Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2006-01-24

00:00:03JdGordonwhats the command to compile 1 .c file to a dll (a rock for the windows sim) so i dont have to do make every time?
00:00:16Bagderamiconn: yes, but a lot simpler
00:00:45amiconnI would expect cvs to be able to trigger external commands on a commit
00:00:55Bagderit can, it sends emails for example
00:01:27linuxstbJdGordon: There is no simple command.
00:01:35JdGordon:( bugger
00:01:49amiconnWhy is 'make' such a problem?
00:01:51JdGordonwhat about a not so simple command that is still faster than make?
00:01:58linuxstbIf you type "make V=1" you will see what happens.
00:02:06amiconnIt only rebuilds those parts which depend on what you've changed
00:02:06 Quit ender` (Client Quit)
00:02:43 Quit Domonoky (Read error: 104 (Connection reset by peer))
00:06:12__peer__hmm rondom?
00:06:18Rondom__peer__: yes
00:06:37__peer__im in rondom
00:06:44RondomI know
00:07:28 Quit petur ("here today, gone tomorrow")
00:09:21 Join ts-x [0] (
00:09:49 Quit _DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
00:13:06 Part ts-x
00:14:01 Join ts-x [0] (
00:18:13 Join RotAtoR [0] (
00:21:19 Join Mongey [0] (
00:22:07 Join ashridah [0] (
00:28:05 Join DJDD__ [0] (
00:28:36 Part ep0ch| ("Kopete 0.11 :")
00:33:04 Quit Kohlrabi ("Fast alle Menschen sind Regenw├╝rmer")
00:37:07 Quit aegray (Read error: 104 (Connection reset by peer))
00:37:21 Join aegray [0] (
00:41:30 Quit __peer__ (Read error: 104 (Connection reset by peer))
00:44:12JdGordonhow do i open up the screen kleyboard with default text?
00:44:17 Join toujours [0] (
00:45:12 Part toujours
00:46:50 Quit Mongey ()
00:47:13 Quit thegeek (Read error: 104 (Connection reset by peer))
00:47:59 Quit tvelocity ("Leaving")
00:50:51 Quit DrumRBoy320|Away (Read error: 110 (Connection timed out))
00:57:18 Quit matsl (Remote closed the connection)
01:00:26***Saving seen data "./dancer.seen"
01:05:06 Quit muesli__ (Read error: 110 (Connection timed out))
01:09:23 Join thegeek [0] (
01:14:58 Part textchimp
01:16:26 Join tvelocity [0] (
01:23:20thegeekoh no!
01:23:23thegeekwhat have you done?
01:24:10Moosvery quiet this night here
01:24:57 Join San [0] (n=test@
01:27:19jf2Two questions for someone: 1) This is not exactly the right fourm for this but I bet someone knows: Due to a few factors I went away from home without my iRiver IHP 120, but I am curious, does anyone know if it will play incripted Windows Media (ie. Napster files?) 2) Rockbox will not work with such files, correct?
01:27:50ashridahthe H120 won't play DRM'ed windows media
01:28:22ashridahand there's no support for wma yet (no-one's working on porting the ffmpeg codec for it atm afaik either), and will probably never be support for protected wma.
01:28:27JdGordondid any1 try the timetable plugin i posted last night?
01:29:25jf2i got it. thanks ash. two answers I didn't want! :) heh. but I can at least add rockbox to make on the fly play lists
01:29:37MoosJdGordon: sorry h140 here
01:31:13jf2i can make multiple on the fly playlists when I install, correct? or is it just one that has to be overwritten when I want to create a new one?
01:31:45JdGordonits bassically finished.. so ill fix it up for the h140 later today
01:32:01ashridahjf2: playlists can be saved to .m3u files
01:32:07ashridahso you can make as many as you like
01:32:35MoosJdGordon: isn't use RTC?
01:32:39jf2good. that is the most appealing feature of Rockbox for me. awesome.
01:33:17JdGordonMoos: not really... it only uses it to jump to the current day/time. but that can be done with the arrows easy enough
01:34:10gromit`would someone have the jtag specification v2.0x ?
01:36:55 Join Paul_The_Nerd [0] (
01:41:43 Quit jf2 ("CGI:IRC")
01:43:24linuxstbPaul_The_Nerd: Do you have time to testing something on your Nano?
01:49:00 Join TCK [0] (
01:49:53 Join Bluechip [0] (
01:49:58 Part Bluechip
01:51:54 Quit Moos ("Glory to Rockbox")
02:02:53Paul_The_Nerdlinuxstb: Certainly. Sorry, wasn't really here for a bit
02:03:19Paul_The_NerdThis client still doesn't notify my well when my name's mentioned.
02:04:32linuxstbnp. I've got the bootloader loading the retailos from a file on the fat32 partition. It works on my Color, but not on my Video.
02:04:48linuxstbSo I was wondering how it did on the Nano.
02:05:09linuxstbPatch here:
02:05:37 Join muesli__ [0] (n=muesli_t@
02:05:50linuxstbHere is the shell script I use to create and install the firmware partition (you will obviously have to adapt it):
02:06:11linuxstbAnd finally, copy "apple_os.bin" to your Nano's FAT32 partition.
02:07:00linuxstbThis probably isn't much of an advantage to you (apart from potentially simplifying installation), but it cuts about 4-5 seconds off my bootup time on my Color.
02:07:40linuxstbBut I need to go to bed now - let me know how you get on (if you try it). I'll check the logs in the morning.
02:08:31Paul_The_NerdI'll try it in a few minutes then.
02:10:10lostlogic*scowl* pcmbuf_request_buffer can give the same buffer twice in a row, unless I misread badly.
02:16:24Paul_The_Nerdlinuxstb: If you're still here, is this all the differences? It breaks the goto retailos statement, if I applied it properly (it got rejected, so I had to do it manually)
02:17:07 Join J32US [0] (
02:17:49JdGordonoh no!
02:17:49 Quit TCK (Read error: 110 (Connection timed out))
02:17:59 Quit Rondom (Read error: 110 (Connection timed out))
02:18:11J32USwho here actually develops the rockbox?
02:19:38lostlogiclol, there are several devs in here
02:19:59 Join Bluechip_ [0] (
02:21:10 Part Bluechip_
02:21:10Paul_The_NerdWhy, what's up?
02:22:06Paul_The_NerdCould someone on a linux box do me a favor?
02:22:18lostlogicPaul_The_Nerd: probably
02:22:57Paul_The_Nerddd if=/dev/zero of=dummy.bin bs=1024 count=1 <−− Just run this, and give me the file. I know I could probably do very nearly equivalent things in windows, but this exact step has the desired result, and it's easier if I just duplicate it exactly.
02:23:48Paul_The_NerdI want to make sure that if it completely screws up the bootloader, it's something other than some quick step I took to avoid rebooting into linux.
02:24:41J32USis anyone here working on the gigabeat project?
02:25:01lostlogicj32us: the only one really developing for it is markun if he's around
02:27:43J32USreally? hmm
02:28:49Paul_The_Nerdlinuxstb: Panic: stkov if you forget to copy the file over. If you do actually not do that completely stupid act, it gets to the Executing... bit, and then seems to stay there. At least, it's been about a minute so far and RetailOS hasn't loaded.
02:30:43 Quit J32US ()
02:34:49 Quit einhirn (Read error: 104 (Connection reset by peer))
02:38:37 Quit Paul_The_Nerd ("Leaving.")
02:39:09 Join einhirn [0] (
02:39:10 Join Paul_The_Nerd [0] (
02:40:11Paul_The_NerdOh, and thanks lostlogic
02:40:38lostlogichappened to have a term on the webserver open ;)
02:45:02 Quit muesli__ (Read error: 110 (Connection timed out))
02:47:28 Quit einhirn (Read error: 104 (Connection reset by peer))
02:48:54Paul_The_Nerdlostlogic: Well, the bootloader he made works for rockbox and linux, so he didn't break anything but the retailOS loader. That's good so far.
02:50:23 Join einhirn [0] (
02:53:55 Quit RotAtoR (
02:53:55 Quit kkurbjun (
02:53:55 Quit merbanan (
02:53:55 Quit perplexity (
02:53:55 Quit Lynx_awy (
02:53:55 Quit midkay (
02:53:55 Quit elinenbe (
02:53:55 Quit kernelsensei (
02:53:55 Quit goa (
02:53:55 Quit BHSPitMonkey (
02:53:55 Quit Rob- (
02:53:55 Quit radsaq (
02:53:56 Quit Rick (
02:53:56 Quit crash_ (
02:53:57 Quit einhirn (Read error: 104 (Connection reset by peer))
02:54:01NJoinRotAtoR [0] (
02:54:05 Join radsaq [0] (n=radsaq@
02:54:06 Join crash_ [0] (n=crash@
02:54:07***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
02:54:14 Join merbanan [0] (i=banan@
02:54:14NJoinmidkay [0] (n=midkay@
02:54:16***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
02:54:25NJoinelinenbe [0] (
02:54:27 Join Rob- [0] (n=robbie@
02:54:28***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
02:54:28NJoinLynx_awy [0] (
02:54:34 Join perplexity [0] (i=heh14756@
02:54:35***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
02:54:36 Join kkurbjun [0] (n=Jim@
02:54:37***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
02:54:37NJoinkernelsensei [0] (n=boris@unaffiliated/kernelsensei)
02:54:55NJoingoa [0] (
02:55:58NJoinRick [0] (
03:00:09 Join hshah [0] (
03:00:14hshahanyone know how to convert 02 to 2 in C?
03:00:29***Saving seen data "./dancer.seen"
03:00:40hshahie ensure all ints are the smallest possible number of digits?
03:03:04ashridah... convert?
03:03:15ashridahwhy would any conversion be necessary? octal '02' is still '2'
03:03:31hshahit just needs to be done
03:03:36hshahcoz something is outputting 02
03:03:38hshahand 01
03:04:13ashridahfrom the output of printf, you mean?
03:04:44ashridahwhat formatting code are you using?
03:05:14hshahprintf("Initial Time: %u:%u\n", ansih, ansim);
03:05:14hshahprintf("Increment: %u:%u\n", ansch, anscm);
03:05:14DBUGEnqueued KICK hshah
03:05:14hshahprintf("Final Time: %u:%u\n", resh, resm);
03:06:58 Quit Strath (Read error: 104 (Connection reset by peer))
03:07:30ashridahwhy are you using $%u?
03:07:44 Join Strath [0] (
03:07:47hshahits for a friend of mine...
03:08:01hshahcan you explain please, because I don;t know C and its not my work...
03:08:43ashridah%u is the formatting code for those variables (which, i assume, are int's)
03:11:24ashridahso, the question is, why is he using %u? are those actually unsigned integers?
03:14:34 Join BHSPitMonkey [0] (
03:14:35***Server message 505: 'logbot_ :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
03:18:10 Quit tvelocity ("Leaving")
03:18:53hshahim not sure - what should he be using?
03:20:15ashridahwell, that depends. if they're normal int's, and not unsigned ints, %d would be the appropriate choice. but using %u doesn't explain why they're printing as '02'.
03:20:17ashridahbecause %u won't be doing that
03:29:39 Join Membrillo [0] (
03:44:57 Quit Rob2222 (Read error: 104 (Connection reset by peer))
03:49:04 Join JoeBorn [0] (
03:57:00 Join TCK [0] (
04:00:41 Join Jungti1234 [0] (n=jungti12@
04:04:13 Quit Nibbler (Connection timed out)
04:14:57 Join jborn_ [0] (
04:24:01 Join Nibbler [0] (
04:27:58 Quit midkay ("Leaving")
04:28:15 Quit Membrillo ()
04:33:02 Quit JoeBorn (Read error: 110 (Connection timed out))
04:35:15 Join dpassen1 [0] (
04:36:05 Quit TCK (Read error: 110 (Connection timed out))
04:43:33 Quit Jungti1234 ()
04:53:23 Join steveb_ [0] (
04:53:55 Nick steveb_ is now known as Pantalaimon (
05:00:31***Saving seen data "./dancer.seen"
05:01:59 Join TCK [0] (
05:04:04 Quit ts-x ("CGI:IRC")
05:23:15 Join ansiVirus [0] (
05:23:53 Join Pingoomax [0] (n=pingooma@d221-109.RES.UMontreal.CA)
05:24:17 Quit TCK (Read error: 110 (Connection timed out))
05:25:18 Join manhattan [0] (
05:26:27PingoomaxI just discover the project
05:27:46 Quit mikearthur (Remote closed the connection)
05:32:50PingoomaxI would like to know if something is under construction for archos xs202.
05:32:57 Quit ze (Read error: 110 (Connection timed out))
05:41:46Paul_The_NerdI've not heard anyone even mention that before.
05:42:02Paul_The_NerdBut the way the project is structured, anyone could conceivably begin an effort toward it.
05:42:04 Nick jborn_ is now known as JoeBorn (
05:50:23 Join jborn_ [0] (
05:53:16 Quit San (Read error: 110 (Connection timed out))
05:56:50 Join Membrillo [0] (
06:00:32 Join Jungti1234 [0] (n=jungti12@
06:07:53 Quit JoeBorn (Read error: 110 (Connection timed out))
06:22:53 Part Paul_The_Nerd
06:30:46 Join ze [0] (
06:31:35 Quit RotAtoR ("zzzzzz")
06:34:40 Join LinusN [0] (
06:40:10 Join Lost-ash [0] (
06:40:22 Quit Membrillo ()
06:41:52 Quit ashridah (Nick collision from services.)
06:41:59 Nick Lost-ash is now known as ashridah (
06:43:08 Join eEye [0] (
06:43:48 Join jborn__ [0] (
06:47:50 Quit saratoga ("CGI:IRC (Ping timeout)")
07:00:35***Saving seen data "./dancer.seen"
07:01:59 Quit jborn_ (Read error: 110 (Connection timed out))
07:05:12 Join TCK [0] (
07:06:08 Quit Jungti1234 ()
07:09:46 Quit TCK (Read error: 104 (Connection reset by peer))
07:11:09 Join EvilDude [0] (i=EvilDude@
07:15:42 Join Jungti1234 [0] (n=jungti12@
07:16:05Jungti1234Gigabeat S series supports Korean.... :)
07:24:21JdGordonwhat retard puts .bmp on the web when the pic is that big???
07:24:37JdGordonplayer looks damn fine tho :D
07:26:30eEyelooks like they are trying to hard to imitate the look of the ipod
07:26:51eEyeinstead of trying to imitate and recreate that kind of look i think the companies should try to make there own style
07:27:16eEyehonestly i still think the x5 looks a lot more aesthetically better then the majority of the players out there
07:31:55DreamTactix291i do too
07:32:00DreamTactix291except maybe the H1xx :)
07:32:12 Quit jborn__ (Read error: 110 (Connection timed out))
07:33:14 Join San [0] (
07:33:19Bgermorning :)
07:34:11CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
07:34:11*Bger doesn't like temperatures like ~ 20 deg Cel. at all
07:34:33JdGordon?? too cold or hot?
07:34:48Bgertoo cold :)
07:34:53Bgerfor me, at least
07:35:03JdGordon:) i prefer the heat also
07:35:14JdGordoncept when it gets up around 40C
07:35:18Bgerit depends on how hot is ... :)
07:35:34 Quit lostlogic (Remote closed the connection)
07:38:43 Join Membrillo [0] (
07:47:40 Join Paul_The_Ner1 [0] (
07:48:00Jungti1234I have question
07:48:51Jungti1234Can H300 int'l model hardware support DRM?
07:49:27ashridahisn't there a chip missing on the international players?
07:49:35Bgerashridah : no
07:49:39Bgerit's the reverse
07:49:54 Quit EvilDude ()
07:49:59Bgerthere are some components missing in the *US* version, but i wouldn't call them chips
07:50:07Bgerthey are some capaciters, resistors ...
07:50:10Jungti1234Iriver spoke that it's impossible.
07:50:27BgerJungti1234 this is a lie
07:50:36Bgerthe problem is in the DRM code
07:50:41Bgernot in the hardware
07:51:01Jungti1234Bad men
07:51:35ashridahaah, yes. the missing key from the unit is the only problem then.
07:52:02Jungti1234Circumstantiate more.
07:53:45Jungti1234Explain more please.
07:54:06ashridahi'm more or less assuming DRM is a variant of a public/private key system.
07:54:49Bgerthere are 2 rumours regarding the code
07:54:57BgerDRM code
07:56:08ashridahJungti1234: you have to understand, DRM is a closed system, and in many countries, it's illegal to try to reverse engineer it. there's not a fantastic amount of real specs about, unless you feel like paying for the privilege to read them
07:56:18Bger1) the DRM code is located in a block of the flash (and the structure of the .hex firmware file allows flashing it on blocks, so some blocks can be skipped)
07:56:30Bger2) the DRM code is located in the EEPROM
07:58:26BgerLinusN is the man who can reveal the truth
07:59:09 Join lostlogic [0] (
08:02:28 Quit elinenbe (Read error: 104 (Connection reset by peer))
08:02:45 Join elinenbe [0] (
08:02:58 Join EvilDude [0] (i=EvilDude@
08:05:53Jungti1234They spoke always.
08:06:06Jungti1234'DRM is impossible by hardware.'
08:10:10 Quit Paprica (Read error: 110 (Connection timed out))
08:10:26Bgerthey said the radio recording is impossible on H100, too :)
08:11:22Jungti1234Liar LiarLiarLiarLiar
08:12:57dwihnoThat reminds me. My IFP-799 has been away for repairs for almost two months now.
08:13:16dwihnoGotta check up on that today.
08:13:28Jungti1234two months?!
08:14:00Jungti1234I like paprica
08:17:12dwihnoJungti1234: Yeah, I returned it in november... I wonder what takes so much time...
08:17:32JdGordonslack... but it is over the "holiday period"..
08:18:26Jungti1234Their omission happens often.
08:20:01 Join midkay [0] (n=midkay@
08:21:46SlasheriHmm, something should be done to the wps loader.. it increases boot time on my player 3-4s
08:23:31 Join HaKuNa [0] (
08:24:52 Quit midkay (Client Quit)
08:25:02 Join midkay [0] (n=midkay@
08:25:10 Quit midkay (Read error: 104 (Connection reset by peer))
08:25:12 Join midkay [0] (n=midkay@
08:26:43 Join ender` [0] (i=ychat@
08:29:31amiconnSlasheri: yes
08:29:48*amiconn whispers "combined bitmap loading"
08:30:04amiconnI know there is a patch for this...
08:30:12Slasherii think the bitmaps should be even compiled to native format before loading them..
08:30:27Slasherii did some timing and noticed the bmp loader is slow
08:31:19midkaysorry, i'm a bit behind the times.. BMP loader, as in, rockbox loads its bitmaps from disk? :)
08:31:21Jungti1234funny picture.
08:31:36dwihnoJungti1234: Erhm. What does that image tell? :)
08:31:44dwihno"You are bad" "Yes I am, mohahah!" ?
08:32:01Slasheriamiconn: and i just put the audio initialization on background, so now the wps loader is almost the only thing which increases the boot time
08:32:34Jungti1234dwihno: It says about H300 LCD problem.
08:33:11Jungti1234"Change LCD!" "No!"
08:33:31dwihnoah, okay
08:34:13Jungti1234Soon New Year's Day comes...
08:34:31Jungti1234Home delivery service stops.
08:35:02amiconnSlasheri: I don't think the bitmaps should be in native format. Would be far too cumbersome for users to create wps'es
08:35:31Slasheriamiconn: ah, true.. then the wps loader should probably automatically convert them to native format..
08:35:32LinusNamiconn: i'm thinking about transparent bitmaps, but i'm not sure how i want to implement it
08:35:33amiconnThe bmp loader could receive some optimisation work...
08:35:49amiconnLinusN: What for? Background image only?
08:35:51Slasheriso just keep the current separate bmp method and convert them runtime at first load
08:36:00LinusNeither a single transparent color, or some blending
08:36:30amiconnLinusN: That's some of the lowest prio in gfx work for me. So many areas which are far from finished.
08:36:31Slasheriand that way create a cache file for the wps which could be loaded really fast at boot time..
08:36:43amiconnReal alpha will probably be really slow.
08:36:54amiconn*Really* slow I mean, by factors
08:37:27amiconnSlasheri the cacheri ;)
08:37:46LinusNSlasheri: speaking of cache, how's the tag cache nowadays?
08:37:58amiconnNah, I wouldn't want rockbox to cache too much in files. There's already a lot of clutter in /.rockbox
08:38:19SlasheriLinusN: currently i haven't had much time for that in the past few weeks.. but maybe at next weekend :)
08:38:25*JdGordon votes for more chached files to speed up load :D
08:38:33ashridahwouldn't you be able to accelerate the alpha mapping using the emac unit?
08:38:40Slasheriamiconn: hehe, probably i could take a look on that too.. shouldn't be hard to implement
08:38:41LinusNcombined wps bitmaps would be a blessing
08:39:38amiconnLinusN: There should be a patch by _Firefly_ in the tracker for this
08:39:39LinusNamiconn: have you had a look at my color bmp loader patch?
08:39:47amiconnDidn't have time to look at it...
08:40:21LinusNit appears to work fine for the misticriver crowd
08:40:38amiconnCurrently my local tree is rather cluttered, as the language cleanup is still unfinished. Then there is memmove(), memset16() for optimising 16bit gfx...
08:40:48Bgerhaha LinusN the betatester comunity
08:41:29amiconnLinusN: ALready checked out greyscale cube on archos? :-)
08:41:30LinusNamiconn: still considering changing the framebuffer alignment?
08:41:40amiconnAh, no
08:41:47LinusNamiconn: no, i haven't tried the cube
08:41:55amiconnNot the framebuffer alignment, but the mono bitmap format
08:42:01LinusNah yes
08:42:26amiconnBtw, changing the framebuffer would be another option, as the LCD allows switching the update direction
08:42:47amiconnThen the framebuffer would go in columns instead of rows
08:43:10amiconnDunno how hard it would be to get this working on iPod though...
08:46:58 Quit Membrillo ()
08:56:43Slasherihmm, i just found ways to optimize the bmp loader.. trying if it has any effect
08:59:33 Join Kohlrabi [0] (
09:00:36***Saving seen data "./dancer.seen"
09:02:11 Join Maxime [0] (
09:02:11 Quit Maxime` (Read error: 104 (Connection reset by peer))
09:03:13LinusNSlasheri: it might be possible to omit the extra seek()
09:03:57LinusNespecially in my color loader, since it loads the palette
09:05:02SlasheriHmm, true. trying that also
09:06:22LinusNwhat was your idea?
09:06:45amiconnLinusN: Btw, if you want to get an impression of how much speedup for H3x0 lcd we need:
09:07:15amiconnTake H1x0 and H3x0, start cube on both and switch to high speed mode. Then compare... :-/
09:07:17Slasherilittle cpu optimization inside the loop. But as i now tried, the open(...) takes the longest time.. so we need only one file to load
09:08:59LinusNamiconn: ouoch
09:15:19amiconnLinusN: Btw, the solid cube on archos now has about the same speed as on H3x0, even though it uses the complex mechanisms of the grayscale library
09:15:39amiconnOkay, archos has far less pixels, but then it also has a way smaller CPU
09:16:08amiconn(and it's a bit faster on Ondio than on the recorders :-P)
09:17:03 Join B4gder [0] (
09:17:15preglowbtw, wps load is waaay faster on ipod than h1x0
09:17:17preglowany idea why?
09:17:32preglowif i understood linuxstb correctly, this also extends to the disk based ipods
09:18:57amiconnpreglow: On nano I suppose?
09:19:08amiconnThis is flash, so no disk seeks involved...
09:19:51preglowread above, please
09:22:09 Nick Lynx_awy is now known as Lynx_ (
09:22:12 Quit Kohlrabi (Read error: 104 (Connection reset by peer))
09:22:23Slasheriopen takes so long, because it needs to scan files on current directory to find the correct subdirectory to open and same again..
09:22:51Slasheriwith dircache that would be much faster, but currently don't have the cache available at boot time..
09:24:18 Join stamppot [0] (
09:24:51Slasheriand the needed information to open the file seems to be the startcluster which dircache stores also
09:30:15 Quit EvilDude ()
09:33:33 Join einhirn [0] (
09:35:53 Join Zagor [0] (
09:38:01JdGordonJungti1234: ?
09:42:05linuxstbpreglow, amiconn: Yes, none of the themes take more than about a second to load on my 60GB ipod.
09:49:04JdGordonlinuxstb: iv added the tt line to the viewers.config file but .tt files still arent shown when i have the supported fikles shown..?
09:49:43B4gdernope, that's another list
09:50:27JdGordonwas that to me?
09:50:40JdGordonhmm... it works on the sim tho?
09:53:11B4gderI kind of agree that it _should_ work like that
09:54:03B4gderbut the list of extensions the file browser recognizes is found in an array in apps/tree.c
09:54:27preglowlinuxstb: yes, which sounds strange
09:54:29B4gderhm, I'll see if I might be wrong
09:54:37preglowsince open shouldn't take any less time on an ipod than on a h1x0
09:54:39preglowbut it does!
09:55:00JdGordonits just, its wierd coz it works fine in the sim (ive just checked that supported is set), but not on the player
09:55:28B4gderJdGordon: that is certainly weird
09:56:13 Quit ansiVirus (Read error: 110 (Connection timed out))
09:56:34*B4gder remembers when apps/tree.c actually looked fine... :-)
09:57:21JdGordonis there another file that it does check? coz it shows my .ss (sudoku files) fine...?
09:57:30JdGordonor.. did untill i deleted thhem
09:57:38B4gderok, then it does work
10:00:29JdGordonLinusN: any idea whats wrong?
10:00:39LinusNnope, no error splash?
10:01:05LinusNdid you write the .tt viewer?
10:01:16JdGordonand its where it should be
10:01:38LinusNis .tt a standard file format?
10:04:06 Join muesli__ [0] (n=muesli_t@
10:10:51JdGordon.. odd.... i just deleted the .rockbox folder and put it back (also the .rivier file) and its working again..
10:12:33LinusNworks for me too, just tested
10:12:58LinusNJdGordon: you did restart rockbox after changing the viewers.config file?
10:16:32JdGordongrr... now its got bootloader usb mode but windows isnt mounting it..?
10:18:04 Join amiconn_ [0] (
10:31:15 Quit HaKuNa ("CGI:IRC (EOF)")
10:36:35 Quit amiconn (Read error: 110 (Connection timed out))
10:36:36 Nick amiconn_ is now known as amiconn (
10:50:20 Quit Zagor ("Client exiting")
10:54:52 Join DangerousDan [0] (
11:00:39***Saving seen data "./dancer.seen"
11:03:33 Join Rob2222 [0] (
11:04:00Bgerasking for a second time, but ... does anyone knows any open source program/project for taking the phone call info from a Panasonic KX PBX ...
11:04:45Bgeri don't know whether the PBX term is correct
11:08:05Lynx_does asterisk do that kind of stuff maybe?
11:08:35BgerLynx_ i really don't know ...
11:08:48Bgerbut asterisk is a PBX itself...
11:09:29Lynx_yes, i know
11:09:52Bgerah, u're in that stuff or ?
11:10:02Bgeri would appreciate any hints ...
11:10:04Lynx_no...just a wild guess
11:13:40linuxstbBger: You could try #asterisk (here on freenode). I don't think asterisk will do the job, but the people there will probably have a better idea than us about PBXs.
11:14:03Bgerah, linuxstb, very good idea :)))))))0
11:14:25JdGordonim bored...
11:14:27Bgerbtw, is the term PBX correct ? :)
11:14:40 Join Zagor [0] (
11:15:39 Join gtkspert [0] (
11:15:55linuxstbJdGordon: Port some more plugins to the H300's screen. There are lots there. Or go buy yourself an ipod and help us with the port :)
11:16:27JdGordonanything else?
11:16:31linuxstbBger: Asterisk describes itself as a PBX.
11:16:43B4gderJdGordon: one coffee for me please
11:16:51JdGordoncoming right up :D
11:17:09amiconnpreglow, linuxstb: I'll check theme load speed on archos later. If it's fast there as well, I suspect a performance issue in the coldfire ata driver
11:17:29linuxstbamiconn: I'll try do to a side-by-side test of my ipod and h140 later as well.
11:17:29amiconnLinusN: Do you know whether our ata timing is optimal?
11:17:30Bgerlinuxstb i'm wondering whether the panasonic thing i'm talking for is a PBX
11:17:33*JdGordon trips on the way over to B4gder's table and pours boiling coffee all over him.. scolding him and sends him to hospital with 3rd degree burns
11:17:51LinusNamiconn: which platform?
11:18:01amiconnH1x0 and H3x0
11:18:09LinusNi'm not sure
11:18:40*amiconn whistles 'logic analyser'
11:19:38LinusNone of those days...
11:24:29amiconnPerhaps reading the datasheet and double-checking the ata reg values is enough...
11:26:03LinusNi remember tweaking those values when i had the cpu overheating problems in the beginning of the h100 porting
11:26:20LinusNthey might be quite conservative
11:26:50 Quit perplexity (Read error: 113 (No route to host))
11:28:19linuxstbI've just tested loading "boxes" side by side on the h140 and my 60GB ipod. With the disk spinning, it takes 1-2 seconds on the ipod, and about 5 seconds on the h140.
11:34:50 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
11:35:12JdGordonany1 got any tips on how to un-blurrify a old crt monitor?
11:36:57LinusNdidn't even know it was possible
11:37:02ashridahJdGordon: sometimes you can play with a focus knob on the back or inside
11:37:22ashridahbut it's probably going blurry for a reason, which means that the magnets inside are dying, typically.
11:37:50JdGordon:'( havta wait 1 more week then till 3 monitors :d
11:41:25JdGordonfucking aye :D mario on my h300
11:41:33JdGordonback when the battery runs out :D
11:42:49 Join perplexity [0] (i=heh21147@
11:45:55 Quit hshah ("Leaving")
11:46:25 Join Paprica [0] (
11:48:03Bgerlinuxstb regarding my asking .. asterisk won't help me
11:50:26 Join Pieter_ [0] (
11:51:07Pieter_hello! is it know that the tagdatabase can crash the ondio? can't find it in the bugreports
11:51:46LinusNPieter_: yes, it can be a little crash prone
11:51:49Pieter_not that the database is very useful on an ondio perhaps, but it doesn't seem nice that it just freezes or gives some strange error :)
11:52:09LinusNwe are working on a new implementation
11:52:36 Nick paugh is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
11:53:10Pieter_ok.. i won't report this as a bug then.. a little crash prone would be not working at all here
11:53:11markunhi Pieter_
11:53:16Pieter_heee markun :)
11:55:37Pieter_anyway, if it's any use, the error it gives is I09: CPUAdrEr at FFFFFFFF
11:58:47amiconnThat's strange. When I last tried tagdb on Ondio it worked...
11:59:55 Join Febs [0] (
12:02:01 Join Membrillo [0] (
12:10:12 Join Moos [0] (
12:11:29preglowanyone got any ideas on how to implement cue file support? :>
12:13:24LinusNpreglow: one idea is to treat a cue file like a playlist
12:14:03preglowyup, i'd definitely do that
12:14:31LinusNthe question is if we want to treat it one long track or several small
12:14:32preglowso just tack on support for file positions to playlists? heh
12:14:39preglowseveral small, i'd say
12:14:46LinusNi'm not so sure
12:15:07LinusNtreating it as a set of tracks may give us other problems
12:15:17LinusNlike how to handle shuffle and resume
12:15:43preglowbut yeah
12:15:52linuxstbWe should also think about "chained" ogg files and MP4 chapters - they are similar to cue files.
12:15:55preglowhow to handle skipping and such if it appears as one big file?
12:16:01LinusNlast time i gave this some thought, i kind of settled for the single-track approach
12:16:26LinusNand adding support for multiple tags and skip-points
12:17:00LinusNwhich would mean that you can't shuffle the tracks in a queue file
12:17:13preglowlinuxstb: how do you discover if an ogg file is chained? you need to seek, i presume?
12:18:06linuxstbYes - the metadata parser should identify that. IIRC each ogg file has a unique serial number - so pages at the end of the file will have a different serial number to pages at the start.
12:18:23preglowogg really isn't the most elegant container format in existence
12:18:28linuxstbIt's a dog.
12:19:56linuxstbThe only good container format I've found is mp4. It simply stores the compressed and uncompressed size of every frame in the header. Some may think it's a waste of space, but it makes life very easy.
12:20:41preglowyes, i agree
12:20:48linuxstbThe uncompressed size is normally constant (apart from the last frame), and this is encoded as simply X frames of length N, plus 1 frame of length M. So that only takes a few bytes.
12:20:52preglowit's negligible
12:23:07linuxstbBut going back to cuefiles, when I looked at implementing it in Rockbox a few years ago, I think I just had a boolean "has_cuefule" attached to each track. If this boolean was true, then the progress bar displayed vertical lines to indicate the track boundaries, and next/previous skipped to the next/previous track within the file.
12:23:54preglowthe seek bar might end up having a very low resolution
12:24:00preglowi'm not convinced that's very helpful
12:24:27 Join Maxime` [0] (
12:24:27 Quit elinenbe (Read error: 104 (Connection reset by peer))
12:24:28 Quit Maxime (Read error: 104 (Connection reset by peer))
12:24:48 Join elinenbe [0] (
12:25:42preglowmy prefered implementation as a user would indeed have every title in a cue file appear as a separate track in all ways
12:25:46preglowbut of course, that might be cumbersome
12:26:24LinusNthe playlist code would be very tricky
12:26:28linuxstbI would still like to know that I am in one long file though, and how far through that file I am.
12:27:14LinusNthe major drawback of the singlefile approach would be that it isn't possible to shuffle, or insert/queue tracks
12:27:19linuxstbI think I would ideally like two progress bars - one for the current track, and one for the cuefile.
12:29:14 Join Maxime [0] (
12:29:14preglowbut yeah, i wonder what to do about the seeking problem
12:29:14 Quit Maxime` (Read error: 104 (Connection reset by peer))
12:29:27preglowwe'll use either fast and very inaccurate seeking, or slow and accurate
12:29:33preglowfoobar uses the last, and is slow even on pcs
12:29:50Rob2222Well, 12:30 MP3 runtime now... at my RB test my battery was empty after 12:30. The Iriver FW switched from 3 to 2 battery indicators.
12:30:04linuxstbI think cuefiles demand accurate seeking. But Rockbox should only ask the codecs to seek - it is up to the codecs to do the best job they can.
12:30:15Rob2222Looks like Iriver FW is still a little better.
12:30:37linuxstbIs this a H300?
12:30:58Rob2222Will see when the player shuts off.
12:31:24linuxstbCan you put your test results in the IriverRuntime Wiki page?
12:31:50Rob2222Yes, i can. With battery log files?
12:32:18linuxstbpreglow: So are you planning to implement cue support, or are you just thinking about how it could work?
12:32:48linuxstbRob2222: Why not. You could add an extra column in the test table for log files.
12:32:50preglowjust thinking, since i just was playing a cue based file on my player
12:32:57preglowand found it incredibly boring, heh
12:34:03 Join Rob2222_ [0] (
12:36:29 Join tvelocity [0] (
12:44:15XavierGrArghhhh! I am gonna kill myself along with advancedmp3players and my bank!
12:45:19BgerXavierGr ah
12:45:23XavierGrI ordered an H340 from there. Price was 274pounds I send them 298 due to bank expenses :x and now they tell me they need another 12 because I didn't select to share the expenses.
12:45:49Bgerbad :(
12:46:03XavierGrWhile I could have paid them in euros at no charge or visa electron but NO their f*cking store won't accept that.
12:46:32XavierGrNow I have paid 40 euros more that the price tag and still no H340
12:46:44Bgeruh :(
12:47:02preglowi never ever pay by bank
12:47:04preglowit's always trouble
12:47:37XavierGrpreglow if I pay online in euros there is no trouble if I go to the clerk they start charging like freaks.
12:48:36XavierGrBger you are here? Are you working on the battery_bench?
12:49:27BgerXavierGr yes and no :)
12:49:38Bgerbut no will go to yes
12:49:40Bger :)
12:49:54XavierGrDid you checked latest version in the tracker? It has your latest changes. Also fixed the yesterday bug in the A/D column now it works perfect, but if you still want to change the disk routine...
12:50:38Bgerno, i didn't checked it, but checked my e-mail :)
12:51:29XavierGrthe mail was from the day before yesterday. ;)
12:51:31 Quit Rob2222 (Read error: 110 (Connection timed out))
12:52:17Bgeri mean that at least i checked my mail
12:54:38 Quit ashridah (Read error: 110 (Connection timed out))
12:56:21 Join ashridah [0] (
12:56:24 Quit ashridah (Client Quit)
13:00:40***Saving seen data "./dancer.seen"
13:10:44 Join ep0ch [0] (n=ep0ch@
13:17:43amiconnXavierGr: You've updated the battery status plugin?
13:17:58Bgeramiconn it's not complete yet
13:18:32 Join Kohlrabi [0] (
13:19:11 Join _FireFly_ [0] (
13:25:58 Quit JdGordon (Read error: 104 (Connection reset by peer))
13:30:05 Quit uwe_ (Remote closed the connection)
13:30:35*amiconn would test the battery status plugin on archos...
13:30:50amiconn*would like to test the ...
13:31:10Bgeramiconn let's finish it first :)
13:31:14XavierGramiconn: you can do it with this version too.
13:31:32BgerXavierGr ?
13:31:56XavierGrwell he can see if at least works. Of course it will be changed a little.
13:32:15XavierGrI don't know if it compiles on archos targets yet.
13:34:51amiconnBger: What's missing?
13:35:54Bgeramiconn i'd like to test a code that at least i think it isn't buggy
13:40:48 Quit gtkspert (Read error: 110 (Connection timed out))
13:42:59 Join Matze [0] (
13:44:03 Join tucoz [0] (
13:44:50tucozHi, sorry for the ot, but this goes out to the nordic speaking rockboxers.
13:45:10tucozIt's basically about the first bug found in a computer. :)
13:45:18tucozwhich was in fact a bug.
13:46:49preglowyeah, a rather clear ethymology for that one
13:46:50Bgeri see dead bugs ... in my code
13:47:18tucozdried moths everywhere :)
13:48:34muesli__yummy :)
13:48:59muesli__fry them..even better ;)
13:49:38Bgerand eat with cream :)
13:49:56muesli__. +s
13:51:51 Quit _FireFly_ (Read error: 110 (Connection timed out))
13:53:03tucoztalking about bugs, and firefly quit :D
13:53:23muesli__mayb he feels guilty ;)
13:54:01 Join ashridah [0] (
13:54:14Bgerhahaha :D
13:54:19 Quit Rob2222_ (Read error: 104 (Connection reset by peer))
13:54:50Jungti1234eat bugs?
13:54:55 Join Rob2222 [0] (
13:55:06Bgerhe shouldn't be afraid of frying ...
13:56:58tucozsorry for the confusion. It was not the first bug, but the first _actual_ bug found. You could read the english version <-here, a bit down the page.
13:58:02 Join uwe_ [0] (
13:59:08Bgerstop talking about fryed flys, guys! :p
13:59:36tucozreminds me, should have some lunch now. bye bye
13:59:43 Part tucoz ("Leaving")
14:01:50 Quit San (Read error: 110 (Connection timed out))
14:02:44Membrillodoes anyone have webspace with lots of bandwidth that they are willing to use? i need a mirror of the h300 optimized builds
14:03:10Bgerask B4gder
14:03:16B4gderI doubt you need lots of bandwidth for it
14:03:31B4gderor rather, it depends on your definition of lots
14:04:20B4gdersorry, but I actually don't want to be affiliated with the optimized build
14:05:04linuxstbMembrillo: Why don't you make the fonts a separate download? That will cut down your bandwidth.
14:05:44linuxstbOr help with integrating the major features of that build into CVS.
14:06:13B4gderthat's what I think is a better way to spend the time
14:06:23Bgerhehe :)
14:06:34Bgerthese remotes ...
14:06:39Membrillolol. well i know no C
14:07:00Bgernobody knows when he starts to learn :))
14:07:14Bger's style answer)
14:07:21Membrilloits not my bandwidth im worried about. the optimized builds URL is strange and i cant use it, so i need a seperate mirror
14:07:41linuxstbWhat's strange about the URL?
14:07:46Membrilloi dont know
14:07:47B4gderMembrillo: so you can mirror it
14:08:01Membrilloi cant mirror it. I dont have web space
14:08:02B4gderor use a better way to download it
14:08:16*B4gder knows someone who made a lib for that :-]
14:08:24*B4gder ducks
14:08:26BgerB4gder really ?
14:08:30linuxstbwget would work :)
14:08:35Bgerhe's your neighbour ? :P
14:08:44Membrilloah well, where do you suggest i start learning C then?
14:08:59Membrilloim getting very fed up with visual basic to be honest
14:09:50Slasherivisual basic is the worst language to start with and it can be very bad for future programming studies.. i recommend you immediately stop using that :)
14:10:06Membrillodone and done
14:10:34Membrilloso where do you suggest i look to get familiar with C?
14:10:48Membrilloall i understand is #define creates a variable i think?
14:10:55linuxstbStart reading the code to the Rockbox plugins to see if you can understand them. I'm sure googling for "c tutorial" will give you one or two websites.
14:10:59Membrillolike Dim blahblah as string in visual basic
14:11:03Bgerveni, vidi, vici, Slasheri :)
14:11:21Membrilloany plugin?
14:11:22Slasherisome good books can help or programming courses in local university or something like that
14:11:38SlasheriBger :)
14:11:45Bgerlike the art of computer programming ? ;)
14:12:17BgerMembrillo start with "helloworld.c"
14:12:20Membrillowow the source is huge
14:12:30 Join TCK [0] (
14:12:31Membrillohelloworld.c. ok
14:12:37Membrilloso what software do i need?
14:12:43linuxstbMembrillo: No, things starting with a # are known as pre-processor directives. These are processed before the compiler sees your program. A simple example is "#define LCD_WIDTH 220" - this means that the pre-processor will replace LCD_WIDTH with 220 everywhere in your code before it's compiled.
14:13:12Membrillooh ok fair enough. and they are always at the top?
14:13:20Bgerthey can be anywhere
14:13:24Membrillothey can be anywhere?
14:13:27Membrillooh ok
14:13:29Membrillofair enough
14:13:34Bgerbut are on lines
14:13:37linuxstbYes - but by convention they are at the top.
14:13:49Membrilloeasy enough
14:13:54Bgerthat is, if a line starts with #, the whole line is considered for the preprocessor
14:14:13Membrilloso out of interest, does C have any similarities to VB?
14:14:18Bgerno :)
14:14:24preglowluckily, no
14:14:30Jungti1234I don't C..
14:14:35Jungti1234I don't know C
14:14:36Membrilloso it will be exactly the same as starting programming from scratch?
14:14:38preglowit's the same type of imperative programming
14:14:46preglowbut the syntax doesn't have too many similarities
14:14:49Jungti1234Know nothing.
14:14:51preglowMembrillo: no
14:15:02preglowit's the same kind of programming
14:15:03Membrillook, thats what i thought
14:15:13preglowjust the language is different and a bit more complex
14:15:13Membrillothe same kind but different syntax
14:15:17Jungti1234'Helloworld.c'is difficult too.
14:15:21SlasheriMembrillo: almost, yes. But after you have learned C i can promise it will be very easy to learn any other language (almost at least)
14:15:21preglownow try learning haskell, and you'll have to learn programming again :)
14:15:38preglowSlasheri: all other imperative languages, at least
14:15:43Bgerpreglow what's this ? :))
14:15:44Slasheriyes, true
14:15:49Slasherihaskell is something totally different
14:15:53preglowfunctional languages are a different cup of tea altogether
14:16:00Membrilloso i can open .c files with a text editor?
14:16:06preglowof course
14:16:14Membrilloso i dont need anything special
14:16:16BgerMembrillo they *are* text files
14:16:18SlasheriMembrillo: sure, but a good editor with syntax hilighting is recommended
14:17:13Membrilloat the moment its MS word :P
14:17:28Bgerdon't use MS Word!
14:17:29lostlogicBger: beat me to it :-P
14:17:29linuxstbMembrillo: Have you compiled Rockbox yourself from the source code?
14:17:40Membrillolinuxstb: yes a little while ago
14:17:50Membrilloi compiled the h300 sim
14:18:04Membrilloive forgotten how to do it, but i can take a look at the guide again
14:18:13Membrillowhats wrong with MS word?
14:18:16BgerMembrillo u're on windows, yep?
14:18:33linuxstbThat's a good place to start. You can change Rockbox and test things in the safety of the sim.
14:18:51BgerMembrillo use this
14:18:58Jungti1234I does it too.. It's easy.
14:19:16B4gderword is not a text editor
14:19:17Membrillodarn i hate setup files
14:19:19lostlogicjedit is also a decent editor
14:19:30Membrillowhy do you have to setup everything you download
14:19:54Membrillomakes no sense to me
14:21:02Membrillook conTEXT is up
14:21:12Membrilloso what does #include do?
14:21:21Membrillojust a reference?
14:21:30BgerMembrillo it includes other file
14:21:41Membrilloi guess at the start of all plugins you have to have #include plugin.h
14:21:46Bgerin most cases so called "header" files (with .h extension)
14:21:57Membrillook nice
14:22:15Bgerheader files contain declaration of structures, functions ...
14:22:31Slasheriand the included header file contains function prototypes, structures etc. so compiler knows function syntax is valid when called outside and it knows structs structures ..
14:23:32 Quit steveb (Remote closed the connection)
14:23:33Membrilloso what other header files will i possibly need to call other than plugin.h?
14:23:42XavierGrMembrillo: A very good editor (small fast and easy) is notepad2.
14:24:05MembrilloXavierGr: ok thanks, ill check it out
14:24:06XavierGrIt will cover all yoour needs for rockbox. Highlighting, whitespaces columnwidths e.t.c
14:24:19 Join steveb [0] (
14:24:52XavierGrIn fact it is a simple exe file like notepad of windows. I have deleted windows notepad and replaced it with notepad2 in order to open all text files with this.
14:25:32DBUGEnqueued KICK XavierGr
14:25:32XavierGr ,04Slasheri,99: visual basic is the worst language to start with and it can be very bad for future programming studies.. i recommend you immediately stop using that :),99. Well I started with visual basic. I know I had a hard time converting, but the logic is the same.
14:26:04Slasherihuh, nice colors :D
14:26:08Membrilloexplain this line to me: static struct plugin_api* rb;
14:26:24Bgeruf, is there something like struct in VB ...
14:26:46Membrilloi dont think so
14:27:00BgerMembrillo struct is something that contains variables from different types
14:27:03SlasheriMembrillo: it creates a new pointer that is used to point to struct plugin_api instance
14:27:07 Quit steveb (Remote closed the connection)
14:27:10XavierGrit must have structs, I just don't remember.
14:27:12Bgerif you have struct { int a; float b;} st;
14:27:24Bgeru can access st.a (which is integer)
14:27:30MembrilloVB has dimensional variables
14:27:33Bgerst.b (which is float, aka real)
14:27:34lostlogic#rockbox Open source Jukebox Firmware and programming lessons in one radical place
14:28:05linuxstbAnd Hungarian notation is strictly forbidden :)
14:29:06B4gderand now, how to do fixed-point math instead of floats...
14:29:08Membrillosplash is the equivalent of msgbox in VB?
14:29:19Membrillopops up a message?
14:29:21BgerMembrillo yep ...
14:29:30SlasheriMembrillo: no, there is no such kind of equivalent..
14:29:39 Join steveb [0] (
14:29:44Membrillooh ok
14:29:46Slasheribut rockbox api has that kind of function which can be used for that
14:30:16Slasherihowever, it most certainly has a different syntax
14:30:33Membrilloso rb->splash(HZ*2, true, "Hello world!"); will display "Hello world!" in what way
14:30:35 Join mikearthur [0] (
14:30:46XavierGrcan you believe this? It is like I lost 433 euros in advancedmp3players.
14:30:58SlasheriMembrillo: yep, that should display "Hello world!" for two seconds
14:31:01BgerMembrillo look at apps/plugin.h
14:31:09Bgerthere is explanation
14:31:15Bgerfor each function
14:31:25linuxstbMembrillo: Do you have an ipod video?
14:31:38Bgerlooks like he has a h300
14:31:43BgerXavierGr ??
14:31:49Membrillooh ok
14:31:57Membrillolinuxstb: yes i do
14:32:07Membrilloi have an ipod video and an h320
14:32:58linuxstbOne useful job (once you get comfortable with Rockbox and C) would be making the existing plugins (including the "plugbox" h300 plugins) work nicely on the video.
14:33:23Bgerhahaha linuxstb
14:33:29linuxstbOr just making the existing Rockbox plugins work on the h320.
14:33:39Membrillohmmm, that might take a while
14:33:40linuxstbI mean work better on the h320.
14:33:53Membrilloit will take me some time to get familiar with C
14:34:47Membrillohow is #if different to #ifdef
14:35:13Bger#ifdef X test whether there was #define X before that
14:35:47Membrillooh ok
14:35:57linuxstb#ifdef X is exactly the same as #if defined(X)
14:36:06linuxstbIt's just a handy shortcut.
14:36:14Membrilloso basically checks wether a 'variable' is empty or not. in a way
14:36:25Bgerhm, no
14:36:30Bgerit's a preprocessor thing
14:37:09Membrillowell #define replaces X with a value. and #ifdef checks wether X has a value? is that rihgt?
14:37:09Bgerit's like whether the preprocessor has defined X [ = value] or hasn't
14:37:23Bgeru can have "unvalued" define
14:37:57Membrilloan example?
14:38:09Bger#define X
14:38:15Bger#ifdef X
14:38:22Bgerthe last one will be true
14:38:43Bgerbut you could have #define X 3
14:39:13Membrilloso what will #define X do? open a spot in the memory for X but leave it empty?
14:39:29BgerMembrillo there's no memory for this
14:39:39Bgerit's just used in the process of compiling
14:39:51Membrillooh ok
14:40:02Membrillowell, what will #define X do without a value after X?
14:40:18Bgerit will just be true in tests like #if defined(X)
14:40:18Membrillowhats to pre-process?
14:40:30Membrilloim with you now
14:40:36 Quit DJDD__ ("Trillian (")
14:40:45 Quit mikearthur (Read error: 104 (Connection reset by peer))
14:40:53linuxstbRockbox uses it for true or false things like HAVE_LCD_BITMAP
14:42:13Membrilloso you could define having a colour LCD screen at the top under X, and later on you could have "ifdef X then <display colours>"
14:42:23XavierGrMembrillio: Look at this example: Some targets have colour screen others don't, so that means different code for each option. But how are we going to compile the correct code?
14:42:27Membrillootherwise its a monochrome lcd
14:42:28XavierGrWe use defines
14:42:44Membrilloyes thats what i thought up there ^
14:42:57XavierGrWe set HAVE_LCD_REMOTE for targets that have remote lcds
14:43:14XavierGrthen when we want code to be compiled only for that option we use an ifdef
14:43:20lostlogicMembrillo: there are several steps to compiling a C program: first the preprocessor handles all of the # stuff which basically generates a new preprocessed C file with all of the #defines processed into the text of the file, the compiler then compiles that into assembly instructions, and finally the assembler turns the assembly into binary machine code.
14:43:54Bgerand after that is the linker .....
14:43:58XavierGrso if the user is going to compile for archos the code that corresponds to the lcd remote will not be compiled
14:43:59Membrilloso for each different version of a plugin you will have a different #define up the top and that will affect the output completely
14:44:19XavierGryes e.g some plugins are not for the sim
14:44:29lostlogicMembrillo: the defines actually end up coming from other files, but in essence yes
14:44:37XavierGrif you open the.c file of such plugin you will see an #ifdef in it
14:44:40linuxstbThe main #defines for all the different targets are in firmware/export/config-*.h
14:45:35Jungti1234I have question
14:46:10Membrilloso you have something like HAVE_COLOUR_LCD? and if thats true (the player has a colour LCD) then you define something or rather up the top saying that?
14:46:59XavierGrmembrillio: open up bejeweled.c
14:47:19lostlogicMembrillo: you don't have to redefine anything, you can just do #ifdef HAVE_COLOUR_LCD code code code code #endif anywhere in your source file. the code will only be compiled if that's defined
14:47:57MembrilloXavierGr: where is bejeweled.c?
14:48:35Membrillooh i get it. So all the definitons for each different build are in a file. and you reference that in your plugin
14:49:01Membrilloim in bejeweled.c
14:49:04XavierGryes different parts of code in there will b compiled for different targets using the #ifdef
14:49:29XavierGrin compile time (as you probably remember) you choose which target you build
14:49:32Membrillothats why the source code is so huge
14:49:38Rob2222Iriver runtime (MP3 128 ABR) now 1h50m over RB runtime and still 2 bars left. muahaha
14:49:42Membrilloyeah i remember that
14:49:58XavierGrso if you build for iriver the defines for the iriver target are created.
14:50:04Membrilloso compiling goes through all the #ifdefs and works out which code to use
14:50:21XavierGrso when the compiler goes to a .c file it will only compile parts of the code that has iriver defines
14:50:32Membrilloi understand it
14:50:40 Quit tvelocity (Remote closed the connection)
14:50:46Membrilloso i guess all the defines are stored in multiple files?
14:51:06lostlogicRob2222: time to make the audio engine faster, go to it.
14:51:10Membrilloso #ifdef HAVE_LCD_BITMAP is has a colour screen?
14:51:25XavierGrno because iriver h100 has that define too
14:51:37B4gderLCD_BITMAP means a bitmapped LCD
14:51:40lostlogicMembrillo: no, that's for if the LCD is rendered by pixel, som eof the archos LCDs are rendered by character instead
14:51:44BgerMembrillo there are 2 rb targets with charcell display
14:51:55Membrillowhats special about an lcd bitmp
14:52:00Rob2222lostlogic: Hehe. This test is my 3rd discharge of my "2200mA" battery. Will take another RB after that iriver test
14:52:03Membrillooh ok
14:52:07BgerMembrillo u can write what you want on them
14:52:08Membrilloi know what you mean
14:52:40Bgerand charcells are based on blocks ... and you tell them "write this character at this position"
14:52:40Membrilloso, it will take a while to learn how to use all the defines
14:53:25Membrillook. i now understand definitions
14:54:04XavierGrMembrillio: better start with C syntax variables structures functions and then pointers
14:55:18linuxstbMembrillo: The only defines you really need to worry about in plugins are CONFIG_KEYPAD, HAVE_LCD_BITMAP, LCD_WIDTH, LCD_HEIGHT and LCD_DEPTH (and maybe HAVE_LCD_COLOR)
14:55:22 Join chopped_pork [0] (
14:55:32linuxstbYou'll find lots of examples of they are used in the plugins.
14:55:46Membrilloso whats LCD_DEPTH?
14:55:53linuxstbHow many bits per pixel.
14:56:06Membrilloah ok
14:56:20linuxstbArchos devices have 1bpp, the H1x0 is 2bpp and all the colour lcds are 16bpp
14:56:40Membrilloso every build has LCD_DEPTH LCD_WIDTH etc, but they are defined under different values
14:57:24linuxstbThe convention is for "optional" defines to begin with "HAVE_" - so you test them using #ifdef HAVE_.....
14:57:50Membrilloso for some H300 code you could use #ifdef LCD_DEPTH 16?
14:58:22Membrilloor isnt it that simple
14:58:28Bgerit isn't
14:58:35Membrilloit never is
14:58:43B4gder#if LCD_DEPTH == 16
14:58:52Membrilloah ok
14:58:56B4gderbut then there are many models with 16 bit depths
14:59:00Membrilloso why two equals signs?
14:59:15B4gderMembrillo: it is a comparion, they use two signs in C
14:59:20B4gderassignments use one
14:59:31B4gdera = 2 vs a == 2
14:59:49Membrilloah ok
15:00:02Membrilloso #if a =2 makes no sense
15:00:13Bgerhm, it makes ...
15:00:14Membrillobut #if a == 2 does
15:00:30B4gderdon't confuse the preprocessor instructions with the actual C code
15:00:30Bgerbut i don't know how's this in preprocessor
15:00:44***Saving seen data "./dancer.seen"
15:00:57Membrilloyeah ive got it
15:01:05Membrilloassignments, one sign comparisons two
15:01:28Membrilloso what if you wanted #if LCD_DEPTH less than 16
15:01:35BgerMembrillo in c code if you have "if(a=2)" this is true ...
15:01:46Bgerbecause (a=2) has value of 2
15:01:53Bgerand 2 != 0 so it is true
15:01:54Membrillowell i have to run now but ill check out some more of the plugins later
15:02:01Membrillosee ya
15:02:04Membrillothanks for the help
15:02:06Bgerin c everything different from 0 is considered "true"
15:02:07 Quit Membrillo ()
15:05:36 Join bshift- [0] (
15:05:36 Quit eEye (Read error: 104 (Connection reset by peer))
15:07:35 Quit steveb ("##uk - The brits still rule the world")
15:07:37 Join steveb [0] (
15:07:53 Quit muesli__ (Read error: 104 (Connection reset by peer))
15:08:07 Quit elinenbe (Read error: 110 (Connection timed out))
15:08:10 Join muesli__ [0] (n=muesli_t@
15:09:26amiconnBger: using (a = 2) instead of (a == 2) has two nasty effects
15:09:40amiconn...for people who come from other languages
15:10:14 Quit Jungti1234 ()
15:10:46Bger2 ?
15:10:51Bgerlike a = 2
15:10:53Bgerand true ? :)
15:11:12Bgeralways true
15:12:00amiconn(1) always true, (2) a will have the value 2 afterwards, as '=' is the assignment operator
15:12:27 Quit steveb (Remote closed the connection)
15:12:48Bgeri just wanted to point him that this is not entirely correct => <Membrillo> so #if a =2 makes no sense
15:12:52amiconnDoing this mistake by accident can be avoided by switching both sides, so that the left side consist of something that can't be an lvalue
15:12:59linuxstbAt least gcc will warn you that (a=1) is always true.
15:13:21linuxstb(if people enable warnings and pay attention to them)
15:13:51 Join steveb [0] (
15:14:12 Part Lxy
15:14:18linuxstbAnd it will even warn you about if (a=1) - forcing you to put extra brackets around a=1
15:14:58Bgersomething like "possibly unwanted assignment" ?
15:15:00 Join tvelocity [0] (
15:15:22linuxstb"warning: suggest parentheses around assignment used as truth value"
15:15:57linuxstbI personally find if (1==a) a lot harder to read.
15:16:11CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:16:11*Bger too
15:16:23LinusNlinuxstb: is there anything special about building the arm compiler?
15:16:34linuxstbNope. But we're using 4.0.2
15:17:26B4gderbut 3.4 was no problem to build either
15:17:31linuxstbWe may have to do something special if we want to mix arm and thumb code - but we haven't reached that stage yet.
15:17:51LinusNgotta go, cu
15:17:53 Part LinusN
15:20:03preglowlinuxstb: had a look at the remapping code?
15:20:32linuxstbNo, not yet.
15:20:35preglowi guess it boils down to what a "page" is for portalplayer
15:20:55linuxstbDo I understand correctly that we can only do a single mapping?
15:21:02preglowno, you can do several
15:21:07preglowthere are four mmap regs, i think
15:21:13linuxstbAh, OK.
15:21:36preglowat least the ipl wiki lists four
15:21:51linuxstbSo in theory at least we should be able to map IRAM to 0x0 and SDRAM close enough to never need long jumps?
15:22:06preglowyeah, so long as we keep our code at the start
15:22:49linuxstbYes, my idea is PLUGINS -> CODECS -> rockbox.bin at the start of SDRAM.
15:23:20 Quit TCK (Read error: 104 (Connection reset by peer))
15:23:20preglowwould be really great
15:23:27preglowi'm bloody fed up of the long calls already
15:30:56 Join mirak [0] (
15:31:00amiconnThat would require a relocatable plugin format
15:31:17amiconnWhy not just map the iram close enough to never need long jumps?
15:31:25preglowbecause it's not possible for all ram sizes
15:31:39preglowiram would always have to be mapped above ram
15:31:45amiconnAfaik, the arm can do short jumps in a 64 MB range
15:31:46preglowand for ram sizes bigger than 32, that wouldn't work
15:31:53preglowyes, +/-32
15:31:54amiconnand the iPod has 32MB ram, correct?
15:31:59 Quit ashridah ("Leaving")
15:32:00 Join gtkspert [0] (
15:32:01preglowipod video 60 gig has 64
15:32:10muesli__not bad
15:32:12preglowso we'll even hit the limit with the current generation
15:32:37amiconnThen map the iram 'on top' of the dram, covering (iram_size) bytes of it
15:32:37preglowwhy would that require a relocatable plugin format, btw?
15:32:54amiconnBecause the rockbox binary size will vary
15:33:15 Join Mmmm [0] (
15:33:15 Quit thegeek (Read error: 113 (No route to host))
15:33:16linuxstbThe idea is for plugins and codecs to be fixed at the very start of SDRAM - before the rockbox binary
15:33:29linuxstbInstead of the very end.
15:33:32amiconnWell, if that's possible...
15:33:52linuxstbWe can then do run-time detection of the available RAM - meaning 1 less target.
15:34:41amiconnIn fact this would be possible on other targets too where we have our own loader
15:34:54amiconn(thinking of H100 vs. H120)
15:36:21amiconnIt would mean that crt0.S has to detect the RAM size, and probably set some flag somewhere too
15:36:38amiconn(H100 and H120 have one slight difference apart from the RAM size)
15:36:43linuxstbWhy crt0.S? Isn't that RAM just used for the audio buffer?
15:37:17amiconnThe RAM has to be initialised properly
15:37:26amiconnNah, it's done in the loader
15:37:40linuxstbAt worse, we can have separate bootloaders.
15:38:29amiconnStill, rockbox has to know the RAM size, for setting the buffer end address and for setting the proper DRAM refresh on frequency change
15:40:06 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
15:41:25preglowlinuxstb: this scheme might break horribly on platforms with no mmap units, though
15:41:33preglowhope i wont work on one
15:41:46 Join San [0] (
15:56:30 Quit AliasCoffee ("Leaving")
15:56:50linuxstbpreglow: Does the PP5002 code in ipodlinux do the same remapping? I can't find any info about a PP5002 memory controller on their wiki
15:57:26coobthere's a memory controller on both
15:57:33coobjust quite limited
15:59:02linuxstbSo the same remapping of SDRAM to 0x0 is done in IPL for the PP5002?
15:59:31preglowlinuxstb: yes
15:59:43preglowi even think both share the same register addresses
16:00:22linuxstbYes, I'm just reading their code now.
16:01:26preglowlooks like the minimum remappable chunk of memory is 512kb
16:01:33preglowif the mask size is anything to go by
16:01:46Rob2222Wow... H300: RB runtime 12,5h, IR runtime now 16h and still 2 of 4 bars left, same volume, same MP3s ...
16:02:53preglowbut what is a portalplayer page?
16:03:14preglowregister descriptions use a page as a unit
16:03:24mirakdo this remmapping stuff on ly for ipod ?
16:03:41preglownone of the other targets are capable of it
16:04:54lostlogicRob2222: some of the recent bugfixes to the audio engine hurt performance (at least in my testing), it needs work.
16:05:17preglownah, i think a lot of other factors than the playback engine or playing in here
16:06:27lostlogichmm... I dunno, the boost percentage makes a pretty darn big runtime difference
16:07:41Rob2222lostlogic: my RB bench build is ca 1 week old
16:07:43XavierGrRob2222: Also amiconn noted that there is a great power consumption on rockbox somewhere.
16:08:29Rob2222maybe by "bad 2200mA" battery starts to rise its power. cause of that i will do a RB bench after that agein
16:09:25Rob2222but I dont think. 1st discharge RB bench got almost identical resulst as 2nd discharge RB bench
16:11:52ep0chwell someone could do firmware battery test by not playing any music at all. That could give an indication as to whether its the playback thats hurting performance or not.
16:12:08ep0chbut that might be a long long test :)
16:13:02Rob2222well i did some power concumption measurements
16:13:14 Quit Pingoomax ("A+")
16:13:17Rob2222but some people meant, that they arent usefull
16:13:48Rob2222which could be right, cause I didnt have professional equipment.
16:14:56preglowlike i said
16:15:02preglowi very much doubt the playback engine is at fault
16:15:12preglowafaik, we get far better battery times for h1x0
16:15:58XavierGryes I get 25 hours with an 1900mah battery on my H160
16:16:33 Join wladston [0] (n=wladston@
16:16:59 Join ModernExecutive [0] (
16:17:05 Quit ModernExecutive (Client Quit)
16:17:28Rob2222Soneone knows that ?=> "One thing I found out. When in PAUSE Mode and switching
16:17:28Rob2222to a new title and staying in PAUSE, the CPU _doesnt_
16:17:28Rob2222switch back. The CPU stays at 124MHz until you start playback."
16:19:09 Join ModernExecutive [0] (
16:19:36ep0chi'll see if i can check that on the 120
16:20:25Bgerit should be the same ...
16:20:29preglowRob2222: interesting
16:20:39preglowif it's like that on h3x0, it's like that on h1x0
16:20:54preglowand it sounds like a likely bug
16:21:01ep0chif you pause when its not boosting its stays not boosting
16:21:11ep0chwhen you pause in boost it stays in boost
16:21:58amiconnRob2222: Are you running your tests on rockbox with the battery status plugin in the background?
16:22:13ep0chwhen you switch song its likely buffering from disk and hence boost is on
16:24:30Rob2222amiconn: Yes, right.
16:24:44 Part wladston
16:24:51amiconnPerhaps there's something wrong in this plugin's thread (not sleeping the CPU)
16:24:58Rob2222CVS + battery_bench patch ca. 1 week old.
16:25:00amiconnI'll check that tonight
16:25:19preglowthe current measurements don't look too bad
16:25:27amiconnI get >12 hours on H300 with rockbox + stock battery
16:25:28Rob2222i changed default background and foreground colors in LCD.h. That shouldnt matter something, or?
16:25:32amiconn(intermittent use)
16:26:50Bgeramiconn the plugin ver he used is very different from current one
16:28:05Rob2222After my IR bench is complete, I make again a bench with the old version, to check if maybe the battery capacity has rised. after that i will make a final RB bench with a actual version without battery_bench plugin. I'll keep you informed.
16:29:39linuxstbIf anyone's interested, I've just tested the speed of starfield on my two ipods (Color and Video) and H140. The ipod color achieves 33fps, the Video 14fps and the H140 does 36fps. This is after removing the sleep(1) from the main loop.
16:31:08linuxstbMy "fps display" patch is here if anyone wants to try other targets:
16:31:19linuxstbI would be curious to know the h300 result.
16:32:09 Join webguest16 [0] (
16:32:55 Quit webguest16 (Client Quit)
16:32:58 Join webguest16 [0] (
16:33:03 Quit ModernExecutive ()
16:33:06ep0chand the nano
16:34:03 Quit XavierGr ("Trillian (")
16:34:17webguest16bin ich?
16:34:31 Quit webguest16 (Client Quit)
16:34:38 Quit gtkspert (Read error: 104 (Connection reset by peer))
16:34:40preglowthink i'll try it on nano
16:35:19ep0chis the X+Y size of the starfield consistent on all platforms?
16:37:04 Join gtkspert [0] (
16:37:34ep0chi think to get a fair comparison the dimensions should be the same
16:39:51preglowholy shit
16:39:53preglow83 fps
16:40:38preglow82 stable
16:41:04linuxstbThat's faster than my monitor....
16:41:35preglowmore people should get nanos
16:41:41amiconnlinuxstb: I expect <10 fps on H300...
16:41:55ep0chwhats the screen size of the nano?
16:41:58preglowis performance that lousy? ipod video has a huge screen
16:42:30amiconnAfaik iPod color resolutio and H300 resolution are the same
16:42:46preglow220x176 for ipod color
16:42:53linuxstbamiconn: Yes. They even use the same lcd controller - but in different ways I think.
16:42:59Rob2222Didnt you think you should patch the plugin to a constant screen size for that test?
16:44:18 Quit Paul_The_Ner1 ("Leaving.")
16:44:20 Join Infirit [0] (
16:44:33preglowstarfield looks really nice, btw
16:44:38 Quit Infirit (Client Quit)
16:44:49 Join Paul_The_Nerd [0] (
16:44:57linuxstbpreglow: I know. I hope that multi-color starfield patch never gets committed.
16:45:18preglownow TRAILS, that would be cool
16:45:38preglowi find myself wanting to hack some feedback on top of this thing
16:45:57preglowlinuxstb: btw, i think i know how to do the remapping now
16:46:26linuxstbCool. So if we mapped IRAM to 0x0, how close to it could we put SDRAM?
16:46:43preglowwithin 512k, if i'm correct
16:46:51preglowif the specs i've got are true, i'm wrong
16:47:30linuxstbSo you're assuming that the top two bits mean something?
16:49:16preglowwell, i hope they
16:49:22preglowi they don't we can't remap iram
16:50:17preglowbtw, it doesn't seem to me like the video lcd can get very much faster without offloading screen drawing routines to the broadcom chip
16:53:59linuxstbAccording to what I've heard the ipl people say, it's the lcd_bcm_finishup() function that takes all the time. But I haven't really looked at it yet - I'm mostly ignoring the video for now.
16:54:45preglowgood :>
16:54:53preglowbut i'm bloody starving
16:55:07preglowgotta go the shops, then see if i can get some quick mmapping going
16:59:16 Quit B4gder ("time to say moo")
17:00:48***Saving seen data "./dancer.seen"
17:01:21 Quit Mmmm (Read error: 104 (Connection reset by peer))
17:03:21Rob2222dumb question: what is IRAM? Internal ram of the cpu core?
17:05:51lostlogicyes, SRAM located on the core, as fast as cache
17:06:00lostlogicbut user programmable so not like cache
17:07:00Rob2222ok. did all targets have iram?
17:07:19Rob2222sure the size of the iram of the different target differs, right?
17:08:08lostlogicno, not all targets
17:08:29lostlogicand yes, the new ipod video has 1/3 more IRAM than the older ipods or iRivers or iAudio, IIRC
17:09:39amiconnYes, all current targets have IRAM
17:09:46amiconnJust the size is different
17:10:18lostlogicoops, my mistake, I thought some of the archos didn't
17:10:51amiconnSH1 (archos player/recorders/ondios): 4KB, coldfire MCF5249 (iriver H1x0, H3x0): 96KB, coldfire SCF5250 (iAudio X5): 128KB
17:10:55Paul_The_Nerdlinuxstb: What is the point of this line in loading retail os? ipod.c:451 entry = fw->addr + fw->entryOffset;
17:11:32amiconnAfaik, iPods have 96KB or 128KB depending on the CPU
17:13:01linuxstbPaul_The_Nerd: It calculates the entry point for retailos - "addr" is the load address, and "entry offset" is the offset of the entry point within the image. fw->addr is normally 0x10000000 and fw->entryOffset is normally 0.
17:13:28linuxstbamiconn: Yes, PP5020 has 96KB, PP5021 (used in Nano and Video) has 128KB.
17:13:50linuxstbRockbox doesn't yet see any difference between PP5020 and PP5021 though.
17:14:13 Quit `3nergy (SendQ exceeded)
17:14:15Paul_The_Nerdlinuxstb: In the old way, you then returned entry, but now "return (void*)fw->addr;" and you don't seem to use entry..
17:14:32 Join `3nergy [0] (
17:15:23Paul_The_NerdThe only place I see entry used is after the infinite for loop
17:15:26linuxstbThat's a bug then - it should be return (void*)entry". At least with my retailos entryOffset is 0, so it doesn't matter.
17:15:42Paul_The_NerdAh, good then. I shall change it and see.
17:16:07 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
17:18:48 Join Mongey [0] (
17:18:59 Quit Mongey (Client Quit)
17:19:01 Join Mongey [0] (
17:19:05 Quit Mongey (Client Quit)
17:20:06linuxstbPaul_The_Nerd: If you look at the definition of struct image_t at the start of bootloader/ipod.c, that documents the header in an Apple firmware image. You can compare that with your retailos image.
17:21:23Paul_The_Nerdlinuxstb: Apparently that's what I'll be doing next then. Stiil no go on the executing step.
17:24:27 Join RotAtoR [0] (
17:28:36 Quit Zagor ("Client exiting")
17:32:37amiconnlinuxstb: What about PP5002 and PP5022 ?
17:37:24linuxstbPP5002 is 96KB and PP5022 is 128KB.
17:38:34Paul_The_Nerdlinuxstb: 'nother question. What is the "padding" for, and why is it subtracted?
17:41:42 Quit Febs ()
17:46:39amiconnOooh, cygwin packages with sh-elf-binutils, sh-elf-gcc, m68k-elf-binutils and m68k-elf-gcc on the rockbox site :)
17:47:02 Nick Lynx_ is now known as Lynx_awy (
17:47:15linuxstbPaul_The_Nerd: That could be the problem.... Give me 5 minutes.
17:47:26linuxstbamiconn: What about arm-elf-gcc? :(
17:48:01amiconnAsk LinusN, he provided these packages. See wiki
17:48:05linuxstbAh, I guess that's why Linus was asking about compiling arm-gcc earlier.
17:48:15Paul_The_Nerdlinuxstb: Hehehe. padding == fw->devOffset which is why I asked. Heh.
17:48:50linuxstbYes, but going back to the original ipodlinux bootloader code, padding should be 0x4600 for the 4g and later ipods...
17:49:55Paul_The_NerdIt's 0x4400 here
17:50:12linuxstbYes, it's 0x4400 in CVS
17:50:20linuxstbAnd devOffset is 0x4400...
17:51:13linuxstbWhich doesn't make sense to me - the actual firmware starts at offset 0x200 in my retailos image. So just try changing that memmove to be memmove16(fw->addr, fw->addr + 0x200, fw->len)
17:52:29Paul_The_NerdUm, shouldn't that be loadbuffer + 0x200? (Just making sure...)
17:52:36linuxstbYes, just about to say that...
17:54:20Paul_The_NerdThat causes it to reboot, load the bootloader again normally, and go into rockbox.
17:55:17Paul_The_NerdIf padding should be 0x4600, and devOffset should be 0x4400, it should be -0x200 no?
17:55:37Paul_The_NerdAt least, judging from the math that was applied before.
17:55:53Paul_The_NerdThat's weird.
17:56:31mirakPaul_The_Nerd: any succes with mpeg2 ?
17:57:18Paul_The_Nerdmirak: I'm sorry, but none. Didn't have the chance I thought I did.
18:00:53Paul_The_Nerdlinuxstb: But, looking at Apple_OS.bin the first non-0x00 value after the header is at 0x200
18:01:38linuxstbYes, which is why I think we need to add 0x200 to the loadbuffer. Those first instructions at 0x200 are the arm exception vectors.
18:03:09Paul_The_NerdWell, I tried it with 0x200 added, and it seemed to just reboot the bootloader.
18:03:16 Join slarti [0] (n=tmartin@gentoo/developer/slarti)
18:04:24 Join DT291 [0] (
18:04:24 Quit DreamTactix291 (Read error: 104 (Connection reset by peer))
18:04:42linuxstbDoes the "execute" cause the reboot, or does the reboot happen before then?
18:05:23Paul_The_NerdIt says executing
18:05:36Paul_The_NerdThen an Apple logo shows up, and the rockbox bootloader runs again.
18:09:07 Quit linuxstb (Read error: 104 (Connection reset by peer))
18:12:46 Join Thus0 [0] (
18:12:49slartidoes anyone know how to force rockbox into leaving, say, a one pixel thick uncoloured border on the outside of the display so that the text doesn't run up against the outside of it? (H300)
18:15:39 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
18:18:05 Join SteL [0] (
18:19:37 Join Philip_0729 [0] (
18:23:26 Join linuxstb [0] (
18:32:51 Join LinusN [0] (
18:35:46 Quit SteL ("Chatzilla 0.9.67+ [SeaMonkey 1.0b/2005121900]")
18:36:26 Join SimonJW [0] (
18:37:00Bagderslarti: that'd take a change all over I guess
18:39:57 Quit chopped_pork ("leaving")
18:40:28slartiBagder: hrm, okay
18:41:01 Quit Matze ("Miranda IM! Smaller, Faster, Easier.")
18:43:25 Join thegeek [0] (
18:44:20 Quit tvelocity (Read error: 113 (No route to host))
18:44:49 Join tvelocity [0] (
18:45:23preglowahh, cygwin tools at our site is great news
18:50:51 Join tucoz [0] (
18:53:10SimonJWGuys, just updated my iriver H140 with new mp3's and the latest daily build, and now it completely locks up about 5, 6 seconds in (no matter what you are currently doing). Tried rolling back to the earliest build I could find, one I'm sure I've used before, and it's still doing it. Any ideas what may be causing it?
18:53:39tucozamiconn: are you around? Just found a fun little project that might be possible to do on the colour targets.
18:53:58SimonJWIf not should I'll try and format, fsck and then completely reload everything onto it, and see if it's still happening?
18:54:15 Quit tvelocity ("Leaving")
18:54:43tucozSimonJW: maybe you should just delete the .rockbox folder and unzip the latest build again?
18:55:07SlasheriSimonJW: probably you have uploaded new "too long" filenames to the player and it locks up while dircache is scanning on background
18:55:32Slasherithat is a known bug somewhere in the rockbox filesystem code
18:55:39SimonJWSlasheri: Ah! Now that may be it, one of the files does have a very long name
18:55:46SimonJWcheers, I'll try removing that, thanks a lot
18:55:58Slasherilets hope that will help
18:56:18tucozit was just that I hadn't updated my build in ages, and I updated recently and got an I04 or something when loading a gameboy file.
18:56:41preglowtucoz: amiconn doesn't much like anti-aliased fonts, heh
18:56:46tucozbut Slasheri is probably right on this problem
18:57:38tucozpreglow: hehe, but this would be in the same ballpark for the colour lcd's as the grayscale lib is for the mono-lcd's
18:57:48 Quit gtkspert ()
18:57:48preglowno, not really
18:57:51preglowwe don't get more colours
18:58:01preglowjust "improved" resolution
18:58:05tucoznah, but you get something you don't have
18:58:08preglowand it might not be pretty on a small display
18:58:15 Join gtkspert [0] (
18:58:35preglowbut sure, sounds like fun to me
18:59:09preglowi proposed supports for alpha-blended fonts yesterday, didn't seem to be the most wanted feature ever
18:59:16Paul_The_NerdSub-pixel rendering makes my eyes water, apparently
18:59:19 Quit gtkspert (Client Quit)
19:00:50***Saving seen data "./dancer.seen"
19:02:05SimonJWSlasheri: That worked! Thank you so much, I was pretty worried there for a sec
19:02:17SlasheriSimonJW: great :)
19:02:21SimonJWDamn Sufjan Stevens and his very-long named songs :)
19:02:30Slasherihehe :D
19:02:50Paul_The_Nerdlinuxstb: You said the retailOS from fat32 works on an ipod belonging to you?
19:03:10 Quit SimonJW ("Thanks for the help, #rockbox")
19:03:22preglowi love sub-pixel rendering when done properly
19:03:25preglowmakes it look damned nice
19:04:13tucozpreglow: in what context have you used or had a look at it?
19:06:28Paul_The_Nerdpreglow: Apparently it was the BGR one that was making my eyes water, even when it was only peripheral.
19:09:06 Part tucoz
19:09:43preglowperhaps because that's not usually the right order?
19:10:44Paul_The_NerdI was looking at the RGB one, but it was still there below it. When I scrolled it off the bottom, my eyes were fine again.
19:11:52Paul_The_NerdSo, how would you do this anyway? A step in creating the fonts?
19:13:00 Join linuxstb_ [0] (
19:13:02 Quit Rob2222 (Read error: 104 (Connection reset by peer))
19:13:16 Quit linuxstb (Nick collision from services.)
19:13:23 Nick linuxstb_ is now known as linuxstb (
19:13:39linuxstbPaul_The_Nerd: Loading retailos from fat32 works on my color, but not the video.
19:14:26Sananyone else getting a database error on Misticriver
19:15:16 Join Rob2222 [0] (
19:15:37 Part LinusN
19:15:45 Join Febs [0] (n=medifebb@
19:15:45 Quit paugh ("Leaving")
19:16:53Sanits fixed now phillip
19:18:59 Join airwin [0] (
19:20:31airwinno idear what this is about :D, just looking for a dev willing to take a look at the remote patch. I made a complete log of the errors is this topic
19:22:10airwinif i do this completely the wrong way, i'm sorry i'll read more about how to later on when i get home. Tanx for all the work anyhow you guys :D!!! Greetz air.
19:22:17 Quit Rob2222 (Read error: 104 (Connection reset by peer))
19:23:00 Join Rob2222 [0] (
19:26:36 Join ghode|afk [0] (
19:26:57 Quit Pantalaimon ("Gone!")
19:27:16 Join steveb_ [0] (
19:29:10 Join _FireFly_ [0] (
19:32:35 Join mikearthur [0] (
19:33:16Paul_The_Nerdlinuxstb: Are you 100% sure you created the bootloader for your color with dummy.bin rather than apple_os.bin?
19:34:40linuxstbYes - it boots about 4-5 seconds quicker with dummy.bin.
19:35:50Rob2222what is in dummy.bin?
19:36:03 Join linuxstb_ [0] (
19:36:21Paul_The_NerdBecause if I create the bootloader with apple_os, it goes into retail properly, which leads me to assume that it's still using the entry point in boot_table to start, rather than in loadbuffer.
19:37:17Paul_The_NerdWhich, if my understanding is okay, makes sense, because rockbox is at 0x4600 with the bootloader/dummy file, and apple_os is at 0x4600 in the apple_os/bootloader version, and in the first the code causes bootloader to be rerun, and in the latter, apple_os gets launched.
19:37:22*Paul_The_Nerd could be seeing this all wrong.
19:38:39Paul_The_NerdSo, I'm guessing that the memmove16()'s results get overwritten at some point (while the apple logo is displayed, I guess?)
19:38:46 Quit darkless (Read error: 104 (Connection reset by peer))
19:38:51 Quit Maxime (Read error: 104 (Connection reset by peer))
19:39:01 Join Maxime [0] (
19:39:20 Join darkless [0] (
19:41:34linuxstb_Paul_The_Nerd: As a test, you could try creating a "dummy.bin" exactly the same size as the apple firmware. Also, are you using "-l" (ell) with ipod_fw, and not -i ?
19:42:46linuxstb_dummy.bin is just there to keep ipod_fw happy - if we can get this to work, we can change ipod_fw so that it just creates a firmware partition containing Rockbox and nothing else.
19:43:40Paul_The_Nerdlinuxstb_: Apparently I hadn't been
19:43:48linuxstb_Rob2222: dummy.bin is just a file of 1024 zero bytes - hence the name.
19:43:55Paul_The_NerdAt least, not today. Yesterday I used the line exactly as in your script
19:44:07linuxstb_I don't _think_ it matters.
19:44:13Paul_The_NerdWith -l it works
19:44:33Paul_The_Nerdmemmove16(fw->addr, loadbuffer + 0x200, fw->len); and return entry;
19:44:41linuxstb_And RetailOS works?
19:45:09linuxstb_Does it make any noticable difference to the bootup time when starting Rockbox?
19:45:44linuxstb_It means the apple bootloader doesn't copy a 5.5MB firmware image to RAM that we don't need.
19:46:49Paul_The_NerdOn a hard reboot it displays the apple logo for a much shorter time, yes.
19:47:59linuxstb_I just need to get it working on the video now then.
19:48:14 Nick steveb_ is now known as Pantalaimon (
19:48:40Paul_The_NerdWell, I did have to change that bit in the memmove, are you sure that's right for the video?
19:49:14preglow5.5 meg firmware image???
19:49:21linuxstb_Yes, I think it's the same as your version. I'll need to do some more tests. The video is different because the apple firmware is in two parts - the normal retailos, and then another firmware that runs on the broadcom chip.
19:49:40linuxstb_preglow: The "apple_os.bin" file - that's how bit mine is.
19:49:51preglowhow big is retailos?
19:49:55preglowthat is reatilos
19:50:05preglowit uses all that ram for the os?
19:50:21linuxstb_It seems to.
19:50:48preglowmine is 4.7 megs
19:50:52 Join lamed [0] (
19:50:53preglowprobably bitmap size differences
19:51:06 Quit mikearthur (Remote closed the connection)
19:51:06preglowit's got a rather large font, for one
19:51:11linuxstb_Yep. The video is about 5.8MB.
19:51:18linuxstb_But that has more features...
19:51:36 Quit linuxstb (Read error: 110 (Connection timed out))
19:52:06linuxstb_But anyway, if you shutdown your Nano from within Rockbox, can you start the RetailOS by pressing MENU (to wake it up) and then keeping your finger pressed on MENU?
19:52:11lamedHi, Guys, It should be a stupid question to ask whether I should edit the wikimanual for inserting the new functions that rockbox now contains and remove some that we no longer have, right? (they might not make it to the release)
19:52:50Paul_The_Nerdlinuxstb_: Yes
19:53:11linuxstb_lamed: Sounds sensible to me. But what features have been removed?
19:53:22 Join mikearthur [0] (
19:53:49lamedacctually, I think only Bidi langs... as unicode automaticly recognize when it should be activated.
19:54:49 Quit uwe_ (Read error: 113 (No route to host))
19:56:51 Quit _FireFly_ ("Leaving")
19:56:55 Quit mikearthur (Read error: 104 (Connection reset by peer))
19:57:01 Join uwe_ [0] (
19:57:08 Join mikearthur [0] (
19:57:41 Join elinenbe [0] (
19:57:46Paul_The_Nerdlinuxstb_: Though, it doesn't go into USB mode any more, as far as I can tell.
19:58:05preglowlinuxstb_: a bit hard to say, i'm not using the button driver you are
19:58:28preglowlinuxstb_: but it certainly seems this one does, yes
19:59:00 Join ed-210 [0] (
19:59:18preglowforget that, this button driver isn't consistent at all
19:59:47 Nick ed-210 is now known as ed-209 (
20:00:29 Join XavierGr [0] (
20:01:38 Part XavierGr
20:05:04 Join elinenbe_ [0] (
20:06:55Rob2222H300: After 20h MP3 playback me battery indicator is blank. Now lets see when it shuts off. ^^ (RB: 12,5h Playback same playlist, same volume) ^^
20:08:11 Quit ed-209 ("CGI:IRC")
20:10:20Paul_The_Nerdlinuxstb_: RetailOS becomes somewhat pointless when loaded from disk, it seems. iTunes doesn't recognize the ipod, and windows won't let you access it as a mass storage device. It also doesn't seem to play music.
20:10:32BHSPitLappy20 hours??
20:11:04Rob2222BHSPitLappy: Yes, IR-Firmware plays my playlist now since 20h and 10 minutes
20:11:23preglowRob2222: and that's with a non-stock battery, yes?
20:11:56 Join ModernExecutive [0] (
20:13:02Rob2222Yes, this battery is a 2200 that should have ca. 1666mA. Calcualted of that a friend of mine gets 10,5h RB (same playlist) with the 1400er stock battery of new player and I get 12,5h with my new cheap 2200er.
20:14:17 Join Rondom [0] (
20:14:18Rob2222This Iriver Bench is now my 3rd discharge of that new battery. 1st and 2nd done in RB and got 12:30 and 12:20.
20:14:39Rob2222Before each bench the battery was fully charged until charger stopped loading
20:15:38 Join tvelocity [0] (
20:18:19 Join Matze [0] (
20:18:54 Part Pantalaimon ("##uk - The brits still rule the world")
20:19:05 Quit elinenbe (Read error: 104 (Connection reset by peer))
20:19:06 Nick elinenbe_ is now known as elinenbe (
20:26:17 Join Bger_ [0] (n=Bager@
20:26:44 Quit Bger (Nick collision from services.)
20:26:47 Nick Bger_ is now known as Bger (n=Bager@
20:26:49 Quit airwin ("CGI:IRC (EOF)")
20:26:49 Quit stamppot ("CGI:IRC (EOF)")
20:26:49 Quit lamed ("CGI:IRC (EOF)")
20:28:01 Join Bger_ [0] (n=Bager@
20:30:03 Quit ModernExecutive ()
20:32:50 Quit Bagder (Read error: 110 (Connection timed out))
20:33:01 Join Bagder [0] (
20:41:18 Quit Kohlrabi ("Fast alle Menschen sind Regenw├╝rmer")
20:43:32 Join amiconn_ [0] (
20:43:41 Join lamed [0] (
20:43:47 Quit amiconn (Nick collision from services.)
20:43:52 Nick amiconn_ is now known as amiconn (
20:44:00amiconnGahbah :(
20:44:21lamedI don't get it. all Hx00 players can do AB repeat. why do we have that |#ifdef AB_REPEAT_ENABLE| at gwps.h then?
20:44:44amiconnCurrent GoogleEarth hard-crashes my machine in DirectX mode :(
20:45:06amiconnThe old version (which told me that I have to update) did not do this
20:45:45lostlogiclamed: because some people could turn it off in their config
20:45:53lostlogicor: why not?
20:46:19 Join Infirit [0] (
20:46:23 Part Infirit ("Leaving")
20:46:44amiconnlamed: There are not only iriver players
20:47:30lostlogicamiconn: the section is with in a #ifdef IRIVER_H300_SERIES || IRIVER_H100_SERIES
20:47:55lamedAmiconn: it's under H100/H300 - yeah what lostlogic just said :)
20:48:37amiconnAh, yes
20:48:51amiconnStill AB repeat can be disabled in the config
20:49:31amiconnIt's the same with the archos recorders and player
20:49:49lamed:/ Anyhow I'm working on a patch
20:50:23lostlogiclamed: I think you should go ahead and separate out the H1 and H3 series button maps for wps in there
20:50:36lostlogicbecause I think they should probably be different, because the button locations are _so_ different
20:50:49 Join __peer__ [0] (
20:54:11lamedI don't own an h300 so I have no idea what's a better map... I'm doing the button_pre trick over the ab repeat function so we could use it over the h100 series. later on i'll do the same thing over the screen scroll patch
20:57:01 Quit Matze (Read error: 113 (No route to host))
21:00:51***Saving seen data "./dancer.seen"
21:05:57 Quit mikearthur (Remote closed the connection)
21:07:05Rob2222finally my battery beeping starts at iriver FW after 21h
21:07:07lostlogiclamed: I just meant that you should separate out the GWPS button mappings for H3 and H1 into separate #ifs because I think they will diverge further soon
21:07:33Rob2222is it ok to stop the iriver bench when battery beeping is starting?
21:08:29 Join mikearthur [0] (
21:08:57 Part ep0ch ("Kopete 0.11 :")
21:09:16 Quit lamed ("CGI:IRC")
21:09:38amiconnlostlogic: In what way?
21:09:44 Quit mikearthur (Remote closed the connection)
21:12:07Rob2222H300 buttons are near. H100 far away from another. That he means i think
21:12:13 Join mikearthur [0] (
21:12:33 Join mymomthelush [0] (
21:12:35Rob2222=> different button combinations for each target
21:12:54 Quit solexx (Remote closed the connection)
21:16:03 Quit mikearthur (Remote closed the connection)
21:21:18 Join mikearthur [0] (
21:29:43 Part mymomthelush
21:30:13 Quit igor47 (Remote closed the connection)
21:30:48 Join igor47 [0] (
21:32:34Rob2222Final iriver FW shutoff after 21,5h
21:33:06 Join Infirit [0] (
21:33:55Paul_The_NerdAlright, I'll bite. I have no idea why RetailOS no longer plays music. As far as I can tell, it's any time it tries to output sound (for example, it'll also freeze if I set the clickwheel to output through the headphones). It also freezes when it goes into USB mode.
21:34:21preglowand this happens when you load it from a file?
21:34:38preglowabsolutely certain you load it where it's supposed to be?
21:34:58 Quit mikearthur (Read error: 104 (Connection reset by peer))
21:35:04 Join mikearthur [0] (
21:35:29Paul_The_NerdI'm fairly certain. Linuxstb wrote the code.
21:35:36preglowwell, is he certain? :)
21:35:42 Quit goa (Read error: 131 (Connection reset by peer))
21:35:46amiconnNow, let's see whether this will become anything useful
21:35:51preglowthis what?
21:35:54Paul_The_NerdWell, retailos works fine until you get to that point.
21:36:01*amiconn is going to install debian for amd64 on his brand-new box
21:36:08 Join goa [0] (
21:39:08preglowwhy not try anything else than debian?
21:39:37 Join Luthion [0] (
21:40:28linuxstb_Paul_The_Nerd: Yes, I've got the same problem with music in retailos. I think I'll just delete the retailos...
21:40:50preglowthat's what i'll do when we've got playback
21:40:58*amiconn wonders how to get debian to recognise the raid controller
21:41:11Paul_The_NerdYeah, I planned on deleting it once I had playback too.
21:41:16preglowamiconn: like i said, why not try something that's supposed to work out of the box?
21:41:27amiconnwindows? ;)
21:41:35preglowlike a distro
21:41:47Luthionsomebody mind helping me a bit with the Simplified Guide To Compiling?
21:41:55preglowubuntu is debian based, has got amd64 binaries, and i'd be very interested in seeing how it fares on your setup
21:42:14preglowdon't know how much you need to download, though
21:42:34amiconnI tried ubuntu vs. debian testing on vmware
21:42:41amiconnUbuntu lost the comparison
21:42:44Paul_The_Nerdlinuxstb_: I still kinda wish I knew what was causing it though. I mean, none of your changes seem to me like they'd do it after a few glances.
21:43:10 Join matsl [0] (
21:44:51preglowamiconn: what kind of comparison?
21:45:14 Quit Bger (Read error: 104 (Connection reset by peer))
21:45:26amiconnWhat works (after *some* fiddling) and what does not
21:46:03LuthionOn the Simplified Guide To Compiling: When I'm trying to do the step "In that folder type ../tools/configure which will bring up a list of all possible devices" I just get an error: "BASH: ../tool/configure: No such file or directory". Any idea about what I'm doing wrong?
21:46:11preglowi'd have thought ubuntu beat the shit out of debian in that regard, though
21:46:30preglowLuthion: what directory are you in?
21:46:40Luthionthe dir I just made
21:46:52amiconnUbuntu even installs a gcc that doesn't match the kernel
21:46:55preglowsure, and "dir .." shows a "tools" dir?
21:47:08preglowamiconn: doesn't match the kernel how?
21:47:29amiconnkernel was built with gcc 3.4.5, installed gcc is 4.0.2
21:47:40Luthionyep, "dir .." shows, among others, a tools dir
21:47:45preglowamiconn: so what? having a separate gcc around for compiling kernels is pretty common for some reason
21:47:54preglowamiconn: and anyway, i don't see why it matters
21:48:15preglowLuthion: what happens if you type "bash ../tools/configure" ?
21:48:45Luthionpreglow: "../tool/configure: ../tool/configure: No such file or directory"
21:49:07preglowLuthion: tool?
21:49:11preglowLuthion: you mean tools, right?
21:49:25Luthiondoh :P
21:49:27Luthionnvm then
21:49:30Luthionthx =)
21:49:36preglowhehe, np
21:56:34Paul_The_Nerdpreglow: As best as I can tell, the firmware is copied to the same place, in the same way, as when loaded from the special partition. The only other thing he did was add INPUT(crt0.o)
21:56:34Paul_The_Nerd to in the PP5020 section.
21:59:01lostlogicHmm... is there a way in the current rockbox to send an event to a thread and sleep until that event has been executed on that other thread before awakening?
22:04:10 Quit goa (Read error: 104 (Connection reset by peer))
22:05:53LuthionI'm looking at the patch . I'd guess that the line "+int onplay_one_click(char* file, int attr, int from,struct tree_context* c){" is the one that identifies which button the code should trigger on. Right? If I want it to trigger on, let's say the record button (H120). Should I then change it to "onrec_one_click(", or something like that?
22:11:41Paul_The_Nerdlinuxstb_, preglow: Interestingly enough, with the old way the bootloader works, the memmove16 step doesn't actually happen...
22:12:34Paul_The_NerdThe one right before/as it loads retailOS
22:14:00 Join SteL [0] (
22:15:16 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
22:18:27 Join solexx [0] (
22:20:15 Join NightCat [0] (
22:21:10 Join Thus0__ [0] (
22:23:00Paul_The_NerdMaybe the codecs don't stay loaded after boot, and it's trying to load them from the hidden partition and failing?
22:24:57 Join NicoFR [0] (
22:28:11preglowstill doesn't explain usb problems
22:28:15preglowand sounds unlike too :P
22:28:20preglowunlikely, yes
22:28:29 Quit linuxstb_ (Read error: 110 (Connection timed out))
22:28:32 Join linuxstb__ [0] (
22:28:51preglowlinuxstb__: didn't slowcoder do this kind of thing too?
22:29:29Paul_The_NerdUnlikely, I agree.
22:32:30Paul_The_NerdWell, I'm eating lunch now, then I guess I'll poke some more.
22:33:55 Join goa [0] (
22:35:09 Quit Rondom ("I'm leaving on a jetplane, don't know when I'll be back again...")
22:35:36RotAtoRanyone with an h300, ipod, recorder, or ondio care to test an updated version of bejeweled?
22:36:22RotAtoRit's been updated for the new bitmap build system and there are (hopefully) some speed fixes for color lcd targets
22:36:25Paul_The_NerdRotAtoR: Nano here.
22:36:48Paul_The_NerdAlso, H120, though I notice that wasn't in your list.
22:36:49Papricah300 here
22:36:53RotAtoRok, patch and bitmaps available here:
22:37:04 Quit paugh ("Leaving")
22:37:07RotAtoRPaul_The_Nerd: i have a h140 that i've tested it on
22:37:21preglowRotAtoR: what's new?
22:37:26RotAtoRi can quick compile a .rock for you too, if that's easier
22:37:37 Join seffy [0] (
22:37:59RotAtoRpreglow: it's been updated for the new bitmap build system and there are (hopefully) some speed fixes for color lcd targets
22:38:00 Quit Thus0 (Read error: 113 (No route to host))
22:38:00 Quit seffy (Client Quit)
22:38:41Paul_The_NerdRotAtoR: Okay, I'll let you know how it goes.
22:38:53RotAtoRthanks :)
22:39:56preglowRotAtoR: what about transparency? :>
22:40:25*Paprica compiling
22:40:45 Join seffy [0] (
22:40:48RotAtoR:) hehe, you want to add the functionality into the graphics api or where ever for me?
22:41:18RotAtoRit would be much nicer to have transparency though...
22:41:18Paul_The_NerdRotAtoR: We got errors..
22:41:30RotAtoRwhat kind of errors?
22:43:06seffyfew weeks ago i download Rockbox ...and any time i update my version once to the Official and once to the Optimized and any time the battery life seems to be not normal...and in the iriver firmwire my bat life is ok
22:43:14seffyi really like RB
22:43:19RotAtoRPaul_The_Nerd: did you try a "make clean" first?
22:43:30Paul_The_NerdRotAtoR: Yessir.
22:43:46RotAtoRi think i've had that error before from an uncleaned build directory
22:43:54Paul_The_NerdLemme rm -rf it and try again
22:44:04preglowseffy: not normal? how?
22:44:21preglowseffy: you mean like iriver firmware and rockbox doesn't give the same battery reading?
22:44:33RotAtoRi have it compiling for nano right now, i can quick put up the .rock
22:44:57seffyyeah...its look like...the bat life of the iriver fd is longer
22:45:05preglowseffy: no surprises
22:45:19RotAtoRit seems to compile fine here
22:45:21preglowseffy: it's a work in progress, you'll have to live with it
22:45:47Paul_The_NerdGive me just a second RotAtoR, we'll see if it just needed a extra-clean build dir
22:45:47 Part Philip_0729
22:46:08seffyok 10x...but how much the bat life shuold last in rb ?
22:46:13 Join Thus0 [0] (
22:46:24seffycuz in the iriver fd it last to me...about 14h
22:46:43Paul_The_NerdRotAtoR: It seems make clean just didn't get everything
22:47:16RotAtoRworking now?
22:47:34Paul_The_NerdCopying it to my ipod now
22:47:40RotAtoRgreat :)
22:48:11Paul_The_NerdOoh, the background is white now.
22:48:12 Quit Luthion ()
22:48:23preglowseffy: like i said, it's a work in progress, it's impossible for us to know how long a battery will last you at all times
22:48:42preglowseffy: rob2222 just did a battery test with a better battery than the default one, and he got 12.5 hours
22:48:48Paul_The_NerdRotAtoR: Everything seems to work just fine.
22:48:52ghode|afkdoes anyone know the command in dos to dump the contents of a dir to a text file?
22:48:55preglowseffy: so battery time is pretty lousy right now
22:48:59seffyi just ask in prectage...cuz in the iriver it can last just 8/7
22:49:05RotAtoRPaul_The_Nerd: how is the speed?
22:49:07 Join stamppot [0] (
22:49:07Bagderghode|afk: dir > file
22:49:07preglowghode|afk: dir > my.txt ?
22:49:17ghode|afkah thanks
22:49:21ghode|afki think
22:49:29Paul_The_NerdRotAtoR: Not noticeably different, but on the Nano it was quite fast.
22:49:39seffyand 24bit wps take more bat power ?
22:49:57Bagderseffy: no
22:50:02Paul_The_NerdAlso, are the keymappings for ipods slightly different?
22:50:14RotAtoRPaul_The_Nerd: because before it would noticably change as the cpu boosted during music playback
22:50:16Bagderand they aren't 24bit
22:50:31Paul_The_NerdRotAtoR: Well, I can't yet test it with music playback. ;-)
22:50:48amiconnRotAtoR: Why did you put single #if conditions in parentheses?
22:51:00 Join Thus0_ [0] (
22:51:01RotAtoRumm, where?
22:51:03seffywhy they are not 24bit ?! wrote in the optimized version..
22:51:13Bagderseffy: that's just isn't true
22:51:16cooball ipods have the same buttons, but in a different combo for 1->3->4g
22:51:30amiconnRotAtoR: #elif (CONFIG_KEYPAD == IRIVER_IFP7XX_PAD) , #elif (CONFIG_KEYPAD == IAUDIO_X5_PAD)
22:51:30Bagderseffy: the bmp files use 24 bit colors, but the screen uses 16bits
22:51:48Rob2222preglow: but my better battery is bad. i got with my old stock 9,5h and my friend with new stock 10,5. i think about if the battery_bench plugin makes some problems at the H3x0
22:51:53seffyok...and no one can "improve" it ?
22:51:59RotAtoRamiconn: hmm, ok, i'm not sure why it's like that :)
22:52:01*amiconn finally managed to understand how sw raid and lvm work together in linux
22:52:10Bagderseffy: it _can_ use 18bit but we don't use that mode
22:52:25 Quit SteL (Remote closed the connection)
22:52:27seffyits really make diffrent?
22:52:33Bagdernot really
22:52:34amiconnRotAtoR: I can test on archos and H3x0
22:52:44amiconn(you already tested on H1x0 so...)
22:52:47RotAtoRamiconn: great, thanks the picrures looks so good in the iriver fd(on Picture mode) ?
22:53:24Bagderhow can you shorten firmware to "fd" ?
22:53:50preglowseffy: that's the magic of colour lcds
22:53:56RotAtoRare the ifp7xx and x5 ports in any state where I should be looking to have bejeweled fully support them?
22:54:06linuxstb__RotAtoR: My intention was that mono bitmaps should also go in the "native" directory. The mono directory is for bitmaps that are mono on all displays.
22:54:13BagderRotAtoR: the x5 sim builds and runs fine
22:54:14preglowRotAtoR: not urgent, if you ask me
22:54:24Bagderbut the port is still far away
22:54:25Paul_The_NerdBagder: The x5sim doesn't build here.
22:54:41linuxstb__RotAtoR: This keeps all the "same" bitmaps together.
22:54:48BagderPaul_The_Nerd: no?
22:54:53Paul_The_NerdWell, it didn't two days ago
22:55:01BagderI fixed it about 2 days ago
22:55:10Bagderjust after the bmp build commit
22:55:11Paul_The_NerdJust tried again with a new build dir
22:55:12RotAtoRlinuxstb__: ok, so will the mono bitmaps be converted as mono then?
22:55:18RotAtoRand not native?
22:55:20Paul_The_NerdDoes it require that I build it in SDL mode?
22:55:31amiconnRotAtoR: Did you try 4-grey bitmaps on H1x0?
22:55:36linuxstb__RotAtoR: "native" and "mono" means the same thing for Archos.
22:55:39BagderPaul_The_Nerd: I only tried sdl mode since I'm on linux and the x11 doesn't do colors
22:55:51RotAtoRlinuxstb__: but not on the h1xx, i use mono on them
22:55:52PapricaRotAtoR, work good with the h300
22:56:04Paul_The_NerdBagder: Ah, well the Win32 version won't build. And SDL sims still don't build in cygwin as far as I can tell
22:56:32Bagderwell, I hope someone fixes those issues one of these days
22:56:43Paul_The_NerdSo, it's not so much that the x5 sim doesn't build, is that I cannot reach the conditions under which it will build.
22:56:54RotAtoRright now i'm planning on using mono on all non-color lcds so I don't have to have the ugliness of the jewel backgrounds overlapping eachother
22:57:01seffyBagder: sorry i want to say fw*....
22:57:31preglowlinuxstb__: do we ever use current_tick in the bootloader?
22:57:37preglowlinuxstb__: just wondering why we need to fake it
22:57:40linuxstb__Yes, but it's the "fake one".
22:57:50linuxstb__The ATA driver uses it.
22:58:13amiconnRotAtoR: Real alpha blending is most probably slowdown hell, but masking should be possible
22:58:19NightCatWhy make rebuild bitmaps only after make clean?
22:58:32RotAtoRamiconn: yes, masking is all I'd really want
22:58:35 Quit manhattan ()
22:58:46*amiconn is planning a function that takes both a mono and a native bitmap
22:58:49Rob2222Does anyone have an idea, why at least me has such a big battery difference between RB and IRIVER?
22:58:54 Nick San is now known as [San] (
22:59:04RotAtoRamiconn: nice :)
22:59:23amiconnMasking is basically drawing a mono bitmap just with the native bitmap being the 'foreground colour'
22:59:24preglowRotAtoR: btw, does bejeweled speed depend on cpu clock now?
22:59:53RotAtoRpreglow: right now i have it boosting during animations so they're consistent
23:00:28preglowRotAtoR: wouldn't it be better to use a max fps and frame skipping?
23:00:30amiconnThat won't happen before my optimisation work though
23:00:37NightCat* is asking some questions but nobody are listening him... All are too busy...
23:00:43preglowa lot of rockbox depends on cpu clock
23:00:53***Saving seen data "./dancer.seen"
23:00:56RotAtoRpreglow: hmm, i suppose that would be better
23:01:08 Quit Thus0__ (No route to host)
23:01:23preglowRotAtoR: well, it is the way most games work, heh
23:01:28preglowand it isn't too complex
23:02:13 Quit seffy (Broken pipe)
23:02:17RotAtoRpreglow: i'll look into that then
23:03:53linuxstb__NightCat: It's a bug (my bug) in the build system. When new bitmaps are added, the bitmap libraries are not rebuilt. But I'm not sure why.
23:04:06 Join ansiVirus [0] (
23:04:16amiconnlinuxstb__: The bitmap archives don't get rebuilt when more bitmaps are added
23:04:32NightCatlinuxstb__: ok.
23:04:33*amiconn feels like an echo :/
23:04:35linuxstb__I just said that :)
23:04:53 Quit thegeek (Read error: 113 (No route to host))
23:04:59NightCatamiconn: It does not rebuilt at all...
23:05:07NightCatonly with make clean...
23:05:16Bagderhey linuxstb, it doesn't get rebuilt when you add new bmps
23:05:20*Bagder ducks
23:05:38linuxstb__What doesn't?
23:06:21NightCatlinuxstb__: I'm talking about graphik libraries...
23:06:27linuxstb__What about them?
23:06:58NightCat? You just said: It's a bug (my bug) in the build system. When new bitmaps are added, the bitmap libraries are not rebuilt. But I'm not sure why.
23:07:08 Quit Thus0 (Connection timed out)
23:09:13BagderI see why
23:09:53amiconnRotAtoR: When building for archos recorder:
23:09:54amiconnMAKE in plugins/bitmaps/mono
23:09:54amiconnmake[3]: *** No rule to make target `bejeweled_jewels.112x64x1.bmp', needed by `/home/jens/rb-patched/build/recorder/apps/plugins/bitmaps/mono/bejeweled_jewels.112x64x1.c'. Stop.
23:09:54DBUGEnqueued KICK amiconn
23:09:54amiconnmake[2]: *** [/home/jens/rb-patched/build/recorder/libpluginbitmapsmono.a] Error 2
23:09:54amiconnmake[1]: *** [rocks] Error 2
23:09:54***Alert Mode level 1
23:09:54amiconnmake: *** [all] Error 2
23:10:16amiconnThat's after rm -r * and reconfiguring...
23:10:19linuxstb__Bagder: I think I know why, but I'm not sure how to fix it. Can you?
23:10:35NightCatI know some more bugs...
23:10:39BagderI'm trying to think of the best fix
23:10:58RotAtoRamiconn: hmm, i'll check it out
23:11:07Rob2222Does ABR/VBR Mp3 files need more power/boost to decode?
23:11:13amiconnRotAtoR: Nm, me silly
23:11:22RotAtoRamiconn: ok :)
23:11:27Bagderlinuxstb__: I'll work on a fix
23:11:34amiconnI didn't even manage to unzip correctly :(
23:11:46linuxstb__Thanks. Don't forget there is the same bug in apps/Makefile as well.
23:11:55NightCatWhen I rebuilt all to sim (win for iriver 3xx) and start sudoku, it does not work...
23:12:29linuxstb__NightCat: That could be the same problem I had with the sdl sim. Let me check.
23:12:44NightCatIt works but there is no pictures displayed on the game field...
23:12:47amiconnIt's still a 'funny' error message - obviously make is trying to build the .bmp if it's not there...
23:12:50preglowlinuxstb__: so ata driver depends a lot of current_tick being updated accurately, then? :>
23:13:30NightCatlinuxstb__: Sudoku uses lcd_bitmap_part!!!
23:13:42linuxstb__I think it's just used for timeouts - so it doesn't have to be perfect I think.
23:13:43preglowjust toying around with making ipod tick timer properly configurable, and at four times the tick rate, stuff hangs at disk access
23:14:42 Join thegeek [0] (
23:14:54NightCatlinuxstb__: I tryed to use all new system (about bmps) and this function (lcd_bitmap_part) and have this problem too...
23:15:43linuxstb__NightCat: What problem are you having with the win32 sim and sudoku? I can't see anything obviously wrong in the Makefile, apart from a missing dependency (which I'm fixing now).
23:17:08NightCatlinuxstb__: Yes win32 sim + sudoku (or with my own plugin)...
23:18:15amiconnRotAtoR: Looking good on archos recorder
23:18:39NightCatlinuxstb__: When uses function loke that: lcd_bitmap_part(src,12,0,0,0,0,12,12) it works like lcd_bitmap_part(src,0,0,0,0,0,12,12)
23:19:36RotAtoRamiconn: great
23:19:55***Alert Mode OFF
23:20:41 Join Thus0 [0] (
23:20:59 Join gantrixx [0] (
23:21:23linuxstb__NightCat: Are you saying sudoku is doing something wrong, or the win32 sim is?
23:21:29gantrixxso we have some savy media player guys here?
23:22:03gantrixxwithout installing new software, which media player would you recommend that supports ogg and m3u and has a good menuing system?
23:22:10NightCatI think function lcd_bitmap_part works wrong on sim and now I'm trying it on DAP...
23:22:35gantrixxI currently have an MPIO HD300 and it's a piece of crap
23:22:53gantrixxthe ipod is great, but it doesn't do ogg or m3u
23:23:13Bagdergantrixx: anything that Rockbox will do ;-)
23:23:17Bagderthat runs
23:23:29linuxstb__Well, apart from all the Archos players....
23:23:38Bagderoh, right ;-)
23:23:43amiconnGah! :(
23:24:02amiconngdm crashes when trying to configure anything
23:24:29 Quit Thus0_ (Read error: 113 (No route to host))
23:24:54preglowamiconn: anything special about your setup?
23:24:59lostlogicgantrixx: iaudio x5 or toshiba gigabeat most likely
23:25:17Paul_The_Nerdgantrixx: Why "without installing new software?"
23:25:44gantrixxbecause if my neice has a problem with it, I won't be available to troubleshoot it for her
23:25:45 Quit BHSPitMonkey (Success)
23:25:59gantrixxand I'd like something that I don't have to fool with too much either
23:26:17 Join g33 [0] (n=g33@
23:26:19Paul_The_NerdI second lostlogic's suggestions then
23:26:22 Quit NicoFR ()
23:26:23gantrixxare there any video media players that support ogg-theora?
23:26:37g33are there any known problems with playing certain mp3 files?
23:26:48g33ive got an album that just wont play in rockbox
23:27:02g33but plays fine in vanilla iriver
23:27:02amiconnLogging in to gnome freezes it as well. Have to hard reset
23:27:04gantrixxa 3rd preference would be one that I can connect to my car stereo
23:27:13BHSPitLappygantrixx: so, any?
23:27:53gantrixxipods are the only ones that the car stereo companies are making so that you can interface to them
23:28:12preglowamiconn: gfx card?
23:28:30Bagdergantrixx: use a line-in or cassette adapter in the car
23:28:32BHSPitLappygantrixx: well, I listen to mine in the car
23:28:35BHSPitLappywith a casette adapter
23:28:40BHSPitLappyin the headphone jack...
23:28:47Bagderso do I
23:28:50gantrixxI don't have cassette or inline
23:28:52BHSPitLappythat I got from charity :P
23:29:02gantrixxand besides that doesn't really keep my media player charged
23:29:02BHSPitLappygantrixx: no cassette deck?
23:29:15gantrixxcassette deck? what is that?
23:29:21gantrixxis that like an 8 track?
23:29:21amiconnpreglow: What could be special with this setup: HP mb, Athlon64 3800+ (no dualcore), 1GB RAM, ATI chipset w/ integrated gfx, SATA controller, 2x 160GB configured as raid1 with lvm
23:29:25BHSPitLappyare you like, 9?
23:29:27Paul_The_Nerdg33: As far as I know, there aren't any currently known issues with MP3 playback. What's happening exactly, and H1x0 or H3x0? Also, tell me about the file itself (metadata, bitrate, etc)
23:30:19g33i think its been encoded with r3mix settings in lame
23:30:41 Join TCK- [0] (
23:31:14gantrixxare there any media players out there that can connect with the ipod connector cable?
23:31:18g33its funny cause it seems to be vbr, but it changes bitrate quite infrequently
23:31:31g33Paul_The_Nerd want me to send you the song?
23:32:05Paul_The_Nerdg33: Doing a quick search, r3mix sounds like it has its own issues
23:32:06g33oh, i havent got the latest rockbox, mine is from mid december if that makes a difference
23:32:10 Join webguest50 [0] (
23:33:16NightCatlinuxstb__: !!!
23:33:25preglowamiconn: the ati chipset, i guess
23:33:29NightCatlinuxstb__: The same problem on DAP not only SIM
23:33:30preglowamiconn: i'd try something else than debian
23:33:51linuxstb__NightCat: So it's a sudoku bug? Are you on the h300?
23:34:00NightCatlinuxstb__: Then it's something with function lcd_bitmap_part...
23:34:03g33does rockbox use more battery than regular firmware?
23:34:26NightCatlinuxstb__: I'm on H300, and it's not sudoku...
23:34:39g33theres a sudoku game???
23:35:01Paul_The_Nerdg33: At the moment, it may use more battery than the stock firmware.
23:35:13gantrixxthe gigabeat does ogg?
23:35:16preglowamiconn: what debian branch, btw?
23:35:39NightCatlinuxstb__: It's a lcd_bitmap_part function works unproperly...
23:36:17NightCatlcd_bitmap_part from plugin.h
23:36:59preglowamiconn: what drivers, btw? opensource?
23:37:29amiconnI just installed debian, so it uses whatever driver is provided
23:37:37preglowopensource, then
23:38:14preglowwell, like i said, i'd try something else than debian
23:38:29preglowi have few fond memories of my trying to use debian
23:38:37amiconnI could try the HP-supplied linux variant, but I don't really want that
23:38:51amiconn(1) It's SuSE (2) It's 32bit
23:39:12preglowwell, you already know what i'd like you to try, heh
23:39:24preglowrunning 32 bit on a 64 bit machine would be a bit of a waste, yes
23:39:26amiconnI really want to try 64bit (not for some magic speedup)
23:39:32preglowno, sure
23:39:42amiconnI'm downloading ubuntu 64bit now, let's see...
23:39:57preglowworked like a breeze on my athlon64
23:40:03NightCatlinuxstb__: What the matter?
23:40:07preglowand again on this 32 bit xeon with some weird hardware
23:40:38NightCatlinuxstb__: I'm looking on
23:41:19preglowbah, i should have kept in mind that build times are long at this time...
23:41:44 Join Bger [0] (n=Bager@
23:41:56elinenbepreglow: what target was that commit for?
23:42:20NightCatlinuxstb__: It seems like sudoku uses lcd_bitmap_part function since the time when new bmp functions was improved...
23:42:33Bagderlinuxstb__: any opinion on this: ?
23:43:05preglowelinenbe: ipod
23:43:13preglowdoesn't it say?
23:43:33Paul_The_NerdIt says PP5020, I think.
23:43:39preglowyeah, which is ipod
23:43:46 Join webguest85 [0] (
23:43:47CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:43:47*preglow wonders if we'll have a h10 port
23:43:58preglowthe number of ports is starting to get dangerously high
23:44:02Bagderthere seems to be some interested guys gathering at least
23:44:18preglowBagder: i had a luck at it some time ago, i think i could recode the decryption routine in c
23:44:32 Quit webguest85 (Client Quit)
23:44:42preglowthat part of the firmware didn't seem to be encrypted
23:44:47preglowthe rest of it was either encrypted, or thumb
23:44:57Paul_The_Nerdpreglow: Define "dangerously"
23:45:31 Join Thus0_ [0] (
23:45:32preglowPaul_The_Nerd: as in we'll break our poor build box soon :/
23:45:59preglowit's struggling as it is
23:46:12NightCatlinuxstb__: Are you here?
23:46:19NightCatlinuxstb__: ;)
23:46:22Bagderone of these days we buy a new box to run the builds on
23:46:46Paul_The_Nerdpreglow: Ah, yes. That means it's time for a donation drive.
23:46:47preglowlets do one of those wikipedia style fund drives
23:46:52preglowPaul_The_Nerd: exactly :PP
23:47:01Paul_The_Nerd"Guys, we'd REALLY love to make an H10 port, but our build box just can't take it."
23:47:09Paul_The_NerdThat is what should be on the front page of MisticRiver, right now.
23:47:45preglowsomeone take a photo of it, and photoshop a tongue sticking out of the floppy drive, crosses for eyes, and so on
23:47:58Paul_The_NerdAnd smoke out the ears.
23:48:04preglowit's got ears? :P
23:48:10NightCatAnybody works with lcd_bitmap_part anywhere???
23:48:26 Quit Thus0 (No route to host)
23:48:29Paul_The_NerdWell, it has eyes?
23:49:01preglowi hope it's got a hat and pants as well
23:49:03preglowperhaps a little tie
23:49:26preglowyou never know with those crazy swedes
23:50:21Bagderbut hey, those are mighty nice pants!
23:50:52NightCatWho works on SUDOKU???
23:50:57preglowNightCat: linuxstb coded it
23:51:07BagderNightCat: feel free to fix bugs
23:51:24linuxstb__NightCat: I'm back....
23:51:26Paul_The_NerdBut he seems to be unresponsive at the moment. I'm sure he's seen your messages, and repeating them won't make him reappear any sooner.
23:51:37Paul_The_NerdSure, make me a liar.
23:51:40NightCatlinuxstb__: Ok :)
23:51:59 Quit Thus0_ ("Leaving")
23:52:04NightCatlinuxstb__: Please read what I wrote to you.
23:52:22Bagderno, try my match instead
23:53:08linuxstb__Bagder: Does your bitmap patch work? i.e. does it fix the dependency problems? I see any difference in the logic.
23:53:17linuxstb__NightCat: I'm reading your messages now....
23:53:28BagderI believe it fixes the problems, yes
23:53:34NightCatlinuxstb__: OK :)_
23:53:44Bagderby making sure it runs make in all bitmap dirs at all times
23:54:53linuxstb__Is that because build-bitmap is a phony target?
23:55:08linuxstb__I understand then.
23:55:30NightCatlinuxstb__: Please, let me know when you solve the lcd_bitmap_part problem (cause I use it too)... OK?
23:55:37linuxstb__I think there's a typo in the start of your patch though - you've written "build-" instead of build-bitmaps.
23:55:55linuxstb__NightCat: I'm not sure what your problem is. Sudoku works fine for me.
23:56:08NightCatOn wich target?
23:56:27linuxstb__ipod - which uses the same lcd code as the h300
23:56:57linuxstb__Are you saying Sudoku doesn't work on the h300?
23:57:58NightCatlinuxstb__: Yes and on sim...
23:58:04linuxstb__No, you're right. Sudoku's broken... My bad.
23:58:51NightCatYes, I knew it I knew it :)

Previous day | Next day