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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2009-01-10

00:01:23 Join solexx_ [0] (
00:02:27 Quit Beta2K (Read error: 110 (Connection timed out))
00:03:05 Quit bluebrother ("leaving")
00:05:10 Nick amiconn is now known as notamiconn (n=jens@rockbox/developer/amiconn)
00:05:59 Nick notamiconn is now known as amiconn (n=jens@rockbox/developer/amiconn)
00:08:27 Quit kerwood ("Leaving.")
00:12:25 Quit {phoenix} (Remote closed the connection)
00:12:42 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
00:13:10 Nick fxb__ is now known as fxb (
00:16:45 Quit solexx (Read error: 110 (Connection timed out))
00:18:43 Quit jgarvey ("Leaving")
00:20:32Bagderthat proved parts of the script working at least
00:31:23 Quit jeffronius (Read error: 104 (Connection reset by peer))
00:31:52mcuelenaereBagder: what exactly does the new script do on red commits?
00:32:11Bagderright now it just extract the committers
00:32:24BagderDaniel Stenberg=1
00:32:37Bagder(I made it display for yellows to get a test)
00:33:01Bagderthe plan is to perhaps make logbot announce it here
00:34:15*mcuelenaere wonders whats up with the colors on
00:35:08Bagderwhat a mess
00:35:27mcuelenaereperhaps it has something to do with the contents of the patch?
00:35:40Bagderprobably, but that feels like a viewvc bug
00:36:12mcuelenaereI don't think so, it seems it escapes < and >
00:36:26mcuelenaeres/it seems it/it seems to/
00:36:30 Quit gregzx (Read error: 104 (Connection reset by peer))
00:41:37 Join tyfoo2 [0] (
00:43:22 Join domonoky1 [0] (
00:49:08 Quit domonoky (Read error: 104 (Connection reset by peer))
00:52:17***Saving seen data "./dancer.seen"
00:54:02 Join gartral1 [0] (n=Gartral@
00:55:49 Quit tyfoo (Connection timed out)
00:55:49 Nick tyfoo2 is now known as tyfoo (
00:56:28Unhelpfulperhaps it could yell if there are enough yellow builds?
00:57:21rasherShouldn't it yell at any amount of yellow?
00:58:02gartral1I'm 90% sure someone's asked this before, but: is there a way to either add a WPS too the radio screen, or merge radio playback and controls into the existing WPS?
00:59:16Unhelpfulrasher: given the amount of spurious yellow lately, ie, at least one sim target per build, i'm not sure that'd be helpful.
00:59:30domonoky1gartral1: if you can code, there is a way: code it :-)
00:59:42rasherUnhelpful: well that should get fixed eventually
00:59:53 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
00:59:56rasherPersonally I think each finished build should result in a report
01:00:13gartral1domonoky1: ive been looking through source... but so far, I'm lost
01:01:00Unhelpfuli wouldn't mind having reports done from build to build... but right now it generates a little too much noise.
01:01:37Unhelpfulbut, per-symbols binsize deltas... that would not be a horrible thing to have :D
01:04:11 Quit robin0800 (Remote closed the connection)
01:05:11 Join Davide-NYC [0] (
01:05:40 Quit Thundercloud (Remote closed the connection)
01:06:16Davide-NYCjhMikeS: ping
01:07:34jhMikeShello New York!
01:08:55Davide-NYCI have my Gigabeat S back and want to know what needs to be tested etc. In particular how do we test for USB transfer bugs and charging bugs. AFAIK those are the two main blocks to afully supported target
01:09:14 Quit mcuelenaere (Read error: 113 (No route to host))
01:09:21*Davide-NYC probably wrong about how close it is to being ready
01:09:23 Join jeffronius [0] (n=kvirc@
01:09:55BigBambiDavide-NYC: I wasn't aware of any charging or USB bugs
01:10:05jhMikeSDavide-NYC: Use it? I think gevearts gave it the green light on transfer bugs. Charging bugs, I know of one possibility of generating an error if pulling the main charger while USB is connected under a full charge.
01:10:09Unhelpfulit's been working pretty much flawlessly for me.
01:10:34BigBambiOfficial supported build was just waiting on manual, install instructions, official downloads, sendfirm windows licence mess etc
01:10:54Davide-NYCNo no! I'm not saying there are any bugs. Just looking for pointers on what to (and how to) test for problems.
01:11:03BigBambijust use it then
01:11:48Unhelpfulthere's a firmware update, believe it or not. toffe(82?) and i have new dual-boot bootloaders patched from it.
01:12:00jhMikeSAnything I've worked on I've already beaten on the code pretty hard to discover stuff before even committing it. What I mentioned above I thought of by considering the logic.
01:12:08BigBambiI long ago ditched dual boot - it just takes too damn long to boot
01:12:13 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
01:13:01gartral1booting into OFW on sansa e200's is painfully slow
01:13:15BigBambiI don't think anyone cares about the OF
01:13:34Unhelpfulgartral1: it's a bit worse than that on S, booting into RB on a dual-boot S is painfully slow.
01:13:41BigBambiBut having a dual-boot bootloader on the beast makes booting Rockbox much slower compared to the non-dual-boot bootloader
01:13:59gartral1well, on the e200's, its the only way to get your files on it, unless youed rather use a microSD card
01:14:02UnhelpfulBigBambi: if it supported AVC, like zune (i think?) instead of just wmv, maybe
01:14:16BigBambiUnhelpful: hmmm?
01:14:31 Quit Bensawsome ("The awsome is gone :(")
01:14:40jhMikeSBigBambi: Unless things are revered and the BL loads OF. That would take some work though.
01:14:54Unhelpfulvideo. you can do better, for compression ratios, than mpeg2. by quite a bit.
01:15:02BigBambijhMikeS: Personally I don't think there is any point
01:15:30BigBambiUnhelpful: I see what you mean (confused me as when I said I don't think anyone cares about the OF I didn't mean what you thought I meant) :)
01:15:46BigBambibut I never managed to create a wmv that worked for the S
01:15:59jhMikeSBigBambi: The dual boot would be as fast as single boot at least and no mknkboot needed
01:16:01UnhelpfulBigBambi: ok, i don't care about the OF, only about things that it can perhaps do better. ;)
01:16:25 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
01:16:46BigBambijhMikeS: Sure, I just meant that I don't think that it is worth the work - the OF isn't very useful so if someone really wants it they can use the slow existing dual boot
01:17:01BigBambijhMikeS: It isn't like some of the OF needed to copy files targets
01:17:13Unhelpfulit would be a "nice to have" if we could support video post-processing on the IPU, and perhaps some acceleration.
01:17:36BigBambiDo bass/treble work yet?
01:18:18jhMikeSBigBambi: Personally, I want to use the HW EQ like an EQ, not cut it to Bass/Treble. It's just silly to force that.
01:18:31BigBambijhMikeS: That'd work too :)
01:19:41BigBambijhMikeS: Is it as fully featured as the existing software one?
01:21:10amiconnUSB itself seems to be rather reliable on the beast
01:21:22 Part pyro_maniac
01:21:22BigBambiyes, I have never had a problem
01:21:46Davide-NYCthis is all wonderful news.
01:22:05amiconnHowever, since charging got committed I can only use usb when the charger is connected. Otherwise I get an endless connect-disconnect-connect-disconnect... sequence
01:22:21amiconnI know I am supposed to test that patch...
01:22:34BigBambiamiconn: on Windows? As I don't get that on linux
01:22:34 Quit CaptainKewl (Read error: 104 (Connection reset by peer))
01:22:40amiconnYes, windows
01:22:54 Join CaptainKewl [0] (
01:23:54Unhelpfulmy experience agrees with bigbambi's
01:24:27jhMikeSBigBambi: Not quite. We could consider it to have five tone controls perhaps.
01:25:06BigBambijhMikeS: I was just trying to think about how to have them both in the ui without being confusing
01:25:11jhMikeSamiconn: I did post a patch just to disallow USB charging if you want in order to see. I'll bet it won't provide the current even though it configures it.
01:25:47amiconn[01:22:21] <amiconn> I know I am supposed to test that patch...
01:26:08amiconnI know that my ports can provide 500mA just fine (root hub as well as separate, powered hub
01:26:55amiconnIt seems that rockbox draws more than requested/configured, I just don't have an idea how to verify that
01:27:09jhMikeSDoes it happen with OF?
01:27:33amiconnOF works fine (as far as you can call MTP fine...)
01:27:56jhMikeSI have measured current draw physically and it matched (after adjusting for the .316 ohm I-sense).
01:32:00jhMikeSThinking again, I might have neglected the device current draw which the I-sense won't include with that charger plugged. I did that very early in the process. It is in the .3xx ohm range for sure.
01:32:34*amiconn is digging for the patch
01:33:29*jhMikeS can post it again
01:33:58amiconnNm, found it
01:34:17jhMikeSmeh, I take that previous statement back. I had the meter connected at the battery, not the charger, so it should be good.
01:34:48*jhMikeS wonders if it will still apply without error
01:34:50amiconnYou would need to measure between usb and device
01:35:15amiconnBattery will see the difference between input power (from usb) and device operating current
01:35:33 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
01:36:13jhMikeSThe I-sense picks up total current when USB is connected, otherwise battery only. If the main charger measurement is right, then the USB measurement is right since it's the same I-sense and I=E/R
01:36:50*amiconn will try a plain current build first
01:36:56amiconnMaybe something has changed...
01:41:00 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
01:42:14 Quit Horscht ("I got raided by the FBI and all i got is this lousy quit message")
01:42:46jhMikeSIt should draw 80-90mA until configured for 500mA.
01:47:17 Join fdinel [0] (
01:48:19 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
01:51:15 Quit n1s ()
01:51:39amiconnHmm, seems like rolo is broken on the beast
01:52:06amiconnIt begins to reboot, but then just shuts down instead of booting (including nasty emergency park)
01:52:39Unhelpfuli've also had trouble with stability after rolo boots on beast
01:53:38 Quit japc (Read error: 104 (Connection reset by peer))
01:54:05amiconnBah, murphy
01:54:30jhMikeSIt might be locking and then it can't service the watchdog.
01:55:07Unhelpful...watchdog? none of my real computers even have one, except for the software kind.
01:55:10amiconnNow the usb disconnect-reconnect problem doesn't occur. Probably the battery charge state is too high atm
01:56:31UnhelpfuljhMikeS: why is the level limited when charging from USB? i recall you saying it's a hardware limitation, but i don't think i've seen anything said beyond that, anywhere...
01:57:35jhMikeSUnhelpful: Because the highest setting draws more than 500mA whereas the USB level keeps it in range and it's the highest setting that does.
01:58:33Unhelpfulhrm... doesn't USB let you ask the host for more than 500mA? or is the USB charging hardware incapable of drawing more, even if our stack could ask for it?
01:59:35gevaertsThe USB limit is 500mA
02:03:50 Join MethoS [0] (
02:05:36 Join japc [0] (
02:06:40jhMikeSUnhelpful: it uses the watchdog timer to shut off in case it loses control of the charging hardware. the charger shouldn't be left on in a lockup even if it is voltage regulated.
02:09:19 Nick fxb is now known as fxb__ (
02:17:05 Quit Xerion (Read error: 104 (Connection reset by peer))
02:19:07 Join Xerion [0] (
02:20:51jhMikeSgevearts: Is there normally some time allowed for a device to lower its power on the bus? It could take a little time if pulling the main charger for it to lower its setting (1/2 second or so).
02:24:27 Quit tyfoo ("Carpe diem")
02:26:45 Quit DerDome ("Leaving.")
02:27:56 Part gartral1
02:30:55Davide-NYCis there a way, in Cywin, to extract from the command line?
02:31:11Davide-NYCTo a separate drive letter that is...
02:31:50Davide-NYCtar -xf rockbox.tar ???? (if I want it to extract to the "H:\" drive for instance?
02:31:51Unhelpfulextract what frome the command line?
02:31:55aurix_lexicoboth of my computers can't connect to my c240 when rockbox is installed
02:32:35aurix_lexicoI have to reinstall the OF, sync stuff, then reinstall rockbox to get stuff on their?
02:32:37pixelmaaurix_lexico: do you connect while Rockbox is running? Or with a microSD plugged?
02:33:14aurix_lexicowhen I plug in the USB with rockbox running, it boots the OF
02:33:25aurix_lexicoand the microSD card is in
02:33:41pixelmatry without
02:35:12Davide-NYCamiconn: I get a bizarre error: "tar: Error exit delayed from previous errors"
02:35:26aurix_lexicoI can't get it out of it's slot -.-
02:36:30pixelmaok, most reliable method for me was to boot the OF manually before connecting to USB
02:36:32Unhelpfulhrm, logf-over-usb is not getting me anywhere :/
02:42:39 Quit fdinel (Read error: 104 (Connection reset by peer))
02:47:10Unhelpfulin fact, it looks like plugging at least the e200 with logf over usb enabled hangs the device :/
02:50:07Davide-NYCIf I wanted to switch my Gigabeat S to "single boot" mode and I already have sendfirm.exe which nk.bin file do I need in order to have full charging support in the bootloader?
02:50:28jhMikeSbootloader doesn't have charge support atm
02:51:04Davide-NYCAh, I see. So the nk.bin file in the file is the latest one
02:52:20***Saving seen data "./dancer.seen"
02:53:02 Quit GodEater (Read error: 110 (Connection timed out))
02:53:38 Join GodEater [0] (n=ge@rockbox/staff/GodEater)
02:54:51Davide-NYCI just replace the nk.bin file correct?
02:54:55Davide-NYCthen reboot/
02:55:10 Quit alexbobp (Remote closed the connection)
02:55:26 Join alexbobp [0] (n=alex@
02:57:19Davide-NYCUm, Houston we have a problem.
02:57:24 Quit kugel ("ChatZilla 0.9.84 [Firefox 3.1b2/20081201080242]")
02:57:35Davide-NYCBig yellow triangle with the number 3 in it.
02:57:53Davide-NYC"Firmware update or restoration required. PLEASE WAIT.
02:57:58jhMikeSahh...hmmm...well. how'd that happen?
02:59:09Davide-NYCdo I power cycle this beast?
02:59:49Davide-NYCwaiting doesn't seem to do anything. Even though I've been instructed to do so.
02:59:58jhMikeSyeah, just do that.
03:00:20Unhelpfulindeed, wait, it's probably formatting one or both partitions, possibly after repartitioning.
03:00:57 Join Thundercloud [0] (
03:01:05jhMikeSI'd think you'd hear/feel it doing that. If it's just sitting dead, it's confused.
03:01:09Davide-NYCOK, I power cycled, it told me (again) to wait.
03:01:23Davide-NYCIt then asked me to connect to the PC.
03:01:28Davide-NYCI've done so.
03:01:38Unhelpfulright, and then it should ask you to send firmware.
03:02:23Davide-NYCNow it wantes me to "Update or restore Portable Media Center using PC application. *Read user manual*
03:03:30Unhelpfulyou can send an nk.bin with sendfirm at this point
03:03:54Davide-NYCso I've executed "sendfirm.exe nk.bin" and it now tells the same thing as before
03:04:03Davide-NYCI am instructed to wait
03:04:30Davide-NYCHow much time should I allot to this procedure?
03:04:57Unhelpfuli'm guessing something's not right with your nk.bin :/
03:05:29Davide-NYCLOLs I do not hear any HDD noise so I'm guessing it's not really foramttinganything.
03:06:36Davide-NYCI got the nk.bin file from here:
03:06:41Davide-NYCis that correct?
03:06:53Unhelpfulhrm... i wonder what the wait *is* for, then. also, i think it has a specific message for formatting the media partition, not that i'm about to test that, right now.
03:07:19*Davide-NYC willing to leave it in this state for a bit
03:07:52aurix_lexicostarting the OF then plugging in the USB worked on my c240
03:08:30Davide-NYCgot bored and powered down
03:09:32 Quit MethoS (Remote closed the connection)
03:12:02Davide-NYCHas sendfirm changed recently?
03:13:11Unhelpfuli think only to keep up with changes in libmtp, for the linux version
03:14:49Davide-NYCI believe I need to re-enable MTP mode on my windows machine but I do not remember how to do this. (sigh)
03:16:30Unhelpfuli'm not sure what that means :/
03:18:42Unhelpfulthe S doesn't offer any mode other than MTP in the OF... and unless you installed some custom driver that does something weird, i don't know any way to "turn off" MTP on the windows side of things
03:18:42Davide-NYCjhMikeS: I'm in a bit of a pickle
03:19:19Davide-NYCUnhelpful: your name is funny as you're being very helpful right now
03:19:32*Unhelpful gets that a lot
03:20:26Davide-NYCWhat I would like to attempt is do is to restore the beast to all original. Then install RB.
03:20:48Davide-NYCThe updater (the recently found one) is not finding the unit
03:20:57jhMikeSDavide-NYC: Are you using the libusb version sendfirm-win?
03:21:22Davide-NYCwell, probably
03:21:36jhMikeSIf so, remove the libusb device in the device manager
03:21:48Unhelpfulthat may be the issue... i didn't know we were distributing a libusb-based sendfirm-win?
03:22:46jhMikeSthat was the original one I did
03:23:59 Quit domonoky1 (Read error: 54 (Connection reset by peer))
03:24:11Davide-NYCjhMikeS: where is that in the device manager exactly? I can;t find it in the list.
03:24:20Davide-NYCAnd I don;t remember exactly what I did.
03:24:23jhMikeSIt wasn't really to be distributed but it got me up and running when the other methods utterly failed to work.
03:25:30jhMikeSDavide-NYC: Right-click My computer, properties, device manager
03:25:43Davide-NYCyes, I'm in the Device Manager.
03:25:52jhMikeSbeast plugged in?
03:27:42Davide-NYCdid it, nut the updater failed to find it.
03:28:09Davide-NYCunplugged and replugged and I think we're in business
03:28:10*Unhelpful wonders if the "modern" sendfirm-win might do the trick, now?
03:28:14Davide-NYCneed to be discovered.
03:29:51Davide-NYCI have successfully restored the gigabeat to what I believe is an updated and original state.
03:30:25jhMikeSnow get the newer sendfirm-win :)
03:32:07Davide-NYCOK I just did a wiki search and came up empty. Where do I get the newer version?
03:33:27Davide-NYCFound it:
03:33:28jhMikeSIt's an attachment on the page last I checked
03:34:49 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
03:36:56 Quit moos (Read error: 110 (Connection timed out))
03:38:43Davide-NYCOK, I downloaded the file. Extracted it. Copied the nk.bin file (the 77KB version) into that folder.
03:38:55Davide-NYCConnected the beast and ran sendfirm
03:39:05Davide-NYCnot I'm back to the error condition!
03:40:08Unhelpfulthat's weird. :/
03:40:41jhMikeSYou never just tried a deliberate restore?
03:41:45Unhelpfulhe just did an update to stock firmware, successfully...
03:42:27Davide-NYCto be precise, the updated stock firmware
03:42:36jhMikeSYeah, but...hmmm. I was thinking where you forcefully retrore it.
03:42:50 Quit XavierGr (Nick collision from services.)
03:42:56Davide-NYCwhat do you mean by "forcefully".
03:43:00 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
03:43:22Davide-NYCI just used the updater: gbs_update_1_2_us.exe
03:43:27jhMikeShold Menu+left cross (iirc)
03:43:31jhMikeSwhen booting
03:43:53Unhelpfuli can't imagine what that would do that he hasn't done yet, besides format it?
03:43:54Davide-NYCbear with me...
03:44:17*jhMikeS would just get brutish at this point
03:44:22Davide-NYCshould I sendfirm the 77KB nk.bin file first?
03:44:29jhMikeSbefore what?
03:44:47Davide-NYCbefore putting the device into recovery mode?
03:45:00jhMikeScould try it. won't hurt anything.
03:45:00Davide-NYCor is it already in recovery mode
03:45:23Unhelpfulit's in "recovery mode" if you've told it to be during startup.
03:46:08Davide-NYCNumber 3 triangle - repair recovery
03:46:18Davide-NYCNumber 4 triangle - removal of media
03:46:25Davide-NYCNumber 1 triangle Connect
03:47:00Davide-NYCWhat does that mean?
03:47:03jhMikeSSo Far So Good?
03:47:17Davide-NYCNumer 2 triangle - Update
03:47:43Davide-NYCINow what?
03:47:48Unhelpfulyou could try building a beast bootloader yourself? although i can't see why that would make a difference.
03:48:09Davide-NYCbuild the nk.bin file you mean?
03:48:21 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
03:49:14Unhelpfuli believe the output is named bootloader.bin when you build it, but, yes, you could try that.
03:49:37 Join casainho [0] (
03:49:49casainhohello :-)
03:50:08Davide-NYCplease refresh my memory: I configure, then make... and then?
03:50:18Davide-NYCcasainho: hello
03:50:22UnhelpfulDavide-NYC: PROFIT1
03:50:50Davide-NYCUnhelpful: really? allcaps like that?
03:50:51casainhoI am doing a printf("Rockbox Player Bootloader"); to my display however I am seeing just 3 little lines... is there a way to put fonts with biger size?
03:51:09Unhelpfulahah! it looks like it's PLUGIN_IRAM_INIT where things go wrong with FS #9770 :D
03:52:18Unhelpfulanybody have an idea why one plugin might hang on PLUGIN_IRAM_INIT and another not?
03:53:02jhMikeSPLUGIN_IRAM_INITs use api, not rb usually
03:54:01UnhelpfuljhMikeS: the patch i'm working on does away with passed API, and has a global one in the header, initialized by the loader. so there is no "api" anymore, just rb.
03:54:03jhMikeSsome delay the IRAM init as well so playback doesn't have to stop
03:54:43Davide-NYCOK, so I compiled the bootloader (I think) and I copied the nk.bin file to the sendfirm directory and "sent" the "firm".
03:55:49Davide-NYCrebooted the payer, saw the RB bootloader, it told me "file not found", so I plugged in USB to restore rockbox and poof! I'm back to where I started with an error conditionl.
03:55:52Davide-NYCany ideas?
03:56:47jhMikeSif you get "file not found" use rockbox.tar within '/Content/0b00/00'. If those don't exist, create the folders and put the .tar there.
03:57:27jhMikeSjust use 'make tar' instead of 'make zip'
03:58:05Davide-NYCThe beast is in MTP mode so all I see is "Storage".
03:58:16Davide-NYCwhat am I not understanding here?
03:58:52UnhelpfulDavide-NYC: it may be enough just to copy the .tar to it in that mode.
03:59:08Davide-NYCnot untar it?
03:59:18UnhelpfuljhMikeS: hrm... well, grep says that rockdoom uses it, as well, and i would assume it's called before you actually start play
03:59:25UnhelpfulDavide-NYC: no, the bootloader should do that.
03:59:38Unhelpfuland i've been able to play doom, but mpegplayer crashes :/
04:00:20Unhelpfuli'll have to see if i can test pacbox, rockboy, or zxbox, i guess.
04:00:46jhMikeSDavide-NYC: /Content/0b00/00 is where the MTP stuff ends up (until it changes that after much activity).
04:02:51Davide-NYCOk I've resent nk.bin
04:04:34Davide-NYCno change on the device screen. Before I reboot I tried to copy the tar file but I get a "Cannot Copy File" error from windows because tar files are not supported
04:05:24Unhelpfultry connecting it in bootloader MSC mode, since you got the bootloader running
04:05:25jhMikeSI guess do it in bootloader USB mode
04:05:39Unhelpfuland then create /Conent/0b00/00, and copy rockbox.tar there
04:05:41Davide-NYCNo, I do not have the bootloader running.
04:05:54jhMikeSDidn't you just send nk.bin
04:06:11Davide-NYCbut I haven;t rebooted yet
04:06:35Davide-NYCUnplugged beast, let it reboot but replugged the usb cable quickly.
04:06:36jhMikeSCable plugged of course :)
04:06:43Davide-NYCNOW I'm in USB disk mode
04:07:06*Unhelpful adds more debug splashes inside plugin_iram_init :/
04:07:50jhMikeSUnhelpful: Did you make sure everything is ok in PLUGIN_IRAM_DECLARE?
04:08:17Davide-NYCjhMikeS: how does one reboot the beast from recovery mode after having "sendfirmed" without unplugging the usb cable?
04:09:10jhMikeSIF you plug the USB while it's off, it will start itself. You can just use the batt switch too though.
04:09:21 Quit Thundercloud (Remote closed the connection)
04:09:59UnhelpfuljhMikeS: that macro looks like it just declares a some arrays?
04:11:39Davide-NYChad a bootloader in USB diskmode. Copied the rockbox.tar file over to the tfat partition, unmounted. POOF! Back to square one.
04:12:12Davide-NYCStaring at the Number 2 triangle - Update or Recover. :-(
04:12:13jhMikeSUnhelpful: I guess so. Been awhile since I looked at that stuff
04:12:14Unhelpfuli put some debug splashfs with filenames/line numbers in start_plugin, the last one that displays is the one right before PLUGIN_IRAM_INIT. i just added some more inside plugin_iram_init, i'll see if i can tell where in that the trouble is. i wish it would just be broken on sim. ;)
04:12:24UnhelpfulDavide-NYC: that's bizarre.
04:12:47Davide-NYCI saw a "File not Found" error just before ending up back in the recovery mode
04:13:28jhMikeSDo you ever switch off the device when retailos is fully running? It can cause problems.
04:14:20Davide-NYCAt this point I'm going to go back to a fully retail situation.
04:14:34saratogaI'm having trouble understanding how ci->advance_buffer works
04:14:44saratogahow does it know how much to advance the buffer?
04:16:35jhMikeSsaratoga: by the "amount" param?
04:16:48Davide-NYCOK, I got it. You have to untar the RB file. The BL does not extract it.
04:16:56saratogai thought the argument was a pointer to the buffer
04:17:13 Join massiveH [0] (
04:17:15Unhelpfulok, now that's even more bizarre, a splashf after the last line of code in plugin_iram_init executes, but one after the call to same does not?
04:17:28jhMikeSDavide-NYC: The bootloader extracts it if it finds it
04:17:30 Quit massiveH (Read error: 104 (Connection reset by peer))
04:17:46jhMikeSIt has to be in the correct directory, that's all
04:17:51Davide-NYCI suspect the BL is not finding it.
04:18:17Davide-NYCWait! Is it supposed to be in the same dir as the nk.bin?
04:18:22jhMikeSIf you place it there in MSC mode, it should. THen you'll see "untarring..."
04:18:47UnhelpfulDavide-NYC: no, it's not even supposed to be on the same partition.
04:19:11Davide-NYCthen I can almost certainly confirm that this untarring business does not work
04:19:14jhMikeSDavide-NYC: ah, you put it with nk.bin? It goes on the media partition.
04:19:40Davide-NYCno I did in fact put it on the media partition.
04:19:50UnhelpfulDavide-NYC: you put it in /Content/0b00/00 on the media partition
04:19:56jhMikeSDavide-NYC: I've used it myself in the past and moos did as well just recently to get rid of "file not found"
04:20:03Unhelpfuland there shouldn't be an nk.bin on the media partition
04:20:56Davide-NYCFor clarity, on my first partition there was just nk.bin and on the second partition there was just rockbox.tar and this did not work.
04:21:24Davide-NYCbut the rockbox.tar was in the root of the media partition.
04:21:45Davide-NYCnow I am back to the error condition.
04:21:53Davide-NYCOy vey!
04:23:46Unhelpfulrockbox.tar should not be in the root of the media partition, it should be in /Content/0b00/00... and the error thing each reboot suggests something else is going wrong.
04:24:32Davide-NYCI will sendfirm again and go into USB disk mode. Then I will place the rockbox.tar file in the correct folder. Stay tuned.
04:25:51*jhMikeS wonders why he didn't just do what was asked in the first place with the directories :\
04:26:39casainhowhat file should I include to be able to use printf()?
04:26:42*Unhelpful assumes that the device is not being simply unplugged after writing the file... right?
04:26:54Davide-NYCOK, SFSG. But I'm afraid to unmount.
04:27:08Unhelpfulcasainho: use DEBUGF
04:27:34casainhoUnhelpful: ??
04:27:48Unhelpfulrockbox defines a DEBUGF for debug output.
04:28:19casainhoI will try
04:28:41jhMikeSDavide-NYC: If you're already in MSC with rockbox.tar in /Content/0b00/00 then you should just unmount it already.
04:28:55Davide-NYCUnmounted, saw the unarchiving message
04:29:06Davide-NYCin rockbox
04:29:18Davide-NYCin rockbox usb mode
04:29:29jhMikeSTurn it off and reboot it to make sure before doing anything.
04:30:25Davide-NYCshut down (using power button), turned back on (using power button.
04:30:47*jhMikeS hopes he's joking
04:30:48Unhelpfulyou're doing the "safely remove" or whatever before you unplug, right?
04:31:46Davide-NYCAFAICT the unit does not recover from a shutdown
04:32:07Davide-NYCnot with the nk.bin I downloaded or the nk.bin I compiled.
04:32:12*jhMikeS wonders if that beast got reflashed somehow with new boot code.
04:32:21*Unhelpful wonders what you're doing wrong
04:32:33*jhMikeS has to go for a bit
04:32:38Unhelpfulkidding... and for that matter, i wonder what i'm doing wrong. :/
04:33:09*Davide-NYC wonders if the devs upgraded their beasts using the new updater first?
04:33:47Davide-NYCmaybe the updater changes some code that make the nk.bin incompatible.
04:34:25 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:35:00Unhelpfuli did my upgrade with sendfirm, sorry
04:35:37Davide-NYCthere are three bin files in the update.
04:36:10Davide-NYCnk, pmcboot_secure, and Recovery
04:38:16Unhelpfulif you dig through old versions of the wiki, i believe there's an archive with those files as ripped from the HD of a "fresh" beast.
04:38:36Unhelpfulit might be worth copying them over in BL USB mode, to see if you can downgrade that way
04:41:08Unhelpfulhrm, so both doom and pacbox seem to be fully working, but mpegplayer is hanging :/
04:41:16Davide-NYCif I just copy them to the first partition, they will be installed?
04:42:27 Join FlynDice [0] (
04:43:00Unhelpfuli've updated the rockbox bootloader that way, i would expect the others to work as well.
04:44:13 Join int21hexster [0] (
04:45:02 Join Barahir_ [0] (
04:50:59Davide-NYCReverted to OF version1.1. Then sentfirm. Then installed tarfile. Then POOF! BACK TO ERROR CONDITION.
04:52:24***Saving seen data "./dancer.seen"
04:52:32Davide-NYCNow I've upgraded/restored to OF version1.2 and am calling it a night.
04:52:46Davide-NYCSingle boot RB is epic fail for me.
04:53:20Davide-NYCDual boot may prove to be luckier, but it won;t be this evening.
04:53:38Davide-NYCThank you both for being so helpful;.
04:53:55 Quit Davide-NYC ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
04:59:35 Join blkhawk- [0] (
05:02:12 Quit Barahir (Read error: 110 (Connection timed out))
05:04:58 Join troy_ [0] (n=toppy@
05:05:28 Quit troy_ (Client Quit)
05:05:32 Join troy_ [0] (n=toppy@
05:09:52 Quit casainho ("ChatZilla 0.9.84 [Firefox 3.0.5/2008121622]")
05:12:28 Join FlynDice2 [0] (
05:13:23 Quit FlynDice (Read error: 110 (Connection timed out))
05:13:45 Quit FlynDice2 (Client Quit)
05:14:12Unhelpfulis it possible that flush_icache could be *causing* the hang, but that the hang occurs after it returns, somehow? that still makes no sense, since other iram-using plugins work :/
05:14:31 Join FlynDice [0] (
05:16:12FlynDice Anyone know if can I somehow use the fuze button-fuze.c code to get the scrollwheel to function on an e200v2 (at all)?
05:16:16 Quit blkhawk (Read error: 113 (No route to host))
05:16:33 Nick blkhawk- is now known as blkhawk (
05:17:09Unhelpfulthe fuze devs i know of aren't in at the momenht
05:17:33FlynDiceI'll read the logs if they stop by...
05:20:59 Quit XavierGr ()
05:26:07 Quit HellDragon (Remote closed the connection)
05:26:59saratogaFlynDice: i don't have an e200v2 but i think it should be similar if not the same as the fuze
05:30:33FlynDiceWell I just tried substituting the button-fuze.c code and got some errors, thought maybe there was more to it
05:31:52 Join MOHDMACH [0] (
05:32:01 Quit Seed ("cu, Andre")
05:34:08MOHDMACHI have added the following to the end of a .wps and it has no effect: %?mv<b|c>
05:35:01MOHDMACHb= volume bitmap and c = blank bitmap
05:36:14MOHDMACHThe volume bitmap is a single progress bar physically.
05:36:29MOHDMACHMust I be using bitmap strips for this?
05:40:15saratogaFS #9775 - Premature and multiple storage_sleep() in buffering.c looks interesting
05:40:24Unhelpfulyour tag does *nothing*? it doesn't display a letter b or c on the screen, even, which is what it looks like it would do?
05:41:18Unhelpfulhrm, removing all the IBSS_ATTR in mpegplayer makes it work. i've been adding them back in and it keeps working, so far...
05:45:38MOHDMACHThat is the thing it has no effect on the physical properties of the the WPS.
05:45:57*Unhelpful wonders if this might have to do with the total size of data put in iram by the plugin
05:45:57 Join HellDragon [0] (
05:48:36MOHDMACHBut all I did was added that tag at the end of a pre-made wps file using its the letters that correspond to the loaded bitmaps. Loaded with the "%xl" tag.
05:50:02MOHDMACHI did NOT set up viewports or added anything else to the tag. I was just playing around with the tag to see if it would have any effect and for simplicity before I get into anything complicated like the tags you helped me with yesterday.
05:51:29Unhelpfulyour tag might be off the screen, and besides that, all it would do is display the letter b or c.
05:52:09MOHDMACHIt would display it based on if your adjusting the volume or not right?
05:52:57Unhelpfultext saying "b" or "c", yes. and again, your tag may not appear on screen at all, if the line before it is at the bottom of the screen.
05:53:41MOHDMACHIt would display b if the volume is being adjusted and c if nothing is touched?
05:54:21MOHDMACHIt is last tag. Might that be an issue?
05:54:47Unhelpfulthose lettels, yes, not some bitmap, and as i just said, it might not display anything, because if you just add a line at the bottom of an existing wps, that tag will be below the line above it, which may place it entirely off the screen.
05:55:31MOHDMACHIt is after: %ac%pc %pr
05:56:33Unhelpfulthat means pretty much nothing in terms of whether the tag you added at the bottom will appear on-screen or not.
05:57:26MOHDMACHSo you are saying that if I position it right it should display the bitmaps corresponding to the letters? Specifically it would display b if the volume is being adjusted and c if it is left untouched?
05:58:25Unhelpfulno, i am saying that if you position the tag as you have quoted it here that it will display the actual text "c" or "b", not a bitmap at all.
05:59:02MOHDMACHAh the actual text.
05:59:21MOHDMACHOk so how do I get it to display the images.
06:00:06MOHDMACHSorry I forgot the ? mark.
06:00:46 Quit saratoga ("CGI:IRC (EOF)")
06:00:54Unhelpfulyou really need to understand CustomWPS for yourself before you're going to get anywhere with this. it tells you exactly how to do what you're trying to do, and why it's not working.
06:02:02 Quit FlynDice ()
06:02:56 Quit aurix_lexico ("Leaving.")
06:03:58MOHDMACHDoes it reference how to use a single bitmap for volume without using bitmaps strips or separate images for volume? Or is that not possible in rockbox yet.
06:05:29MOHDMACHSimilar to the progress bar.
06:05:52Unhelpfuli think the progress bar may be the only thing that works that way right now, but i can't be certain of that.
06:06:00 Join nibbler_ [0] (
06:06:03MOHDMACHOk then.
06:06:27MOHDMACHThanks for the help. :)
06:07:29 Quit MOHDMACH ("Bye")
06:15:54 Join pixelma_ [0] (n=pixelma@rockbox/staff/pixelma)
06:16:02 Quit amiconn (Nick collision from services.)
06:16:04 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
06:16:08 Quit pixelma (Read error: 110 (Connection timed out))
06:21:03 Quit nibbler (Read error: 110 (Connection timed out))
06:22:03 Quit HellDragon (Read error: 104 (Connection reset by peer))
06:30:08 Quit midkay_ ("Leaving")
06:52:26***Saving seen data "./dancer.seen"
06:59:31lucentwhat changed recently in the SD handling code? anything?
07:00:07lucentUnhelpful: btw I got distracted and did not yet test the 8->16
07:00:27Unhelpfulno problem, thanks for being a guinea pig :)
07:00:53Unhelpfulit may neet to go quite a bit higher before there's trouble again, actually, so maybe we should start from the other end?
07:01:30Unhelpfultry, say, 188 and see if you have trouble, and go 8 up or down from there?
07:10:10 Join Mahees [0] (
07:10:34 Join HellDragon [0] (
07:10:58 Part Mahees
07:11:19lucentUnhelpful: wtf though, 188 is a multiple of 8? I don't understand
07:11:57Unhelpfulyou're right, it's not :/
07:12:03Unhelpful184, then?
07:12:23Unhelpfulit really wouldn't matter unless you used a theme that had mono bitmaps
07:12:27Unhelpfulthen it might break something terrible
07:12:46 Join zeleftikam [0] (
07:13:07zeleftikamI need help removing the Rockbox bootloader from my Mac so I can give my Sansa C240 to someone else. Help?
07:13:36zeleftikamI removed the .rockbox directory and all the files, but the player flashes the boot screen over and over now, I suspect because the RB bootloader is still there
07:14:00scorcheuninstall instructions are in the manual...
07:15:00zeleftikamcouldn't find :(
07:15:03zeleftikamwhat chapter?
07:15:55zeleftikamoh, did find, in the "Installing" section
07:16:18lucentUnhelpful: you're asking a guy who doesn't know, if you have a suggestion about what to regression test, then I'll go do it
07:16:36lucentI'm not going to be very productive trying to guess at what the test should be
07:19:37Unhelpfullucent: i don't know what the test should be, either. read sizes somewhere in between 8px and 224px cause failure. kugel had all of his cover art load successfully, so it's possible that the max size for that is our lower bound. if we don't have him around to ask... actually, i might have a better idea, which doesn't involve rebuilding over and over :)
07:20:20Unhelpfulis the rockpaint plugin working on fuze?
07:20:35lucentya... kind of
07:21:04lucentthere's a button problem on the fuze 8gb where the wheel causes a ghost keypress
07:21:16lucentkugel says it's weird and doesn't happen on his fuze
07:21:27lucentsomething ain't right on my fuze
07:21:29Unhelpfulok, but can you load a bitmap in it, and exit it, and not crash the device? that will be good enough :)
07:21:52lucentwill try now
07:22:42lucentUnhelpful: shoot, no
07:22:52Unhelpfulbecause if that much works, instead of messing around with rebuilding and reinstalling, you can bump BM_MAX_WIDTH up to 224, do just one build, and then test by loading various sized bitmaps in rockpaint
07:22:56lucentghost buttonpress exits the app
07:23:36lucentUnhelpful: kugel could though
07:24:03lucenthe says stuff works
07:24:19Unhelpfulright... i'll prep some for next time i see him, then. probably nice solid blocks of colors in vertical stripes will make failure obvious :)
07:26:12lucentI want an expert who knows about SD code
07:26:35lucentI want to sit down and figure out why simple access of data is so nutters on the fuze 8gb
07:27:37 Part zeleftikam ("Leaving")
07:29:27Unhelpfulthat may be specific to the particular hardware rather than SD in general :/
07:29:31Unhelpfulis his 8GB as well?
07:30:20lucentkugel owns a 4GB fuze v1, I own a 8gb fuze v1, and I've confirmed all weirdness that happens on my player also happens on a seperate fuze 8gb v1
07:31:04lucentUnhelpful: how can i get a listing of buttons being activated on the Fuze?
07:31:22lucentis there a plugin or test mode which displays active buttons?
07:31:34Unhelpfulnot to my knowledge. i'm not even sure how you'd exit such a thing :/
07:31:43lucentyou'd power off
07:32:10lucentI really need to know why the ghost buttons are happening, and which button it is
07:33:15 Join Owner [0] (n=4cf498a9@gateway/web/cgi-irc/
07:34:02Unhelpfuli don't know of such a thing, and i gather that reading all of the buttons on the fuze involves some weirdness
07:34:39OwnerDoes anyone know if there is or has ever been any thought given to porting this to an LDP600? or is there no way of doing that?
07:34:59lucentI don't know what an LDP600 is
07:35:11OwnerLexar LDP600 MP3 player
07:35:23LloreanOwner: Ports happen by interested owners of the devices doing quite a large amount of work.
07:35:23lucentno, what chipset
07:35:47lucenthave you taken it apart and made high-resolution pictures available?
07:35:53Ownerhow do i find out what chipset
07:35:57LloreanOwner: The best way to encourage a port for a specific player is to go somewhere where owners of the port gather (fan forum, or similar) and try to drum up more people to work on it with you.
07:36:02Unhelpfulin other words, somebody has to have the device, the skills, the time, and the interest. and i've not heard anybody else mention having this device.
07:36:31Ownerits the black one
07:36:40lucentthe specifications on it don't sound very competitive vs. the SanDisk Sansa Clip
07:37:17Owneri would be willing to take a crack at it but I have never heard anyone trying to reverse engineer the firmware to figure the thing out
07:37:26Ownerso I wouldn't even know where to begin
07:37:39Ownersounds like it probably wouldn't be worth the effort
07:37:49lucentyeah, I agree
07:38:14Ownerwell the answer is always no if you don't ask... so i figured I would at least ask
07:38:16lucentOwner: concensus is that unless you do it, it's not even likely to get a developer's attention with that thing
07:38:48lucentit doesn't look familiar to any chipset already supported by Rockbox
07:38:54Ownerhow do i find out what chipset it is? to see if there is even a development environment for it
07:39:00lucentit could be, but then, it's so limited
07:39:05Lloreanlucent: You found internal scans?
07:39:15lucentoh you'd crack it open and take some high resolution photos
07:39:17Ownerthat's what I need? internal scans
07:39:34lucentthen call a lot of people and break some kneecaps with kindness
07:39:37OwnerIve not had the need to search for someone cracking one open
07:40:20Lloreanlucent: You really shouldn't make suppositions about its chipset just from "look".
07:40:30lucent"Hi I'm doing my grad student project thesis thingy over at University of Amazing Ivy League School and need to request a datasheet on FOO, how much will this cost me?"
07:40:57Ownernice =)
07:40:59Lloreanlucent: Please don't encourage people to lie on behalf of the project.
07:41:20lucentOwner: if you want a replacement, we could always use more bodies focused on the Sansa Clip
07:41:28lucentit very nearly runs rockbox
07:41:32lucentbut not quite
07:41:45LloreanIf we can legally use a datasheet, they'll provide it given the real reason we want it. If we can't use it legitimately, acquiring it like that just makes it even more problematic.
07:41:45OwnerMaybe I can ask them to release the source code to the latest firmware since it doesn't seem they are doing anything more with it
07:42:13kadobanOwner: i would be shocked if they even respond
07:42:50kadobanbut i suppose it never hurts to ask, hehe
07:42:50lucenthe's right that you never get anywhere without asking
07:42:50lucentbut uh... don't bet your lunch money on getting a response
07:42:50Ownersorry about the nick, it was supposed to say Owner of LDP600
07:43:25 Join Darksair [0] (n=user@
07:43:27Owneryeah, but you don't know unless you ask
07:43:45Owneror at least get the internal specs from them
07:43:46Unhelpfulspaces aren't allowed in nicks, along with numerous other characters.
07:43:59Owneryeah I should have known that
07:43:59lucentOwner: Fuze and Clip are my favorite hopefuls for rockbox development
07:44:35LloreanOwner: Your best bet is to get an official firmware update, and see if it's in any format we already know as well as opening it and identifying chips. Most players can be opened and re-closed without permanent damage (at least beyond superficial)
07:44:52OwnerI actually stopped carrying my LDP600 around when work upgraded my blackberry, now I use my blackberry for music, I wish it had a better system for playing music
07:45:17Ownerlloreean I have the last firmware update
07:45:50UnhelpfulLlorean: any pointers on where i might find out just how much iram a particular target has? and, would the plugin linker catch it if a plugin tried to stuff too much into iram?
07:45:55Ownerdoing a strings on it right now
07:46:32LloreanUnhelpful: There may be a define for it in the target config, since it's shared between core and codecs usually. Otherwise, I don't know, sorry.
07:46:33Ownerat the very beginning of the rom it says TCC720 DEMO
07:46:55LloreanOwner: We do have some tools for working with TCC devices, though there's no finished port for any of them yet.
07:47:17Unhelpfulno problem, figured i'd just try asking somebody who's awake before i went too crazy digging for it.
07:47:26lucentmaybe Telechips?
07:48:00Unhelpfuldemo?! that's fantastic. :)
07:48:53LloreanProbably just means they're using reference code or something.
07:49:55lucentYep, looks like that's the data sheet for it
07:50:48lucentit's the datasheet for the TCC76X which has references vs. the older TCC720
07:53:40Ownerfrom the wiki it looks like it is the dual CRC firmware version
07:54:21Ownermaybe not
07:55:01lucentOwner: there are easier ways to get involved with Rockbox project :)
07:55:08lucentdo you have a background in programming by chance?
07:55:14 Part int21hexster
07:55:19 Join at0m [0] (n=a548c80b@gateway/web/cgi-irc/
07:56:31OwnerI used to be a programmer, still program but mostly system stuff and j2me when i find a program i want to make work on my blackberry
07:57:11 Quit Darksair ("Use the Force, Luke!")
07:57:28 Join potato_ [0] (
07:57:28Ownermaybe dabbler is a better word for me =D If it interests me I will learn it to do what it takes to make it work
07:57:44*lucent grins
07:58:21lucentdoes maintaining the windows rockbox tools sound interesting? I know some folks here that would consider it "challenging"
07:58:23Ownerside note, this rom says it is Ver:161
07:58:57Lloreanlucent: Which windows tools are you referring to?
07:59:12LloreanMost of the tools are cross-platform.
07:59:24OwnerI do windows and unix/linux
07:59:52lucentthe code for that I think is in rbutils/ in the source tree
07:59:57Ownerwell i should word it the other way around
08:00:04 Quit BHSPitLappy (Remote closed the connection)
08:00:23Owneri've been doing unix since 1982
08:00:50Ownerlucent: jumping back to the web page to take a look
08:01:08lucentOwner: great! if you look at the Twiki doc index, there's a whole list of Wiki docs
08:01:15 Join bertrik [0] (
08:01:18lucentone or three of them say Develop* something
08:03:12potato_i have a plugin which i aim to make runable on all targets, how many screen shots do i need in the manual?
08:03:38Lloreanpotato_: One for each unique screen, basically.
08:04:00Unhelpfulvery, very strange... it's not *purely* about the size of data the plugin puts in iram :/
08:04:06LloreanIf your plugin's mono, that's each screen size. If your plugin uses color, you'll need a few more for cases where there's color and mono screens with the same dimension.
08:04:33Unhelpfulmpegplayer uses far, far more than doom, though
08:05:19potato_oh dear =O. is there a way to quickly obtain screen shots from all the targets without having to do individual builds?
08:05:46Lloreanpotato_: Not really, you'd basically need to do a simulator build for each unique screen.
08:07:16potato_ok, i hope my plugin actually gets included in rockbox when im done =]
08:07:30Lloreanpotato_: Is it in the tracker?
08:07:52potato_yep, its missing a manual, which i havn't had time to work on till now
08:08:45lucentautomating the builds wouldn't be too terribly difficult, pipe some answers into tools/configure and so on
08:09:02lucentwith sim... oh wait no that would take a while
08:09:17LloreanBuilding the sims ought to be quite quick with ccache.
08:09:28 Quit sarixe ("Connection reset by the motherfucking peer")
08:09:32LloreanI'd imagine it's the actual creation of the screenshots that'd take a while since that requires user interaction.
08:10:08Unhelpfultools/ can do some of the work for you
08:10:14Unhelpfulif i recall the path to it properly
08:10:24LloreanIs that the build-everything script?
08:10:40potato_i dont really know what all of this means. im on windows/cygwin so im quite unfamiliar with pipe and such
08:10:49kadobantools/release/ perhaps?
08:12:02 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
08:12:03UnhelpfulLlorean: yes. i have a few local changes that should probably be put on flyspray for comment at some point... it can already to selected builds, though, instead of the whole range.
08:12:15OwnerThanks everyone! I will definitely look into rockbox some more!
08:18:33lucentOwner: happy hunting
08:19:00Unhelpfulok, it's definitely not about size. i've narrowed it down to three structs that i can remove IBSS_ATTR from to make it work
08:24:15Ownerrom firmware 107 for anyone who would care to look
08:24:37Ownerthanks again! have a good night/morning/afternoon =D
08:25:43Unhelpfuland down to one. i'm starting to think that i've exposed a bug, rather than creating one...
08:31:35 Join Rob2222 [0] (
08:38:38Unhelpfulany idea why i need to remove the IBSS_ATTR from the struct mad_synth in apps/plugins/mpegplayer/audio_thread.c to make FS #9770 work? if i don't it hangs on returning from plugin_iram_init - though not before returning, as a splash placed after the last line of code in that function still executes.
08:41:36 Quit Owner ("CGI:IRC (Ping timeout)")
08:42:23 Quit bertrik (Remote closed the connection)
08:44:39 Quit obo (Remote closed the connection)
08:44:57Unhelpfulmake me think that maybe the stack or return address is getting corrupted... but it's just another chunk of data, with a bunch of other chunks that are being handled the same way.
08:49:24 Quit tessarakt (Read error: 110 (Connection timed out))
08:52:27***Saving seen data "./dancer.seen"
08:59:20 Quit Rob2223 (Read error: 110 (Connection timed out))
09:02:35 Quit jhulst (Read error: 60 (Operation timed out))
09:02:38 Join Darksair [0] (n=user@
09:22:03 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
09:25:56 Nick fxb__ is now known as fxb (
09:28:02 Nick Barahir_ is now known as Barahir (
09:30:56 Quit potato_ ("Trillian (")
09:35:22 Quit write__erase ("Leaving")
09:43:50 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
09:47:18 Join nplus [0] (n=nplus@141.25.Globcom.Net)
10:10:40amiconn_Unhelpful: On what target are you testing your patch (where you have the trouble with mpegplayer)?
10:11:07amiconn_If it's PP, I bet the problem is COP related, as mpegplayer is one of the few plugins using both cores
10:11:25 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
10:11:41Unhelpfulamiconn: e200. it's very bizarre, if i move either of the two larger data structures out of iram, it works.
10:12:52Unhelpfuland, especially weird, if i move the mad_synth structure out of iram, and pad out the large array the video decoder uses so that the ibss section is the same size as before - it works.
10:13:20Unhelpfuli don't think it's COP related, since it doesn't ever get that far
10:14:17Unhelpfulif i put a return in plugin_start, right after PLUGIN_IRAM_INIT, it still hangs. :/
10:16:15Unhelpfulalso, if you want to test the greylib scaler, and the aspect ration compensation, on your ondio, i have a build for it.
10:20:11amiconnTry putting the *(hdr->api) = &rockbox_api; a few lines up in plugin.c. It must go *before* the invalidate_icache();
10:20:26amiconnI would put it directly before plugin_loaded = true:
10:21:15amiconnRight now fs #9770 does not guarantee that the cop sees the proper api pointer
10:23:33Unhelpfulhrm... and that would matter, even if the plugin never makes it to where it actually *does* anything?
10:24:13amiconnIt's definitely a bug. Maybe it's not the one that crashes mpegplayer though
10:28:30amiconnAhem, what is 'rb' in the PLUGIN_IRAM_INIT "calls" in your patch?
10:29:40amiconnIiuc this resolves to rb->plugin_iram_init(....) but then your patch removes rb...
10:30:50 Quit Horscht ("electromagnetic radiation from satellite debris")
10:31:25amiconnHmm, it seems plugins still use rb-> to call api functions
10:31:34*amiconn is trying to understand how that works
10:34:20amiconnAh, you put that declaration into PLUGIN_HEADER. But how does that pointer get initialized?
10:35:35*amiconn thinks he understands now
10:36:12*amiconn however does not understand why Unhelpful didn't put the api pointer directly into the header, instead of using a pointer to a pointer
10:37:18Unhelpfuli wasn't sure how to do that, without the plugins having to access it via the header structure... although, i imagine, a linker script change could fix that up?
10:38:05 Quit at0m ("CGI:IRC")
10:40:04amiconnI'm thinking about a variable aliasing trick
10:40:37amiconnBut it should work either way as-is, except for the necessary dualcore fix...
10:42:17 Join Asterisk_ [0] (n=George@
10:42:56Asterisk_I'm getting an error message here. Cygwin Setup Unable to get from <>. how do i fix it?
10:45:45amiconnUnhelpful: Your patch ignores most test plugins. I also think that codecs should be changed the same way when the method works properly
10:45:47linuxstbAsterisk_: I think you need to run "setup.exe -X" from the command line
10:46:31Asterisk_how so
10:47:20linuxstbYou need to open a command prompt, use the "cd" command to change to the directory containing setup.exe, then type "setup.exe -X"
10:47:22 Quit Darksair ("(define zero (lambda (f) (lambda (x) x)))")
10:47:47 Join Jaykay [0] (
10:49:36 Quit JdGordon|afk (Read error: 104 (Connection reset by peer))
10:49:48Unhelpfulwhich test plugins? all the test_*.c files under plugins have the new signature in my local repo... very odd if it's different in the patch.
10:50:02Asterisk_should it be or <>
10:50:20linuxstbThe first one
10:50:37Unhelpfulthe new signature for plugin_start, that is
10:50:47linuxstbAsterisk_: Which instructions are you following? This page is the easier one -
10:51:24 Join JdGordon|afk [0] (n=jonno@rockbox/developer/JdGordon)
10:51:27Asterisk_that did the trick.
10:51:34Asterisk_hmm. that was odd
10:52:25Asterisk_can I idle here if i have more questions later?
10:52:26 Join Ryc [0] (n=cb3bc60b@gateway/web/cgi-irc/
10:52:30***Saving seen data "./dancer.seen"
10:52:52linuxstbAsterisk_: Of course - there are 132 people idling here...
10:53:02Asterisk_heh just checking. :)
10:56:31 Join rvvs89 [0] (n=ivo@pdpc/supporter/base/rvvs89)
10:56:59RycGreetings, wondering if someone please give me write permission to wiki, for updating CFModGuide with a successful install. WikiName is RycPat.
10:57:04Unhelpfulhow would you suggest i arrange that __header.api have the same address as rb?
11:00:30 Join moos [0] (
11:00:39Unhelpfulamiconn: before the very last invalidate_icache(), at line 741? or one of the others further up?
11:00:54amiconnHmm, the test_ plugins are indeed in the patch
11:02:07*amiconn wonders why firefox' search function was lying to him
11:02:49Asterisk_cuz it's mean
11:02:51amiconnWeird patch though - I'm used to use -p0. This one needs -p1
11:03:23Unhelpfulgit generates them that way... but i really don't know if i could live without local branches
11:03:27amiconnHmm, and it doesn't apply
11:04:04Unhelpfullet me see if it needs a resync, and move the api init up as well
11:05:29 Join {phoenix} [0] (
11:05:43amiconn4 failed files:
11:05:59amiconnThis is against latest svn without any other change
11:07:24Unhelpfulweird... perhaps the branch was not synced to master when i generated the patch :/
11:09:27JaykayJdGordon: wheres the advantage of r19622? it caused quite a lot of bugs.....
11:10:08JaykayjdGordon|afk: ping (i used the wrong name)
11:12:48amiconnUnhelpful: Yes, before the last one. But it would be more logical to move it a bit further up, directly before plugin_loaded = true; (line 727)
11:13:00pixelmaRyc: is that your real name? (sorry, hopefully you take no offense it just sounds unusual)
11:13:52Unhelpfulamiconn: ok. i need to clean up some whitespace changes i've introduced while changing some of the pluginlib function signatures, i'll move that call and put a new patch up.
11:14:26amiconns/call/assignment/ ?
11:15:04Unhelpful...indeed. it's been a long, and rather unusual, night.
11:15:49 Join at0m [0] (n=a548c80b@gateway/web/cgi-irc/
11:16:30Unhelpfuli see that your proposed location is still after the target/sim split in the loader
11:17:09 Join tessarakt [0] (
11:17:12 Join flydutch [0] (
11:21:53 Quit BHSPitLappy (Remote closed the connection)
11:27:50amiconnWhy split code when it's not necessary?
11:30:54 Join Darksair [0] (n=user@
11:33:03Unhelpfuli suspect at least one of the conflicts to be due to how git handles deletions :/l
11:34:26 Nick JdGordon|afk is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
11:39:30 Join pyro_maniac [0] (n=jens@
11:52:50 Quit Asterisk_ ("KMSMA")
11:56:04Unhelpfulamiconn: new patch up. your fix sadly doesn't fix mpegplayer... i'm thinking i should add some debug stuff in plugin_iram_init to see if the stack is getting messed up somehow, because i can't see how else it could reach the end of the function, but not the next statement in the caller
11:56:47 Quit Ryc ("CGI:IRC")
11:57:51JaykayJdGordon: now around?
11:59:15JdGordonyes, but if you ask the same question im going to ignore you again
12:00:40amiconnUnhelpful: I'll build for a PP target (ipod color) and a single core target (H300) and test
12:01:08Unhelpfulis beast dual?
12:01:17amiconnNo, single
12:01:24amiconnBut the beast uses no iram
12:01:55Unhelpfulhrm, right, so assuming the problem is strictly iram init, the beast would work
12:02:05*amiconn should also try the greylib, i.e. build for H180 and ipod mini g2
12:02:48JdGordondoes anyone know why audio_stop() 0's the current track id3 struct? Why can't it just be left so if playback is stopped the current track is actually the track which was playing at the stop time?
12:04:31kadobansorry if this is a stupid question, but are floats not used in rockbox because they are slow, or because they won't work at all? in other words: can i do basic float assignment and addition in a plugin?
12:04:54Jaykayjdgordon: ?
12:04:56Unhelpfulkadoban: there is no hardware floating point on most (all?) of our targets
12:05:30linuxstbkadoban: They will work, but are slow. Some plugins use them though I think, but you should always try and avoid them.
12:05:43linuxstb(especially if you want your plugin committed to SVN)
12:05:51moosUnhelpful: the beast is the exeption IIRC
12:05:56Unhelpfulyou should be using fixed-point binary or fixed-divisor rational if you're going to be doing any substantial non-integer math
12:06:19kadobanokay, thanks. it's for something small, basically just storing a setting so it shouldn't matter that it's slow
12:07:02Unhelpfulmoos: is it? i know there are some vector and floating extensions supported by various arm implementations, and i *think* the beast does iwmmxt...
12:07:30Unhelpfulno vectors of 32bit values, though, if i recall correctly, which is the only type that would help in any code that i've touched :/
12:07:49linuxstbmoos: That's not a target yet ;)
12:08:01mooshaha :D
12:09:15Unhelpfulwhat?! did my primary DAP get un-supported? ;)
12:09:21mooslinuxstb: what is the conclusion about sendfirme then? In case I miss some episods
12:09:38moosUnhelpful: hehe :)
12:09:48 Join MysticKitsune1 [0] (n=MysticKi@
12:10:23linuxstbmoos: I don't think there's been any - mcuelenaere was around yesterday, and I think he's trying to find alternative code to use.
12:11:03mooslinuxstb: ok, let's hope so then.
12:11:29MysticKitsune1add preset doesnt add the entry too preset, it replaces all other presets with that preset
12:12:18jhMikeSUnhelpful: If I allocate rb using the .lds mpegplayer works.
12:12:19linuxstbmoos: I'm sure we can solve it for a standalone sendfirm, but I would be surprised if we find a way to incorporate windows MTP support into rbutil (without it needing to call an external .exe)
12:12:56jhMikeSThat also ensures proper order (header _then_ pointer)
12:13:34UnhelpfuljhMikeS: huh. that's... very very weird. can we use the .lds to get the visible-in-plugins pointer aliased to a header field, as well?
12:14:36mooslinuxstb: we can live with this maybe (I mean without windows MTP rbutil support)
12:14:47 Join tyfoo [0] (
12:15:39 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
12:16:32*linuxstb sees that Songbird uses the GPL with an exception for some explicitly mentioned binary add-ons (gstreamer plugins and NPAPI browser plugins), but not an exception for the closed-source MTP add-on
12:17:16 Join MethoS [0] (
12:17:45 Quit JdGordon (Read error: 104 (Connection reset by peer))
12:17:45 Join JdGordon_ [0] (
12:17:48jhMikeSnvm the last bit, I just got back here and checked it out. I'm about half asleep. header has its own seciont. :p
12:18:40jhMikeSUnhelpful: You said that simply using plugin_iram_init is enough to crash it?
12:19:40jhMikeSmpegplayer still works in any case
12:20:54UnhelpfuljhMikeS: *mpegplayer* never makes it out of plugin_iram_init on my e200, unless i move synth or video_stack out of iram. doom and pacbox work fine.
12:22:11Unhelpfulbut i would very much like a pointer in the header for the api, instead of the pointer-to-a-pointer nonsense i'm using now... i just have too little linker-fu to even know where to start with arranging that.
12:22:44amiconnLinker script could work, but only for target, not for sim
12:23:23Unhelpfulis there any way to arrange it strictly within the C source?
12:23:54Unhelpfuli didn't want things to be doing __header.api->, or to #define rb __header.api somewhere
12:25:03amiconnI thought this would be possible with aliases, but it looks like I was wrong
12:25:08 Join casainho [0] (n=chatzill@
12:25:39*jhMikeS wonders about the .map file and what it indicates
12:25:44casainhohello :-)
12:26:38casainhoI am writing the message printf("Rockbox Player Bootloader"); but it is longer than my LCD width... is there a way to it draws in 2 or more lines?
12:26:52amiconnThe pointer-to-pointer stuff needs a bit more ram, but at least it works in plain C
12:27:03casainhoor I must work with smaller strings?
12:27:17kadobancasainho: \n?
12:27:51amiconnNo \n in the mini-printf available in rockbox bootloaders
12:27:51Unhelpfulit's four bytes, i'm more concerned about the inelegance of it than i am about four bytes... :/
12:28:02 Nick JdGordon_ is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
12:28:09kadobanah, sorry
12:28:43casainhoamiconn: so, I need to work with smaller strings?
12:28:44jhMikeSIn pondering it awhile back, I didn't really see anything better than pointer to pointer
12:29:20 Quit gevaerts (Nick collision from services.)
12:29:33 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
12:29:34amiconnWith the linker script, you have two options. Either delcare 'rb' extern and let the linker script fill in the value as an offset from 'hdr', or separate 'rb' from 'hdr' and let the linker script put them in sequence by using appropriate .section attributes
12:30:42amiconnBoth methods would break for the sim though, as we don't use a linker script there, and doing so would be quite difficult (if possible at all)
12:30:43Unhelpful... make the pointer the first element in the header, name the header rb, and make the header a union of struct plugin_header and struct plugin_api* ;)
12:31:22amiconnThat would work, but it would be hackish. No more magic there...
12:31:33amiconnActually, you could reuse the magic...
12:31:46Unhelpful...yes, yes you could.
12:31:59Unhelpfulthat's both a nasty and a beautiful hack, i think :)
12:32:08jhMikeSrb ends up on COMMON as the first variable
12:32:49Unhelpfulyou could always initialize it differently in the sim, and just fetch the address for rb and init it - sim plugins load via dlopen, correct?
12:32:56amiconnjhMikeS: Yes, right now, as there's no .section attribute
12:33:42Unhelpfulwe find the header differently in the sim loader, anyway
12:33:43amiconnUnhelpful: dlopen on *nix, LoadLibraryW() on windows
12:34:24amiconnsee uisimulator/common/io.c
12:35:15Unhelpfulok... but we can still use a fixed offset on target, and treat it as a separate exported symbol in the sim, right?
12:35:35jhMikeSit used to be in .bss
12:36:28amiconncommon is essentially also bss
12:36:54*Unhelpful hasn't tested this patch on beast...
12:37:32Unhelpfulmaybe i should, if it works on beast we know the mpegplayer issue is something that differs between 1) sim and PP 2) imx31 and PP
12:37:39jhMikeSI just wonder if that causes a clobber somehow
12:38:11MysticKitsune1add preset doesnt add the entry too preset, it replaces all other presets with that preset
12:39:51jhMikeShmmm...iram sections get loaded to the same address as .bss and then get overwritten.
12:40:36amiconnArgh, of course...
12:40:48amiconnSo 'rb' essentially must not go into bss
12:40:51*Unhelpful suspects amiconn has just cracked it.
12:40:55JdGordonpixelma: got time to test a patch to fix the wps statusbar issue where its shown on one screen but on on both?
12:41:14jhMikeSIt was in .bss before but initialized after the IRAM init.
12:41:56amiconnNow the loader inits it, therefore altering iram stuff (code, whatever) before it is copied
12:42:10jhMikeSWas I right? I don't really recall.
12:42:30 Quit tessarakt (Read error: 110 (Connection timed out))
12:42:51 Quit GodEater (Remote closed the connection)
12:42:59 Join tessarakt [0] (
12:43:07Unhelpfulamiconn: would moving it to its own section behind the plugin header fix that?
12:43:43Unhelpful...time to pack up here. i'll check back this afternoon or evening.
12:43:49amiconnFor a test, you can force it into the .data section
12:43:53jhMikeSyeah, it's only used api until after iram init
12:44:11amiconnJust init it (you need to use anything else than NULL)
12:44:33pixelmaJdGordon: not today
12:44:45Unhelpfulwill test when i get a chance. :)
12:45:09amiconnE.g. const struct plugin_api *rb = (struct plugin_api *)4;
12:45:30 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
12:46:41amiconnUnhelpful: This problem potentially hits any plugin that uses iram. It's pure luck if it works. Plain plugins or targets without iram (usage) are fine
12:46:54JdGordonB4gder: you actually wrote up a commit blame script?
12:47:37B4gderyes, but it's not actually in use yet, I need to decide on what it should do once it finds someone
12:48:26*amiconn hides so the script won't find him ;)
12:48:51B4gderthe changelog html pages now include the svn user name in a html comment ;-)
12:49:08casainhoB4gder: the LCD on Rockbox Player is working :-)
12:49:10jhMikeSwriting any data to .bss sections before iram init is rather risky when reusing it like that
12:49:22B4gdercasainho: great!
12:49:35amiconnjhMikeS: Yeah, hence that variable needs to go somewhere else
12:50:59jhMikeScould just define an .api section and put it after the header in the .lds
12:52:20casainhoB4gder: however there is some questions that I dont' know if the problem are from my drivers or Rockbox. I need to first right strings in all LCD, like doing scroll down until next strings apear again on top... looks like it's a problem of some sort of scroll engine... ??
12:52:33***Saving seen data "./dancer.seen"
12:53:12 Join BXCracer [0] (
12:53:12amiconnB4gder: Imho it would also be useful to announce the commits themselves (though probably not in full like the bot in #quassel does)
12:53:44B4gdercasainho: I don't understand what the problem is
12:54:13B4gderamiconn: possibly, yes. I think we can start with getting a "red alert" system and see how that feels and then possibly add from there
12:54:44BXCracerHi everyone, i don't see that clip bug that we were unable to access more than 2gb's anymore. does that means that now we can access all of it ?
12:56:06casainhoB4gder: the first messages appears on LCD but quiclly disappears!! they just appears steadily after LCD is filled with messages...
12:56:21B4gdercasainho: that sounds like an error in your driver or something
12:56:31B4gderrockbox fills data in a framebuffer, how would that vanish?
12:56:34 Join uu [0] (n=cut-jasm@
12:57:23scorcheuu: stop spamming please..
12:57:24B4gderuu: stop that crap please
12:57:39uuhi how r u
12:57:51casainhoB4gder: what file should I include for use the printf()?
12:57:52B4gderuu: see topic
12:58:11B4gdercasainho: we have no actual printf, only a few bootloaders using a kind of printf() thing
12:58:14uutell tell
12:58:15*MysticKitsune1 thinks uu is a bot
12:58:16scorcheuu: before you say anything else, i suggest that you read the channel guidelines as linked in the topic
12:58:56 Part uu
12:59:02scorchethat works too
12:59:15MysticKitsune1smart bot, lol
13:00:12 Join blkhawk- [0] (
13:02:35*jhMikeS 's contribution included (conflicting files in FS patch probably aren't all correct):">
13:03:02 Quit casainho (
13:03:02 Quit pyro_maniac (
13:03:02 Quit {phoenix} (
13:03:02 Quit pixelma (
13:03:02 Quit blkhawk (
13:03:02 Quit linuxstb (
13:03:02 Quit idshark (
13:03:02 Quit merbzt (
13:03:02 Quit HBK (
13:03:02 Quit feisar (
13:03:30NJoin{phoenix} [0] (
13:03:30NJoincasainho [0] (n=chatzill@
13:03:30NJoinpyro_maniac [0] (n=jens@
13:03:30NJoinpixelma [0] (n=pixelma@rockbox/staff/pixelma)
13:03:30NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
13:03:30NJoinidshark [0] (
13:03:30NJoinmerbzt [0] (n=benlar@
13:03:30NJoinHBK [0] (
13:03:30NJoinfeisar [0] (
13:04:06amiconnjhMikeS: No need to invent an extra section. Just define rb so that it goes into .data
13:05:10 Join Thundercloud [0] (
13:05:37 Join CaptainKwel [0] (
13:05:54*jhMikeS just felt like keeping them together at the start (if it mattered)
13:06:13casainholinuxstb: thanks :-)
13:07:19 Quit CaptainKewl (Read error: 104 (Connection reset by peer))
13:08:29 Part MysticKitsune1
13:08:41 Quit Horscht ("I am root. If you see me laughing, you better have a backup")
13:10:18 Join merbanan [0] (n=banan@
13:13:45 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
13:24:35 Quit AndyI (Read error: 60 (Operation timed out))
13:27:54ZambeziWhich Rockbox-supported player works with CompactFlash? My iPod Mini 2gen works fine, but there's glich in the battery so I'm looking for a player which you don't need to tear open. I know there a Sansa Exx, but just v1 which only found as refubrished.
13:28:12Unhelpfulamiconn: but if we add a section, we can specify its offset, and if we can do that, we can kill the **api member in the header
13:28:16 Join AndyI [0] (i=AndyI@
13:28:24B4gderZambezi: they're not compact flash though
13:28:38B4gderjust nand flash
13:28:43amiconnUnhelpful: Only on target. For the sim we'd need a different method to look up the symbol
13:29:11ZambeziB4gder: It's SD right?
13:29:18Unhelpfuli notice that we map dlsym to a GetProcAddress on win32... please tell me that doesn't mean it can *only* find functions?
13:29:28B4gderZambezi: uSD, yes
13:29:50amiconnRight now sim_plugin_load() returns the address of 'hdr'. With more than one symbol, this simple method won't work anymore
13:29:50Unhelpfulamiconn: right, so we export it for sim and find it the same way we do __header
13:29:50B4gderZambezi: and the AMS sansas are hopefully supported in the future
13:30:31ZambeziB4gder: Modelnumber on that?
13:31:04amiconnUnhelpful: GetProcAddress obviously works for 'hdr'. The problem isn't looking up symbols, but how to return more than one symbol address
13:31:29*Unhelpful slaps his forhead
13:31:38ZambeziB4gder: I got a rip off. A 32 GB CF-card for less than ten dollars so I would really like to use the card in another player when mine says goodnight.
13:32:14B4gderZambezi: then your only choice is a hdd based player and doing a hdd to CF replacement
13:32:47Unhelpfulyes, obviously it's not limited in that fashion. i see the real problem now... it's fixable in a few ways, we really just have to decide how :)
13:34:13ZambeziB4gder: That's the catch. I don't want to disammble it since the battery broke. I was a bit reckless when I had do reopen it for ten+ time due to the batterycable was a bit lose.
13:34:40Unhelpfuli think i have it :)
13:35:10B4gderZambezi: catch 22
13:35:20B4gderthere's no such thing that you look for
13:35:24Unhelpfuldid you say it *would* be possible to have the linker alias rb to a header field?
13:36:28 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
13:36:44Unhelpful(obviously only on sim)
13:37:27ZambeziB4gder: I think I can replace the cable on the battery with some assistance. I might do that. I know a friend which can solder. He's about the extend another cable already so I don't think he mind to get two more.
13:37:31amiconnI've added an updated patch that puts 'rb' into data (just a one-line fix)
13:38:15casainhohere is a picture of the first messages writeen on LCD from Rockbox Player:
13:38:20Unhelpfulbecause if we can, i think it can be kept very simple. we make header.api a void *, or a union of plugin_api * and **.
13:38:28 Quit B4gder ("It is time to say moo")
13:38:33 Join Schmogel [0] (
13:38:49Unhelpfulon sim, it's initialized in PLUGIN_HEADER as a pointer to rb
13:39:27mcuelenaerecasainho: nice! you're getting progress :)
13:39:29ZambeziIs these player AMS b4dger spoke about?
13:40:05Unhelpfulon target, it's left uninitialized in PLUGIN_HEADER, it's aliased to rb by the linker, and the loader just sets the field, via the plugin_api* member of the union, to the address of the actual api struct
13:40:39 Join funman [0] (
13:41:08Unhelpfulit's a few lines in source, and somebody who groks ld scripts to arrange the appropriate aliasing
13:41:11 Join miepchen^schlaf [0] (
13:48:24amiconnUnhelpful: Aliasing in the linker script would obviously only work for targets, not for sims
13:48:51Unhelpfulright... did you see how i suggest handling it for sims?
13:49:30Unhelpfulmake the field a union, and on sims, initialize it with a pointer to rb, basically exactly as happens on sim in the existing patch
13:50:13 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
13:50:35Unhelpfulon targets, rb is just an alias for the field, and we store the address directly to it
13:51:31amiconnThat could work, but it makes target and sim more different
13:52:22Unhelpfultrue, but not by much, and it saves 4B on target, not to mention feels to me like how the target version "should" work...
13:52:59casainhomcuelenaere: :-)
13:53:58UnhelpfulPLUGIN_INIT_IRAM should lose its argument, too, shouldn't it?
13:55:36amiconnYeah it could indeed, although it makes no difference in the final code
13:55:37 Quit mcuelenaere (Read error: 104 (Connection reset by peer))
13:56:51 Quit Jaykay (Read error: 110 (Connection timed out))
13:58:06 Join Jaykay [0] (
14:05:17 Join fdinel [0] (
14:05:47domonokycasainho: congrats for a working lcd, now use show_logo() and display the rockbox logo on this :-)
14:06:13Unhelpfuli'd like to at least take a stab at getting rid of the **api on target... i don't have time now, but i should probably learn to deal with linker scripts at some point, anyway
14:07:11casainhodomonoky: thanks!! I didn't know about that show_logo() :-)
14:07:40amiconnjhMikeS: Weird. I can't make the disconnect-reconnect cycling bug show up anymore
14:08:14domonokycasainho: take a look at the different bootloaders in bootloader/, they all use this show_logo() function :-)
14:08:17amiconnPerhaps I need to discharge further...
14:09:30*amiconn starts playing some ape tracks
14:12:02JdGordonamiconn: is the statusbar working on the player (specifically in the wps?)
14:13:56 Join PaulJam [0] (
14:14:36*JdGordon assumes it is.. or he would have been buzzed by now
14:17:32 Quit merbanan ("Leaving")
14:20:08amiconnJdGordon: I don't know. With that many targets I have to decide which one to use, and I didn't try the player recently
14:20:59 Quit rvvs89 (Remote closed the connection)
14:24:42domonokyJdGordon: yes, the wps on the archos player has a statusbar (but thats special chars in the charcell display)
14:25:04JdGordonok good
14:25:23Unhelpfulgot it working now... but surely there's a more elegant way than . = . - 4; rb = .; ?
14:25:47*amiconn usually decides for a main target for a few weeks. Currently that's the H180
14:26:19amiconnUnhelpful: . = . - 4 moves the actual current address. Just rb = . - 4; should work
14:27:03*amiconn only rarely decides for a colour target as the main target. They're just too annoying
14:28:36 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
14:29:56Unhelpfulright, i'll put up a patch with that shortly, then it's bedtime.
14:31:48 Join kerwood [0] (
14:32:37Unhelpfuland the weirdness in mpegplayer when i tried to fix it was just a result of ld changing its mind about placement for symbols when i changed their sizes and/or which one went in iram?
14:33:10 Quit Xerion (" ")
14:33:20amiconnYou moved stuff in iram, so the rb initialisation overwrote something different
14:33:40amiconnFwiw, with my fix mpegplayer is working fine both on H300 and ipod color
14:33:56casainhodomonoky: I am getting this error: rockbox/bootloader/show_logo.c:27:33: error: bitmaps/rockboxlogo.h: No such file or directory
14:34:19Unhelpfulyes, i tried basically the same fix, as i'd read that suggestion before i left to head home, and found it worked on e200
14:34:46domonokycasainho: you need to add the bmp file to the SOURCES for your target. (in apps/bitmaps/native)
14:36:14casainhoSOURCES of ifrmware?
14:36:44domonokycasainho: take a look at: apps/bitmaps/native/SOURCES
14:38:38casainhodomonoky: okok :-)
14:39:10JdGordonyou probably dont want the logo to display now anyway...
14:39:31JdGordoncomment out that show_logo() line and add a verbose = true line so printf works
14:39:45domonokyJdGordon: why not ? it would prove that the lcd is really working, and it looks nice :-)
14:40:04JdGordonyeah, but he needs to be able to debug stuff...
14:42:04casainhodomonoky: however I still get this error: error: bitmaps/rockboxlogo.h: No such file or directory
14:42:33 Quit funman ("leaving")
14:42:58domonokycasainho: if you corrected the SOURCES (and maybe make clean) it should generate this header for you.
14:43:18casainhoah, I didn't make lcean :-)
14:52:37***Saving seen data "./dancer.seen"
14:54:19casainhodomonoky: I still get that error, my LCD_DEPTH 16 and LCD_WIDTH 132, and thtere is already and for this values: #elif (LCD_WIDTH == 132) && (LCD_DEPTH == 16)
14:55:34 Quit fdinel (Read error: 104 (Connection reset by peer))
14:55:56casainhothe SOURCES looks correctly
14:56:41domonokycasainho: to make sure it really uses this line, insert a "#error i got here" after your #elif, then you know for sure that its correct.
14:57:02casainhodomonoky: ok
14:59:15 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
15:00:25 Join fdinel [0] (
15:01:13casainhodomonoky: I got this error:
15:01:37casainholooks like I am not getting that error of that SOURCES
15:01:50domonokyso your defines are not what you think.
15:02:22 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
15:02:54domonokyrecheck what you have defined in your config-<traget>.h
15:03:52 Join Seed [0] (
15:08:17casainhodomonoky: well, Ican't make it owrking. If I send you a patch, could you please look at it?
15:08:56casainhoI am having problems with my LCD drivers, show_logo() could help me, as an image is different that the printf()
15:09:01domonokyyes, i could take a look, if you want.
15:11:16casainhodomonoky: here is the patch:
15:12:23 Quit PaulJam (Read error: 145 (Connection timed out))
15:12:37 Join PaulJam [0] (
15:17:55domonokycasainho: a few remarks. the configure entry is not correct. (do you really have 64MB of ram, and the bmp2rb and toolset commands are misssing)
15:18:57casainhoyes, I have 64MB of SDRAM
15:19:24 Nick jfc^2 is now known as _jfc (
15:19:41casainhoI don't know how to setup bmp2rb and tollset...
15:19:46 Join MethoS- [0] (
15:20:18domonokybut you need the bmp2rb rules, to convert the bmps. (just steal them from a target with similar display)
15:20:42 Nick _jfc is now known as jfc (
15:20:59casainhook, I will do that
15:21:09domonokyand why does the config-rockboxplayerlittle redefine ROCKBOXPLAYER_LITTLE ? (it should already be defined by the Makefile (from configure))
15:22:16casainhodomonoky: so, I will remove that line.
15:25:33domonokywith the correct bmp2rb rules, it builds the bootloader with logo without problems :-)
15:27:23casainhodomonoky: okok - it's build here also ;-) −− thakns :-)
15:29:58casainhodomonoky: ah, okok. Looking at logo image... looks like my LCD have half of resolution... some problem on drivers, I need to work on it :-)
15:30:35domonokyhave fun :-)
15:34:06 Join aurix_lexico [0] (
15:35:37linuxstbdomonoky, bluebrother: Have you done any work on plugins for rbutil? I'm wondering if Windows MTP could be implemented as a plugin, and we add a GPL exception to allow that specific plugin.
15:35:41 Quit MethoS (Read error: 113 (No route to host))
15:36:37domonokylinuxstb: i had some play with Qt plugins in rbutil (for python), and it isnt really difficult. So that could be a solution.
15:37:04bluebrotherhaven't done anything on Qt plugins −− todo list is simply too long :(
15:37:12 Quit PaulJam (Read error: 104 (Connection reset by peer))
15:40:12bluebrotherIf I add a function to struct plugin_api I need to increase the PLUGIN_API_VERSION, right?
15:42:09 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
15:44:40*linuxstb realises Qt itself is GPL'd, so that approach won't work...
15:46:16*domonoky think it really maybe the best to just call a sendfirm.exe out of rbutil.
15:47:12linuxstbdomonoky: Yes, I think so too.
15:47:49domonokywe can even embedd the .exe in rbutil, so the user wont notice :-)
15:48:00linuxstbWhat about Linux/OS X - what about incorporating libMTP/libUSB there?
15:48:09bluebrotherwe can? Wouldn't that create new license issues?
15:48:22bluebrotherlibusb is already linked.
15:48:42marazis anyone using the capability of rockbox?
15:48:43Lloreanbluebrother: Surely just because resources share the same file, they don't need to share the same license.
15:48:51LloreanOtherwise GPL would be infectious across .zip files.
15:49:01linuxstbbluebrother: It is? So nothing is stopping us from supporting the e200r install in Linux/OS X (apart from Someone needing to do it)?
15:49:10marazand if you are, what are you using to sync it to
15:49:15bluebrotherlinuxstb: yes, that is my understanding.
15:49:30bluebrotherwe use libusb for the USB ID detection and statically link it for that.
15:49:51linuxstbAnd that works fine on OS X and Linux? And I assume you use the windows API there?
15:49:51bluebrothermaraz: I believe there is a wiki page on this. qtscrobbler can sync to
15:50:33bluebrotherlibusb works fine at least on linux. No idea about OS X as I cannot test anything. On windows the IDs are resolved using setupapi, which is w32api.
15:50:36linuxstbdomonoky: Or it could grab it from the download server - which might be easier (it will be a tiny download)
15:50:38bluebrotheri.e. part of MinGW.
15:50:56domonokylinuxstb: yes downloading it is also possible.
15:51:04bluebrotherfetching sendfirm from the download server would also have the advantage that we can update it independently
15:51:13linuxstbbluebrother: But the OS X builds at least include the code and statically link libUSB?
15:51:32bluebrotheryes, afaik.
15:51:46domonokylinuxstb: on osx we dont link statically. We use so called "Frameworks"
15:52:08marazbluebrother: and it works with utf-8? i've been using logscrobbler but it doesn't seem to understand it
15:53:36bluebrothermaraz: I guess so −− Qt uses unicode internally. Check the LastFMLog wiki page −− I don't use it ...
15:54:30 Join GodEater [0] (n=ge@rockbox/staff/GodEater)
15:56:02linuxstbdomonoky: But the libUSB "Framework" is included?
15:57:05bluebrotheris there a way to distinguish between e200 and e200r btw?
15:57:39Bagderas is so common with sansas, the OF is pretty much the only real teller
15:57:44domonokyhm, doesnt copy a libusb-framework, maybe we really link this statically ? :-)
15:58:08linuxstbbluebrother: sansapatcher won't find the e200r, as the firmware partition isn't visible via UMS - it uses the partition layout/contents to identify it.
15:58:22linuxstbI don't know if the USB IDs are different.
16:00:08bluebrotherhmm. Might be a starting point (but could also cause issues with users lacking the required permissions)
16:00:28marazbluebrother: um, doesn't rockbox filter things going into scrobbler.log at all? i remember audioscrobbler waiting to 50% or 2:40 of track length (whichever comes first) before scrobbling the track
16:00:35marazbluebrother: rockbox seems to add everything on start
16:00:41*bluebrother considers a linux-deploy script
16:01:07linuxstbbluebrother: The bootloader install requires the e200r in manufacturing mode anyway...
16:01:50bluebrothermaraz: what makes you believe this? I'm pretty sure the Rockbox implementation works correctly
16:02:01 Part kerwood
16:02:43linuxstbbluebrother: So it will be more a case of 1) User clicks on "install bootloader" button; 2) rbutil gives user instructions for manufacturing mode; 3) User puts it in manufacturing mode and attaches USB; 4) patching application is uploaded with libUSB
16:03:29 Join PaulJam [0] (
16:03:29bluebrotherwell, we have quite a bunch of dialogs now, so that wouldn't be the issue. But how to handle windows?
16:04:10linuxstbOr download e200rpatcher.exe and run it.
16:04:25linuxstbBut that requires the libusb driver to be installed...
16:04:49bluebrotherhmm. If users need to use a separate application on windows anyway, would it make sense adding e200rpatcher for other OSes? The windows uses are still the majority
16:06:14bluebrotherwe could simply make the bootloader installation for e200r output the instructions and do nothing. Everything else can still get done using rbutil
16:06:27*domonoky thinks its possible to install/uninstall drivers from the command line, so maybe rbutil could do that for the user ?
16:06:47*linuxstb is confusing himself between sendfirm and e200rpatcher...
16:06:51bluebrothersounds nasty.
16:07:27bluebrotherbut feel free to figure out how that works. If it works it _could_ give a way around the sendfirm / libmtp issue on windows ...
16:07:27linuxstbI think for Windows, you simply need to tell the user to install that driver. Because it uses the manufacturing mode USB IDs, it shouldn't conflict with normal use, so no need to uninstall them.
16:07:44linuxstb(talking about e200rpatcher now)
16:07:46bluebrotherand that ID differs?
16:08:09bluebrotherok, that's good.
16:08:10linuxstbIt's 0x0781 0x0720 for manufacturing mode.
16:11:00bluebrotherthat's also listed as c200 recovery mode on the DeviceDetection page. Shouldn't do any harm though.
16:11:06 Join MethoS-- [0] (
16:14:18 Join MethoS_ [0] (
16:16:20 Quit at0m ("CGI:IRC (EOF)")
16:20:40 Join MethoS [0] (
16:20:40***Alert Mode level 1
16:20:40DBUGSent KICK MethoS- to server
16:20:40DBUGSent KICK MethoS-- to server
16:20:40***Alert Mode level 2
16:20:40DBUGsent MODE #rockbox +b *!*n=clemens@*
16:20:40DBUGSent KICK MethoS_ to server
16:20:40DBUGEnqueued KICK MethoS
16:20:40***Alert Mode level 3
16:20:41Kick(#rockbox MethoS- :*bang* too many joined users) by logbot!n=bjst@gateway/web/cgi-irc/
16:20:41***Alert Mode level 4
16:20:41Kick(#rockbox MethoS-- :*bang* too many joined users) by logbot!n=bjst@gateway/web/cgi-irc/
16:20:41***Alert Mode level 5
16:20:41Mode"#rockbox +b *!*n=clemens@* " by logbot (n=bjst@gateway/web/cgi-irc/
16:20:41Kick(#rockbox MethoS_ :*bang* too many joined users) by logbot!n=bjst@gateway/web/cgi-irc/
16:20:41***Alert Mode level 6
16:20:42DBUGQ-Sent KICK MethoS to server
16:20:42Kick(#rockbox MethoS :*bang* too many joined users) by logbot!n=bjst@gateway/web/cgi-irc/
16:20:42***Alert Mode level 7
16:22:26marazbluebrother: i just skipped a few songs and they all are on .scrobbler.log
16:22:42marazthen again
16:22:50maraz they have a Played value of S
16:22:59marazand only the one i tested to play fully is L
16:23:04marazguess that's the difference then
16:25:34BigBambiIf the limit isn't met, it gets marked as S (skipped)
16:26:03BigBambiand then when you submit they don't appear
16:30:43***Alert Mode OFF
16:30:48 Join Asterisk_ [0] (n=George@
16:31:20Asterisk_I've got a sans e280, would that fall under the banner of "sans e200?" I'm compiling rockbox now
16:31:50BigBambie200 is the series
16:31:55Asterisk_just checkin
16:32:36 Quit {phoenix} (Remote closed the connection)
16:35:19Asterisk_bash: 59 command not found
16:36:02 Join Gnason [0] (n=4d8813dd@gateway/web/cgi-irc/
16:36:28Gnasoncan anyone help me compiling rockbox?
16:36:43Asterisk_tryng to do that myself
16:36:47Gnasoni get the following error
16:36:49domonokyGnason: just shoot your question.
16:37:06BigBambiGnason: If it is going to be a big paste, please use
16:37:06Gnasonarm-elf/bin/ld: /Users/robert/rockbox/build/apps/codecs/mp3_enc.elf section .ibss will not fit in region PLUGIN_IRAM
16:37:16Gnasonarm-elf/bin/ld: region PLUGIN_IRAM overflowed by 1244 bytes
16:38:15domonokyGnason: and which target is that ?
16:38:24GnasoniPod Photo 60GB
16:39:05Gnasoni'm using arm-elf-gcc 4.3.2
16:39:53Asterisk_is there a list ont eh site of the platform numbers?
16:39:56Asterisk_on the
16:40:02domonokyis it a clean svn or did you apply a patch ?
16:40:13Gnasonclean svn
16:40:36domonokyAsterisk_: why do you need that ? configure lists all.
16:41:07Asterisk_cuz it's telling me command not found when I type in what I think is the right number
16:41:43domonokyGnason: then i thinks its you compiler which generates to big code. i think the recommended version (and installed by is 4.0.3
16:42:25domonokyAsterisk_: if it tells you 50 is not a command, you pressed enter one time too much :-)
16:43:32 Part linuxstb ("Leaving")
16:44:47 Join gregzx [0] (
16:45:19Gnasonhmm, where do i find ^^
16:45:33 Join _lifeless [0] (n=lifeless@
16:45:43 Quit nibbler_ (Read error: 60 (Operation timed out))
16:45:50domonokyGnason: in rockbox sources under toos/
16:46:05domonokys/toos/tools/ :-)
16:47:21Gnasonk, thx
16:48:37Gnasonperhaps this should be mentioned on
16:49:04 Join {phoenix} [0] (
16:49:13 Join kugel [0] (n=chatzill@unaffiliated/kugel)
16:50:07CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
16:50:07*domonoky points to
16:51:51 Join merbanan [0] (n=banan@
16:52:16 Join Xerion [0] (
16:52:39***Saving seen data "./dancer.seen"
16:59:13 Quit Asterisk_ (Read error: 104 (Connection reset by peer))
16:59:24 Join Asterisk_ [0] (n=George@
16:59:46 Join petur [50] (n=petur@rockbox/developer/petur)
17:00:43 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
17:00:54 Join Horschti [0] (
17:01:10 Quit Horscht (Nick collision from services.)
17:02:08 Quit Asterisk_ (Read error: 104 (Connection reset by peer))
17:03:33 Join mirak [0] (
17:05:32casainhoI want to make a rockboxlogo for my 130x130 display... is there any source image? or should I resize from a bigger one?
17:06:10kugelcasainho: I think there's a svn available
17:06:21kugelBagder: might know
17:07:36casainhokugel: I don't understand... where? on apps/bitmaps?
17:07:39domonokymanual/frontpage has a rockboxlogo.svg i think thats the source
17:08:25kugeloh, and I meant svg, not svn, sorry
17:08:29 Quit CaptainKwel (Read error: 110 (Connection timed out))
17:08:39casainhookok :-) thanks :-)"> is the original source - not sure where that can be found
17:15:04bluebrotherthere's also a rather big png in the manual sources
17:20:08 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
17:21:55 Join MarcGuay [0] (
17:22:22MarcGuayIs updating the bootloader on the Gigabeat S as simple as replacing the nk.bin file?
17:22:52MarcGuayAKA sendfirm not required once you have MSC access via RB USB?
17:23:20LloreanMarcGuay: Yes.
17:23:32MarcGuayLlorean: Thanks.
17:23:47linuxstbcasainho: I thought your LCD was 128x128 (according to the RockboxPlayerPrototype wiki page) ?
17:24:00linuxstb(the same size as the iriver H10 5GB)
17:25:34casainholinuxstb: oh, looks like that information is not correct
17:26:15casainhothe size is 132X132 however the first and last ones pixels of collums and rows can't being seen
17:26:22casainhoso, it's just 130x130
17:26:39kugelcasainho: I'd say if it's 130x130, (and if possible), just treat it as 128x128, that will save you lots of work
17:27:06casainhokugel: well, I have it working now, showing the rockboxlogo
17:27:19kugelthere's a bunch of other bitmaps you need to convert
17:27:24Lloreankugel: "Take the easy way out" isn't exactly a project philosophy here.
17:27:35MarcGuayGig-s bootloader is 12MB OF vs. for 72KB Rockbox?
17:28:00LloreanIt might make sense to use the images for 128x128 plugins in many cases, but be careful what you mean by "treat it like 128x128"
17:28:24LloreanMarcGuay: Not quite sure I understand the question. For dual boot you'll still need a large file.
17:28:29kugelLlorean: he's got a lot of other stuff to do. all the work with converting bitmaps will just cost time he can use for e.g. getting audio work
17:28:37casainhokugel: well, I didn't know about that others bitmaps... then maybe its smart to stick with 128x128
17:28:49kugelplus, as far as I know, this lcd is a prototype anyway, and not going to be used in the final version
17:29:02Lloreankugel: Most of the converting bitmaps is plugins and doesn't need to be done now anyway.
17:29:07MarcGuayLlorean: I just built a single-boot RB bootloader. The size difference is quite large.
17:29:16LloreanMarcGuay: Ah, as long as it's single boot, this is expected.
17:29:19BigBambiMarcGuay: the 12 MB isn't just the OF bootloader though
17:29:23BigBambiIt is the OF
17:29:39MarcGuayBigBambi: Ah. That makes sense. Thanks guys.
17:29:50Lloreankugel: My point is, be clear when you say stuff like "don't bother" that you don't mean "it doesn't need to be done" but rather "focus elsewhere for now"
17:30:01casainhowell, that's true, I don't really need that 2 pixels (130-128)
17:30:05Lloreankugel: If it does end up with a 130x130 screen, it _does_ need to be supported properly, not just treated as 128x128
17:30:11kugelLlorean: that's what I was trying to say
17:30:21Lloreankugel: Then *say* it.
17:31:02Lloreankugel: You said "treat it as 128x128" and "it will save you lots of work." It *won't* actually save lots of work if that screen is used, so say "it will save you lots of work if you really are planning on getting rid of the screen"
17:31:13kugellook, his lcd is temporary (he said he's not going to stick with it for the final product), so that's likely not needed anyway
17:31:43Lloreankugel: You're missing the point.
17:32:03LloreanSay *why* so it's clear why you said that. So it's clear the message *isn't* "cutting corners is okay if it's close enough"
17:32:16rashercasainho: if the final LCD is not going to be 132x132, it's probably easier to treat it as 128x128.
17:33:05bluebrotherdid he say what size the final LCD have before at all?
17:33:08casainhookok, so to do it, on my config, I should define as 128 and on my drivers ignore the other missing pixels?
17:33:30rasherSomething like that - but if the final LCD is going to be 132x132, you should probably not cut corners
17:33:31bluebrotherplus, if he figures how to properly do it now it will save work when adjusting for the final LCD ...
17:33:53casainhofinal size is not defined, for now we have this hardware, the future is open... :-)
17:34:21bluebrotherthen doing it properly sounds like the better solution to me ...
17:34:30casainhowell, I do prefer to learn...
17:34:49LloreanYou don't need to convert the bitmaps to fit the screen, but all code should treat it as its full size at least, then.
17:34:57LloreanThe bitmaps can come later.
17:36:39casainhomy idea is to use this temporary prototype for showing that is possible to make our own hardware and I wanted the prepare myself to support a future hardware
17:36:50 Part linuxstb ("Leaving")
17:37:04casainhoso, I will then stick with the 130x130
17:37:55casainhowith all information and tools I have, will be quick to make this port!! even If I have to generate bitmaps and so on...
17:37:58bluebrothercasainho: what display are you using?
17:39:10casainhoNokia LCD 132x132 : −− but we can get it for $10 from other site
17:40:16 Quit saratoga ("CGI:IRC (EOF)")
17:40:18bluebrotherhehe −− I was suspecting something like that. But why does that page tell it being 128x128?
17:40:51kugelgood question
17:42:18casainhowell, I think is a mistake −− if you see the source codes, like the first one listed on that page, they use as if it have 132x132...
17:43:09casainhothat LCD is referenced in almost every cheap open source project as being 132x132, tha'ts why we get $10 for it :-)
17:43:28casainhoand it is working here as 132x132 :-)
17:43:50LloreanSo what's wrong with the first and last row and column so that you must use it 130x130?
17:46:26casainhosimple they don't appear: "The Nokia 6100 display has 132 x 132 pixels; each one with 12-bit color (4 bits RED, 4 bits GREEN and 4 bits
17:46:27casainhoBLUE). Practically speaking, you cannot see the first and last row and columns."
17:47:05 Quit Seed ("cu, Andre")
17:47:08casainhoit is written on James P. Lynch tutorial for that LCD
17:47:28LloreanWell, on the Nokia 6100 the casing may block the visibility of them.
17:47:35LloreanAre they actually visible on the raw display?
17:49:43bluebrotheri.e. you can light up the pixel at (0,0) and nothing appears? Even not when looking close?
17:51:24casainhowell, I took a picture now, wait a bit please :-)
17:52:14Jaykayill try again with and, i suggest both to be committed....
17:53:30 Join CaptainKwel [0] (
17:55:11 Join sko [0] (
17:56:43 Quit sko (Client Quit)
17:57:18Jaykayaaah they are ignoring me again :(
17:58:12casainhobluebrother, Llorean : here the rockboxlogo on LCD:
17:58:41LloreanJaykay: don't expect a response to every thing you say.
17:58:47bluebrothercasainho: nice :)
17:59:12casainhothat logo have missing small pixiels at left side
17:59:13LloreanJaykay: Seriously, if you think a patch should be committed, bring it up on the mailing list and have patience. Have you tried the mandelbrot one? From the description, it sounds like it would actually be bad.
17:59:48Jaykayllorean: i dont expect it, i just stated that nobody answered.
17:59:54Jaykayand i tested it, it works fine
18:00:33LloreanJaykay: It sounds like it would drop key presses
18:00:48Jaykayand i already wrote a mail to the mailing list with a few patches i suggested for committing, the only answer i got was "write a comment to the patches"
18:00:50LloreanLike, if I press "zoom" 10 times faster than it can zoom 10 times, it'll clear the button queue between each zoom and lose some of them.
18:01:31LloreanThat sounds bad to me.
18:01:53Jaykayit just prevents your dap to be unusable for the time of the zooms....
18:01:56LloreanIt sounds like it basically is assuming you *don't* press the button the right number of times, and punishes you if you do...
18:02:18LloreanYou always have to wait for zooms. All it does now is prevent you from zooming in a fixed amount without watching it and counting how many button presses it *actually* registered.
18:02:26bluebrotherwouldn't it be better to have an abort button for that?
18:02:44bluebrotheri.e. stop calculating and wait for the next user interaction.
18:02:46Jaykayyes that would be best
18:03:05Jaykayor just a button for clearing the button queue
18:03:32Jaykayok then this patch can be closed :)
18:03:35 Join Davide-NYC [0] (
18:04:00LloreanJaykay: Well the "iter" being reset aspect of it may be a bug that could use fixing. It's possible the patch can be improved.
18:04:55Jaykayhm i dont know what iter does, but i noticed nothing except the cleared button queue....
18:05:34casainhobluebrother, Llorean : and about the mixing pixels, you can also see this image that shows a printf(), you can clearly see pixels missing on the top of the LCD
18:06:13Lloreancasainho: that looks like 2 rows of pixels are missing at the top
18:06:58bluebrotherthat could also be caused by an offset calculated wrongly ;-) I though it's missing 1 pixel on each border?
18:07:22Davide-NYCLast night I had a hell of a time attempting to install the single boot nk.bin bootloader onto an updated (all original) gigabeat S with v1.2 firmware. I want to file a bug report as I believe there is something incompatible with an updated gigabeat S and the single-boot nk.bin file in it's current state.
18:08:11Davide-NYCcasainho: that's amazing. keep up the great work@
18:08:40MarcGuayDavide-NYC: That would be more of an issue with sendfirm, no?
18:09:05Davide-NYCMarcGuay: I'm not sure. Is that the likely culprit?
18:09:54MarcGuayDavide-NYC: Just a guess. What was the problem?
18:10:04Davide-NYCI suspect most devs updated their bootloaders using the sendfirm utility *before/without* updating the entire player with the updater utility from toshiba.
18:11:00Davide-NYCI used the updater to get an "original" fresh install on my beast and then attempted to run through the install procedure for a single boot rockbox install.
18:11:28Davide-NYCAfter a (seemingly) sucessful install I'd end up in recover mode after any reboot.
18:12:00MarcGuayjhMikeS: The battery meter reading is a bit weird on my beast (USB-plug, climbs to 99%, disconnect, falls down to original % and then slowly climbs up again?)
18:13:08Davide-NYCAny attempt at starting up the ploayer from shutdown would result in a forced recovery mode.
18:13:48MarcGuayDavide-NYC: Hmm. I'm having the same problem. Fresh bootloader + build, charged, turned off, turned on -> error 3.
18:14:08Davide-NYCOK. That's what I wanted. Confirmation.
18:14:25Davide-NYCDo we need to post a big report on flyspray?
18:14:48MarcGuayDoes the dual-boot work?
18:15:14Davide-NYCI haven't had the time to try and probably won;t until next weekend.
18:15:16MarcGuayNot sure what to do with this error#3. Any advice?
18:15:39MarcGuayBigBambi: Any problems installing the single-bootloader?
18:15:50Davide-NYCI've restored to factory v1.2.
18:16:14Davide-NYCIf I tell you anymore I'd be off topic and Llorean would berate me. (just kidding)
18:17:00MarcGuayThis feels pretty Rockbox related to me.
18:19:22 Join toffe82 [0] (n=chatzill@
18:19:25*MarcGuay is glad the beast wasn't called "supported" earlier...
18:20:16 Join karashata [0] (n=karashat@
18:22:09Davide-NYCI'm writing a bug report on flyspray.
18:23:07casainhobluebrother, Llorean : −− shows pixels missing from prontf() letters
18:24:16MarcGuayI'm having a tough time recovering, here. Got error#1 (insert USB). Sendfirm is failing, though.
18:24:32bluebrothercasainho: why don't you just draw colored lines / boxes? That way you can quickly spot which parts aren't displayed.
18:25:38casainhowell, maybe is better to stay with 128x128... I can't do this for all kind of this Nokias clones LCDs
18:26:05casainhobecause other developers bought from other online shop, so, it can be different...
18:26:39*domonoky also thinks that using a (centered) 128x128 might be good for this project.
18:26:50casainhofor example, the same clone Nokia LCD appear with 2 different controller, or Epson or Phillips, which have different commands,etc...
18:27:34Bagdercasainho: then you're up to doing run-time probing or two ports...
18:28:32Davide-NYCHere is maybe the most poorly written bug report of all rockbox history:
18:28:47bluebrotherwell, that LCD was used in some series of Nokia phones. I bet they wouldn't have liked different behaviours between the series of the LCD
18:29:30 Quit Davide-NYC ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
18:31:11casainhoI must try to mix the commands for the LCD, as the commands have different values and controller just ignores wrong commands... at least this is done on initialization, that is the same for both controllers...
18:31:32bluebrotherso the controllers are different?
18:31:34casainhobut again, this is a temporary prototype
18:32:20 Quit petur (Remote closed the connection)
18:32:21casainhoyes, controllers are different and looks like they have different values for commands, and ignores no valid commands...
18:32:40casainhoso, we can make 2 in 1 :-)
18:32:45bluebrotherurgh :(
18:33:02casainhoat least that is working at lcd_init()
18:33:09domonokyas this is a selfmade target, runtime detection can be easy. just ground a specific gpiopin if you have one type, or set it to vdd if you have the other type.
18:33:48casainhodomonoky: good idea :-)
18:34:40Bagdermy advice would be to avoid two controllers though
18:34:54Bagderand who says there won't be a third soon... and a 4th...
18:35:39casainhoeheh - but even the sellers can't identify them....
18:35:51Bagderthen buy from those who can
18:35:53casainhoit's like a Chinese
18:36:28 Join ender` [0] (
18:36:32domonokyBagder: for this nokia lcd runtime detection is good. these two types are very common in home-brew projects.
18:37:29*domonoky thinks even the nokia phone supports both types :-)
18:37:31 Join Ryc [0] (n=7a6f8eb9@gateway/web/cgi-irc/
18:37:45Bagderwell, I'd rather call it necessary evil than good...
18:37:46casainhoon sparkfun seller page: "Uses the Epson S1D15G10 or Philips PCF8833 Controller"; "What we've found over in China is a 'copy' of the Nokia LCD found in models 6100, 6610, 7210, 7250, and 5100? We don't know what phones it fits, and we don't really care."
18:38:42MarcGuaytoffe82: Got that link for the beast OF handy?
18:43:55Bagdercasainho: so how are things progressing on the audio front of your device?
18:44:27casainhoI don't have yet buttons!!
18:44:36Bagderoh, haha
18:44:46casainhowell, after LCD I will put buttons working
18:45:08 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
18:45:12casainhoafter, mount the SD card FAT32 and load RB firmware...
18:45:20casainhoaudio can be the last, no?
18:45:42Bagdersure it could, I was just curious on what you've decided regarding the hw for that
18:46:06 Quit MarcGuay ("ChatZilla 0.9.84 [Firefox 3.0.4/2008102920]")
18:46:14Rycpixelma: You know how some folks wanted their kids to have memorable names? "Ryc" is their version of "Rick".
18:46:15casainhowe are making an expansion board for buttons + audio IC
18:46:42casainhoand I can alwyas start working on SD card drivers....
18:46:55 Quit Ryc ("CGI:IRC")
18:47:28domonokycasainho: yes, you will need buttons and storage, before you can play music :-)
18:47:36*LambdaCalculus37 summons a fellow manual writer
18:47:51bluebrotheronly storage. Who needs buttons to select music? ;-)
18:47:58gevaertsdomonoky: I don't see why? :)
18:48:04bluebrothercasainho: what audio codec are you using? Have you decided yet?
18:48:16casainhoI have one button on the dev. board...
18:48:31casainhoues, TLV....
18:48:33Bagderone press for down, two for up, three for left ...
18:48:49casainhoyes, that TLV very well know, with drivers on Rockbox :-)
18:48:52domonoky... 100times for shutdown :-)
18:48:58krazykit5 for select? ;)
18:49:29bluebrotherI think he means the codec used in the h100 series (and others)
18:49:59casainhoyes, that ic
18:50:11LambdaCalculus37pixelma: Here?
18:50:23casainhoagain, there is a lot of resources for it, drivers for the board and MCU I am using :-)
18:50:45Bagderthe X5 and family use the TLV320
18:51:00bluebrotherhmm. Wasn't someone searching for that chip a while ago in the forums and it hard to obtain?
18:51:25casainhosimple buttons are very easy to wire on dev. board but that TLV no... very thin pitch, we need a PCB for that...
18:51:41casainhobluebrother: hard to buy?
18:52:13bluebrotheras in hard to find a distributor. But if digikey has them it shouldn't be a problem. Also, my memory could be wrong on this ;-)
18:52:43***Saving seen data "./dancer.seen"
18:52:43Bagderwow, 12 config-*.h files have TLV320 as defined audio codec
18:53:20Bagder... but only 3 of them are running Rockbox fine
18:53:21casainhoyes, digikey, farnell, etc - but again, is temporary prototype, the company that will build the hardware will select a cheap one IC, etc
18:53:28casainhocan be anyohter
18:53:32 Join bertrik [0] (
18:53:54bluebrotherthat's bad −− it will require you to rewrite the driver ...
18:54:14casainhoyes, but is a strategy
18:54:36casainhoI don't have money to invest, to make and sell the hardware
18:54:36Bagdersurely the TLV is a cheap ic?
18:55:11casainhoI don't know if is cheap or not, it's more important for me that have drivers on Rockbox an so on...
18:55:31bluebrotherbut why spend time getting a chip up and running that will get replaced later anyway?
18:56:00 Quit Darksair ("Do you hear that? This is the sound of inevitability. This is the sound of your death, Mr. Anderson.")
18:56:03bluebrotherI somewhat fail to see the point in that approach
18:56:15casainhowell, I will not spend to much time, because drivers are already done and tested.
18:56:30casainhobut, yes, I will spend but I will also learn ;-)
18:56:31 Join nibbler_ [0] (
18:57:00casainhoI would say the same for the working on Sansa V2 for example....
18:57:33 Join gfather [0] (n=g@
18:57:37gfatherhello guys
18:57:53gfathercan i update my rockbox through the installer ?
18:58:05bluebrothercasainho: why? The hardware for that port is fixed, thus everything developed doesn't need to get rewritten.
18:58:08bluebrothergfather: yes
18:58:36gfatherany link for llatest releas ?
18:58:50bluebrothersame link as ever ...
18:58:59casainhobluebrother: but that hardware can have a short window of life... like the V1, for example. Who knows if after V2 RB is working, Sandisk makes the V3?
18:59:06gfatherwhich is www.?? :)
18:59:23casainhojust Sandisk controls the Sansa hardware, they decide
18:59:27bluebrotherjust follow the white rabbit ... eeeh the release link on the rockbox website
18:59:49Lloreancasainho: Yes, but the difference is: You intend for your hardware not to have _any_ life.
19:00:01bluebrothercasainho: that's true, but in your case you're porting to a hardware you're going to trash
19:00:05LloreanOr so you've said.
19:00:11gfatheri found the rabbit
19:00:14gfatherlets kill him :)
19:00:43casainhoyes, thats why I choosed cheap hardware and with tested drivers available...
19:00:43bluebrotherpoor rabbit
19:01:00 Quit kugel ("ChatZilla 0.9.84-rdmsoft [XULRunner]")
19:01:10bluebrotherisn't the final hardware supposed to be cheap as well? ;-)
19:01:42casainhobluebrother: yes, but I can't get cheap prices buying pieces at units
19:01:52domonokybluebrother: the problem is, that nobody knows what the final hardware will be.
19:01:54casainhonor I have an electronics company
19:02:45bluebrotherdomonoky: but why does nobody know?
19:02:50casainhofor example, Olimex have knowledge and stock materials, and relations on the market of semiconductors... then they can tailor an hardware also for price
19:02:51gfatheri think we should make an opensource mp3 player based on rockbox :)
19:03:07 Quit BXCracer (Connection timed out)
19:03:14gfatherforget about installing it on devices that dont have 100% compatability
19:03:31bluebrothermy Ipod is compatible :P
19:03:42gfatherme to , its nott that what i mean
19:04:02gfatherimagine if all the components in the device is compatable
19:04:14gfathersee how much video playback and stuff will be great
19:04:14casainhogfather: read here about our ongoing efforts:
19:04:40gfatheryes thats what i was thinking :D
19:05:04gfatherFind a company to assembly and sell It online, to worldwide (already received some offers); i can help in this matter
19:05:21casainhobluebrother: just one company with money to invest can decide what hardware will produce and sell!! I can't do it!
19:05:58casainhogfather: how?
19:06:09bluebrothercasainho: yep, but my point is: shouldn't that company do that decision first? Otherwise you'll have lots of work you can trash, and porting won't be faster at all
19:06:27bertrikcasainho, will it look as cool (e.g. casing of the player) and be as cheap as currently available players?
19:07:19gfathercasainho i can see some chines factories , and companies who offer oem and stuff
19:07:43gfathersee what can we do with them , becouse i dont think it will be built outside china :)
19:08:37casainhobluebrother: first I need to show that it is possible and that I can help port Rockbox to some custom hardware... then I think I can convince some companies
19:09:04casainhobertrik: that do not depends on me, I don't have a company, I am not a investor...
19:09:26bluebrotherwell, I disagree that someone need to prove it's possible porting Rockbox, given the number of targets it already has been ported to.
19:09:35casainhogfather: yes, China for sure, as every one else :-)
19:09:59gfathercasainho didnt u think to build it on freerunner for showcasing ?
19:10:21gfatherit would bee too much cool if you showcase it on freerunner or something similar
19:11:03domonokygfather: freerunner is a phone, not really a ideal target for rockbox i would think.
19:11:18casainhobluebrother: yes, that's true. I think the same, that's way I am not worried in using a better LCD, anyone can see an Ipod running Rockbox. But I coudn't say that I would commit my energies to port rockbox to some custom hardware If I didn't make any port before, you know, this is my first time with ARM and 32 bits...
19:11:55gfatherwell its only for showcasing , it has a high resolution touch screen , its not ideal as a mp3 player , but for showcasing it would look cool
19:12:57casainhogfather: I selected and MCU and dev. board with a lot of information, code, and tutorials... freerun is not that hardware
19:13:51gfatherwell im just suggesting :)
19:13:59gfatherif u need any help , just tell me
19:14:01casainhofor showing case, than any "good" player as Ipod
19:14:13gfatherthats true too
19:14:18casainhogfather: we need help for sure!!
19:15:13gfatherwell anything other than money , i can help :)
19:15:13 Quit Gnason ("CGI:IRC (EOF)")
19:15:26casainhowell, no money :-)
19:16:09gfatherwell if u need someone to think of cool desigens and stuff , i mean the gui for users
19:16:10casainhothe plan is to make the prototype and after talk to some company that wants to invest on it - we need that someones sells it online to worldwide
19:16:38gfatheri can sell online
19:16:46gfathercan do allot of stuff ;)
19:17:05casainhoahah - but do you have any shop? −−
19:17:09gfathermarket it in the middel east region with arabic language support
19:17:22casainhoanyway, it's an open hardware, everyone can sell... Ebay and so on
19:17:43gfathersaudi arabia and such countries can be big hit for you ;)
19:17:56casainhofor me no, for the project :-)
19:18:03gfatherfor the project ;)
19:18:13gfatherwhy theme site is down ;(
19:18:20casainhofirst we need to a company to invest on it
19:18:37gfatheris the prototype ready ?
19:19:03domonokygfather: themes site is down, because it gets rebuilt. But not enogh people work on it .. :-)
19:19:36gfatherwe can contact some venture captalist to invest in it ;)
19:19:54gfatheri can make 100% that i can get a company for the idea
19:20:18gfatheri mean a venture to be interested in the project
19:20:43gfatherooff the topic , where do i get the themes from ?
19:21:06domonokygfather: from the wiki WpsGallerys.
19:21:40gfatheryes , found it
19:22:08casainhono prototype is not finished and prototype is not the final hardware, but we can suggest the hardware
19:23:09gfatherthats cool , ill pm u my contact , so you can feed me with needed info
19:23:25 Join LambdaCalculus3_ [0] (
19:25:28 Join funman [0] (
19:25:36 Join kugel [0] (n=chatzill@unaffiliated/kugel)
19:25:55gfatherhow can i download all themes ?
19:26:04gfatheri dont think i want to download 1 by 1
19:26:37scorcheyou cant until we get the new theme site up and running
19:27:03gfather:S damn
19:27:41LloreanScrolling down the page and downloading just the ones that look nice is probably much, much faster than downloading them all, then trying them all on the player to see how they look.
19:28:03 Quit Bensawsome ("The awsome is gone :(")
19:30:28funmanI just found that the Sansa Clip OF uses CMD6 for SD cards, and CMD6 isn't listed in the v2.00 specification (could it be a cmd for bank switching?)
19:30:28 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
19:31:56 Quit LambdaCalculus37 (Read error: 110 (Connection timed out))
19:32:23funmanand also CMD23 (also undocumented)
19:32:43 Nick fxb is now known as fxb__ (
19:32:55bertriknice find, /me googles for those commands and checks the spec
19:34:10funmannice : e200v1 uses CMD6 to switch banks
19:34:32bluebrotherhow does Rockbox do that?
19:34:58funmansee firmare/target/arm/ata-sd-pp.c
19:35:12funmanthe SD_SWITCH_FUNC is defined in firmware/export/sd.h
19:37:10bertrikmy version of the sd spec does mention CMD6
19:38:02funmanoh right, it's to _enable_ bank switching
19:38:02 Quit CaptainKwel (Remote closed the connection)
19:38:31 Join likemindead [0] (
19:38:37funmane200v1 uses cmd35 (proprietary) to actually switch banks
19:39:12funmancmd23 is marked "reserved"
19:39:37bertrikparagraph of the physical layer simplified specification version 2.00 says "Switch function command (CMD6)1 is used to switch or expand memory card functions."
19:40:03 Part likemindead ("Making like a tree...")
19:41:10funmanoops I can't list all commands used because some are loaded from a table at an (still) unknown location
19:42:09funmankugel: did you already have a look at the code which switches SD banks in ata-sd-pp.c ?
19:45:03 Quit PaulJam (Read error: 113 (No route to host))
19:46:31 Join CaptainKewl [0] (
19:46:40 Join Zagor [242] (
19:47:15kugelfunman: a short one yes
19:48:53funmanI also see commands 20 & 11 being mentioned in an array : { write_multiple, write, read_multiple, read, 20, 20, 11, 11 }
19:49:41kugelfunman: i guess you're looking at clip v1's OF?
19:49:51funmankugel: could you determine exactly how many sectors you can access on your fuze' storage ?
19:50:16funmankugel: yes, version 29 (the version which introduced ogg playback, but not flac playback)
19:51:09kugelit has been determined already, I'll go search the logs
19:51:39 Quit kugel (Remote closed the connection)
19:52:06funmanAlso I'm not sure how to determine the real capacity, since the CSD always indicates a bit less than 4GB
19:55:32 Join kugel [0] (n=chatzill@unaffiliated/kugel)
19:56:49 Quit killan ("( :: NoNameScript 4.22 :: )")
19:58:24kugel0x1DADFF is the last sector you can read
19:59:27funmanso the bank size is 0x1E9E00 (adding the space reserved for the OF)
20:00:05 Join killan [0] (
20:00:09kugelonly the first bank needs to reserve space for the of
20:01:43 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
20:03:53funmanI can't find this number in the fuze or clip firmwares
20:07:28 Join PaulJam [0] (
20:08:47funmankugel: I'd try to mimic the code used in ata-sd-pp.c !
20:08:53funmangood luck, see you next time
20:08:55 Quit funman ("leaving")
20:14:45 Quit Acky (Read error: 54 (Connection reset by peer))
20:15:07 Join Acksaw [0] (
20:28:53 Quit LambdaCalculus3_ ("Ka-chunka")
20:29:20 Quit pyro_maniac ("Leaving.")
20:47:17 Quit casainho ("ChatZilla 0.9.84 [Firefox 3.0.5/2008121622]")
20:50:55 Quit aurix_lexico ("Leaving.")
20:52:34 Join aurix_lexico [0] (
20:52:45***Saving seen data "./dancer.seen"
20:53:37 Join toffe82_ [0] (
20:54:48 Quit kugel (
20:54:48 Quit toffe82 (
20:54:48 Quit {phoenix} (
20:54:48 Quit GodEater (
20:54:48 Quit domonoky (
20:54:48 Quit miepchen^schlaf (
20:54:48 Quit pixelma (
20:54:48 Quit merbzt (
20:54:48 Quit HBK (
20:54:48 Quit feisar (
20:54:48 Quit idshark (
20:54:48 Nick toffe82_ is now known as toffe82 (
20:54:53 Join tvelocity [0] (
20:55:34NJoin{phoenix} [0] (
20:55:38 Quit toffe82 (Killed by (Nick collision))
20:55:39NJoinkugel [0] (n=chatzill@unaffiliated/kugel)
20:55:39NJointoffe82 [0] (n=chatzill@
20:55:39NJoinGodEater [0] (n=ge@rockbox/staff/GodEater)
20:55:39NJoindomonoky [0] (n=Domonoky@rockbox/developer/domonoky)
20:55:39NJoinmiepchen^schlaf [0] (
20:55:39NJoinpixelma [0] (n=pixelma@rockbox/staff/pixelma)
20:55:39NJoinidshark [0] (
20:55:39NJoinmerbzt [0] (n=benlar@
20:55:39NJoinHBK [0] (
20:55:39NJoinfeisar [0] (
20:55:57 Join m0f0x [0] (
20:56:17 Join toffe82_ [0] (n=chatzill@
21:05:22 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
21:06:43 Join robin0800 [0] (
21:09:33 Join gregzx [0] (
21:10:29 Quit toffe82 (Connection timed out)
21:23:09 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
21:23:36 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
21:23:48 Join mcuelenaere [0] (i=mcuelena@rockbox/developer/mcuelenaere)
21:24:06 Join Schmogel [0] (
21:29:19 Join allele [0] (
21:29:52alleleso has anyone else considered a spore-like game for rockbox?
21:30:21scorcheconsidering might be nice and all...but someone has to do the actual writing...
21:31:08alleleagreed, just seeing if anyone else has already started such a project
21:32:41alleleand now if it hasn't its at leas on the log, eh Llorean?
21:33:22mcuelenaereallele: I don't think anyone considered anything like that, otherwise you would've at least seen some patch or a forum thread..
21:35:21alleleyeah, well I was thinking something a bit more primitive considering the platform
21:36:31LloreanWell, most of the actual gameplay of spore is simple enough that it could be represented in bitmapped graphics, etc.
21:37:00LloreanYou could reproduce the gameplay experience well enough, you'd just lose the procedural animation and such. Most of that's just gloss anyway
21:40:43 Quit allele ("Java user signed off")
21:41:16 Quit kugel ("ChatZilla 0.9.84-rdmsoft [XULRunner]")
21:49:39 Quit bmbl ("Woah!")
21:53:10 Quit karashata ("G'bye everyone!")
22:03:27*Zagor prepares to commit FS #9703, slightly modified (bigmem targets get to keep their big pcm buffer for now)
22:08:12 Join archivator [0] (i=foobar@
22:08:44 Quit merbanan (Remote closed the connection)
22:08:46 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
22:11:03 Quit archivator (Remote closed the connection)
22:14:17 Join archivator [0] (i=foobar@
22:16:20 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
22:24:38B4gderBjörn Stenberg/zagor=1
22:25:04Zagoryeah, missed some sim stuff
22:25:20B4gdergreat time for me to verify my script!
22:25:55Zagorof course. you don't think I did it by mistake, do you?
22:26:06B4gderthanks for volunteering! ;-)
22:26:48 Join gregorovius [0] (
22:27:43 Nick Horschti is now known as Horscht (n=Horscht@xbmc/user/horscht)
22:27:46*scorche wonders if Zagor has done anything on the google money front
22:28:53Zagorno, but I'm setting an alarm in the phone for monday now. this is getting silly :-(
22:29:36scorcheit would be nice to have that by next devcon ;)
22:30:54saratogadoes anyone have a good resource for the mpeg4 specs ?
22:31:01saratogagoogle turns up a lot of not very detailed stuff
22:31:43mcuelenaeresaratoga: ?
22:33:18 Quit gregorovius (Remote closed the connection)
22:33:57saratogamcuelenaere: Part 14 - MP4 File Format has no document unfortunately
22:36:32 Join Riku [0] (
22:36:33 Quit Lss (Read error: 104 (Connection reset by peer))
22:37:41 Quit ender` (Read error: 54 (Connection reset by peer))
22:41:51saratogado we support playing the audio track of MP4 movies in Rockbox?
22:45:10B4gderZagor: so as you may understand, we're close a "blame" feature
22:48:16*scorche wonders if it will blame perl
22:48:40B4gderits the other way, this is perl doing the blaming!
22:48:46saratogahow come the clip didn't see a drop in memory use from this change?
22:50:08 Join tusjen [0] (
22:51:01Zagorbecause the pcm buffer is taken from the file buffer, i.e. "the rest of the ram"
22:51:30Zagorthe change just moves ram from pcm to file buffer, hence no change for the program
22:51:36saratogaah ok
22:52:19 Join kugel [0] (n=chatzill@unaffiliated/kugel)
22:52:39ZagorB4gder: yeah I saw in the log. do you plan to do the logbot modification too?
22:52:49***Saving seen data "./dancer.seen"
22:53:33B4gderno, I don't have any such plans atm
22:55:09 Quit B4gder ("It is time to say moo")
22:57:56 Quit XavierGr (Nick collision from services.)
22:58:07 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
22:59:02Lloreansaratoga: No, but there's probably no reason we shouldn't
22:59:07Lloreanre: mp4 movies
23:00:23saratogaLlorean: i just tried it and it seems to work
23:00:30saratogaso i guess we do, at least for some movies
23:02:12LloreanProbably just chance.
23:02:14 Quit mirak (Remote closed the connection)
23:02:23LloreanI'd imagine most would be large enough files to have the mP4 parser issue
23:02:37 Quit m0f0x (Read error: 60 (Operation timed out))
23:03:19saratogayeah probably
23:03:46 Quit Jaykay (Read error: 110 (Connection timed out))
23:03:53amiconnJdGordon|zzz: One of your recent wps changes broke "follow playlist" again when the playlist ends.
23:03:58Lloreansaratoga: Do you know if that's a case of being "very difficult to fix" or a case of "nobody's gotten around to really even looking at it"?
23:04:05amiconnIt works when stopping manually
23:05:41saratogaLlorean: I'm looking at it now
23:06:02saratogathe problem is that the parser seems to allocate a huge table of packet lengths upfront, and then steps through them
23:06:14saratogai'm looking for a way to compute them on the fly instead
23:11:12 Join Jrf [0] (
23:11:39 Nick Jrf is now known as BigBambi_ (
23:12:54kugelI wish rockbox would boot on my clip
23:23:58Unhelpfulkugel: can you open bitmaps in rockpaint on your fuze, and then exit? i think i know a quick way to test different read sizes in the loader.
23:24:26Unhelpfulfantastic. :)
23:24:34 Quit archivator (Read error: 104 (Connection reset by peer))
23:24:40BagderZagor: is the automatic reconfigure really that needed?
23:24:57Bagderwhat if you have altered the main makefile, do svn up and then make ?
23:25:08Bagderzap, changes gone
23:25:36Bagder(and the help text wasn't removed either)
23:25:37Unhelpfulit'll overwrite changes in autoconf.h as well, right?
23:25:55Bagderhm wait
23:26:17Bagderdoesn't this change only make you have to type "make Makefile" instead of reconf?
23:26:31 Join archivator [0] (i=foobar@
23:26:35Zagorno, it runs it automatically
23:27:11Bagderwhat depends on the Makefile? the entire build?
23:27:51Zagormake always checks its' source file for dependencies and if so re-reads it. that's how the dependency file works too.
23:28:22kugellol, make now runs configure, asking for the values
23:28:39kugelI thought it would at least use the old values
23:28:44Bagderkugel: the first time it has to
23:29:14Zagorkugel: the new configure stores the values. the old didn't.
23:29:18Bagderas the old values are stored in the Makefile
23:29:21kugelah right
23:31:11ZagorI'm split about the automatic aspect.
23:31:32*amiconn thinks it's unnecessary
23:32:04amiconnA 'configure' change doesn't necessarily mean that I have to reconfigure, e.g. if just a new target was added
23:32:10Zagorhow about it just saying "You should run 'make reconf'" instead of doing it
23:32:30BagderI'd be fine with that
23:33:00Zagorif it does nothing people will overlook it and it will have lost part of the purpose
23:37:38kugelhaha, playback doesn't work at all anymore on my fuze
23:37:56kugelhissing noise at flac+ogg, crash/reboot on at aac/mp3
23:39:09Zagorkugel: since when?
23:40:33kugelsince your commit of the watermark patch, as far as the hissing noise is concerned
23:41:14 Join Xerion_ [0] (
23:41:46 Quit Thundercloud (Remote closed the connection)
23:42:40Zagorthat's interesting. the fuze has 8MB ram, so it didn't even get the changed pcm buffer.
23:43:41 Quit Xerion (Read error: 60 (Operation timed out))
23:43:41 Nick Xerion_ is now known as Xerion (
23:44:39Bagderoh foswiki released 1.0.0 ...
23:47:53bluebrotherZagor: does configure now save _all_ its values, i.e. also the advanced options?
23:48:10kugelUnhandled IRQ 00: INT_WATCHDOG
23:48:28kugelthat shouldnt happen I suppose, or did I miss a watchdog isr?
23:49:00Zagorkugel: I use the watchdog to reboot.
23:49:23Zagorother than that, I'm not aware of any use of it
23:49:32kugelyes, but that's not an interrupt handler. I don't remember a watchdog isr (i.e. no INT_WATCHDOG function)
23:50:03Zagorno. that's what I'm saying too :)
23:50:20kugelwhich makes me curious why that happens
23:50:29kugelmaybe the VIC is setup slightly wrong?
23:53:00Bagderbluebrother: but it should be fairly simple to add in the same fashion it currently works
23:54:15bluebrotherBagder: sure. It's just that someone(tm) needs to do it −− and I can't use reconf on my h100 rtc build as of now.
23:54:46bluebrotherlet's see if I get around looking at it myself or if someone else is faster ...
23:55:04BagderI'll be the fastest off to bed now!
23:57:15 Join solexx [0] (

Previous day | Next day