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 2007-02-19

00:00:16*Mouser_X doesn't have a wiki account.
00:00:22Mouser_XAlso, ADX is missing.
00:00:27 Quit redbreva ("Chatzilla 0.9.77 [Firefox]")
00:00:28Mouser_XAs I said, it needs updating.
00:00:32peturtime to get one
00:01:10Mouser_XI don't recall if ADX is compressed, uncompressed, lossy, or lossless...
00:01:11 Quit FOAD (Read error: 104 (Connection reset by peer))
00:01:23Mouser_XI'm pretty sure it's compressed, but not a whole lot.
00:01:44Lunar_LampYay, it seems that faad2 has been integrated since last I checked out rockbox. However, are all supported codecs in all firmwares for all devices?
00:01:45Mouser_XAlso, I think it might be *somewhat* lossy, but not a whole lot, if any at all.
00:02:33Mouser_XCri Middleware makes ADX.
00:02:41SoapIsn't ADX more like SID or SPC?
00:02:42 Join rasher [0] (n=rasher@rockbox/developer/rasher)
00:02:58Mouser_XNo, not really.
00:03:01Shaidthere's an idea.
00:03:01Mouser_XIt's closer to FLAC.
00:03:13Soapsee why I haven't updated the page!
00:03:32rasherHey, I can't build rbutil (with patch #6643). I get a bunch of "cannot convert ‘const char*’ to ‘const wxChar*’". What gives?
00:03:39Mouser_XSID and SPC are both sequenced formats. Except that SPC is a memory dump, and SID isn't.
00:03:51Mouser_XADX is a streamed format, like FLAC, or MP3, or OGG.
00:03:59linuxstbrasher: Is it the unicode problem again?
00:04:11Bagder"ADX is a lossy proprietary audio storage and compression format"
00:04:25Mouser_XAh, lossy then.
00:04:27Bagder(according to wikipedia)
00:04:38Mouser_XThere's an encoder for ADX available.
00:04:51rasherlinuxstb, not sure. I compiled wxwidgets myself this time (with standard options), but I had libwxgtk2.6-dev installed. Purged it and trying again
00:04:57Mouser_XI'd convert many of my OSTs to ADX, if I could find a surefire way to get the looping correct.
00:05:12rasherWhy oh why are there no Debian packages of 2.8
00:05:37Dunkelschubwhat do you guys think of Ubuntu?
00:06:04rasherDunkelschub, it doesn't run very well on portable audio players
00:06:35Mouser_XShaid: ADX is a video game format. It supports looping of music, since many games have BGM that lasts as long as you are in the stage.
00:06:44BagderShaid: apparently ADX has a loop feature
00:07:10Mouser_XMost streamed formats don't support looping.
00:07:12Lunar_LampDunkelschub, I love it really, simple to use generally, works nicely, though not as optimised as distro's like gentoo, the performance loss is not significant to me.
00:07:13Shaidoh, nice
00:07:19BagderMouser_X: so get working on a adx codec for rockbox!
00:07:26bospaadjeI found out recently that, on my ipod nano, after switching off the hold button, buttons don't react until scroll is used, which is very annoying.
00:07:27Mouser_XIt's already in there.
00:07:42Mouser_XBagder: HCS ported the codec
00:07:46bospaadjethere's a bug report in flyspray already (for a long time):
00:07:56Bagdershows how much I'm awake
00:07:56Dunkelschubyeah, I tried a bunch of different Linux flavors, found Ubuntu to be the best for ease of use yet wasn't dumbed down
00:08:02linuxstbWe have an ADX codec?
00:08:11*linuxstb is sleeping too much as well...
00:08:13bospaadjeany idea wether this would be easy to address?
00:08:25Mouser_XLike I said, the soundcodecs page needs updating.
00:08:29midgeyadx works decently well from what i've tried
00:08:36Mouser_XWorks great for me.
00:08:42Mouser_XTakes less battery than MP3.
00:09:27linuxstbbospaadje: I seem to remember that ipodloader2 had similar issues, but tempel fixed them. So a comparison of Rockbox's button driver with the driver in loader2 might help.
00:09:42Mouser_XNot sure by how much, but when playing ADX, my player has lasted longer than when playing MP3.
00:09:53rasherlinuxstb, okay, it was a combination of having wx2.6-dev and setting my PATH wrong. Now I get errors about fwpatcher. Should I copy fwpatcher into the rbutil dir?
00:10:08linuxstbNo, you shouldn't need to...
00:10:09rasherlinuxstb, that's the patch, of course, which I realise you're not the author of
00:10:24linuxstbAre you using the Makefile provided? I'm not sure that's updated.
00:10:30midgeyMouser_X: i get around 1% boost ratio on my h300
00:10:41linuxstbDomonoky is working with the code::blocks thing.
00:11:30powriter25Gah. My ipod refuses to load rockbox.ipod.
00:11:37linuxstbmidgey: I think it's just a simple adpcm type format - nothing complex to decode.
00:11:44rasherlinuxstb, YEah, I'm simply running make after applying the patch
00:11:51Mouser_Xmidgey: I wouldn't know how to check. HCS says that ADX playback takes less CPU (boosting? I can't recall now) than MP3 does. I take his word on it.
00:12:08linuxstbrasher: You'll need to add the new object files.
00:12:15Mouser_Xmidgey: Even so, I have gotten more playtime with ADX than MP3 (not sure by how much though).
00:12:54midgeyyah, my adx files are straight from one of my dreamcast games
00:13:09*Mouser_X has a ton of ADX files.
00:13:17Shaidhang on, is it THAT adx?
00:13:33Mouser_XWhat other ADX is there?
00:13:49Mouser_XMany (if not all) of SEGA's games use ADX,.
00:13:52ShaidI hadn't clicked yet.
00:13:58ShaidSo do quite a few PS2 games
00:14:05Shaida bunch of the capcom games
00:14:05Mouser_XI have F-Zero GX in ADX.
00:14:11Mouser_XSega helped develop that one.
00:15:12Mouser_XShaid: See my PMs?
00:15:15Mikachuis it similar to .gym files? (that music dumps from mega drive use)
00:15:21Shaidyes, and that second one is true
00:15:29Mouser_XMikachu: No.
00:15:29ShaidI tried to say 'sure'
00:15:32Shaidbut it wouldn't let me
00:15:36ShaidI think I'll go register.
00:15:43Mouser_XGYM files are logged. ADX isn't logged.
00:16:03preglowweird... i give the alarm menu a time, and it always sets a time one hour and fifteen minutes in the past
00:16:16Mouser_XMikachu: ADX is a streamed format, like MP3, OGG, FLAC, etc.
00:19:24rasherlinuxstb, can it be that the problem is, that the fwpatcher files require windows stuff, and I'm compiling for linux?
00:19:53rasherIn file included from ../tools/fwpatcher/iriver.c:23:
00:19:53rasher../tools/fwpatcher/iriver.h:20:19: error: tchar.h: No such file or directory
00:19:55rasherand so on
00:20:06 Join |Judo| [0] (
00:20:12rasherWhen trying to make iriver.o
00:20:34 Nick |Judo| is now known as Judo (
00:20:35linuxstbrasher: Are you using the latest patch? It compiles file for me under Mac OS X.
00:20:45rasherlinuxstb, yes, I am.. strange
00:21:17rasherlinuxstb, I changed the path to fwpatcher in the OBJS line in the Makefile, and got that
00:21:39linuxstbAre you using tools/fwpatcher/* ?
00:22:08rasherhang on
00:22:11linuxstbYou shouldn't be - the code should be inside the rbutil now - Domonoky extracted what he needed...
00:22:39rasherWell it isn't
00:22:55rasherI guess it got lost in the patch or something
00:22:56linuxstbYou shouldn't need iriver.c - I think it's in irivertools.cpp or something like that.
00:22:59 Quit Mouser_X (Nick collision from services.)
00:23:12 Join Mouser_X [0] (
00:23:19rasherI see
00:23:27rasherso I need to remove the fwpatcher objects?
00:23:39 Quit lee-qid ("aufwiederbyebientotsayonara")
00:24:38linuxstbrasher: Ah, I see the confusion now...
00:24:52linuxstbYes, revert back to the SVN version of the makefile and start again.
00:25:04linuxstbThe files that have been added there aren't used any more.
00:25:24rasherWhat does your OBJS line say?
00:26:18linuxstbI don't have one (only on my Mac, which I am not in front of). But I think you should remove the two fwpatcher objects, and add irivertools.cpp and md5sum.cpp
00:26:47linuxstbI'm just looking at the diff at the moment.
00:27:14linuxstbPlus bootloaders.o...
00:27:29rasherThat Bootloaders.o is there
00:27:33 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
00:27:34rasherI think it works now
00:27:37rasherAt least it finished without error
00:27:41rasherLet's see if it runs
00:31:07 Quit petur ("Zzzzz")
00:31:44rasherlinuxstb: Okay. It wants to run, but it can't find my libwx.. and I can't remember how to tell it where to look
00:32:10 Part nls
00:32:21Mikachuis this linux?
00:32:37Mikachu(not including the file name)
00:32:54rasherAnd lo, it runs!
00:33:09Mikachuyou can also call the linking step with -rpath /path/
00:33:25Mikachuif you're bored
00:33:34 Join jaebird [0] (
00:33:35rasherThis will do
00:34:59midkayJdGordon: around?
00:35:17JdGordoni is
00:35:33*preglow waits for his nano to wake up
00:36:15midkayJdGordon: what's the current situation on root menu.. what needs to be done for it to be committed? a few keymaps?
00:36:40bospaadjepreglow, do you have the hold-button problem i described earlier as well? (after switching hold button off, buttons don't react until you've used scroll)
00:36:59Mikachui do
00:37:05 Quit idnar (Read error: 145 (Connection timed out))
00:37:09Mikachui usually just scroll a bit after turning hold off
00:37:16preglowbospaadje: known bug
00:37:17JdGordonmidkay: that, and fix the bugs in the tree browser
00:37:27midkaywhat are the current bugs?
00:37:34 Join FOAD [0] (
00:37:48preglowlinuxstb: didn't slowcoder fix the hold switch bug?
00:37:52bospaadjeyeah, i found it in FS already, just wondering wether ne1 is actually trying to fix it :)
00:38:14 Quit Judo ("I'm sorry I can't stay, but I don't want to spread the AIDS any more than necessary.")
00:38:24JdGordonmidkay: trying to go from wps > file browser will put you one directory higher than you want, and posibly not let you exit
00:38:45 Join idnar [0] (i=mithrand@unaffiliated/idnar)
00:38:58 Join noldon [0] (n=noldon@
00:39:01rasherAren't these the same for the Iriver bootloader as well: ?
00:39:04noldonhi i need help
00:39:43bospaadjenoldon, with what? just ask your question.
00:40:33Mouser_XYes. Unfortunatly, all our telepaths have found better jobs elsewhere.
00:41:00Mouser_XSometimes, I wish they had gone into the tech-support area...
00:41:13Mouser_XBut I don't think I would wish that on anyone...
00:41:14noldoni installed the bootloader for my ipod video and now nothing happens it just restarting all the time
00:41:24noldonand all i hear is a click
00:41:48linuxstbpreglow: I remember having a conversation with tempel about it in #ipodlinux-dev, so I thought he fixed it...
00:42:02noldonit dosnt load the rockbox bootloader
00:42:05rasherlinuxstb, it worked, and created a proper patched firmware. Adding a comment now.
00:43:19preglowlinuxstb: might be, but we never took that code did we?
00:43:27noldonwish bootloader should i use for ipod video
00:43:28linuxstbNo, I don't think we did.
00:43:40preglowmy nano doesn't want to wake :/
00:43:49linuxstbnolden: Just run ipodpatcher without any arguments, and it will install a built-in bootloader for you.
00:44:21linuxstb(assuming you've downloaded the latest ipodpatcher)
00:44:22JdGordonmidkay: if your not sick of making icons yet, could you do one for the eq menu please? the +'s are no good :)
00:44:43JdGordonpreglow: reset the stupid thing...
00:44:48midkayJdGordon: haha. gladly. also what happened to the file view one?
00:44:49 Join barrywardell [0] (n=barrywar@
00:44:52midkayi can give you the code for it..
00:45:06JdGordonhehe, keep forgetting about it
00:45:10midkayi just don't know how to put it in the actual menu as the icon, i can't find where it goes.
00:45:14preglowJdGordon: not like that, it doesn't want to wake up when the rtc alarm goes off
00:45:19noldonlinuxstb i downloaded the one from the manual
00:45:20Lunar_LampHmm, m4a file playbakc doesn't seem to work properly for me. It freezes every few seconds for a blip. How do I check I'm using the latest rockbox version, and is this a known issue? (Player: iaudio x5)
00:45:26JdGordonpreglow: ah :)
00:45:28Lunar_Lamp(other files play fine: ogg and mp3)
00:45:34Mikachupreglow: the alarm works when it's turned off? :)
00:46:11JdGordonmidkay: in the MAKE_MENU() macro for the item, file fiew should be in menus/settings_menu.c
00:46:29midkayJdGordon: would you mind doing it? otherwise i've gotta start up vmware...
00:46:31 Join Thundercloud_ [0] (n=thunderc@
00:46:55JdGordonoh alright :p
00:47:31preglowMikachu: isn't that the whole point?
00:48:20Mikachui don't know, i guess it would be much more useful
00:48:24*Mikachu tries
00:48:59noldonshould i unzip the directly onto the ipod root
00:49:11noldonand then after that install the bootloader
00:49:39midkayJdGordon: got the bitmap?
00:50:00JdGordonall done
00:50:08midkaygreat, and..">
00:50:33midkayseems the best to me (a couple of sliders).. lots of ways you could go on this one but this one gives best visual indication i think.
00:50:35Mikachuthat said, where is the wake up alarm?
00:50:38 Quit Lunar_Lamp (Remote closed the connection)
00:50:48Mikachusearching "alarm" in the manual only says it isn't voiced
00:50:51JdGordonyeah, thats good
00:50:56rasherlinuxstb, I can't install Rockbox. It claims it was unable to unzip. Where do I report?
00:51:14midkayJdGordon: cool, commit! :)
00:51:20midkayJdGordon: do you need bitmaps for the root menu?
00:52:22JdGordonwhatcha got for it? I tihnk ill just use one of the notes from rasher's page (assuming he doesnt mind?)
00:52:43midkayi mean for any individual items...
00:52:54noldonall that appears on my ipod is a folder whit the adress under it
00:52:56JdGordonthe browsers...
00:53:00midkayi think the submenu icon would look good for the title icon.
00:53:05rasherJdGordon, I don't. The page even says they're public domain :)
00:53:06midkayi'm on it!
00:53:09preglowMikachu: i'm trying to implement it
00:53:15JdGordonim using the folder icon atm, but that is.. meh
00:53:27Mikachupreglow: i just realized that :)
00:54:23noldoncan anyone help me
00:54:44JdGordonnoldon: reset it
00:55:01noldonJdGordon how
00:55:22JdGordonhold select + menu
00:55:25JdGordonfor about 10s
00:55:35noldonive done that
00:55:51Mikachuthen hold select+play before the apple shows up
00:56:01noldonall i got was a folder whit a warningsign on it and the adress to apples support site under it
00:56:54 Quit Nico_P (Remote closed the connection)
00:57:35midkayJdGordon: i think we don't need any new icons, here's what i'd propose:
00:57:59 Join Thundercloud__ [0] (n=thunderc@
00:58:11midkaybrowser: playlist icon. database: musical note icon. now playing: PLAY icon. settings: general settings.... spanner... icon. :) everything else should be already there..
00:58:41JdGordonwe are going to add a playlist option tho, so dont want the same icon twice
00:58:44 Quit Thundercloud (Operation timed out)
00:58:54JdGordonmusical note is for wps...
00:58:56midkayoh, alright.. how about the folder icon then... that works even better.
00:59:07midkayright, i suggest the PLAY icon = wps and musical note = database/music library.
00:59:13noldonnothing is working
00:59:15midkaynow PLAYing? :D
00:59:26JdGordon... maybe...
00:59:45midkay*shrug* i think it'd work well.
01:00:08*JdGordon working on eq menu today, will play with root again after this is done
01:00:20midkaysounds good.
01:00:34 Join Lunar_Lamp [0] (i=lunar@unaffiliated/lunarlamp/x-038437)
01:00:49preglowi write the alarm regs, i set the pcf to wakeup on RTC alarm, i even bloody unmask the alarm interrupt
01:00:50 Nick Everybody is now known as Everybody|away (
01:00:52preglowno dice
01:00:54noldoni have the ipodpatcher v 0.8 is that the latest?
01:01:09JdGordonpreglow: :( you'll get it...
01:01:40preglowproblem is i don't have too much time to get it
01:01:46Lunar_LampOk, I'm using the latest daily build on an iAudio X5. m4a playback is not working in realtime - that is, it glitches every couple of seconds, and then catches up again a few seconds later. MP3 and OGG playback works fine. Is this a known issue and is there a workaround (other than converting all m4a formatted files to another format)?
01:02:08pixelmalinuxstb: with your chopper-archos patch it is really better playable... I even think it's a bit slow in the beginning now...
01:04:26linuxstbpixelma: The speed is easily changes - just change the definition of CYCLETIME near the top.
01:04:48 Quit secleinteer (Read error: 110 (Connection timed out))
01:05:12linuxstbnoldon: Yes, ipodpatcher 0.8 is the latest. What does your ipod do now?
01:05:28noldonit dosnt do anything
01:05:45noldoni got it to reinstall to the original bootloader
01:06:01noldonbut he wont install the rockbox
01:06:09linuxstbWhat did you do to your ipod before you tried to install Rockbox?
01:06:13pixelmathe heli looks a bit odd but that's a general problem (just a bit more obvious)
01:07:21noldoni unzipped the rockbox to the root on my ipod
01:07:31noldonthen installed the bootloader
01:07:50 Quit zylche (Read error: 54 (Connection reset by peer))
01:08:08noldonand either i got a screen with a folder on and the adress or he just restarting all the time with a click sound
01:08:57linuxstbWhich instructions did you follow to install the bootloader?
01:08:58 Join zylche [0] (
01:09:39noldonthe manuall for ipod video
01:10:08noldonand now hes only booting the original firmware
01:10:20linuxstbAre you installing from Windows?
01:10:54noldonno ubuntu
01:11:46linuxstbSo your ipod is working now? i.e. it's booting the original firmware as normal?
01:12:04linuxstbIs it a FAT32 or HFS ipod?
01:12:25rasherRestarting with a click sound.. could that be a charging-from-usb loop thing?
01:13:19noldonlinuxstb yes now its working as always
01:13:37noldonso wish package should i install and wish bootloader should i use
01:14:58noldonand btw how do i remove the .rockbox folder and the rockbox.ipod file from my ipod
01:15:21***Saving seen data "./dancer.seen"
01:15:36linuxstbIs it just me or is unavailable?
01:15:38bospaadjenoldon, probably using rm? (or nautilus, since you're on ubuntu)
01:16:12 Join aflac [0] (
01:16:21noldonbospaadje i tried to remove it but when i disconnect it and connect it again it still there
01:16:38rasherlinuxstb, not responding to me either
01:16:48*linuxstb pings LinusN and Bagder
01:17:01Lunar_Lamp(apologies for my earlier complain - as usual, pebkac error - I forgot to copy the .rockbox folder across)
01:17:13bospaadjenoldon, doesn't it give you an error message when trying to remove?
01:17:29LinusNlinuxstb: pong
01:17:46noldonbospaadje no
01:18:34linuxstbLinusN: is dead
01:18:46*linuxstb takes that back - it's alive...
01:18:53bospaadjenoldon, it should be just as easily removable as music files or anything else on your ipod
01:19:00 Quit Thundercloud_ (Read error: 110 (Connection timed out))
01:19:08bospaadjehave you tried removing them from command line?
01:19:20bospaadjethat usually spits out more useful error messages
01:19:47rashernoldon, to remove .rockbox from nautilus, you'll have to enable showing hidden files
01:19:54aflacIs there a way to change the auto-turn-off-point when the battery is getting low? My 4g ipod is crashing before it shuts down, and I have to either wait on the battery to FULLY drain or open it up and unplug the battery to un-freeze it.
01:20:07noldonrasher i know
01:20:09linuxstbnoldon: You should just need to download this file - - then type "chmod +x ipodpatcher" and then type "./ipodpatcher". It should then ask you if you want to install the bootloader.
01:20:13noldonbut it wont dissapear
01:20:34rashernoldon, are you unmounting it correctly before disconnecting?
01:20:40noldonlinuxstb i tried that
01:20:52noldonrasher i dont know
01:20:54linuxstbaflac: Holding MENU+SELECT doesn't work?
01:21:00noldonhow do i unmount it
01:21:14noldonmaybe thats the problem
01:21:20rashernoldon, right click the icon on your desktop and select unmount. That's most likely the problem
01:21:31aflaclinuxstb: nope, completely frozen
01:21:44linuxstbnoldon: Also, when you install the bootloader, look at the disk icon in the top-left of your ipod's LCD, and wait for it to stop spinning before disconnecting.
01:23:11 Join taerte [0] (n=piespy@
01:23:17aflacseems to work fine as long as I leaved it plugged in, but trying to run off of batteries is a risky propsition, as I never know when it's going to freeze up.
01:25:19 Part LinusN
01:25:42noldonnow i got it to work
01:25:56preglowwhich is more than i did
01:26:02noldonwere do put the music
01:26:10preglowwherever you want
01:26:16preglowas long as you can find it again, it's up to you
01:27:20 Quit pearldiver ("some games are better left unplayed")
01:27:38noldonwhy cant it find the music i synced with itunes?
01:27:53preglowbecause itunes adds the music in a retarded way
01:27:55preglowbut you still can
01:27:58preglowby using the database
01:27:59rashernoldon, because it's in a hidden directory. If you show all files, you can find it
01:28:17preglowor that, but the file names are screwed up
01:28:19rashernoldon, or by using the Database (which is preferable in this case, since the filenames are mangled beyond recognition)
01:28:27rasherI need to type faster.
01:28:36 Quit powriter25 (Remote closed the connection)
01:28:49rasherAnd go to sleep. Goodnight.
01:28:55 Quit rasher ("leaving")
01:31:30 Quit Thundercloud__ (Remote closed the connection)
01:32:16preglowcan retailos wakeup on alarm?
01:33:22Mikachui don't think so
01:33:51 Quit aflac (Read error: 145 (Connection timed out))
01:33:53linuxstbpreglow: It's not working?
01:34:08preglowlinuxstb: can't make it work, no, i can set the alarm time and read back the values
01:34:12preglowbut that's that
01:34:26dan_apreglow: RetailOS does have an alarm function
01:34:37preglowjust making sure it's possible
01:35:09linuxstbAnd I assume you've set the RTCWAK bit when entering standby?
01:35:23preglowlinuxstb: indeed
01:35:33preglowin pcf50605.c
01:35:47linuxstbI wonder if that's still being used...
01:35:59preglowi hope not
01:36:34preglowi concluded it was anyway
01:36:41preglowlooks like it is
01:37:42preglowif it's actually called or not, i don't know
01:37:45preglowbut it looks like it is
01:38:36preglowyeah, it definitely is
01:38:45preglowi get weird behaviour on shutdown without it
01:41:03preglowi even unmask the alarm interrupt, but i don't think that matters
01:44:09noldonhow come that i dont have permissions to do anything on my ipod
01:44:37preglowhmm, no, that would actually be important, it seems
01:44:44preglowi should double-check i'm doing that right, then
01:45:45 Quit ender` (" Just when you get really good at something, you don't need to do it any more.")
01:47:01JdGordonwtf? how did adding 1 icon add 20kb to the rec fm build?
01:48:08noldonwhy dont i have any permissons on my ipod
01:48:19preglownoldon: linux?
01:48:23preglowJdGordon: sheer luck
01:48:30noldonpreglow yes
01:48:43noldonbut i had just a moment ago
01:49:12preglownoldon: that's an issue with your linux, not rockbox
01:49:23preglownoldon: btw, you do unmount it properly, yes?
01:50:58linuxstbnoldon: If Linux detects errors, it will remount a disk read-only.
01:51:21preglowif you just unplug it without unmounting and ejecting properly, you will very probably corrupt the file system and linux will do what linuxstb just said
01:51:34noldonpreglow yes
01:51:59preglownoldon: ok, then do a "dmesg|less" and see if anything is mentioned towards the end
01:52:02noldoni press eject
01:52:35preglowlinuxstb: and now i bloody well enabled all interrupts on the entire thing and i still get nothing
01:52:42preglowthis was supposed to be a quick deal
01:53:02noldoni unmount it by press eject
01:53:19noldonhow do i make it mount writeable
01:54:11preglowfirst check if linux has a reason for mounting it read-only
01:54:13preglowby doing what i said
01:54:39noldonFAT: Filesystem panic (dev sda2)
01:54:40noldon[17273753.716000] fat_get_cluster: invalid cluster chain (i_pos 0)
01:55:07linuxstbpreglow: Nothing turns out to be simple on the ipods...
01:55:12noldonshould i unmount it by doing something else then eject
01:55:13 Quit entheh ("^~")
01:55:18linuxstbWe should have learnt that by now.
01:55:25preglowapparently not
01:55:33preglownoldon: it's corrupted, you need to restore it
01:55:56preglowyou can do that by using dosfsck when it's unmounted
01:56:51noldondosfsck /dev/sda?
01:57:39noldonmedium not found
01:58:10XavierGrADX is in SVN?
01:58:20noldoni unmounted it by pressing eject
01:58:29Mouser_XXavierGr: Has been for months.
01:58:41Mouser_XSince around Sept. or August.
01:58:48pixelmaJdGordon: looks like there is something wrong in the table there (to me)
01:58:55midgeybarrywardell: are you able to build any of the hwcodec builds on mac?
01:59:20linuxstbMouser_X: SVN logs say 25th September 2006.
01:59:54Mouser_XThen it was probably started around August.
01:59:54XavierGrMouser_X: You said that hcs did that?
02:00:01Mouser_XHe did.
02:00:38XavierGrstrange didn't know that
02:00:50 Join web-taz [0] (
02:00:52XavierGrit is my first time I hear the name of that codec
02:00:56Mouser_XHe put NSF and SPC into the SVN as well.
02:01:12web-tazhello everyboday
02:01:14noldondosfsck /dev/sda2 dosnt work when its not mounted
02:01:30 Join pearldiver [0] (
02:01:40linuxstbnoldon: Try not ejecting, but unmounting - e.g. with "umount /dev/sda2"
02:01:44*Mouser_X hopes that HCS writes a GBS codec as well.
02:01:47 Quit Weiss (Read error: 104 (Connection reset by peer))
02:01:48XavierGrMouser_X: Yes I know about them
02:02:34noldonshould i use first fat or second fat
02:03:12XavierGrYes you are right
02:03:24XavierGrCommited by Rani Hod on 25 of August
02:03:30noldonit leaves it unchanged when i choose correct
02:03:34XavierGrno September
02:04:07XavierGris ADX considered to be emulated music?
02:04:48linuxstbIt's a form of adpcm
02:05:23XavierGrso it should be under loosy or other codecs?
02:05:28preglowthe bloody thing triggered!
02:05:35preglowso why the flaming hell doesn't it reboot
02:05:46preglowat least that means my alarm setup code is correct
02:06:09linuxstbXavierGr: I think it's lossy, at least it throws away information compared to 16-bit PCM.
02:06:17noldondosnt work
02:06:27preglowlinuxstb: i found a pcf register debug menu patch in the tracker so i sat staring at the screen unblinking for two minutes waiting for 0xc0 to flash for a second
02:06:33preglownot what i intended to do this night
02:07:09XavierGris it known if it is realtime on all targets?
02:07:11linuxstbIs RTCWAK defined correctly?
02:07:28linuxstbXavierGr: It should be very efficient - someone mentioned 1% boost earlier...
02:07:29preglowlooks like it
02:07:34 Join Weiss [0] (
02:07:43noldonit wont work to run dosfsck on /dev/sda2
02:07:45preglowbit 4, which is 1 << 4
02:07:54XavierGrlinuxstb: ooops I should read the logs more carefully then
02:08:25linuxstbThere's no complicated maths in the codec.
02:08:33 Quit wooo (Read error: 110 (Connection timed out))
02:08:35XavierGrwell I updated the wiki page, but I am very suprised that I discovered a codec that I didn't know that Rockbox supported
02:08:43preglowXavierGr: do you know about spc? :>
02:08:48XavierGrof couse!
02:09:14XavierGrSPC was on my personal preference list and updated my build with the patch (before the commitment)
02:09:15linuxstbXavierGr: You're not alone...
02:09:47linuxstbnoldon: You could simply reformat it.
02:10:21Mikachunoldon: "doesnt work" isn't very useful
02:10:21noldonyeah and how do i get the apples system on it?
02:10:43Mikachudoes it find any errors?
02:10:56Mikachuby default it doesn't write any fixes to disk, i think you have to say -r
02:11:18linuxstbYou don't need to - the Apple firmware is in the first partition (/dev/sda1), so reformatting the FAT32 partition won't touch the Apple firmware. You will of course have to copy your music back.
02:11:20XavierGrPSF would be cool to have too
02:11:25XavierGrvery nice emulation
02:11:25preglowXavierGr: word up, and it'll get better soon too, jhmikes has some optimisations going on that'll allow us to reenable echo
02:11:34preglowXavierGr: psf is a bit cpu heavy
02:11:45XavierGrI've been following your conversations about that
02:11:48Mikachuis gba music as heavy as psf?
02:11:53Mikachui know the formats are similar
02:12:02XavierGrgba and psf?
02:12:10Mikachui forget the extension, hang on
02:12:20noldonlinuxstb its no problem
02:12:23XavierGrgba is for game boy while psf for playstation
02:12:24Mikachuah, .gsf
02:12:31Mikachuno, gba as in game boy advance
02:12:38XavierGryes you are right
02:12:49XavierGris for game boy
02:13:23Mikachufor example the whole soundtrack for final fantasy tactics advance is 904kB :)
02:13:59LloreanThe GBA sound format may be simpler than SNES.
02:14:00Mikachuhrm, it uses 11% of my 2ghz cpu to play though
02:14:16Mikachubut i can't vouch for playgsf 0.07
02:14:21XavierGrI think hcs said that psf would be possible
02:14:25LloreanI know at least that SNES games ported to GBA always have horrible sounding music because the chip can't reproduce it wholly. That doesn't require that it's easier to emulate the GBA chip, but it suggests it may be.
02:14:26XavierGrat least on gigabeat! :D
02:14:28Mikachuits optimizedness rather
02:14:56preglowXavierGr: well, you can see how much cpu spc uses now
02:15:27preglowpsf needs to emulate a mips cpu at a possible max of 33mhz, as for audio there are more samples with higher quality, i believe they're also adpcm encoded
02:15:28XavierGrI think it has dropped considerably
02:15:31preglowalso, there's the reverb
02:15:37XavierGrsince it was struggling on coldfire too
02:16:05preglowi'd love to see psf support, though, but for now i've got the few soundtracks i like encoded as ogg :>
02:16:17preglowthat would be the chrono cross soundtrack which everybody should have
02:16:45XavierGrchrono cross? any relation to chrono trigger?
02:16:49preglowyep, it's a sequel
02:17:03preglowsame guy who made the music too
02:17:11preglowbut he did an even better job, which is no mean feat
02:17:37XavierGryasunor mitsuda
02:17:42*Mikachu is listening to Yasunori Mitsuda - The Brink of Death (0:14 / 2:48)
02:17:43preglowyasunori :>
02:17:59XavierGryeah the guys is one of my favorites
02:18:08preglowthat too is great
02:18:10XavierGrincredible soundtrack
02:18:14preglowbut not as great as chrono cross
02:18:15XavierGr(and game)
02:18:16 Quit YouCeyE ("Leaving")
02:18:23noldonnow i got it to work again
02:18:23XavierGrI think I am going to get it then
02:18:25noldonthank u
02:18:51Mikachui like legend of mana's soundtrack too
02:18:54noldonanyway the themes is just to unzip into the root folder right?
02:18:56preglowyeah, me too
02:19:05preglowalso, that game has absolutely lovely graphics
02:19:10XavierGrMikachu, I once tried to play it, but in the end I quit
02:19:18preglowthe game isn't all that good, no
02:19:24Mikachuit's a bit weird, but i liked it
02:19:37preglowi liked secret of mana and seiken densetsu 3, though
02:19:43XavierGrthe best
02:19:48Mikachui probably wouldn't have liked it if i wasn't a die hard secret of mana fan :)
02:20:03preglowsecret of mana was pretty much my first encounter with an rpg
02:20:10XavierGrI've said it many times, but nearly all my childhood memories are filled with legend of zelda and secret of mana
02:20:11preglowi absolutely love it
02:20:12Mikachusword of mana for gba sucks though
02:20:21preglowXavierGr: haha, yeah
02:20:26preglowhad tons of nice times with those
02:20:33XavierGrand imagine that I didn't have a snes
02:20:39Mikachui remember i never managed to get the axe to level 9
02:20:41safetydanXavierGr, while you're updating SoundCodecs, you could always add Speex
02:20:44Mikachuall the other weapons, but not the axe
02:20:48preglowMikachu: i thought only the sword did level 9
02:20:49XavierGrI was playing it at summer in my village when Genre9mp3 would come down in Crete :P
02:21:10preglowoh, i played them on snes
02:21:12Mikachupreglow: if you roamed the fortress you'd get orbs for all weapons
02:21:24preglowMikachu: yeah, don't think i ever bothered for long
02:21:25XavierGrso time consuming though
02:21:29preglowbut i don't remember
02:21:32preglowlong time ago i completed it now
02:21:42XavierGrOn an emulator I bothered to level all spells to level 9 though
02:21:42Mikachui think i've spent about 1 month effective time just in the fortress of mana :)
02:21:46Mikachu(when i was little)
02:21:49preglowi'm doing seiken densetsu 3 an hour here and there these days
02:21:53Mikachuthat music is burnt in my retina, as it were
02:22:21midgeyhmm, gmake detection in is broken on mac...
02:22:21XavierGrI still remember the sound of secret of mana when you turned the game on
02:22:22preglowthe music gets more and more esoteric the further towards the end you get
02:22:32Mikachui've played two of the three main quests on sd3, but i have no idea which so i never bothered playing it a third time since it might be the wrong one
02:22:42preglowXavierGr: i also remember the sound i made when the fucking ntsc->pal adapter glitched and all my saves were deleted...
02:23:10XavierGrthat would be painfull
02:23:14Mikachui traded my som cart for shadowrunner :P
02:23:14preglowit was, several times
02:23:24preglowMikachu: also i enjoyed that a fair bit
02:23:29Mikachuin retrospect, not the best trade
02:23:33preglowMikachu: i traded mine too, for god knows what, i've regretted it since
02:23:40Mikachubut at the time i had played som > 234 times and shadowrunner 0.5 times
02:23:54Mikachushadowrunner in swedish at that
02:24:07preglowswedish cyberpunk
02:24:09preglowwwhat a treat
02:24:35XavierGrI had a sega mega drive, now I understand the major mistake I did when I sold it
02:24:49Mikachunow that there are emulators i don't regret it as much
02:24:56preglowi've still got my snse
02:25:03XavierGrbut I still have an original game boy with 10 cartridges
02:25:04preglowbut i have no idea where most of the games are
02:25:09preglowor at this moment, even where the console is
02:25:10Mikachui only have three snes carts, no snes (zelda, fzero and sim city)
02:25:17XavierGrah fzero!
02:25:25XavierGrjust great
02:25:40XavierGrgreat music too
02:25:42 Quit Mouser_X (Read error: 110 (Connection timed out))
02:25:46preglowhighhats at 500bpm
02:25:59Mikachutook many years until i finished it at the secret 4th skill level
02:26:10XavierGrI don't know if I ever did that :P
02:26:23XavierGrmario kart was awesome too
02:26:25Mikachuiirc my favorite tracks were the ice level and silence
02:26:25XavierGrvery fun
02:26:40XavierGrand zsnes has netplay
02:26:48XavierGrcrazy stuff!
02:26:56XavierGrI did that once
02:27:08XavierGrI enjoyed it
02:27:12Mikachuthe only games i played over netplay are goof troop and tetris attack :)
02:27:17Mikachuboth of which also have good music
02:27:24XavierGrnever played them
02:28:01XavierGrdid you know that game boy had a game that could have up to 16 players?
02:28:04XavierGrit was an fps too
02:28:07XavierGrcrappy game though
02:28:12preglowgod, the title track of chrono cross is really something
02:28:19XavierGrbut imagine the scene, lan party with 16 game boys
02:28:45Mikachuheh, i just remembered the title theme in legend of mana was sung in swedish
02:28:50Mikachuit really surprised me when i played it
02:28:50preglowhaha, yeah
02:29:37preglowbut the bloody chick from rednex, i think
02:29:56Mikachulettuce see what wikipedia has to say
02:30:03preglowa band that is better left forgotten
02:30:18Mikachuit agrees
02:30:24MikachuAnnika Ljungberg
02:30:30Mikachushe sings that song well enough thouh
02:31:01Mikachui need someone to give me an nds :)
02:31:02preglowi've got a couple of final fantasy tracks that are sung in saami
02:31:05preglownow that's esoteric
02:31:44Mikachunintendo ds
02:32:11Mikachuthere is actually already an emulator that half works
02:32:31XavierGris there an emulator for psp?
02:32:48Mikachui haven't heard of one
02:33:09Lunar_LampHow do I add fonts that I don't have? That is, if a wps requires a certain font, how do I provide it?
02:33:18MikachuLunar_Lamp: put it in /.rockbox/fonts
02:33:29noldonso were do i put the music?
02:33:36Lunar_LampMikachu, and where do I get the fonts from? I just google for them>
02:33:37Mikachunoldon: anywhere! (on the ipod)
02:33:46MikachuLunar_Lamp: they should come with the theme
02:34:18Lunar_LampMikachu, what if they don't?
02:34:29preglowMikachu: like i said, put it wherever you want. rockbox doesn't care as long as you can find it
02:34:40 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
02:34:51Mikachui'll assume you meant noldon there
02:35:20noldonpreglow ok
02:35:40XavierGrin home village arni: the *swissing* sound is it supposed to be the guitar fretboard? LOL
02:35:41Lunar_LampMikachu, if /.rockbox/fonts doesn't exist, should I create it myself?
02:35:54 Quit taerte (Remote closed the connection)
02:36:03MikachuLunar_Lamp: if it doesn't exist, you haven't installed rockbox
02:36:40Lunar_LampMikachu, the /.rockbox folder exists, and I boot rockbox and it runs. /.rockbox/fonts does not exist.
02:36:41preglowXavierGr: yes
02:36:52MikachuLunar_Lamp: you want
02:37:02MikachuLunar_Lamp: i'm sure the install instructions tell you to install that
02:37:27Lunar_LampMikachu, hmm, I must have missedthat.
02:38:06Lunar_LampMikachu, it does *sigh*: "If you are installing Rockbox for the first time, or reinstalling after removing the .rockbox directory it is recommended you install the fonts package from"
02:38:19Lunar_LampDo I just unzup that folder in the /.rockbox/ folder?
02:38:28Mikachui think it has the full structure, so unzip it to /
02:41:39 Join secleinteer [0] (
02:43:00web-tazor just unzip it in a seperate folder and find out, how the structure looks like
02:43:30web-taz(is this sentence, especially the 2nd part, right English?
02:43:47Mikachuchange how to what
02:44:14Dunkelschuband remove the comma
02:44:17Mikachuyou can say "how X looks" or "what X looks like", but they don't mean exactly the same
02:45:00Lunar_LampOk, I did that.
02:45:04Lunar_LampThanks, it worked fine.
02:45:18Lunar_LampOn a related note - what happens if you load a theme that can't be viewe dfor whatever reason?
02:45:22Lunar_LampHow do you "undo" it
02:45:46web-tazyou could just load the default theme?
02:45:57web-tazor reset the settings
02:46:01Lunar_Lampweb-taz, how though - I can't see any of the menus
02:46:08Mikachureboot and enable the hold switch
02:46:16Lunar_Lampenable the hold switch?
02:46:25Mikachuturn on hold? activate hold?
02:46:43Lunar_LampThat didn't do anything.
02:46:50Lunar_LampI turned off, turned on, put it on hold.
02:46:56web-tazwhat's the effect? I never heard of this
02:46:56Lunar_LampThe "hold" switch is on/off.
02:47:08Mikachuyou want to turn on hold before rockbox finishes booting
02:47:16Mikachubut not before the bootloader finishes
02:47:17Lunar_LampOk, i did that.
02:47:21Lunar_LampAh, right.
02:47:28Mikachuor you'll go to the apple firmware
02:47:30Lunar_LampI was putting it on when the bootloader was still going.
02:47:56Lunar_LampIf I get apple firmware on an iaudio I think I'd get a pretty amazing prize for firmware hacking ;-)
02:48:03Mikachuyou can also just plug it in and edit .rockbox/config.cfg
02:48:06Mikachuor remove it
02:48:24Mikachuoh heh, sorry i just assumed you were using an ipod, must have mixed you up with someone else
02:48:30Mikachuthey probably have another settings reset button
02:48:33Lunar_LampNo worries :-)
02:48:54preglowgah, this won't work out
02:48:56web-tazwhat's the effect of turning on hold while booting? is that available on every player with rockbox?
02:49:11Mikachuon ipods it boots the apple firmware or resets settings
02:50:15Lunar_LampHmm, are you sure it's "config.cfg" Mikachu?
02:50:22Lunar_LampIt's just a single line file - that looks like a comment to me.
02:50:40Lunar_Lamp# .cfg file created by rockbox r12384-070218 -
02:50:43Lunar_LampThat's the entire contents.
02:50:55Mikachuthen it should be using the default settings (ie readable menus)
02:50:57Mikachutry removing it anyway
02:51:54web-tazwhats with the viewers.config
02:51:58barrywardellmidgey: sorry, missed your message earlier. My mac is in being repaired at the moment. I haven't tried compiling HWCODEC targets in quite a while
02:52:11Mikachuweb-taz: that is somewhat unspecific
02:52:37Lunar_LampMikachu, nope.
02:52:52Mikachuthen your last resort is to look what the settings reset button is in the manual :>
02:53:15web-tazokay :D i'm just learning, though having vacances
02:53:32 Quit barrywardell ("Leaving.")
02:53:40Mikachuweb-taz: it associates file types with plugins
02:54:07web-tazah well the viewer plugins like mpegplugin and so on...
02:55:41 Quit miepchen^schlaf (Read error: 60 (Operation timed out))
02:59:00 Join miepchen^schlaf [0] (
02:59:16Lunar_LampI had to delete a couple of files related to the theme - but it's viewable again now.
02:59:46Lunar_LampI suspect it was to do with the theme I was using and changed from using an all-black backdrop
02:59:55noldoni cant find the music on the ipod
03:00:05Mikachuthe itunes music is a bit hard to find
03:00:18noldoni put it in an own folder under .rockbox
03:00:38Mikachuit's better if you don't put your music in there
03:00:40Llorean.rockbox is hidden in most file view modes.
03:00:47noldonyea but im looking for the one i copyed to the ipod now
03:00:55noldonok were should i put it then
03:01:00noldonin the ipods root?
03:01:03LloreanAnywhere else.
03:01:13LloreanJust as long as it's not in a hidden folder, and it's someplace you can remember
03:01:26LloreanHidden folders are okay, but they're a little more annoying to get to.
03:02:55web-tazto find it: why don't you use the database function?
03:03:19Lloreanweb-taz: Many people prefer file tree navigation.
03:03:48web-tazme too, but in order to search a file i often use the database
03:04:58web-tazi managed my music in folders from artist -> album -> MP3s and all files are named in the same way,
03:06:00noldonok i think o got it now
03:06:04web-tazbut samplers for example are (damn i miss a word^^fuck vocabulary) difficult to find...
03:06:19noldonis it possible to boot the original bootloader without reinstall it
03:06:35noldoni mean have both apples and rockbox on the ipod
03:06:41noldonlike a dual boot
03:07:08web-tazit's possible i think
03:07:22LloreanIf you followed the official install instructions, you have dual boot, yes.
03:07:25 Join bk [0] (
03:07:37LloreanWell, strike that, I don't know which player you have.
03:07:45LloreanOn anything but the X5 or the Gigabeat, you have dual boot.
03:07:49web-tazi remember so telling that there are bootloader where you also can choose between rockbox, apple FW, and ipodlinux
03:08:19pixelmaehmm... he asked about appleOS and rockbox ;)
03:08:20Lloreanweb-taz: You can do that with the official button, depending on which button you hold down while booting.
03:08:28noldonLlorean i have the ipod video 30gb
03:08:43Lloreanpixelma: Ah, yes, I just read the "boot the original bootloader" and didn't read the next line.
03:08:58Lloreannoldon: Just turn on hold right after turning on the iPod.
03:10:05LloreanA question mark alone gives me no information about what you don't understand...
03:10:26noldoni got it
03:10:29noldonforget it
03:11:43noldondamn this is so cool
03:12:31web-tazYes, it is
03:15:17noldonwell time to sleep
03:15:20noldonsee ya
03:15:23***Saving seen data "./dancer.seen"
03:16:19 Quit web-taz ("( :: NoNameScript 4.02 :: )")
03:17:21 Part noldon ("Ex-Chat")
03:17:30JdGordonwhos server is ?
03:19:12 Join sound_fx [0] (
03:19:26sound_fxHello out there
03:19:39pixelmaJdGordon: Nimdae's
03:20:20JdGordonhmm. he doesnt have paths set correctly :(
03:20:35sound_fxI have a few simple Rockbox questions.
03:21:19sound_fxIf I have a song paused, it will turn off eventually, right?
03:22:27LloreanIt should, I believe. I've never tested, as I just stop the music, since you can resume anyway.
03:22:58sound_fxUm, is there a different button to stop it versus pause?
03:23:24LloreanWell, I don't know what player you have. On MOST of the players, there's a stop button.
03:23:33LloreanOn the iPods, you hold down Play/Pause for a bit longer.
03:23:48sound_fxAh, ok
03:24:58sound_fxI see... Additionally, does it turn off when I have it plugged in to charge? (I have an iPod 5.5g)
03:25:17LloreanProbably not.
03:25:34LloreanIf it does, it'll immediately turn back on anyway, iPods wake on charger insertion.
03:26:00sound_fxOk, so do you normally charge it fully, unplug, and then turn off by holding play?
03:26:26LloreanOr just leave it plugged in.
03:26:55LloreanUsually I boot mine into the apple firmware for charging, since it charges faster there, at least when hooked up to a computer.
03:27:02sound_fxOk... I see... I'm just trying to maximize battery life.
03:27:17sound_fxI saw a post saying that it charges faster in iPod firmware, right?
03:27:17Mikachuit'll turn off when paused
03:27:41Lloreansound_fx: That is what I just said, yes...
03:27:54sound_fxAh, sorry.
03:27:59Lloreansound_fx: And what does 'maximizing battery life' have to do with charging. Do you mean battery lifetime?
03:28:18LloreanBecause it will stop charging on its own when the battery is full, whether you disconnect it or not.
03:28:36sound_fxOk. I apologize for these questions, I just picked up the iPod today.
03:28:58robbanyone with a archos 604 wifi? i have some questions about network speeds
03:29:14Lloreanrobb: The Archos 604 is not a rockbox target.
03:29:55 Join [shodan] [0] (
03:30:11robbthat's appreciated, Llorean, but doesn't really have much impact on my question :)
03:30:29Lloreanrobb: Well, you're asking in #Rockbox, so your question doesn't really belong here.
03:30:45robbwell would it matter if i was using rockbox or not?
03:31:19robbif i was using rockbox, and asked the same question would it belong then?
03:31:29LloreanIf the hardware was supported *in* Rockbox, yes.
03:31:33sound_fxOk... Thanks for answering my important questions. I installed some themes, uploaded some gb games, and tried out the video player, is there anything else that I can do that I haven't tried yet?
03:31:45Lloreanrobb: But this is not "#questions about MP3 players"
03:32:05LloreanYes, and you'll notice the 604 isn't among that list.
03:32:13LloreanSo I have no idea what "point" you're trying to make
03:32:28Lloreansound_fx: Play music?
03:32:30robbjust says jukeboxes?
03:32:37Lloreanrobb: Yes, that's a specific archos model. The Jukebox
03:32:41sound_fxyeah, did that one too.
03:32:52robbwell my bad then
03:33:01Lloreanrobb: The list of supported targets is *very* clear on the front page of the site
03:33:09robbremove rider from high horse, and continue please
03:33:11 Part robb
03:33:12pixelmaJdGordon: I've just discovered that the order in the "backlight settings" list is the wrong way around, meaning "off" is at the bottom...
03:33:59JdGordonI thought it looked funny, but wasnt 100% sure :p
03:34:07JdGordoncan we live with it like that?
03:34:49JdGordongrr... gonna need to figure out the nicest hack to fix it then
03:34:59LloreanIs there some reason why it got inverted?
03:35:11sound_fxOk, well, I think I am using rockbox to its fullest extent, thanks for your wisdom Llorean.
03:35:19JdGordonyes, it uses the set_int() "screen" instead of set_option() now
03:35:34 Join nickv111 [0] (
03:35:34JdGordonset_int put 0 at the bottom, set_option puts 0 at the top
03:35:36pixelmaI mean it wraps so it doesn't hurt but it is strange somehow
03:36:12nickv111Hey all. Anyone here who is a developer: where are math functions defined? If I wanted to add a new mathematical floating-point function, would I just put it in firmware/include/math.h, or is there another place where it would be coded in?
03:36:33 Quit sound_fx ("Leaving.")
03:37:02JdGordonnickv111: apps/plugins/lib i think
03:37:11JdGordonwe dont really need floating point math in the core...
03:37:29nickv111I see.
03:37:46*JdGordon back in 10
03:38:02nickv111Where is the floating point math in apps/plugins/lib
03:38:14nickv111Does it not exist, and I have to make it?
03:40:39 Part pixelma
03:42:43safetydannickv111, there is no libm in rockbox
03:43:24nickv111Hmm. Should I add math.h to apps/plugins/lib?
03:43:44nickv111math.h already exists in another location though. Maybe floatingpoint.c and floatingpoint.h?
03:44:08safetydannickv111, why do you need another math.h?
03:44:22nickv111What should I do?
03:44:31safetydanWell I'm not sure what you want
03:44:40nickv111I want to add a few functions to calculator.c
03:44:47safetydanSo add them there then.
03:44:48nickv111But a friend suggested I just add them to the math library
03:44:58nickv111So other plugins can use them.
03:45:18safetydanThere is no floating point math library in Rockbox. It's very unlikely there ever will be since they won't be fast enough for general use.
03:45:48nickv111So you think that I should just make the functions in calculator.c?
03:45:50safetydanSo I'd just say make them part of calculator.c
03:47:32 Quit shodan- (Read error: 113 (No route to host))
03:48:19nickv111BTW, in calculator.c, I'm getting rid of all of that power crap. The upper bound of a double is ±1.7976931348623157×10^308
03:48:28nickv111That's more than enough.
03:50:02safetydanWell it's unlikely anyone is going to be relying on a Rockbox calculator for extended precision.
03:56:30 Quit nickv111 (Nick collision from services.)
03:56:40 Join nickv111 [0] (
03:56:46nickv111Sorry, computer froze
04:07:05 Join Mouser_X [0] (
04:12:30JdGordonLlorean: ok, I got it fixed... can you tihnk of any other settings which are backwards?
04:13:27LloreanNot that I'm aware of right now
04:18:28 Quit nickv111 (Read error: 110 (Connection timed out))
04:19:01 Quit Lunar_Lamp ("Leaving")
04:20:31safetydanJdGordon, what's up with the 8 space indents in eq_menu.c? Are they tabs?
04:20:37 Quit jhMikeS (Read error: 104 (Connection reset by peer))
04:21:03 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
04:21:34safetydanErr... and most of those constants you've moved out to a header are supposed to be private to the eq menu system
04:22:34 Quit jurrie ("Leaving")
04:23:26 Quit jhMikeS (Nick collision from services.)
04:23:32 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
04:24:00safetydanOh, I guess they need to be public for settings_list.c
04:24:39 Join wacky [0] (
04:25:43wackyhey guys, why isn't the scroll-margins patch merged ?! almost all WPS for the X5 require it!!
04:26:46Soaponly when ALL the WPSs for the X5 require it will it be commited! Get to work!
04:28:23Lloreanwacky: Well, that's *their* fault for using it.
04:29:17wackyI don't understand why it's not been merged.
04:29:35LloreanBecause 'people use it' isn't a good enough reason?
04:29:37wackyIs there a reason ?
04:29:59safetydanwacky, either no one has got around to it, or we're waiting for something better
04:30:05LloreanSomething better
04:30:39wackyWhat are supposed to be viewports ?
04:30:55safetydanwacky, the ultimate in customisability probably
04:31:07LloreanBasically, like virtual screens.
04:31:57wackybut in the mean time, the X5 only has a couple of horrible default "supported" wps
04:32:31Lloreanwacky: So tell the theme authors to make themes that don't require it.
04:32:41wackywhy not merge the patch, since the ultimate isn't there yet.
04:32:52LloreanBecause then it'll have to be removed later, and that'll take even more work?
04:33:52LloreanI can't read words that aren't there, question marks don't mean anything on their own.
04:34:19wackyI guess those viewport will trash away the whole WPS concept, why fear a tiny margin'ed scrolling thing ?
04:34:27LloreanThey won't
04:34:39LloreanThe planned method should be compatible with all existing WPSes, at the moment.
04:35:09 Quit Mouser_X (Read error: 110 (Connection timed out))
04:36:23wackyand the viewports will introduce natural scrolling alignment ?
04:38:04wackyI'm just guessing that if all designers use that patch to make nicer layouts (and they are), then that patch should sound important..
04:38:59LloreanThey chose to use it, we can't be blamed if they wanted to use an unsupported feature.
04:39:53JdGordonsafetydan: grr... i dunno wtf my text editor did :p ill fix them
04:40:23wackyI'm not blaming anyone, just wondering if there was a reason for that unsupportedness
04:41:15Lloreanwacky: And you've been told.
04:41:24wackyYour the main developer in that area of the code ?
04:41:39wackyWho does that ?
04:41:56safetydanI think part of the problem is there's only a few devs interested in eye candy. So patches like that just sit there.
04:42:02safetydanwacky, no one person works on that
04:42:28wackyOr, should I ask, what part do you work on ?
04:42:49Lloreanwacky: It doesn't matter. This is why that feature is not accepted so far, and the reason given for why it's unlikely to be accepted.
04:44:31wackyLlorean - you part of the dev crew ?
04:44:55Lloreanwacky: Before I answer, why does it matter?
04:46:21wackyIt matters in how I'll consider your answer, and I understand that you aren't.
04:46:48LloreanActually, I'm a core member of the team with commit access, I'm in charge of the forums, and so on and so on.
04:47:01LloreanThough I'm not fond of the assumption that people would lie to you in here.
04:48:46LloreanRockbox hasn't been released for the X5 yet, and there's very little point in adding a feature that may, or likely will, be removed before an official release.
04:51:47wackysafetydan - I understand :)
04:52:14 Quit wacky (Remote closed the connection)
04:52:35thegeek_before a release, the primary goal is _not_ usability/features, it's having _good_ code/design
04:53:34Lloreansafetydan: Did you say something to him, or did he just get the name wrong for response?
04:54:02safetydanI've no idea.
04:54:23safetydanLlorean, I think maybe he was responding to the "no devs interested in eye candy" bit
04:56:28pearldiverpeople requesting the album art, scrolling margins, and custom line height patches can be seen everyday
05:00:28 Join Mouser_X [0] (
05:03:00Lloreanpearldiver: They should get to work on clearing out the things blocking those features (or features like them) from being implemented then.
05:04:49 Join bonbonthejon [0] (n=jon@
05:06:11bonbonthejonrockbox is giving me an error on boot, "Can't load rockbox.ipod: Read failed (image)"
05:06:45LloreanAnd have you tried re-copying the file to the device?
05:07:23Mouser_XXavierGr: I'm late on this one, yes, but GSF playback takes more CPU than PSF playback.
05:07:25bonbonthejonLlorean: yes, but when my computer mounts the ipod, it says its a read only filesystem, when i boot into disk mode
05:07:49Lloreanbonbonthejon: So, have you tried checking your disk for errors?
05:07:52Mouser_XXavierGr: In other words, I would expect to see PSF support (full-speed) before I see GSF playback.
05:08:06Mouser_XThough, some games would run better than others, of course.
05:08:43bonbonthejonLlorean: no, i;ll look up how to do that
05:08:56bonbonthejonLlorean: it was working just before this
05:09:21 Quit Mouser_X (Nick collision from services.)
05:09:30 Join Mouser_X [0] (
05:09:31Lloreanbonbonthejon: Well, the disk mode isn't part of Rockbox.
05:09:50Lloreanbonbonthejon: So if that's not working properly, you need to check up on things.
05:09:57bonbonthejonLlorean: yeah, but i need to get into it to update the files
05:10:36LloreanSo do a filesystem check, and see if the disk is okay
05:10:48Mouser_XAt least, this is my experience with Highly Advanced.
05:10:49Mouser_XHighly Experimental (PSF) ran much smoother on a 333 than Highly Advanced (GSF) did.
05:11:41Mouser_XPSF playback was full speed on my 233. PSF2 was not. On my 333, PSF2 playback was full-speed.
05:13:05Mouser_XJust wanted to point out that PSF is more likely than GSF, in my experience.
05:13:36Mouser_XSexyPSF is the only open-source PSF player there is. Anyone want to port it?
05:13:58Mouser_XI don't think HCS is to interested in doing it...
05:14:18 Quit printfXh4 (Remote closed the connection)
05:14:55 Join printfXh4 [0] (
05:15:28***Saving seen data "./dancer.seen"
05:18:37XavierGrMouser_X: I meant gbs at that time
05:18:46XavierGrMikachu was asking about gsf
05:19:26Mouser_XHCS is considering GBS.
05:19:34XavierGryes I know
05:19:35Mouser_XActually, he's said that might be his next project.
05:19:45*Mouser_X hopes for it.
05:19:49XavierGrthough it would be good to have PSF, but yeah it would be hard indeed
05:20:04Mouser_XHowdu HCS.
05:20:10Mouser_XYes, PSF would be difficult, at best.
05:20:26hcsI'd like to do gbs as it'd round out the 8-bit audio emulation support
05:24:46Mouser_XIt'd be nice if someone did VGM as well... I would think that'd be far easier than PSF support.
05:25:23Mouser_XI've still got a few GYM files, but I use VGM basically exclusively now.
05:25:55 Nick [Tesser] is now known as [Tesser-Away] (
05:51:42 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
06:11:34 Quit rotator ()
06:12:56 Quit Mouser_X (Nick collision from services.)
06:13:07 Join Mouser_X [0] (
06:15:43 Quit [Tesser-Away] (Read error: 110 (Connection timed out))
06:17:55 Quit xNibbler (Read error: 110 (Connection timed out))
06:18:23 Join xNibbler [0] (
06:18:29 Join [Tesser-Away] [0] (
06:22:23 Join Llorea1 [0] (
06:22:42 Quit Llorean (Nick collision from services.)
06:22:45 Nick Llorea1 is now known as Llorean (
06:24:04 Nick [Tesser-Away] is now known as [Tesser] (
06:24:43 Nick [Tesser] is now known as [Tesser-Away] (
06:25:10 Part [Tesser-Away] ("Leaving")
06:25:29 Quit Mouser_X (Nick collision from services.)
06:25:38 Join Mouser_X [0] (
06:36:01 Part safetydan
06:40:07 Quit Mouser_X (Nick collision from services.)
06:40:14 Join Mouser_X [0] (
06:54:28 Quit Mouser_X (Nick collision from services.)
06:54:37 Join Mouser_X [0] (
07:00:14 Join decayedcell [0] (
07:08:26 Quit Mouser_X (Nick collision from services.)
07:08:37 Join Mouser_X [0] (
07:15:31***Saving seen data "./dancer.seen"
07:22:22 Part voltagex
07:22:52 Quit toffe ("MegaIRC v3.27")
07:58:14 Join daurnimator [0] (n=fake@unaffiliated/daurnimator)
07:59:53amiconnpreglow: Unfortunately your sw tone control patch no longer applies :(
08:01:25amiconnBagder: Did you spot the errors at 23:51 and adjust the build server capabilities accordingly, or is it just by chance that all is well again?
08:02:38amiconnOh, the last 5 builds were run on 4 build servers only?!?
08:05:35 Quit Mouser_X (Read error: 110 (Connection timed out))
08:09:45 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
08:09:57midkayJdGordon: seems ready? :)
08:10:04midkayi better try it out. :)
08:11:02amiconnLinusN: Good morning :) Do you know what's wrong with the build system?
08:12:52LinusNamiconn: is anything wrong?
08:13:28LinusNhmmm, only 4 servers
08:14:57amiconnYes, after the 23:51 build had strange effects: didn't find m68k-elf-gcc, and used a stone-age sh-elf-gcc
08:15:30amiconnThe size difference for fm recorder was quite shocking
08:17:02 Quit bzavala ("CGI:IRC")
08:17:16 Join Mouser_X [0] (
08:19:00LinusNwell, we have little control over the build servers out there
08:20:48 Join daurn [0] (n=fake@
08:29:22 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
08:36:28 Quit daurnimator (Connection timed out)
08:39:16 Quit decayedcell (Read error: 104 (Connection reset by peer))
08:39:22 Quit daurn (Read error: 110 (Connection timed out))
08:41:31 Join markun_ [0] (n=markun@rockbox/developer/markun)
08:42:20 Join Xerion [0] (
08:42:31 Quit markun (Read error: 104 (Connection reset by peer))
08:43:33 Quit Xerion_ (Read error: 104 (Connection reset by peer))
08:45:57 Quit Mouser_X (Read error: 110 (Connection timed out))
08:47:12 Join miepchen^schlaf [0] (
08:48:59 Join Rob222241 [0] (
08:53:50 Join kubiix [0] (
08:54:00 Quit kubiix (Client Quit)
08:54:23 Join ender` [0] (
08:56:51 Join fejfighter [0] (
09:05:09 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
09:05:46 Quit Rob2222 (Read error: 110 (Connection timed out))
09:15:35***Saving seen data "./dancer.seen"
09:25:07 Join Vyrus001 [0] (
09:25:21 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:25:27 Join BHSPitMonkey_ [0] (
09:25:43 Quit GodEater (Client Quit)
09:25:46 Join BHSPitLappy_ [0] (
09:26:08 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:26:43 Quit BHSPitMonkey (Nick collision from services.)
09:27:08 Quit BHSPitLappy (Nick collision from services.)
09:27:17 Nick BHSPitMonkey_ is now known as BHSPitMonkey (
09:27:34 Nick BHSPitLappy_ is now known as BHSPitLappy (
09:29:46 Join Winchester [0] (
09:29:51 Quit Everybody|away ("( :: NoNameScript 4.02 :: )")
09:36:24 Join decayedcell [0] (
09:40:36 Join pondlife [0] (
09:44:26 Quit safetydan (Remote closed the connection)
09:47:47 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
09:48:07 Join miepchen^schlaf [0] (
09:56:11*GodEater watches some tumbleweed blow through the channel
09:56:36 Quit GodEater ("CGI:IRC")
09:56:51 Join Everybody [0] (
09:57:01 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
09:58:58 Part decayedcell
10:00:29 Join tick [0] (i=c27f0812@gateway/web/cgi-irc/
10:01:04tickBagder: any thoughts/opinion about the feature matrix (see yesterday's log at 19:15:09)
10:06:18 Join decayedcell [0] (
10:07:25tickhmm... no traffic today
10:11:02 Quit midkay ("Leaving")
10:14:54pondlifeHmm, first 7 build servers go missing, then most of the IRC users...
10:16:09LinusNthe 'net has a bad day today :-)
10:16:41pondlifeAs long as it's not some kind of abduction thing.
10:16:57 Join wooo [0] (
10:16:58 Quit BigBambi ("Leaving")
10:17:36LinusNpondlife: i'm safe, my foil hat is on
10:17:43pondlifeWould anyone else like to see the size delta table go back much further (maybe on the daily builds)?
10:18:02pondlife"Keeps out the alpha waves, you don't get old."
10:18:21*pondlife waits for foil to come into fashion.
10:18:35tickLinusN: any thoughts/opinion about the feature matrix (see yesterday's log at 19:15:09)
10:18:44*LinusN looks
10:18:46tickSorry for nagging...
10:18:50 Join midkay [0] (n=midkay@rockbox/developer/midkay)
10:19:35 Quit fejfighter (Read error: 104 (Connection reset by peer))
10:20:34 Quit BHSPitLappy (Read error: 104 (Connection reset by peer))
10:20:48pondlifeWhy is #ifdef being replaced by #if?
10:21:03LinusNi'm not sure
10:21:15 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
10:21:24LinusNtick: interesting idea, let me think about that
10:21:26hcswe're enabling warnings for #if SOME_UNDEFINED_THING
10:21:52 Quit Shiiny ()
10:21:56pondlifeOK, but #ifdef wouldn't trigger a warning would it?
10:22:00LinusNhcs: so it warns even for #ifdef?
10:22:07LinusNsound odd
10:22:15Llorean#ifdef doesn't trigger a warning.
10:22:16 Quit Winchester ("Download Gaim:")
10:22:21hcser, crap, I misread
10:22:26LinusNso why are we replacing it?
10:22:34LloreanBut the #ifdefs are only being used for the HAVE_ things, right?
10:22:35hcsin that case I don't know
10:22:47pondlifeNot a biggie, but I find #ifdef more readable than #if defined..
10:22:59Lloreanpondlife: #if defined() shouldn't be used.
10:23:03 Join Vyrus [0] (
10:23:14LloreanThe ones being checked with #if have a value of 0 when they're not supposed to be true for that target.
10:23:26LloreanThat way if they haven't been defined, then you get the warning.
10:24:24LloreanIf my understanding is correct over the last few days
10:24:26tickLinusN, hcs: IMHO it's good if you don't have to guess what you should use: #if or #ifdef. #if is a more general mean hence we should only use it. But that requires to assign values to symbols.
10:24:26pondlifeAha, to avoid CONFIC_CODEC?
10:24:34LloreanBasically, yes.
10:24:56 Join Mouser_X [0] (
10:25:48 Join bluebrother [0] (i=2BC1r78a@rockbox/staff/bluebrother)
10:26:01LinusNi tried to build a h120 bootloader yesterday, and it was too big to fit :-(
10:26:10LloreanThe bootloader is too big?
10:26:13LloreanHow'd that happen?
10:26:47pondlifeThe H300 bootloader has been too big for quite a while, hasn't it?
10:27:11LinusNyes, but now the h120 is too big as well
10:27:29LloreanWhat's been causing the bootloader to grow?
10:27:40LinusNthe eeprom stuff among others
10:27:47pondlifeLlorean: Am I right to think that HAVE_... is used for features and CONFIG_... for hardware?
10:28:04Lloreanpondlife: I honestly don't really know.
10:28:17*midkay thinks HAVE_LCD_COLOR
10:28:27pondlifeIt's a bit odd. HAVE_RTC_RAM too.
10:28:31safetydanI think it's just HAVE_ is bool and CONFIG_ is multivalued
10:28:39midkaythat makes sense.
10:28:39safetydanor at least most of them work that way
10:28:46tickpondlife, Llorean: you wouldn't have such conversation if only #if's were used
10:28:51midkayplus HAVE sounds better sometimes, i think is another factor ;)
10:29:00pondlifeHAVE_ is still used with #ifdef then.
10:29:08LinusNshould be, yes
10:29:09Lloreantick: Look, we get your point, you've said it before, repeating it isn't going to magically make it any more or less valid just by saying it over and over.
10:29:23tick#ifdef means that something is considered to be default. But for example, is HW_CODEC or SW_CODEC default?
10:29:27LloreanYou spent quite some time trying to convince me yesterday already
10:29:53LinusNtick: that was a bad example
10:30:29 Quit Vyrus001 (Read error: 60 (Operation timed out))
10:30:31tickLlorean: ok. But nobody said anything to it, not even 'I don't get it'. I'll be more quiet now.
10:30:32LinusNtick: since SW_CODEC is not supposed to be tested with #ifdef
10:31:20Lloreantick: Yesterday I said that I thought it was unnecessary to include every single feature in every single config file.
10:35:03tickLlorean: yes, but that was (IIRC) before I suggested to do in a systematic way, i.e. with a feature table
10:35:41LinusNwhat problems will this feature table solve?
10:36:22peturit will wipe half of the FS bugreports </sarcastic>
10:37:34tickLinusN: it would serve two puposes. 1. An overview of all supported platforms along with their features and 2. all features would be tested with #if since every feature symbol would have a meaningful value (even if that values means 'feature is not present')
10:38:06LinusNtick: and why would #2 be a good thing?
10:38:29BagderI think #2 is a bad idea
10:38:39peturchange #ifdef hell into #if hell
10:38:53tickLinusN: it would make thing more consistent and simple. Only IMO as I see :-)
10:39:00Bagderand also force all targets to have to be updated as soon as one gets a new feature
10:39:24LinusNi don't see how it would be simpler with #if
10:39:27Bagderor have a default "no feature" define, when then defeats the idea of using #if
10:39:54Mikachui think the idea was that the target-specific .h files define stuff, then config.h has some ifndef + define 0
10:40:05Bagderbut why?
10:40:15Mikachuso gcc warns if you typo the symbol name in .c files
10:40:40LinusNMikachu: to catch #ifdef MISSPELED ?
10:40:46tickLinusN: maybe not simpler but more consistent. Now, if you want to test for a feature you can't say for sure how it should be tested: with #if or with #ifdef. But I admit that the appropriate compiler flags might solve the problem quite well.
10:41:01BagderI'm against it since it is not how these things are usually done
10:41:04MikachuLinusN: it won't catch #ifdef typo, but #if MISSPELED
10:41:14LinusNof course
10:41:26Bagderand we've not so far had suffered from #ifdef MISSPELLED problems, afaik
10:41:30 Nick lex_ is now known as lex (
10:41:43Bagdersince they are often quite noticable
10:41:53Mikachuthere was an #if TYPOE = FOO that triggered all this
10:42:00Mikachu== maybe
10:42:00Bagderyes, those we've had
10:42:15Bagderso we should still do so for multi-value defines
10:42:44 Join libuxstb_ [0] (i=5343d4aa@gateway/web/cgi-irc/
10:42:44LinusNbtw, i think #ifdef for includes is quite unnecessary
10:43:04Bagderyeah, those are unfortunate cygwin work-arounds
10:43:07LinusNi.e conditionally include a file
10:43:21Mikachubetter to have the ifdef in the include then
10:43:39Mikachuunless it's a <> include
10:43:42tickOk, if converting everything to #if is discarded then we probably shold document (a wiki page?) all the symbols and their meaning and how they should be tested
10:43:48LinusNi guess it's there to improve cygwin build times
10:43:55libuxstb_tick: You do now know for sure (or should,if things are correct), whether to use #if or #ifdef - you use #if CONFIG_XXX and #ifdef HAVE_XXXX
10:44:25Bagdertick: I don't think we have a problem with this but sure, if you feel like writing a wiki page so feel free
10:44:25 Nick Everybody is now known as Everybody|ingame (
10:46:31ticklinuxstb: that was my point. Since it doesn't seem to be accepted I take it back (me waiting to be poked by llorean :-)
10:47:28Lloreantick: Linuxstb is saying that the way things work now is fine, assuming things are defined properly currently. That wasn't exactly your point...
10:48:37 Quit libuxstb_ ("CGI:IRC")
10:48:48 Join daurnimator [0] (n=fake@
10:49:15Lloreantick: You were suggesting checking them all with #if, while he was saying it's clear which ones should be checked with which, right?
10:50:12JdGordonLinusN: you mentioned having to #if around includes... some are needed because the include isnt in the path for certain builds
10:50:34JdGordonI cant remember which tho :p I tihnk possibly background.h or something
10:50:49BagderJdGordon: no, they are #ifdefs ;-=
10:50:50 Join Ribs [0] (n=ribs@
10:51:08JdGordon.... that too
10:51:14Bagderbut most of them are there for cygwin
10:51:28Bagdersince it is so painfully slow
10:51:37Bagderwhen including things that might not be needed
10:51:42JdGordonah alrighty
10:51:51JdGordonso dump cygwin :)
10:51:58Mikachuincluding a nonexisting file isn't an error, i think
10:52:05Mikachubut it's not nice of course
10:52:13*petur gives JdGordon an angry look
10:52:18tickLlorean: ehh... sorry, I've read 'you do NOT know for sure' :-))
10:53:25Lloreantick: Alright then. Just making sure we all know which side we're all on. ;)
10:56:16tick Has anyone counted the total number of HAVE_xxx symbols that are used in RB?
10:56:41tickJust out of coriousity
10:56:43BagderI don't think so
10:57:06Bagderwe do have a significant amount of imported code too, that I'm sure add a bunch
10:57:11 Join Nico_P [0] (
10:57:43tickBagder: but I think that code doesn'r rely on RB's symbols
10:57:58Bagderno, but they are in the rockbox source tarball and are #ifdefed
10:58:36BagderI'm just saying, so if someone _does_ count it should be taken into account
10:58:36tickI meant only RB's "own" symbols
11:00:22tickAnother question (I can't use grep right now): is #ifndef used somewhere? That would reverse the handling of defaults
11:00:34BagderI'm sure it is
11:00:49 Quit Mouser_X (Read error: 110 (Connection timed out))
11:01:55Bagderfind . -name "*.[ch]" | xargs grep "^#ifndef" | wc -l
11:02:13Bagder(again, that's a lot of codecs and stuff too)
11:05:38midkayanybody know what's the real way to get $Id: to work right? for example on a newly-committed file.
11:05:42Bagderfind . -name "*.[ch]" | xargs grep "^ *#if" | wc -l
11:05:52 Join barrywardell [0] (
11:07:34tick bagder: would it be easy to count the #ifndef's only in the RB code, e.g. in firmware and apps?
11:08:19Bagderwell, do you count plugins?
11:08:35tickI.e. how do we undertsnad the world? Nothing is present unless a symbol is defined? And only then is some code produced (gets compiled)?
11:08:56Bagderfind . -name "*.[ch]" | grep -v codecs | xargs grep "^ *#ifndef" | wc -l
11:09:12tickOr do we deliberately produce code even if a feature is absent (symbol is not defined)?
11:09:24Bagderwe define features/things that are that exist, yes
11:09:42Bagdersome stuff then need to be done differently if a specific feature does not exist
11:09:49Bagderhence #ifndef
11:10:43bluebrothermidkay: you need to set the svn:keyword property
11:10:51tickIMHO the latter case should be tested with #if. Or is #if used only for (potentially) multi valued symbols whereas #ifdef for boolean?
11:11:04midkaybluebrother: hm? what's that?
11:11:29 Join robin0800 [0] (
11:11:44Bagdertick: since we don't define them when missing, we must use #ifdef
11:11:51Bagderor #ifndef
11:13:02Bagderthis is how C programs work all over and have done for ages
11:13:20tickSo #ifdef is used for boolean, right? And we should not have something like #if XXX but only #if XXX == yyy, right?
11:13:48bluebrothermidkay: you need to svn propset svn:keyword "Author Date Id Revision" <file>
11:14:10Bagderalthough "#if XXX" implies "#if XXX != 0" which can be meaningful
11:14:14bluebrotheras that is stored in a property along with a file
11:14:28midkaybluebrother: interesting, i'll have to look into that, thanks.
11:14:51bluebrothercheck out the svn book −− the properties concept is quite interesting.
11:15:08tickOk, I got it now. I mean the logic behind the symbol usage in RB. #ifdef for booleans (or what would be boolean if programmed in C), #if for multivalued
11:15:28Mikachuactually there aren't booleans in c :)
11:15:36***Saving seen data "./dancer.seen"
11:15:44 Join BigBambi [0] (
11:15:50BagderMikachu: there are in C99
11:15:52tickC is not the best language to explain that since there's no proper boolean
11:16:09tickMikachu: you type faster than I
11:16:16 Quit wooo (Read error: 110 (Connection timed out))
11:16:34Bagderthere is "proper" booleans in C
11:16:48Bagderbut we speak preprocessor here, which is slightly different
11:17:12tickBagder: oh? Proper bool's in C?!
11:17:50tickBagder: since when? And how are they defined/used?
11:17:52Bagderthe standard is called C99
11:18:05bluebrotherC99 has some really nice extensions
11:18:22Bagderit does
11:18:27tickBagder: ah, ok. I don't know that. Only the "old good standard"
11:18:30BagderI love the named struct fields the best
11:18:36bluebrotherlike variable length argument lists for preprocessor macros
11:18:39 Quit linuxstb (Read error: 113 (No route to host))
11:18:41Bagderwhen initializing them
11:18:55*bluebrother likes that too
11:20:07 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
11:22:19tickThere hasn't been much talk about the root menu recently. Has everyone agreed on how it should work and it gets implemented now?
11:22:51decayedcellI think JdGordon has just gone with linuxstb's suggestion
11:23:04BagderI'll just whine afterwards ;-P
11:23:07JdGordonpretty much
11:23:58JdGordonbluebrother: getting resuming in the DB to work would add a fair bit of "safety" code to stop crashes which I dont think is worth it
11:24:01JdGordonbut it can be done..
11:24:22*decayedcell waits for Slasheri's GUI optimisation commit :P
11:25:03midkaydecayedcell: which?
11:25:44decayedcellhe said something about fixing the scroll lag issues on certain targets such as the iPod 5G by putting a thread on the cop
11:26:47decayedcellbut he wasn't commiting yet because he didn't know if his current implmentation was the most efficient
11:27:22 Quit atsea-196 ("using sirc version 2.211")
11:27:43 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
11:27:59 Quit tick ("CGI:IRC")
11:29:34 Join ZeroOrDie [0] (
11:35:03bluebrotherJdGordon: I usually use the file browser, so not a big deal to me
11:35:38bluebrotherbut where did the startup screen setting go? Or was I just missing the entry?
11:35:42LinusNwhat is the issue with db resume anyway, does it work correctly in svn?
11:36:37bluebrotherLinusN: resume in terms of you keep the position in the db tree view when switching wps - tree view
11:36:55bluebrotherthat works without the root menu patch
11:37:22bluebrotherand I guess a lot of people will whine about it missing with the patch
11:37:38LinusNhow come the root menu patch kills it then?
11:38:05bluebrotherask JdGordon. According to his last comment on the patch to prevent lockups
11:38:58LinusNi think we can't commit it if it breaks the db browsing experience
11:39:03midkaydecayedcell: cool, i wish i could try that..
11:39:06 Join atsea- [0] (i=ariel@gateway/tor/x-d0a0bba69aaa6d51)
11:39:17decayedcellmidkay me too haha
11:39:20bluebrotherit works fine with the file browser.
11:39:33midkaywe need coprocessor support in svn :D
11:39:33bluebrotherbut I agree that it should still work with the root menu
11:40:02LinusNthe root menu is a controversial thing, and we don't want to give a bad first impression of it
11:40:29decayedcellah yes how close is the cop patch getting to into SVN? dan_a claims to have fixed the recording freeze
11:41:13LinusNso when we commit the root menu, it should work as expected, otherwise we will get a whole lot of whining
11:43:54 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
11:44:39linuxstb_decayedcell: Are you sure Slasheri was using the COP with his GUI optimisations? I've only read that he, just that he was minimising screen redrawing, and skipping frames.
11:44:54linuxstb_(plus some separate work on clickwheel acceleration)
11:45:37Slasheriyes, i was not using cop
11:46:25Slasherihmm, need to commit those optimisations soon..
11:47:50 Quit linuxstb (Read error: 60 (Operation timed out))
11:47:51bluebrotherI also think the root menu should stop playback with the stop button −− the file browser did that before
11:48:14bluebrother(unless the stop button is needed for the menu due to lack of buttons, of course)
11:48:24LinusNof course
11:49:33 Nick Everybody|ingame is now known as Everybody|away (
11:50:34amiconnSlasheri: If you're talking about the wheel stuff - for m it feels significantly worse than cvs
11:50:53amiconnNavigation gets really jumpy and imprecise when moving faster
11:51:14Slasheriamiconn: with the latest version it should be better
11:51:17 Join PaulJam [0] (
11:51:31Slasheribut it still has some issues to be fixed before committing that
11:51:44Slasherii will commit the gui optimizations first
11:52:31Slasheriamiconn: btw, when moving faster, shouldn't it be imprecise? with _fast_ speeds no one can be that precise =)
11:53:26amiconnWell, the problem is that the bar jumps around the screen even when scrolling in just one direction, and that makes it feel very imprecise
11:53:45Slasherinow it no longer jumps
11:53:51amiconnI don''t know whether that was the latest version though; didn't check that patch for a while
11:53:55Slasheriit keeps at the top or bottom
11:54:04amiconn...simply because I don't have any need for it on my mini
11:54:10Slasheriand just the list jumps as necessary
12:02:24LloreanDoes this have a tracker entry?
12:03:29 Quit ZeroOrDie ("[ HIX-Script v2.2 ]:::[ Download from ]:::[ ]::")
12:04:23decayedcellLlorean don't think so
12:15:46pondlifeNot meaning to be churlish, but what's the reason for the root menu? I don't see it adds anything useful, apart from fixing some keymap inconsistencies maybe...
12:15:54*pondlife dons flame suitg
12:16:52pixelmalinuxstb_: I tweaked your chopper patch a little (decreased cycle time for Archos screens again a bit - the other minor changes are about readabilty of the status info at the top of the screen)
12:17:01*decayedcell prepares flamethrower
12:17:58*safetydan curses vorbisfile.c for its unreadability
12:18:49Lloreanpondlife: It also makes it 100% clear what the 'home' screen is, and is much more of a device interface than a computer-like interface.
12:20:58Shaidis there a new layout for the root menu, or is it rows of text?
12:21:23*decayedcell suggests putting TWiki Rootmenu link in IRC Topic
12:21:26LloreanIt's still the normal menuing system.
12:23:04 Quit pearldiver (Read error: 60 (Operation timed out))
12:23:57linuxstb_pondlife: IMO it greatly helps usability - users can choose things from menus rather than having to know the correct button combination. There are also related changes which make buttons act more consistently.
12:24:51linuxstb_pondlife: But you can make Rockbox act very similar to the current way - you can set the "startup screen" to be the file browser, and there are still shortcuts between screens, such as NAVI/Select from WPS->Browser, and PLAY from browser back to WPS.
12:25:23linuxstb_And it makes both the file browser and database browser accessible without needing to change a setting.
12:25:49*pixelma has to try that with Ondio (especially the shortcut part)
12:26:40Lloreanlinuxstb_: Do you think we could put the "File View" modes on the context menu of the filetree entry in the root menu, and then may use that part of the quickscreen for something else, or do you think it needs to still be more accessible than that?
12:27:05 Join lee-qid [0] (
12:27:38linuxstb_The file view settings would be on the "file browser" context menu. I'm not sure if it should be replaced on the quick-menu though - what would be more important?
12:27:59 Quit directhex|work (Read error: 104 (Connection reset by peer))
12:28:29 Quit Ribs ("eh eh ehhhh!")
12:28:34 Join kaaloo [0] (
12:28:38Lloreanlinuxstb_: I'm honestly not sure off the top of my head, but with it becoming 'relatively' easier to get to that way, something might become so.
12:29:24LloreanThere may not be anything though.
12:29:39LloreanFile view has just almost always been, for me, something I set once and change on the rare occasion I need to get into the .rockbox folder.
12:30:14linuxstb_I would rather not think too far ahead - i.e. let the root menu embed itself for a while, and then see what needs to be changed.
12:30:31LloreanThat's fine.
12:31:11LloreanHonestly, I can't see _much_ of an alternative for it unless something like the right-to-insert patch gets accepted.
12:35:10pondlifeThanks guys, sorry - phone rang
12:35:35pondlifeI haven't tried the patch yet but will do shortly.
12:38:29pondlifeOn another topic, has anything low-level changed recently on the H300, maybe in the LCD updating? I've had 2 crashes today, both times with a part-scrolled filename on the WPS.
12:38:39pondlifei.e. top half of text scrolled
12:39:19pondlifeThis happened a while ago when the LCD timings were changed, but that got resolved. I better look into SVN...
12:40:09LinusNmaybe something broke when i updated the PLL code for the outletmp3 players?
12:41:24pondlifeI'll keep it running and report back. Wish I knew which version I was running previously ;)
12:42:31 Join Mouser_X [0] (
12:48:28 Quit safetydan ("Ex-Chat")
12:52:09 Part decayedcell
13:01:12pixelma~2500 bytes binary increase with the root menu patch for backlight modded Ondio... I thought it should flatten everything and make binary size smaller in the end (?)
13:03:06pixelmaas a side effect
13:03:23LinusNpixelma: iiuc, a lot of the browser code can be removed when the root menu is finished
13:03:42LinusNpixelma: but i don't think anyone has claimed that the root menu patch would reduce binary size
13:03:47 Join Thundercloud [0] (n=thunderc@
13:05:35peturor... JdGordon takes back what he gave first ;)
13:05:45pixelmawould be nice though
13:06:02preglowamiconn: ok, i'll do a new one
13:07:43midkaywoohoo.. *just finished bitmapping all the clock modes for Recorder*.. that mono LCD is a pain. :)
13:07:46midkay7seg mode:
13:08:08preglowLinusN: you wouldn't happen to be skilled with the pcf50605, would you?
13:08:42 Quit printfXh4 (Read error: 145 (Connection timed out))
13:08:59LinusNpreglow: still struggling with the alarm?
13:11:42LinusNhow far have you come? does it set the alarm bit?
13:12:02preglowi set the alarm time registers, i set the wake on rtc bit, i unmask the interrupt
13:12:17preglowand i've even seen the interrupt trigger
13:12:32preglowbut it seems the wakeup just doesn't happen when it's shut off
13:12:38LinusNok, i see
13:13:28LinusNwhich target?
13:13:33midkayquick question for anybody who knows about grayscale LCDs + rockbox... how should bitmaps be saved/formatted? saved as a grayscale BMP and it's automatically worked out to fit Rockbox's shades, or?
13:13:51 Join My_Sic [0] (
13:14:05LinusNyes, grayscale bmp
13:14:29pondlifeBagder: Any chance that the delta size table could be addedonto the daily builds page? It would be nice to see if things are really getting better...
13:14:56preglowLinusN: ipod nano
13:15:03midkayhm... photoshop asks whether it shall be saved as 4 bit or 8 bit?
13:15:37***Saving seen data "./dancer.seen"
13:15:45LinusNpreglow: could it be that it in fact triggers, but the nano bootloader is double-checking the buttons?
13:15:46Bagderpondlife: I haven't planned anything like that, no
13:15:58LinusNmidkay: 8-bit methinks
13:16:09pondlifeBagder: Probably not a simple bit of reuse... :(
13:16:29midkayLinusN: thanks, i'll try it.
13:16:32Bagderpondlife: correct, it would probably be better to write separate script to check the sizes of the dailies
13:16:42preglowLinusN: hmm, possibly, but that would be weird, since retailos also has an alarm feature
13:16:48LinusNi see
13:16:59preglowas a matter of fact, i'll just try that out to see it does what i want
13:17:23linuxstb_preglow: But retailos only sleeps...
13:17:29preglowalso a good point
13:17:31LinusNpreglow: it could be that the ipod can't wake up from the OFF state, only the sleep state
13:17:48Mikachuso if you set the alarm more than 24 hours in the future it doesn't work? that doesn't sound like a feature
13:18:06preglowi bet you can't do that
13:18:11linuxstb_Maybe if you have an alarm set, it doesn't leave sleep mode.
13:18:11preglowdoesn't it switch off more properly before 24 hours?
13:19:23LloreanI thought it was something like 72 hours before it moves from sleep to off.
13:19:24preglowbhargh, can't you set the alarm from retailos itself?
13:19:31linuxstb_Bagder: Build sizes on the daily builds would be useful as they are always built with the same compiler/binutils version - unless you're planning on distributing them.
13:19:42Mikachui've never seen any alarm function there...
13:19:51Bagderlinuxstb_: i agree, and no I don't have any such plans
13:20:02preglowfound it
13:20:08Mikachuwhere is it?
13:20:50preglowi'll just have to do some more testing to see that i can consistently trigger the interrupt
13:20:58preglowat least then there is just one more source of errors
13:21:28linuxstb_BTW, have people been reporting that with Rockbox installed, the original firmware on the 5g doesn't save its settings?
13:22:26Mikachuthe of doesn't save when you reset with menu+select
13:22:30Mikachuat least not on nano
13:22:40Mikachuyou have to wait some time, no idea how much
13:22:52linuxstb_Probably until it powers-off...
13:23:16preglowi re-read the alarm regs after retailos has set them, and they're not what i expect them to be
13:23:34midkaywait a sec, H100 has no RTC.. *slaps self*
13:23:52LinusNpreglow: they are supposed to be BCD, are you aware of that?
13:24:25preglowLinusN: yes
13:24:32preglowperhaps there's some silly bug there
13:24:47*preglow hates bcd
13:26:41pixelmamidkay: there are some H1x0s that have it now...
13:26:41LinusNmidkay: some h100's have it (RTC mod)
13:27:10midkayoh, cool.. well i still need to keep the gray bitmaps for ipods, so they'll automatically be used.. <3 the bitmap system. :)
13:27:16Bagder /echo off
13:27:46 Quit Mouser_X (Nick collision from services.)
13:27:55 Join Mouser_X [0] (
13:30:12 Quit PaulJam (".")
13:33:34w1ll14mmorning all
13:36:12 Quit kaaloo (Read error: 113 (No route to host))
13:36:55preglowwould someone PLEASE tell me the difference between logical and bitwise operators
13:36:58preglowit's obviously needed
13:37:02*preglow shoots forehead
13:38:03*LinusN shoots preglow
13:39:51*Bagder offers bullet-proof equipment to members of the channel
13:40:23*midkay shoots Bagder before anybody gets any
13:40:25 Join linuxstb__ [0] (
13:40:55 Nick linuxstb__ is now known as linuxstb (
13:41:09 Quit linuxstb_ (Read error: 110 (Connection timed out))
13:41:24 Join PaulJam [0] (
13:41:29*pixelma goes hiding from violence in the channel and blames doom support
13:42:04*LinusN shoots pixelma and reaches the next level
13:42:31amiconnmidkay: 4 bit bmp is sufficient, if your gfx app is able to save as 4-bit greyscale. bmp2rb can handle all standard formats, but some pc gfx are braindead
13:42:49amiconn(pc means (at least) both windows and linux here)
13:43:06LinusNthe bmp format is somewhat braindead
13:44:32amiconnWhat I mean is that some gfx apps insist on using the standard palette in conjunction with 4-bit bmp, although 4 bit supports arbitrary colours in the palette as normal
13:45:11amiconnSame goes for 1 bit bmps. Some programs insist on b&w, even when the palette e.g. specifies green & red
13:45:17 Join MonkeyTamer [0] (n=chatzill@
13:46:57preglowjhMikeS: i thought you said you unrolled the channel loop
13:47:08*bluebrother notices a huge amount of blood in this channel and calls the ambulance
13:47:25preglowi have alarm!
13:47:32 Quit GodEater ("CGI:IRC (Ping timeout)")
13:47:51*w1ll14m rolls on the floor laughing his ass off
13:48:03preglowLinusN: it was, of course, me messing up with bcd and some rather nasty operator precendence :>
13:48:15*preglow hands w1ll14m his ass back
13:48:40LinusNpreglow: congrats
13:48:44*w1ll14m takes some super glue, and glued it back on
13:48:56preglowtrivial problem made into one requiring hours, courtesy of me
13:50:03midkayamiconn: gotcha.. 8 bit worked so i'll probably just leave it there.
13:50:13 Quit MonkeyTamer (Client Quit)
13:51:00Lloreanpreglow: The important thing being that it works now, though.
13:52:11preglowsure, hadn't it been for that, i surely would be tying myself a snug noose as we speak
13:52:31Shaideh, I vaguely recall logical operators only evaluate things if they have to
13:52:40Shaidwhile bitwise ones will evaluate the whole thing no matter what
13:52:46Shaidor vise-versa.
13:52:51*Shaid shrugs.
13:53:11LloreanI can't remember which is which.
13:53:31LloreanBut one only evaluates one side, and if the condition is false, doesn't bother with the other, right?
13:53:33preglowalso, they don't really end up as the same assembler
13:53:46preglowanyway, i know the difference, my fingers just don't care that i do and write funky stuff here and these
13:53:54preglowand my eyes refuse to pick it up
13:54:43LloreanI don't think I've ever thought "Should I be using a bitwise operator to do this better?"
13:54:58ShaidYou do if you want to do it on a bit basis.
13:55:08jhMikeSpreglow: initially, but faster with it rolled :)
13:57:04Shaidand I remember bitwise not (~) and logical not (!) working differently again.
13:57:11ShaidI should do a C refresher class.
13:59:18*w1ll14m thinks that google is way better than class
13:59:31Shaidgoogle is my friend, but sometimes he annoys me.
13:59:40w1ll14mtrue, true :)
14:00:23*bluebrother suggests "C in a nutshell"
14:01:26 Join Obsys [0] (
14:05:41linuxstbamiconn: (or anyone else) Can you have a quick look at my chopper-on-archos patch and let me know if you think it's commit-worthy? Ideally the whole of chopper needs making more LCD-size independent, and my patch doesn't do that, it just scales everything by 50% for the 112x64 LCDs -
14:06:05pixelmalinuxstb: seen my comment about it?
14:06:42linuxstbpixelma: Yes, thanks for fixing those problems. How is it now from a usability viewpoint?
14:06:59preglowjhMikeS: really now
14:07:27 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
14:07:29preglowjhMikeS: anywho, i'll about starting the arm asm today, i've just accumulated a couple of other rockbox things to do as well
14:09:13jhMikeSpreglow: I'm mac'ed with things to do myself. Saw another optimization in the upsampling but I guess it can wait a few days.
14:09:21 Join web-taz [0] (
14:10:07pixelmalinuxstb: I think it's quite ok... for the info it uses DRMODE_COMPLEMENT now so that it is readable (though it looks a bit broken sometimes but you can always see it - the only thing is that I can't test on a colour target, looks right in a sim)
14:10:12 Join kaaloo [0] (
14:10:21pixelmaI could provide the patch too
14:10:41 Part kaaloo
14:11:15linuxstbThanks. I can't think of a better way to scale it properly for the Archos, so I'll probably commit it as it is. I also don't want to spend any more time on it...
14:11:57 Nick HardDisk_WP is now known as HD|NA (n=Marco@wikipedia/harddisk)
14:12:27linuxstbI'll test it on a colour target and check it's OK.
14:13:53pixelmasure - I thought the status was a bit too short - so I chose "Dist" and "Hi" for it and I noticed that the full version was a bit too long for the small H10 too, so I let it check for LCD_WIDTH <= 128 now (only for the text)
14:14:20web-tazoh srxy
14:14:31web-tazwrong chan...
14:14:44pixelmame wants some ;)
14:17:17 Join juxtap [0] (
14:22:18 Quit Mouser_X (Read error: 110 (Connection timed out))
14:25:06 Join kaaloo [0] (
14:26:57 Quit My_Sic (Read error: 104 (Connection reset by peer))
14:28:57amiconnlinuxstb, pixelma: I think it's better to keep white text for colur targets, set it to black for greyscale and only use DRMODE_COMPLEMENT for monochrome
14:28:58*SaLoMoN slaps web-taz around a bit with a fat walrus.
14:30:33*web-taz whips SaLoMoN with a mouse cord
14:30:47ShaidGet a room, you two.
14:31:03*petur wonders if the above can be blamed on global warming
14:31:40LinusNJdGordon: what makes the db resume so tricky?
14:32:22JdGordonbecause it doesnt like going between files and db without restarting at root
14:32:28PaulJamis there a known bug, that the pcm buffer stays really low (at ~9% of the available buffer) and it is not managing to fill the codec buffer once it is enpty? or should i try if i can reproduce it with a svn build?
14:32:44 Quit Thundercloud (Remote closed the connection)
14:33:08*JdGordon thinks Llorean is going to kill PaulJam one day :D
14:33:13pixelmalinuxstb: guess amiconn is right... if you take my patch as basis you chould also revert my change in line 834 back to LCD_DARKGRAY
14:33:13 Join Thundercloud [0] (n=thunderc@
14:33:34JdGordonPaulJam: its possible its staying in low latency mode...
14:33:44linuxstbpixelma, amiconn: OK, unless either of you want to commit it, I'll look at it this evening.
14:35:12JdGordonLinusN: its also possible im over-complicating the whole thing... Im going to redo the whole tree loading bit tomroow
14:35:31*JdGordon has to be awake in 6 hours, so goes to bed
14:35:34 Quit JdGordon ("Leaving.")
14:36:47*GodEater waves his latest forum post ( at any artists currently paying attention.
14:37:54LloreanGodEater: Can you go in and hand-edit one circle-on-magenta then copy and paste it?
14:38:24bluebrotherGodEater: convert it to 1bit bw, and then make one color transparent
14:38:34bluebrotherand color the circles green afterwards
14:38:52*GodEater goes to try that
14:40:12 Quit GodEater ("CGI:IRC")
14:40:25 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
14:41:13GodEaterIf I do 1bpp, I'll lose any sort of aliasing though right?
14:41:18 Join fasmaie [0] (
14:41:42midkayyou mean antialiasing, and yes.
14:42:32midkayyou should be able to redo the circles sans anti-aliasing...
14:42:42midkaygimp should have a checkbox or somesuch to enable/disable AA./
14:43:00GodEaterbut I want them anti-aliased - just with a white background, not a magenta one
14:43:16pixelmaGodEater: you wouldn't have real antialiasing anyways (like I understand your second image)
14:43:33midkaythat won't work.
14:43:57midkaybecause anti-aliasing will "blend" the green with the 255,0,255 pink transparent background..
14:44:01Shaidyou can't anti-alias onto transparent very easily.
14:44:13midkayand only the 255,0,255 parts will be removed, and the blended pixels will remain.
14:44:19GodEaterShaid: I know that's what I've found
14:44:21midkayeven if they're *very* close to the transparency color.
14:44:38midkayit's either antialiasing and a solid background, or a transparent background and no antialiasing.
14:44:40Shaidyou might be able to do it with varying alpha levels
14:44:46Shaidbut not easy.
14:44:57GodEatermidkay: correct - thats why I want them green circles to be anti-aliased against white in my source image, and then have those pixels left that way when I paste it over a magenta background
14:44:59midkayShaid: varying alpha levels? what do you mean?
14:45:01bluebrotherI don't think rb supports more than 1bit alpha
14:45:17GodEaterit doesn't
14:45:34Shaidinstead of blending colour, you blend alpha.
14:45:42ShaidIt's late at night here, I could be talking crap.
14:45:43GodEaterthat's why I'm trying to fake it
14:45:49pixelmaGodEater: did you already finish the progress bar on white?
14:45:51midkayShaid: won't work.
14:46:10GodEaterpixelma - yes - but it doesn't look quite right
14:46:21*Bagder learned the phrase "chroma key" for the transparency concept Rockbox uses
14:46:26midkayGodEater: oh, ok.. well.. you'll need to do a bit of work for that. create a green antialiased circle against white, then just somehow cut it out how you need (magic wand tool?) and put it on top of the magenta like that.
14:46:31Bagderas it isn't an alpha
14:47:01pixelmaGodEater: don't know gimp but in other graphic programs you could just change the clear white to pink...
14:47:28midkaythat's another way.. create the circles on a white background then use the paint bucket tool to fill in the pure white parts with pink.
14:47:46GodEatermidkay: that's the problem I think
14:47:54*linuxstb notices ipodwizard are distributing both ipodpatcher.exe and a hacked version of the Apple firmware, with a "YOU MUST NOT REDISTRIBUTE" statement...
14:48:07GodEaterGIMP is actually taking the edge pixels as have an alpha channel it seems, rather than an actual colour
14:48:15Bagderlinuxstb: !
14:48:17LloreanDoes GIMP leave the circles as vector or does it immediately raster them so that they're essentially 'done' with antialiasing effects at that point?
14:48:41LloreanGodEater: Do the circles on white, actually save it as a 24-bit bitmap, load the bitmap, and then fill pink?
14:48:46GodEaterLlorean: they're rastered, but as I say - I think the it keeps an alpha channel with them
14:48:55GodEaterhmm - that could owrk
14:49:04linuxstb (seems to be a kind of side-project from the ipodwizard people)
14:49:09*petur advises MSPaint ;)
14:49:34midkayGodEater: then just save it out as a bmp, and then re-open it, do the paint bucketing, and voila.
14:49:38pixelmamidkay, GodEater: I meant some programs can replace a colour with another
14:50:15midkaypixelma: right.. both work :)
14:50:19midkaymany ways you could do it.
14:50:33Lloreanlinuxstb: I thought the new background changing was a feature of iPodWizard itself?
14:50:36pixelmaok... thought that would be faster :)
14:50:57Lloreanlinuxstb: I'd been led to believe that since they couldn't change the background, they figured out how to extend the status bar, which they could skin, to full screen.
14:51:12GodEatermspaint sadly doesn't run on linux ;)
14:51:18Shaiddoes in wine
14:51:45midkaypixelma: if you know where it is and how to use it, yeah, it would be :)
14:52:13midkayGodEater: create your image of green circles on white, save out as a BMP. close it. open the BMP, use the paint bucket or color replacer.
14:52:19midkaythat will do it.
14:52:19linuxstbLlorean: It seems to require a hacked version of the firmware. My guess is they've managed to hook in a function that displays a background, rather than clearing the background to all-white.
14:52:38linuxstbLlorean: Seems to be a lot of skill wasted there...
14:52:49GodEaterLlorean: that seems to have done the trick ;)
14:53:36Lloreanlinuxstb: It seems an awful lot of effort goes into making things shiny.
14:54:01GodEaterwitness compiz / beryl
14:54:27 Join webtaz [0] (
14:54:54*midkay thinks less shiny and nice looking, more toy/demolike
14:55:03midkayfor compiz et al.
14:57:21 Quit fasmaie ()
14:59:01 Quit himitsu_ (Read error: 111 (Connection refused))
15:03:13preglowlinuxstb: should the drivers pcf50605.c file be target treeified?
15:04:00 Join fasmaie [0] (n=yohann@
15:05:10 Quit PaulJam (".")
15:07:25 Quit fasmaie (Remote closed the connection)
15:07:33amiconnpreglow: No it shouldn't
15:07:55preglowamiconn: it already is for x5, why not for ipod then?
15:08:09amiconnThat is, if there are target dependent parts how to control this chip, these parts should be in target tree
15:08:24amiconnThe chip-dependent part should stay in firmware/drivers/
15:08:27preglowthe whole file is target dependent
15:08:58amiconnThen that's a mistake
15:09:17amiconn(afaiu the target tree concept)
15:09:29 Join himitsu [0] (n=himitsu@
15:09:32preglowi'll see if i can be bothered to fix it
15:09:33LinusNamiconn: yes?
15:09:44 Quit web-taz (Read error: 110 (Connection timed out))
15:10:38amiconnThe target tree questions above...
15:11:38LinusNi think the pcfxxxx.c files should be split in one pcfxxx.c part which deals with the register writes etc, and a pmu.c file which deals with the target specific stuff
15:12:48LinusNi think we could benefit from a generic i2c driver layer
15:13:27LinusNin the spirit of the generic i2c stuff i wrote once
15:13:51LinusNwhere you map device ID's to a lowlevel driver
15:15:24LinusNthe pcfxxx driver would then only be a tiny layer above the i2c driver layer
15:15:40***Saving seen data "./dancer.seen"
15:16:10LinusNand pmu.c would contain all the target specific stuff, like setting up voltages, handling ADC and charging etc
15:16:19amiconnThe generic i2c stuff is way too slow, unfortunately
15:16:37LinusNi wasn't talking about that part of the generic driver
15:17:10LinusNi meant the idea to map device id's to a low level driver, being either asm bitbang or dedicated i2c controller
15:17:18LinusNso you have a common api
15:18:17LinusNthen pcfxxxx.c would be target independent, with calls to the generic i2c driver
15:19:10LinusNand also the uda, tlv and eeprom drivers
15:19:32preglowwould be sweet
15:19:45LinusNmy generic i2c concept had that mapping
15:20:19LinusN(together with the slow bitbanging stuff)
15:20:21preglowshould i combined the two rtc_pcf* files in drivers/rtc, btw?
15:20:27preglowthey do the same stuff, basically
15:21:22preglowalso, i should complete the ipod i2c driver
15:21:29preglowlots of crufty ipod code around
15:24:02preglowLinusN: and obvious benefit would be x5 getting rtc alarm support too
15:24:24LinusNand the h300, when the bootloader supports it
15:24:48*petur would love to use his h300 as alarm clock when traveling :D
15:25:02petur(soon to be H380 :) )
15:25:11*LinusN feels guilty for not working on the h300 bootloader
15:25:21 Join Kasperle [0] (i=kasperle@
15:25:58*petur is affraid to ask LinusN again about tracing ID and VBUS for usbotg
15:26:16preglowi just look forward to using my nano as an alarm clock
15:26:17*LinusN feels guilty for that too :-)
15:26:24BagderLinusN: we offer you absolution for your sins, just pay $199 to the rockbox fund and say 3 ave Maria
15:26:29preglowno more bLOODY beeping and me jumping around
15:26:33LinusNBagder: lol
15:26:54petursome beers will do too ;)
15:26:55preglowtwenty pater nosters too
15:27:02LinusNand 100 pushups
15:27:09 Join ackbahr [0] (
15:27:12 Quit ackbahr (Read error: 104 (Connection reset by peer))
15:27:18preglowbut i've gotta go
15:27:21Kasperlewho was the guy working on ipodpatcher again? (i think i've talked to him here before, don'T remember the nick :/)
15:27:37*preglow prods amiconnn and whispers "sw-tone-controls"
15:27:46*LinusN looks at linuxstb
15:29:01 Join relaxed [0] (n=relaxed@unaffiliated/relaxed)
15:31:17 Join Arathis [0] (
15:31:19KasperleLinusN: thanks :D
15:31:40 Quit linuxstb (Read error: 110 (Connection timed out))
15:31:46amiconnpreglow: Not possible to test before tonight
15:32:55LinusNgtg, cu folx
15:32:58 Part LinusN
15:34:33 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
15:34:39linuxstbKasperle: That's me...
15:34:54 Join My_Sic [0] (
15:35:11 Join Winchester [0] (
15:37:11 Join rickeyjt [0] (i=4c10d57a@gateway/web/cgi-irc/
15:37:15Kasperlei'm having some issues with the latest version on FreeBSD atm
15:38:05Kasperleit takes ages blocked on i/o and then responds "/dev/da0: Input/output error"
15:39:26Kasperlehrm. it magically works after reconnecting the iPod
15:39:46*Kasperle shrugs
15:39:53webtazdid you say sth like: abra kadabra?
15:39:54 Join PaulJam [0] (
15:40:05linuxstbWhat options were you using? It now automatically does a scan if you don't specify a device explicitly.
15:41:05Kasperlelinuxstb: this was with options "/dev/da0 -l", but it also hung without any options. it's fine now, though, so i guess it might've been something else, possibly fbsd's fault
15:44:37linuxstbOK, I'll ignore you unless it keeps happening...
15:45:06linuxstbI can't recall anything changing in the low-level disk access since the initial versions.
15:45:41 Join ackbahr [0] (
15:49:52 Quit Winchester (Remote closed the connection)
16:01:58 Join Mouser_X [0] (
16:07:53 Quit Obsys ("Chatzilla 0.9.77 [Firefox]")
16:10:27Kasperlelinuxstb: quite likely hald messing things up :/
16:14:02 Join ackbahr_ [0] (
16:17:07 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
16:17:42 Join shriphani [0] (n=shriphan@
16:25:53 Quit shriphani ("Leaving")
16:26:08 Quit ackbahr_ (Remote closed the connection)
16:26:35 Quit ackbahr (Read error: 110 (Connection timed out))
16:29:03 Quit petur ("stkov")
16:32:29 Quit Vyrus (Read error: 104 (Connection reset by peer))
16:32:38 Join Vyrus [0] (
16:33:58 Quit PaulJam (".")
16:40:22 Quit linuxstb (Read error: 110 (Connection timed out))
16:44:00 Join toffe [0] (
16:46:36 Quit Mouser_X (Read error: 110 (Connection timed out))
16:50:13 Quit toffe (Read error: 131 (Connection reset by peer))
16:53:06Kasperlei have a small patch for to make it usable on FreeBSD, too:
16:53:44 Quit GodEater ("CGI:IRC")
16:53:58 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
16:55:03 Quit GodEater (Client Quit)
16:55:23 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
16:55:45 Join _Veseliq_ [0] (
17:02:03 Join Domonoky [0] (
17:04:27 Join lukaswayne9 [0] (
17:05:49bluebrotherKasperle: does this mean that bsd's shell can't expand variables inside of quotes?
17:07:13bluebrotheror is it simply bash not available?
17:07:27Kasperlebluebrother: first thing i came across is that there is no /bin/bash on FreeBSD. next thing was that test ([) wouldn't work correctly for != if either side is empty (which happened to me)
17:08:01Kasperlein line 22, to be precise
17:08:16bluebrotherah. Does -ne work?
17:08:36bluebrotheriirc this is supposed to do the same as !=
17:08:46Kasperleum, that's for ints
17:09:09Kasperlebasically, i just made it prefix an "r" to $SVNALT, which is the same you did for $SVNVER
17:09:25Kasperle(and changed the shell to /bin/sh)
17:12:38 Join rp- [0] (n=rp@
17:13:43 Join fasmaie [0] (
17:15:43***Saving seen data "./dancer.seen"
17:17:45 Quit YouCeyE (Read error: 110 (Connection timed out))
17:18:26 Quit BigBambi ("Leaving")
17:25:06 Join tick [0] (i=c27f0812@gateway/web/cgi-irc/
17:25:22 Join psychoid1 [0] (
17:25:28preglowamiconn: ok, just a small warning unless i fix it before you can test it, you need to adjust the values before playung unless you want a short burst of white noise
17:26:01psychoid1are you breaking the law
17:26:11psychoid1by using your own firmware?
17:26:17psychoid1or is that not illegal
17:26:37psychoid1as it seems its being used on ipods
17:26:42preglowno, it's
17:26:43psychoid1im sure they have a copyright :/
17:26:44preglownot illegal
17:26:56preglowcopyright has nothing to do with this
17:27:02tick Llorean: hi, that's me again. I've faount a couple of places where HAVE_xxx symbols are not just defined (for later testing with #ifdef) but are also assigned a value. Which is not quite in the line with the purpose.
17:27:08psychoid1so you have had it legally checked
17:27:14psychoid1to make sure its all above board?
17:27:18tickE.g. firmware\export\config-e200.h(14): #define HAVE_LCD_BITMAP 1
17:27:24preglowpsychoid1: there's no need to legally check it, copyright has nothing to do with running your own firmware
17:27:33psychoid1yes it does
17:27:38psychoid1if you put it on an ipod
17:27:52psychoid1anyway this stuff reli does stink... id advise no1 2 use it...
17:27:56preglowif we were to rip apple's firmware, then put it on the internet
17:27:56 Part psychoid1
17:28:00preglowfuck off
17:28:24preglowfor his sake, i hope he's a troll
17:28:43preglowyesterday's performance surely gave me that impression
17:29:15tickIt should either be converted to just "#define HAVE_LCD_COLOR" or be defined as CONFIG_LCD_COLOR (if it's a bit flag)
17:29:41Kasperlehaha. what an idiot
17:30:35preglowisn't #define XXX equivalent to #define XXX 1 ?
17:31:45preglowi may be really wrong on this one...
17:32:27tickpreglow: wrt #ifdef −− yes. But if it's #define XXX 1 it can also be tested with #if whereas if it's just defined it can't
17:33:05preglowah, ok
17:33:21tickAnd we've said today that all the HAVE_xxx symbols should be tested with #if
17:33:29tickerr.. with #ifdef
17:33:45 Quit GodEater ("CGI:IRC (Ping timeout)")
17:34:35preglowfeel free to cook up a patch :)
17:35:34 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
17:35:46tickBTW, speex (for example) defines the constants as 1 so that it can be used both ways. But it's not good IMHO.
17:36:13preglowthe problem isn't that big in the codecs, though
17:36:22preglowbut i do agree, but there we've also got the concern that we don't want to modify them too much
17:36:23 Join Winchester [0] (
17:36:30preglowto ease merging in new versions
17:36:49tickpreglow: yes, they are beasts of their own. That's why I only care about RB's own code
17:37:10preglowbut consistency is good, so do feel free to make a patch
17:37:13preglowdinner time
17:37:49 Quit lukaswayne9 ("Ex-Chat")
17:40:59 Part tick ("l(a)unch time")
17:47:21 Join kubiix [0] (
17:50:36 Quit kaaloo (Read error: 110 (Connection timed out))
17:57:36 Quit robin0800 ("Now if you will excuse me, I have a giant ball of oil to throw out my window")
17:58:10 Join kaaloo [0] (
17:58:33 Part kaaloo
18:01:54 Quit XavierGr ()
18:05:06 Nick JB_away is now known as JoeBorn (
18:09:01 Join BigBambi [0] (
18:13:20 Quit secleinteer (Remote closed the connection)
18:14:16 Quit zylche (Read error: 104 (Connection reset by peer))
18:15:33 Join zylche [0] (
18:15:58 Join nls [0] (
18:17:36 Join pearldiver [0] (
18:22:44 Join Mouser_X [0] (
18:23:05 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:28:20 Join ArrowStomper [0] (
18:28:35ArrowStomperHello, is anyone in here?
18:29:30ArrowStomperI attempted to put rockbox on my sansa e250 and didn't copy the .mi4 file properly. is there anywhere I can get a copy of this file?
18:34:30 Quit pearldiver ("some games are better left unplayed")
18:34:52 Join webguest08 [0] (i=90208071@gateway/web/cgi-irc/
18:35:38 Quit GodEater_ (Read error: 60 (Operation timed out))
18:35:54 Part webguest08
18:37:40 Join PaulJam [0] (
18:38:41 Join Hall_9000 [0] (
18:39:29 Quit fasmaie (Read error: 110 (Connection timed out))
18:41:06 Quit PaulJam (Read error: 104 (Connection reset by peer))
18:41:11 Join PaulJam [0] (
18:41:57 Quit Mouser_X (Nick collision from services.)
18:42:06 Join Mouser_X [0] (
18:42:32 Join GodEater_ [0] (
18:42:48 Part Hall_9000
18:42:51 Join Hall_9000 [0] (
18:42:57 Part Hall_9000
18:43:32 Join pearldiver [0] (
18:44:45 Quit bonbonthejon (Read error: 131 (Connection reset by peer))
18:46:59 Quit pearldiver (Client Quit)
18:47:31 Join ackbahr [0] (
18:51:04 Join pearldiver [0] (
18:54:45 Quit bluebrother ("leaving")
18:55:53 Quit Mouser_X (Nick collision from services.)
18:56:00 Nick webtaz is now known as taz|away (
18:56:02 Join Mouser_X [0] (
18:58:01 Join petur [0] (n=petur@rockbox/developer/petur)
19:02:05 Quit ackbahr (Remote closed the connection)
19:03:33amiconnpreglow: The sw tone control has no effect on my mini other than changing the volume
19:04:04amiconn...and for some reason the range is -6...+9dB (?)
19:04:25 Join redbreva [0] (
19:05:18preglowamiconn: did you define HAVE_SW_TONE_CONTROLS, then?
19:05:21preglowamiconn: my guess is not
19:05:30amiconnAh, no
19:05:41*amiconn expected the patch was ready to use as-is
19:05:41preglowmentioned it yesterday, i think, but yes, do that :)
19:06:02preglowamiconn: ahh, no, like i also said yesterday, it's very preliminary, heh
19:06:03amiconnIn the platform config?
19:06:08preglowamiconn: yes
19:07:39preglowamiconn: that reminds me i think it probably might be a good idea to enable it on most ipods, the tone controls only seem to affect volume
19:07:51preglowsame thing on my nano, at least
19:08:00amiconnI though the hw eq does work on the G5?
19:08:10preglowyeah, therefore "most ipods"
19:08:15amiconnAh ok
19:08:35*amiconn enabled it for mini2g and X5 for testing now
19:10:08preglowamiconn: any comments on how it sounds are welcome, though without the prescaling it'll be somewhat hard to test for extreme settings. right now, the midpoint boost frequencies are set to 200 hz for bass boost and 6000 hz for treble boost
19:10:15preglowamiconn: i suspect we might want to lower the treble boost freq a bit
19:11:42PaulJamshould bugs with the midi plugin be reported on flyspray? or is the plugin in a too early stage.
19:12:12preglowLear: there?
19:13:02preglowLear: any reason for all the gain == 0 checking in replgaygain.c get_replaygain_int() and co? if you pass 0, it bugs out, and there are already checks for zero in fp_exp and co
19:14:46 Join toffe [0] (
19:15:45***Saving seen data "./dancer.seen"
19:17:11preglowLear: also, the bounds checks in convert_gain, have you checked those yourself?
19:17:27LearHm... Not sure about the one in convert_gain, but as I recall it, make sure it stays zero, even if there's arounding error, thus avoiding processing in dsp.c.
19:17:46preglowi wouldn't exactly call 17 decibels "unreasonably large"
19:17:59 Nick taz|away is now known as web-taz (
19:18:27amiconnpreglow: Working nicely, apart from the missing prescaler, which makes it limited use
19:18:35amiconnNo white noise, at least on mini2g
19:18:38preglowLear: make sure it stays zero? if you pass a gain of 0 db to those right now, they return zero, and that's definitely not right in any way
19:18:48preglowamiconn: i got some on h120 today, hmm
19:18:48Lear17 is unreasonably large in the context of replaygain.
19:19:05preglowLear: i guess
19:19:16 Quit Juddy ()
19:19:48preglowLear: right now i removed the zero checks in both convert_gain and get_replaygain_int and everything works nice
19:19:56LearI did some bounds checking in convert_gain when they where -17 and 23, then you went about and changed them. :)
19:19:59preglowi haven't exactly gived replaygain a thorough spin, though
19:20:13preglowdid i...
19:20:21 Quit Winchester (Read error: 110 (Connection timed out))
19:20:36preglowamiconn: so, but how about the sound? any tweaking to suggest?
19:20:43amiconnpreglow: Hmm, if the values are completely uset, it produces full-range white noise on x5
19:20:51 Join XavierGr [0] (
19:20:57amiconn...until I set either bass or treble, even setting 0 is enough
19:21:05preglowamiconn: yeah, that's what i got too
19:21:11preglownot a problem to fix
19:21:23preglowit's probably just me missing an init somewhere
19:22:24Learpreglow: replaygain.c was written for replaygain, and in dsp.c, it only actually do anything if not zero. In that context, the zero checks makes sense. Then you started using that code for other purposes, and the original assumptions aren't always valid. :)
19:22:56amiconnpreglow: No white noise on ipod
19:23:01preglowLear: in that case removing the zero checks won't hurt replaygain at all, then
19:23:05preglowamiconn: weird
19:23:21 Join printfXh4 [0] (
19:23:37 Join Thundercloud_ [0] (n=thunderc@
19:23:51 Quit Mouser_X (Nick collision from services.)
19:24:00 Join Mouser_X [0] (
19:24:23Learpreglow: as long as a replaygain of 0 doesn't activate the gain code, then yes.
19:24:41preglowLear: won't be changing the dsp.c replaygain code at all
19:25:00preglowso unless dsp.c actually checks the result for zero instead of the argument, nothing should happen
19:25:04preglowi'll have a look anyway
19:25:31Leardsp.c does check for zero, yes.
19:25:37amiconnpreglow: I also have the feeling that the setting isn't live on x5
19:25:38 Join mattzz [0] (
19:26:16LearThough maybe not in that way...
19:26:54preglowamiconn: definitely is on h120
19:27:10preglowamiconn: if it isn't live it's not my fault anyway, i haven't changed that part of the code
19:27:22preglowamiconn: still use set_int(), i just change the callback
19:27:27 Join Winchester [0] (
19:27:39Learpreglow: No, shouldn't break anything, actually...
19:27:44amiconnMaybe it was just the dsp delay
19:29:41preglowamiconn: if it's that big, we're going to have some fun "bug reports" with the prescaler indeed
19:30:08 Quit mattzz (Client Quit)
19:30:16preglowif you've spent a couple of seconds in the setting, the low latency mode should have had enough time to kick in
19:38:30 Quit XavierGr (Nick collision from services.)
19:38:33 Join XavierGr [0] (
19:39:43Nico_PLear: is it normal that long M4A files don't play ?
19:40:05Nico_Pi've got one that lasts approx. 4 hours and it just gets skipped
19:40:22Learnico_p: Yes, known limitation.
19:40:28Nico_PLear: ok
19:40:55Nico_PLear: where does it come from ? not enough mem ?
19:42:12LearSort of. Could perhaps be worked around, but the current "simple" implementation requires much memory.
19:42:20luckzanybody seen obo or anyone else with a clue of the scrobbling feature?
19:42:54 Quit Thundercloud (Read error: 110 (Connection timed out))
19:43:25 Join redbreva_ [0] (
19:43:54 Join darksoulk [0] (
19:44:11 Join wooo [0] (
19:44:51 Join bezlemin [0] (
19:45:16preglowdoes nyone know what happened with the fm recorder build when jdgordon commited an icon? 20kb extra gained that just vanished at the next commit?
19:45:33 Join GreyFoux [0] (
19:45:46amiconnpreglow: Build server with sh-elf-gcc from the stone age
19:45:57preglowhas it been fixed?
19:46:00amiconn(3.3.6, unpatched)
19:46:04 Quit GreyFoux (Client Quit)
19:47:42amiconnNot sure; the build system went crazy after that. Only 4 build servers for several rounds
19:47:54amiconnLast build seems okay again
19:48:22Nico_Paah, that explains the disappearance of the warnings...
19:48:54amiconnYes, the bad-gcc build servers were among the dropped ones
19:49:06amiconnThey're back now
19:49:17 Quit nls (Remote closed the connection)
19:49:24preglowtold bagder to tell whoever is responsible to fix them?
19:49:25Nico_Pcouldn't they be fixed ?
19:50:07amiconnFunny distro that still has this problematic gcc snapshot
19:50:29 Join nls [0] (
19:50:29amiconndebian had that last summer, long fixed ...
19:50:44 Quit redbreva (Read error: 60 (Operation timed out))
19:51:04amiconnThe stone-age sh-elf-gcc should of course be fixed asap
19:52:05preglowwasn't that gentoo or something?
19:52:24amiconnThe bad native gcc? I think so
19:53:07 Quit Mouser_X (Nick collision from services.)
19:53:16 Join Mouser_X [0] (
19:53:49 Join funky [0] (n=repulse@unaffiliated/funky)
19:54:51preglowis bagder still the only one having access to the build server system?
19:55:21 Quit rickeyjt ("CGI:IRC (Ping timeout)")
19:55:54pixelmaI believe is Galois'
19:57:36 Quit SaLoMoN (Remote closed the connection)
19:59:42 Join rasher [0] (i=3e4f4094@gateway/web/cgi-irc/
20:00 is missing m68k-elf-gcc it seems
20:00:45rasherAnd just when I wanted to download a build, too
20:02:50 Quit Shaid (Read error: 110 (Connection timed out))
20:02:52 Nick darksoulk is now known as Shaid (
20:03:26 Quit daurnimator (Read error: 110 (Connection timed out))
20:04:05 Join secleinteer [0] (n=scl@
20:05:02 Join SaLoMoN [0] (
20:05:04Soapluckz: what is your question?
20:05:44luckzSoap: I have bothering submission issues with, plus rockbox occasionally is very weird about what gets added to the log and what doesn't
20:05:53 Quit rasher ("CGI:IRC 0.5.7 (2005/06/19)")
20:06:04luckzand within the last month it twice just didn't give me a log of any value
20:06:26Soapon the submission issue side, audioscrobbler themselves have been having server issues.
20:07:02 Join PaulJam_ [0] (
20:07:04Soapand (used to?) has a problem where the totality of the log gets cleared upon submission, if the submission fails or not.
20:07:26nicktasticI have problems with qtscrobbler
20:07:43luckzit has a problem where the log gets half-skipped upon submission when failing
20:07:50luckzso the current track that failed plus the next 2 or 3 or so
20:07:53nicktasticI have to edit the log by hand to remove the submissions that have succeeded, then re-submit....over and over and over until it is finished
20:08:11luckznicktastic: that gets me"Spam protection triggered: You submitted a track dated earlier than your last submission.
20:08:11 Quit PaulJam (Read error: 104 (Connection reset by peer))
20:08:17 Quit Mouser_X (Nick collision from services.)
20:08:19Soapluckz: yes, you can work around that by keeping a copy of the log open, and constantly editing and resubmitting, but it is a pain.
20:08:22nicktasticluckz: I just wait a few minutes and try again ;)
20:08:25 Join Mouser_X [0] (
20:08:32nicktasticSeems to work
20:08:35nicktasticMost of the time anyways
20:08:36Soapexactly nicktastic - that is what I used to do.
20:08:47nicktasticSoap: *pain*
20:08:52Soapvery much so.
20:09:05nicktasticSoon I'm going to write a boo or python script to do this all for me
20:09:14SoapI contacted the author both via email and via the mysticriver forums, and never got a response.
20:09:30nicktasticSoap: From what I can tell, many, many such softwares have this problem
20:09:30SoapSo now I use the .net application for windows and haven't had a single problem. is just very sensitive
20:09:52nicktasticSoap: What's the name of the app?
20:10:01nicktasticBeing a .NET app, it might run on Mono on Linux
20:10:24Soapnicktastic: LogScrobbler - it's listed in the wiki.
20:10:29nicktasticSoap: Thanks!
20:10:43Soapluckz: on the "spam protection" warnings - are you using a target with a real-time-clock?
20:11:16nicktasticSoap: I.e., does my player have a RTC? I believe it does (iAudio X5)
20:11:24nicktasticI have timestamps
20:11:35nicktastic...if that's what you're getting at
20:11:35 Join taz|away [0] (
20:11:39luckzSoap: ja, X5.
20:11:55Soapnicktastic: are you running the perl script with the RTC command line option? IIRC it defaults to treating your log as it you don't have a RTC.
20:12:04 Quit web-taz (Nick collision from services.)
20:12:08luckzthe .net app used to have a *huge* problem with nonenglish characters
20:12:08 Nick taz|away is now known as web-taz (
20:12:08nicktasticdidn't see luckz: there ;))
20:12:15Soapthat would explain why you get continuing failures after the first failed submission.
20:12:27nicktasticluckz: major bummer
20:12:29luckzI guess it would.
20:12:41luckzI can fake a million play times.
20:12:41nicktasticSoap: I was using qtscrobbler
20:12:45luckzI mean, I have 10 days I can spread my submissions over :p
20:12:46SoapI haven't used that one.
20:13:07nicktasticSoap: It is very just triggers flood protection constantly
20:13:48Soapluckz: also the perl scrip, when it fails, appears to mangle your .log file in one of a dozen different ways - I would use another app for a while and try and determine what problems you see are Rockbox's, and what problems are the script's
20:14:19nicktasticThat's just terrible
20:14:22luckzI know what problems are whose because I occasionally look at stuff before submitting ;p
20:14:25luckzan example:
20:14:42Soaphave you tried running the PHP script locally (assuming you don't trust someone's website with your Last.FM password)?
20:14:44luckzif I stop playback when a song has been playing for 3 seconds, it usually gets added as fully played. that's something I can live with.
20:15:04luckzthe perl thing seems to have issues with the server, so I guess it's the server that's being borky
20:15:08Soapreally? you don't get the "S" in the log?
20:15:14luckzI get L.
20:15:17nicktasticluckz: The server will have a problem with that, iirc
20:15:19luckzor at least it submits.
20:15:28nicktasticluckz: not sure if it will error, but it won't count the track
20:15:48Soapthe perl script has issues with the server even when the server is behaving. I swear it runs too fast. If I bog down my computer with other tasks at a higher priority it fails less often.
20:15:52luckznicktastic: I had several tracks successfully submitted that I essentially only listened to for a second
20:16:17nicktasticluckz: haha /me makes a note of it
20:16:34luckztry shutting the player off just as a new song starts, see if it gets L'd
20:16:46SoapI look out for that luckz - I seem to get /more/ Ss than I think are right, not less. hmm.
20:16:55luckzyes, I also get more S's than are right.
20:17:03luckzbut I get Ls if I turn off at the "right moment"
20:17:09preglowwhen switching my nano off, backlight pulses for a bit. am i the only one getting that?
20:20:10 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
20:21:27nicktasticLogScrobbler seems to be working perfectly
20:21:31nicktasticThanks for the tip
20:21:41luckzlogscrobbler just hangs up for a while and gives me an error
20:21:55nicktasticYeah it blocks the UI...lame.
20:22:49bospaadjepreglow, when i shut my nano off, backlight goes off normally, though a horizontal line of random junk, a few pixels high, stays at the LCD screen for a while :(
20:23:13preglowbacklight goes off here, then on again before it's finished fading out, then off
20:23:47linuxstbpreglow: Is there much work to do before the RTC alarm is commitable?
20:24:08preglowlinuxstb: depends to which degree i'm willing to ignore the miserable state the i2c driver is in
20:24:10bospaadjepreglow, haven't got that problem here
20:24:13preglowlinuxstb: but no, i'd say not
20:24:54 Quit Mouser_X (Nick collision from services.)
20:25:03 Join Mouser_X [0] (
20:28:03preglowlinuxstb: also depends if i'm going to try to merge the x5 and ipod drivers
20:28:18 Quit toffe ("MegaIRC v3.27")
20:29:00 Quit Winchester (Read error: 104 (Connection reset by peer))
20:29:23luckzeither qt or log again did not manage things like characters at all :/
20:32:09luckzmh, qt.
20:32:18luckzlogscrobbler seems to cope with them just fine
20:35:26 Join PaulJam [0] (
20:37:26luckzit definitely does not do
20:37:29luckzbut it does do
20:37:32luckzor something like that :/
20:38:06 Join PaulJam__ [0] (
20:38:13 Join toffe [0] (
20:38:35 Quit PaulJam (Nick collision from services.)
20:38:41 Nick PaulJam__ is now known as PaulJam (
20:39:09 Quit kubiix (Read error: 131 (Connection reset by peer))
20:39:12 Nick Everybody|away is now known as Everybody (
20:39:22 Join fasmaie [0] (n=fasmaie@
20:39:46rp-barrywardell or dan_a here?
20:40:36dan_arp-: I'm here
20:41:12 Join pseudo_ [0] (
20:41:25pixelmaluckz: it could have to do with how your music is tagged. IIRC id3v2 tags can be in utf-16 and need to specify that, some taggers don't do that correctly
20:41:33rp-fine, 1. question do you know why the buttons for the sansa are broken in mpegplayer plugin?
20:42:14rp-if i play i file the only way i can stop it is a hardware power off
20:42:15dan_aNo - in what way are they broken, and since when?
20:43:08 Quit toffe (Read error: 104 (Connection reset by peer))
20:43:12rp-couldn't say since when, but i know they worked, when i tried my(copied) blit code
20:47:03preglowlinuxstb: any reason pcf50605_write_multiple() is not implemented?
20:47:06barrywardellrp-: I'm her now too
20:47:57dan_arp-: According to the code, the power button should stop playback
20:48:17 Join toffe [0] (
20:48:25preglowlinuxstb: also, why don't the pp i2c functions use the same names as the i2c.h ones?
20:48:28rp-and my 2. question, i'm currently working on the battery code, i noticed when i read out the battery level and play a movie, it gets stuck for a moment, when adc_scan() is called
20:49:00dan_arp-: You can read the battery without the Sansa locking up?!
20:49:10barrywardelldan_a: I managed to do that too
20:49:13 Quit PaulJam_ (Read error: 110 (Connection timed out))
20:49:26rp-dan_a: yes dan_a, but it doesn't, looks like the whole key thread doesn't work
20:49:27barrywardellit's just a matter of doing it after rockbox has already started
20:49:49dan_abarrywardell: Ah.
20:49:53barrywardellrp-: does the scrollwheel backlight still work?
20:50:05rp-yes there was some issue with the adc kernel thread
20:50:11rp-i'll check
20:50:20barrywardellthat's a good indicator of whether it really is the button thread that has died
20:50:26dan_arp-: Are you using SVN rockbox, or have you patched it at all?
20:50:55rp-dan_a: i'm using the svn, without any patch
20:50:55amiconnbarrywardell: There's a separate button thread on sansa?
20:51:54rp-barrywardell: scrollwheel backlight is working
20:52:14barrywardellso it's detecting the scroll wheel then
20:52:25rp-but i can't enter the menu or exit or....
20:52:51barrywardellamiconn: no, I mean whatever thread the button code runs on
20:53:11barrywardellthe button backlight code is run at the same time as the button detection
20:53:18amiconnHmm, buttons are read and put into the queue by the button tick task
20:53:36amiconn...and the gui thread reads the button queue
20:54:35 Nick Everybody is now known as Everybody|ingame (
20:54:49 Nick Everybody|ingame is now known as Everybody (
20:55:45amiconnIf buttons die on sansa and other adc readouts also cause problems (e.g. battery), the conclusion would be that something is wrong with the adc driver
20:56:17barrywardellthe problem with the adc driver is that we don't have one ;)
20:56:29amiconnOf course dying buttons might also indicate that tick tasks are no longer working, e.g. because the tick interrupt is disabled or somesuch
20:56:43barrywardelli think we should remove the adc code now that you mention it. It clearly doesn't do anything it's supposed to
20:57:00amiconnbarrywardell: Hm? rp-'s comment earlier suggested that there is one
20:57:06 Join mathgl [0] (
20:57:13dan_arp- has one - nobody else does
20:57:15 Quit pondlife ("disconnected has pondlife")
20:57:20barrywardellthe current SVN code is copied from the H10 driver
20:57:23barrywardelland doesn't work
20:57:32rp-it isn't really a adc driver :)
20:57:35barrywardellit only barely works on the H10!
20:57:51rp-i just put i2c code for the battery there
20:58:04 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
20:58:58barrywardellrp-: did you remove all the existing adc code too?
21:00:02 Quit printfXh4 (No route to host)
21:00:26rp-barrywardell: yes, i commented out the most of it
21:00:39 Nick pseudo_ is now known as printfXh4 (
21:01:54 Join Winchester [0] (
21:02:18barrywardelldo you have a copy of the code you used?
21:02:33barrywardelli mean, can you post a copy?
21:03:27rp-of the adc-e200.c file?
21:03:54barrywardellyeah, the code you have for reading the battery level
21:09:07luckz20:41:28 < pixelma> luckz: it could have to do with how your music is tagged. IIRC id3v2 tags can be in utf-16 and need to specify that, some taggers don't do that correctly <- we are reading from text files here. how my files are tagged has no relevance.
21:09:37barrywardellrp-: still here? can you post a copy of that code?
21:09:54rp-ah i remember, i put the i2c code directly to the adc_read function, because when the i2c code was called within the thread rockbox crashed
21:09:58luckzsaid text files use "ANSI as UTF-8", notepad++ claims
21:10:09rp-barrywardell: just a second
21:10:28luckzI don't think that how files are tagged strongly influences the creation of the .scrobbler.log file
21:11:22pixelmaah ok... I have no idea how that works ;)
21:13:08Soapluckz: are the files failing to appear in the file, failing to submit, failing to be recognized, or causing a crash of the submitting program?
21:13:30Soap(when I say failing to be recognized, I mean by Last.FM itself)
21:13:35luckzSoap: the current issue is only that logscrobbler fails at grasping the concept of unicode
21:13:40luckz"H?ngen Und B?ngen "
21:13:55luckzso -> ?
21:14:32SoapI can't load said page, but do you realize that Last.FM does not display the actual submissions you make?
21:14:48luckzI'll say it in simple terms:
21:14:54luckzit shows up fine if I use the perl thing.
21:14:58luckz(on the page)
21:15:03SoapIt displays the track it believes (through a process of peer moderation) you attempted to submit.
21:15:10luckzlogscrobbler seems to implode something in midway.
21:15:17peturkernelsanders already admitted logscrobbler doesn't do unicode, he said it's on his list
21:15:26barrywardellrp-: hmmm. and you don't get the stopping when you remove the code from adc_read()?
21:15:35luckzon his list means that it has some form of priority?
21:15:46***Saving seen data "./dancer.seen"
21:15:49peturI don't know if qtscrobbler is any better
21:15:55luckzthe "shift after last played" feature of logscrobbler is really helpful
21:15:59luckzas it lets me actually submit stuff
21:15:59Soapahh, well I guess that explains it then, I don't use UTF-16 ID3v2.3 tags, they break too many players.
21:16:03luckzbut if that then becomes all garbled..
21:16:05rp-barrywardell: you mean within mpegplayer?
21:16:27peturlukz: ask him on MR
21:16:43Soappoopy that problem hasn't been addressed yet.
21:16:46luckzMR being, here?
21:16:50luckzthe iriver page?
21:16:58petur1 sec
21:17:16barrywardellrp-: i'm referring to problem 2. you said playback stutters
21:17:20rp-barrywardell: if i make that part of the code unreachable it doesn't stop
21:18:18 Join inversions [0] (
21:18:18barrywardellrp-: ok. it could be that the i2c stuff takes a while to run, and adc_read gets called too frequently
21:18:53barrywardellwhat about moving the reading code into adc_scan() and have adc_read just output the value that was previously read in?
21:19:44rp-barrywardell: wait, i tried to playback a smaller( 128x128) mpg file and it doesn't stutter
21:20:31barrywardellthat would fit in with what i suggested. maybe the i2c stuff is holding the cpu up too much and it can't keep up
21:21:33peturluckz: you can also try and see if that works better
21:21:44rp-or i had a wrong version of rockbox on the player :)
21:21:50 Quit fasmaie ()
21:22:41luckzpetur: that does not have the nice fancy "submit with date after last submission" featurey
21:23:01barrywardellrp-: ?
21:23:36peturluckz: yes
21:24:28peturluckz: use the 'recalc Date/Time' button bottom-right
21:24:51luckzI thought that did some form of "change date relative to now mysteriously"
21:25:03rp-barrywardell: now i'm confused, seems the adc code had nothing todo with the stutter, i had some debug stuff in the lcd_yuv_blit function
21:25:11 Join fasmaie [0] (n=fasmaie@
21:25:12rp-barrywardell: now the playback is smooth
21:25:28peturluckz: it recalcs date/time to start from now and count backwards
21:26:09luckznot as fancy as since last submission. mh. guess I'll try that with some calculating around.
21:26:35barrywardellrp-: haha. i hate when that happens!
21:26:36 Join kopf [0] (
21:26:59peturluckz: qtscrobbler somes with source so you can always change it ;)
21:27:02kopfdoes everyone else have to restart their 4th gen iPods with rockbox regularly (like 3 times a day)?
21:27:13 Join webguest27 [0] (i=557ca9ca@gateway/web/cgi-irc/
21:27:22 Quit webguest27 (Client Quit)
21:27:23desowinluckz: does have way to check what was the last submission date ?
21:27:58 Join webguest1354178 [0] (i=557ca9ca@gateway/web/cgi-irc/
21:28:01rp-barrywardell: yeah, i'm sorry, but i'm not quite sure if it really was that, because battery is at 100%
21:28:22 Join Thundercloud__ [0] (n=thunderc@
21:29:29peturnice nick ;)
21:29:47barrywardellrp-: is your button problem also ok now?
21:29:49Soapdesowin: apparently it does as LogScrobbler will use your last submission date to fake your playback times for you.
21:29:53 Quit redbreva_ ("Chatzilla 0.9.77 [Firefox]")
21:30:11Soapdesowin: I don't know if that is part of the plugin API, or if it scrapes the website.
21:30:20rp-barrywardell: nope thats still there
21:30:21desowinLogScrobbler is open ?
21:30:37peturdon't think so
21:30:55SoapIt's on googlecode - aren't projects hosted there supposed to be OS?
21:31:01rp-barrywardell: i could make a patch of the lcd_yuv_blit(), but actually it's just a copy of the gigabeat one
21:31:20luckzpetur: I doubt my talent at dealing with source of anything beyond pascal/delphi :(
21:31:35barrywardellrp-: that would be handy. I'd like to look at it and maybe commit it
21:31:44webguest1354178rockbox permanently hangs on my ipod photo, does anyone know how to fix that?
21:31:45peturSoap: seems I'm wrong
21:32:17luckzdesowin: does not want to have anything submitted played before the last submission, if you mean that
21:32:17 Quit Mouser_X (Read error: 110 (Connection timed out))
21:32:30desowinluckz: yes, I know
21:32:38luckzwhat was your question exactly?
21:32:47luckzlogscrobbler seems to query the page for what was last submitted
21:32:55desowin <- I can get last submission date from it, so I'll implement it soon
21:34:03kopfLogscrobbler won't submit for me at all
21:34:07kopfi have to use that php script
21:34:13 Quit wooo (Read error: 110 (Connection timed out))
21:34:14peturdesowin: as you're taking featurerequests, any chance of 1) remembering the location (ie don't open browse dialog) and 2) progressbar?
21:35:09desowinyou scrobbler so much so you need progressbar ? :D
21:35:19petursometimes, yes
21:35:32desowinI'll add it to todo
21:35:33peturit 'll get worse when I upgrade my battery ;)
21:35:38 Join taz|away [0] (
21:38:50rp-barrywardell: <−− i forgot to select a operation system :>
21:39:32hcsI can edit it
21:40:10hcsdone, though it is already obvious from the title
21:40:18rp-thank you
21:40:29 Quit Dunkelschub (Read error: 104 (Connection reset by peer))
21:40:49barrywardellrp-: great. i'll have a look now
21:40:49luckzkopf: did you try qtscrobbler?
21:40:53 Quit Lear ("Chatzilla 0.9.77 [Firefox]")
21:41:30toffegood news I have the console port connected on the gigabeat S30
21:41:53toffecheck the boot sequence on the wiki page :
21:42:24toffeIt is only in receive for now, just trying to fond the transmit
21:42:49toffefind ^^
21:43:10hcsjhMikeS: I saw your DSP stuff, cool; any further thoughts about SPC?
21:43:55toffethe boot sequence must be the same for the zune .. more or less ;)
21:43:56preglowshouldn't spc run better now?
21:44:11hcsyeah, it relies on the dsp resampler
21:44:22hcsooh, we haVe chopper now?!
21:44:58hcscool, I've got to get around to porting parachute/sabotage
21:45:00preglowthink i'll try to reenable echo now
21:45:05preglowspcs sound so bloody dead without it
21:45:05kopfluckz: no, never heard of it.
21:45:13kopfi'd need Qt installed though wouldn't i
21:45:23luckzyou just download and run it
21:45:44kopfdownloading now
21:45:46luckzthe exe at least. dunno where the tar.gz would get you
21:46:20kopfcause every time i use that blasted php script, i get authentication errors for about 4 hours on my foo_audioscrobbler plugin
21:46:24kopfit drives me fucking insane
21:46:50luckzsounds.. cool
21:46:53Soapit that a timezone issue?
21:47:15kopfthey eventually get scrobbled
21:47:24kopfit's an authentication issue, according to the foo_audioscrobbler logs
21:47:27luckzis that an issue, them being scrobbled slightly later?
21:47:29kopfwhich makes no sense
21:47:56kopfthink about it
21:48:07kopfespecially in the context of scrobbling from both an ipod and foobar
21:48:22preglowhcs: seems like it works just nice with echo on h1x0 now
21:48:41preglowboosts alot, but works
21:49:56 Quit fasmaie ()
21:50:24hcspreglow: glad to hear it, but I expect no success on ipods, since he really did mostly coldfire optimization
21:50:43hcspreglow: Uncharted Waters 2, by any chance (happens to be one of the most demanding)?
21:50:53preglowhcs: yeah, i'm going to do the arm resampler soon
21:51:00preglowprobably tomorrow, we'll see
21:51:16preglowhaven't got that, no
21:51:19 Quit Thundercloud_ (Connection timed out)
21:53:06 Quit web-taz (Read error: 110 (Connection timed out))
21:53:52preglowwhat's up with the volume, btw?
21:54:24hcsI have it set to not do any amplification... it actually uses the full 16-bit range for some tracks, it seems
21:54:32preglowahh, ok
21:54:33preglowthen it's all good
21:54:34 Join daurnimator [0] (n=fake@
21:54:53rp-could anyone provide me a guide on how to get the battery code working for rockbox? how do i get the capacity(min/max) and the scale factor?
21:55:02preglowquiet music should be allowed to be quiet
21:55:27barrywardellrp-: look in firmware/powermgmt.c
21:55:45hcspreglow: yeah, I understand, it's just annoying with all my too-loud MP3s
21:55:48barrywardellyou basically set a scale factor to convert the adc reading to centivolts
21:56:11barrywardellrp-: also note that the adc value is supposed to be 10 bit
21:56:16preglowhcs: someone should code a dynamic normaliser
21:56:43preglowif i follow my trend in coding stuff i'll never use, it'll probably end up being me doing it
21:57:18Mikachuhcs: you know about mp3gain?
21:57:51barrywardellrp-: mpegplayer works nicely :) I wonder what is causing the buttons not to work.
21:57:56*hcs is just lazy enough to not do anything about it but adjust volume between tracks
21:58:45rp-barrywardell: ahh you calculate it to centivolts
21:58:54 Quit _Veseliq_ (Remote closed the connection)
21:59:02preglowahh, ff5 with echo
21:59:08preglowthe way the good lord intended it to be
21:59:23rp-barrywardell: nice, yes its odd, it worked some time ago, don't know what broke it
21:59:24preglowthe good lord being nobuo uematsu, of course
21:59:29barrywardellrp-: yes. the code is a bit confusing.
22:00:15 Part juxtap
22:00:25barrywardellrp-: was it more than 7 weeks ago when it was working?
22:00:27 Join fasmaie [0] (n=fasmaie@
22:01:13rp-barrywardell: yes it is, is there a reason why you convert it to centivolts? because i guess every controller what return you a max(full) and a min(empty) value or not?
22:01:21hcspreglow: one of the reasons uncharted waters 2 is so intensive is that it does its own echo by just repeating the notes on a bunch of channels
22:01:42preglowtastes like .mod
22:02:02Mikachuyoko kanno, she did the gits soundtrack
22:02:19preglowshe also did the cowboy bebop soundtack
22:02:22preglowwhich fucking rocks
22:02:23rp-barrywardell: i think so, i could check the irc logs, because i talked with linuxstb about the lcd_yuv_blit, because it had a bug, when i implemented it
22:02:25barrywardellrp-: i'm not sure what the reason is. maybe for consistency between targets
22:02:27Mikachui bet she did lots
22:02:28preglowin absolutely all ways anything can rock
22:02:37hcsyep, but she started with koei video games :)
22:02:56barrywardellrp-: I'm asking because audio support was added about 7 weeks ago, and I wonder if that is causing problems
22:02:56preglowbest anime ever with the best soundtrack ever
22:03:56Mikachuthe title is funny at least :)
22:04:20hcsto lunch, enjoy the echoes
22:04:36rp-barrywardell: i think it was in mid december, because that time i started rockbox hacking
22:05:59barrywardellrp-: it is indeed the audio support that was added
22:06:05barrywardellthat breaks it
22:06:43rp-odd how could that be? hm ok the sansa hast no audio yet ;)
22:08:04barrywardellyep, exactly!
22:09:52 Quit atsea- ("using sirc version 2.211")
22:13:29rp-hm it seems 0x30F is the value when it's full charged
22:14:39 Join mattzz [0] (
22:16:08barrywardellrp-: what order are you taking the bytes in?
22:16:22 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:17:37barrywardellrp-: I think I'll commit your mpegplayer patch if that's ok? The button thing is a separate issue, really
22:18:28rp-barrywardell: yes no problem, but i only copied the code from the gigabeat and changed the frame buffer it writes to :)
22:18:42 Quit fasmaie ()
22:19:44 Join atsea- [0] (i=ariel@gateway/tor/x-181e3d72502f6f91)
22:19:52pearldiverwhat does the patch do?
22:20:31rp-pearldiver: video playback for sansa e200
22:22:50rp-barrywardell: according to the BL firmware it should work like this, have disassembled it −−>
22:24:36barrywardellrp-: ok. so bat[1] is the LSB, bat[0] is the two msb's
22:24:40 Quit amiconn (Nick collision from services.)
22:24:40 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:24:55barrywardellthat could be done a bit simpler
22:25:14barrywardelllike this: battery = (bat[0]<<8) | bat[1];
22:25:38 Quit mattzz ("Leaving")
22:27:48rp-barrywardell: what is lsb and msb?
22:28:19barrywardellLSB = least significant byte, msb is most significant bit
22:28:26rp-ah ok ;)
22:28:40barrywardelli'll have a scale factor for you now in a sec...
22:29:02 Nick Everybody is now known as Everybody|away (
22:30:24rp-whats odd in the BL, they read that value 3 times in a row
22:31:07 Join fasmaie [0] (n=fasmaie@
22:31:40amiconnrp-: The conversion to centivolts is for, ahem, displaying the voltage, and sticking with integers
22:32:08 Part LinusN
22:33:06rp-ah ok
22:33:29barrywardellrp-: try setting BATTERY_SCALE_FACTOR to 5358 in config-e200.h
22:33:33amiconnThe voltage vs. battery percent scale isn't linear anyway
22:33:59barrywardellamiconn: do you think the voltage conversion is really necessary for any reason?
22:34:09 Nick HD|NA is now known as HardDisk_WP (
22:34:20 Quit BigBambi ("Leaving")
22:34:31amiconnbarrywardell: I'm not seeing it as a conversion - we're measuring a voltage after all
22:34:44barrywardellrp-: do they average the 3 values? or just taking the last one read?
22:34:46rp-barrywardell: ok done, and should i set min, max?
22:35:19rp-barrywardell: they just take the last, as far i could see
22:36:03rp-but maybe they do something with the values on the stack, didn't look further
22:36:18barrywardellmax and min are not essential, really
22:36:58amiconnThe powermgmt code calculates a sliding average of the battery voltage
22:37:35barrywardellamiconn: but why not just work with the value read from the adc directly? why do we need to work in voltages?
22:38:19amiconn(1) because we want to display the voltage. (2) In order to be able to reuse voltage->percent mapping tables
22:38:25barrywardellamiconn: yes, rockbox probably doesn't need to read 3 times. but maybe the OF isn't as sophisticated, hence the 3 reads
22:38:35 Join wooo [0] (
22:38:38 Quit JoeBorn (Nick collision from services.)
22:39:00 Join jborn_ [0] (
22:39:04amiconnIirc rockbox reads once per second, and calculates a sliding average. No need to read multiple times at once
22:41:26barrywardellfor (1), the only place I see actual voltages being output is in the debug screen and battery_bench.
22:41:32 Quit fasmaie ()
22:41:43barrywardellfor (2), couldn't it just be changed to an adc_value->percent table?
22:42:35amiconnVoltage is also shown in the charging screen, for targets where that exists
22:43:48barrywardellI didn't know about the charging screen. Is it useful to have a voltage reading there?
22:44:29amiconnIt's a little extra
22:44:44amiconnThe bootloader also shows the voltage, at least on the irivers and the x5
22:45:45*amiconn wonders how tomal builds rockbox for ifp
22:45:57 Join mirak [0] (
22:46:02amiconnThere must be uncommitted changed, svn doesn't compile
22:46:23amiconn"region IRAM is full"
22:46:35barrywardellhmm. ok. It just seems there are a lot of conversion functions in the powermanagement code, and I wondered if they were really worth the space
22:46:42barrywardellbut anyway, I gotta go now.
22:46:59amiconnThe powermgmt code is a mess, I agree
22:48:09barrywardellone of these days, I'm going to have a go at cleaning it up.
22:48:18 Quit barrywardell (Remote closed the connection)
22:48:31 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
22:49:50 Nick Everybody|away is now known as Everybody (
22:54:55 Quit darkless ("Leaving")
22:55:01 Join darkless [0] (
22:55:41 Quit inversions (Read error: 113 (No route to host))
22:56:04 Nick Everybody is now known as Everybody|Lost (
22:56:28 Nick markun_ is now known as markun (n=markun@rockbox/developer/markun)
22:57:33desowinluckz: can you compile from source ?
23:01:28 Join ber1 [0] (
23:06:39 Join Mouser_X [0] (
23:06:43 Join fasmaie [0] (n=fasmaie@
23:10:24 Quit bospaadje ("great minds run in great circles")
23:14:23luckzdesowin: I do not have anything that compiles anything else
23:14:43desowinyou're on windows ?
23:15:49***Saving seen data "./dancer.seen"
23:16:22 Join web-taz [0] (
23:18:48 Join bospaadje [0] (
23:19:49 Quit rp- ()
23:20:50 Join safetydan [0] (i=cbca159f@rockbox/developer/safetydan)
23:22:26 Join entheh [0] (
23:23:59desowinso, I guess you'll have to wait for next qtscrobbler release to get feature you wanted ...
23:34:22 Quit taz|away (Read error: 110 (Connection timed out))
23:35:29 Quit Winchester (Read error: 60 (Operation timed out))
23:35:46Soapand the PHP doesn't do it for you?
23:36:01SoapI'd run it on my site, but who would trust me?
23:36:46 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
23:36:50 Part ber1
23:38:59 Nick jborn_ is now known as JoeBorn (
23:41:27 Part nls
23:41:37 Nick Everybody|Lost is now known as Everybody (
23:42:06 Join Everybody|away [0] (
23:42:11 Quit Everybody|away (Client Quit)
23:42:18 Quit fasmaie (Read error: 110 (Connection timed out))
23:43:02 Quit midkay ("Leaving")
23:43:07*petur votes against changing nick to indicate status
23:43:30*bospaadje seconds petur
23:43:31 Join ]RowaN[ [0] (
23:43:56bospaadjestatus in nick sucks
23:44:45markunis there a way to discourage it?
23:44:49]RowaN[none of my mpegs are playing in my sansa.. maybe i need smaller mpegs
23:45:24Mikachumarkun: kick anyone who does it
23:45:31]RowaN[on a sidenote.. seems to be no icon for .mpEg files.. although maybe thats not a proper file extension anyway
23:46:24 Part LinusN
23:47:10 Quit webguest1354178 ("CGI:IRC (EOF)")
23:47:12 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
23:48:51markunMikachu: and ban nicks with | in them?
23:49:15idnar/nick idnar[breathing_in]
23:49:16idnar/nick idnar[breathing_out]
23:50:41idnarpetur: the disturbing thing is, some people don't think that's absurd
23:50:48 Join stoffel [0] (
23:50:51idnarit's just that it's too hard for them to type it out the whole time, so they don't bother :P
23:53:01peturI would only notify people I'm away/back if I think somebody is interested (ie I was talking to somebody)
23:53:11idnarI use /away
23:53:28idnarif I'm in the middle of a conversation, I'll say something like "gotta run, probably be back in a few hours" or whatever
23:53:35 Quit jhMikeS (Nick collision from services.)
23:53:37peturI use /quit ;)
23:53:41 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
23:54:04idnarI never /quit ;)
23:55:08*petur is no longer surprised there are 135 in the channel
23:58:28Mikachumarkun: banning isn't useful since you can change to a banned nick but not from

Previous day | Next day