blank/something that expects to be loaded nontrivially
12:01:57 Quit BigBambi (Read error: 110 (Connection timed out))
12:02:11Torneit will branch to it then the hardware will probably just crash
12:02:25Tornewhereas currently if rockbox.ipod is missing or otherwise useless the bootloader tells you so
12:02:58JdGordoncan anyone think of a screen which forces the statusbar off ever?
12:03:07linuxstb_Why would that happen? you can tell if nothing is there. (load address of the bootloader)
12:03:35Tornelinuxstb_: is that information still around by the time you get to the bootloader's main?
12:03:49Torneit'll've relocated itself by then, no?
12:04:16bluebrotherJdGordon: can't the statusbar be only redrawn if it actually changes?
12:04:17*Torne shrugs, it just seems nicer to mirror what it does for retailos
12:05:05linuxstb_Yes, I guess it could be simpler.
12:05:09JdGordonbluebrother: well yes, but it still needs to be called to check if it needs to redraw
12:05:23*linuxstb_ hasn't looked at all this for years....
12:06:37Tornelinuxstb_: if it puts in, say, "rockbox" and the four byte model id, it can check that really trivially
12:06:59Tornethat omits the size and crc checking that load_firmware normally does
12:07:19kugelJdGordon: so you don't do anything about FS #9603 for now (since it breaks your local code)?
12:07:23bluebrotherJdGordon: true, but that could be handled by a global message variable which the changing code sets when it changes the contents
12:07:31Tornebut inserting size and crc into the binary would be nontrivial to start with :)
12:08:44JdGordonkugel: no, I've just started fixing this again because your patch is a bad design descision...
12:09:17w1ll14mTorne: if the rockbox tring is ok, but somewhere it went wrong and the binary misses 15KByte of it's binary, we'll never know and we don't know what's happening
12:09:24linuxstb_i don't think we need to worry about verifying the image - ipodpatcher should make sure the firmware partition is always sane
12:09:37w1ll14mso i think it's pretty smart to have some information in the binary as it has now ....
12:10:36Tornew1ll14m: would make the build more complicated, though...
12:10:46kugelJdGordon: is it? It may not be the best design (it doesn't intend that), but it certainly better than svn code. But anyway, there's other benefit parts in that patch other than the hide_bars param
12:11:06Tornew1ll14m: to say nothing of the fun involved in putting a checksum inside the data being checksummed :)
12:11:07kugelI begin to wonder why you even introduced it if it's such a bad thing
12:11:18JdGordonand if it was 2 patches the good parts would be in already
12:11:20Torneanyway i need to go actually do RL things
12:11:25Torneso, i'll have a fiddle
12:11:32Tornethanks for the comments :)
12:11:33 Part Torne
12:13:11 Quit bs66_1 (Connection timed out)
12:13:15 Quit einhirn_ (Read error: 60 (Operation timed out))
12:13:34 Quit linuxstb_ ()
12:15:06kugelJdGordon: another thing. with my patch you can finally get real full screen viewports, and that's a good thing imho
12:15:07 Join miepchen^schlaf [0] (
12:15:35 Join bs66_ [0] (
12:15:48JdGordonumm what? svn can do that also
12:16:15kugelnot with viewport_set_defaults
12:16:32kugelyou need to fix bars after (like it's done it star plugin)
12:16:58kugeland I don't really see why viewport_set_defaults shouldn't be able to do that
12:18:43kugelmaybe you explain the "bad design decision" part a bit more detailed so I can understand it
12:21:53JdGordonviewport_set_defaults was always meant to setup the viewport for the "usable screen" which (ignoring the archos recorder) means the whole screen minus the statusbar if enabled
12:22:06JdGordonthe buttonbar is another story so ignored for now...
12:22:31JdGordonscreens should NOT disable the statusbar unless its needed, so to discourage this it should be extra work to do it
12:22:39kugeland now it can do both
12:23:09kugelwhich makes perfectly sense to me for a function which sets viewports up
12:24:19kugelgo for hide_bars = true and safe the manual viewport fixing after if you're going to hide the bars with do_menu, or go for hide_bars=true and get the old behavior
12:24:32kugelunless it's the buttonbar which is problematic
12:25:04kugeland btw, how does the buttonbar not reduce the usable screen in the same way as the statusbar does?
12:25:05JdGordonhide_bars was origionally a mistake and was put in to band-aid another issue...
12:26:23kugelso I move the band aid from the ard to the shoulder (and the shoulder is blooding). I move the band aid, you fix the shoulder, it's fine afaics
12:27:25kugelthe point is, the main cleanup part in the patch is to fix the multiple re-inits in menu.c, and that
12:27:39kugeland that's not possible without taking the bar-fixing out of it
12:27:58JdGordonoh FFS... I've said 6 times if it was just the reint thing and the menu_vp remoal it would go in...
12:28:41kugelyea, and I said it's pretty much not possible without moving the bar fixing out of it (which I did with expanding viewpot_set_defaults)
12:29:05kugeland that doesn't remove the band aid either
12:29:47JdGordonso if you stop arguing and just let me stay in my text editor things could speed up so it could be done without fidling with the bars
12:29:52kugelif the menu_vp and reinit is to be removed, the viewports must be bar-fixed before calling do_menu
12:32:13 Join einhirn [0] (
12:32:20kugelJdGordon: ok, then I'm going to get it working by manually fixing the bars before and leave viewport_set_defaults untouched (even though I'd really welcome it to create a fullscreen viewport)
12:33:20JdGordonwhat manually fidling with bars? no screens should be disableing the statusbar, and barely any enable the button bar...
12:33:46kugelthere are some, plugins mainly iirc
12:34:20JdGordonplugins are an entirely different story
12:34:33JdGordonand second class citizens.. so it SHOULD be hard for them
12:36:17*JdGordon would like to know how button can be mangled between the call to default_Event_handler() in tree.c and the actual funciton
12:36:33kugelJdGordon: then I don't understand r16812 at all
12:37:26JdGordonlike I said... thats the band aid...
12:38:29kugelas if second class citizens deserved band aids ;)
12:39:22amiconn"[12:33:20] <JdGordon> what manually fidling with bars? no screens should be disableing the statusbar..." <== I would in fact agree for anything except plugins, but obviously some poeple would disagree, probably including yourself, when it comes to the wps
12:39:24kugel"what manually fidling with bars? no screens should be disableing the statusbar, and barely any enable the button bar..." - that was the explanation I needed to understand your doubts
12:39:47kugelah right, wps
12:40:11JdGordonamiconn: yeah, came across the WPS already... thats all i could tihnk of also
12:40:19kugelwould be indeed ease things if you could have a fullscreen wps setup by a single function for the wps too
12:40:45amiconnJdGordon: There might be other screens where themability might be desired.
12:41:04 Quit einhirn (Read error: 104 (Connection reset by peer))
12:42:14JdGordongrr... got a 64bit bug here :(
12:42:20amiconnImo themability isn't just related to design (that's something I don't deem very important), but also to choice of information.
12:42:33JdGordonwhy are events using "long" instead of int?
12:42:59amiconnJdGordon: Events need at least 32 bits, that's why
12:43:11kugelamiconn: so what's your opinion on viewport_set_defaults being able to ingore bars to set a real fullscreen vp up?
12:43:46JdGordonint isnt 32 bits on all targets?
12:43:46amiconnThere once was the (unfinished) calmrisc16 port. Calmrisc16 is a 16 bit architecture
12:44:00kugelif it was only for plugins I'd agree to manually fixing the viewport to ingore bars, but - as you said - screens with actual or possible themeability surely would want that
12:44:22amiconnThere were many fixes for this going in. Back then we didn't have the intNN_t types
12:44:52JdGordonAH... my problem was it was unsigned instead of int...
12:45:03JdGordonworking again :)
12:45:25amiconnRegarding the choice of information, the radio screen probably doesn't need themability, but the recording screen does, at least on targets with smaller screens
12:45:54JdGordonI imagine radio screen will eventaually get themability.. if only for looks...
12:46:09JdGordonrec and radio are opposites :) one is all looks, the other is all info
12:46:36***Saving seen data "./dancer.seen"
12:46:57amiconnNot necessarily. On hwcodec targets, radio recording happens *in* the radio screen. Imo that's much more convenient than switching to an entirely different screen
12:47:54JdGordonyeah, but if you're recording from FM there is less info you need on the screen ( I would tihnk anyway)
12:48:05JdGordonyou dont need the fine grained control?
12:48:16kugelso I figure my viewport_set_defaults would be useful for non-plugins too
12:49:05 Quit tophe (Read error: 104 (Connection reset by peer))
12:49:31amiconnJdGordon: Depends. Stations often differ in level, and that's one thing I don't like that much in hwcodec fm recording. They use a completely fixed recording level.
12:51:23JdGordonI guess we could allow recording from either screen (with fast switching between them) so if you want the looks you have it, and if you want the info its there also
12:51:26kugelwhat about RDS if it happens? afaics the gigabeats radio supports that. I imagine you'd like to theme that too
12:51:28JdGordonwith a selectable default
12:51:44*JdGordon wants AA in the radio screen :/
12:52:29amiconnYou don't need the input selection and mono/stereo setting (that should be linked to the radio's "force mono"), but all other settings still apply: gain, format (swcodec), quality (hwcodec), sample rate, time split, prerecording...
12:52:54kugelso, wps and radio (maybe recording screen too) are screens which possibly hide the bars for themeability, and they're first class citizens
12:53:12 Quit miepchen^schlaf ()
12:53:27JdGordonno... wps/radio is a class by itself somewhere close to G-d :p
12:54:07 Quit bluebrother ("leaving")
12:54:45JdGordoncan anyone else not edit apps/player/keyboard.c?
12:55:17amiconnBesides, I don't think of plugins as second-class, just as addons which provide extra functionality. Sometimes they need (and have) their specific support functions in the core, e.g. lcd_blit_yuv() / lcd_blit_grey_phase(), user timer support (although that is also used for core functionality on some targets - backlight fading on H1x0 and several ipods)
12:55:23*kugel assumes JdGordon's opinion didn't change
12:56:45kugelyour opinion about viewport_set_defaults. I actually think your "band aid" turns out useful
12:56:57JdGordonamiconn: yeah, what I meant was that they are expected to do more work when drawing anyway... they are allowed to break rules like keeping the bars shown if enabled
12:58:02*JdGordon has a patch which probably halves the number of redraws of the statusbar ready
12:58:11kugelamiconn: couldn't the pwn fade also use the backlight tick-task, like the sw fade does?
12:58:53 Join BigBambi [0] (
12:59:10 Join devurandom [0] (n=devurand@unaffiliated/devurandom)
12:59:59devurandomI would like to ask how far the support for the new Sansa devices is. (Fuze&Co)
13:00:14kugelfar enough
13:00:39kugelWe actually don't like people coming here asking for the status. It's done when it's done
13:00:50JdGordonamiconn: care to give this patch a glance?
13:02:23devurandomkugel: Well, I can understand that. And I do not want to push you. I would just like to know whether it would make sense to buy one to play a bit with it. And that depends a little bit on whether I will be able to eventually play a little bit of music with it in a few weeks, or not.
13:03:22devurandom(If it is not perfect, I don't mind, as long as I can poke it a bit, and maybe even use it.)
13:04:18kugelIt's not in a usable state
13:04:21w1ll14mdevurandom: maybe there's information on the wiki about that specific port
13:04:33kugelbut you can of course use the OF to play some music
13:04:53devurandomOF=original firmware ?
13:05:00devurandomSo it's dual boot, I guess?
13:05:25amiconnkugel: That wouldn't be possible. PWM needs much more fine-grained timing than tick tasks can provide, and they need it 100% stable
13:06:03kugelah ok
13:06:11devurandomOk, I'll have a look at it then. :)
13:06:13devurandomThanks. :)
13:08:07amiconnTick tasks are run every 10ms, while PWM uses a 5ms *period*, with 2 interrupts occuring diring this period. The interval between the 2 interrupts is adjusted and can be as short as 50µs (0.05ms)
13:08:48 Nick w1ll14m is now known as w1ll14m_ (
13:09:38kugelI see
13:12:33 Join gregzx [0] (
13:12:34*JdGordon is annoyed that this patch sometimes introduces a ~200ms pause before redrawing the statusbar on some screens
13:13:37*JdGordon found more statusbar disabling screens!
13:13:43JdGordonfull screen debug screens
13:16:39 Quit BHSPitLappy ("Ex-Chat")
13:23:36 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
13:24:01kugelJdGordon: so I ask you a last time (now that there's some more bar hiding screens): Do you (still) think viewports_set_defaults shouldn't really be able to give a fullscreen viewport/should never ignore bars?
13:24:09 Quit AndyIL (Read error: 60 (Operation timed out))
13:24:11JdGordonhmm... anyone want to test this patch out? otherwise I guess ill just commit and wait for the backlash
13:24:37JdGordonkugel: it DOES give the fullscreen viewport
13:24:56JdGordonif the statusbar is enabled then the full screen is 8 pixels smaller...
13:25:15kugelyea, and if it's hidden, but enabled, you don't get the fullscreen you want
13:25:41 Join AndyI [0] (i=AndyI@
13:25:48kugelenabled in the global settings, but hidden for the screen
13:26:23kugelwhich is exactly why star.c needs to fix the viewports after viewport_set_defaults
13:26:39JdGordonof course.... if you dont want the bar you have to disable it before calling viewports set
13:27:27kugelyou mean touching the global settings, just for a single screen?
13:27:51kugelwhat if you shut down while you're in the screen, that means your statusbar is disabled at the next boot for all screens
13:28:17kugelstar.c doesn't disable the bar, it's just not drawn
13:28:49JdGordonno i dont...
13:29:20kugelwhat do you mean with "disable before" then?
13:29:57JdGordonby calling void viewportmanager_set_statusbar(bool enabled);
13:30:10JdGordonwhich apart from debug_menu.c is only called once
13:30:12kugelgui_statusbar_height() returns the global settings
13:30:32kugelah what your patch does
13:30:38kugelI'm at svn
13:30:44JdGordonthe penny finally drops!
13:30:52kugelit's not in svn
13:32:21kugelbtw: I think your patch should disable the tick task, or at least return before the if (TIME_AFTER.. if there's no statusbar enabled
13:32:42kugelonce your patch is in svn, of course, the question is obsolete
13:34:07kugelanyway, do you commit it now? Then I'll quickly edit the patch (and fix star.c in the run)
13:34:11JdGordonthats not the full patch...
13:35:17JdGordonhmm... I tihnk Ill disable the autoredrawing before going into any plugin... that will make things easier
13:35:56 Join faemir [0] (
13:38:58 Join t0mas [0] (n=tomas@rockbox/developer/t0mas)
13:41:29kugeljust saying, what you do is basically the same what I did, just in a seperate function
13:42:12 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:44:12JdGordonhmm... I'm a bit undecided on commiting this...
13:45:58kugelwell, you might at least commit the viewportmanager_set_statusbar part
13:47:05JdGordon... hardly any point... its all or nothing really... it works but im not sure I can be bothered fixing screen glitches it might cause right now
13:48:55JdGordonbugger... the rec screen statusbar handling is wierd and causing problems
13:54:04JdGordonumm... no its not :/
13:56:59 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
14:04:36 Join n1s [0] (n=nils@rockbox/developer/n1s)
14:08:01JdGordonkugel: have a look at that patch ive added to 9603...
14:09:07kugelJdGordon: nice!
14:10:14JdGordonnow imagine using this to draw a partial-screen WPS along side the menu...
14:11:03 Quit t0mas (Remote closed the connection)
14:11:09kugelJdGordon: if you've been that clear (and sometimes a patch is clearer than 1000 words), we could've saved the wasteful discussion
14:11:21kugelno question that yours does it bettere
14:11:44kugelif you've been that clear from the start* I mean
14:12:24kugelthe "gui_syncstatusbar_draw(&statusbars, false);" around all places always bugged me too
14:12:43 Quit jhMikeS (Nick collision from services.)
14:12:49 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
14:14:30 Join stoffel_ [0] (
14:15:13JdGordonthat patch adds some drawing glitches so they need to be tracked down
14:16:42kugelI can imagine 250ms isn't enough if the screen changes
14:16:58 Join t0mas [0] (n=tomas@rockbox/developer/t0mas)
14:17:32JdGordonit entirely depends on how the incoming screen is coded, but 250ms is the upper limit
14:17:37JdGordongoing into the menus is instant
14:18:03 Join Schmogel [0] (
14:18:03JdGordonand for some reason the SB is drawn ontop of the logo on boot !
14:18:41JdGordonthe time/date screen is actually the only one ive found with the pause
14:18:47kugelwell, the tick task already kicks in during the logo I'd think
14:19:08JdGordonit shouldnt... the order of init should make it impossible
14:19:36kugelI also noticed that that the time&date screen doesn't update the time as it's supposed to be too
14:20:00JdGordonit does here...
14:20:02JdGordone200 sim
14:21:03pixelmastatusbar drawn a bit later than the menu/browser is already visible in current SVN (I notice it quite often when returning from WPS and on Ondio, it is visible on my other targets too but usually not that obvious)
14:21:06amiconnJdGordon: I thought that the new time & date screen also handles time & date setting, but obviously it doesn't
14:21:14kugelhm, it didn't when I looked at it recently
14:21:34JdGordonpixelma: oh ok..
14:21:38amiconnTime & date setting still uses the old separate screen
14:21:47JdGordonamiconn: ?
14:21:59pixelmaJdGordon: maybe your patch makes it worse though ;)
14:22:06JdGordonyou mean the actual screen to set the time?
14:22:17JdGordonpixelma: haw haw haw... care to test the patch to find out?
14:22:29JdGordonamiconn: no.. that "coming soon" :p
14:22:55amiconnThe status bar is delayed several seconds on hdd targets in some screens when the disk needs to spin up. Rather irritating
14:23:52JdGordonif thats because the main thread is being blocked by ata then this wont change it
14:23:57amiconnFurthermore the whole startup became a bit irritating on slow targets at some point (somewhat longer ago, pre-3.0)
14:23:58pixelmaJdGordon: sorry, busy with other stuff currently. But I hoped that this would be fixed one day
14:24:46amiconnOriginally the browser (or whatever startup screen is configured) used to replace the logo screen. Now the logo screen is cleared, and the new screen only appears after a fraction of a second
14:25:29JdGordon... there are so many things loading.. it could really be anything
14:25:47JdGordonicons, filetypes, viewers, colours, wps, config....
14:25:58amiconnJdGordon: I think it's a matter of call order. Imo the gui should redraw anything that will indicate something is happening *before* entering a potentially blocking function
14:26:30JdGordonlike I said before.. its up to the entering screens coding
14:27:01amiconnBesides, what has loading stuff to do with an irritatingly blank screen? All that's needed is to delay the screen clearing. I suspect there's just an extra clearing which is unnecessary
14:27:41amiconnUnfortunately I'm not familiar with the startup sequence since configurable startup screen got added
14:27:48JdGordonanyway... you OK with a 4 hertz statusbar update?
14:28:07amiconnImo 2 Hz would be sufficient
14:28:08JdGordonconfigurable start screen is the very last thing in the startup order
14:28:41amiconn(given that an immediate redraw on activity avoids irritating delays)
14:29:13kugelamiconn: 2Hz would be rather irritating, with regards to volume/playback state changes
14:29:22amiconnNo it wouldn't
14:29:57 Join miepchen^schlaf [0] (
14:30:13kugelI think the tick task is supposed to make manual redraws unneeded
14:30:15*amiconn is reminded of his universal indicator idea, which would put the multicolour led on M3 to good use, get rid of special handling of the Player's "status bar", and probably simplify a bunch of other stuff
14:30:26kugelelse we could just stick to how it's now
14:30:31 Join LambdaCalculus37 [0] (i=44a04303@gateway/web/ajax/
14:31:13amiconnkugel: Right now the status bar redraws as necessary, which can be much more often than 4 times per second if things like volume change.
14:31:25kugelyes, I know
14:31:35JdGordoneven with nothing happening I tihnk its redrawn much more often than that
14:31:43amiconnBut if nothing changes, redrawing too often just costs cpu cycles and hence battery power.
14:32:27amiconnJdGordon: Iirc some screens use 2HZ and some use 4HZ right now. But they use button_get_w_tmo(), so redraw more often when buttons are pressed
14:33:02amiconnThis behaviour has to stay, otherwise it will become *very* irritating.
14:33:04JdGordonyes, but also, I think the bars are actually redrawn in the list wigdet code also! so it could be drawn more than once per loop
14:33:40JdGordoncan you test the patch and see how it feels?
14:33:47amiconnI'm not talking about bars, but about things like plugging/unplugging the charger, changing playback state, or adjusting volume
14:33:48JdGordonlast comment in fs#9603
14:34:22*JdGordon wonders when the topic changed?
14:34:43JdGordonah.. im tired
14:34:59JdGordonsupper then bed... gnite
14:35:14 Quit anewuser (Read error: 54 (Connection reset by peer))
14:35:17LambdaCalculus37Good night!
14:36:18kugelamiconn: so manually redraw on such events + redraw ever 2Hz?
14:38:04n1sthe only thing that change without user activity should be clock/battery level/playback state so 2Hz would be plenty
14:39:09kugelyea, the infos that are displayed should issue redrawing, not all the screens
14:39:37kugelbut then you get into trouble again with screens that hide the bar and screen changes
14:43:13kugelthere should probably some kind of "screen change manager" which would handle bar draws and fullscreen updates
14:45:18JdGordonyou mean the root menu?
14:46:40***Saving seen data "./dancer.seen"
14:46:53kugelif that's involved with all the arbitary screen changes
14:47:44 Quit jhMikeS (Nick collision from services.)
14:47:50 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
14:49:18 Quit t0mas (Remote closed the connection)
14:55:56 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
14:59:38 Join anewuser [0] (i=anewuser@unaffiliated/anewuser)
15:01:52 Join fdinel [0] (
15:14:35 Join reacocar1 [0] (i=reacocar@
15:15:01 Quit reacocard (Nick collision from services.)
15:16:16 Quit faemir (Read error: 104 (Connection reset by peer))
15:16:49 Quit reacocar1 (Client Quit)
15:16:53 Join itcheg [0] (i=41d59de2@gateway/web/ajax/
15:17:00 Join reacocard [0] (i=reacocar@
15:19:31 Nick Sir_Lewk is now known as jgreco (
15:25:38kugelfdinel: hi
15:27:14 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.0.5/2008121623]")
15:29:29 Quit stoffel_ ("leaving")
15:31:11fdinelhi kugel
15:34:26 Join Aurix_Lexico [0] (
15:38:11 Join kugel [0] (n=chatzill@unaffiliated/kugel)
15:40:29 Quit CaptainKewl (Read error: 110 (Connection timed out))
15:43:07 Quit Seed ("cu, Andre")
15:58:46 Part LinusN
16:09:12 Quit kugel (Nick collision from services.)
16:09:15 Join kugel_ [0] (
16:09:19 Nick kugel_ is now known as kugel (
16:10:40 Join tessarakt [0] (
16:14:00 Nick devurandom is now known as dev|away (n=devurand@unaffiliated/devurandom)
16:16:36 Join GNUtoo [0] (
16:17:41 Quit gromit`` ("Coyote finally caught me")
16:18:28 Join nplus [0] (n=nplus@141.25.Globcom.Net)
16:19:30 Join stoffel_ [0] (
16:20:00 Join Casainho [0] (
16:20:23Casainhohello :-)
16:20:54Casainhois there a way to build the bootloader and using the -g option for GCC, for debug with gdb?
16:21:15CasainhoI mean, in the time of ../tools/configure, to create the makefile?
16:21:36kugelyou can edit the resulting makefile
16:21:51GNUtoohello, is there a hardware comparison table for the suported players? that compare things such as battery life,capacity,flash||hdd etc...
16:24:26 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
16:24:53Casainhokugel: okok :-) I am doing that, however maybe there were an option for do that, since for JTAG debug we must do that...
16:25:43GNUtooLambdaCalculus37, thanks a lot
16:27:03kugelCasainho: you can also edit configure if you're tired of that
16:27:45kugelI'm not sure, but have you looked if you find something under (A)dvanced build?
16:29:57Casainhokugel: no, I didn't look at it :-) −− I must do that :-)
16:39:22 Quit Casainho ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
16:42:27 Quit daurnimator ("Cyas later...")
16:42:27 Quit BigBambi (Read error: 104 (Connection reset by peer))
16:44:39 Quit Acksaw (Read error: 104 (Connection reset by peer))
16:46:41***Saving seen data "./dancer.seen"
16:52:15 Join gromit` [0] (
16:56:20 Nick Barahir_ is now known as Barahir (
17:04:22 Part dev|away
17:04:34 Join Acksaw [0] (
17:04:38 Join toffe82 [0] (n=chatzill@
17:35:31 Quit jhulst (Remote closed the connection)
17:35:34 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
17:50:38 Quit XavierGr (Nick collision from services.)
17:50:49 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
17:51:23 Join gregzx [0] (
17:51:53 Join BigBambi [0] (
18:00:14 Quit itcheg (" ajax IRC Client")
18:00:25 Join itcheg [0] (i=41d59de2@gateway/web/ajax/
18:13:09 Join tyfoo [0] (
18:14:34 Quit stoffel_ ("leaving")
18:16:56 Join faemir [0] (
18:17:55 Join ChAoS_PeNNa_DeLu [0] (
18:18:07 Quit faemir (Remote closed the connection)
18:18:22ChAoS_PeNNa_DeLujemand da?
18:19:14 Quit ChAoS_PeNNa_DeLu (Client Quit)
18:19:16pixelmathis is an English only channel though (and you do yourself a favour if you want help)
18:19:31pixelma...and patient too...
18:19:48LambdaCalculus37pixelma: He took off.
18:19:58*LambdaCalculus37 sighs
18:21:36 Join faemir [0] (
18:25:58 Join tophe [0] (
18:29:03 Join karashata [0] (n=karashat@
18:36:44 Quit n1s ()
18:46:42***Saving seen data "./dancer.seen"
18:58:40 Quit faemir (Remote closed the connection)
19:03:29 Join faemir [0] (
19:16:48 Join tyfoo2 [0] (
19:18:48 Join stoffel_ [0] (
19:25:30 Quit Darksair ("(define zero (lambda (f) (lambda (x) x)))")
19:30:27 Join MethoS- [0] (
19:31:13 Quit tyfoo (Connection timed out)
19:33:29 Join BigBambi_ [0] (
19:34:31 Quit BigBambi (Read error: 110 (Connection timed out))
19:39:42 Quit MethoS- (Remote closed the connection)
19:45:38 Join gartral [0] (
19:46:26gartralhello, does anyone know if the RButil has been patched, or do i still need my ming10.dll dummy?
20:03:11 Quit Horscht ("We don't make mistakes, we just have happy little accidents")
20:03:45 Join shelby [0] (
20:04:45shelbyhey can you guys help me with something? i have a folder at / called Music, and in there is,. yea, all my music.... all of it works, but for some reason when i put my pink floyd in there, it dosent show up! weird, its only for pink floyd.... also, i have a folder with some avi movies that dont show up either! what in the hell is going on?
20:05:10krazykitshelby, do they have the "hidden" attribute set?
20:05:25shelbydo you mean like '.something'
20:05:30 Part GNUtoo ("Leaving")
20:05:45shelbyim not sure what you mean by hidden attribute set, but i dont think they do...
20:06:13shelbyhow could i check that sh*t?
20:07:07shelbyits really annoying not being able to listen to pink floyd or watch movies... :P
20:07:33shelbyits just a normal folder like any of the others....
20:07:50shelbymaybe the dude who made rockbox has something against avi encoded videos and old rock stars
20:08:03gevaertsrockbox does indeed not handle avi
20:08:13agaffneyavi isn't an encoding
20:08:17agaffneyit's a container format
20:08:34*gevaerts suspects the "view files" setting, or whatever it's called
20:10:12gevaertsshelby: can you check what Settings > General Settings > File View > Show Files is set to?
20:10:58shelbysupported... yikes...
20:11:30shelbyi changed it to all
20:11:34shelbyand still no pinky
20:13:31 Quit stoffel_ ("leaving")
20:13:37*gevaerts doesn't know then
20:13:57gevaertsMaybe a corrupted filesystem☕?
20:15:47Lloreanshelby: How did you put the pink floyd music on there, and are you looking for it through "files" or "database"?
20:21:02 Quit Llorean (Read error: 54 (Connection reset by peer))
20:21:25 Join Llorean [0] (
20:23:09 Join lilmackie [0] (n=483546c5@gateway/web/cgi-irc/
20:23:23lilmackiehey there i'm looking for some help
20:24:12krazykitwell, you'll want to ask a question or mention a problem if you want it
20:24:53lilmackiejust want to get some feedback on how easy it is to remove the firmware if i want to go back to my stock
20:25:19krazykitwhich device?
20:25:42lilmackiesansa e200 series
20:25:52lilmackieits actually the e270 but i don't think that matters here
20:26:20LloreanIt's just as easy to remove as it is to install.
20:26:21LambdaCalculus37It doesn't matter. And to answer your question, it's quite easy.
20:26:26*LambdaCalculus37 is too slow
20:26:31lilmackiei know it has a dual boot, but does that mean i have to press the keys to go back into sansa mode every time?
20:26:50lilmackieLlorean what do i do to remove it?
20:27:01LambdaCalculus37Use Rockbox Utility.
20:27:47lilmackieis that the installer?
20:28:06LambdaCalculus37It's also an uninstaller.
20:28:32lilmackiewhat are the chances ill brick it? >.< that's the only thing i'm really afraid of
20:29:17scorchethey are quite hard to brick...worst case scenario, you go through our documented recovery procedure
20:29:20LloreanYou can't really brick it. It can require some restoration procedures if you do something like unplugging it during the installation at the wrong time or something
20:30:01*lilmackie sighs
20:30:11lilmackiei feel like a chicken but i just want to be really sure before i do it
20:30:47 Join BigBambi [0] (
20:31:40 Quit BigBambi_ (Read error: 104 (Connection reset by peer))
20:32:29lilmackiewhich do i download, the build or the release?
20:35:54scorchewhichever you wish to...they are both "builds", but the latest build is what most of us use, i would imagine
20:36:47lilmackiethanks a bunch guys
20:37:08 Quit lilmackie ("CGI:IRC")
20:37:27shelbygrr im getting seriously pissed at rockbox now
20:37:40shelbyit wont show my pink floyd, or cypress hill, or any movies.....
20:37:52shelbyi dunn o whats wrong but its really, really, reallyreallyreally pissing me off
20:38:00scorcheand you are sure you changed the aforementioned setting?
20:38:52LambdaCalculus37And we already told you that if your movies are AVIs, then they won't show up nor work.
20:39:16scorchewell, they will show up in the file browser
20:42:15Lloreanshelby: I may have missed your answer.
20:42:19LloreanHow'd you put the album on there?
20:46:44***Saving seen data "./dancer.seen"
20:47:58shelbyi just put the music in the music folder... im sorry but thats just it
20:48:09shelbywell the avis i can handle
20:48:15shelbyfine, but the music... jeez....
20:48:23Lloreanshelby: I said "how" not "where". As in, did you use a file manager, a sync program, or what?
20:48:37shelbyi used the gnome file manager....
20:48:51LloreanAnd you've changed "file view" to all?
20:48:53scorcheand you are sure you changed the aforementioned setting?...and that you are looking in "files" and not using the database?
20:49:21LloreanThere were three questions, was that to all three?
20:49:55LloreanWell, I guess two questions since one was redundant.
20:49:55scorchethough 2 of those 3 were the same
20:50:01shelbyi changed file view to all, i know the music is on there because when you plug it in it shows up in the music directory, and yes, im looking in files....
20:50:24Lloreanshelby: "Files" when it's set to "all" shows the exact same folders and files as visible on the PC in MSC mode.
20:50:37scorcheshelby: and are you sure you unmounted cleanly?
20:50:45shelbyyu−− no....
20:50:50shelbyi just unplugged it
20:51:00*scorche ahems
20:51:17shelby:P okay let me try that one more time....
20:51:31shelbycan i install avi support or something?
20:51:44Llorean"avi" isn't a format.
20:51:53scorcheno...and as we said before, there is no such thing as "avi support" is just a container
20:51:58LloreanIt's just a container, there are many, probably dozens, of things it can have in it.
20:52:04shelbyoh, alright...
20:52:08shelbyi never knew that! lol
20:52:20shelbyonemmore thing... can i compile code written in C to work in rockbox?
20:52:25LloreanAnd you'd need to convert your files to match the screen size anyway.
20:53:10scorcheof course...keep in mind though that rockbox is not POSIX compatible, does not have a general malloc, and is in fixed may want to check out the "for developers" section in the documentation index
20:53:21shelbyalright... lol
20:53:33shelbyyou guys are really helpful, thanks a lot
20:56:34shelbyme an taylor gonna hotbox my cellar lol
20:56:39shelbyoops... wrong window... lol
20:56:54shelbyi believe that is known as a FAIL....
20:57:46shelbyima go now... thanks again guys
20:57:48 Quit shelby ("Lost terminal")
21:08:47 Quit __lifeless (Remote closed the connection)
21:08:55 Join __lifeless [0] (n=lifeless@
21:12:19 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
21:15:09 Quit DataGhost (Nick collision from services.)
21:15:09 Join DataGhost_ [0] (
21:15:11 Nick DataGhost_ is now known as DataGhost (
21:21:48 Join robin0800 [0] (
21:22:11 Join _Auron_ [0] (
21:30:08 Join Zagor [242] (n=bjst@rockbox/developer/Zagor)
21:30:18 Join casainho [0] (n=chatzill@
21:30:39casainhohello :-) −− I am getting an error: /home/cas/Documentos/rockbox_player/rockbox/firmware/panic.c:51: warning: implicit declaration of function ‘disable_fiq’
21:30:53*kugel got a more reliable wheel
21:30:55casainhocan someone help me?
21:31:06kugelZagor: I think your usb_detect commit broke the fuze
21:31:30kugelcasainho: that's a warning, not an error
21:31:44Zagorkugel: regardning your FS #9584 report: the diff statistics is simply the output of "diffstat". if you feel it is buggy a report is better aimed there.
21:31:57Zagorkugel: really? what happens?
21:32:10kugelit boots with the usb inserted logo
21:32:33 Join stoffel_ [0] (
21:32:41kugelcasainho: I suggest looking existing targets to find out what disable_fiq exactly is
21:32:52 Join fml [0] (n=4fd3c294@gateway/web/cgi-irc/
21:32:52casainhokugel: do you think that warning will not give any trouble?
21:33:09kugelI don't dare to predict that
21:33:19kugelbut if it compiles it's worth a try ;)
21:34:10fmlkugel: hello. Have you thought more about the general font cache? The more I think about it the more I get to the opinion that we need to implement a rather general memory management algorithm operating on the given buffer.
21:34:27casainhokugel: it do not compile :-( −− this error: rockbox/firmware/panic.c:54: error: ‘HIGHEST_IRQ_LEVEL’ undeclared (first use in this function)
21:35:13fmlThe buffer should also contain the index and the info needed for LRU (if the LRU is really needed, which I'm not sure of).
21:35:34Zagorkugel: ok, sounds like it is active high there then. can you try simply inverting the check (remove the !) and test?
21:36:18kugelZagor: it boots the of if r1 != 0
21:36:30*kugel 's reading dualboot.S
21:36:47kugelso yes, high indicates USB
21:37:20fmlZagor: hi! Would it be possible to make a switch in the FS viewer to show/hide the patch stats in the comments? Sometimes they are too many and it's hard to read the comments.
21:37:34 Quit kachna (Read error: 113 (No route to host))
21:37:44Zagorhmm, how does it work for clip then
21:38:06Zagorfml: not without considerable effort, I'm afraid.
21:38:23kugelZagor: no idea
21:38:35fmlZagor: can't that be done by just redefining the CSS?
21:38:42Zagoroh, I have a guess. checking...
21:38:48kugeldid you make sure it boots the OF with usb inserted, but boots rockbox without USB?
21:39:04Zagorfml: a switch requires at least some javascript
21:39:32kugelZagor: other than that, you're missing setting GPIOA_DIR, which might give unpredictable results
21:39:35Zagorand I expect those who disable it want it to stay disabled, hence requring either cookies or modified database schema
21:40:11fmlZagor: yes. If I'd provide the script, would you include it? Is it possible at all?
21:40:48kugelcasainho: have you already implemented interrupts? I guess so
21:40:54Zagorfml: I might, if it is cleanly done
21:41:12casainhokugel: no, I think
21:41:15kugelcasainho: again, I highly recommend looking at existing code, that's likely to give you a clue
21:41:26casainhoI already looked
21:41:26kugelcasainho: didn't you say you have a tick task running?
21:42:25casainhokugel: no, I don't have it. I started again to creat a new port, and now the code do not builds.. it's the first time I am getting this errors
21:43:03kugelcasainho: what file did you include?
21:43:12kugelsystem-target.h or what?
21:43:37fmlZagor: ok. I'll try and if I'll like it I'll provide the code to you
21:43:44kugelif you include system-arm.h you should be getting the macros and symbols you need
21:43:48Zagorfml: ok
21:43:57kugelbut don't sue me if it doesn't work for your specific port
21:45:20casainhokugel: yes, system-arm.h, looks like is waht is missing, however I don't know why is missing
21:46:09kugel'#include "system-arm.h"'?
21:46:50casainhokugel: this erros apears when building kernel.c....
21:46:58kugelcasainho: actually, if I got this right, you're only supposed to include system.h, not all the others system-*.h's, except for some situations
21:47:44kugelcasainho: so you have that included?
21:48:00kugeljust a wild guess: did you define ARM_ARCH for your target?
21:48:11casainhokugel: I didn't include any of that files. I just included kernel.h and after do kernel_init()
21:48:26casainhokugel: where do I define that? :-)
21:48:48kugelconfig.h I think
21:48:55 Join creamycenter2003 [0] (
21:49:03kugeli.e. firmware/export/config.h
21:50:15kugeljhMikeS, amiconn: ping
21:50:33creamycenter2003the latest rockbox installer (rbutilqt 1.0.8) is asking me for mingwm10.dll
21:50:56 Join einhirn [0] (
21:50:58casainhokugel: yes, I did: #if defined(CPU_TCC77X) || defined(CPU_TCC780X) || (CONFIG_CPU == DM320) || (CONFIG_CPU == AT91SAM9260) #define ARM_ARCH 5 /* ARMv5 */
21:51:45kugeland you defined that cpu for your target too?
21:53:06casainhokugel: yes, I think: #define AT91SAM9260 9260, on config.h
21:53:56kugelcasainho: ok, it was a wild guess
21:54:32kugelcasainho: You should probably inserted some #warning here and there to make sure the code you want to use is actually compiled
21:55:07casainhokugel: but the problem should be something about the kernel files, because I am getting this errors: implicit declaration of function ‘enable_irq’
21:55:21kugelcasainho: yes I understand
21:55:36kugeland you don't have system-arm.h included for some reason I don't know
21:56:14casainhokugel: strange be cause I just checkedout the source....
21:57:01kugelcasainho: ok I think I know what's the problem
21:57:16casainhowhat? :-)
21:57:19kugeldo you have created a system-target.h in your target tree?
21:57:28casainhoyes, I did
21:58:03kugelwhat's in there?
21:58:21*Zagor adds a note about mingwm10.dll to the front page
21:58:24kugelit should have nothing but '#include "system-arm"'
21:58:26 Quit LambdaCalculus37 (" ajax IRC Client")
21:58:44pixelmaZagor: good idea
21:59:26*pixelma should put it in the wiki too
21:59:36casainhokugel: this:
21:59:51creamycenter2003pixelma: thanks, do you know if it is related to Quicktime? i see QT mentioned in some of the fora, and i do not have it installed
22:00:21kugelcasainho: that one looks weird, what's all that stuff?
22:01:14pixelmacreamycenter2003: no, you don't need this. It's for building the Rockbox Utility, you shouldn't need to care about that as a user (and it's not quicktime)
22:01:15ender`creamycenter2003: the talk is about the Qt toolkit, not QuickTime
22:01:31creamycenter2003ok, thanks'
22:01:32casainhokugel: eheh - I took it from other sources, I was hoping to use it for my code...
22:01:51creamycenter2003there are too many recursive abbreviations out there! LOL
22:01:58casainhokugel: so, I should clean it and put there the #include system-arm.h?
22:02:08kugelcasainho: try that, yes
22:02:38casainhokugel: your name is Kugel in Austrian?
22:04:55kugelkeanu: ping
22:05:05keanukugel, pong
22:05:36*kugel might finally five Zagor's watermark patch a try now that he's got a acceptable scroll wheeöl
22:05:41casainhokugel: it did buils, almost everything, at least the kernel.c :-) −−- I am just being listening in my work of the word "kugel", we have a client in Austria...
22:06:05Zagorkugel: I'll just commit a fixed usb-as3525.c so you can boot :-)
22:06:08kugelcasainho: well, kugel is a common German word
22:06:26casainhokugel: ah :-) okok :-)
22:06:41casainhokugel: thank you - you are my hero of today ;-)
22:06:55kugelno problem :)
22:06:58casainhoI will continue now building the code ;-) - thanks
22:07:13kugelkeanu: might try a better scrollwheel?
22:07:30keanuin a little bit - busy at the moment
22:14:08 Join Acky [0] (
22:14:39 Join dfkt_dt [0] (i=dfkt@unaffiliated/dfkt)
22:14:39 Join Llorean1 [0] (
22:14:51 Quit dfkt (Nick collision from services.)
22:14:55 Nick dfkt_dt is now known as dfkt (i=dfkt@unaffiliated/dfkt)
22:14:57 Quit Llorean (Nick collision from services.)
22:15:00 Nick Llorean1 is now known as Llorean (
22:15:15 Join will [0] (n=c77eae75@gateway/web/cgi-irc/
22:15:17 Join Xerion_ [0] (
22:17:45 Quit fml (
22:17:45 Quit Acksaw (
22:17:45 Quit nplus (
22:17:45 Quit fredddy (
22:17:45 Quit pixelma (
22:17:45 Quit goffa_ (
22:17:45 Quit Xerion (
22:17:50 Nick Xerion_ is now known as Xerion (
22:18:38 Join goffa [0] (n=goffa@
22:18:40Zagorkugel: can you try the new commit on your fuze?
22:19:27 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
22:20:16 Quit will ("CGI:IRC (Ping timeout)")
22:20:49 Quit tessarakt (Read error: 110 (Connection timed out))
22:21:07 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
22:21:22 Join tessarakt [0] (
22:21:22NJoinpixelma [0] (n=pixelma@rockbox/staff/pixelma)
22:21:24 Quit stoffel_ ("leaving")
22:21:42NJoinnplus [0] (n=nplus@141.25.Globcom.Net)
22:23:24NJoinfredddy [0] (
22:24:39 Join Torne [0] (
22:25:37Tornew1ll14m_, linuxstb: well, i implemented the boot-from-osos bootloader change. seems to work, and it saves at least five seconds on boot for me.. (not timed it very accurately)
22:27:02kugelZagor: sure, wait a second
22:28:48linuxstbTorne: Is that 5 seconds compared to retailos in the firmware partition, or just the Rockbox bootloader in the firmware partition (and loading rockbox.ipod from disk) ?
22:29:32ZagorI didn't know ipods were so slow to boot
22:30:11linuxstbBy default, they load about 5MB of OF into RAM from disk first, which Rockbox then discards (unless dual-booting).
22:30:14kugelZagor: before I try, I might add that gpio a3 apparently shared with the lcd somehow which sets that to output, that could possibly lead to problems
22:30:41kugelon the fuze
22:31:20Zagordoes our fuze lcd driver change that pin?
22:33:06Tornelinuxstb: compared to loading rockbox from disk
22:33:14kugelZagor: at init, yes
22:33:14Zagorwhee, red build for m200v4 boot
22:33:27Tornelinuxstb: retailos is stupidly slow to boot wherever you load it from :)
22:33:33kugelit sets it ;)
22:33:46kugelZagor: your commit works nevertheless
22:34:10Tornelinuxstb: and retail is 7.3MB for ipodvideo, compared to 611K for rockbox
22:34:18Torneso it loads less than a tenth of the data
22:34:20casainhoB4gder: are you there?
22:34:26B4gderI am
22:34:27kugelapparently the lcd needs it only once at init, that it should be fine
22:35:43casainhoB4gder: don't know if you can, but, I did make a patch: −− but I am getting an error witout sense:
22:36:00kugelZagor, B4gder: What do you think about adding the ams sansas to the build table?
22:36:02Tornelinuxstb: it should preserve all existing behaviour, and if rockbox.ipod exists on disk it loads that rather than the one in osos
22:36:13kugelthe main builds that is
22:36:17Tornelinuxstb: so if you upgrade rockbox the normal way you do get the new version :)
22:36:18B4gderkugel: sounds like a good idea
22:36:21Zagorkugel: what happens if you remove that?
22:36:35casainhoB4gder: /firmware/target/arm/at91sam/rockboxplayerlittle/crt0.S:(.text+0x0): multiple definition of `reset_handler'
22:36:39kugelZagor: from the lcfd?
22:36:44Zagorkugel: yea
22:36:48*kugel tries
22:36:52linuxstbTorne: I remember seeing a very noticable speedup when replacing retailos with the Rockbox bootloader, so I'm surprised you get a further 5 seconds - IIRC, the total boot time was down to about 6 or 7 seconds.
22:37:03kugelit's basically what the of does though
22:37:03linuxstb(I think I tested on an ipod Photo)
22:37:17TorneOh, sorry, i don't mean compared to just having hte rockbox bootloader
22:37:40Tornei mean comparing retail+bootloader -> rockbox.ipod, to rockbox+loader
22:37:50Torneit's probably not a big difference compared to just using the bootloader
22:38:04linuxstbTorne: That's what I was saying earlier - hence me wondering why you are bothering to do this...
22:38:05Tornenot tried that, as it honestly didn't occur to me :)
22:38:12B4gdercasainho: well it means you get the symbol twice, why is that no sense?
22:38:12Tornei missed what you meant
22:38:27 Join webguest67 [0] (n=43bfb401@gateway/web/cgi-irc/
22:38:56 Nick webguest67 is now known as horned_reaper (n=43bfb401@gateway/web/cgi-irc/
22:38:59casainhoyes, it makes sense BUT I don't have it twice!! that same code did build before...
22:39:01Tornelinuxstb: i guess whether it speeds that up or not depends how efficient the apple rom is at loading compared to the rb loader
22:39:20 Quit tessarakt (Connection timed out)
22:39:25kugelcasainho: tick tasks don't work without interrupts
22:39:26B4gdercasainho: well, apparently something has changed...
22:39:27 Quit bmbl ("Woah!")
22:39:49 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
22:40:01 Join tessarakt [0] (
22:40:01casainhokugel: what you mean? I am missing some code? - what should I do?
22:40:13casainhokugel: what you mean? I am missing some code? - what should I do?
22:40:17linuxstbTorne: Sorry if I wasn't clear earlier. I don't have any ipods with me at the moment - can you do a comparison?
22:40:48kugelcasainho: I'm not really in the moods to explain interrupts right now
22:40:49Tornei can do but not right now, my other half will get annoyed if i keep walking in front of him playing halo to plug and unplug my ipod from my dev box :)
22:40:53Tornei'll try it later
22:40:54casainhokugel: can you please try build that code?
22:41:17*kugel is in mood for that even less
22:41:36casainhokugel: yes, I know the interrupts... I will seach better... because I already did some code for interrupts
22:41:42kugelZagor: lcd seems to work
22:42:18 Quit horned_reaper (Client Quit)
22:42:27Tornelinuxstb: i was biased by having already read the thread with the other guy who tried this, i think, otherwise i should've realised what you meant :)
22:42:33kugelI'm not sure if it's related to your commits (I don't think so), but I get lcd weirdnesses again
22:42:34Zagorkugel: good
22:42:49casainhokugel: I did the __timer_register, unregister and __timer_set
22:43:07linuxstbTorne: What thread?
22:43:10Zagorkugel: try commenting out usb_init_device() and see if it makes a difference
22:43:22Tornelinuxstb: some guy on the forums did it by just replacing hte code that loads rockbox from disk
22:43:24kugelI got those too before you committed the GPIOA_DIR change, so I doubt that's related
22:43:34Tornelinuxstb: i.e. breaking the ability to load rockbox.ipod normally
22:43:53kugelI'm gonna try some revisions, I actually didn't touch my fuze since before christmas
22:43:57Tornelinuxstb: i've just done the same thing but backward-compatibly
22:45:56kugelZagor: before you do anything in the lcd driver, that it still works doesn't necessarily mean anything since I didn't update the bootloader
22:46:32Zagorkugel: I wont touch that until I can test it personally
22:46:47***Saving seen data "./dancer.seen"
22:47:00 Join beta [0] (
22:47:48 Nick beta is now known as Beta2K (
22:48:25 Quit w1ll14m_ ("leaving")
22:48:48kugelI guessed so :)
22:52:44kugelhmm, a fuze build shouldn't fail because of a missing #include in serial.c, should it?
22:52:56 Quit casainho ("ChatZilla 0.9.84 [Firefox 3.0.5/2008121622]")
22:53:22 Nick Naked is now known as Hadaka (
22:53:45kugelor, maybe it should (I thought this file was added with IAP)
22:54:33Zagornew clips are coming with single-color screens:
22:54:48ZagorI wonder what else has changed
22:57:48 Quit BigBambi (Read error: 110 (Connection timed out))
23:00:27mcuelenaeregevaerts: are you here?
23:01:05 Quit karashata ("G'bye everyone!")
23:02:27B4gderclip v2s have much larger OF images than v1 it seems
23:02:39kugelUnhelpful: ping!
23:02:59kugelUnhelpful: Not sure how, but r19692 somehow broke my fuze's display
23:05:32gartralhowed you obtain a build from the future...?
23:05:47kugelI mean r19592, sorry
23:06:55Zagorkugel: clip and fuze added to build table
23:07:02Zagorfuze is yellow
23:07:49kugelZagor: yea, I'm going to fix that with the button driver work
23:08:13linuxstbZagor: I think funman said that the clip v2 has a slightly different AMS SoC, plus 8MB RAM.
23:09:20 Join fml [0] (n=4fd3c294@gateway/web/cgi-irc/
23:09:34creamycenter2003question, rockbox on the ipod nano, it powers off fine, but how do you get it to turn back on?
23:10:09fmlZagor: ping
23:10:42Zagorfml: pong
23:10:46gartralcreamycenter2003: have you hit play?
23:10:54linuxstbcreamycenter2003: You should just need to press either the middle or menu buttons. If they don't work, try holding them both together (to force a reset).
23:11:16gartralhe wants to turn it on, not off
23:11:18 Join Thundercloud [0] (
23:11:34linuxstbgartral: Yes, I know.
23:12:12fmlZagor: here's ma take at toggling the diff stats in FS viewer. Place the script somewhere (directly in the page or in a linked .js file) and then add a button or something that calls the function "toggleDisp()".
23:12:39fmlI did it like this: <span onclick="toggleDisp()">Toggle style!</span> but it can be done in a different way
23:12:42bertrikcool to read that the clip and fuze are now on the build table. I kind of like the dual-color clip screen, pity it's a single-color screen with new clips
23:12:55ZagorI wonder where to get a clip v2. not even seems to sell them.
23:13:28creamycenter2003huh, well now it works
23:13:52bertrikthere was a forum post with some info about the chip numbers of the v2 clips IIRC
23:13:55creamycenter2003earlier it brought up the Safe to unplug dialogue, even though it wasnt plugged into
23:14:08fmlZagor: the script works with Mozilla 3. I have not tested it with other browsers but if they don't support some properties nothing should happen (it's programmed in a defensive way)
23:14:29linuxstbcreamycenter2003: It will do that if you hold select+play together as it turns on.
23:14:54Zagorfml: it needs cookie support. turning it off for just a single view of a single page is rather pointless.
23:15:55fmlZagor: hmm... you're right. But that was just a proof of concept. I didn't know for sure if it's possible at all.
23:18:32Zagorbertrik: do you remember anything more to search for? I can't find such a post.
23:19:02gartral <−− a mix of v1s and 2s
23:19:16gartralclip v1s and 2s* excuse me
23:20:22Zagorgartral: where do you see a v2?
23:20:26bertrikZagor, it's the latest post in the sansa v2 forum thread:
23:20:47Zagorbertrik: thanks
23:21:12gartrali thought the v2s had the shiney screen
23:21:30gartralim sorry
23:22:04Zagorgartral: from the pictures on the telltale sign seems to be that v2 has a uniform-colored display. no more split yellow/blue.
23:22:09 Quit perrikwp (" ajax IRC Client")
23:24:26gartralthanks, ill keep looking
23:24:55_Auron_split yellow/blue?
23:25:53 Quit fml ("CGI:IRC (Ping timeout)")
23:26:10bertrik_Auron_, yes, it's basically a 'monochrome' OLED display where the top 16 lines are yellow, then some empty lines and the rest is blue
23:26:22_Auron_oh weird
23:26:34_Auron_guess they did that to save on power
23:27:04bertrikI don't think so, I think it's done so they can give the status bar a different color
23:27:28_Auron_meant the monochrome bit, not specifically the color
23:29:52gartrallol, are we sure there selling these yet?
23:32:17bertrikI wonder if the single-color is a mock-up and if the real clip v2 actually have a single-color display
23:32:35Zagorbertrik: that thought crossed my mind too
23:32:57gartralits right
23:33:08gartralthird photo down
23:34:10scorchefifth? ;)
23:34:28gartralok, ok, there loading bassackward on my comp
23:35:39B4gder90 builds now...
23:35:52Zagordoes anyone have a v1 with this "flower" logo on the back?
23:36:00B4gderany takers on guessing when we hit 100? ;-)
23:36:09Zagorgartral: thanks btw. good to get that cleared up.
23:36:58Zagorodd that they market the thing using so very wrong images
23:37:30gartralZagor: np, only 3 discrepencies disscussed are; 1. brighter OLED screen. logos on both front and back. and 3. "better" plastics
23:37:58*amiconn reminds Bagder of the speculative parallel build idea
23:38:19Zagorgartral: the front "sansa" logo is not v2 though. my v1 has it.
23:38:22 Quit Thundercloud (Remote closed the connection)
23:38:26B4gderyeah, the idea is as fine as ever...
23:38:28bertrikmy v1 clip says "sandisk 1 gb" on the location of the flower
23:38:45gartralbut the font has changed
23:39:30Zagorgartral: ah, true. the 's'-es are a little rounder
23:40:00Zagorbertrik: yeah, mine too. looks like there may actually be an external visual cue this time...
23:41:34gartralhaving a 32" monitor helps.... somtimes
23:41:36Zagoruh, no. at least one person claims he got a v1 with the flower logo.
23:42:18gartralwell... my e250 standard has the rhpsody back... so, thats not uncommon
23:43:14ZagorI know. but it's annoying.
23:43:15 Join RockRabbit [0] (n=3aac9a01@gateway/web/cgi-irc/
23:44:30RockRabbitIs there a command line tool in Linux to copy vorbis tags on flac files over to wave files? I have converted my flac files to wave, but the tags are lost.
23:44:33 Join ChadFM [0] (
23:44:48gartraltrue, i flipped out after trying to patch the BL failed 3 times with the rhapsody instrutions before i just tryed installing rockbox, and it worked
23:47:17kugelZagor: I think I have an acceptable button driver in a few minutes
23:47:46kugelstill no irq for the wheel, but it's surprisingly well anyhow
23:48:16gartralRockRabbit: why convert from flac, anyway?
23:48:21kugelwell, only home button missing, needs more investigation
23:49:01 Quit B4gder ("It is time to say moo")
23:53:26Tornelinuxstb: ok, it's a bit fiddly to time by hand.. but it looks to me like it is faster to do it my way :)
23:53:39gevaertsmcuelenaere: for a few minutes
23:54:05 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
23:54:06mcuelenaeregevaerts: 'The usb serial implementation doesn't seem complete enough to be usable from Windows.' -> what does this mean?
23:54:18Tornelinuxstb: replacing appleos+loader with just loader takes off about 400ms, but replacing it with rockbox+loader takes off about 900ms
23:54:57Tornelinuxstb: if that's accurate it would imply the disk reading code in the apple rom is faster than the code in the rb loader :)
23:55:11 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
23:55:58Tornelinuxstb: all subject to unfortunate variance from my reaction time, though
23:56:15Tornelinuxstb: i think my estimate of 5 seconds before was a bit optimistic :)
23:56:23 Join solexx_ [0] (
23:56:27gevaertsmcuelenaere: just that :). Basically it just implements one bulk in and one bulk out endpoint, which is enough to make linux talk to it. A complete implementation also needs some control commands and (maybe, I don't remember) an interrupt endpoint, and windows seems to depend on those being present
23:56:53gevaertsIIRC bertrik worked on it a while ago, but I don't know/remember how far he got
23:57:01mcuelenaerehmm ok, so how's Windows supposed to react on the current implementation?
23:57:42gevaertsI think it recognizes it, starts installing the driver, and then decides that it doesn't work
23:58:28mcuelenaereI get the standard 'the USB device doesn't get recognized'
23:58:39mcuelenaereeh isn't recognized*
23:58:59*gevaerts digs deep in his memory

Previous day | Next day