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 2006-08-31

00:00:23BagderI published all my mi4 files and my stumbling efforts
00:00:35Bagderhe came bursting in with answers to all questions ;-)
00:00:44Nico_Pwow... providential man
00:01:35 Join linuxstb_ [0] (i=5343d4aa@gateway/web/cgi-irc/
00:01:53Bagdergetting that TEA encryption nailed is in fact quite a feat in my eyes
00:02:05blxthe folder names on my iaudio turns lowercase somehow
00:02:20Bagderblx: it is probably the way your OS shows them
00:02:22 Quit _Veseliq_ (Remote closed the connection)
00:02:30blxBagder, debian gnu/linux
00:02:37Bagder... can be told to do so
00:03:00blxmight be the thing that's wrong..
00:03:12BagderI doubt that
00:03:31bluebrotheris there a way to "reset" the settings in the sim before startup?
00:03:42linuxstb_Would we be able to add the mi4 decryption to our bootloaders - to load original firmwares?
00:04:25bluebrotherah, found it. sector3D.bin
00:04:32Bagderlinuxstb: yes sure
00:04:45linuxstb_blx: I use the option "shortname=mixed" in my fstab entries for vfat disks.
00:04:57blxlinuxstb, thank you
00:06:10blxthere we go
00:06:40barrywardellBagder: what do you think about providing compiled mi4 bootloaders?
00:06:53 Quit lodesi_ (Read error: 110 (Connection timed out))
00:07:00blxis x5v_fw.bin a correct name for the image?
00:07:01Bagderwe do the similar for other targets
00:07:17Bagderbarrywardell: see
00:07:32linuxstb_barrywardell: BTW, I now have scramble generating a .mi4 file my Elio accepts. Next will be the 0103 files, then I'll commit.
00:07:38barrywardelloh, cool
00:07:54barrywardelllinuxstb_: nice
00:08:38linuxstb_I'm planning to call the options -mi4v2 and -mi4v3 - can anyone think of anything better?
00:08:48Bagderfine enough for me
00:08:58linuxstb_I'm not sure what's needed for the v3 files yet though - will they all be the same?
00:09:15 Part tucoz ("Leaving")
00:09:15barrywardellBagder: I'll send you mi4 files for the two H10 models. Or you can compile them yourself if you want
00:09:23merbananlinuxstb_: the original bootloader can decrypt the original firmware, just concatenated the rb fw and the original firmware
00:09:52Bagderlinuxstb: yes, they'll be pretty much like v2 but with the dummy DSA as well
00:09:57merbananand then set the plantext lenght to the end of the rb firmware
00:10:22linuxstb_merbanan: OK. But it would be nice if installation just meant renaming pp5020.mi4 to something like "original.mi4" and then copying the Rockbox bootloader as pp5020.mi4.
00:10:32markunBagder: should I integrate the gigabeat firmware scambler with the scramble tool or just leave it as a seperate program?
00:10:42linuxstb_Rather than running any tools to concatenate the bootloader and original firmware.
00:11:04Bagdermarkun: as you wish, I personally like having one tool that does the job but its not really a big issue
00:11:10uskilinuxstb_: hello, it's me again >:) just a quick question, do you think that the MBR of the compactflash card matters ? (the first 512bytes sector)
00:11:41linuxstb_markun: I'll be adding the mi4-building code to a new file (mi4.c) which will be linked to scramble. You could do the same.
00:11:50barrywardelllinuxstb_: that's how I was thinking of loading the original firmware
00:11:51merbananlinuxstb: then you need to either call code in the bootloader or code your own decryptor and add the decryption keys to the rb firmware, or decrypt the original firmware
00:11:53 Quit Kohlrabi ("Quit")
00:12:43barrywardellmerbanan: you can have the original.mi4 be an already decrypted firmware
00:12:59Bagderbut that makes it slightly less conveniant
00:13:05blxok, it worked with the other version of the bootloader.
00:13:22Bagderthe other version?
00:13:36linuxstb_uski: The MBR (first sector) contains the partition table. So in that respect it matters. You could try taking an MBR from the Rockbox wiki (IpodInstallationFromMacOSX page), dd it to your disk, then fix the partition table part using fdisk.
00:13:42barrywardellBagder: true
00:13:54uskilinuxstb_: thanks again
00:13:56 Join stripwax [0] (
00:14:11BagderI think having a decrypt function with the key built-in a fair idea
00:15:28blxBagder, there was two versions on the X5 dl page
00:15:40Bagderyes, one for X5 and one for X5V
00:16:15 Quit lodesi (Read error: 110 (Connection timed out))
00:16:41blxok, that was not very clear
00:17:12Bagderat least it explains why it didn't work ;-)
00:17:49blxtime to copy lots of music to the player now :)
00:18:07Bagdernah, just play the games ;-)
00:18:24blxyou have to download them separatley right?
00:18:35Bagdernope, they come with Rockbox
00:19:05Bagderbut rockbox is separate from the bootloader, if that was what you meant
00:20:13lorijhoBagder: I'll send you the H10.mi4 for the 6GB version
00:20:15Bagderwe miss ipod3g on the devicechart
00:20:31lorijhoBagder: where to ?
00:20:52Bagderdaniel a t rockbox . org
00:21:57barrywardellBagder: i just sent you mine too
00:22:27BagderI just noticed
00:25:23amiconnBagder: Hmm, your new configure makes it look like archos has the most supported targets :)
00:25:50amiconnBut that's in fact not true - the 2 gmini targets aren't really supported :/
00:26:00Bagderwell, targets in configure aren't necessarily "supported"
00:26:05lorijhoBagder: the bootloader is sent.
00:27:15Bagderthe new menu looks a lot less cramped though
00:28:49Bagderlorijho: so what is the exact target file name for the h10 5GB?
00:28:59Bagderseems appropriate to use that
00:29:48 Join Snyper [0] (
00:29:52preglowme bed
00:30:10BigMacanyone talk to senab on msn messenger?
00:31:15dan_aBagder: If I do "mi4code decrypt -s SKU_E-PP5022.mi4 SKU_E-PP5022.mi4-decrypt ; mi4code build SKU_E-PP5022.mi4-decrypt PP5022.mi4", should PP5022.mi4 be a bootable image?
00:32:24dan_aOh, no, I need to sign it first.
00:32:31barrywardelldan_a: you also need to do mi4code encrypt -pall SKU_E-PP5022.mi4 PP5022.mi4 default
00:32:51barrywardellyou don't need to sign it but the crc has to be right
00:32:58barrywardellencrypt takes care of that
00:33:27barrywardellor maybe you do need to sign it too if its a 010301 version
00:34:01Bagdersign yes
00:34:45dan_aDoes doing a "build" also do an "encrypt"?
00:35:14 Quit bluebrother ("Leaving")
00:35:21Bagderbuild, encrypt then sign
00:35:24lorijhoBagder: H10.mi4
00:35:40Bagderlorijho: right, thanks, I noticed you had it that in the tar
00:35:56lorijhoBagder: yep
00:36:02Bagderboth H10 mi4 files are now available here =>
00:36:34belzethx alot
00:36:40barrywardellthanks Bagder. Saves a lot of explanations about how to compile :)
00:36:58Bagderlet the hoards come
00:37:14stripwaxIs the 64MB patch for ipod 5g in cvs?
00:37:42 Quit Siku ()
00:37:44barrywardelldan_a: using -pall with encrypt leaves the mi4 unencrypted but fixes the crc
00:37:56Bagderstripwax: but linuxstb has expressed an intention of commit
00:38:26stripwaxcool. was just wondering since the tools/configure script has changed and I'm trying to merge in the patch with the changes
00:38:53Bagderthe patch should merge fairly easy
00:39:09Bagderif its the memory-size option version
00:40:34barrywardellmarkun: I've committed my change for the H10. It's very small so should be easy to fix your wmcodec patch
00:40:56stripwaxBagder - ah, that's probably the problem.. let me check out the latest version of the patch...
00:41:34dan_aYay! The re-signed Sansa firmware boots (though I'm sure that's been proved before)
00:42:07Bagderyes ;-)
00:49:01linuxstb_I've just created a .mi4 using the "h10" option to, and the only difference apart from the version number is a 0x01 at byte 0x2f - is that all I need to do?
00:49:19linuxstb_^the only difference to the 0102 files
00:49:39Bagderapart from the header number
00:49:48Bagderah you said that
00:50:10linuxstb_Looks like I can now clean up the code and commit then.
00:50:25 Quit [San] (Read error: 110 (Connection timed out))
00:51:55 Quit Febs (Read error: 110 (Connection timed out))
00:53:11 Quit ender` (" Liar, n: one who tells an unpleasant truth.")
00:55:11 Quit alberink (Read error: 60 (Operation timed out))
00:57:02markunbarrywardell: ok, that was easy to fix
00:57:07BigMacbadger can you hand out artist badges?
00:57:11BigMacor who can
00:57:16 Join [San] [0] (
00:57:32BagderBigMac: llorean (in the forums) is badge master
00:57:33stripwaxhrm, how do I exclude building rockdoom in a dev build? seems it isn't listed in apps/plugins/SOURCES
00:57:34barrywardellmarkun: yes, easy when you know how
00:58:05Bagderstripwax: apps/plugins/SUBDIRS
00:58:42markunbarrywardell: I meant the conflict, not the h10 bug :)
00:59:12stripwaxBagder - ah, doh. thanks
00:59:56 Quit MarcoPolo (Remote closed the connection)
01:02:32BigMacanyone got a sec to help me fix my wps
01:06:10 Quit rretzbach (Read error: 104 (Connection reset by peer))
01:06:15 Quit ootput ("leaving")
01:06:48uskiany chance to reprogram the bootloader stored in the flash memory of an apple ipod mini ?
01:07:22uskithe idea is to make a bootloader that will be less picky about the CF card used
01:08:47 Part stripwax
01:08:48linuxstb_uski: Very little chance.
01:09:34uskiok, ty
01:10:58BigMaclinuxstb:how would i go about installing this?
01:16:32 Quit TeaSea (Remote closed the connection)
01:17:15*linuxstb_ commits mi4 format support to scramble
01:17:20 Quit ghode|afk (Read error: 110 (Connection timed out))
01:17:43 Quit [San] (Read error: 110 (Connection timed out))
01:20:48 Join [San] [0] (
01:20:52BigMacanyone know of an alternative snc site then since it is down
01:20:59markunlinuxstb_: and you broke my gigabeat support which I just added..
01:21:17 Join Pyromancer [0] (
01:21:28linuxstb_markun: How did I manage that?
01:21:31barrywardelllinuxstb_: works great here
01:22:08linuxstb_barrywardell: Good. The files should be identical to the output of
01:22:32markunlinuxstb_: conflicts in the Makefile and scramble.c
01:23:48markunlinuxstb_: you forgot to add mi4.h after scramble.o:
01:24:24linuxstb_Ah yes. Will you fix in your commit?
01:26:30markunis the scramble tool only used for the bootloader?
01:26:51linuxstb_It's used for the main firmware for the Archos devices.
01:27:04 Join Paul_The_Nerd [0] (
01:27:14markunAnd can I set different output filenames for the bootloader and the firmware?
01:27:17linuxstb_In fact, it's used for everything.
01:27:40linuxstb_The rockbox.iriver etc files are generated using scramble.
01:28:00linuxstb_I'm not sure how to integrate it into configure though.
01:28:20markunWhy would you scramble rockbox.iriver?
01:28:51linuxstb_It just adds a small header - four bytes identifying the model, and a simple checksum.
01:29:40linuxstb_It's the -add option.
01:35:07linuxstb_barrywardell: Is sound now perfect on the H10 after your GPIOL commit?
01:36:44barrywardelllinuxstb_: pretty much, yes. the cpu can't keep up with peak meters on though
01:37:44Paul_The_NerdThat's pretty much the same as the iPods
01:39:16 Join _belze [0] (
01:39:27lorijholinuxstb_: flac and wav playback works smoothly. MP3 and ogg/vorbis playback stutters when wps is displayed (peak meters)
01:39:43linuxstb_As Paul_The_Nerd just said - the ipods are exactly the same.
01:40:43linuxstb_BTW, has anyone run the test_fps plugin on the H10 to test LCD update speed? test_fps.c is in CVS, but you need to add it to apps/plugins/SOURCES for it to get compiled.
01:40:49 Join ben_ [0] (
01:41:31 Quit bbroke ("Chatzilla 0.9.75 [Firefox]")
01:41:48 Quit [San] (Read error: 110 (Connection timed out))
01:42:13 Quit lightyear (Read error: 110 (Connection timed out))
01:42:37barrywardelli'll try it now
01:44:08 Join [San] [0] (
01:44:28linuxstb_Results for all the other targets are here:
01:44:48linuxstb_Also, have you tried mpegplayer?
01:45:58barrywardellwe don't have the blitting yet so it doesn't work
01:46:20barrywardellthe lcd driver needs work in general. it is very minimal
01:47:34***Saving seen data "./dancer.seen"
01:47:59barrywardellhow do I change the cpu frequency in test_fps?
01:48:29linuxstb_You change it in the debug menu before starting test_fps.
01:48:47barrywardell30MHz: 1/1 19.5 1/4 77.5
01:49:33linuxstb_Almost identical to the ipod Photo - which has a 220x176 LCD.
01:50:12barrywardell24Mhz: 1/1 18.0 1/4 72.5
01:50:24barrywardellroom for optimisation then
01:50:49linuxstb_How about 75MHz?
01:51:22barrywardelllol.i only have down on the scrollpad.i'll have to recompile to try it
01:51:54linuxstb_24MHz on the ipod only gives 15.5/60.5
01:52:32 Quit belze (Read error: 110 (Connection timed out))
01:56:08barrywardell75Mhz: 1/1 37.5 1/4 148.0
01:56:09markunbarrywardell: will you add variable scrollspeed for the scrollpad?
01:56:58barrywardellmarkun: it would be nice eventually. for now all we can to is tell that the scrollpad was pressed. we can't even tell where it was pressed
01:57:56linuxstb_barrywardell: Ouch at your LCD speeds. Hopefully they can be improved.
01:57:57 Join Drkepilogue [0] (i=478df32d@gateway/web/cgi-irc/
01:58:13barrywardellouch exactly
01:58:26barrywardelli'm sure they can be improved
01:58:59linuxstb_Just making lcd_send_data2 and lcd_send_data inline will probably help - unless gcc has done it automatically.
01:59:38 Quit BHSPitMonkey (Connection timed out)
01:59:56billytwowillyhey, how does rockbox get around the licencing issues with mp3? you guys obviously didn't buy a licence...
02:00:11zewhat would they need a license for?
02:00:21billytwowillyis the mp3 patent expired/not valid where the server is hosted?
02:00:28zeyou think the various 3rd-party implimentation free decoders have licenses?
02:00:37barrywardellthe lcd is fairly glitchy too
02:00:40zedo you think they need to?
02:00:46billytwowillyI'm not sure how they get around taht either.
02:00:54blxze, in countries with softwae patents, yes.
02:00:58billytwowillywell, under U.S law they would, and probably under Canadian law.
02:00:59zewell afaik
02:01:04barrywardelleg. the statusbar is at the bottom!
02:01:05zethey're non-commercial
02:01:08zewhich might not count for anything
02:01:16billytwowillyoh yah. ok, I forgot about that. most of europe doesn't have software patents.
02:01:22 Quit ben_ (Remote closed the connection)
02:01:22blxze, most people ignore the decoding bit of the patent afaik.
02:01:23zebut then they're also generally developed in countries that don't care
02:01:29blxbut legally it is infrnging
02:02:04zehonestly i'm not sure if independant implimentations are really covered by the patent
02:02:11zeianal, and i don't really know
02:02:21billytwowillyin the U.S, if you can play mp3 you own someone money.
02:02:24Paul_The_Nerdbillytwowilly: If I recall, the possessors of the patent have even explicitly said that they don't intend to pursue "free" decoders for MP3.
02:02:24blxsome gnu/linux distributions have removed mp3 support
02:02:41Paul_The_NerdThere's no legal basis to it, so they still *can* but they've explicitly stated that they don't intend to, if I recall.
02:02:44zebut if there's no trace of the reference implimentation
02:02:50billytwowillyI don't care really. I was just curious. the lack of software patents in europe satisfied my curiousity.
02:02:51zeand especially if it uses different methods to accomplish the same goal
02:03:00zethen i dunno if it could be covered by the patent
02:03:09zeof course as general of patents as they give away these days, it could be
02:03:25billytwowillyyep. most patents are lame.
02:03:51zei'm totally for patents in their original intent
02:03:57zeas an option to an inventor/innovator
02:04:04zebut shits just ridiculous these days
02:04:09zeand totally out of the spirit of the idea
02:04:47zebut yeah, anyway
02:04:48billytwowillyI'm totally for copyright in its original sense as well. i.e 14 year time period with sever limitations on just what can be copyrighted
02:05:19zeand fair use
02:05:40billytwowillyI wouldn't even care too much about fair use if the term was short enough.
02:05:44billytwowillyi.e 5 years or less.
02:05:55zei would
02:05:59zei want backups of shit i buy
02:06:23zeand i want to be able to transfer it to other formats as appropriate for my use
02:06:25 Nick JoeBorn is now known as JBonBike (
02:06:41billytwowillyI just wouldn't buy anything that could be copyrighted... I'd wait for the copyright period to end.
02:07:00zei dunno
02:07:03 Quit [San] (Read error: 110 (Connection timed out))
02:07:07zewaiting 5 years after a cd i want comes out
02:07:12billytwowillythe more restrictive the discoverer of the copyrighted material wants the copyright to be the shorter the term should be.
02:07:15zeto buy it so that i can put it on my mp3 player
02:07:19zeseems kindof ridiculous
02:07:50billytwowillyif I only had 5 years to make money off of something I would be damn sure to have it available in any format you wanted.
02:08:06BigMacanyone know whats up with
02:08:31billytwowillydidn't it get shut down by the RIAA?
02:09:09billytwowillythat's interesting.
02:09:24billytwowillygoogle cache of the webpage url saying it is suspended...
02:09:34zeif you only had 5 years to make money off something and were to be damn sure to have it available in any format one wanted
02:09:48zeyou'd probably also be sure to charge seperately for every format
02:09:57zeso if i wanted to play it on cd and in ogg, i'd have to pay twice
02:10:01zewhich is lame
02:10:02 Quit GliGli ("Miranda IM! Smaller, Faster, Easier.")
02:10:33 Join [San] [0] (
02:10:50billytwowillymaybe ze, but you'd get it after 5 years. there is so much music out there that is good right now anyway.
02:10:58BigMachow can i get snc files then?
02:11:03BigMaccan i make them?
02:11:15 Quit linuxstb_ ("CGI:IRC")
02:11:19BigMacits a lyric file
02:11:26BigMacthat can be read on the ipod
02:11:28billytwowillyno idea. google it.
02:11:38BigMacalready have
02:11:44zebillytwowilly: where is it all? :p
02:12:05zebillytwowilly: how do i seperate it from all the tons and tons of horrid junk out there drowning it out?
02:12:58billytwowillyhmm. using filetype:snc only brings up
02:13:15BigMacthat blows
02:13:30BigMacthen there is no point in the snc viewer
02:13:43amiconnThe h10 lcd frame rates are indeed ouch
02:13:54amiconn(as are the rates on the colour ipods except nano)
02:13:57billytwowillythere has to be more out there. I dunno, I don't care about lyrics though, so I have little knowledge.
02:14:13amiconnMy observation is that things start to get smooth around 50..60fps
02:14:27 Nick BigMac is now known as Bigmac|shower (
02:14:40amiconnThat's the goal - preferably achieved w/o boosting
02:15:20amiconnThat's also why I settled for ~60fps for the iriver remote lcd frame rate with the ticking fix enabled
02:15:31zei always just google for lyrics, personally
02:15:42zei've never found any single lyric site that was always 100% complete and correct
02:15:54zeand honestly most of them have the exact same set of lyrics with the same errors too
02:17:27 Join XavierGr [0] (
02:17:45zeand googling you might stumble upon an artist-specific site by someone who cares about the accuracy of those particular lyrics and might even have them transcribed from an insert (and maybe that alongside a "corrected" version of whats actually being sung, where it differs from the insert)
02:18:52barrywardellamiconn, linuxstb: inline improves things a lot
02:19:12barrywardell30MHz: 1/1 26.0 1/4 104.0
02:20:02barrywardell75MHz: 1/1 39.0 1/4 155
02:20:49barrywardell24MHz: 1/1 22.5 1/4 88.5
02:22:22billytwowillyso does bagder ever hang out here?
02:24:42amiconnbarrywardell: ~+30%, not bad for a relatively simple improvement you can do in C
02:25:13barrywardelli'm sure there's more potential in there
02:25:36amiconnOn some other targets we use assembler to improve things - but I doubt that asm would help much on pp
02:26:04 Part Paul_The_Nerd
02:26:31amiconnIt seems the pp chips have a special interface to the lcd controller
02:26:59amiconnI think this interface is software configurable
02:27:37*amiconn would really like having pp docs available
02:28:30barrywardellamiconn: if you have any tips for optimisation they are welcome. you seem to know a lot about it
02:28:34 Quit sharpe (Read error: 60 (Operation timed out))
02:29:00billytwowillyoooh.. pp=portal player? The chip in the sansa e200 series stuff?
02:30:02amiconnI would like to improve the rates on the greyscale ipods as well. I have one pp target, and ipod mini g2. Improving lcd framerate would allow running the grayscale library w/o boosting
02:30:43 Quit [San] (Read error: 110 (Connection timed out))
02:31:43 Join [San] [0] (
02:32:20amiconnbarrywardell: Well, the lcd update is 'just' pumping data. so there are 2 main spots of possible optimisation: (1) reading from RAM. (2) writing to the lcd controller.
02:33:14amiconnOn pp, (1) will have less effect than it had on coldfire. PP does have a data cache, coldfire does not.
02:34:14amiconn(2) is hard on pp since the interface isn't documented. We stay well below of the lcd controller specs (as far as they are known), especially on the greyscale ipods
02:34:36 Join linuxstb_ [0] (i=5343d4aa@gateway/web/cgi-irc/
02:36:05barrywardellit's slower than ipods with a larger screen and the same cpu so there must be room for optimisation somewhere
02:37:25 Join webguest97 [0] (i=deefbf35@gateway/web/cgi-irc/
02:37:39amiconnI didn't look at the h10 code yet. The interface between pp and lcd controller is slightly different between the various models
02:38:01markunamiconn: I want to make a wiki page that lists which scripts are supported by each font, do you have a suggestion for the name of the page?
02:39:05linuxstb_barrywardell: I've no idea if it will work, but I've just done a quick lcd_yuv_blit implementation for the H10 -
02:40:16amiconnThe old, (comparatively) slow C implementation...
02:40:21barrywardellamiconn: the interface is very similar to the nano
02:40:33barrywardellonly the register defines are different
02:40:54barrywardelllinuxstb_: testing now
02:41:29 Nick _belze is now known as belze (
02:41:30 Nick webguest97 is now known as edwardjin (i=deefbf35@gateway/web/cgi-irc/
02:43:47barrywardellit kinda works
02:44:23linuxstb_How kinda?
02:44:28barrywardellthe colours are all funky playing stacys-176x128.m2v
02:45:18barrywardellabout 14.5 fps
02:45:43linuxstb_Ah, you've got an rgb565 byte-swapped LCD?
02:46:57linuxstb_In the two calls to lcd_send_data - add a swap16() around the parameter.
02:47:03barrywardellpossibly? it's the same lcd as the x5
02:47:38amiconnIt's probably the lcd hookup (the interface I talked about)
02:48:05linuxstb_Looking at lcd-h10.c, you're sending a byte a time to the LCD - so you can choose the byte order yourself.
02:48:12amiconnIirc the ipod photo has the same lcd controller as the H300, yet it also has the rgb data byte-swapped
02:49:44 Quit [San] (Read error: 110 (Connection timed out))
02:50:43 Quit gopp ("gop: god giveths and democarts taketh away!")
02:51:01 Join [San] [0] (
02:51:47amiconnlinuxstb: Btw, is there a reason why lcd_yuv_blit() for ipod calculates rgb555 instead of rgb565 ?
02:52:05barrywardelliirc i did have to change the order of bytes when adapting the x5 driver
02:52:15linuxstb_amiconn: Does it?
02:53:04amiconnI stumbled across it because I wondered why the green bits are only >>15, not >>16
02:53:10linuxstb_I should pay more attention to the patches I commit...
02:53:18barrywardellyeah, bytes were swapped
02:53:24barrywardellworks fine now
02:53:26amiconnAnd btw, there's a lot of unnecessary shifting
02:53:40 Join Davide-NYC [0] (
02:53:52 Join ghode|afk [0] (
02:53:56barrywardell~19 fps
02:55:06amiconnThis yuv_bitmap_part() the C implementation of lcd_yuv_blit() was derived from was never intended for high speed.
02:55:20amiconnIt was intended for reasonable speed and good quality
02:55:33 Join BHSPitMonkey [0] (
02:55:35amiconn...and it's also a bit off for mpeg yuv
02:56:52amiconnThe (normalised) yuv formula is the same for jpeg and mpeg, but the integer ranges are different
02:57:29Davide-NYCQuestion about startup screen:
02:57:47Davide-NYCI've changed the startup info screen bitmap to read:
02:57:49Davide-NYC Please Return To:
02:57:49amiconnjpeg uses 0..255 as (0.0...1.0) for Y and -128..+127 as (-0.5..+0.5) for U and V
02:57:50Davide-NYC Gene Simmons
02:57:52Davide-NYC (555) 555-5555
02:57:53Davide-NYCBut it up for so little that you don't even have time to read it.
02:57:55Davide-NYCCan I insert a delay of x seconds?
02:58:33amiconnmpeg uses 16..235 for Y and -112..+112 for U and V
02:59:14linuxstb_So you're saying we need to completely rewrite lcd_yuv_blit()?
03:00:35linuxstb_Davide-NYC: Yes, add the line sleep(x*HZ); (where x is the number of seconds you want to sleep for) after the logo is displayed.
03:01:07Davide-NYC(total noob) what file should I be looking in?
03:01:45Nico_Pthis makes me think of the jpeg viewer... do you think it's possible to optimise it so that pictures load faster ? or is it already optimised ? and would it be possible to preload the next (and why not the previous) picture in the background so that it displays instantlty ?
03:02:15 Join BHSPitLappy [0] (
03:02:17linuxstb_Davide-NYC: Around line 294 of apps/main.c.
03:02:32Davide-NYClinuxstb: thanks
03:03:24amiconnlinuxstb: The coldfire asm versions already take the difference into account...
03:04:29*amiconn thinks this dreaded inl() and outl() stuff should be KILLED
03:04:49Davide-NYCthere are five instances of show_logo(); in main.c
03:05:11Davide-NYCafter which conditional statement is the show_logo(); I anm looking for?
03:05:22linuxstb_amiconn: I'm all for that, but am no good at inventing names for all the unknown registers...
03:05:41linuxstb_Davide-NYC: The one at line 294...
03:06:07dan_aunknown_0 through to unknown_FFFFFFFF
03:06:15dan_a(for the register names)
03:06:31amiconnThey're not completely unknown
03:06:33 Quit Bigmac|shower ()
03:06:57amiconnAt least we know what hardware block they're associated with
03:07:23amiconn(lcd, ata, ...)
03:08:20linuxstb_True. But anyway, it's already long past my bedtime. Goodnight all.
03:08:38linuxstb_barrywardell: Feel free to commit that lcd_yuv_blit() if you want to.
03:08:44 Quit linuxstb_ ("CGI:IRC")
03:09:48barrywardellwill do along with the inline optimisation
03:10:40barrywardellamiconn: any idea why I'm getting glitchy lcd display?
03:10:51barrywardelleg. the statusbar is at the bottom
03:11:21barrywardellscrolling text is 4 lines lower than where the selected text is
03:12:20Davide-NYCanyone know if that sleep(x*HZ); function works witht he simulator?
03:16:42blxlistening to the smiths on my new iAudio X5L with rockbox. neato! :)
03:18:35 Quit [San] (Read error: 110 (Connection timed out))
03:18:41Nico_Pbarrywardell: could i possibly be of any help on the scrollpad ?
03:18:41 Quit edwardjin ("CGI:IRC (EOF)")
03:19:07barrywardellyou can try :)
03:19:17barrywardelllorijho is working on it at the moment
03:19:18Nico_Pwhere should i start ?
03:19:18 Join [San] [0] (
03:19:36barrywardelltwo possible places are gpio and the adc
03:19:39Nico_Pi'm looking at button-h10.c but there's not very much there
03:19:45barrywardellboth are connected to the scrollpad
03:20:10barrywardelladc channel 4 gives pressure sensitivity
03:20:46barrywardellthe gpio is documented here:
03:22:42barrywardelllorijho has also just found that gpiod is conected to the buzzer that makes the click sound when you touch the scrollpad
03:24:41*lorijho has a metronome with a low volume click :)
03:29:38Nico_Pthanks for the info...
03:30:10Nico_Pbasically you're trying to find which GPIO value is changed when you go up or down with the scrollpad ?
03:30:44barrywardellyes, although we don't know if we will even find one
03:31:38barrywardellthe adc is another possibility
03:32:03 Join JdGordon [0] (
03:32:16Nico_Phmm i'll have a shot at tommorrow when i'm less tired ;)
03:32:34Nico_Panyway congrats on the progress
03:33:15barrywardelli need sleep too
03:33:27 Quit Nico_P (Remote closed the connection)
03:33:49 Join JoeBorn [0] (
03:34:11 Quit JoeBorn (Read error: 104 (Connection reset by peer))
03:34:57 Quit barrywardell ()
03:37:16 Join JoeBorn [0] (
03:38:05Davide-NYCis there a way to "make fullzip" so that the rockbox.iriver (or whichever) file is stored *inside* the .rockbox folder?
03:38:42Davide-NYC(this is of course after extraction)
03:40:16 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox]")
03:40:23XavierGrDavid: So basically you want to move rockbox.iriver inside the .rockbox and inside the zip file right?
03:40:36XavierGryou should see the make code
03:40:52XavierGrI am sure that it can be on command before the final zip command
03:41:13 Quit spiorf ("Read error: 104 (Connection reset by peer)")
03:44:08 Join Davide-NYC [0] (
03:45:52XavierGr[04:40] <XavierGr> David: So basically you want to move rockbox.iriver inside the .rockbox and inside the zip file right?
03:45:52XavierGr[04:40] <XavierGr> you should see the make code
03:45:52XavierGr[04:40] <XavierGr> I am sure that it can be on command before the final
03:46:19XavierGrthis is when you left and I didn't see it David
03:47:02Davide-NYCyeah I had some wierd freeze up and decided to shut down a bunch of apps.
03:47:37***Saving seen data "./dancer.seen"
03:47:48 Quit [San] (Read error: 110 (Connection timed out))
03:49:12Davide-NYC$(SILENT)$(TOOLSDIR)/ $(VERBOSEOPT) -t \"$(ARCHOS)\" -r "$(ROOTDIR)" -f 2 -o $(TARGET) $(BINARY)
03:49:26Davide-NYCthat's some greek to me! :-/
03:49:39 Join [San] [0] (
03:50:37Davide-NYCand for clarity, should all these ARCHOS references really be TARGET or something more generic
03:51:11*Davide-NYC thinks his random thought must seems dangerous to some in here
03:52:01Davide-NYCOK I figure I need to put the BINARY in side a directory
03:53:47 Quit lorijho ("leaving")
03:54:13XavierGr<Davide-NYC> that's some greek to me! :-/ <- LOL I wish those would be Greek, then I could help more :P
03:54:54XavierGrBetter ask the script master (Bagder)
03:55:17 Join _jhMikeS_ [0] (
03:55:51Davide-NYCyup I'm not messing with it. It's too cryptic
04:00:55 Join andy_ [0] (n=andy@
04:01:59 Quit andy_ (Client Quit)
04:03:35 Join Whiffle [0] (n=andy@
04:03:42 Join pixelma_ [0] (i=pixelma@
04:08:42 Quit jhMikeS (Read error: 110 (Connection timed out))
04:09:18 Quit Whiffle ("Leaving")
04:11:55 Quit [San] (Read error: 110 (Connection timed out))
04:11:58 Join [San] [0] (
04:12:59 Quit scorche (Read error: 110 (Connection timed out))
04:20:43 Quit pixelma (Read error: 110 (Connection timed out))
04:20:44 Nick pixelma_ is now known as pixelma (i=pixelma@
04:25:43 Quit blx (Read error: 110 (Connection timed out))
04:27:28 Quit Genre9mp3 (Excess Flood)
04:28:54 Join Genre9mp3 [0] (n=yngwiejo@
04:33:54 Quit Snyper (Read error: 110 (Connection timed out))
04:34:05 Join Snyper [0] (
04:36:34 Quit [San] (Read error: 110 (Connection timed out))
04:38:39 Join [San] [0] (
04:38:47 Quit dan_a (Read error: 110 (Connection timed out))
04:41:10 Quit Genre9mp3 ("I don't suffer from Rockbox psychosis. I enjoy every minute of it.")
04:42:12 Part pixelma
04:45:23 Nick _jhMikeS_ is now known as jhMikeS (
04:52:13 Join Davide-NYC_ [0] (
04:52:13 Quit Davide-NYC (Read error: 104 (Connection reset by peer))
04:52:15 Nick Davide-NYC_ is now known as Davide-NYC (
04:53:54Davide-NYCI have a smal bug report to file on flyspray on the new FM recording screen, does it go in recording or user interface?
04:55:12 Quit CriamosAndy (Read error: 104 (Connection reset by peer))
04:55:41Davide-NYCAlso FS #5508 — Recording interference with remote plugged in should be closed. Mmmm commited a workaround a while back
04:56:06 Join Criamos [0] (
04:59:31 Quit [San] (Read error: 110 (Connection timed out))
04:59:36 Join [San] [0] (
05:01:20jhMikeSCheck out this bug (it's got me goin' wtf?):
05:01:32 Join lucas42 [0] (
05:06:20Davide-NYCwoah. I don;t even understand what he's saying in that bug report. re-reading
05:07:43jhMikeSWhat needs clarifying? I'll edit it to make it better.
05:09:43Davide-NYCLoaded it in CoolEditPro −− which one the copy or the original?
05:10:00Davide-NYCalso, did you save the copy to your X5 or to your PC?
05:12:00Davide-NYCjhMikeS: ciould it be some sort of header information that's being saved incorrectly?
05:12:30*Davide-NYC I tells myself to be quiet.
05:12:42jhMikeSthe copy
05:12:59jhMikeSno...I said the files were identical right?
05:13:01Davide-NYCthat's pretty clear from context
05:13:29jhMikeSSo how could they play differently unless the filesystem gives playback inaccurate info?
05:14:20Davide-NYCso if you play the X5 stored file on the PC it's 16s and the PC stored file played throught the same software is 18s.
05:14:25jhMikeSI saved the copy right back on top ... all done on the x5 through a drive letter in explorer
05:14:41jhMikeSI played back both times on the x5
05:14:56jhMikeSA PC based player it doesn't matter
05:15:13Davide-NYCis it always 2 seconds difference?
05:15:24*Davide-NYC needs to do some testing
05:15:26jhMikeSYes on all recorded formats
05:15:55jhMikeSI used WAV cause it's easy to verify in a hex editor too...
05:16:15jhMikeSRecording writes the file itself perfectly.
05:16:22Davide-NYCdoesn't a binary file comparison to the trick?
05:16:29jhMikeSDid that too
05:16:47jhMikeSTextPad lets you compare two files and it said "identical".
05:17:31Davide-NYCI'm still unclear on something. In what instance does the file play 'short'?
05:20:33jhMikeSThe one directly made by the recording feature that is
05:20:45Davide-NYCOK I test
05:21:02Davide-NYCjhMikeS: did you try other inputs aside from FM?
05:21:15Davide-NYCbecause that is fairly new (sort of)
05:21:17jhMikeSWell...if you get that 1/100 chance that is shows to the end, try again.
05:21:23jhMikeSI wrote it
05:21:33jhMikeSThere's no encoding difference.
05:21:44jhMikeSThe keys is that the files are the matter what source
05:22:13Davide-NYCOK, what and how should I test. I'm willing to put in an hour
05:22:39Davide-NYCI have the tools to do all types of iRiver H1xx recording (including optical)
05:23:07 Quit [San] (Read error: 110 (Connection timed out))
05:23:09jhMikeSJust record anything at all...then exit recording and play it back. Watch the progress compared to the file length displayed
05:23:23jhMikeSDo you have an x5? This could be an x5 only bug btw
05:23:46Davide-NYCI'm an iRiver ihp-140 only kind of guy
05:23:52jhMikeSSPDIF should be functional
05:24:10Davide-NYCyou mean bug-less?
05:24:26jhMikeSShould be...yes...others have tested it for me
05:24:35Davide-NYCquestion: Does the rockbox simulator display any of the new FM recoding features?
05:24:44Davide-NYCMine does not
05:24:59 Quit lucas42 ("time for bed")
05:25:00Davide-NYCI have a small UI bug to report directly to you then.
05:25:04jhMikeSno radio input!
05:25:12 Join [San] [0] (
05:25:18Davide-NYCbut I can go to the FM screen otherwise
05:25:34jhMikeSYou can select FM Radio from the recording option too
05:25:36Davide-NYCmenu −−> FM Radio
05:26:08Davide-NYCI only get mic, line and digital
05:26:21jhMikeSIt's defined for your player...ok now that's a bug
05:26:29 Join MrShlee [0] (
05:26:48Davide-NYCshould I put it up on flyspray?
05:26:55Davide-NYC(my pleasure)
05:27:05jhMikeSNah...I'm gonna look right now about it.
05:27:18jhMikeSBUt you can still do it from the radio screen?
05:27:34Davide-NYCI can enter the MENU −−> FM Radio screen no problem
05:27:35MrShleeHey, Ive just got my Iriver H340 back from repair - This is my third Iriver and I've always had Rockbox.. Should I install it again?
05:27:52jhMikeSYou should probably use the latest CVS yes.
05:28:11jhMikeSOther's have not had that behaviour to my knowledge
05:28:22Davide-NYCI just DLed the latest
05:28:40Davide-NYCI'll rename the rockbox-devel dir and reDL from scratch
05:28:53 Quit XavierGr ()
05:29:12Davide-NYCoh wait! it might be the remote button patch that's breaking the sim.
05:29:18Davide-NYCtesting... :-/
05:30:08jhMikeSI think I found the reason you're not getting the FM Radio input option.
05:30:16Davide-NYCyou're fast
05:30:54jhMikeSI call radio_hardware_present() to decide to add the menu item. You're player must report it as not present.
05:32:34 Join infamis [0] (i=4b0377fc@gateway/web/cgi-irc/
05:32:53*jhMikeS is going to have to ask on which players that function should be used
05:33:13infamiscygwin gcc only goes up to 3.4.4
05:33:21infamiswiki recommends 3.4.6
05:33:37infamiserror in mp3_enc w/ cygwin
05:34:19Davide-NYChow can I test if the player report radio hardware or not? debug?
05:35:49Davide-NYCmenu −−> info −−> debug −−> FM Radio == HW detected: yes, Philips regs: 000037B0
05:36:02 Join Crackerizer [0] (n=root@
05:36:05Davide-NYCthis is on an iRiver H140
05:36:48JdGordonDavide-NYC: what r u trying to do?
05:37:10 Join _Veseliq_ [0] (n=veseliq@
05:37:31Davide-NYCI'm just poking around in the debug menu and reporting back what I read about FM hardware
05:38:15Davide-NYC(in case that helps) the "−−>" mean 'go into this menu' in my horrible lingo
05:38:18jhMikeSthat's wierd...I'll look at the debug menu code
05:38:26jhMikeSx5 has no trouble
05:40:05Davide-NYCjhMikeS: I'm using today's version 20060830-2126
05:41:06jhMikeSthat'll work
05:41:31jhMikeSIf you can record the radio or whatever see if the shortened progress thing pops up
05:42:07infamisanybody know the define that tells what gcc version you have ?
05:42:11Davide-NYCI'm just first confirming that my sim bug report has nothing to do with the sim-remote-buttons patch first
05:42:41 Quit MrShlee ("Changed by popular Request - "I am the heart of your darkness, you stinking oompa-loompa morons!"")
05:44:42 Quit [San] (Read error: 110 (Connection timed out))
05:44:48 Join [San] [0] (
05:46:13Davide-NYCConfirmed: using a clean CVS downlaod, building the sim, I get no option for FM Radio in the Recording "Source" menu.
05:46:58jhMikeSYou mean on the H140 right? Not sim? (I shouldn't ask such a dumb q)
05:47:07 Join Variable [0] (
05:47:14Davide-NYCno no, I'm specifically talking about the sim
05:47:22Davide-NYCI see this stuff on the actual target
05:47:40***Saving seen data "./dancer.seen"
05:47:44jhMikeSThe sim shouldn't have the menu item
05:47:58Davide-NYC(humbly) why not?
05:48:01jhMikeSIt's specifically not defined on sim builds
05:48:11Davide-NYCthe FM Radio screen exists in the sim
05:48:23jhMikeSIt would just be soft of a dummy menu item then
05:48:34Davide-NYCyeah, just for UI testing
05:49:17Davide-NYCI'm focused on recording and trying to get the recording experience unified (enough) across all targets and remotes
05:49:32jhMikeSI'm trying to remember...there's the actual source switch details that get included if HAVE_FMRADIO_IN is defined...but I could #ifndef SIMULATOR out the hw parts.
05:49:50Davide-NYCgreek bro, I no speaky C that good
05:50:19Davide-NYCbut I feel that (with a little patience on the dev end) I can be of actual help
05:50:35jhMikeSjust try recording something! pretty please! I want to see if you get that progress bug
05:50:43Davide-NYCam doing!
05:50:56*jhMikeS shuts his mouth
05:51:16jhMikeSI'm adding bit rate selection for SWCODEC right now
05:51:28jhMikeSI mean sample rate
05:51:35infamisanyone know where to find the list of defined registers for gcc 3.4.4 (m68k-elf)?
05:52:15jhMikeSMFC5249.h? I think
05:52:41jhMikeSmcf5249.h more correctly
05:54:29infamisoops, sorry I meant internal registers for gcc
05:54:48*jhMikeS thinks that it's wierd that in late June he knew nothing more that knowing _about_ Rockbox
05:55:31jhMikeS%acc0 - %acc4
05:55:34infamiswhere can I find that in gcc though
05:55:52 Join mubix [0] (
05:55:59jhMikeSscheiss...I never looked there I just peeped at various assemly files.
05:56:14infamislike gcc 3.4.4 m68k-elf recognizes registers a, b, c, d, e, etc...
05:56:38mubix[INFO] Seeking to sector 2009291924
05:56:38mubix Seek error An attempt was made to move the file pointer before the beginning of
05:56:38mubix the file.
05:56:47infamis...while 3.4.6 recognizes g, h, and i in addition to a - e...
05:56:56mubixmeant to ask the question before posting
05:57:06jhMikeSinfamis: now I'm not sure what you mean...what do they represent...I'm fairly new to gcc anyway
05:57:27infamisok I'll start over.
05:57:53infamis3.4.4 has a problem finding a certain register that mp3_enc uses some inline assembly (inside the .c file)
05:58:16jhMikeSa file I added...
05:58:46infamisok, so I'm trying to find out which is the "bad" register. going through gcc's docs to see what registers are recognized, but I can't find said docs
05:59:16jhMikeSwhat version comes with the CVS?
05:59:42jhMikeSWhat does yours spit out when encountering that code?
05:59:59 Nick [H5N1] is now known as H5N1 (
06:00:21*jhMikeS knows you must mean the "filter_subband" function
06:01:08jhMikeSNot on their site?
06:01:49infamisyup, filter_subband
06:02:03infamis"can't find a register in class 'addr_regs'"
06:02:40infamisas for the version, only up to 3.4.4 is available through cygwin
06:03:25jhMikeScomment out: asm volatile ("movem.l %%d0/%%d2-%%d7/%%a2-%%a7,%0\n" : "=m" (*(int*)reg_buff));
06:03:27 Quit [San] (Read error: 110 (Connection timed out))
06:03:29jhMikeSand asm volatile ("movem.l %0,%%d0/%%d2-%%d7/%%a2-%%a7\n" : : "m" (*(int*)reg_buff));
06:03:50 Quit Rudy4Pez (Read error: 110 (Connection timed out))
06:04:35 Join [San] [0] (
06:04:40infamisalright, trying
06:04:49jhMikeSIf that's not it it's probably the (e)mac stuff
06:04:57infamisbut I don't think that's it cause those registers seem normal
06:05:34infamisstill get the same err
06:05:35jhMikeSbut you'll know for sure it's not that and can proceed
06:06:28jhMikeSis it in the first or second sequence of assembly?
06:06:54jhMikeStry commenting one then the other
06:07:22 Quit Crackerizer ("using sirc version 2.211+KSIRC/1.3.12")
06:11:16infamisstill trying...
06:13:24Davide-NYCjhMikeS: did some tests only on the iRiver H1xx
06:13:38Davide-NYCI recorded mic, line and FM to WAV. they play back *on the player* correctly as far as I can tell.
06:13:50Davide-NYCI recorded mic, line and FM to MP3. they play back *on the player* correctly as far as I can tell.
06:14:24 Quit idnar (Nick collision from services.)
06:14:25Davide-NYCthey play correctly and report postion correctly *on the player*
06:14:27 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
06:14:55jhMikeSI have gcc version 3.4.6 from CVS
06:15:04Davide-NYCI'm now going to play them from the H1xx through some PC software.
06:15:21Davide-NYCwhat else should I do to try to confirm that this bug is not present on the H1xx
06:15:36jhMikeSDavide-NYC: Like I said could be unique to x5
06:15:51Davide-NYCI understand, what can I do to narrow it down further?
06:16:07jhMikeSDavide-NYC: I don't think there anything to do but play a file back on the player that was recorded on the player
06:16:28Davide-NYCand you files ended 2sec early when played back on the X5?
06:16:45jhMikeSThe progress bar should stop advancing about 2s before the end
06:16:55jhMikeSThe file plays all the way through
06:17:08jhMikeSThe progress bar on the WPS stops moving 2s early
06:17:28jhMikeSBut not on the identical resaved copy
06:18:33Davide-NYCMy progress bars move steadily throughout.
06:18:45jhMikeSYou've done this many times?
06:19:06jhMikeSIt _always_ happens on the x5
06:19:19jhMikeSThat does narrow it down :)
06:19:23Davide-NYC3 to WAV 3 to MP3 one source type per recording.
06:19:40Davide-NYCI'll try some short recordings, mine were all over 30sec
06:19:48jhMikeSThe source makes no diff...nor the length
06:20:09jhMikeSBut to be sure...
06:20:17*jhMikeS is sure for his player
06:20:23Davide-NYCyou should post in the recording section of the forum, tell plp what to do and report back...
06:20:50Davide-NYCI'll do whatever you tell me to do as long as I believe it to be rockbox related
06:21:41Davide-NYCI'm compiling a bug report, your new encoding stuff has intoduced a few tiny bugs in the WRS.
06:25:21jhMikeSThe quality of the display of the format was not worked on in detail.
06:26:00jhMikeSDavide-NYC: thanks for taking the time with this
06:26:48Davide-NYCare you crazy? thank you for making me 100% RB now that real-tine encoding exists!
06:27:18jhMikeS:) It's going to be real-time-er soon
06:27:33Davide-NYCwhat's not RT? MP3?
06:27:41jhMikeSDoing the sample rate stuff is hard
06:27:59jhMikeSMP3 is about the fastest...wavpack is the slow one
06:28:40 Quit [San] (Read error: 110 (Connection timed out))
06:28:56 Join [San] [0] (
06:28:59jhMikeSpreglow said to implement a new digital resampler for non-native rates cause he's using linear interpolation atm
06:29:08Davide-NYCno need for WV, just low MP3 for meetings, otherwise I go WAV
06:29:35Davide-NYCsounds fancy! let me know when you all need me to do some manual heavy lifting
06:29:51Davide-NYCwhat about the FM sim business? Flyspray?
06:30:00jhMikeSBeing able to use 11025 would help for that stuff but that encoder doesn't support that sampling rate.
06:30:21jhMikeSI have a just want the menu item displayed for the sim and that's it?
06:32:01jhMikeSIt's an mpeg 1 encoder and you need mpeg 2.5 to support the super low sampling rates
06:32:31jhMikeSI'll have to see if I can get a more advanced GPL encoder
06:33:12 Nick H5N1 is now known as [H5N1] (
06:34:08 Nick [H5N1] is now known as H5N1 (
06:34:27 Nick H5N1 is now known as [H5N1] (
06:34:50 Nick [H5N1] is now known as h5n1 (
06:34:57jhMikeSDavide-NYC: I can add the menu item right now! ;)
06:35:13 Nick h5n1 is now known as [H5N1] (
06:35:20 Nick [H5N1] is now known as h5n1 (
06:35:23Davide-NYCsweet, thanks. I'm going to post three smal bugs in the forum. stay tuned
06:35:51jhMikeSForum or FS?
06:36:01jhMikeSPost the task real quick here
06:37:01 Nick h5n1 is now known as [H5N1] (
06:41:55jhMikeSmenu option: done
06:42:58 Quit _Veseliq_ (Remote closed the connection)
06:45:14infamiswell, this is what I got....
06:47:04Davide-NYCso if i "cvs up -dP" and re-make the sim I get FM now?
06:47:25 Quit [San] (Read error: 110 (Connection timed out))
06:47:36infamisthe input operands to the asm (line with [yint] "+a" (int), [xint] "+a") around 2476-ish, seems it doesn't have any registers left?
06:48:07infamisif I took off the a5 reg clobber declaration, it compiles fine.
06:48:36infamissame result if I only had one "+a" output operand
06:48:46 Join [San] [0] (
06:49:03jhMikeSinfamis: in which block?
06:49:23infamisfirst one
06:50:03jhMikeSDavide-NYC: The recording screen for the radio on software codec platforms is the same screen...
06:50:19jhMikeSYou just don't get a source option
06:50:53jhMikeSinfamis: hmmmm....shouldn't matter since regs are restored any way after that block
06:51:30jhMikeSIt calls recording_screen(false) for normal operation and recording_screen(true) when in FM radio screen
06:51:43jhMikeStrue means = "no source selection"
06:52:00Davide-NYCin the SIM, if I go to menu−−>recording−−>recording settings−−>source I do not get a FM Radio option?
06:52:19jhMikeSSo when you're doing it directly from the fm radio screen you cannot change sources...this must be thae case
06:52:28Davide-NYChow to I test the UI for the WRS when source = FM Radio?
06:52:48jhMikeSinfamis: why not just get the 3.4.6 compiler and assembler from the CVS tree and use that?
06:53:11jhMikeSDavide-NYC: ??
06:53:18infamisthere is none for cygwin...
06:53:28Davide-NYCOK I think we're having two *slightly* different conversations
06:53:31jhMikeSYou can go to the recording screen from the FM Radio menu
06:53:52jhMikeSLong select for x5
06:54:04jhMikeSDunno what that is for you
06:54:15Davide-NYCchacking the manual
06:54:17jhMikeSinfamis: I recommend using vmware
06:54:59jhMikeSI set my vmware memory to 48MB. You can edit this in it's config file before starting it.
06:55:08jhMikeSEven lower may work too emulate linux?
06:55:55jhMikeSinfmais:'s fast as hell...I did cygwin at first and though I would die waiting for each compile
06:56:17infamisI actually have a dual boot with gentoo but I'm 90/10 windows/linux (when it comes to actually doing stuff).
06:56:20jhMikeSThere's info in the wiki about's really nice
06:57:02jhMikeSI think RB has moved somewhat beyond the cygwin setup
06:57:13jhMikeSYou need the newer tools
06:57:50Davide-NYCfirst time making manual
06:57:51jhMikeSOk...are you laughing with me or at me?
06:57:52infamisI think RB has fully embraced cygwin ;)
06:58:12Davide-NYCyou said 'tool'
06:58:37Davide-NYCmaking manual for the first time, what happens after 'make'?
06:59:02infamisI laughed cause it sounds like you're saying cygwin is old & outdated like I'm still using windows 95
06:59:42jhMikeSinfamis: well...if you could build the latest TOOLs from the CVS on cygwin maybe that would work?
06:59:58 Quit webguest70 ("CGI:IRC 0.5.7 (2005/06/19)")
07:01:23infamiswhat latest tools? I always get fresh cvs updates
07:01:54 Quit Criamos ("( :: NoNameScript 4.02 :: )")
07:01:59infamislatest tools stuff I got was for the gigabeat & mi4
07:02:00jhMikeShmmm...I think you need to get the source and gcc m68k-gcc-elf and m68k-gcc-as on cygwin
07:03:21infamisahhh! forgot all about the crosscompiler is actuall m68k-blablablah
07:03:27infamisversion of that is 3.4.5
07:04:26jhMikeSThe version on wmware is 3.4.6
07:04:58 Quit infamis ("CGI:IRC (EOF)")
07:05:01jhMikeSThe assembler is 2.16
07:05:16 Join infamis [0] (i=4b0377fc@gateway/web/cgi-irc/
07:05:59*jhMikeS is gonna read Davide-NYC's report now
07:06:05infamisoops... so you're saying grab m68k-elf-gcc-3.4.6 source
07:06:30Davide-NYCleave him alone so he can read my report! (just kidding)
07:07:35 Quit Drkepilogue ("CGI:IRC (EOF)")
07:08:04 Quit [San] (Read error: 110 (Connection timed out))
07:10:09infamishey jhMikeS, got a new problem for ya....
07:10:18Davide-NYCme too!
07:10:31infamisjust playin....knew you'd like that Davide-NYC....
07:10:47 Quit rotator ("zzzzzzzzz")
07:12:43infamiswell, if you care jhMikeS, you could use this for those of us on cygwin & m68k...3.4.5; only have that reg clobber list #if gnu >= 3.4.6 (since they get restored from the reg_buf[] anyway)...
07:12:59infamis#if __GNUC__ > 3 || (__GNUC__ == 3 && (__GNUC_MINOR__ > 4 || __GNUC_MINOR__ == 4 && __GNUC_PATCHLEVEL__ >= 6))
07:13:09infamis : "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7", "a2", "a3", "a4", "a5"
07:14:53infamiswell, I have school in 7 hours....time to sleep
07:14:59 Quit infamis ("CGI:IRC 0.5.7 (2005/06/19)")
07:16:08 Join ksure [0] (
07:20:17 Join Crackerizer [0] (n=root@
07:20:44Davide-NYCthere it is.
07:20:47 Join [San] [0] (
07:26:26 Quit [San] (Read error: 104 (Connection reset by peer))
07:26:35jhMikeSinfamis: of course I care...this is the only problem you've encountered I take it?
07:27:14jhMikeSoh...he quite
07:27:33jhMikeSDavide-NYC: I'm responding to the other thing! :)
07:28:09Davide-NYCwhich one, the forum post or the FS bug report?
07:32:02jhMikeSOk...FS after I answer nature's call....:)
07:35:13 Quit ksure ()
07:36:53 Join [San] [0] (
07:39:32Davide-NYCjhMikeS: you are the man ( assuming you are male) thanks fo rall your effort and expertise.
07:39:46Davide-NYCMe, I gotta go to sleep... 'night.
07:39:57 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox]")
07:40:18jhMikeSyou're welcome
07:40:49*jhMikeS should read those "has quit IRC" things :D
07:46:50 Join carini [0] (i=chatzill@gateway/tor/x-897b86d353ad6751)
07:47:43***Saving seen data "./dancer.seen"
07:48:36 Part carini
07:48:53 Join carini [0] (i=chatzill@pdpc/supporter/active/carini)
07:49:37 Join Jazz00006 [0] (
07:50:26 Quit carini (Client Quit)
07:50:43 Part Jazz00006
07:50:51 Join _Veseliq_ [0] (n=veseliq@
07:50:58 Join Jazz00006 [0] (
07:53:49 Join webguest82 [0] (i=ca068a29@gateway/web/cgi-irc/
07:54:27 Quit _hotwire_ (Read error: 110 (Connection timed out))
07:54:34 Quit webguest82 (Client Quit)
07:56:03 Quit Jazz00006 ()
08:01:47 Join itsmeguys [0] (
08:02:01itsmeguyshello eveyrone
08:02:40 Join theli_ua [0] (
08:03:04itsmeguysi now have rockbox running on my h10 :D
08:03:58itsmeguysits pretyy kool
08:04:06itsmeguyshavent had much of a chance to use it but
08:04:12itsmeguysi had no pbattery :(
08:04:24 Quit JdGordon ("Konversation terminated!")
08:04:50 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
08:05:05 Quit [San] (Read error: 110 (Connection timed out))
08:06:54 Join [San] [0] (
08:07:46itsmeguysicatcher wps
08:07:49itsmeguyswhere can i find that?
08:09:05tuwe|awayisn't it included?
08:09:12 Nick tuwe|away is now known as tuwe (n=the_fuch@
08:09:22itsmeguysdw, found it :D
08:10:09itsmeguyscan someone help me?
08:10:17itsmeguyswhere do i put this wps?
08:10:39 Quit Snyper (Read error: 110 (Connection timed out))
08:10:55 Join Snyper [0] (
08:11:11itsmeguysi assume i just out them into the same folders as rockbox
08:11:50amiconnjhMikeS: I had a look at the register list problem in mp3_enc.c
08:12:51amiconngcc 3.4.4 not being able to compile this is clearly a gcc glitch. There is exactly one A register left, which is necessary to keep the stack pointer
08:13:43amiconnBUT - the way this block and its surrounds deal with clobbered registers is way odd, and afaics reduces performance
08:14:46amiconnAn asm block specifies a clobber list _for gcc_, i.e. so that gcc knows which registers to save/restore
08:15:09 Join Coldtoast [0] (
08:15:16amiconnSo I'm puzzled about the 2 blocks dealing with reg_buff
08:15:21Coldtoasthey. a friend of mine just installed rb on his h300
08:15:33amiconnThey shouldn't be necessary at all
08:15:46Coldtoasthe says in usb bootloader mode, his computer isn't detecting the h300
08:15:59Coldtoastby default, does the h300 enter Charge mode or something?
08:16:32LinusNColdtoast: that is a bug in the bootloader
08:16:33amiconnThe computer not detecting the h300 in bootloader usb mode happens for a few people, and it's a mystery
08:16:48amiconnLinusN: You found it?
08:16:57Coldtoastoh? ok. does it happen EVERY time usually?
08:17:26amiconnAfaik it does, for those few devices where it happens.
08:20:46Coldtoasthis first impressions aren't very favourable unfortunately :/
08:21:02Coldtoasthopefulyl he'll stick with it and find what rb offers to MORE than make up for things
08:21:26Coldtoasthe's not actually gotten to USING it yet. just the interface he's not enjoying
08:22:07ColdtoastI'm finding rb is better suited to the ipod then my iriver actually
08:22:26Coldtoastseems to be more intuitive to use on the ipod
08:27:58LinusNamiconn: no, i haven't found it
08:28:55 Quit [San] (Read error: 110 (Connection timed out))
08:29:37tuweColdtoast: maybe you could recommend a different theme to your friend?
08:31:21 Join JdGordon [0] (
08:31:32 Join [San] [0] (
08:33:08Coldtoastit's not the WPS
08:33:21Coldtoastit's the button mappings really
08:33:31Coldtoasthe's reading the wiki now tho
08:33:45Coldtoasthe already knows the wps is skinnable
08:34:24itsmeguyswhat do i do with that?
08:34:32 Join ender` [0] (i=null@
08:34:55itsmeguysor, how do I update rockbox on my h10 :D
08:35:14tuweis this a build log?
08:35:59Coldtoasthey... is rb ever likely to implement something like smartplaylists?
08:36:27itsmeguystuwe, I have no idea
08:36:28 Quit Crackerizer ("using sirc version 2.211+KSIRC/1.3.12")
08:36:34tuweitsmeguys: seems as if you just typed make in the console. try make zip afterwards, and unzip the resulting zip file into the root dir of your player.
08:37:45itsmeguysbasically, Rockbox on h10 is farely new, right? so, I download rockbox etc, and there is no sound. but barry (the dev of rockbox fro h10) said it has been fixed. I was wondering if it is in some update or something. has that got anything to do with cvs?
08:38:02Coldtoasth10 onlu JUST got sound
08:38:05Bagderthere is sound
08:38:21Bagderget the package from the cvs build page
08:38:34tuweColdtoast: i think they are already implemented. i made some changes and now i have "search by artist", "search by album", etc., "most played tracks", ...
08:38:36itsmeguyswhat exactly do i click?
08:38:52Bagderdownload the package for your player
08:39:16itsmeguysis this where i am supposed to be:
08:39:33Bagder"Bleeding edge builds"
08:39:39Coldtoasttuwe: that's not smart playlists
08:39:39itsmeguysok thank you man ...
08:39:59itsmeguysbut I have no idea what to do with the text etc.
08:40:27Bagderthe text?
08:40:42itsmeguysdamn why am i soo stupid. :S
08:40:52 Join stamppot [0] (i=d4883814@gateway/web/cgi-irc/
08:40:59tuweColdtoast: what is it then?
08:41:03itsmeguysok, so i click the highest zero next to h10, right? do i right click save as or something
08:41:31itsmeguysyou see
08:41:32tuwecoldtoast: ok search by.. is not smart playlist, but i thought stuff like most played tracks is.
08:41:37itsmeguysive already downloaded the stuff down the bottom
08:41:43itsmeguysand i have no sound
08:41:44Bagderthen you have it
08:41:51itsmeguysbut no sound?
08:42:01Bagderthis package is updated at every change
08:42:01 Join Rudy4Pez [0] (
08:42:04tuweColdtoast: ok i see.
08:42:07Bagderit depends on _when_ you downloaded it
08:42:15Coldtoasta little app to generate them would be neat, rather than rb doing it I guess.
08:42:21itsmeguysbut i downloaded it this morning
08:42:26itsmeguysso its not that old, is it?
08:42:36Bagderwhen was "this morning" to you?
08:42:53itsmeguyswell its 4:13pm now
08:43:04itsmeguysso bout 7 hours ago
08:43:04Bagderthen I'd say it has been fixed since
08:43:05JdGordonitsmeguys: adelaide?
08:43:25itsmeguyswell thank you very much bagder :D
08:43:28itsmeguysi hope it works :D
08:44:26 Quit Variable (Read error: 110 (Connection timed out))
08:45:29 Quit Seed (Nick collision from services.)
08:45:37 Join Seed [0] (i=ben@
08:45:52tuwehas anyone tried out the new album art patch yet?
08:48:40itsmeguysit workde
08:53:59 Join webguest35 [0] (i=c0647cda@gateway/web/cgi-irc/
08:54:44 Join Molotov1 [0] (n=Molotov@
08:54:56 Quit Molotov1 (Client Quit)
08:57:33 Quit [San] (Read error: 110 (Connection timed out))
09:01:04 Join [San] [0] (
09:04:23 Join dan_a [0] (n=dan-mirc@
09:09:56 Quit Pyromancer (Read error: 104 (Connection reset by peer))
09:15:04tuwei wonder how i can simulate the hold switch in the ipod mini 2g simulator.
09:15:06Lynx_wow, does rockbox already run on the H10 in a useable form?
09:16:57Lynx_that was fast...
09:18:04Lynx_i remember a time when portal player chips were a bad thing to have ;)
09:18:22amiconnThey're still not the best choice
09:18:54Bagderwe don't have as good run-time on those players
09:20:26Lynx_is that because there would be a single fix, like on the h300, that has not been found/done yet?
09:20:51theli_uatuwe, you cant
09:21:03BagderLynx_: very possible, yes
09:21:29 Join xorAxAx_ [0] (n=xorAxax@moinmoin/coreteam/alexander)
09:21:53 Quit xorAxAx (Read error: 104 (Connection reset by peer))
09:21:54amiconnI don't think it's a single fix for pp
09:21:57 Nick xorAxAx_ is now known as xorAxAx (n=xorAxax@moinmoin/coreteam/alexander)
09:22:22amiconndan_a: When you did the scanrate measurements, how warm was your unit?
09:22:26BagderI mostly mean that I don't think it is a question of optimizing
09:22:35amiconnI am asking because the refresh rate is temperature dependent
09:22:54amiconnBagder: I think it's mainly a question of optimising
09:23:01Lynx_Bagder: that's what my question was, ok
09:23:15Bagderamiconn: I think the difference is too big for that
09:23:30amiconn...and maybe some hardware features we don't know how to use yet
09:24:15amiconnIiuc the PP5022 (and probably PP5021) have voltage scaling to reduce core voltage when running at low clock, and that voltage scaling isn't automatic
09:24:43 Quit ToyKeeper (Read error: 113 (No route to host))
09:25:12amiconnBut our mp3 decoder clearly needs optimisation; on arm mp3 currently needs more cpu power that ogg vorbis, which is odd
09:27:21amiconnOn coldfire, mp3 uses only about half the cpu power of ogg vorbis with an optimised decoder
09:27:36 Quit [San] (Read error: 110 (Connection timed out))
09:28:01dan_aamiconn: I can't remember. I think I have done tests both immediately after booting up and when it has been playing MP3 for a while, both with the temperature in the room being around 22C
09:28:45amiconnOk, so medium temp
09:29:31 Join [San] [0] (n=San@
09:31:01amiconnI get 87.9Hz immediately after bootup (room temp around 23C), 87.7Hz when 'heating' the mini above 30C (putting it on top of the laptop for a while), and 88.3Hz after putting it in the fridge for a while
09:31:38amiconnSo I'd say the scan rate is the same for the full size & mini greyscale LCDs
09:32:09amiconnThere are tolerances of course; lcd controllers usually use simple RC oscillators
09:32:27amiconnThe scan rate on archos varies a lot more with temperature.
09:34:23 Join ootput [0] (n=ootput@tor/regular/five0)
09:40:34 Join aliask [0] (
09:43:11 Join pondlife [0] (
09:44:52 Join eGen [0] (i=generat0@
09:45:02 Quit eGen_ (Read error: 104 (Connection reset by peer))
09:47:47***Saving seen data "./dancer.seen"
09:47:51 Quit Seed (Nick collision from services.)
09:47:57 Join Seed [0] (
09:51:38 Join vo|t [0] (
09:53:15 Quit Coldtoast ()
09:58:16 Quit [San] (Read error: 110 (Connection timed out))
09:58:35 Join [San] [0] (n=San@
09:59:29jhMikeSamiconn: are can the clobber list be removed safely? infamis said it worked without it. those registeres are deliberately saved/restored anyway.
10:01:14 Quit [HO]vo|t (Read error: 110 (Connection timed out))
10:05:44jhMikeSamiconn: I do find it odd to have to retore those registers if they're on the clobber list as it is.
10:06:52 Join Siku [0] (
10:07:21JdGordonamiconn: which actions need to be #ifdefed out for the archos targets? do i just do all the ones that havnt been used in the keymap file for the target? also did u get a chance to try the shutdown fix?
10:08:36 Join webguest64 [0] (i=c27f0812@gateway/web/cgi-irc/
10:09:13webguest64markun: Are you here?
10:10:05 Join _FireFly_ [0] (
10:10:43webguest64markun: if yes: please change the last line in chardef of W in nimbus-12 from F380 to FF80. That way it looks better.
10:11:01markunI'll try it
10:11:32markunwebguest64: do you have more suggestions or is it the only problem you found so far?
10:14:08 Join safetydan [0] (n=dan@
10:14:56markunwebguest64: you are right, it makes the W the same as the w and the W and w from the nimbus-14 font
10:16:33 Quit [San] (Read error: 110 (Connection timed out))
10:16:41webguest64markun: hmm... I see something else. The lower right curves of V and W are different. I think we should unify them. I like V better. So we should make W like V. That means, the last two lines in W should be CC80 and FF00. In both nimbus-12 and nimbus-14. What do you think?
10:16:41 Join [San] [0] (n=San@
10:18:54 Quit dan_a (Read error: 110 (Connection timed out))
10:19:27webguest64markun: if we'd make V like W (i.e. the other way around) then V would be almost indistinguishable from U
10:23:27webguest64markun: BTW, I like your changes (tried them)! The only thing I'm not sure of is the width of the space char (4 -> 3).
10:26:58 Join tuwe_ [0] (n=the_fuch@
10:30:28 Quit tuwe_ (Read error: 104 (Connection reset by peer))
10:31:04 Join tuwe_ [0] (n=the_fuch@
10:31:07markunwebguest64: yes, space is perhaps a bit small now
10:31:07webguest64 markun: the same (make W like V) applies also to v and w, i.e. the last two lines for w should be CC80 and FF00
10:31:30markunand also to all the derived W glyps in nimbus 14
10:31:53markunI'm also adding some nice quotation marks now
10:32:13 Quit tuwe_ (Remote closed the connection)
10:32:43webguest64markun: fine!
10:33:04 Join the_fuchs [0] (n=the_fuch@
10:34:09 Join dan_a [0] (
10:35:05pondlife.amiconn: around?
10:35:13pondlifeamiconn: around?
10:35:34pondlifeOr anyone who really likes to keep the binary small...
10:35:54JdGordonyou got an idea to shrink it?
10:36:09pondlifeNope, sadly... just a question. Is there a problem with introducing lots of logf() calls?
10:36:18pondlifeI assume not, as they are #defined out in non-logf builds
10:36:24pondlifeBut just wanted to check
10:36:33JdGordonye, i dont tihnk there i
10:36:44JdGordonbut it makes the code messy (imo)... but no big deal
10:37:16pondlifeI have put logging on all the audio/voice/codec queue handling in playback.c and it has been very useful in finding (and hopefully fixing) a crash case
10:37:44pondlifeI think it might be useful for me to put this logging into CVS for future use
10:38:06pondlifeBut wanted to check it wouldn't break anything
10:38:09 Quit Caliban (Read error: 104 (Connection reset by peer))
10:39:40 Quit tuwe (Read error: 110 (Connection timed out))
10:41:19 Join solarflare [0] (
10:43:31 Quit [San] (Read error: 110 (Connection timed out))
10:43:46LinusNpondlife: playback.c isn't used on the archos platforms, and those are the most critical when it comes mto binary size
10:44:05 Join [San] [0] (n=San@
10:45:32pondlifeYou wuoldn't have any objections to logfs going into CVS?
10:45:47LinusNnot really
10:45:47pondlifei.e. any other reason (code bloat?)
10:46:18pondlifeI think I've fixed the problem where natural end of playback causes a crash with voice
10:46:55pondlifeBut that will be a seperate patch
10:47:17pondlife(Just removing 2 lines from the existing code)
10:47:33pondlifeWill do some more testing today
10:47:49 Quit solarflare ("Leaving")
10:48:11itsmeguysok guys n gals, when i create my own wps, do i have to include any images or anything, or could it just be a simpel one, like only one .wps file?
10:48:33itsmeguysto which question? :D
10:48:42itsmeguysbtw, where ya from gordon?
10:49:01JdGordonyou can make the wps as pretty or boring as u want... and melbourne
10:49:16itsmeguyswell howdy do Gordon from Melbourne :)
10:51:00 Quit the_fuchs (Read error: 110 (Connection timed out))
10:51:14JdGordondoes anyone object to some list scrolling acceleration?
10:51:17itsmeguysok, since you seem to know about wps gordon, for each, like, command or whatever you acll it, do i need to put it on a new line?
10:52:15JdGordoni dont know about wps's... check the wiki or manual...
10:52:33itsmeguysyeh i have, it doesnt tell you much about the structure or anythinhg :(
10:52:35itsmeguysohwell, ty
10:53:09JdGordonits pretty much layout the .wps as you want it to be on the screen... so each line in the wps file is a line on the screen...
10:53:46itsmeguysyou sure about that?
10:54:03itsmeguysi have downloaded one from
10:54:05itsmeguysand opened it
10:54:10itsmeguysand there are way lines :D
10:54:15itsmeguyslike, 100 or so
10:54:28itsmeguysok maybe 50 :D
10:54:33JdGordonlots will be loading the pics and stuff
10:55:32itsmeguysdoes a / mean like a new command?
10:55:38webguest64pondlife: the only problem with logf is that it can't be activated selectively. If you place it at a frequently executed place in code the logf will be full quite fast. And if someone wants to see logf output from another specific place he'll be searching very long
10:55:58Bagderitsmeguys: try this =>
10:56:09jhMikeSShould I even worry about coding sample repacking for little endian right now?
10:56:10itsmeguysty kind sir!
10:58:26itsmeguyshow do i change through wps's?
10:58:30itsmeguyson my actualy h10 i mean
11:00:34 Quit [San] (Read error: 104 (Connection reset by peer))
11:00:43 Join [San] [0] (n=San@
11:01:13jhMikeS?? It matters when working with longs and doing downmixing fe
11:01:20 Join the_fuchs [0] (n=the_fuch@
11:01:24pondlifeAh, ok
11:01:29 Nick the_fuchs is now known as tuwe (n=the_fuch@
11:01:56pondlifeI'll put a #ifdef in for LOGF_QUEUES or somesuch
11:02:29pondlifeI didn't realise there was a limit to the number of logfs allowed - I'm just dumping them on the sim to stderr.txt
11:03:25markunwebguest64: I'm not so sure about the W anymore..
11:03:54webguest64markun: what exactly?
11:05:47LinusNpondlife: well, there is no limit, it's just that the logf buffer wraps
11:06:07pondlifeOK. I'll enable this on the sim debug build only (by default)
11:07:14jhMikeSLinusN: no op. on that? I just want to know if I should bother atm or do something more productive.
11:10:06LinusNjhMikeS: what problem are you trying to solve?
11:12:08jhMikeSLinusN: Well...when it comes to rearranging and mixing audio samples in long chunks, endianness matters. Is it a good idea for me to implement some of that for le too right now?
11:13:18JdGordondoes tagcache not work in the sim?
11:13:21markunwebguest64: can you try them? and nimbus-14.bdf
11:13:24jhMikeSor is it not too likely we'll encounter in the near future?
11:14:16markunJdGordon: works fine here
11:14:47LinusNjhMikeS: you mean reading two 16-bit samples as a long etc?
11:15:02JdGordonhmm... not working here.. i wonder why
11:15:23jhMikeSLinusN: Yes...especially mixing to mono
11:16:01LinusNjhMikeS: i'd say solve the problems when you face them
11:16:20jhMikeSI'm not facing it except in "what if" terms. :)
11:16:30LinusNthen don't do it
11:16:32 Quit webguest64 ("CGI:IRC")
11:16:41 Join webguest64 [0] (i=c27f0812@gateway/web/cgi-irc/
11:16:43jhMikeSlinusN: sage advice
11:16:54markunwebguest64: got my last message?
11:17:27webguest64markun: what exactly should I look at? I don't like the W. IMHO, the bottom line should be FF.
11:18:28markunI don't think that looks very good, but perhaps we should ask other people
11:19:29LinusNmarkun: show me some screendumps
11:19:52webguest64markun: I checked the TTF version of Chicago. The 'hole' at the bottom of W is very very small, visible only at large sizes.
11:20:29markunwebguest64: didn't know what font it was based on..
11:20:44markunLinusN: I can make you some screenshots to compare later
11:20:49markunHave to go to work now
11:21:23markunwebguest64: you are right, no gap
11:21:26JdGordonare we still having problems with scrolling in the list making playback stop?
11:22:30 Join MarcoPolo [0] (
11:23:02JdGordonbecause i just scrolled through the entire track list in tagcache without it stopping at all and i dont know if its my list acceleration or the issue has been fixed?
11:23:25webguest64markun: actually I've seen nimbus as something looking different. But there is also chicago and xtal in RB that llok the same as nimbus, only containt less glyphs. I don't know why they exist.
11:23:49JdGordonhmm.. TC doesnt let you know if the has updated :p
11:23:50linuxstbBagder: Are you happy with this patch to add different "tool" and "output" definitions for the bootloaders in configure?
11:24:10 Quit [San] (Read error: 110 (Connection timed out))
11:25:00 Join [San] [0] (n=San@
11:25:28itsmeguysfurk making a wps :)
11:25:41itsmeguysit doesnt work
11:25:47itsmeguysit always loads the default rockbox one
11:26:23linuxstbmarkun: What is the scramble command you need to run to build the Gigabeat bootloader?
11:27:56markunscramble -gigabeat rockbox.gigabeat FWIMG01.DAT
11:28:29linuxstbYou mean bootloader/bootloader.bin ?
11:28:49markunwell, right now it doesn't generate a bootloader.bin file
11:29:26linuxstbIt should be in the bootloader/ directory of your build directory. Looking at the existing configure script, I think that's just copied to rockbox.gigabeat.
11:29:38 Part eGen
11:29:41markunah, ok
11:29:45linuxstbSo rockbox.gigabeat and bootloader/bootloader.bin should be identical.
11:29:51markunyes, they are
11:29:56linuxstbI'll replace that copy with the scramble command.
11:32:51linuxstbPatch is here if you want to test:
11:36:28 Join Caliban [0] (
11:36:51jhMikeSwhere'd amiconn go? He started talking about that gcc glitch and I wanted to work out a solution but I'm tired now and must sleep. bback later though.
11:40:03 Join Mmmm [0] (
11:40:22 Quit jhMikeS ("Client exited")
11:42:03pondlifeHmm, I think I might try my first commit. .. This doesn't actually do anything and should only affect sims ;-). LinusN or anyone want to quell my fears? I'll wait a while....
11:44:02 Quit [San] (Read error: 110 (Connection timed out))
11:44:04 Join [San] [0] (n=San@
11:46:52pondlifeWiki is giving me error 500 - Error message: Premature end of script headers: view
11:47:06pondlifeAttempting to do a search
11:47:49***Saving seen data "./dancer.seen"
11:48:13 Join barrywardell [0] (i=892b7a6e@gateway/web/cgi-irc/
11:50:25barrywardelllinuxstb: nice changes to configure
11:51:03barrywardelljust one thing is that the H10 bootloader filenames are slightly wrong
11:51:09linuxstbPlease fix.
11:51:32barrywardelli'm about to commit some other changes now so I'll include the fix for that too
11:53:24JdGordoni tihnk i should start naming my rockbox source trees a bit better.. atm i have rockbox rockbox_ rockbox__ rockbox_not_sure_whats_here rockbox_menu :p
11:54:26dan_apondlife: I like that patch - it's going to make my life a whole lot easier when I come to running the codec thread on the second core on PP machines.
11:54:36 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
11:55:15webguest64JdGordon: yes, you should come up with a rule for the number of underscores! E.g. depending of how unsure you are about what's in :-)))
11:55:15pondlifedan_a: I'm just working out how to use CVS and then I'll commit it
11:55:43tucozwhat should we call the direction keys on the x5? left/right/up/down or prev/next/plus/minus?
11:56:15*amiconn would prefer left/right/up/down
11:56:36tucozok. that's fine.
11:56:38amiconnlinuxstb: RED
11:56:47markunamiconn: yes, but the builds didn't fail
11:56:58markunmaybe a problem with the buildtable script?
11:57:22pondlifeWith CVS commit, does the message after -m need to be quoted? I assume so, but don't want to put extraneous quotes in!
11:58:07dan_apondlife: Yes
11:58:41JdGordonBagder: while you were fiddling with the configure script, why not remove the sim choice?
11:59:44tucozamiconn, is select done by clicking the joystick on the x5?
12:01:14tucozAnd also, what are the buttons in clockwise direction starting from the top of the player?
12:02:14amiconntucoz: Yes, select is clicking the joystick
12:02:28tucozpower, hold, play, rec?
12:02:31amiconnAt the right side there are, starting from the top:
12:03:00amiconn(1) The combined power/hold slider, (2) Record, (3) Play
12:03:00linuxstbamiconn: All I can think is that the scramble tool isn't returning the correct error code - so the build script thinks it's failed...
12:03:23tucozamiconn, thanks! the one on top, what is that?
12:03:25amiconnThe left side has no buttons
12:03:42amiconnNo button at the top either, it's the charging LED
12:03:44 Quit [San] (Read error: 104 (Connection reset by peer))
12:03:57tucozok. thanks for the help
12:04:02amiconnIn fact it's a software controllable two-colour led
12:05:01 Join [San] [0] (n=San@
12:05:12tucozcool, like the sw controlled power led on the amiga :)
12:05:24tucozbut that wasn't two-colour though
12:05:38amiconnLinusN: What do you think: Should we use the X5 LED for the same purpose as the of does, or should we e.g. use the red led for hd activity?
12:06:38amiconnThe red and green led chip can be switched individually, so we can also produce yellow, and afaics they can also made blinking (w/o further cpu intervention)
12:09:53amiconnThe amiga power led isn't really software controlled, or rather, not without changing other hw settings
12:10:03amiconnIt's coupled with the audio lowpass filter
12:10:06JdGordonhave fun with it... do orange for disk activity, red flashing when the battery is getting low.. green when the cpu boosts (if x5 has it), yellow for pea meter..
12:11:31pondlifeHmm, how long will it normally take forf a CVS commit to appear on the website? i.e. on the CVS build page or homepage?
12:12:08pondlifeAh, never mind
12:13:23linuxstbArgh, the build system doesn't seem to run when the tools are edited. Someone please commit something....
12:13:45linuxstbAh, someone has...
12:18:04preglowamiconn: did anyone ever use the lowpass filter??
12:18:15 Join TCK [0] (
12:19:11 Quit JdGordon ("Konversation terminated!")
12:20:46webguest64 /part
12:20:50 Join spiorf [0] (
12:21:05 Join JdGordon [0] (
12:21:51 Quit webguest64 ("CGI:IRC")
12:22:03 Join webguest64 [0] (i=c27f0812@gateway/web/cgi-irc/
12:22:42 Quit [San] (Read error: 110 (Connection timed out))
12:23:10 Join [San] [0] (n=San@
12:23:14 Quit pondlife ("lunch")
12:26:32 Join GliGli [0] (
12:26:42 Part itsmeguys
12:27:33 Quit GliGli (Client Quit)
12:29:13 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
12:29:27 Join scorche [0] (
12:29:33 Join GliGli [0] (
12:29:51 Quit ghode|afk (Read error: 110 (Connection timed out))
12:32:04linuxstbBagder: Can you look into the red bootloader builds? I think it's a build-script problem - the bootloaders build fine.
12:32:26*linuxstb afk
12:34:58LinusNamiconn: using the x5 leds for hd activity might be a bad idea, since the leds are controlled via the pcf50606
12:35:11LinusNlots of i2c
12:35:23amiconn_Not necessarily lots...
12:35:27LinusNperhaps not
12:36:09JdGordondoes anyone object to list scrolling acceleration?
12:36:25LinusNnot me
12:36:39webguest64markun: your changes are nice! I like them! One note though: IMHO the U+201a (low single comma) and U+201e (low double comma) should be placed two points deeper
12:39:23amiconn_LinusN: We could use a delay mechanism like we do for the 'emulated' disk led
12:39:33LinusNamiconn: yes
12:40:56 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
12:43:22 Join Kohlrabi [0] (
12:43:24 Quit Mmmm (Read error: 110 (Connection timed out))
12:45:13 Quit [San] (Read error: 110 (Connection timed out))
12:45:25 Join [San] [0] (n=San@
12:47:34 Quit amiconn (Read error: 110 (Connection timed out))
12:47:34 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
12:49:27markunwebguest64: ok, I was not sure about that
12:49:38markunThey should be under the baseline?
12:50:51markunJdGordon: the gigabeat, h10 and ipods have ways to control the scrolling speed. Any idea how/if this could be implemented in rockbox?
12:51:43webguest64markun: IMHO they should be exactly as the normal comma. Just type ,Hello
12:52:05JdGordondiffenbach has a patch for the ipod doesnt he?
12:54:27markunwebguest64: they agree with you:
12:57:10webguest64markun: for U+201a: "This character aligns on the baseline at similar height as the comma."
12:58:05webguest64markun: it's weird but all glyphs (e.g. T) are somewhat below the base line which probably explains the problem with U+201a.
12:59:14webguest64markun: try to switch on the base line in fontforge (view -> H. Metrics IIRC) Then you'll see...
12:59:41markunI'll do that when I get home
13:00:36markunMaybe I'll add greek to the fonts
13:00:38 Quit uski (Read error: 54 (Connection reset by peer))
13:01:54webguest64markun: OK. I never use that quotation marks (who does?) so it doesn't matter to me.
13:04:05barrywardellit was suggested to me that we should be calling the rockbox firmware for the H10 rockbox.iriver instead of rockbox.h10.
13:04:28barrywardellI don't really know which would be better, but am happy enough with rockbox.h10
13:04:51Bagderto me it doesn't matter at all
13:05:40barrywardellI'll leave it as it is so. It will help to avoid problems with people downloading h300 firmware by accident, etc
13:06:01webguest64 Bagder: wouldn't it be better to generally name the loadable file rockbox.<model> to avoid confusing?
13:06:17Bagdermight be, yes
13:06:38 Quit [San] (Read error: 110 (Connection timed out))
13:08:54 Join [San] [0] (n=San@
13:10:15 Join bbroke [0] (
13:10:53JdGordonis it bad to boost the cpu if its not _really_ needed?
13:11:11webguest64Bagder: that would also kill the problems when people download RB for wrong platform
13:11:37Bagderwebguest35: it would possibly reduce them, not kill
13:11:52BagderJdGordon: yes
13:11:52dan_aJdGordon: It might waste battery
13:12:23tucozi think that would be hard to avoid. I think the images above the download links are good enough (unless you are blind)
13:12:44Bagderpeople just don't pay attention
13:12:48Bagderthey just click and install
13:12:53Bagderand then come asking
13:13:02Bagder"uh, what I don't know what I downloaded"
13:13:12tucozyes. and that can not be helped by renaming the extension
13:13:13JdGordonim trying to do list acceleration, it works fine now except while fast scrolling it will always select the 2nd item in the list if there is more than one screen.. OR or i can do it but it need cpu boosting to look better...
13:13:32Bagdertucoz: exactly, but it might be quicker to diagnose the mistake
13:13:33webguest64 Bagder: Einstein said: there are two endless things: the universe and human stupidity. I'm not sure about the universe though... :-) So you're right: reduce, not kill.
13:14:01JdGordonno he said "only 2 things are infinite.. the universe and human stupidity, and im not sure about the former"
13:14:18tucozsure. or ask them to go to the daily builds page and identify the model by looking at the picture, combined with the name of the build
13:14:46Bagder"there are two endless things, the universe and people trying to put quotes in my mouth for centuries after my death"
13:15:22JdGordonhe has only been dead for 20odd years...
13:15:25JdGordonhardly centuries
13:15:29Bagdernot yet
13:15:38Bagderits a prophecy
13:15:51 Join uski [0] (
13:16:07tucozBagder is now known as Nostradamus
13:16:57tucozwe could have a fortune plugin in rockbox name Nostrabox or Rockdamus
13:18:07theli_ualinuxstb, ping :)
13:21:05 Join Brock [0] (
13:26:58 Join merbanan_ [0] (
13:27:59JdGordonis xor faster than timesing by -1 to negate a number?
13:28:48LinusNxor should be faster, and also easier to understand in the code
13:29:03Bagderyou don't really negate by xoring, you one-bit complement
13:29:04LinusNespecially if you are dealing with unsigned numbers
13:29:27LinusNBagder: exactly
13:29:42Bagderbut of course, perhaps that's what you meant
13:29:52 Quit ShadowdogMU (Read error: 110 (Connection timed out))
13:29:55 Quit [San] (Read error: 110 (Connection timed out))
13:29:56LinusNJdGordon: to negate a number: x = -x;
13:30:11JdGordon:p why didnt i think of that
13:30:16*JdGordon is asleep obvously
13:30:28*Bagder pokes JdGordon with a stick to wake him up
13:30:37 Join [San] [0] (n=San@
13:30:43Bagderfeeling better? ;-)
13:30:56JdGordonnot really...
13:31:04*scorche hands Bagder a Long Pointy Stick (TM)
13:32:48JdGordonanyone want to try my list aceleration?
13:33:21JdGordonim not sure if im happy with it or not :p
13:37:16JdGordoncmon.. someone wants to... or should i just commit it and fix it if people dont like it? ...">
13:38:57tuweJdGordon: ok i'll try
13:40:57tuwedid you notice a strange behaviour in the iCatcher.wps which is included? it doesn't show the bmp files correctly in today's cvs. i also checked yesterday's source, both without patches.
13:42:05tuweit seems as if it doesn't load themes correctly. if i try my own wps, it loads the font as specified in the theme/*.cfg file, but it doesn't display the new wps but the last one.
13:42:42tuwemy wps says %wd in the first line, and "Hallo" on the second. yet, the status bar still can be seen.
13:42:44 Quit Snyper (Read error: 110 (Connection timed out))
13:42:53 Join Snyper [0] (
13:42:55tuwebut it doesn't show the "Hallo"
13:45:55 Join lightyear [0] (
13:46:55 Join ElCabron [0] (i=5144f420@gateway/web/cgi-irc/
13:47:15ElCabronHi all,..
13:47:44 Join Genre9mp3 [0] (n=yngwiejo@
13:47:52***Saving seen data "./dancer.seen"
13:48:31 Part webguest64
13:48:41JdGordontuwe: did u try the patch?
13:49:16tuwenot yet, but i just put the newly build rockbox on my ipod.
13:49:50ElCabronI'm currently trying to update my iPod vid. 30gb to use rockbox. Problem is that ipodpatcher cannot find my ipod. I have windowsXP as OS. Anyone who can help me out?
13:52:09ElCabronOh, and my current firmware is apple's 1.1.2 firmware update.
13:52:38tuweJdGordon: no strange behaviour yet. although my lists aren't very long...
13:53:03JdGordonswitch to tag browser (if u use it)
13:53:24tuwei don't remember how it was before that patch, i spent most of my time figuring out how to get album art to work on ipod mini :)
13:53:32JdGordonthe thing that bugs me is that the selection stays at the top of the screen... i dont know if u should bother fixing it or not tho
13:53:49tuwethe list that shows search by artist etc.?
13:54:09tuweyou mean it doesn't wrap?
13:54:58 Join bluebrother [0] (
13:55:27JdGordonno, (it doesnt but...) try scrolling a list shorter than the screen.. the selection moves down as expected... do a list longer than the screen and selection will jump from the middle somewhere to the 2nd top item
13:55:35 Quit MarcoPolo (Read error: 54 (Connection reset by peer))
13:55:47 Quit [San] (Read error: 110 (Connection timed out))
13:56:10 Join [San] [0] (n=San@
13:57:28 Quit safetydan ("Ex-Chat")
13:57:43tuweok i have a list that contains most of my songs. if i scroll down from the start, the selection is on the second line, counting from bottom. if i scroll upwards, the selection is on the second line, counting from top.
13:58:06JdGordonye.. thats what im complaining about... do i try to fix it?
13:58:58 Join pondlife [0] (
13:59:05*amiconn wonders what JdGordon is trying to do...
13:59:12JdGordonlist acceleration
13:59:35*JdGordon buzzed you earlier about removing waste code from actions but u didnt respon...
14:00:04tuwewell, i thought that was the intended behaviour. i like it that way. even with a relatively big font (i use nimbus 12 on an ipod mini)
14:00:22tuweand i like that it doesn't wrap.
14:00:36tuweit behaves correctly at start and end of the list
14:01:11bluebrothertuwe, ?
14:01:21JdGordondoes it stutter a bit while scrolling?
14:01:38JdGordoni guess its stil better for the list to stutter than playback stop :p
14:02:17tuwebluebrother, ?
14:02:40tuweoh. you're in
14:03:07bluebrotheryeah. You just noted that name right after I came in.
14:03:28bluebrotherso I was wondering if this was meant to say something to me ...
14:03:53tuwebluebrother: yes, i went to university in karlsruhe
14:04:01bluebrotherah. Nice.
14:04:28 Quit barrywardell ("CGI:IRC (EOF)")
14:05:01tuweJdGordon: i don't see any stutter... playback is clean, too. (as far as i can tell, maybe i should actually play music :))
14:05:10tuwebluebrother: well... never mind. :)
14:05:39JdGordontuwe: ok, in that case i guess my stuttering is because audio is playing.. and id rather the gui stutter than the audio to stall
14:07:08tuweJdGordon: no it is playing something, but it's rather noisy :)
14:07:32tuwei think playback is good while scrolling around in the list
14:08:16amiconnlist accelleration?
14:08:26bluebrotherdamn. Has someone else problems building logf-enabled?
14:08:31tuweJdGordon: ok it is not as good as i thought
14:08:48JdGordonthe longer u hold down next in the list, the faster the list ttravels...
14:08:53JdGordontuwe: how?
14:09:12tuweif you play a song and do excessive scrolling, the playback stops, and/or the scrolling stutters.
14:09:23*amiconn wonders what that should be good for
14:09:25JdGordonwhich ipod?
14:09:34tuwei'm not sure if it is the same without patch
14:09:36JdGordonamiconn: scrolling through long lists...
14:09:40tuweipod mini 2g
14:09:43amiconnFor me, scrolling through lists was always fast enough
14:10:00JdGordondo u use tagcache?
14:10:05bluebrothersomeone an idea? Building h120, only logf enabled.
14:10:07JdGordonlots of loooong lists in TC
14:10:28tuweumm. i just put the default install with your patch on my ipod
14:10:46 Join ghode|afk [0] (
14:11:07JdGordonsorry the TC q was to amiconn, tuwe im gonna guess scrolling in lists will pause playback on the ipod 2g anyway
14:11:25amiconn...even in lists with 1000s of items
14:11:28 Join ^BeN^ [0] (
14:11:44amiconnAnd there's always the option to use paged scrolling (on many targets)
14:12:09tuweJdGordon: playback doesn't stop when entering the list, only after about 5s of scrolling around.
14:12:18JdGordonye, expected
14:12:24tuwei'll try to change the anti-skip buffer time
14:12:35JdGordontry doing the same without the patch and it should be the same
14:12:40amiconnThis has nothing to do with the anti-skip buffer
14:13:34tuwei didn't find it anyway.
14:14:33tuweJdGordon: another thing is, if i scroll down to the bottom of the list, then put my hand off the scroll wheel for a moment, and scroll down further, it wraps the list.
14:16:06tuweJdGordon: if i choose a track from my list, it plays the track below instead.
14:16:09JdGordonthats also expected
14:16:18JdGordonthat is odd tho
14:16:33JdGordonare you sure??
14:16:42 Quit [San] (Read error: 110 (Connection timed out))
14:16:45tuwei don't know if it's your fault, i think someone reported it in the forums.
14:17:22 Join [San] [0] (n=San@
14:17:24tucozbluebrother, just to let you know. same thing happens here
14:18:50tuweJdGordon: no i wasn't able to reproduce it.
14:21:01tuweok i tried it after wrapping and scrolling downwards and upwards and stuff, i wonder if it was just in my mind :)
14:24:25JdGordoncool.. div by 0 doesnt crash the player!
14:26:31tucozbluebrother, according to some old logs try to add -fomit-frame-pointer to the Makefile
14:27:20tucozbluebrother, read from 08.54.41
14:27:58tuwegeneral question: you can go from the wps to the file browser by pressing the select button. but pressing the select button there plays the highlighted file, to return to the wps again, you have to press the play/pause button. why is it not the other way around?
14:28:18tuweon ipod mini.
14:28:36 Join Arathis [0] (
14:35:33bluebrothertucoz, thanks, that did the trick.
14:36:46 Quit [San] (Read error: 110 (Connection timed out))
14:37:13 Join [San] [0] (n=San@
14:38:00tucozlooks like that option should be set when building logf builds (according to amiconn)
14:38:18tucozthat is, there is no point not to set it
14:38:25JdGordonLinusN: Bagder, anyone else.. anyone object to me commiting list accel (amiconn, it adds a whopping 200bytes to rockbox.bin with gcc 4.0.3)
14:39:53 Quit TCK (Read error: 104 (Connection reset by peer))
14:43:56amiconntucoz: The frame pointer (sometimes) makes debugging easier.
14:44:13amiconnSo it makes sense to not set -fomit-frame-pointer for debug builds
14:44:20JdGordonruddy hell... rearranging code to make it reuse more code didnt touch the size at all :p
14:44:23tucozi see. but for logf builds?
14:44:34 Quit aliask ("Chatzilla 0.9.72 [Firefox]")
14:44:53ElCabronHi guys I finally got it up and running!
14:44:56amiconnFor logf builds, the framepointer is of no use imho
14:45:47tucozok, so it could be wise to include -fomit-frame-pointer in logf builds (by default) then.
14:45:50amiconnThe problem is that the framepointer blocks one extra register, so asm blocks which are 'eating' registers might trigger errors
14:46:24tucozso that is why gcc complains that it is out of registers
14:46:56 Join lorijho [0] (n=lorijho@
14:47:15tucozand by omitting the frame pointer, a register is freed and the build compiles as it should
14:47:19amiconnWe could set -fomit-frame-pointer for logf builds, but I wonder whether this would be a good idea
14:48:08amiconnNot setting it reveals places which are tight on registers. This might be useful for optimisation
14:49:05 Join TCK [0] (
14:49:08amiconngcc sometimes starts to do weird, ineffective things when it's short on registers
14:49:27amiconn..even if it doesn't error out
14:50:11tucozanother topic, wonder if the A-B repeat mode is functional at all on the ipods. No real answers to my question in the forums.
14:50:23*amiconn has a similar problem with his grayscale lib scrolling rework on sh and coldfire
14:50:45tucozIt is possible to select A-B mode, but as it seems it is not possible to set the A and B markers.
14:51:21tucozanother huge asm-block in the works?
14:53:09amiconnYes. Unbuffered grayscale lib scrolling already uses asm, but scrolling with bitshift _and_ byte shift (i.e. scroll amount >8 and not a multiple of 8) is currently doen in 2 stages
14:53:25*bluebrother starts to hate gnome-mount
14:53:28 Join [IDC]Dragon [0] (i=d90a3c55@gateway/web/cgi-irc/
14:53:47amiconnOoohh rare guest!
14:53:52*amiconn greets [IDC]Dragon
14:54:30[IDC]DragonI'd like to know how fast my jpeg decoder is, on an arm target
14:54:49[IDC]Dragon(for some estimation which I need)
14:55:15[IDC]Dragonanybody with an arm target around?
14:55:30amiconnIirc decoding a ca. 900x650 image to grayscale takes ~1.0 seconds
14:55:34amiconnat 75MHz
14:55:45[IDC]Dragoncolor, sorry
14:55:54amiconncolour is around 50% slower iirc
14:56:19[IDC]Dragonah, I forgot that this does yuv->rgb ontop
14:56:38[IDC]Dragonso, the pure decoding is not so easy to benchmark
14:56:46amiconnyuv_rgb isn't included in the rockbox jpeg figures
14:57:00amiconnIt's done live for displaying
14:57:34[IDC]Dragonthe comparison I need is a 2MP image, downscaled by factor 4 decoding
14:57:56amiconnSomeone with a colour arm target (ipod, h10 etc) should be able to give exact figures
14:58:03[IDC]Dragon(not full res)
14:58:05*amiconn only has a greyscale ipod
14:58:17*[IDC]Dragon has no ipod
14:58:27*Bagder only has a non-supported ARM target
14:58:36 Join sm007h [0] (
14:58:40*[IDC]Dragon has no sw-codec target
14:58:48*daurnimator also has an unsupported ARM target
14:58:54sm007hmm, I'm having a problem with the latest build
14:59:09*amiconn pings linuxstb
14:59:23*tuwe has a greyscale ipod
14:59:34sm007hI'm using a 4gb nano
14:59:37Strogghrmm I think I have an arm target.
14:59:43*tuwe is still wondering about album art display on ipod mini2g
14:59:57StroggI just installed yesterday though, so I haven't had a chance to try displaying photos on it
14:59:58[IDC]Dragonnano is color arm
15:00:03sm007hwhen it boots up, all I see is "rockbox", I don't have a way to get to my media
15:00:49Bagderit says rockbox all the time?
15:01:02sm007hyeah, like it's in the root directory
15:01:09Bagderyes it is
15:01:22Bagderand you've tried reading the manual?
15:01:54sm007hI don't remember it doing this on earlier builds. If it changed since 0829, then I may need to reread the manual
15:02:11Bagderso how is your music stored on the device?
15:02:13 Quit [San] (Read error: 110 (Connection timed out))
15:02:16 Join [San] [0] (n=San@
15:02:23sm007hI have it in a music folder inside of .rockbox
15:02:39Bagderthen select "show all" and enter .rockbox
15:02:58Bagderalthough I must say having music within .rockbox sounds like a weird choice
15:03:17sm007hwhat do you recommend?
15:03:24amiconn[IDC]Dragon: The rockbox jpeg decoder has a slight asm optimisation for all target archs now (sh, coldfire and arm) - the clamping to [0..255]
15:03:40Bagdersm007h: a music dir in the root
15:03:48Bagderor whatever really
15:04:49[IDC]DragonI remember to have done such a clip in C, compiled to the same on ARM
15:05:17amiconnWhat compiler was that??
15:05:34[IDC]Dragonfrom ARM themselves
15:05:54*[IDC]Dragon peeks on how you clip
15:06:03*amiconn thinks that it couldn't have been gcc
15:06:06 Quit spiorf (Remote closed the connection)
15:06:15sm007hregardless of where I put my music files, why is the .rockbox.ipod file showing up? When I click on it, it reloads.
15:06:47Bagdersm007h: that's rockbox
15:06:58Bagderand its not starting with a dot
15:07:16Bagderso when you click on it, it is loaded and started - again
15:07:25sm007hright, isn't that strange?
15:07:30Bagdernot to me
15:07:35Stroggwhat's strange about that?
15:07:44StroggIf you open a door, and the door opens.. is that strange?
15:08:11bluebrothersm007h, set "show files" to music, then you'll only see music files.
15:08:14sm007hwhen I start my iPod, I immediately see a screen that lets me select a music folder (based on Badger's suggestion) or if I click on rockbox, the whole thing reloads
15:08:32Bagderbut why do you click on the firmware file?
15:08:37Stroggsm007h: maybe you should copy some music to it?
15:08:40sm007hStrogg, I don't think you're making an appropriate analogy
15:09:05sm007hbefore I moved my music directory, it was the only thing there. so I clicked on it.
15:09:32Bagderbut its still a useful feature
15:09:36Stroggsm007h: think of the rockbox.ipod file as an exe file... when you run it, it runs.
15:09:43sm007hand now it's one of two things. so I may click it on accident or I may not click it but always be bothered by it being there
15:10:06StroggSo don't show it. go to your settings and click on "show only music files"
15:10:06sm007hok, if you think it's useful, I'd like to hear the reasoning. I think it's counter-intutative.
15:10:26StroggI've used rockbox for 10 minutes and I figured that part out.
15:10:27Bagdersm007h: build 22 different rockbox versions, put them all there and voila, you can run whichever you want
15:10:28 Join XavierGr [0] (
15:10:39sm007hthat's true. I could do that. If so, I think it shold be default. Or there should be a media only settings
15:10:42StroggBagder: what about 23 versions?
15:10:43Bagderhow is that not useful?
15:10:52sm007hit's not a confusion thing
15:11:08sm007hBadger, that is...I agree...for devs, testers, and hackers.
15:11:25Bagder... which is about 95% of all rockbox users
15:11:25sm007hbut not for general public. I can see your point, I guess I'm raising a usability issue.
15:11:49 Join Pyromancer [0] (
15:11:57sm007hok, I'll leave it at that. I agree with what you're saying. So hopefully you'll see where I was coming from.
15:12:00StroggI'm not a dev, tester or hacker(of ipods) but I could see a user for that
15:12:27Bagdersm007h: I see it, I just don't see why it is so bad that you got to see a file you could run
15:12:42 Quit Snyper (Read error: 110 (Connection timed out))
15:12:49 Join Snyper [0] (
15:13:00sm007hit's not, it's just counterintuative. When I turned on my iPod, it would show me media files. not the entire file structure in my root directoy.
15:13:01*Bagder is long-time rockbox damaged
15:13:44Bagdersm007h: rockbox is traditionally for file tree based music using
15:13:53Stroggsm007h: so then you installed the wrong thing. It says right in the docs.. rockbox is file based.
15:13:54sm007hreminds me to much of windows 3.1. I'd boot it up, then I'd have to run win.exe, lol
15:13:59tucozthat is the exact number one reason I like rockbox. to have control over my disk.
15:14:05Strogghell. that's the whole reason why I switched to rockbox
15:14:17 Join Carroarmato0 [0] (
15:14:23Stroggthough the blue on black text is a nice bonus perk :)
15:14:24sm007hlol, way to take my comments out of context
15:14:42Bagdernot at all
15:14:54Bagderit starts in the file browser much because of that
15:14:58bluebrotheris there a maximum line length in logf? For the output I mean.
15:14:59StroggI don't think so.. if you want an itunes like interface, then you should probably restore to itunes..
15:15:03sm007hall I questioned was whether the dang firmware should be visible at boot, lol
15:15:11Bagderbluebrother: rather short, I don't remember the exact limit though
15:15:12StroggI mean.. what feature did you choose rockbox for?
15:15:19sm007hitunes? wtf are you talking about
15:15:36Stroggeehh.. itunes, ipod.. what ever the normal interface on an ipod is.
15:15:45Bagdersm007h: it doesn't have to be that you can 1) change that and 2 ) resume on start
15:15:45sm007hok, youre bored, have fun flaming someone else.
15:16:04*Strogg sort of lumps everything ipod'ish into a bix box labeled "Itunes"
15:16:28StroggThat reminds me...
15:16:32*Strogg uninstalls Itunes. :)
15:16:49sm007hBadger, yeah, but I chose Supported files instead of all
15:16:49tucozsm007h, it is useful to have it there by default. If you do not want it, you can hide it. Or you can use tagcache, and browse by tags instead.
15:17:14sm007hand that should hide it. if I choose music, then I lose other media files, such as, pictures and videos
15:17:19markuncan rockbox on the ipod also load the firmware from .rockbox?
15:18:15tucozirivers can do that right?
15:18:50*tuwe displaying albumart bitmap: x|y=39|10, w|h=60|60
15:18:58bluebrotherthe iriver bootloader can do so since version 6
15:19:13sm007hso what I'm asking is whether it not make more sense to have "all files" be all kinds, including execs, "supported" be media, and "music" be music. or if execs are included in the "supported files" then we should have a "media files" setting
15:19:46tuwesm007h: you can change all of that in the settings.
15:20:05tucozsm007h, in the not so distant future the default will be to load rockbox.iriver from .rockbox. maybe that is the case for the ipods as well.
15:20:11StroggWhat if you use your mp3 player as a portable harddrive (like I do) and you just want to check if a file is on there?
15:20:14sm007htuwe, are you referring to actually editing the config files manually?
15:20:41Carroarmato0Is there a way to force Rockbox on the Ipod nano to set the battery at regular 340mAh instead of the default 1300mAh? I think that that would saveup some battery time.
15:21:08sm007hStrogg, if you're seriously asking me, then I think the obvious solution is to set it to show "All files"
15:21:12JdGordonthat value does nothing but guess the reamina time
15:21:29tuweno, go to the menu, general settings, file view, show files. if you choose id3 database it only shows your songs.
15:21:42tuweyou have to populate the id3 database, of course.
15:21:50 Join lodesi [0] (
15:21:50sm007htuwe: right, but that hides my pictures
15:22:13tuweumm, i didn't think of that.
15:22:24sm007hI would like *only* media files to show
15:22:52sm007hCarroarmato0: is that the battery in all of the ipods? I have a 4gb but couldn't find the battery spec
15:23:02bluebrothersm007h, why don't you just put your files in subfolders? That way you can completely ignore the root folder (and maybe place your configurations there for easy access)
15:23:27 Quit Arathis ("[rl_bot quit]")
15:23:33Carroarmato0sm007h I searched the specs on the ipod nano wiki:
15:23:38sm007hbluebrother, that's what I had and then badger suggested I move them to the root folder ;)
15:23:48tuwewell, if i switch to show files->supported, it shows images, but no rockbox.ipod file. i don't know if that's just because i'm on the simulator at the moment.
15:24:03bluebrothersm007h, not exactly. Bagder suggested to put them somewhere else than in .rockbox
15:24:13Carroarmato0sm007h:It says 340mAh Li-poly battery
15:24:44sm007hbluebrother, what distinction are you making? I had them in subfolders under .rockbox. he said that didn't make sense. the only other place is / (root)
15:24:49 Quit nudelyn ("At Argon, we're working to keep your money.")
15:24:51JdGordonCarroarmato0: that value does nothing but guess the reamina time
15:25:08tuwesm00th, put a folder called media or something inside your root folder
15:25:12tuweand start from there.
15:25:20 Join Arathis [0] (
15:26:03Carroarmato0JdGordon: Oh ok, thanks. Guess that changes to the code would be the only solution to lighten the battery usage.
15:26:04tuwestupid question: how do i exit from the image viewer?
15:26:10bluebrothersm007h, for example I have /Albums which I use for CD albums. As you told, you have folders like /.rockbox/SomeAlbum which is weird.
15:26:19sm007htuwe: that's basically what I have now
15:26:30sm007hno I don't
15:26:39sm007hI said I had /.rockbox/music/
15:26:40bluebrotheralso, Bagder told to use a folder in the root dir.
15:26:56sm007hchrist, who do think I am, lmao
15:26:58bluebrotherin that sense he told you to use /music/ instead.
15:27:24sm007hthat's what I'm doing, I need to reread my comments because I'm obviously not coming across making much sense
15:28:25tuwetucoz, even on the simulator?
15:28:47sm007hyeah, that's exactly what I said. That I have a music folder under .rockbox and he said this makes more sense:
15:28:47tucozbut then the keys are simulated
15:28:48sm007h<Bagder> sm007h: a music dir in the root
15:29:05tuwedoesn't work here...
15:29:20tucozhmm, you don't get it to show a menu?
15:29:32 Join nudelyn [0] (
15:29:39tucozselect+menu? not sure what keys those are in the sim though.
15:29:53sm007hbluebrother, regardless, my files are in a music folder in my root, how do I ignore the root directory at boot?
15:29:59tuwei'm on the ipod mini simulator though
15:30:20tuweif i run it with −−background, select is 5 and menu is .
15:30:32amiconn[IDC]Dragon: Will we ever see the 10x8 idct for correcting the archos pixel aspect?
15:30:46sm007htuwe: I don't know, when I check supported files, it shows me rockbox.ipod
15:30:50*amiconn ponders adding an interpolation method instead, for all zoom levels
15:31:23bluebrothersm007h, you can't ignore the root. When using the file view you'll startup in the root folder.
15:31:25 Quit Carroarmato0 ("Leaving, bye bye!")
15:31:32tuwesm007h, i'm afraid i can't help you then.
15:31:41 Join D_C [0] (
15:32:00sm007h<bluebrother> sm007h, why don't you just put your files in subfolders? That way you can completely ignore the root folder (and maybe place your configurations there for easy access)
15:32:07sm007hI was referring to that
15:32:31bluebrotheryou can ignore that by just entering the music folder upon startup. That's mostly the way I do.
15:33:00bluebrotherexcept from the fact that I have multiple music folders in the root, so I can select which "type" I want to browse.
15:33:29 Nick D_C is now known as D_C_ (
15:33:33sm007hwhere this becomes a problem is when I erased my config files. and then when I started all I had access to via the file view was the rockbox.ipod
15:33:56D_C_can anyone see this
15:34:07bluebrotherD_C_, this thing is on.
15:34:12BagderD_C_: about 135 persons can
15:34:24D_C_lol ok, just that it said "can't join #rockbox (auth'ed only (+r))!"
15:34:42 Nick D_C_ is now known as D_C (
15:35:31Bagderlinuxstb: are you are of the red builds since your configure commit?
15:35:39D_Cso rockbox uses gnuboy right? so does that mean if i can obtain some gameboy ROMs, i can play these?
15:35:54BagderI bet its the new output name
15:36:03Bagderthen the problem ends up in my lap...
15:36:07 Join MarcoPolo [0] (
15:36:19 Nick D_C is now known as D_C_ (
15:36:26*Bagder shuts up and whistles
15:36:36 Quit theli_ua ("Leaving")
15:37:17StroggPoo.. I think I crashed it
15:37:33 Join Quelsaruk [0] (n=kvirc@
15:38:06Bagderlook, an old-timer!
15:38:28D_C_can anyone answer my question?
15:38:43BagderD_C_: tried the manual?
15:39:10bluebrotherD_C_, surely everyone can decide if he can answer your unasked question ;-)
15:39:12Quelsarukhi Bagder :)
15:39:49amiconnWow, the second rare guest today!
15:39:52amiconnhi Quelsaruk
15:39:55LinusND_C_: the answer is yes
15:39:57Quelsarukhi amiconn
15:40:12Quelsarukwho came today? another old-timer?
15:40:38 Quit TCK (Read error: 104 (Connection reset by peer))
15:40:42 Join pixelma [0] (i=pixelma@
15:41:35 Join JazzBone [0] (
15:41:42amiconnQuelsaruk: [IDC]Dragon.
15:41:49D_C_so i just place the .gbc .gb files anywere and just open them?
15:41:52 Quit Snyper (Read error: 110 (Connection timed out))
15:42:05LinusND_C_: yes
15:42:07Quelsarukamiconn: i thought that he was still working in rockbox
15:42:15 Join Snyper [0] (
15:42:40bluebrotheris there anything I need to take care of when adding a new char[] to the user_settings struct
15:44:15tuwetucoz: i can exit from the image viewer when i look at correct bitmaps, i seemed to have a corrupt one.
15:44:48LinusNbluebrother: that question wasn't very specific
15:45:21bluebrotherLinusN, yes. I added a char[MAXFILENAME+1] at the end of struct user_settings.
15:45:49bluebrotherstrangely it does work only in some functions. When placing the array somewhere else in the struct it works.
15:46:11LinusNcan i see the patch?
15:46:43bluebrothercurrently I don't have one but I can make you one ... give me a few minutes.
15:47:56***Saving seen data "./dancer.seen"
15:48:02 Join linuxstb_ [0] (n=linuxstb@
15:48:24bluebrotherwhen adding new files can I simply diff -u /dev/null new/file.c >> my.patch?
15:50:20amiconnbluebrother: You can also add the new file to your local working copy, then use cvs diff -uN
15:50:50[IDC]Dragon<amiconn> [IDC]Dragon: Will we ever see the 10x8 idct for correcting the archos pixel aspect?
15:51:01[IDC]DragonI have a 10*8 IDCT
15:51:16[IDC]Dragonbut no 5*4
15:51:34[IDC]Dragonand no 2.5*2, 1.25*1
15:52:35bluebrothersee the two changes in settings.h −− with the line now //-commented out it doesn't work
15:52:38amiconn[IDC]Dragon: Then we could use the 10x8 idct for full scale, and interpolation for the smaller sizes
15:52:47bluebrotheron h120 sim it's 8 chars off, on h300 sim 4.
15:52:59amiconnI guess 10x8 idct will look the same as 8x8idct plus interpolation
15:53:04D_C_oh my having pokemon red on my iriver is gonna make school a whole lot more fun
15:53:24[IDC]DragonI planned for this, but nobody complained about the aspect
15:53:31[IDC]Dragonso I skipped that
15:53:38 Join theli_ua [0] (
15:53:46[IDC]Dragonbesides, it would neen more memory
15:53:55amiconn5x4 idct would be the only size with a real benefit
15:54:21 Quit bbroke (Read error: 104 (Connection reset by peer))
15:54:31amiconnInterpolation can be done realtime for display
15:55:42[IDC]DragonI can send you the 10*8, if you like
15:55:50 Quit webguest35 ("CGI:IRC")
15:55:55LinusNbluebrother: define "doesn't work"
15:56:16LinusNbluebrother: do you know for sure that all settings fit in one 512-byte sector after your addition?
15:56:57bluebrotherLinusN, no, haven't checked that. I assumed there is enough room left.
15:57:20LinusNcould be worth checking
15:58:09bluebrotherbut does this also apply for that struct? Even when not saving the setting to disk?
15:58:20amiconnNo it doesn't
15:58:45 Join leftright [0] (
15:59:28leftrightany manual editors around ?
15:59:34bluebrotherthen I don't understand why I sometimes have a shifted position.
15:59:37bluebrotherleftright, yes.
16:00:36bluebrotherLinusN, the problematic line is if(global_settings.theme_file[0] != 0) in smart_open −− when I append theme_file to the end of the struct on h120 the value starts at theme_file[8].
16:00:59leftrighthi there, looking at section 5.2 of the manual, it shouldnt be neccessary to NAVI/Right to make that selection, NAVI imples press down, and that isn't neccessary
16:01:56leftrightbluebrother: thats at section 5.2 Navigating the Main Menu
16:02:16bluebrotherleftright, what manual you are referring to? Iriver I guess?
16:02:32leftrightyes, the H140 html manual
16:02:36bluebrotheryou can use that two buttons alternatively so they should be named.
16:02:36Quelsarukguys, it was nice to see you again after so many months. See you another day. Enjoy coding.
16:03:02 Part Quelsaruk ("Time makes no sense")
16:03:17leftrightbluebrother: description of NAVI implies press down
16:03:48bluebrotheryeah, but it isn't written as Navi + right.
16:04:14bluebrotheras it is an alternative, not a button combination. How else should this be written?
16:04:16leftrightyes but thats what it means to me, press down+right
16:05:45bluebrotherdo you think it would be better writing it as Navi, Right? This could be misunderstood to be a key sequence.
16:05:54LinusNbluebrother: no, i can't see any immediate flaws in your code that could cause this behaviour
16:06:46bluebrotherwhile the Play+Up should be clear.
16:06:50leftrightbluebrother: problem is that you are refering to the joystick as "Navi" and the action of pressing down on the joystick as Navi as well
16:07:34bluebrotherLinusN, thanks. I wanted to submit that patch to the tracker but only after solving that :(
16:08:10LinusNbluebrother: i guess resetting the settings doesn't help?
16:08:24bluebrotherunfortunately not.
16:08:35bluebrotherI also tried it on the target (h120), same.
16:08:44LinusNbtw, do you clear theme_file when resetting the settings?
16:09:23LinusNah, yes i see that you do
16:10:07bluebrotherbut when I swap the location it works. Currently I can load tarred themes on my player but it doesn't get remembered after reboot.
16:10:29leftrightbluebrother: i think that there should be two distinct functions for the joystick, i.e. joystick right and NAVI which implies press joystick down
16:11:29bluebrotherleftright, we're referring to the joystick as navi? As far as I understood we simply name the "buttons" of the directions and the press.
16:11:36leftrightQuote: Pressing the joystick down is labelled Navi throughout this manual. Its directions are labelled Right, Down, Left and Up.
16:11:58bluebrotheralso there is no combination that uses press + direction.
16:12:50leftrightaccording to that statement in the manual, whenever i see NAVI I must press down on the joystick
16:13:33bluebrotherleftright, sure, but NAVI / Right is an alternative. I find this naming pretty obvious.
16:14:01leftrightNAVI/RIGHT implies press down and right, not right select
16:14:11 Quit JdGordon ("Konversation terminated!")
16:14:17tucozleftright, Navi+Right implies that
16:14:29tucozNavi/Right implies Navi or Right
16:14:30bluebrotherhmm. Seems to be something that could be misunderstood.
16:14:35leftrightno NAVI implies press down
16:14:52leftrightquote: Pressing the joystick down is labelled Navi throughout this manual. Its directions are labelled Right, Down, Left and Up.
16:15:16tucozhmm, that sentence could perhaps be improved
16:16:38LinusNbluebrother: so it happens in the sim as well?
16:17:07bluebrotherLinusN, yes. I first had this in the sim and found it to be the same on the target.
16:17:21LinusNok, should be easy to debug then
16:17:29 Quit sm007h ()
16:18:03bluebrotherusing gdb I discovered that for the h120 sim the string itself starts at an offset of 8, for the h300 at 4.
16:18:42bluebrotherI also tried creating a pointer to global_settings.theme_file. It pointed to a different memory location.
16:18:55bluebrotherbut I have absolutely no clue what's happening there.
16:21:43leftrightperhaps it should be noted in the description of Navi that Navi+right means press down and right, and, Navi/right means joystick to the right
16:22:12StroggI don't know anything about the docs or the source.. but that sounds confusing to me.
16:22:21bluebrotherleftright, I thought about adding something like that in the getting started chapter, 1.3
16:22:26Genre9mp3Does anyone know if bmp progress bars work on rwps? To be more accurate, on X5 remote?
16:22:38amiconnLinusN, bluebrother: I guess some .h file redefines struct packing. So .c files which include that .h use one definition, and those which don't use another
16:22:58LinusNamiconn: i'm thinking that too
16:25:23LinusNbut i can't see where that would be done
16:25:40leftrightproblem is that Navi has two meanings, 1) joystick. 2) press joystick down, the manual does not diffirentiate between these two
16:27:21bluebrotheras far as I understand it iriver refers to clicking the joystick as navi. As on h300 it actually is a button.
16:28:07leftrightand that is the problem, on the H1xx's its a joystick
16:28:12 Quit MarcoPolo (Remote closed the connection)
16:28:55Genre9mp3Nevermind my previous question...
16:29:30leftrightNavi on the H3xx descirbes the button function well, but on the H1xx it has dual meanings
16:29:40LinusNleftright, bluebrother: why are we even discussing this? afaik, none of the iriver players can combine select and directions
16:30:12leftrightits the noobs that i worry about, the manual is ambiguous for the H1xx with regards to Navi
16:31:13leftrightyes, developers might know that those selections aren't possible, but not all users know that
16:31:24bluebrotherhow about simply noting "a+b" means press both buttons and "a/b" means an alternative? I think this should do it.
16:31:28*Strogg enjoys his new freedom :)
16:32:01LinusNbut when and why do we combine navi with a direction in the manual?
16:32:26leftrighterr, isn't the horizontal scroll thing an example
16:32:55bluebrotherno, that is holding the buttons and was play + direction before.
16:33:08bluebrotherwhich is a point where the manual isn't up to date.
16:33:22leftrightthat was incorporated to a hold down the joystick and right select
16:33:32LinusNno, not hold down
16:33:37LinusNthat is impossible
16:34:19leftrightLinus, I threw that feature away after trying it once, not a nice feature at all, to me
16:34:52LinusNi believe you, since you obviously tried to hold down the joystick while scrolling
16:35:01LinusNthat must have been a pain
16:35:02bluebrotherleftright, I just looked into iriver's manual. "Press the NAVI/MENU button"
16:35:20 Join lowlight [0] (i=c730190b@gateway/web/cgi-irc/
16:35:40LinusNleftright: what made you think that you had to hold SELECT as well?
16:35:44bluebrotherso they're calling it similarly.
16:35:49leftrightyes it says "Press"
16:36:12bluebrotherhmm. Do we really need to explain how to use buttons?
16:36:20leftrightand again i quote: Pressing the joystick down is labelled Navi throughout this manual
16:36:45LinusNwhich is correct imho
16:36:49leftrightNavi/Rignt means
16:36:54bluebrotherand where it said you need to use Navi + Right?
16:37:06LinusNwhere does it say navi/right?
16:37:41 Quit ElCabron ("CGI:IRC (EOF)")
16:38:13leftrightsection 5.2
16:38:24LinusNjust found it
16:38:37bluebrotherleftright, but how do you understand things like "Up/Down" (4.1.1)
16:38:48LinusNthe other sections say "or" instead of /
16:38:53 Quit theli_ua ("Leaving")
16:39:17bluebrotherhmm. We could just replace all those lines to use the word "or".
16:39:19leftrightI "think" it means select joystick up or down
16:41:22lowlightLinusN: Hi. I've got a question about gwps...
16:41:42lowlightLinusN: I don't see the need for format_align being an array as it never seems to be referenced without doing a format_display first
16:41:43 Join Moos [0] (i=c4ceecc8@gateway/web/cgi-irc/
16:42:11*Strogg wanders around and hugs everyone.. then goes back to playing with his ipod. :)
16:42:15MoosHello all !
16:42:20StroggDid I mention how much I hated itunes? lol
16:43:18lowlightLinusN: the code never seems to reuse the alignment if it's already been formated
16:43:33leftrightStrogg: be sure to tell Mr Jobs that
16:43:49LinusNlowlight: it's been a long time since i looked at that code
16:44:34Stroggalso cool that the photo stuff is actually useful now.
16:45:17LinusNlowlight: i have looked at your <no files> patch
16:45:26 Quit idnar (Nick collision from services.)
16:45:29 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
16:46:18 Join Hansmaulwurf [0] (
16:46:44lowlightLinusN: my "wannabe" patch :)
16:47:13 Quit merbanan_ ("Lämnar")
16:47:22LinusNlowlight: i haven't tried it though. what happens if you try to select the <no files> entry?
16:48:50LinusNi'd rather have the patch to not add a dummy entry, but instead display the <no files> separately, so to speak
16:48:56lowlightLinusN: Nothing...if I recall the action is based on the number of files in the dir (which is still zero) not the number of entries in the list
16:49:46LinusNi have to go
16:49:54LinusNcu later, i hope
16:50:06 Part LinusN
16:54:15 Join Paul_The_Nerd [0] (
16:54:50Strogghail Austin!
16:56:48StroggDon't mind me.. just saying hi.
16:57:21*Strogg is in that sleepy, caffienated sleepy state where he usually makes no sense
16:57:45 Join rretzbach [0] (
16:58:24 Join webguest04 [0] (i=c3ec4b1b@gateway/web/cgi-irc/
16:58:55 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
16:59:26*Paul_The_Nerd is in that state where he doesn't want to spend several hours today learning STL but has to due to silly requirements.
16:59:26 Quit jaebird (Read error: 104 (Connection reset by peer))
16:59:40 Quit stamppot ("CGI:IRC (EOF)")
16:59:44 Join jaebird [0] (n=jae@
17:01:25leftrightbluebrother: the playlist "Insert" function at 4.1.3 is described incorrectly in html, the correct description is "Insert positions file after previous file added to playlist, (if previously added file was Insert last then, Insert positions new file after that one)"
17:03:02 Join webguest58 [0] (i=c27f0812@gateway/web/cgi-irc/
17:04:17 Quit Moos ("CGI:IRC (EOF)")
17:06:52bluebrotherleftright, I don't understand your corrected version ...
17:11:00Genre9mp3Hmmm...It seems that using Unifont breaks bmp loading in WPS completely on X5 remote
17:11:19Genre9mp3At least this happens on the sim
17:11:20leftrightbluebrother: "insert next" was used on ther previous file then "Insert" will add new files after that one, if "Insert Last" was used for previous file then 'Insert' will position files after that one
17:12:09tuweGenre9mp3: is unifont the default font?
17:12:32bluebrotherhmm. What will happen if none of those was used before?
17:12:48tucozleftright, it would be great if you could sum up your findings in a bug report. (To make sure we do not forget them)
17:13:16bluebrotherI always found "insert" to be strange and use only the other options ;-)
17:13:19 Join dpassen [0] (n=dpassen1@resnet-236-163.resnet.UMBC.EDU)
17:13:28 Quit _FireFly_ ("Leaving")
17:13:41Genre9mp3tuwe: No, but as I experience it on the sim, bmp rwps on X5 are impossible when using unifont
17:13:47leftrightone has to understand that insert depends on what was used previously or last
17:14:01Genre9mp3I'm doing some tests right now to be sure
17:14:54tuweGenre9mp3: nervermind, i just overlooked the filename rockbox_default.fnt in the fonts dir. :)
17:15:03leftrightif you use "insert next" to add files to playlist then "Insert"will position new files after those
17:15:35tuwei have problems with bmp display on the ipod mini simulator, though.
17:16:22leftrightIf you use "Insert Last" to add files to the playlist then use "Insert" it will position additional files last
17:16:39 Part Paul_The_Nerd
17:17:13 Part webguest58
17:17:39leftrighttucoz, i'll do so
17:18:00Genre9mp3Hmmm...It seems that something else is wrong.... I just managed to load a bmp on rwps while using unifont
17:19:58 Join solarflare [0] (
17:20:08 Quit solarflare (Remote closed the connection)
17:20:12 Join solarflare_ [0] (
17:30:00 Quit dpassen1 (Read error: 110 (Connection timed out))
17:33:27 Quit solarflare_ ("Leaving")
17:35:56pondlifelinuxstb: What do the "ld" errors in the build table mean?
17:39:27leftrightthe subject of simplifying the Playlist "Insert" menus has been brought up before but shot down, I think that the "Insert" menu can be removed, as most folks have no understanding of how it actually works, and it seems that most folks use either "Insert Next" or "Insert last", rendering "Insert" superfluous, it also would unclutter the Playlist Menu
17:40:19amiconnInsert isn't superfluous at all
17:40:54leftrightyes but one has to understand how it works to utilise its potential, which isn't obvious
17:43:27leftrightobviously when creating an Initial playlist "Insert" is required, but thereafter "Insert Next" and "Insert Last" are accurate and easily understood.
17:43:47Genre9mp3Heh...unifont is too small for Gigabeat's screen!
17:43:56linuxstb_pondlife: It means Bagder needs to update his build scripts.
17:44:15pondlifeOK, nothing much to worry about then
17:44:28linuxstb_No - the builds are fine, just not what Bagder's scripts expected any more.
17:44:34tuweGenre9mp3: why does it affect bmp loading?
17:45:32Genre9mp3Tuwe: No, I was wrong, It works ok...For a moment I thought it was a bug but this hopefully is not the case
17:47:57***Saving seen data "./dancer.seen"
17:50:38 Join HyperNotActive [0] (
17:53:18 Part HyperNotActive ("Leaving")
17:53:39 Join MarcoPolo [0] (
17:54:21 Part linuxstb_ ("Leaving")
17:55:28Arathistucoz: I read about a manual for the H10 you made. where can I find it?
17:56:10 Join ToyKeeper [0] (
17:57:25bluebrotherArathis, the manual will appear on the manuals page once it gets included in the manual build process.
17:57:48bluebrotherso up to now you need to build it yourself.
17:58:34Arathisis it in den 'rockbox' cvs? or do I need 'rockbox-devel' ?
17:59:07bluebrotheryou need the module "manual". I don't think rockbox or rockbox-devel include it.
17:59:15 Join apo` [0] (
17:59:42Arathisthanks anyway.
18:00:19 Quit apo (Read error: 110 (Connection timed out))
18:01:56 Nick tuwe is now known as the_fuchs|weg (n=the_fuch@
18:02:13 Nick the_fuchs|weg is now known as tuwe|away (n=the_fuch@
18:06:27 Part [IDC]Dragon
18:19:52 Join bluey- [0] (
18:26:23 Join Criamos [0] (
18:27:23 Join barrywardell [0] (n=barrywar@
18:30:01 Quit _Lucretia_ (Read error: 110 (Connection timed out))
18:31:27 Join _Lucretia_ [0] (
18:34:01 Quit bluey- (Remote closed the connection)
18:34:02 Join CriamosAndy [0] (
18:34:39 Part leftright
18:36:57 Join Febs [0] (
18:37:54lorijhowhat's the difference between GPIO*_ENABLE and GPIO*_OUTPUT_EN ?
18:50:16 Quit Criamos (Read error: 110 (Connection timed out))
18:54:30 Quit JoeBorn (Remote closed the connection)
18:55:02 Quit XavierGr ()
19:01:55 Quit dan_a (Read error: 110 (Connection timed out))
19:05:56*Bagder poked the build scripts, might be less reds on next commit
19:06:57lowlightamiconn: I have a patch that reduces the code enough to build the FM rec
19:07:05lowlightamiconn: I *don't think* that gwps.format_align[][] needs to be an array. But I'd like another set of eyes to verify.
19:07:50lorijhowhat's the difference between GPIO*_ENABLE and GPIO*_OUTPUT_EN ?
19:11:14 Join bluey- [0] (
19:12:38 Quit uski (Read error: 110 (Connection timed out))
19:13:01 Join uski [0] (
19:13:11 Quit Rob2222_ (Read error: 104 (Connection reset by peer))
19:13:15preglowi've got this ipod mini 1g here that just displays an electrical socket when i switch it on
19:13:19preglowand sometimes a battery
19:13:31ArathisDo I understand the license the rockbox manuel ist distributed with allows to print the manual for personal use even through a service like ?
19:13:36 Join Rob2222 [0] (
19:14:29preglowwhat's up with that?
19:15:59 Quit MarcoPolo (Remote closed the connection)
19:17:08amiconnpreglow: Probably discharged
19:17:41 Join Nico_P [0] (
19:17:49preglowamiconn: then why doesn't it do anything when i plug in usb?
19:17:56preglowdo minis need a wall wart or something?
19:18:12amiconnIt should charge from USB
19:18:16 Join Lear [0] (
19:18:23 Join bbroke [0] (
19:18:55amiconn...but when battery is low, it won't connect to USB, until enough battery charge built up
19:19:08amiconnThat's how the mini g2 does it
19:19:23preglowok, we'll see
19:19:30preglowbut they do charge from usb?
19:19:48amiconnI even have no extra wall wart, charging exclusively from USB
19:20:05preglowa mate wants rockbox on his mini 1g, which won't happen unless it starts acting up, heh
19:20:12preglowstops acting up, even
19:21:12Joey_Di charge my mini via firewire
19:21:32 Join dan_a [0] (n=dan-mirc@
19:21:40amiconnlowlight: I cannot see how this patch would reduce code size enough to make rombox fit again for the fmrecorder
19:21:59amiconnBut it saves RAM, so it should still be a good thing to do
19:22:01 Quit pondlife ("byebye")
19:23:50amiconnBasically a test should tell whether everything still works like it should. Afaics the 2 main points to test are (1) Different alignment on several lines. (2) whether it also still works with alternating sublines
19:24:08lowlightamiconn: ajbrec.ajz is 236528 with patch
19:24:25Nico_Ppreglow: have you seen the musepack seeking update ?
19:25:13amiconnlowlight: In order to make rombox fit again, we need to get rombox.bin down to or below 233472 bytes
19:25:54amiconnThe array is uninitialised data, i.e. BSS and not included in the binary
19:26:10amiconnBut as I said, saving RAM is also a good thing
19:27:15lowlightok. (but it builds...otherwise it complains that flash is full)
19:27:48tucozArathis, around?
19:28:13amiconnThat's something I don't really understand... the linker doesn't complain immediately when the binary becomes too big
19:28:21 Join dishnecto [0] (
19:28:57preglowNico_P: i have indeed, i will commit it tonight after having a look at it
19:29:12dishnectohey guyz how cud i watch a video file on my rockbox
19:29:37amiconnIf you had checked the build logs for the fm before it started to complain about region flash, you would have seen that the build scripts checked the binary size and decided to remove it again because it was too big
19:30:00amiconnBut that means that your patch also reduces binary size at least a bit
19:30:15amiconnNot enough to make it fit again, but enough to stop the linker complaining
19:31:08lowlightyes, I saw the warning (3056 bytes too large)
19:32:02tucozArathis, in case you plan to print the H10 manual, i would suggest you wait for some time with that. It is far from complete.
19:33:22tucozArathis, and I believe it is perfectly fine to sell the Rockbox manual
19:33:34tucoz(from a licence point of view)
19:34:19 Quit dishnecto ()
19:35:16 Quit barrywardell ()
19:35:54lowlightamiconn: looks like format_align is only used after calling format_display (around line 1779 gwps-commmmon.c) which reinitializes the struct. The only references to format_align follow before incrementing to the next line/subline.
19:37:10 Quit lostnihilist (Read error: 54 (Connection reset by peer))
19:37:14amiconnYes it looks like that
19:37:31amiconnAnd removing the array saves almost 2KB of RAM on archos, more on targets with larger LCD
19:38:02amiconn13 lines * 12 sublines * 3 pointers (12 bytes)
19:41:29lowlightWhy not reduce the MAX_IMAGES for they need 46 images?
19:42:38amiconnI dunno... some wps designs use lots of images, even if they're small
19:42:41bluey-why isn't rockbox ignoring hidden Windows files?
19:43:10amiconnbluey-: It does, unless you have set 'View files' to 'All'
19:43:16dan_aI'm about to commit a change moving all the PortalPlayer stuff out of crt0.S into a separate file, unless there are any last minute objections...
19:43:40tucozYou can at least the remove the number of codecs the sw platforms have from MAX_IMAGES
19:43:48amiconnImho crt0.S should be split into each target architecture's part
19:44:13dan_aamiconn: I agree - is there any reason not to do that in small steps, though?
19:44:37amiconn...i.e. all arm stuff to firmware/target/arm/, with the pp stuff either ifdefed or in a second file etc
19:44:48amiconnOf course it can be done in small steps
19:45:11 Quit Arathis (Read error: 110 (Connection timed out))
19:45:27dan_aThe PP stuff is going to firmware/target/arp/crt0-pp.S - the rest (including the non-PP arm code) I'm leaving for now
19:45:45 Join Arathis [0] (
19:45:57amiconnWhy not move the arm stuff as well?
19:46:24tucozArathis, did you get my reply?
19:46:37Arathiswhich one?
19:46:45*amiconn already moved some asm code to target tree, and found that .S files usually are the easiest ones to move there
19:46:49tucoz in case you plan to print the H10 manual, i would suggest you wait for some time with that. It is far from complete.
19:46:56dan_aamiconn: OK - though if I'm doing that I may as well do all of crt0.S
19:47:04tucoz and I believe it is perfectly fine to sell the Rockbox manual
19:47:12Arathistucoz: nope, didn't get that. thx
19:47:19*amiconn should split lcd.S completely
19:47:25tucoz(from a licence point of view, as long as the licence is kept in the manual)
19:47:35amiconnI already took out h300 - also sorta incomplete...
19:47:42Arathisbut I thought that before. I was just wondering if it *would* be allowed
19:47:58***Saving seen data "./dancer.seen"
19:48:52tucozAs long as the licence allows it, it's allowed
19:49:28Arathistucoz: that's what I was asking about. I don't really get the whole license :)
19:50:24tucozArathis, i understand you. I don't either. But the GFDL is a copyleft licence, which basically says: 'Do what you want, as long as you keep the licence in the manual'
19:51:08Arathisgreat. 'cause I don't like reading 130 pages at the screen xD
19:51:19StroggYou could always ask a lawyer.. the $30 investment would probably be worth it
19:51:58Arathisnot if I just do it for personal use and won't redistribute it
19:52:04tucozArathis, if you want to help with getting the manual complete, all help is greatly appreciated.
19:52:59tucoze.g. complete the key-tables, add screenshots for the h10 5GB manual etc
19:53:08Arathisbut $7.13 (5.81€ for me) would be a nice price for it :)
19:53:44Arathisatm the moment I have other things todo for myself, but perhaps I'll take a look at it
19:54:26tucozI just wouldn't want you to spend $7.13 on something incomplete
19:54:48Arathisbut I wanted to wait till the H10 port is more "mature" before I test it myself
19:54:59Arathisme neither ;)
19:55:59dan_aWhat architecture is a TCC730?
19:57:47amiconndan_a: calmrisc16. Used in gmini 1xx and sp
19:58:01amiconnThe port has stalled for more than a year or so
19:58:11amiconnVery difficult to work with
19:58:25amiconnharvard architecture, 16 bit
19:58:36dan_aDoes that need a new target tree? Or should I leave just that in firmware/crt0.S?
19:59:43ArathisI'm downloading rockbox-all and it stucks at "cvs checkout: Updating manual/rockbox_interface/images"
19:59:48amiconnGood question...
20:00:45ArathisI just want to build the manual (for H10, testing reasons), but don't know which parts I need for this. That's why I get -all
20:01:29amiconnIt's very probably not stuck
20:01:38amiconnimages/ just takes a while
20:02:33dan_aIf I leave it in firmware/crt0.S then that means I can add a comment to say where the rest of the file has gone
20:03:02Arathisamiconn: I did not stop the time, but I think it's hanging there for about 30 minutes or more
20:03:54 Quit lowlight ("CGI:IRC (EOF)")
20:04:02amiconnLinus already thought about throwing out the stalled port
20:04:27amiconnI think it's easier for a possible reactivation to keep the code around
20:04:39bluebrotherArathis, if you want I can build you a manual
20:04:41amiconnBut if we do, it should also be moved to target tree
20:04:53Arathisfor H10?
20:05:09amiconnThen someone needs to build the hyper-special gcc and check whether it still builds
20:05:12bluebrothersure, if you want that ...
20:05:23Arathiswood be great, thanks
20:05:31bluebrotherthe 5bg one?
20:05:37 Join darkless [0] (n=darkless@
20:05:39Arathisbetter the 20GB
20:05:43amiconnArathis: 30 minutes really sounds a bit much
20:06:32Arathisamiconn: yes. but do I really need all files? and is there a way to redownload all but the already downloaded files?
20:08:42amiconndan_a: Afaiu the correct target path would be firmware/target/calmrisc16/ with sub-dirs archos/gmini1xx/
20:09:02bluebrothercvs will notice if you have already checked out files
20:09:12Arathisbluebrother: uuh, a german fellow :) thanks
20:09:29 Join damaki [0] (
20:09:33amiconnBut probably leave it in firmware/ for now and we discuss this later
20:09:44bluebrotherI'll remove the pdf in some time ... I'm always low on web space ;-)
20:10:15Arathisbluebrother: saved to desktop :)
20:10:39preglowis it just me or is the rockbox server lagging badly?
20:11:27Arathisamiconn: it's proceeding after a restart now
20:16:11 Join MrShlee [0] (
20:16:20Arathisbluebrother: good you made me a manual. for me it says "WARNING: The compiler you must use (arm-elf-gcc) is not in your path!" :(
20:16:53MrShleeHey, I just got a H340.. is Rockbox safe on that edition?
20:17:34Arathisbluebrother: I can't find that in my repos at all. (ubuntu dapper)
20:17:42bluebrotherwhich page that happened?
20:17:52bluebrotherah, on configuring.
20:18:09bluebrotherYou can ignore this, you only need LaTeX for the manual, not a compiler ;-)
20:18:17 Quit MrShlee (Client Quit)
20:18:22belzeMrShlee: i would say so
20:19:18 Nick JBonBike is now known as JoeBorn (
20:19:41Arathistucoz: which parts of the manual are missing/incomplete?
20:21:26Arathisbluebrother: but I would need it if I'd want to build rockbox, right? too bad. but good that there are daily builds :)
20:21:55bluebrothercorrect, but you can also use that build-environment-building script
20:22:33Arathishaven't even latex installed XD
20:22:43dan_aI've not done a "cvs add" before - what do I need to do with the $Id: lines?
20:22:50bluebrotherfor building the manual you should ;-)
20:22:59bluebrotherbut it doesn't hurt
20:23:11Arathisbut which package do I need?
20:23:13billytwowillyholy carp. bagder is here... but he's away.. too bad
20:23:26*billytwowilly was hoping for an update on sansa e200 rockbox port
20:23:29 Join MrShlee [0] (
20:23:47bluebrotheron linux the "standard" latex package, in most cases called tetex-latex (and all its dependencies)
20:24:12dan_abillytwowilly: It's pretty much as it says on the forum thread. We can't write to the screen or the disk yet, so we can't tell if our code is working...
20:24:36 Quit Genre9mp3 ("I don't suffer from Rockbox psychosis. I enjoy every minute of it.")
20:24:38billytwowillyI was hoping there would be a break through on that;)
20:24:46dan_aGive it time!
20:25:09MrShleeLets try this again.. any issues with the H3XX releases?
20:25:31belzeMrShlee: still no
20:25:36bluebrotherdan_a, the lines with $Id:$ get replaced by cvs, so you don't need anything do about them
20:25:48 Quit MrShlee (Client Quit)
20:25:53dan_abillytwowilly: Bagder is pretty busy at the moment, so it's down to people with less experience to try things out
20:25:58dan_abluebrother: Thank you
20:26:16billytwowillydan_a: yah, no doubt. I appreiate all the work you guys do on this stuff. I'm just excited because rockbox on that player would give me bookmarking, which I need in any new player I wish to buy.
20:26:32Arathisbluebrother: that isn't in my repos. only ttex-base and things.
20:27:26 Join TeaSea [0] (
20:27:30billytwowillydan_a: and with the wps and theming stuff I think rockbox is the better firmware
20:27:49bluebrotherttex sounds good. I'd try ttex-latex ...
20:30:29 Part JazzBone
20:30:41Arathisneither ttex nor ttex-latex is included :(
20:31:54Arathistucoz: the H10/20GB manual would even be $6.65/€5.42 instead as it is with 106 pages :)
20:32:58 Join MrShlee [0] (
20:33:25Arathisbtw, are the manuals updated parallel to the CVS changes for every port?
20:33:38MrShleeQuick question: is there any plans to release an updated firmware bootloader for H3XX
20:33:44MrShleeor just when its done
20:35:39 Quit MrShlee (Client Quit)
20:37:02bluebrotherArathis, the manuals are updated daily (but not their content ;-)
20:42:56bluebrotheryou could try installing latex-ucs and waiting for the dependencies it pulls
20:43:11Arathisthere are no. i tried that before
20:43:34bluebrotherthe ubuntu repository layout seems pretty strange to me. Hard to find packages, even guessing their names.
20:44:14Arathisthat's right. perhaps I'll switch the distribution someday
20:44:35Arathisbut atm I like ubuntu
20:44:40bluebrotherbut I found tetex-base and tetex-bin packages. Maybe you're missing some repositories?
20:44:58Arathisperhaps, but than they would be non-standard
20:45:20tucozArathis, in ubuntu it's tetex-base tetex-extras
20:45:28Arathisoh, thanks
20:46:01tucoztetex-extra not extras :)
20:46:23Arathistucoz: but you don't know where to get the arm-elf-gcc for ubuntu, do you?
20:46:34tucozArathis, you build it yourselft
20:46:43bluebrothertetex should be part of every serious linux distribution.
20:46:46Arathiswhere to get it?
20:47:01tucozjust run the script from the tools folder in the rockbox cvs tree
20:47:02bluebrotheror use the script I mentioned earlier
20:47:31tucozyou can edit that to set the path to where it is installed
20:48:03tucozin case you want it to be installed in your /home/arathis/arm directory for intanse
20:48:13ArathisI'll try that when the port has developed to a point I'd use it
20:48:53Arathistucoz: did you get my question about h10-manual prograss erlier?
20:49:12*dan_a thinks he has spotted a small Sansa bug
20:49:51tucozjust read it. You will see some white space in the key tables. That is, where the keys are defined.
20:50:09tucozinstall instructions for the H10
20:51:16tucozthe H10 was added to the daily builds a couple of days ago, and the same is true for the manual. So, I only added the proper definitions so that you can build a manual.
20:52:30 Join Cris_Cr0ss [0] (
20:52:43 Quit Snyper (Read error: 110 (Connection timed out))
20:52:46 Join Snyper [0] (
20:54:22tucozall places marked in red in the manual needs attention as well
20:55:17amiconndan_a: What is it?
20:55:57*tucoz is afk
20:56:09 Quit bluey- ("Leaving")
20:56:51 Quit ender` (" You're not drunk if you can lie on the floor without holding on.")
20:57:11dan_aamiconn: Nothing that will be affecting us yet, AFAICT. CONFIG_CPU on the Sansa is defined as PP5024, and there is nothing in for that, so the linker would give incorrect memory locations
20:58:25Cris_Cr0ssman, i give up.. rockbox is too difficult to navigate
20:58:43linuxstbdan_a: Yes, I think Bagder is just using bootloader builds at the moment - i.e.
20:59:20dan_alinuxstb: Ah, yes, that's set up correctly
20:59:47*linuxstb is about to add Rockbox target #25
21:00:29 Join ender` [0] (i=null@
21:00:45linuxstbIronically, with 12 buttons, the Elio doesn't have anything I would like to call BUTTON_LEFT and BUTTON_RIGHT...
21:00:46dan_aDid you ever spot what the problem might be with the COP patch?
21:01:03linuxstbAnd Rockbox assumes all targets have buttons with those names...
21:01:20*amiconn wonders hwo long it will take to figure out the flash disk interface on pp5024
21:01:52linuxstbdan_a: No, I've given up for now. But I'll go back to it soon.
21:02:21 Quit [San] (Read error: 110 (Connection timed out))
21:02:59dan_aamiconn: Quite a while, I should think. Though the recovery mode being in the bootloader should help, for those who can read thumb code well.
21:03:20linuxstbdan_a: Are you doing any work on the Sansa?
21:04:02dan_aI am - I'm trying to get some sign that code we're running is working. Either by resetting it, or writing something to the screen
21:04:30 Join jhMikeS [0] (
21:04:35*amiconn is running a test to verify a possible pp bug
21:04:50amiconnNothing serious - the timer is probably off-by-one
21:05:25jhMikeSamiconn: you were talking about that gcc glitch earlier but weren't around. what's up?
21:05:35linuxstbdan_a: So "DEV_RS |= 4;" doesn't seem to work?
21:06:27amiconnI referred to the fact that m68k-elf-gcc <= 3.4.5 fails to find a free register despite the fact that there is one
21:06:42amiconnOtherwise building wouldn't work on any gcc version
21:06:56dan_alinuxstb: Not as far as I can see. The bootloader brings up a Sandisk logo, and then whatever I try to do it seems like the device has hung.
21:07:05amiconnBut the construct with the register save and restore block is really odd anyway
21:07:21amiconnI guess it leads to registers being saved twice, thereby wasting cycles
21:08:02linuxstbdan_a: But the same mi4 creation procedure creates a working original firmware?
21:08:35amiconndan_a: Is it possible that the sansa tries to start the main firmware in thumb mode?
21:08:43dan_alinuxstb: It does, and I'm sure Bagder has run one with modified strings
21:09:18jhMikeSamiconn: Is it not odd that the registers are deliberately being saved but also on the clobber list?
21:09:24dan_aamiconn: I've wondered that, but the OF *seems* to start in normal mode
21:09:40amiconnjhMikeS: Yes, that's really odd
21:09:43jhMikeSamiconn: What to do about it? Save regs for older gcc and use clobber list for newer?
21:09:54amiconnThe extra save and restore blocks should be completely unnecessary
21:09:58*dan_a adds "explicitly come out of thumb mode" to list of tests to perform
21:10:01 Join rotator [0] (n=e@rockbox/developer/rotator)
21:10:37amiconnI would use the clobber list exclusively, and drop the save-and-restore blocks
21:10:41jhMikeSyes...but I'd like it to build for those using cygwin
21:10:49jhMikeSOr just forget it?
21:10:51amiconnIf doing that makes it stop working properly
21:10:59amiconnI build on cygwin and have no problems
21:11:18jhMikeSwhat gcc version?
21:11:21Arathisbluebrother: I build the manual, but it's bigger than yours. ..
21:11:30amiconnI am using m68k-elf-gcc 3.4.6 though, built myself
21:11:49jhMikeSAhhh...I recommended that infamis do exactly that
21:12:03amiconnIt takes a while, but is worth the effort
21:12:35amiconnIn case he doesn't have special requirements for the compiler path, I could just zip mine and send it to him
21:12:46amiconn(my prefix is /opt/m68k )
21:13:14jhMikeSTold him "maybe rb has surpassed cygwin" ;D. Got some yucks from that one.
21:13:52amiconnIf I knew how to make cygwin packages, I could make a replacement package for the package hosted on (which still is 3.4.5)
21:14:14amiconnThis has nothing to do with cygwin itself
21:14:28jhMikeSMight be a good idea to keep the tools in sync there.
21:14:42jhMikeSIt was sort of a joke. :) block *poof*
21:15:54amiconnI wonder whether we would be better off using a more modern gcc for m68k as well
21:16:13amiconnWe switched to 4.x for all targets except coldfire
21:16:50amiconnBtw, the parameter list for this asm block is slightly wrong
21:17:21jhMikeSI haven't looked to much at it...what'd you find?
21:17:28amiconnenwindow is changed within the block, hence it is a combined in/out parameter
21:18:05jhMikeSit should be "+a" (enwindow) in the out block?
21:18:44amiconnI guess the save/restore blocks are there for covering more mistakes like that - in that case removing them will probably make the encoder glitch in some way
21:19:39jhMikeSthe second block changes yint but it's not in the out list
21:21:36*linuxstb does the dance of the 25th target
21:22:06ender`25th target..?
21:22:30dan_aAww, and I'd just made sure my CVS tree was up-to-date too!
21:22:57linuxstbender`: Tatung Elio TPJ-1022
21:23:10linuxstbdan_a: The changes were harmless (I hope)
21:23:24ender`Tatung... that's the brand of VGA monitor i had on my 286 :)
21:23:54 Quit uski (Read error: 110 (Connection timed out))
21:23:56linuxstbI also had a Tatung Einstein computer many years ago...
21:23:57mirakthis is the 100€ question of the day ((a >> 8) * 181) = ?
21:24:18mirakthat's a/sqrt(2)
21:24:19*dan_a tries to do an add from a directory checked out anonymously - oops!
21:24:35jhMikeSDoesn't the out block tell the compiler the value must be preserved subsequent to the block?
21:24:47jhMikeSasm block that is
21:24:47 Join uski [0] (
21:25:17amiconnIt just tells the compiler which values are output values
21:25:20dan_aIs there any way to change user in CVS - from read-only to read-write?
21:25:49jhMikeSWhat if it's not an output for use in C codec but simply altered in the asm?
21:25:54amiconnIf these values aren't used subsequent to the block, it won't preserve them
21:26:25mirakdoes anyone ever heard of a a solver, which can factorise or decompose equations like what some Texas Instruments calculators can do ?
21:26:30jhMikeSSo anything that's c var written to must be in the out block?
21:26:50amiconnBut if you don't tell gcc that the value is changed, gcc might try to reuse the register value, thinking it still holds the same value as before the block
21:26:54*jhMikeS is gett block and list messed up
21:27:59amiconnIf that doesn't hold, all sorts of weird things might happen, even crashes and freezes it tells in not to assume the value of something sort of like volatile in c
21:28:26amiconnIt becomes more likely to happen at higher optimisation levels
21:29:16amiconnWell, you can pass an asm block an intermediate result, e.g. [val1]"d"(a * b)
21:29:57amiconnIf you don't tell this value gets changed in the block, the compiler will probably try to reuse the register value if something after the block also needs the result of a * b
21:29:58crwlmirak, have an example equation?
21:30:07jhMikeSmakes sense
21:30:42amiconnSo you need to tell gcc it gets changed... and in turn you need to put the result of a * b in a temporary variable
21:31:03amiconn...because an expression is not an lvalue and hence cannot be put in the output list
21:31:47mirakcrwl: block[8*5] = (d0 - W6 * d1 - d2 + W2 * d3 - ((W7 * d7 * 1/sqrt(2) + W1 * d4 * 1/sqrt(2) - W3 * d5 * 1/sqrt(2) - W5 * d6 * 1/sqrt(2) - W7 * d4 * 1/sqrt(2) + W1 * d7 * 1/sqrt(2) + W3 * d6 * 1/sqrt(2) - W5 * d5 * 1/sqrt(2) ) >> 17;
21:32:12jhMikeSlike I said, the asm changes y_int in the second block but does not use it outside the second block again
21:32:14mirakcrwl: that's simple to do manually but I have 4 other lines, I am just lazy
21:33:00crwlmirak, ugh, that's longish, and i'm not much at all knowledgeable with mathemathics, but i *think* have factorized some equations with qalculate! ;)
21:33:25amiconnjhMikeS: You're sure? The second block is used within a loop... should probably be in out too
21:34:31amiconnPutting it in out will probably change nothing in the binary, but it's cleaner
21:34:39mirakcrwl: ok I remember this one, I didn't knew it could factorise
21:34:57jhMikeSwhat's the "=" again? (I'm new at the gcc asm stuff)
21:35:10amiconn= is output only
21:35:16amiconn+ is combined in/out
21:35:22jhMikeSbut it's in the "in" list
21:36:29amiconnNo, it's in the out list
21:36:36amiconnThe order is out - in - clobber
21:37:22dionoeahello world
21:37:27 Join barrywardell [0] (n=barrywar@
21:39:20jhMikeSOh... I missed the ":" in the middle there!
21:45:41 Join Paul_The_Nerd [0] (
21:46:05*dan_a does a commit which breaks his own patches
21:47:05FOADYou tramp.
21:47:18 Join lee-qid [0] (
21:47:57dionoeadan_a: do you have the second kernel runing on the COP now with interaction between COP and CPU threads ?
21:48:02***Saving seen data "./dancer.seen"
21:49:14 Join MarcoPolo [0] (
21:49:31dan_adionoea: Only on 3G and only with the cache turned off, so it's not in CVS yet. There's a patch on FS which I've just broken, so I'll get that fixed and posted back up soon.
21:50:03dionoea3G is pp5002 ?
21:50:38dionoeaso basically some addresses need to be changed and it should work on other pp targets ?
21:51:00dan_aThe patch I posted *should* have worked
21:51:27dan_aHas the CVS build page hung?
21:51:42 Join ChekS [0] (i=54f5165c@gateway/web/cgi-irc/
21:51:48ChekSgood day....
21:51:53StroggHah. funny.
21:52:31*Strogg notices that ChekS' day isn't based in the same universe that his is... hehe
21:52:59ChekSI have a question... would it be possible to make a firmware for a Medion mp3 player? It's the same chipset as the Ipod from Portalplayer ....
21:53:54ChekSI use the webclient .. since my laptop does not have Mirc
21:54:04dan_aChekS: Not unless somebody who has one is willing to do some hard work - see
21:55:02ChekShmmm i am a computer engineer... but i have not programmed for many years.. hehe... not my thing anymore.
21:56:04dan_aThere are 3 ports in progress for other PortalPlayer DAPs, so it might not be too hard
21:56:20 Quit Arathis ("[rl_bot quit]")
21:56:25ChekSbut since its the same inside as an Ipod... and Ipod has a Rockbox ... it cannot be that hard?
21:56:56linuxstbChekS: The first thing to do is to physically disassemble it and identify all the chips inside - especially the audio codec (DAC) and the LCD controller.
21:57:12linuxstbIf you can identify those, then a port will be easier.
21:57:17*Strogg wishes he had a DAC
21:57:36ChekSyes. I will do this. I think this is a nice project to start .. i have all the time in the world next to work and two childeren.. lol
21:57:58 Quit Siku ()
21:58:03linuxstbChekS: But you're right - porting Rockbox to other PortalPlayer DAPs is easier than porting to a completely new architecture - but no-one is going to do it without owning such a player.
21:58:19ChekSthe kewl thing about the medion.. it cost virtually nothing to buy second hand... i bought mine... 20 GB for 75 euro
21:58:20StroggBuy the dev team a box of players. hehe
21:59:13ChekSno.. i will do it myself if people can point me in the right direction.. i think i start reading how the ported it to the ipod.
22:00:05jhMikeSamiconn: If you have the time, check out
22:00:13jhMikeSThis bug is very strange indeed
22:00:57ChekSi wanna find more people that have that player.. and help me ..
22:01:05 Join carini [0] (i=chatzill@pdpc/supporter/active/carini)
22:01:22ChekSbut if not.... i will do it myself and screw my player up
22:01:58StroggI don't know if the same applies to your player.. but I screwed up my ipod 4-5 times when installing rockbox. hehe
22:02:13Stroggbut I was able to recover it kuz I kept a copy of the original firmware..
22:03:48dan_aVery few people have permenantly damaged their players by developing Rockbox on them
22:04:37ChekSyeah.. it's kinda strange way of upgrading.... it has a system folder where the is a formware file called "jukebox.MI" or something... and it loads that when it's start up on a FAT32 partition.
22:05:12ChekSyou can just replace that file when it's on USB disk mode.. and it's upgraded
22:05:23 Part Paul_The_Nerd
22:06:28ChekSi already installed a firmware from a 440 version.. on my 420 version and that worx fine...
22:06:30 Join lowlight [0] (i=c730180a@gateway/web/cgi-irc/
22:06:48ChekSso i think the insides of all the Medion Apods are the same.
22:07:27barrywardellChekS: is it a .mi4 file?
22:07:33dan_aI guess the 440 is the 40Gb version and the 420 is the 20Gb?
22:07:48ChekSno both 20 gb .. but a different look
22:07:55ChekSthe 440 is newer
22:08:26ChekSit's called JUKEBOX.mi4""
22:08:44lowlightjhMikeS: for your bug, are you using dircache?
22:09:04barrywardellCheckS: have you seen this page:
22:09:25ChekSa german board but for the picture
22:10:08dan_aChekS: Is there a recovery mode for it? Some way to correct it if you put a bad firmware on?
22:11:29ChekSdan_a yes there is a recovery_cd.exe file that does that
22:11:40jhMikeSlowlight: you suggest I try disabling it? There's another problem related to that. If you start recording while it's still'll see a dup file name in "recordings"
22:11:54ChekSbarrywardell thank you forthat site.. that helps a lot!
22:12:47barrywardellChekS: a port for the medion would work a lot like the Sansa, H10 and Elio ports
22:12:55jhMikeSlowlight: The files always play back that way on x5 anyhow...even after reboot so I have doubts about that being the source of the progress problem
22:13:06jhMikeSBut will try...who knows.
22:13:12barrywardellyou can probably already build a mi4 file and start testing code on your device
22:14:23lowlightjhMikeS: dunno, just a suggestion
22:15:02 Join Mmmm [0] (
22:15:15 Quit mirak (Connection timed out)
22:15:31ChekSbarrywardell . yes i think a lot of work is already done... i think i''ll do some investigation and some reading.. and going to try this.
22:15:51 Join _FireFly_ [0] (
22:17:14 Quit bluebrother ("Leaving")
22:18:47 Join mirak [0] (
22:21:21 Join XavierGr [0] (
22:22:38miraklinuxstb: the idct in the optimisation pdf is the same than in libmpeg2 but they optimised it to death
22:22:58*billytwowilly tries to picture optimizing something to death
22:23:08ChekSmi4 files are signed with DSA keys. I wonder how this works... does the compile tool generate them...
22:23:45 Quit MarcoPolo (Remote closed the connection)
22:23:50barrywardellCheckS: there's a trick to that
22:24:04barrywardellyou can use a dummy key - dsa(1,0) iirc
22:24:24barrywardellthe scramble tool will build a mi4 file for you
22:24:28mirakthe main trick is to remove the multiplication by cos(4pi/16) which is 1/sqrt(2) by multiplying with sqrt(2). Since it's 1 you don't need to multiply anymore
22:24:35barrywardellit's in the tools directory in Rockbox
22:24:37linuxstbChekS: You can probably just run the existing Elio TPJ1022 bootloader that I'm in the process of developing. It's a very very basic firmware that writes some debugging text files to your disk.
22:24:42miraksaving multiplications
22:24:51 Quit carini (Nick collision from services.)
22:25:15 Join carini [0] (i=chatzill@pdpc/supporter/active/carini)
22:25:24linuxstbChekS: If that works, then you are already a long way into a port. Next step would probably be to write an LCD driver.
22:25:58mirakonly thing I don't get is this : d0 = (block[8*0] << 11) + 65536;
22:26:07 Quit carini ("Chatzilla 0.9.75 [Firefox]")
22:26:18 Join carini [0] (n=chatzill@pdpc/supporter/active/carini)
22:26:30ChekShmm yes. This sounds good... this is going to be a nice project. I think the driver for the display is the same as any of Portalplayer?
22:27:10linuxstbNot really. Some seem to use the internal LCD controller, others have one of various external controllers.
22:27:50linuxstbThe H10 port was lucky - the driver is a mixture of two existing ones - the ipod and the iaudio X5.
22:28:04dan_aCould anyone give me a quick overview of what steps are required to send some output from a GPIO port?
22:28:04ChekSok.. first let me me open the thing up..hehe... i'll do this tomorow.. if you enjoy this like i do.. i can keep you posted with some pics.
22:28:51 Quit Mmmm ("Byeee")
22:29:46 Join Massa [0] (n=Massa@
22:31:16mirakbillytwowilly: that's a common expression here
22:31:20belzeChekS: i think its best of you post them on the wiki
22:31:31belzeso everyone can have a look at
22:32:20ChekSyes will do so.
22:32:32ChekSThanks everybody for your time and help
22:32:37ChekSBye Bye
22:33:31billytwowillymirak: optimized to death?
22:33:43mirakbillytwowilly: yes
22:34:05mirakbillytwowilly: dunken to death, whatever extreme but to death
22:34:20barrywardelldan_a: what do you want to do specifically?
22:34:43billytwowillyI suppose it's better than my mom's catchphrase "To die for"
22:35:28barrywardellis Bagder here? I have a third mi4 file for the downloads section.
22:36:19barrywardellBagder: the 5/6GB MTP H10s are 010301
22:36:21billytwowillybarrywardell: are you working on the sansa e200 port?
22:36:52Bagderbarrywardell: aha, do we get a name collision now then or do they use a different file name?
22:37:07barrywardellyes, name collision unfortunately
22:37:23barrywardellbillytwowilly: no, H10 port
22:37:33merbananbarrywardell: isn't it possible to boot both with the same firmware ?
22:38:05 Quit ChekS ("CGI:IRC (Ping timeout)")
22:38:13barrywardellbillytwowilly: dan_a is doing some sansa stuff
22:38:56billytwowillyyar. It sounded like you had a firmware ready to be tried or something;) Which would be great if it was teh sansa player;)
22:39:31barrywardellmerbanan: unfortunately not. there are two different 5/6GB versions - UMS and MTP. they both use different mi4 versions
22:39:43barrywardelland the 20GB has a different LCD
22:39:57 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
22:40:46merbananbarrywardell: ok, I forgot that the bootloader check the version number otherwise you could have just signed and calculate the checksum
22:40:56mirakI figured out the left shift 11 in d0 = (block[8*0] << 11) + 65536; now I am not sure for the 65536
22:40:59barrywardellBagder: i sent you the third mi4. maybe put them in different directories to avoid the naming issue?
22:43:04MassaIs it possible to display a bitmap and make sure it's _always_ in foreground and may not be overlapped e.g. by text?
22:43:07FOADH300 broken?
22:44:33jhMikeSlowlight: problem still occurs with no dircache :\
22:46:22dan_abarrywardell: There are a few GPIO pins which MrH has suggested we try to get the LCD working on the Sansa - so I just need to know if all I have to do is GPIOX_OUTPUT_VAL |= 1<<pin, or if there is more to it than that
22:46:48barrywardelldan_a: yes, that is all you have to do
22:47:03barrywardellyou may need to enable them first
22:47:16barrywardellGPIOX_ENABLE |= 1<<pin
22:47:36barrywardelland if you want to use them for output GPIOX_OUTPUT_EN |= 1<<pin
22:48:23dan_aThank you! I'll give that a go
22:48:37 Quit _Lucretia_ (Read error: 101 (Network is unreachable))
22:49:39FOADUhm help?
22:49:54 Join _Lucretia_ [0] (
22:49:56FOADI just installed new daily on my H300 and I don't get sound anymore.
22:50:11FOADAnd the second counter goes way too fast.
22:50:18FOADOf the song position.
22:50:30Bagderbarrywardell: that third mi4 is for "H10_5GB-MTP" ?
22:50:42Bagderor what is a suitable name?
22:50:46barrywardellyes, 5 and 6GB
22:51:00Bagderyeah, but I call them 5GB for simplicity
22:51:05barrywardellthat's fine
22:51:27lowlightjhMikeS: odd...and the copy that plays ok has the same checksum?
22:52:19Bagderbarrywardell: that one is now stored in a subdir on the download server
22:52:29barrywardellgreat, thanks Bagder
22:52:38lowlightMassa: I think it depends on the order things are sent to the lcd
22:53:12jhMikeSlowlight: a binary file comparison says they're identical so by implication...
22:54:07 Quit Febs ()
22:54:14 Join linuxstb_ [0] (i=5343d4aa@gateway/web/cgi-irc/
22:54:34 Quit vo|t (Read error: 110 (Connection timed out))
22:54:42mirakok my guess is that 65536 is added to compensate a lost of precision
22:54:56LinusNFOAD: 1) did you unzip the *entire* zip file? 2) have you tried to reset the settings?
22:55:09*jhMikeS just caused: "internal compiler error: in emit_move_insn, at expr.c:2809, Please submit a full bug report"
22:55:22Massalowlight, so actually I should be able to bring a bitmap in foreground by putting it in before the text in the WPS file?
22:56:21lowlightMassa: no, last
22:57:21Massalowlight, so "%xdasome text" should display the text in front of the bitmap if it overlaps and "some text%xda" should keep the text in background?
22:59:30 Quit webguest04 ("CGI:IRC 0.5.7 (2005/06/19)")
23:00:00lowlightMassa: I think the wps always displays all the text first, then does the images
23:00:38FOADLinusN: thanks, fixed.
23:00:49MassaHmm, it seems that scrolling text is always displayed last...
23:00:59lowlightMassa: try to follow gui_wps_refresh
23:01:33lowlighta scrolling text line is always updated in a separate thread
23:02:16Massalowlight: aah - that maybe the problem - so it's not possible to have a scrolling text scroll behind a bitmap?
23:02:32LinusNlowlight: how should the tar file look like?
23:03:48lowlightLinusN: tar file?
23:03:58LinusNyes, the .rbt file
23:05:02lowlightI'm not sure what you mean
23:05:17LinusNi'm an idiot
23:05:21linuxstb_Bagder: Thanks for fixing my tpj1022 target_id
23:05:36*Bagder bows
23:05:47LinusNlowlight: i mistook you for bluebrother
23:06:29linuxstb_Bagder: And sorry for breaking your build scripts this morning.
23:06:54linuxstb_And I was blamed for the redness.... :)
23:07:26Bagderlinuxstb: so have you gotten the LCD to show anything?
23:07:57linuxstb_No, not yet. I did some easy work instead... I think I need to dig into the bootloader disassembly.
23:08:24 Part LinusN
23:08:30linuxstb_(or kidnap MrH and lock him in a room with my Elio)
23:09:10MassaAnother question (related to that): when an image will be cleared, the function clear_image_pos is called - this just fills a rectangle with the background color (as far as I understand). This has some sideeffects for text which will be displayed on same position.
23:09:16Bagderand we can't even manage to reset the sansa...
23:10:05MassaIsn't there another possibility to clear an image?
23:11:56linuxstb_Bagder: What about branching to address 0, or somewhere else in the bootloader? Just to try and confirm your code is running?
23:12:12Bagderdan_a works on that
23:12:31Bagderhe even added code into the original firmware and ran that... with no success
23:12:54Bagderbut it might be that the reset isn't working properly
23:13:57linuxstb_Yes, that's what I mean - trying something else rather than reset.
23:14:24Bagderyeah, we should write code that changes a string or something
23:14:39Bagderor a bitmap
23:15:58dan_aAfter I've done these GPIO tests I want to try changing the location of the first jump in the original firmware to some non-code location, and then jumping from there to where the first jump should have landed.
23:16:32Bagdera good idea
23:18:43*Bagder has a bad case of "I'm just gonna clean up this..." and now my system dies hard on boot
23:20:33lowlightMassa: clear_image_pos doesn't fill a rect with the background displays the bitmap with inverted colors so it clears the original bitmap (unless there was text overtop of it)
23:20:51 Join [HO]vo|t [0] (
23:21:10 Quit preglow (Client Quit)
23:21:40lowlightMassa: forget that...I'm wrong
23:22:04 Join preglow [0] (n=thomjoha@rockbox/developer/preglow)
23:28:26 Quit lowlight ("CGI:IRC")
23:29:29dionoeadoes the docs system used by the manual allow translations through stuff like .po files or something ? (just curious for another project)
23:30:21BagderI don't think so
23:30:30linuxstb_It's just plain LaTeX - I've no idea if there are facilities in LaTeX for translations.
23:30:49dionoeaok, thanks
23:35:33dionoeabtw, rockbox must be the only opensource project with such a good documentation. I wonder how you got people motivated :)
23:35:54preglowtranslations as in how?
23:36:00preglowtranslate button names or stuff?
23:36:09preglowjust whatever is translated in rockbox?
23:36:11Bagderdionoea: its all the glory and appreciation you get from your family! ;-)
23:36:24dionoeapreglow: no, translate the full manual
23:36:42dionoeaBagder: hehe :)
23:42:32 Quit lee-qid ("Trillian (")
23:48:05***Saving seen data "./dancer.seen"
23:48:15 Quit Hansmaulwurf ("( :: NoNameScript 3.81 :: )")
23:52:02 Quit bbroke ("Chatzilla 0.9.75 [Firefox]")
23:56:44 Quit tucoz ("Leaving")

Previous day | Next day