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 2006-04-29

00:00:59 Join BobMarley [0] (n=BobMarle@
00:01:24BobMarleywill there ever be a rockbox for Sony MP3 players?
00:01:42Paul_The_NerdBobMarley: If someone who owns a sony mp3 player decides to port it.
00:02:00BobMarleyif only i knew how to do that :/
00:02:21BobMarleyi have one and the firmware + sonic stage sucks beyond belief
00:02:46Paul_The_NerdMost original firmwares suck compared to Rockbox.
00:02:51Paul_The_NerdHeck, most suck standing on their own.
00:03:19BobMarleyhas anyone ever tried to port it to a sony?
00:04:24Paul_The_NerdNot and reported back to us about it, I believe
00:04:39Paul_The_NerdBasically, almost all work anyone has done on any player (at least, that they've told us about) can be found here:
00:05:01Paul_The_NerdAnd information that may be helpful in getting a port started is here:
00:05:39BobMarleyis it possible it just won't work on a sony?
00:06:11Paul_The_NerdWell there's no "Just will" or "Just won't" about it.
00:06:22Mikachunot all players have updatable firmware
00:06:25Mikachuit won't work on those
00:06:28BagderBobMarley: "it depends"
00:06:48BobMarleymy sony does have updatable firmware
00:06:49Bagderit can still work
00:06:54Bagder1) buffer overflows
00:06:59Bagder2) chip replacements
00:07:19amiconnThe gmini port used a buffer overflow afaik...
00:07:19Bagderjust a lot harder
00:07:34BobMarleyi don't think there's any way i can try to port it
00:07:34Mikachuso everytime you boot it you have to run the exploit first?
00:07:50BagderMikachu: well, during development at least
00:08:02BobMarleywhat would one need to know to try to port it to a different player?
00:08:31MikachuBobMarley: that's all detailed in the page you got earlier
00:08:37BobMarleyoh yeah
00:08:45BobMarleyi'm definately not up to the task :(
00:08:46Bagderwhen you need further details, come back and ask
00:08:54BagderI think it'll take you a few days ;-)
00:08:57 Quit chi (Remote closed the connection)
00:09:22 Quit hotwire (Connection timed out)
00:09:26 Join _Lucretia_ [0] (
00:09:49Paul_The_NerdBobMarley: If there's a forum somewhere of fans of the player you're interested in, you could bring up the topic there. Maybe there are others more skilled who would be interested on working on it, if you don't feel up to it. :)
00:09:49BobMarleyi could do the PCB scans i suppose
00:10:20 Join hotwire_ [0] (
00:10:22 Nick hotwire_ is now known as hotwire (
00:10:57BobMarleyif it was just a case of reverse engineering the existing firmware then maybe i could do it
00:11:29BobMarleybut reverse engineering is so boring...
00:12:51 Join chi [0] (n=chi@tor/session/external/x-a44683eb612b3623)
00:12:53*ze checks out that page
00:13:12amiconnHmm, now on to another search strategy
00:13:21Bagder"just a case" of RE
00:13:35Bagderits a massive case of RE if that's all you have
00:13:54*amiconn decided to replace all firmware/ .o files by -Os - compiled versions one-by-one, and check where it breaks
00:14:38amiconnAll app/ modules compiled with -Os work as long as the firmware/ modules are compiled with -O (or -Os but gcc 3.3)
00:15:36 Part ep0ch_ ("Kopete 0.11.1 :")
00:16:39 Quit DarkJesus ("Leaving")
00:17:10 Quit PaulJam (".")
00:18:23 Join JoeBorn [0] (
00:18:29 Quit TCK (Read error: 110 (Connection timed out))
00:18:45Paul_The_NerdWow, I would've expected the open bugs to outnumber the open patches for some reason
00:19:04Bagder254 is a crazy amount of patches
00:19:11Mikachuisn't feature requests the winner?
00:19:17preglowsweet lord, thsoe open * lists are long
00:20:25 Join PaulJam [0] (
00:20:34*amiconn silly
00:20:35Mikachui think some of mine are still open
00:21:36amiconnMikachu: Btw, gcc has an option to do what you did automatically: -finline-functions-called-once
00:21:41BobMarleyi have done reverse engineering, but i've never done any low level programming
00:22:02Bagderreverse engineering that wasn't low level?
00:22:25Mikachuamiconn: ah, it's not in 3.4
00:22:26Bagderamiconn: neat little bugger!
00:22:43Mikachuamiconn: and that's the man page in /usr here
00:22:44BobMarleywell, what i mean is, i know how to look for certain things when reverse engineering, but i couldn't actually write any ASM
00:23:06BagderBobMarley: a Rockbox port take very little asm, it is 99% plain C
00:23:40amiconnBagder: Are you saying I'm only doing 1% of the work? ;)
00:23:48*amiconn likes asm coding
00:23:52BobMarleybut the foundations will have to be ASM
00:23:57Bagderyou must WORK HARDER
00:23:59*preglow adores asm coding
00:24:10BagderBobMarley: no
00:24:16amiconnasm has the great advantage to avoid the deficiencies of gcc
00:24:22 Join macdonalder [0] (
00:24:36 Quit DpK|Cronic (Read error: 104 (Connection reset by peer))
00:24:43BagderBobMarley: asm is almost exclusively used for performance reasons
00:26:08Bagderand a port doesn't have to run very fast in the first phase
00:26:26Bagderbut I'll run fast now
00:26:28 Join Alcyone [0] (
00:26:29Bagderto bed!
00:26:34 Join stevenm [0] (
00:26:35Alcyonehi all
00:26:39stevenmHey guys
00:26:54amiconnBagder: wazdat?
00:27:08Alcyonei have seen a log with an clone of me .... an another Alcyone .... funny !
00:27:21Alcyonehe was connected here
00:27:24Mikachuamiconn: is build/Makefile the only place that sets CFLAGS for apps/ ?
00:27:43stevenmlong time havent been here, trying not to fail my classes. i read that someone got the midi player to run fast. very cool guys
00:27:47 Part midgey34
00:27:54amiconnIt sets the flags for all core sources. Plugins and codecs define their own settings
00:28:01 Join DpK|Cronic [0] (
00:28:02Mikachuif so, i didn't get any difference in rockbox.ipod
00:28:09Mikachumaybe i already caught all the functions then
00:28:23amiconnHmm, I presume you did a full rebuild?
00:28:39Alcyonebye all
00:28:42Mikachuls -l rockbox.ipod; (make clean;make)>&/dev/null; ls -l rockbox.ipod
00:28:44 Part Alcyone ("Time makes no sense")
00:29:01 Quit hotwire (Connection timed out)
00:29:14Mikachubut the option doesn't exist for 3.x it seems
00:29:23amiconnMikachu: How do you know make didn't error out?
00:29:37 Join rotator [0] (n=e@rockbox/developer/rotator)
00:29:39Mikachuamiconn: i wouldn't have a rockbox.ipod then
00:30:21 Join hotwire_ [0] (
00:30:23 Nick hotwire_ is now known as hotwire (
00:31:32 Quit Acksaw ("I'm off, see ya later!")
00:32:51 Quit ender` (" Never trust a computer you can't throw out a window")
00:35:00 Part BobMarley ("Leaving")
00:36:25 Join _chang [0] (
00:37:00 Nick _chang is now known as chang (
00:41:26 Join _Lucretia [0] (
00:42:08 Join grapefruit [0] (
00:42:44grapefruitcan you still play video on the ipod 5th gen with rockbox firmware?
00:44:07Paul_The_NerdNot *in* Rockbox
00:44:16Paul_The_NerdBut it's dual boot, so you can just boot into the Apple firmware to play video
00:44:33grapefruithow would you switch between firmwares?
00:45:05grapefruitcould I just turn off my ipod and back on?
00:45:16Paul_The_NerdYou hold menu while booting. These things are covered in the FAQ, and possibly even the Manual...
00:45:22*dwihno comes to think of a "build a ipod database from here"-plugin for just video... :(
00:46:45dwihnoOnce I get some spare time, I'll probably end up doing that if nobody beats me to the task
00:48:08 Quit Kohlrabi ("Fast alle Menschen sind Regenwürmer")
00:51:21 Quit lee-qid_ (Read error: 110 (Connection timed out))
00:51:51 Quit SereR0KR ("XChat Aqua")
00:53:22 Quit _Lucretia_ (Read error: 110 (Connection timed out))
00:55:08 Part chang
00:58:05 Join sharpe [0] (
00:58:35sharpei figure that i have around 120000 files over the span of three drives...
01:02:57 Join romanos_melodos [0] (n=romanos_@
01:03:45romanos_melodoshave i got to register each time i'm gettin to this chanell in order to send pm's?
01:03:57 Quit stevenm ("Connection reset by beer")
01:06:11Paul_The_NerdYou only register once, but you have to identify each time you connect to the server unless you configure your client to do it for you.
01:11:15romanos_melodospaul, are u always here? :P
01:11:37 Quit hotwire (Connection timed out)
01:14:20 Join hotwire_ [0] (
01:14:23 Nick hotwire_ is now known as hotwire (
01:14:28Paul_The_Nerdromanos_melodos: Not *always*
01:15:03Paul_The_Nerdromanos_melodos: But I spend a lot of my free time online anyway doing various things, so it's not hard to sit in the channel and watch for questions to answer. It costs me a little time, and it helps people
01:19:28 Quit _Lucretia (Read error: 110 (Connection timed out))
01:20:22 Quit grapefruit ("CGI:IRC")
01:20:29 Join _Lucretia [0] (
01:28:13 Join thomari [0] (
01:28:41***Saving seen data "./dancer.seen"
01:31:34 Quit actionshrimp ("( :: NoNameScript 4.01 :: )")
01:34:47 Part thomari
01:42:49 Join scf [0] (
01:42:59scfhow to enter prefs on x5? :)
01:43:09 Quit DpK|Cronic ("get satisfied! • :: ««« (Gamers.IRC) »»» ::")
01:43:13 Part PaulJam
01:43:37Paul_The_Nerdscf: Is it not covered in the manual?
01:44:02scfsorry dont habe it
01:44:12 Join whatboutbob [0] (
01:44:36Paul_The_NerdHmm, then you might want to fix that.
01:45:29 Part Paul_The_Nerd
01:45:35scfargh it's rec button :)
01:46:06amiconnWow... the problem with gcc 3.4 / 4.x -Os on archos is ata.c - something I wouldn't have expected...
01:49:36 Quit hotwire (Success)
01:50:05 Quit _Lucretia (Read error: 110 (Connection timed out))
01:50:08*preglow wishes amiconn welcome back from the futile gcc bug hunting
01:50:20 Join hotwire_ [0] (
01:50:22 Nick hotwire_ is now known as hotwire (
01:50:50preglowdoesn't sound like you'll ever get to the bottom of the gcc troubles...
01:51:11amiconnIt seems to me that we might do some details wrong in the code
01:51:26amiconn..something like a missing 'volatile' here and there
01:51:43preglowi tried another approach for my ata.c nano hack yesterday
01:51:44preglowstill no luck
01:51:55preglowthis time i tried to just disable the soft reset code the flash gate chip doesn't seem to like
01:52:02preglowit still hangs on wps usb sometimes
01:52:38amiconnI have built for recorder v1 with gcc 4.0.3 -Os. *Most* modules compiled that way cause no problems (except system.c which I will deal with later), but ata.c -Os breaks things badly
01:53:01amiconnATA accesses ports, and ports are volatile...
01:53:50amiconngcc 3.3 -Os reduces code size by ~8K, and gives back rombox for fm recorder
01:54:07amiconnHowever, gcc 4.0 -Os reduces code size by another 6KB...
01:55:07amiconnRegarding the switch() corner case bug, I have a patch that fixes it. It's not "the right way" but it ensures correct operation
01:55:29amiconnThis patch is necessary for any sh gcc version
01:56:35whatboutbobpreglow: need me to test anything for optical recording?
01:56:53 Quit lostnihilist ("Leaving")
01:57:29amiconnWow, I can't believe it!
01:57:47amiconnata.c is indeed the only file causing the problems
01:58:11amiconnJust compiled everything with -Os except ata.c - and it's running like a charm
01:59:16preglowwhatboutbob: were it you that had the satelite receiver thing?
01:59:19preglowah, no, that was mborus
01:59:58preglowwhatboutbob: not much has happened, really, i worked some more on the sample rate detection, but that's that, i still don't understand correctly why recording hangs
02:00:12amiconnSmall wonder I got so many strange effect (rolo checksum errors and the like) - if disk access isn't working correctly...
02:00:14preglowit's almost like dma switches itself off
02:00:55whatboutbobpreglow: yeah, that was mborus w/ the sat receiver.
02:01:05 Join TCK [0] (
02:01:43preglowsuddenly PDIR2 is always full and recording stops
02:01:45whatboutbobpreglow: i made my first 'real' optical recording w/ rockbox a couple of nights ago...everything worked well.
02:01:47preglowand i have no idea how to detect it
02:02:01preglowwhatboutbob: what's cool is that i can record just fine with no cable plugged in
02:02:04preglowwhile most other people can't
02:02:08preglowdoesn't really make it easy for me to debug
02:02:17whatboutbobthat's odd.
02:02:32preglowsometimes, that is
02:02:33amiconnPerhaps there are different chip revisions of the coldfire?
02:02:34preglowsometimes not
02:02:39whatboutbobcould it be a setting thing?
02:02:43preglowamiconn: nah, i think it's something else
02:02:46preglowwhatboutbob: doubt it
02:02:54preglowwhatboutbob: like i said, it happens for me sometimes too
02:03:30preglowthe only thing i know to happen consistently is that the PDIR2 full interrupt is fired each time it happens
02:03:36amiconnMight it happen that the recording dma tries to write to an unaligned address? If that happens, you'll get a dma error
02:03:40preglowbut why the hell does that happen? dma should empty it for me
02:03:43amiconnOr do you use auto-align?
02:04:01preglowamiconn: i don't see how that can happen
02:04:34*amiconn palns to experiment with dma, auto-align & ata
02:04:46preglowamiconn: did you get that converter yet? :>
02:04:51whatboutbobthis may make no sense...but is there perhaps a reason why iriver f/w doesn't offer optical recording option unless a signal is detected?
02:05:13amiconnpreglow: Nope. Had no time to visit that shop during its opening hours
02:05:39amiconnWe have some dma channels to spare. Ata dma would be a nice thing...
02:05:39preglowwhatboutbob: hell, i'll do that too, but it still doesn't explain why some people get hangs while recording
02:05:50whatboutbobfair point.
02:06:04preglowi consider it probable it's the same problem
02:06:43preglowamiconn: i really know very little about dma
02:06:50preglowi need to study the datasheet
02:06:53whatboutbobpreglow: the failures just seem so completely random.
02:07:17preglowwhatboutbob: which is exactly why i gave the patch a break, heh
02:07:28whatboutbobcan't blame ya. :-)
02:07:43preglowthe only consistent way i have of triggering the fault is inserting and removing the cable a lot of times
02:07:48preglowthen recording will hang eventually
02:07:57preglowleave screen then reenter, and hey, it works again
02:10:08preglowanyone know what type ssize_t is?
02:10:19preglowsigned size_t ?
02:12:29 Quit whatboutbob ("CGI:IRC (Ping timeout)")
02:13:15preglowamiconn: didn't linus say ata dma wouldn't be practical some time?
02:14:07amiconnIt needs testing. I believe we can get the same speed as now with dma + auto-alignment, but yield during transfer so the cpu can do other stuff meanwhile
02:14:35amiconnThe only limitation is that not the whole iram supports dma
02:15:06amiconnOne block does not, so we have to switch to cpu transfer in this case
02:15:24amiconn(but I don't believe that current code does ata transfers to/from iram)
02:17:50 Join DarkJesus [0] (
02:18:11 Quit pixelma (" HydraIRC -> <- \o/")
02:22:02*preglow laments at the poor sim performance
02:22:08preglow200% pitch with mpc playback and it skips...
02:22:37 Join ashridah [0] (
02:22:42 Join nave7693 [0] (
02:23:46 Part nave7693
02:24:22preglowahh, i finally managed to break the resampler
02:25:50rotatorpreglow: did you see my musepack results from yesterday?
02:26:29 Join _Lucretia_ [0] (
02:26:51preglowrotator: yeah, you got a skip?
02:27:03rotatoryeah, I did manage to get a tick, but could only find it with one track
02:27:10preglowi've got an entire album of them
02:27:13rotatorneed anymore files to test with? :)
02:27:19preglowincidentally the only bloody album i've got as musepack
02:27:23rotatorhaha, nevermind then
02:27:30preglowencoded especially since it's very gapless
02:27:45rotatorvery nice :)
02:27:50preglowi'm doing some testing as we speak, to see if the gapless problem ever went away
02:28:13rotatorthe problem seems to be different from before, it ticked on far fewer tracks than before, at least as far as I can tell
02:29:00rotatormust be some caused by some musepack stream format voodoo
02:29:33preglowgod knows
02:29:36preglowi'm not too familiar with it
02:29:48preglowbut i'm fairly certain the clicks didn't use to be there with this album
02:29:58preglowthe lack of seeking makes testing a bit... tedious
02:30:07preglowespecially since the tracks are all pretty long, heh
02:30:09rotatoryes, very
02:30:09*amiconn caused the problem due to his infinite lameness....
02:30:20*preglow pats amiconn
02:30:52amiconnIt's better to not lie to gcc about what an asm() really touches...
02:31:13amiconnAs long as not optimising too heavily, all is fine.
02:31:43amiconnBut if I pretend to take a parameter as input only, and then modify it within the block...
02:32:27preglowmy patch isn't gapless either
02:32:49preglowtime to see if reverting my commit does it, then
02:34:53preglowanyone know of any direct cvs way to revert without actually reverting the file? that's caused pains in the past...
02:35:19Mikachucvs diff -r before -r after file | patch -R; cvs commit -m hello
02:35:47preglowahh, without the commit part
02:35:48preglowi'll try
02:37:08preglowcan i just concat patches?
02:37:44Mikachui think so
02:40:53preglowMikachu: worked like a charm, thanks
02:41:59Mikachuglad to help
02:44:09preglowin which case i've also got a bunch of dishes that need doing
02:44:10preglowfeel free
02:44:55preglowi didn't break gapless
02:44:56preglowall is fine
02:44:58preglowfine like wine
02:50:07preglowi wonder how the flaming hell i avoided noticing that gapless still didn't work perfectly
02:50:25Mikachutoo much beer?
02:51:26preglowunfortunately not
02:51:38preglowjust a general wish to think it was fixed, i guess
02:52:35 Join nave7693 [0] (
02:52:54 Part nave7693
03:07:38goffathese phones rule... shure e5c...
03:09:25preglowcould someone please just kill me now, before i have to try using audacity again
03:09:37 Quit muesli__ (Read error: 110 (Connection timed out))
03:10:03*goffa puts preglow out of his misery
03:10:10*goffa thinks
03:10:22*goffa resurects preglow... cracks whip
03:10:37goffamusepack :)
03:11:23 Join dj-fu [0] (
03:12:04preglowgreat, now i'm a zombie and STILL have to use audacity
03:13:45*rotator condemns preglow forever to the hell that is linux multimedia support!
03:13:49goffaAudacity, Inc. provides janitorial services to homes and businesses in the Seattle, WA metro area.
03:14:22preglowrockbox output is bit identical to what libmpcdec outputs on a pc
03:14:34preglowexcept the first frame
03:14:49rotatorand it still ticks?
03:15:43preglowi think perhaps this might be the tick sourcd
03:15:55preglowthe first half of the frame is not identical
03:16:41*preglow wonders how to concat two sound files in linux...
03:17:16*preglow cuddles sox
03:17:42ashridaharghl. for a second there i thought that said * preglow cuddles sex
03:18:06 Quit romanos_melodos ()
03:20:16preglowalways with the interesting interpratations
03:27:50*goffa regrets going on that bike ride... allergies
03:28:14preglowrotator: yep, i pretty much confirmed it now, that's the gapless problem
03:28:39rotatorcool! so do you know you to fix it? :)
03:28:45***Saving seen data "./dancer.seen"
03:28:48rotator*know how
03:29:36preglowknow how and know how
03:29:38preglowi know where, heh
03:30:26HotfusionI plan on ordering some better phones someday. default x5l ones are ok but hurt years and have a low threshold for loudness
03:31:24goffaHotfusion: try er4s or e5c... those rule... and are comfortable...
03:31:54preglowi think i've got the bug now
03:32:05goffai have a pair of e5c... probably going to ebay the e4rs... but they sound good
03:32:06Hotfusionyeah, or sennheier PX-100 or on of the $50 Audio Technica ones that clip on the ear
03:32:18*rotator breaks out the champagne
03:32:34 Join jbauman [0] (
03:32:44preglowas usual, it's probably my bloody fault
03:33:10goffayeah.. just depends on the budget
03:33:38*goffa is thinking of heading to the audiologist and getting some molds
03:33:46rotatorgoffa: go for the UE10s ;)
03:33:54goffahe he he... too late
03:34:00rotatoronly $900!
03:34:37goffai thought the $300 or whatever was rediculous... but i did it anyway because i've got an addiction
03:34:47goffai just have to promise myself i'll never listen to the UE
03:34:55goffaor i'll probably have to shell out the cash
03:35:16Galoisyou can't anyway, they're all custom fitted
03:35:37goffathat's probabyl where the majority of the cost is then
03:35:47goffaor do you pay to get fitted?
03:35:47 Quit scott666 (Read error: 104 (Connection reset by peer))
03:35:49 Join scott666 [0] (
03:35:56Galoisyou pay to get fitted!
03:36:01goffahe he he.. nm then
03:36:04rotatorI know what you mean, it is an addiction, I'm planning to order some Beyer dt990s soon :>
03:36:04Galoisthe low end UEs are only $500
03:36:07Hotfusionwhat do you send them your ear then they fit it
03:36:10preglowoooh, beyers
03:36:13rotatorI already have the e3c for portable use
03:36:25goffavan gough it to them
03:36:38rotatorand I've promised myself I won't go for the e4 or e5
03:36:55goffalol... don't listen to them :)
03:36:57rotatorthis music hobby can be far too expensive
03:37:22Hotfusionwhy get a player and that's really good but crappy phones
03:37:26Hotfusionno sense
03:37:36rotatorand then you need a portable amp too...
03:37:36goffatrue Hotfusion
03:37:46goffayeah... not with the e5c
03:37:54rotatoryeah, probably not
03:37:55preglowif this isn't it...
03:37:58goffaholy hell they are loud
03:38:02HotfusionI mean why these comapanies at lest apply better pones if I spend $300 on a player
03:38:02preglowgod, it's fun to wait seven minutes to find out if you're right
03:38:55 Quit _Lucretia_ (Read error: 113 (No route to host))
03:38:58goffaactually... considering the crap that comes with most players i thought that the ones that came with the x5 were ok
03:39:04goffabut i didn't do any extended listening to them
03:39:09 Quit DarkJesus ("Leaving")
03:39:16goffaand my sister inherited them after about 30 mins :)
03:39:46Hotfusionwell goff if I turn my music to volume 21 the phones on my x5l begin to clip
03:39:50goffanot that they compare to my etys or my shures
03:39:52Hotfusionit's not nice
03:40:06goffayeah... i didn't crank them
03:40:16 Quit dj-fu ("Leaving")
03:40:18 Join _Lucretia_ [0] (
03:40:22goffai was in my house.. so i didn't need to
03:40:37goffahad i been in a car i probably would have come to the same opinion as you
03:42:03goffawhat i like about the shures and etys is the isolation... i remember when i flew down to san fran...
03:42:24goffai didn't realize how loud planes were
03:42:37goffai put in my etys and didn't plug them in.. .i noticed a HUGE improvement
03:42:56preglowbut, but
03:42:58preglowthat has to be the bug
03:42:59rotatorthe isolation is also very good for not realizing you're about to be hit by a car when crossing the street :>
03:43:08goffai don't think i'll fly without earplugs from now on
03:43:12goffatrue rotator
03:43:17goffai don't use them when i ride my bike
03:43:39goffaits fun putting them in and going to lala land i have to admit... you are in your own little world
03:43:59goffamusic feels like you are thinking it ... like its coming from your brain
03:44:04rotatordefinitely, off rocking out in another world :)
03:44:53Hotfusionwhile your in your world you hear *clip* *clip*....*clip*....*clip* *clip* *clip*. Better phone time
03:45:04goffaHotfusion: indeed
03:45:05Hotfusionhard not to spend to much but still get good quality than the default ones
03:45:24 Quit Una^ ()
03:45:38BHSPitLappysome portable speakers would be nice on a bike
03:45:39goffayeah... i think the 6i's are probably a good value
03:45:54goffayear.. or open ear phones
03:45:57preglowrotator: looks like i've got it
03:45:58BHSPitLappyif I biked, I'd have the planets booming from my coby's
03:45:58Hotfusionyeah just mount them on your bike
03:45:58goffaso you can hear outside noise
03:46:01Hotfusionthe ladies would like it
03:46:09rotatorpreglow: hooray! need me to test?
03:46:16BHSPitLappydon't forget a thumpin' woofer
03:46:22BHSPitLappyand hydraulics, and spinners...
03:46:23preglowjust doing a final test for the satisfaction of hearing it gapless
03:46:40BHSPitLappyI'd love to see someone with a huge sub on their bike
03:46:47goffalol... im7 riding on the shoulder reminiscant of the boom box days
03:47:19goffaBHSPitLappy: like a hsu vtf3
03:47:24goffariding in a basket
03:47:59Hotfusionyeah a Dr. Hsu would be great. Or a velodyne
03:48:01BHSPitLappyI need to take my coby's up to school on my shoulder, with my iPod
03:50:22goffaweld a couple of stands to the handle bars and mount the cobys
03:50:30goffawire a sub to a basket on the back
03:50:37Hotfusiona couple klipsch's would work
03:50:39Hotfusiontheir small
03:50:45goffahave a BoomBIKE <TM>
03:51:06goffaor would that be an iBoomBIKE <TM>
03:51:26goffayou'd have to rig the pedals to a generator
03:52:12*preglow dons the fancy pants of success
03:52:22goffayay preglow
03:53:56goffagrr... i knew i was going to put linux on this laptop
03:54:05*goffa wonders why he got an ati card
03:55:34Hotfusionnothing wrong with my x800
03:56:48ashridahgoffa: if it's a laptop, getting a geforce go-based chipset isn't too much more stable, really.
03:57:03ashridahalthough it's probably more widely supported than ati's mobile chips on linux
03:59:36Galoisit's a lose-lose proposition
03:59:47Galoisnvidia == binary-only drivers == non-free
03:59:50Galoisati == no drivers
04:00:24preglowrotator: please test, the seek patch should still apply and now works fine here
04:02:10rotatorpreglow: cool, will do
04:03:59goffawell aparently ati doesn't have a driver for xorg 7?
04:04:16Galoisyou're expected to use 6.8.1
04:04:28goffai see that
04:04:31Galoisalthough, if you're happy with 2d support only, xorg7 has those built in, I think
04:04:48goffawhat's the driver name?
04:04:53goffabecause 2d is just fine for me
04:05:29Galoisare you sure? some of the extra eye candy is actually useful
04:05:34Galoislike transparency
04:05:40Galoiscube desktop animations, etc.
04:06:08goffagood point
04:06:36preglowcube desktop animations???
04:06:49goffai was thinking transparency
04:06:54goffathats nice for console
04:07:14rotatorpreglow: still got a tiny static click on the problem track from before :>
04:07:49Galoispreglow that's when your four virtual desktops are placed on the side of a cube and the computer displays a cube rotating animation when you switch desktops
04:08:12Galoisthis is actually useful for n00bs, because it eliminates the "OMG my desktop disappeared" complaints
04:08:12*goffa could care less about that
04:08:38preglowrotator: you absolutely sure that's not the track?
04:08:41Galois has like one sentence on 2d support
04:08:46pregloweverything works just fine and dandy here now
04:09:00preglowrotator: and is that at a track transition?
04:09:04rotatorpreglow: i'll check to make sure
04:09:11rotatoryes, at track transition
04:10:05rotatorpreglow: definitely not there in foobar
04:10:46preglowrotator: and you have updated your cvs, yes?
04:10:57preglowwhat track?
04:11:15rotatori got the update to mpc_decoder.c
04:11:33rotatordream theater - vacant
04:11:36 Quit _Lucretia_ (Read error: 110 (Connection timed out))
04:11:45rotatori don't think i sent you that track
04:12:01rotatori can do that now...
04:13:44preglowi have perfect gapless even while seeking and at really challenging track transitions here now :7
04:16:09goffawonder if debians package is similar to gentoos
04:16:11goffatime to see
04:18:23goffai guess i could go back to 6 too
04:18:32preglowwhy would they be similar?
04:18:45preglowgentoo's packages are mostly based on the compiled version of the downloaded source archive
04:18:55goffai meant package names
04:22:10preglowit's getting exceedingly late
04:24:05goffa8:30 here
04:24:39preglow04:24 here...
04:26:30goffagetting early then :)
04:26:45preglownot getting light yet
04:26:47preglowbut not long
04:40:41 Quit thegeek ("( :: NoNameScript 4.02 :: )")
04:40:53 Quit TeaSea ("Leaving")
04:45:19Hotfusionwhoo draft all day tomorrow
04:49:02 Quit TCK (Read error: 110 (Connection timed out))
04:53:43 Join Jungti1234 [0] (
04:54:02 Quit Jungti1234 (Client Quit)
04:54:07 Join Jungti1234 [0] (
04:55:32 Quit Jungti1234 (Client Quit)
05:01:43 Join Rob2222 [0] (
05:06:55 Join piroko [0] (
05:07:10pirokorockbox won't boot on the 4g anymore... :(
05:07:44pirokoi think it became broken after the upgrade to libmad
05:07:50pirokobut I'm not 100% sure
05:08:05pirokoit just locks up at the rockbox screen
05:08:35scott666have you tried with a daily build from a few days ago too?
05:08:59scott666you could probably pinpoint which change borked it by just going back until you have a version that works
05:09:38pirokoi'm not sure how to download previous versions :(
05:10:18scott666click on 'old'
05:10:48pirokoahha. thanks
05:14:57pirokoyesterday's build only works if I flip hold to clear the settings... I tried clearing the settings, then shutting down and rebooting and it locks up without the hold switch...
05:15:50pirokovery strange...
05:18:45 Quit Rob2222_ (Read error: 110 (Connection timed out))
05:26:41 Join RotAtoR_ [0] (
05:28:04 Quit rotator (Nick collision from services.)
05:28:09 Nick RotAtoR_ is now known as rotator (
05:28:46***Saving seen data "./dancer.seen"
05:58:00 Quit BHSPitLappy (Read error: 110 (Connection timed out))
06:01:22 Quit piroko ("telinit 0")
06:03:13 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
06:12:06 Join zagor [0] (
06:12:25zagorhey guys, anybody awake?
06:12:34scott666whats up
06:12:54zagorwho's running the jv24 build server? it's slowing down the builds... :-)
06:15:20zagorit's doing 300+ second builds while the rest of the machines are at <100 seconds
06:16:06*zagor is idling at an airport and has nothing better to do :)
06:16:26scott666airport? where ya headed?
06:16:36scott666or are you just hanging out at an airport for fun?
06:17:50Mikachuhe's there for the free wlan
06:18:02 Join thegeek [0] (
06:18:07jncand the human emotion
06:18:10 Quit zagor ("CGI:IRC (EOF)")
06:18:13jncala dogma
06:18:30 Join zagor [0] (
06:18:49scott666haha, good reference
06:19:00zagorheh, going back in cgiirc is a bad idea :-)
06:19:20zagori'm headed for france for a few days. company trip.
06:21:49 Quit amiconn (Read error: 110 (Connection timed out))
06:21:49 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
06:22:05zagorbtw, I think we should use "boxes" as the default wps in 3.0 rather than the old text-only one
06:22:38scott666boxes doesn't work on 5G ipods
06:22:44Mikachuipods aren't in 3.0
06:22:52scott666oh :-(
06:23:26scott666sorry, im still kinda finding out where rockbox is since i left
06:23:35zagorscott666: oh, ok. I thought boxes were the only wps that was adjusted for all targets. apparently not ipod5g yet.
06:24:11scott666yeah, there are about 5 versions of it
06:24:22scott666just none in the right resolution
06:24:28zagorgotta go
06:24:40 Part zagor
06:27:52 Join gursikh [0] (
06:29:54 Join TCK [0] (
06:33:15 Quit rotator ("zzzzzzzzzz")
06:39:35 Join hotwire_ [0] (
06:42:12 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
06:57:08 Quit hotwire (Read error: 110 (Connection timed out))
07:02:35scorchequiet tonight
07:04:22*ashridah bounces around the room
07:04:25ashridahit works!
07:04:53scorcheashridah: glad you were finally able to get it up
07:05:09ashridahencrypted too :)
07:06:22ashridahall i've got left to do is spawn a timeout thread, (easy) and handle expired certificates.
07:06:29 Join powr-toc [0] (
07:24:24 Quit TCK ("well, if you say so.")
07:28:47***Saving seen data "./dancer.seen"
07:48:52 Quit arf-arf ()
07:52:12 Part gursikh
08:16:30midkayanybody here who feels strongly about per-file Viewer settings?
08:16:38midkayi doubt it and hope not, but i better make sure..
08:16:45 Quit midkay (Read error: 104 (Connection reset by peer))
08:28:15 Join ender` [0] (i=ychat@
08:34:02 Quit macdonalder ("bye bye ;D")
08:37:04 Join midkay [0] (n=midkay@rockbox/developer/midkay)
08:44:26 Quit ashridah (Read error: 110 (Connection timed out))
08:44:41 Join ashridah [0] (
08:49:21 Quit ender` (Read error: 113 (No route to host))
08:51:24 Join nave7693 [0] (
08:51:52 Part nave7693
08:53:47 Quit ashridah ("Leaving")
08:53:51 Join Acksaw [0] (
08:54:05 Join ashridah [0] (
09:05:00*scorche takes a step to the left
09:09:15 Join Kohlrabi [0] (
09:20:14 Quit scf ("Lost in disaster")
09:27:09 Join wehn_ [0] (
09:28:49***Saving seen data "./dancer.seen"
09:30:03 Join lostnihilist [0] (
09:37:35Presence Rock - Hard Rock
09:37:36Presence Indexing 666 Torrents
09:43:46midkayfar from as cool as that screenshot i took of Counter-Strike Source's Steam properties window which said that it was using 1337mb of disk space!
09:46:02scorche1627 for me =(
09:47:12midkaythis was.. months ago.
09:47:20scorchei assumedthat
09:48:15midkay !
10:10:44Presenceokay, this is gotta be a lamer question..
10:11:06Presencebut I just successfully did cygwin install + cvs of everything for rockbox-devel so I can do the UI simulator
10:11:14Presenceand I have my lovely little "archos" directory...
10:11:30Presencebut, uh, when I run the ./rockboxui.exe
10:11:33PresenceIT BOOTS PERFECTLY
10:11:41Presenceand I get right into the rockbox ui.
10:11:57Presenceexcept, THERE'S NO FUCKING .rockbox/ folder or rockbox.ipod or whatever binaries.
10:12:25Presenceits like its "magically" running.... its this normal procedure?
10:16:44 Join ProgramZeta [0] (
10:18:01Presencenevermind, I'm a tard.
10:18:02 Join aliask [0] (
10:18:04midkayPresence, is 'show files' all?
10:18:09Presencels was hiding .dot directories!
10:18:18midkaywhat about the rockbox.ipod file?
10:18:27midkayi don't think those are hidden even in supported mode.
10:18:36Presencenaw, there's no "rockbox.ipod" regardless.
10:18:52midkayah, right.
10:18:57 Join speacial_ed [0] (
10:19:15Presencebut the ".rockbox" dir was being hidden by ls cause I never exported alias ls="ls -G -F -A"
10:19:21PresenceI'm a fucking genious.
10:19:35midkayhey, genius. you spelled genius wrong. ;)
10:19:40Presencethanks for the heads up!
10:23:12 Quit speacial_ed (Client Quit)
10:23:34 Join JdGordon [0] (
10:24:17Presencethe uisimulator is great. now to draw themes, yo!
10:24:44 Join PaulJam [0] (
10:27:23 Quit damaki__ (Read error: 104 (Connection reset by peer))
10:33:36 Join ender` [0] (i=ychat@
10:41:16 Quit JoeBorn (Read error: 110 (Connection timed out))
10:45:03 Join bluey [0] (
10:45:54JdGordonhaha @ the open bugs and patches msgs in the dev list..
10:47:26 Join muesli__ [0] (n=muesli_t@
10:48:23aliaskThe patches one is too long for people to bother looking at each one, methinks.
10:49:17 Join chang [0] (
10:50:56wehn_Anyone else noticed that resume playback on startup doesn't work? renders it inoperable.
10:52:38wehn_requires a reset to escape, though not truely hung.
10:54:26amiconnYes, it's mentioned on the SoftwareCodePlayback wiki page
10:54:38amiconn* SoftwareCodecPlayback
10:54:52wehn_ah, sorry.
10:58:17 Quit chang ("So long, and thanks for all the fish !")
11:01:11*midkay proposes a source-tarball thumb to replace the rockbox logo on the daily/cvs builds pages:
11:01:49*JdGordon slaps midkay for being silly...
11:02:04JdGordon.. na just joking..
11:02:05*midkay deletes viewer.c with updates and fixes
11:02:08midkayoh. me too.
11:06:19*amiconn now has an -O2 build running on his iriver, and a gcc4 -Os build (except system.c) running on his recorder −− both stable :-)
11:07:02scorcheso what were you doing wrong?
11:08:07markunamiconn: did you have to change rockbox or gcc?
11:08:46 Quit jbauman (Read error: 110 (Connection timed out))
11:09:19 Quit HCl (Client Quit)
11:09:19midkaytwo other sizes: and - all of them are for Bagder if he pops in. :)
11:09:46 Join hcl [0] (i=hcl@rockbox/developer/HCl)
11:10:02 Nick hcl is now known as HCl (i=hcl@rockbox/developer/HCl)
11:12:19amiconnscorche, markun: Check my latest commit
11:13:05amiconnFor SH1, my gcc patch is needed as well (will publish it soon, as it will soon be required to build rockbox for archos)
11:13:53midkayamiconn, so does that get rombox under the limit for the v2/fm recorders?
11:13:55markunamiconn: is that the old patch that didn't get accepted by gcc or a new one?
11:13:56amiconnI just want to fix system.c for sh-elf-gcc 4.x before −− the weak alias problem. I now know the way to fix it :)
11:14:34amiconnmidkay: Yes, quite a bit
11:14:48midkayamiconn, coool :) time to bloat!
11:15:32amiconnsh-elf-gcc 3.3.x -Os already gives back rombox for fm/v2, and gcc 3.4.x -> gcc 4.x reduce code size further
11:15:39*amiconn slaps midkay
11:16:19midkayhm, i just got a great idea!
11:16:49midkayit's actually not that great! but still! an option to autohide the scrollbar when not scrolling!
11:16:49 Quit aliask ("no reason")
11:17:09amiconnmidkay: That's a bad idea
11:17:21midkay'not great' would be less harsh. :)
11:17:32amiconnThe scrollbar is the visual indication that there is more than one page
11:18:00amiconnI will also test gcc 4 for coldfire, and if that works well, I think we should switch to gcc 4 for coldfire too
11:18:19midkayamiconn, so archos will soon be compiled with gcc 4?
11:18:26amiconnThe advantage being that we use the same gcc version for all targets...
11:18:36amiconnmidkay: Yes
11:27:08ashridahamiconn: i seem to recall some of the codecs being 'too big' when compiled with gcc 4 on coldfire
11:27:13ashridahprobably fixable tho
11:27:51markunamiconn: we didn't switch because most of the codecs were slower with gcc 4 on coldfire. Maybe you can firgure out why.
11:28:50***Saving seen data "./dancer.seen"
11:30:32amiconnThat's actually strange. I would expect gcc 4 to produce better code than 3.x
11:31:09*amiconn has 3 different sh-elf-gcc installations in his debian vm, 3.3.6, 3.4.6 and 4.0.3
11:31:19amiconnEasily selectable by replacing a symlink
11:31:43amiconnI'll do the same for coldfire (of course only 2 different versions)
11:32:43amiconnmarkun: Was that with -O2?
11:32:56 Join SereR0KR [0] (
11:33:08amiconnI know the codecs use -O2 nowadays, but maybe the core -O level has some influence as well
11:34:45amiconnOh, the vmware development image will need updating as well
11:35:41JdGordondoes any1 know the specs of the build servers?
11:36:07amiconnThey're all different, but I think all are 2GHz+
11:36:27amiconnSome are also smp machines
11:36:48amiconnI only know the specs of mine
11:43:58 Quit midkay ("Leaving")
11:44:10 Join petur [0] (n=petur@rockbox/developer/petur)
11:45:56 Join Moos [0] (
11:54:18 Quit JdGordon (Read error: 104 (Connection reset by peer))
11:57:31 Quit bluey ("Leaving")
12:04:20amiconnashridah: You were right. Building for iriver with m68k-elf-gcc fails in Tremor:
12:04:55amiconn. /opt/m68k-4.0.3/bin/../lib/gcc/m68k-elf/4.0.3/../../../../m68k-elf/bin/ld: region PLUGIN_IRAM is full (/home/jens/rb-patched/build/i300/apps/codecs/vorbis.elf section .ibss)
12:06:56amiconn936 bytes too large...
12:06:58 Join JdGordon [0] (
12:07:19 Join Stel [0] (
12:10:07JdGordonthe build page is too bloody wide... :p
12:10:47JdGordonit doesnt fit on a 1024x768 screen..
12:11:03amiconnOn my screen 16 more builds would fit before I had to scroll...
12:11:14JdGordonwhat res u at?
12:11:58 Quit Stel ("Miranda IM! Smaller, Faster, Easier.")
12:12:41markunJdGordon: you could choose a smaller font of course.
12:12:53*JdGordon is blind,
12:13:15JdGordonwell not really blind.. but this font @ 10234x768 is about as small as id liek
12:14:30 Join Stel [0] (
12:16:05 Join RedBreva [0] (
12:19:39 Quit JdGordon (Read error: 104 (Connection reset by peer))
12:20:03 Join mirak [0] (
12:20:21 Join tvelocity [0] (
12:28:52 Nick solexx_ is now known as solexx (
12:34:59 Join ep0ch_ [0] (n=ep0ch@
12:36:34ep0ch_amiconn: i can remember compiling rockbox with gcc4 for ihp a while back. all codecs grew in size and were slower with the exception of aac
12:36:53 Quit Stel (Read error: 104 (Connection reset by peer))
12:37:04amiconnYes, mpa.codec is still slower with gcc4
12:37:44amiconn4% boost with gcc 3.4.6, 13% boost with gcc 4.0.3 (lame -.preset standard)
12:38:17amiconnI really wonder why
12:38:26amiconnpetur: MPegAudio
12:38:32ep0ch_you'll have to compare the asm ;)
12:38:57peturright - was confusing with m4a
12:39:08ep0ch_i'm surprised you got boost with 3.4.6, you have any dsp on?
12:39:11amiconnI have a suspicion though. Gcc is more aggressive at inlining, and if there's a function that should go to iram but is inlined into a dram function...
12:39:24amiconnI don't have dsp on
12:39:39ep0ch_what was your test file?
12:40:01amiconnAn ABBA album, encoded with lame −−preset standard
12:41:21ep0ch_hmm on ihp-120 320kbps doesnt boost at all even with crossfeed on (after preglows last commit)
12:42:29ep0ch_are ihp-120 and h320 cpus clocked at the same speed in rockbox?
12:43:18amiconnClock speeds are the same, however, H340 is a bit slower because (1)tThe button isr draws a few percent. (2) gfx is slower because of more data to modify, and the framebuffer being in dram
12:43:34ep0ch_ahh good to know
12:45:58 Quit Acksaw (Read error: 104 (Connection reset by peer))
12:46:36 Join Acksaw [0] (
12:46:44peturgwps.c contains more #ifdef lines as code
12:47:55 Quit Kohlrabi (Nick collision from services.)
12:48:05 Join Kohlriba [0] (
12:48:11*petur adds two more
12:48:15amiconnep0ch_: Especially mono bitmap drawing on 16bit LCDs is still in need of more optimisation. I have a few ideas, but first I need to write a benchmark because I can only test on H340 myself. I would need testers for ipod color, nano and video, and iaudio X5
12:48:49ep0ch_well i can help with the video
12:49:03Mooshere with the X5
12:49:07amiconn(the iaudio being the least important because it's also coldfire based, hence should gain the same speedup)
12:49:33amiconnBut first I need to finish my adaptions for using gcc4 for sh
12:49:43Moosgood luck ;)
12:49:48amiconnSeems I have to write the whole vector allocation in asm
12:50:11Moosamiconn: still have in mind the boost cpu during ATA?
12:50:13*amiconn curses silly gcc decisions
12:50:34amiconnMoos: Yes, but perhaps I have something even better in the pipe: ATA DMA
12:51:34Mooshard to chose what you are working on, I guess you have a huge Todo list : )
12:57:43ep0ch_can ata dma work on ipod? the 5g just seems sooo slow reading from disk in comparison to ihp
12:58:13 Join SereRokR [0] (
12:58:15amiconnSomeone would need to figure out how ata dma works on portalplayer
12:58:25amiconnMy ata dma will be coldfire only
12:59:06amiconnI had the impression that ata is quite fast on ipod... I should probably perform tests
13:01:13amiconnep0ch_: Where do you observe the slowness? When rebuffering?
13:01:50 Join TeaSea [0] (
13:02:35ep0ch_amiconn: just generally, even usb is really slow for me (not rockbox fault)
13:03:23amiconnHmm, then maybe apple's emergency diskmode is slow on video as well.
13:03:34amiconnDid you also try usb in retailos?
13:04:02 Join romanos_melodos [0] (n=romanos_@
13:04:21amiconnI would expect roughly the same usb performance as on iriver, since they use the same disk type (1.8")
13:04:53amiconnMy mini2g seems quite fast via usb, even though it only has a microdrive (0.85")
13:07:11ep0ch_hmm i might do a quick speed test between winxp and freebsd then, maybe its freebsd being slow...
13:09:23romanos_melodoshello petur! Fancy helping me out sorting out that rockboy-nano thing?
13:09:46peturi don't even have a nano
13:10:07peturand i'm not quite interested in rockboy...
13:10:17romanos_melodossorry.. i meant amiconn
13:10:46peturyou probably meant linuxstb
13:11:07petur /he not here
13:14:50 Quit SereR0KR (Read error: 110 (Connection timed out))
13:25:17 Quit tvelocity ("Ex-Chat")
13:25:52 Join damaki [0] (
13:25:54 Quit powr-toc ("Leaving")
13:28:03RedBrevaamiconn: Try Only the top row, and there is still alot a graphical optimizing, so some are still quite slow to load..
13:28:52***Saving seen data "./dancer.seen"
13:29:24amiconnWorks quite well, although the behaviour of the leftmost pic is different from the others
13:30:15peturI still like the overlib way better
13:30:53RedBrevaI will have a look at overlib, but wanted a quick way (that I already know) as proof of concept...
13:31:21peturoverlib is very easy to use - done in minutes ;)
13:31:45amiconnThe leftmost is probably different because it opens under the mouse pointer, so it doesn't go away when releasing the mouse button
13:32:25RedBrevaI know, but it should dissapear on mouse moving off it... as I say, some work left to do yet
13:32:32romanos_melodosjust wondering. How can iriver u10 play compressed avi but ipod nano can't? Is it just a cpu problem? Or the algorythm is really tricky?
13:33:14peturRedBreva: onmouseover="return overlib('<IMG SRC=bigimage.jpg>');" onmouseout="return nd();"
13:33:57RedBrevaWill try that soon... OK, that's my free time all used up till Tuesday :( [House full of visitor to look after now]
13:34:25peturin IE, the leftmost goes away even with the mouse over it
13:34:43 Join actionshrimp [0] (
13:34:55 Quit RedBreva ("Chatzilla 0.9.72 [Firefox]")
13:39:14peturhmmm... talk_buffer_steal() is only called when using prerecording, but afaik, pcm_recording always uses the complete audio buffer
13:39:45peturcould this be the reason voice stops working after recording for some people?
13:40:39amiconnSounds likely, but I wonder why pcm_recording does this
13:40:58peturalways use the buffer?
13:41:05 Quit ProgramZeta (Read error: 110 (Connection timed out))
13:41:05amiconnOn archos, the recording code doesn't touch the buffer before something is actually recorded
13:41:42amiconnSaves reloading the voice file if you just enter the recording screen without recording anything (without prerecording of course)
13:42:11peturin pcm_recording, as soon as you init recording it starts the audio input and dma, which run around the complete circular buffer
13:42:20amiconnBut otoh, the voice file is stored in a separate buffer for swcodec, so what does talk_buffer_steal() do then?
13:42:43amiconnpetur: Sounds like a waste.
13:42:52peturprobably is
13:43:02peturbut I always use pre-recording :)
13:43:15 Join Membrillo [0] (
13:43:16amiconnIt should only switch to monitoring but not enable the dma and not touch the buffer
13:43:52 Quit Membrillo (Client Quit)
13:44:04peturha, but monitoring is done during the dma cycle
13:44:15amiconnWHy is that?
13:44:20peturpeak meter
13:44:52peturwait let me check
13:44:53amiconnStill no reason to use the whole audio buffer
13:45:32amiconnWe could use a small buffer in IRAM for that (provided the core IRAM area includes only the DMA capable IRAM block)
13:45:50amiconnI would actually want that for fm as well
13:46:02petursorry, peak detection is done in the callback, not the dma handler
13:48:12peturaaargh voice is hanging at startup :( I'd better try to fix that first
13:48:46 Join chang [0] (
13:49:44peturamiconn: why does somebody want peakmeters for fmradio? I don't even want them for playback. bloody waste of cpu cycles
13:50:38peturunless you don't know what you're listening at and need to see if it's acually playing?
13:51:25ravonDo I always have to send 4096 bytes to pcmbuf_insert?
13:52:51amiconnpetur: We do want recording directly from the fm screen. Maybe peakmeters there are too much of a waste for the swcodec targets as long as we're not recording
13:53:01amiconnOn archos they come almost for free
13:54:37preglownot for swcodec
13:54:43preglowthey're pretty expensive there too
13:54:43amiconnOn archos, we switch the MAS to recording mode immediately when entering the FM screen
13:55:04amiconnThat's why we can't voice menu items in the radio screen
13:55:16amiconnHmm, that's an idea...
13:55:50amiconnWe shoudln't always switch to recording immediately when entering the FM screen, only when prerecording is enabled
13:56:01amiconnOtherwise we should switch on start of recording
13:56:11preglowbtw, we don't currently voice the recording screen, no?
13:57:07amiconnThat's because it's impossible on archos
13:57:17amiconn...for the same reason I've just mentioned
13:58:19amiconn...and in the recording screen we cannot delay the mode switch, because the MAS needs to be in recording mode for setting levels
13:58:28amiconnFor radio, the level is fixed
13:58:49 Join akaidiota [0] (
13:59:27preglowif it's ever going to be consistently possible in rockbox, we need to record spdif for monitoring
13:59:28peturfor swcodecs recording screen, we need recording active for the peakmeters, so no voice there
13:59:33preglowright now i hijack the playback channel
14:00:16peturamiconn: changing this is past 3.0 I hope?
14:01:26amiconnpetur: On swcodec it should be possible to do full duplex
14:03:22peturif wa also allow voice during recording we reduce recording buffer
14:04:10amiconnNo more than for playback
14:05:13 Quit mirak (Connection timed out)
14:05:21peturyou also need the playback buffer, no? (i admit i know almost nothing of the playback system)
14:08:01 Join mirak [0] (
14:08:21peturwoohoo - solid crash!
14:09:14peturwithout pre-recording and with voice enabled: going from recording screen to settings: freeze
14:15:51 Quit akaidiot (Read error: 110 (Connection timed out))
14:17:21 Part YouCeyE ("Leaving")
14:21:41 Quit eaz ("leaving")
14:22:48 Quit akaidiota (Read error: 110 (Connection timed out))
14:26:47peturamiconn: is it possible we are already doing full duplex? I just went from recording to settings, had voiced menu, went back and started a recording (pre-rec = 30s). Listening to the wav, I think I hear myself tapping in keys in the menu.
14:27:10peturthe ultimate test: try to record the voice
14:29:28 Join DpK|Cronic [0] (
14:30:34peturI just recorded a voiced menu :D
14:30:50Mikachutoo bad you can't record this monologue
14:31:07peturthe logbot does it for me
14:31:31peturand I'll earn myself a nice place in the stats for talking to myself
14:32:08peturbut it's not completely ok
14:32:09amiconnI have a 100% gcc 4.0.3 build running on my recorder!
14:34:22peturthere's some stuff seriously wrong
14:35:10peturafter the recording experiment, my battery indicator says 0% and I get a freeze. Bootloader says 4.03V
14:35:40peturand the first part of the recording is noise - probably overwritten by the voice code
14:36:24amiconnIt's probably that recording doesn't obey the voice buffer
14:36:50peturclashing is more the word
14:37:38peturmaybe if I fix that we have voiced recording screen?
14:37:54amiconnNot automatically
14:38:19amiconnThere are no voice clips for recording, and the code doesn't try to access them either
14:38:36peturat least I *do* have a recording of the voice output here :)
14:38:45amiconnBut it should be relatively straightforward to add these things, conditional for all swcodec targets
14:38:52amiconn...which have recording
14:39:40amiconnpetur: Yes, from the menu, which itself is voiced because it can be used from outside the recording screen as well. Then it also speaks on archos
14:40:09 Join midkay [0] (n=midkay@rockbox/developer/midkay)
14:42:12 Join tvelocity [0] (
14:46:57 Quit ashridah (Read error: 110 (Connection timed out))
14:46:59 Join daurnimator [0] (
14:47:43daurnimatorhaven't been here in a while
14:49:04midkayew. ew. it's you. get out.
14:49:15 Join Spacksaw [0] (
14:49:22midkayamiconn, that's a pretty big oops.
14:49:32*amiconn almost topped preglow's score...
14:49:32 Quit Acksaw (Read error: 104 (Connection reset by peer))
14:49:36daurnimatorthats ops
14:49:38daurnimatorhe wants a +o
14:49:46preglowamiconn goes for the kill
14:49:50 Join ashridah [0] (
14:49:58midkaywhat's preglow's record
14:50:01midkay? :)
14:50:01preglowbut no, you're still 20k short
14:50:03*amiconn broke all bootloaders, tsss....
14:50:11amiconnarchos bootloaders that it
14:50:19preglowi got something around 24000 with a nasty asm glitch
14:50:34midkayunbeatable. damn.. :)
14:50:46amiconnHmm, why ist that ?? :puzzled: ??
14:52:41amiconnProbably a function order thing... bootloaders uses function sections
14:54:09 Quit ep0ch_ (Read error: 110 (Connection timed out))
14:54:15amiconnAnyway, we gain a total of ~15KB by switching to gcc 4.0 and -Os
14:54:43Mikachudoes anything happen if you add that -finline-functions-called-once switch?
14:55:11amiconnmidkay: No, uncompressed code wise
14:55:57amiconn...which equals .ajz size except for recv1
14:56:41midkayneato :)
14:57:43amiconnThe 15KB are on recv1, but the gain should be roughly the same percentage (6.6%) for the other archoses
15:00:09 Quit tvelocity ("Ex-Chat")
15:01:14 Quit wehn_ ()
15:05:22 Join sucka [0] (
15:05:48 Quit sucka (Client Quit)
15:05:55 Quit PaulJam (".")
15:27:12*preglow wonders what to do about the malloc in libmusepack
15:28:13preglowi think i'll perhaps just disable the seek table
15:28:54***Saving seen data "./dancer.seen"
15:28:54preglowthe seek patch has a seek table that will break for big files
15:28:56Moosbtw, is the seek patch apply cleanly?
15:29:01Moosah ok
15:29:05preglowi've put an updated seek patch on the tracker page
15:29:08preglowthat should apply clean
15:29:13Moosok thanks
15:29:19preglowit now work 100% here
15:29:45 Part daurnimator
15:30:06preglowrotator gave me a couple of tracks that glitched with gapless, but they work just perfect here
15:30:12preglowso i might commit the seek patch today
15:30:28Mooswee \o/ !
15:30:52Mooslet me know if you need I test something specific
15:31:03preglowno, i don't think i'll need further testing
15:32:08 Join PaulJam [0] (
15:33:54*amiconn wonders why the optimised binary size has only marginal influence on the audio buffer size
15:34:55*amiconn wonders why he's wondering
15:35:20Mooshehe :)
15:35:23amiconnWhen I have rockbox in flash, there'll surely be almost no influence on audio buffer size
15:38:29 Join DarkJesus [0] (
15:59:23preglowshitty resampler...
16:12:59 Quit DarkJesus ("Leaving")
16:14:01 Join DarkJesus [0] (
16:14:44 Join Lear [0] (
16:15:57preglowbah dsp
16:16:06preglowit breaks everything into chunks of 128 samples
16:16:12preglowgood by cache usage
16:16:16preglowgood bye, too
16:17:12Learpreglow: about to commit, or what?
16:17:29preglowcommit what?
16:17:38Leardsp change
16:17:44preglowthat's a 3.1 thing
16:17:53preglowi can't be bothered to possibly introduce bugs there now
16:18:05preglowit might also introduce new yield() problems
16:18:14preglowin short: i don't feel i need the bugs right now, heh
16:19:43 Quit PaulJam (".")
16:20:11LearHe, sim link command is more than 4100 chars here. :)
16:21:28 Join julius [0] (n=julius@
16:21:45DarkJesusIs there an ipod style music quiz for rockbox?
16:22:40preglowno <-
16:23:19ravonGonna try again now that people are awake. Is there a recommended data size to send to pcmbuf_insert, or doesn't hit matter how much I send?
16:23:47Learnot really, whatever comes naturally, so to speak... :)
16:23:53DarkJesusDo people want one?
16:24:07ravonLear: Was that reply to me? :)
16:24:12preglowDarkJesus: someone probably does
16:24:20Learravon: yes
16:24:26ravonLear: oki, sweet
16:24:34ravonI must be doing something else wrong then :/
16:29:54 Join Paul_The_Nerd [0] (
16:30:22LearHm... When looking for a file in a directory, should filenames really be compared with strcmp?
16:30:39Mikachuwhat else?
16:34:12 Part Paul_The_Nerd
16:35:16 Join Paul_The_Ner1 [0] (
16:35:34 Nick Paul_The_Ner1 is now known as Paul_The_Nerd (
16:48:57LearAt tree.c:1236, why add 1 to the returned pointer from playlist_peek()? As far as I can see, the only "benefit" of that is that strrchr might fail, which isn't a very good thing...
16:51:23 Join PaulJam [0] (
16:57:59 Quit scottder (Read error: 110 (Connection timed out))
16:58:48 Quit damaki (Remote closed the connection)
17:04:53 Quit chang ("So long, and thanks for all the fish !")
17:09:09 Join damaki [0] (
17:11:25 Quit DarkJesus ("Leaving")
17:14:06 Join scott666_ [0] (
17:14:06 Quit scott666 (Read error: 104 (Connection reset by peer))
17:15:42 Join _Lucretia_ [0] (
17:22:19 Join JoeBorn [0] (
17:23:20preglowi have gcc for coldfire!
17:26:30LearAhh... Now how did you manage that?
17:27:50preglowjust commented away some shit that didn't work :>
17:28:57***Saving seen data "./dancer.seen"
17:29:17preglowmore specifically the check in gcc that prevented us from compiling libgcc
17:31:01Paul_The_NerdDoes it produce working binaries though?
17:31:22Paul_The_NerdIt could hate you in some more subtle way.
17:31:36 Join TCK [0] (
17:34:36 Join solexx_ [0] (
17:35:27 Quit ashridah ("Leaving")
17:36:59 Quit romanos_melodos ()
17:44:46 Join macdonalder [0] (
17:51:00 Quit solexx (Read error: 110 (Connection timed out))
17:51:01 Join rotator [0] (n=e@rockbox/developer/rotator)
17:52:37 Quit TCK (Read error: 110 (Connection timed out))
17:54:35 Quit mirak ("Ex-Chat")
17:56:51 Quit DpK|Cronic (Read error: 110 (Connection timed out))
18:00:20 Join chang [0] (
18:00:25 Quit chang (Remote closed the connection)
18:01:56 Join quobl [0] (n=quobl@tor/session/external/x-5c1849354c0661c5)
18:04:56 Join jborn_ [0] (
18:11:17SlasheriLear: hi
18:11:40LearHi, are you up to date with playback.c?
18:11:56Slasherino, absolutely not
18:12:22LearAh, hoped you would be, but suspected not. :)
18:12:31Slasherii haven't had time to figure out how everything has been changed
18:12:35LearTrying to fix a bug in it...
18:12:46Slasherihehe, good luck with that
18:15:13 Quit JoeBorn (Nick collision from services.)
18:15:19 Nick jborn_ is now known as JoeBorn (
18:20:00LearBut I think I found it... I suspect initialize_buffer_fill didn't back out quickly enough if filling was true
18:22:30 Join ProgramZeta [0] (
18:29:03 Quit PaulJam (".")
18:32:59 Join Ribs [0] (
18:41:23 Join akaidiot [0] (
18:42:12 Join dpro [0] (
18:46:18 Join akaidiota [0] (
18:54:17 Quit markun (
18:54:17 Quit slimeball (
18:54:17 Quit Ironhand (
18:54:44NJoinslimeball [0] (
18:54:44NJoinmarkun [0] (
18:54:44NJoinIronhand [0] (
18:55:53juliushi guys, is there a way to send out key presses on the serial port from an ipod? what I'm trying to do is take a look at here: , there is a iPod Remote mode (Mode 2), is this the one used for this kind of things?
19:01:54 Quit Paul_The_Nerd (Read error: 104 (Connection reset by peer))
19:02:36 Join Paul_The_Nerd [0] (
19:04:33 Quit akaidiot (Read error: 110 (Connection timed out))
19:05:12 Join akaidiot [0] (
19:05:15 Quit crwl ("Reconnecting")
19:05:46 Join crwl [0] (
19:07:40 Quit akaidiota (Read error: 110 (Connection timed out))
19:08:02 Join akaidiota [0] (
19:25:48 Quit akaidiot (Read error: 110 (Connection timed out))
19:25:54 Quit Ribs ("Leaving")
19:26:13 Join akaidiot [0] (
19:28:58***Saving seen data "./dancer.seen"
19:29:01 Join Ribs [0] (
19:32:44 Join TCK [0] (
19:33:07 Join Farpenoodle [0] (n=solo84@
19:35:08 Quit akaidiota (Read error: 110 (Connection timed out))
19:35:57 Join akaidiota [0] (
19:37:03 Join DpK|Cronic [0] (
19:42:27 Join lee-qid_ [0] (
19:52:52 Quit akaidiot (Read error: 110 (Connection timed out))
19:53:40 Quit DpK|Cronic (Read error: 104 (Connection reset by peer))
19:54:15 Quit _Lucretia_ (Read error: 110 (Connection timed out))
19:55:37 Join _Lucretia_ [0] (
20:04:58 Join chang [0] (
20:17:36 Part _Lucretia_ ("Leaving")
20:22:26 Nick slimeball is now known as _slimebal (
20:22:58 Nick _slimebal is now known as slimeball (
20:25:05 Quit Lear ("Chatzilla 0.9.72 [Firefox]")
20:28:57 Quit ProgramZeta (Read error: 110 (Connection timed out))
20:31:31 Join Rondom [0] (
20:35:32 Join Febs [0] (
20:35:39 Join Bger [0] (n=Bager@
20:35:43 Join powr-toc [0] (
20:36:43 Quit Paul_The_Nerd (Read error: 104 (Connection reset by peer))
20:39:19 Quit Bger (Client Quit)
20:43:21 Join BHSPitLappy [0] (
20:47:35 Join jbauman [0] (
20:48:31 Quit midkay ("Leaving")
20:55:53preglowamiconn: the amd64 hack worked fine for you?
21:03:40preglowgcc 4.1 for m68k is a warning feast for rockbox...
21:04:25 Join amiconn_ [0] (
21:05:25amiconn_preglow: Yes, the hack worked for me
21:05:46amiconn_...but I don't use it for "production"
21:05:47preglowme too
21:06:05preglowi'm checkin out 4.1 now
21:06:17preglowa ton of pointer signedness warnings
21:06:18amiconn_I would like to do some further tests, and preferably find out what goes wrong
21:06:37preglowit seems to generate bigger code
21:06:42amiconn_But that has to wait until after 3.0
21:06:43preglowtwo ibss full so far
21:06:57amiconn_Yes, ibss full with 4.0 -O2 as well
21:07:07amiconn_(in vorbis.codec)
21:07:25amiconn_The code is slower than that generated by 3.4.6
21:07:29preglowrockbox.iriver is 1k5 bytes smaller
21:07:46amiconn_On H340: 4% boost with 3.4.6, 13% boost with 4.0.3
21:07:56amiconn_(mp3 lame −−preset standard)
21:08:49preglowthe button driver is COMPLETELY screwed with 4.1
21:09:00preglowit always presses down
21:10:01amiconn_We should get it smaller by adding -fno-align-functions -fno-align-jumps -fno-align-loops -fno-align-labels
21:10:34amiconn_This alignment probably doesn't make much sense with our direct-mapped cache
21:11:05amiconn_preglow: What target?
21:11:18preglowi've used gcc 4.1 for arm for a month now
21:11:29amiconn_Usually this points towards a bug in our code
21:11:52preglowboth button.c and adc.c compiled with no warnings
21:12:01preglowbut there are warnings _ALL OVER_ the rest of the code
21:12:05amiconn_I suspected gcc 3.4 + too on SH1. Turned out that it was a bug in ata.c
21:12:15preglowand the few i've looked at are valid warnings
21:12:18amiconn_The signedness warnings are also there with 4.0
21:12:27preglowi don't get them with arm gcc
21:12:32preglowno wonder
21:12:32amiconn_We just disable them in the Makefile
21:12:35preglowthere is no error on arm
21:13:05amiconn_Maybe configure specifically checks for 4.0, and doesn't add -Wno-pointer-sign for 4.1 ?
21:13:28amiconn_Or perhaps you just forgot to reconfigure...
21:14:02preglowdidn't reconfigure, no
21:14:11preglowwe don't we fix it, btw? just boring?
21:14:17amiconn_You should reconfigure when switching compiler versions
21:14:38amiconn_I had to learn that during my SH1 tests
21:15:03amiconn_Btw, I did not yet try to build sh-elf-gcc 4.x on amd64... hope there are no nasty surprises
21:15:52preglowbut yeah
21:16:07preglowi don't think i'll stick with gcc 4.1 ...
21:16:20preglowat least not for m68k
21:17:55amiconn_Too bad that 4.0 produces slower code for coldfire than 3.4. Would be nice to use the same gcc version for all targets...
21:28:59***Saving seen data "./dancer.seen"
21:34:24 Join damaki_ [0] (
21:41:11SlasheriHmm, when i tried some speed optimization flags with gcc a long time ago, that also screwed up the button driver
21:43:12 Join arf-arf [0] (
21:44:20preglowSlasheri: oh, hmm
21:46:14Slasheriyep, it might be a bug that is encountered only with some special compiler settings
21:47:13Slasherilike a timing issue or something like that
21:47:31preglowahh, yeah
21:47:36preglowsounds probable
21:47:40preglowbut i know nothing about that code
21:47:53preglowwell, at least not the adc part
21:47:55Slasherihehe, me neither :)
21:48:17Slasheribut might be interesting to look into that someday
21:48:28amiconn_Hmm, now that you mention it...
21:49:12amiconn_When looking into the performance issue in the H300 PCF driver, I noticed that the timing in the H1x0 button adc driver might be a tad too aggressive
21:49:48amiconn_I have no problems running an -O2 rockbox core on H300
21:49:56 Quit damaki (Read error: 110 (Connection timed out))
21:50:04amiconn_(after my ata.c fixes of course)
21:51:04Slasheriah, then it should be trivial to fix
21:52:03preglowrotator: yo
21:52:13amiconn_We should probably use the same adaptive timing system as in the H300 PCF driver
21:52:14 Quit damaki_ (Read error: 104 (Connection reset by peer))
21:52:18rotatorpreglow: what's up?
21:52:27preglowrotator: i can't for the life of me get a click out of the two tracks
21:52:40rotatorhmm, let me do some more testing then...
21:52:57preglowrotator: the new rockboux output is bit identical to the decoder source compiled on a pc as well, with the sample.cpp program
21:53:13preglowrotator: even when comparing both tracks, complete with the track transitions
21:53:35rotatorperhaps the problem was that I was seeking to the end of the first track?
21:53:44rotatormight that still cause a glitch
21:53:45 Quit Rondom ("I'm leaving on a jetplane, don't know when I'll be back again...")
21:53:52preglowi've tried it zillions of times
21:53:54preglowdoesn't click
21:53:54 Join damaki [0] (
21:54:02preglowit only clicks if the track after the first track is some other track
21:54:04preglowbut that i can't fix
21:54:08preglowwe'd need some micro crossfade for that
21:54:32rotatorprelgow: let me try it again, i know i got it to click consistently last night
21:55:12 Quit amiconn_ ("CGI:IRC (EOF)")
21:55:37rotatorpreglow: still getting a click :/
21:55:49rotatorI just fastforwarded to the end
21:56:09preglowi'll try it on target, then
21:56:24rotatorI'm letting it play through to the end now...
21:58:03 Quit damaki (Read error: 104 (Connection reset by peer))
21:59:12rotatorpreglow: no click when letting the track play through
21:59:12 Join damaki [0] (
21:59:43preglowyou're right, i got a tick
21:59:48preglowwhy the flaming hell don't i on the sim
22:00:28preglowamiconn: sure your memmove works fine? :>
22:01:19preglowno click on the sim
22:01:40preglowhahah, and i just electrocuted myself on the usb cable
22:02:08rotatorhaha, didn't know that was possible :)
22:02:29preglowseems it is
22:03:40rotatorprelgow: just tried a couple more things...
22:04:03rotatorseeking to the middle of the first file still produced a click at the end
22:04:18preglowwell, i figure if seeking breaks it, it does so at all places
22:04:36preglowbut it did not click at all if not seek, no?
22:04:36rotatorseeking only a few seconds at the beginning then letting it play through didn't
22:04:59rotatorno click if I didn;t seek at all
22:05:11rotatorbut it seems seeking might not always cause a click
22:06:46preglowok, but what the hell might trigger on target and not in sim
22:07:08preglowi triggered it on my h120 easy as nothing
22:09:00preglowi'm gonna stop asking you for problem tracks in the future, heh
22:09:10preglowall my other ones work fine on target as well
22:09:14rotatorhaha, ignorance is bliss ;)
22:09:39preglowbliss and sweetness
22:11:10rotatormaybe it's just a fluke with this one track, I haven't had any problems with the dozens of others I've tried so far
22:11:44preglowwell, it does sort of point to there being a problem...
22:11:52preglowthink i'll try it on ipod as well
22:14:28preglowno go on ipod
22:14:33preglowwhich does give valuable pointers
22:14:43preglowprobably some emac specific shiznittery
22:15:15rotatorhooray... :/
22:15:24preglowbut why only when seeking
22:15:48preglowor, hmm
22:15:52preglowand endianness issues
22:17:11preglowi think i'll just have to ignore this issue for now
22:17:28preglowi've used enough time on musepack for now, and i still haven't commited the seek patch
22:17:31preglowi also need to work on resume
22:17:46rotatorhehe yeah, it should be good enough for now
22:17:58rotatorit's definitely much better than it used to be
22:23:46arf-arfmusepack is working SO nicely now. Muchos thanks, preglow
22:26:27preglowyou're welcome
22:27:38 Join akaidiot [0] (
22:34:04preglowplease test
22:34:09preglowif it works fine, i'll commit now
22:35:56preglowbackwards seeking doesn't seem to work...
22:36:56preglowanyone know how large the malloc buffer is, currently?
22:40:20preglowbiosphere at 200% pitch
22:40:23preglowspeed ambient
22:41:29goffadamn ... tired
22:41:50goffajust got done digging trenches for a campsite in town... we're burying power lines
22:45:23 Quit akaidiota (Read error: 110 (Connection timed out))
22:46:17arf-arfpreglow: when did you last update the musepack_seek_updated.patch?
22:46:34preglowright now
22:46:50arf-arfokay, in that case I'll rebuild & test it
22:46:50preglowi disabled the seek table
22:47:00preglowbut i see that that breaks rewiding
22:47:14arf-arfhmmm -maybe I won't then/
22:48:17preglowit doesn't work anyway here
22:48:36*preglow kicks snowgoon
22:49:24 Quit julius ("Leaving")
22:49:43goffawho's snowgoon?
22:50:02Mikachuis he yellow?
22:50:40preglowsnowgoon's the guy who wrote the musepack seek patch
22:51:56 Quit Kohlriba (Read error: 104 (Connection reset by peer))
22:52:02 Join Kohlrabi [0] (
22:55:28 Quit Moos (Read error: 110 (Connection timed out))
22:56:33 Join bluebrother [0] (
22:57:41 Quit bluebrother (Client Quit)
23:00:57*petur likes to avoid campsites that have power lines
23:01:35goffayeah.. more an rv park
23:01:40goffaits in town
23:01:53preglowcan anyone actually make rewinding work?
23:02:16 Quit TCK (Read error: 104 (Connection reset by peer))
23:02:30goffarvs are fancier than my house these days
23:02:50preglowwhat's an rv?
23:03:03goffarecreational vehicle.. like a big camper
23:03:12peturoversized camper
23:12:11 Join changsito [0] (
23:13:25 Quit changsito (Client Quit)
23:13:48 Quit chang (Read error: 110 (Connection timed out))
23:13:59peturis there some doc that describes how the audiobuffer is used by voice/playback/codecs?
23:14:37 Join chang [0] (
23:15:40preglowdon't think so
23:16:22peturi'll use the source then...
23:16:58 Join bernd_b [0] (
23:22:59peturmy h340 is acting up...
23:23:27amiconnNo linuxstb around...
23:23:54amiconnsudoku.c gives a strange warning with gcc 4.x - but only on SH1
23:24:24peturI play a file, press stop, go to menu. voice reads the first entry twice, then the battery indicator goes from 100 to 0% in a few seconds and the player freezes
23:25:16amiconnVoice sometimes eats itself on swcodec after stop
23:25:27amiconnDidn't get freezes recently though
23:25:31peturoh nice
23:25:42peturthe weird thing is the battery
23:25:50amiconnyes indeed
23:26:11peturthe indicator is 0% and blinking rather quickly
23:26:25amiconnThere's a couple of weirdnesses introduced on archos probably caused by some uncautious swcodec changes :/
23:26:33amiconnI know at least 2
23:26:57peturmaybe something gets overwritten that crashes another thread?
23:27:09amiconn(1) On Ondio, voice doesn't work after boot. It does work after playing music
23:27:30amiconn(2) On Player, the battery icon doesn't animate anymore when charging, but is shown empty instead
23:28:16thegeekdoes anyone know if it is remotely worth it to attempt to fix a broken h120 (bad harddrive connection)
23:28:33preglowthegeek: sure, why not
23:28:51thegeekI think linus once said it was very hard
23:28:56thegeeksmd chips
23:29:00***Saving seen data "./dancer.seen"
23:29:17 Quit powr-toc ("Leaving")
23:29:27peturI thought it was an easy soldering job
23:29:31thegeekI've already fixed it once(an smd resistor was knocked loose from a baad fall)
23:29:32thegeekit is?
23:29:35thegeekif so
23:29:48thegeekis there anywhere I could get some more info?
23:29:54peturthen again, I do have a very bad memory
23:29:56thegeekI've already done a little smd work
23:30:08thegeeksuch as motherboard voltmods and small fixes
23:30:25thegeekatleast I want to try before I give up on it;)
23:32:03peturhave you opened up and looked at it?
23:32:11thegeekmany times;)
23:32:20peturmaybe there's info at MR?
23:35:36 Quit chang ()
23:37:47 Join mirak [0] (
23:42:13 Quit lee-qid_ ("Trillian (")
23:46:43 Join sebastianbf [0] (
23:47:04sebastianbfplease i need some help
23:47:37Bagderso ask
23:47:55sebastianbfi'm planning to buy an iaudio x5l
23:48:16sebastianbfand I've read that you can upgrade the firmware to rockbox in it
23:48:24Bagderyes you can
23:48:40sebastianbfthe thing is where can i take a look at what features it would add t the iaudio
23:49:43Bagder... although not all those features are yet working on the x5
23:50:07sebastianbfoh ok
23:50:11sebastianbfone more thing
23:50:31sebastianbfi was also looking at zen vision:m
23:50:32jncsebastianbf: rockbox has rather difficult default menu systems, if you ask me
23:50:48jncthe audio decoding is very good quality
23:50:50Bagderjnc: compared to what?
23:50:57sebastianbfto the iaudio x5l
23:51:04sebastianbfoh sorry
23:51:09jncany idiot can learn to use an iPod
23:51:23peturidiots buy them :P
23:51:24Bagderthe X5 original FW is not easy to understand
23:51:27jncrockbox of course can be learned, it is a little bit more complicated.
23:51:31BagderI actually had to bring out the manual...
23:51:42sebastianbfi can learn
23:51:42Bagder... to find myself away from the recording screen!
23:51:51sebastianbfi'm more interested in sound quality
23:51:51BHSPitLappyjnc: the iPod's interface is easier because it's so restricted and limited.
23:51:59Mikachui think having three different settings menus in the toplevel menu is a bit crazy
23:52:14Bagdersebastianbf: the zen is not likely to run rockbox for a good while
23:52:21 Join TCK [0] (
23:52:22sebastianbfoh ok
23:52:46sebastianbfbagder: so you think that for sound quality i should go for the x5l?
23:52:55sebastianbf(over the ipod)
23:53:01BagderI have no idea
23:53:09BagderI have an X5, no ipod
23:53:19sebastianbfdoes it sond good?
23:53:48Bagderbut hearing is very subjective
23:53:59sebastianbfi have a Himd player now but im gonna sell it to buy the x5, and i want something that sounds with the same or better quality
23:54:10BHSPitLappyit's not as if one player is going to have horrible sound quality...
23:54:34Mikachureally cheap ones could have a sucky dac
23:54:35jncBHSPitLappy: some iPods have notoriously crappy sound
23:54:40sebastianbfyeah i know but flac support is very nice
23:54:46BHSPitLappythey all play back what you give them, the setbacks are codecs, settings, etc
23:55:08BHSPitLappyjnc: none that I've encountered, so doesn't bother me.
23:55:10amiconnjnc: Huh? WHich one would that be? My impression is that the ipod audiocodec is less noisy than the iriver one.
23:55:21BHSPitLappya defective model is easily replaced.
23:55:33amiconn(but then I know only one ipod audiocodec - the one in the mini 2nd gen)
23:55:50jnci was referring to the DAC
23:55:59amiconnyes, me too
23:56:02jncbetter DACs are used in 3G and newer
23:56:39jncjust the original iPod
23:56:50Bagderits funny how the chip makers usually call them "codecs"
23:56:59markunBagder: The zen vision:m has a 200 MHz ARM, right?
23:57:03amiconnWhy is taht funny?
23:57:18Bagdermarkun: its a DM320 one, I don't know at what clock
23:57:35Mikachuyou could call analogue and digital encodings, which would make a converter a codec
23:57:43Bagderits funny because it creates confusions
23:57:50Bagdermarkun: yes I know, its the same one
23:57:59BagderI think at least
23:58:10BagderI'm not aware if the DM320 exists in different versions
23:58:32Bagderno 100% free code exists for the DM320
23:58:52amiconnBagder: I now have rockbox #1 patches for gcc 3.3.6, 3.4.6 and 4.0.3. How to proceed with the switch now?
23:58:53markunWhy not? You always have to use the DSP?
23:58:53Bagdereven though they (all=) run Linux on them

Previous day | Next day