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 2005-06-03

00:00:43HClsoon we'll have a database generator that can read ogg
00:00:47HClbut its in java :x
00:00:57amiconnLinusN: I get a feeling that this is the cause for so many broken recordings with my older, asm-optimised recording transfer
00:01:12Chamoislinus : i have md5 ffdf317356cc0a964a2708d6b90f59cb
00:01:23Chamoiswith 1.65Eu and bootloader 2 is it ok ?
00:01:25amiconnNow that I have it asm optimised again, I also get broken recordings again...
00:01:28Bagderlucky me, I don't have any oggs
00:01:38preglowi have only oggs ://///
00:01:39jochenhey, oggs are cool!
00:01:51Bagderperhaps, but java is not
00:02:02LinusNamiconn: congrats...not!
00:02:02*amiconn has 99.9% mp3s and 0.1% oggs
00:02:04jochenat least vorbis supports gapless
00:02:19preglowso does (some) mp3s
00:02:23Bagdermp3 "supports" that too
00:02:28pregloweveryone with a brain uses lame anyway
00:03:03jochenwell, but it wasn't meant to support gapless because of the frame-based structure.
00:03:15preglowvorbis too uses frames, you know
00:03:19pregloweverthing uses frames
00:03:27Bagderframes R us
00:03:45preglowit doesn't support gapless because the exact original track length is usually not saved anywhere
00:03:47jochenwell, you know what i mean. *g*
00:03:51tucozpreglow, the high pitched sound is like a faint beep (like on a hearing examination)
00:04:02preglowtucoz: i usually didn't hear those beeps :P
00:04:20Stryke`fwpather.exe won't seem to patch 1.65-US
00:04:29LinusNChamois: yes
00:04:37LinusNStryke`: i am an idiot
00:05:03tucozI have a pair of sennheiser px200, and they seem to be very sensitive to stuff like that
00:05:20preglowdoes the iriver fw do this?
00:05:35jochensry, can't ack that LinusN ;-)
00:05:59XavierGrokay now I have to port another plug-in for the iriver, maybe i will give snake or snake2 a try...
00:06:08preglowtucoz: can you record it somehow?
00:06:10XavierGrany suggestions?
00:06:32tucozBut, i don't know if the iriver or the headphones do this, but when I pass those security things (!) in stores I also hear a faint beep
00:06:38tucozhmm, can try
00:06:53Stryke`tucoz: i hear that as well (the security things)
00:07:13preglowtucoz: that's fairly normal, but i don't think this is electromagnetic interference
00:07:22tucozpreglow, in rockbox?
00:07:33preglowno, i mean the security things
00:07:43tucozthe sound I'm experiencing?
00:07:49preglowyes, that's normal
00:07:51preglowbut not from rockbox
00:07:58preglowfrom security things ;)
00:08:40tucozhmm, never recorded anything in linux. Will look around for a bit
00:08:44preglowi've got a set of great big headphones with a long wire, so i always hear it as well
00:08:52tucozat stores?
00:09:05jochentucoz: take a look at audacity
00:09:06tucozok, so is that the wire causing it?
00:09:27amiconnIt seems that rockbox doesn't correctly initialise something on iriver (UDA?)
00:09:30preglowcombination of wire and headphone impedance
00:09:53tucoznice to know. Not much of a electronics-knowing person
00:10:08preglowamiconn: well, we've got people complaining rockbox sometimes refuses to play sound after booting
00:10:09amiconnThere is a strong 'pop' when I plug or unplug my earphines when rockbox is running
00:10:24preglowsounds like a dc offset
00:10:32amiconnThat doesn't happen with the iriver fw, even at max. volume
00:10:49preglowbut that should never happen
00:11:19tucozI also hear when the harddrive spins up
00:11:31tucozIn my headphones that is
00:11:57amiconntucoz: Possibly caused by the dc offset problem... then you'll hear the power ripple caused by the hd
00:12:43tucozamiconn: in plain english? ;)
00:12:46preglowi thought most output drivers were dc coupled
00:13:00tucozdc like in direct current = battery?
00:13:45jochengtg, gn8
00:14:15 Quit jochen ("Verlassend")
00:14:37amiconntucoz: yes
00:15:02tucozand something is wrong with that? A hw error?
00:15:22amiconnI just measured it: Rockbox outputs +1.1 V at the left channel, and +0.3 V at the right channel
00:15:32preglowbut how?
00:15:43amiconnWith iriver firmware it's perfect zero (<0.01 V)
00:15:52amiconn(measured without load)
00:15:55LinusNcan someone help me, i don't have the korean ihp_120.hex files
00:16:10preglowLinusN: and the site doesn't work?
00:16:21LinusNthe zip files contain windows executables
00:16:21amiconn...and it's independent whether it's fed through the remote or not
00:16:29preglowLinusN: ahhh
00:16:46preglowhave you tried unzipping them?
00:16:52preglowthey might be sfx executables
00:17:00markunLinusN: Wine?
00:17:07amiconnLinusN: sec...
00:17:26LinusNmarkun: sure, but i don't want to install wine just now
00:19:13LinusNok i have 1.65k now
00:19:31LinusNanybody has 1.63k?
00:20:02preglowamiconn: does your meter support picking up the dc component in an audio signal as well? would be fun to see if it's there all the time
00:20:52preglowmine just goes wild
00:20:59preglowi guess it doesn't filter it first
00:22:20Stryke`LinusN: i'm uploading 1.63K now if you still need it to my ftp
00:22:30Chamoistoo late
00:22:40Chamoisi gave it few sec
00:22:57amiconnpreglow: I didn't apply the audio patch
00:24:23preglowi keep forgetting not everyone has it
00:24:26tucozpreglow, didn't show anything when I recorded.
00:24:43preglowtucoz: tried normalizing it?
00:24:53tucozit's just a flat line
00:25:04preglowthe signal is probably too weak for you to see it
00:25:06preglowbut i bet it's there
00:25:25amiconnpreglow: Do you also get the dc offset (with silence)?
00:25:25tucozok, I zoomed in in audacity
00:25:51 Quit RED_M_CHIU (Read error: 104 (Connection reset by peer))
00:26:58preglowamiconn: haven't got an extension jack cable, and it's pretty hard to fit the probes in the hole...
00:27:22preglowtucoz: do normalize in audacity
00:27:23amiconnSuch a cable was part of the accessory...
00:27:30preglowamiconn: i know! can't find it
00:28:35tucozpreglow, where do I find that?
00:29:20preglowtucoz: effect->normalize
00:29:26 Join leftright [0] (
00:29:39tucozfound it, thanks anyway
00:30:09preglowtucoz: do you hear/see it now?
00:31:01tucozargh, don't know. I will try to record it again. Some program I run has 'stolen' the audio device
00:32:20 Join ashridah [0] (
00:34:23tucozpreglow, I guess I didn't record anything. I hear the iriver, but audacity don't record it
00:34:32preglownot much i can do then, heh
00:34:33amiconnLinusN: Any objections against changing the LED defines
00:34:34tucoznow playing song
00:35:33amiconnI'd change that to a multi-value setting. CONFIG_LED { LED_PHYSICAL, LED_VIRTUAL, or undefined }
00:35:58amiconnLED_PHYSICAL would be for a software-controlled LED (archos player /recorder)
00:36:07amiconnLED_VIRTUAL is for Ondio
00:36:27amiconnFor iriver it would be undefined as it has a hardware controlled LED
00:36:53amiconnThat would get rid of the MMC icon on iriver
00:37:22LinusNfwpatcher should work better now
00:37:38LinusNamiconn: fone with me
00:37:54preglowiriver has mmc icon?
00:38:09amiconnIt has... currently
00:38:17preglowhow... misplaced...
00:38:29amiconnThe little icon that pops up in the to right corner on hd access
00:39:14MoosCamaroLinusN: yes, work fine now
00:39:17LinusNtime to sleep
00:39:21LinusNMoosCamaro: great
00:39:28MoosCamarogood night
00:39:34LinusNnite all
00:39:52 Part LinusN
00:41:32 Part leftright
00:42:11tucozok, now the beep is showing
00:42:12 Quit _aLF ("Leaving")
00:45:36tucozpreglow: if you would like, you can have a look at it at
00:46:55preglowthis happens when?
00:47:10tucozwhen I enter a directory for instance
00:47:25tucozlike, when the dir contents are loaded to memory
00:47:26CoCoLUSoriginal firmware -did- the same
00:47:34CoCoLUSthey fixed that, though
00:47:37tucoznot when I return to a dir
00:47:42preglowso it's disk interference
00:47:54tucozok, my player or?
00:48:06tucozsomething wrong with it that is?
00:48:33preglowthe left channel's got massive interference
00:48:43preglowi'll try this on my player
00:48:58tucozit doesn't happen all the time though.
00:49:07XavierGrgood night all!
00:49:35preglowdoesn't happen at all here
00:50:08preglowcan't say i've ever heard it
00:50:22 Quit XavierGr ()
00:50:42tucozI haven't either using iriver fw
00:51:20CoCoLUSi have
00:52:06tucozNow I rebooted to iriver fw, shut down, and booted rockbox. Then it's completly silent
00:52:11 Part MoosCamaro
00:52:30 Join TCK [0] (
00:58:20tucozok, got to go. But, there could be an idea to collect stuff like this on a wikipage.
01:00:02hacimanyone have a good source for purchasing hardware? Ebay is not my favorite method, but thats the only place I know of to go to.
01:00:31tucozthen again, as the iriver rockbox still is in early-mid development, I guess bug reports will be as for the archoses.
01:00:49 Part tucoz
01:04:46preglowwell, what better sources are there?
01:04:51preglowit's all second-hand anyway
01:09:24 Quit ashridah ("out")
01:09:39 Quit TCK- (Read error: 110 (Connection timed out))
01:11:09HCli suppose a lot of people would mind if i upgraded the database to include some more fields, rendering the perl version not usable?
01:11:26HClits just that its really easy to get playtime into place with the java one :/
01:11:38HCland i need that one in the database soon...
01:12:29preglowthen do that in a local copy
01:12:31amiconnNo java please
01:12:41HClwhat should i do then?
01:12:47preglowdestroying the only means people have of updating their database is not clever
01:13:07HClwell, i could adapt the perl version to fill in 0 for the songlength
01:13:14HClwould still make it work, but limit it.
01:13:48preglowthat would of course work, but anyway, requiring java will destroy the flexibility people have now
01:13:54preglowwhat with just having perl.exe on their player
01:14:05HCli know that.
01:14:14preglowi don't exactly LOVE that solution myself
01:14:16HCli leave it open for someone to update the perl variant..
01:14:19preglowbut it does work
01:14:39HCli agree that it limits things a bit
01:14:48HCli'm hoping to find a java to .net compiler somewhere would probably be more useful
01:15:08HClsince it creates a .exe
01:15:17preglowand requires .net to be isntalled
01:15:18HClthat you can run under linux too, with mono
01:15:32HCli'll just make the songdb write in not availables
01:15:34HClfor all stuff added
01:15:35preglowc would by far be nicest :P
01:15:37HClso it will still work.
01:15:41HClnot for me, the coder
01:15:42HCl :)
01:15:43thegeekfewer people have mono than java
01:15:50HClthegeek: maybe
01:15:53preglowno, but the coder isn't the one who's supposed to be having an easy time
01:15:59HCl :P
01:16:04HClscrew that. anyways :P
01:16:10thegeekI'm pretty sure very few actually have mono installed
01:16:19HCli agree with that
01:18:15***Saving seen data "./dancer.seen"
01:19:28thegeekis it absolutely impossible to do it in perl?
01:19:50HClnot impossible, but it will be unmaintainable.
01:20:15HClits already totally unreadable
01:20:16preglowperl isn't unmaintainable by default
01:20:22preglowthen fix it
01:20:30pregloweverything doesn't have to be in one file
01:20:31HCli don't know perl.
01:20:33preglowfor one
01:20:44HClplus i'm | | this close to a working java solution
01:20:49HClthat will also be able to work the runtime databas
01:20:54HClwithout a sweat
01:21:03HClbeing able to read the old database in and adapt it
01:23:08preglowdoes gcj create standalone java executables?
01:23:19preglowthat might be sort of redeeming
01:23:29HCli think it does.
01:26:53preglowlooks like it can
01:28:17HCli'm starting to get very confused
01:28:22HClwhere is my old
01:28:28HCli'm trying to check it out from cvs
01:28:31HCland i'm getting an old version
01:29:56HClstupid cvs.
01:39:08preglowgotta catch some hours of sleep
01:39:10preglowlater all
01:39:16 Quit preglow (":-)")
01:43:14 Quit markun ()
01:50:42 Quit cYmen ("zZz")
01:52:16HClyay! i love oo :3
02:14:48 Join QT_ [0] (as@area51.users.madwifi)
02:17:35 Quit QT (Read error: 148 (No route to host))
02:22:13 Join StrathAFK [0] (
02:22:21Stryke`anyone try bootloader v2?
02:23:16Stryke`i'm always a bit nervous when it comes to flashing
02:24:05HClyou and me both
02:25:39Stryke`i have a dumb question, as well
02:25:58HClask, ask
02:26:05Stryke`when updating the rockbox build over usb, its perfectly okay to load up the rockbox firmware
02:26:34Stryke`it will let you overwrite rockbox.iriver and .rockbox
02:27:09Stryke`the first time i did it i booted the original firmware
02:30:09Stryke`sorry - i didn't really ask my question
02:30:35Stryke`rockbox will let you overwrite rockbox.iriver and .rockbox, right?
02:37:02HClyay, my java implementation pooped out a database
02:39:09*HCl goes to alter it to make it poop out an v2 database..
02:40:27thegeekwhat changed?
02:40:31HClyea, the current database.
02:40:39thegeekI meant the bootloader;)
02:40:42 Quit Strath (Read error: 110 (Connection timed out))
02:41:02thegeekI know you did v2 of the database:)
02:41:07Stryke`no longer reads the hard drive if the hold switch is on
02:41:33HClmmm... where is my database tester..
02:46:47*HCl slaps java.
02:46:52HClno. bad java! no unicode!
02:47:28HClstupid fancy living in the future programming language :3
02:49:52gromit`would the rockbox project be interested in participating to a conference in the RMLL 2005
02:51:44gromit`hmmm for the moment i have only a french website to present
02:53:48gromit`but people from the whole world are coming
02:54:12gromit`it is a free software meeting
02:56:27gromit`pv me for information
02:56:30 Join lostlogic [0] (
02:57:20gromit`going to bed
02:58:44HClyay, i think i nailed the last bug in my java code
03:02:41HClthats what i like to believe anyways ;p
03:02:50HClbut alas, its not true. where does it go wrong..
03:12:30 Join tvelocity [0] (
03:17:00*HCl scratches his head..
03:17:07HClboku wa baka.
03:17:11HCli'm such an idiot.
03:18:16***Saving seen data "./dancer.seen"
03:20:18HClhm. o.o.
03:28:33 Join amiconn_ [0] (
03:37:01 Quit amiconn (Read error: 145 (Connection timed out))
03:37:02 Nick amiconn_ is now known as amiconn (
03:38:42HCli'll have to prod markun tomorrow
03:38:47HClto see whether it reads oggs properly
04:03:38 Join einhirn [0] (
04:05:17 Join QT [0] (as@area51.users.madwifi)
04:12:10 Quit Bagder (Read error: 60 (Operation timed out))
04:14:46 Nick StrathAFK is now known as Strath (
04:17:03 Quit QT_ (Read error: 110 (Connection timed out))
04:37:10 Join asdsd____ [0] (
04:41:37 Join midk [0] (
04:43:21 Join bipak_ [0] (
04:43:54 Quit midk (Client Quit)
04:43:57 Join midk [0] (
04:44:26 Part asdsd____
04:44:37midktopic STILL hasn't changed? haha
04:44:51midkhm. and plugh's back...
04:44:59midkand that darn logbot. :P
04:45:32 Quit xen` (Read error: 148 (No route to host))
04:47:24 Quit midk (Client Quit)
04:47:37 Join midk [0] (
04:49:25 Quit Stryke` (Read error: 110 (Connection timed out))
04:57:03 Quit midk ("later...")
04:59:44 Quit bipak (Read error: 110 (Connection timed out))
05:18:18***Saving seen data "./dancer.seen"
06:15:14 Join DMJC [0] (
06:15:25 Quit DMJC (Read error: 104 (Connection reset by peer))
06:38:57 Join ashridah [0] (
06:58:38 Join LinusN [0] (
07:11:37amiconnLinusN: check the Ondio forum....
07:13:13LinusNhehe, oops
07:14:18amiconnThis is in fact crucial; if someone accidentally reformats an Ondio with FAT32 _and_ also has an old firmware release, the Ondio will be bricked...
07:14:40amiconnOnly rescue would be uart boot + reflash
07:15:59amiconnRead , notes for 1.31f
07:18:21***Saving seen data "./dancer.seen"
07:25:37Bgermorning :)
07:25:38 Part LinusN
07:25:53 Join LinusN [0] (
08:19:05 Quit Chamois (" HydraIRC -> <-")
08:23:51 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
08:24:53 Join Harpy [0] (
08:31:23 Join oxygen77 [0] (
08:43:28 Join Bagder [0] (
08:52:53 Join einhirn [0] (
09:00:24 Quit oxygen77 ("CGI:IRC")
09:18:02 Join t0mas [0] (
09:18:22***Saving seen data "./dancer.seen"
09:27:54 Quit [zmaj] ("Serverwechsel")
09:28:09 Join [zmaj] [0] (
09:46:26 Join bobTHC [0] (
09:46:36bobTHChi mates !
09:47:02Bagderhi ho
10:02:31 Join martijn [0] (
10:11:55t0mashey martijn
10:12:26Bagdernow what is the difference between Holland and The Netherlands? ;-)
10:13:15t0masBagder: nothing... except that Holland has some association with weed ;)
10:13:34t0masoh and Holland is not the whole country...
10:13:45t0masit's the kingdom of the Netherlands... containing some islands too...
10:14:02t0masand Holland is split in 2 parts... Noord and Zuid Holland... and that's just the west part of the country
10:14:12t0mas(Holland is the part where Amsterdam is ;))
10:20:18*LinusN is even more confused
10:20:27 Quit martijn (Read error: 60 (Operation timed out))
10:22:18 Quit t0mas (Remote closed the connection)
10:26:03bobTHCand seems to confused them too, they leave the chan....
10:28:22*Bger defiantly stays in...
10:40:42 Join Mirfle [0] (
10:45:52*HCl rubs his eyes
10:46:30HCltodays the day :3
10:47:04 Quit Mirfle (Read error: 104 (Connection reset by peer))
10:47:17 Join preglow [0] (
10:47:57preglowshift key
10:48:22HClhey preglow
10:48:22preglowi'm insane, so i sometimes like to write the channel name in mixed case to see if anyone notices
10:48:27HClyou had a lot of oggs right
10:48:45HCli don't suppose you have java? :/
10:48:54preglowespecially not now
10:48:56preglow64 bit linux, etc
10:49:17HClthen i'll have to wait for markun to wake up
10:49:25HCli need someone with a lot of oggs to give my new tool a spin..
10:49:27preglowthink 64 bit ubuntu's got java problems
10:50:35 Join niobos2 [0] (
10:50:43 Nick niobos2 is now known as niobos (
10:52:00LinusNusing java is asking for problems ;-)
10:52:36*HCl stretches and doesn't see why :)
10:52:47*niobos 's favorite quote: Saying that Java is nice because it works on all OS's is like saying that anal sex is nice because it works on all genders
10:52:58crwlgetting java to work properly on a random linux box almost always requires significant amount of handwork
10:53:19HClum.. right o.o.
10:53:33HClanyways, i'm not gonna debate it.
10:53:33 Join Mirfle [0] (
10:55:05Bgerniobos lol!
10:55:30niobosstole it from
10:56:00Bgerheh, that's really good hit ...
10:59:26Mirflei've been following this project for awhile, I had a jbrfm for a couple months but then it broke so I got an Iriver, and now I can use your great firmware again :)
10:59:57Mirflehowever, it seems there is a bug with the remote:
10:59:59preglowcrwl: ahahahah
11:00:21crwlpreglow, what.
11:00:35preglowcrwl: misread, i mean:
11:00:39preglowniobos: hahahahah
11:00:52Mirfleif the main player is on hold, and you try to turn in the player with the remote (not on hold)
11:01:18Mirfleyou get a "hold" message, and the player shuts off
11:01:48LinusNMirfle: does that happen even if you hold the remote ON key a little longer?
11:01:58Mirfle*turn on
11:02:10Mirfleahhh, no.
11:02:17ashridahLinusN: i've been getting that one occasionally, when trying to start the original firmware from the remote as well.
11:02:22Mirflewhat's the difference?
11:02:32ashridahdoesn't seem to matter if i hold down record or on or what
11:02:35LinusNashridah: even with the latest boot?
11:03:14Mirflei just flashed the latest bootloader
11:03:39ashridahLinusN: i *think* i've gotten it with the most recent version
11:03:46ashridahi'm not sure. i'll keep an eye out
11:04:15LinusNthe problem is probably that the bootloader is configured to run the original firmware when it is started with the remote
11:04:17ashridahi mean iriver's firmware thinks the unit is on hold, btw, not rockbox's bootloader
11:04:59LinusNand it takes a while to reboot to the original
11:05:21LinusNand you have time to release the remote on button during that time
11:05:57LinusNso when the original firmware starts, it doesn't see the remote ON key anymore, thinking you started with the main ON key
11:08:26LinusNthat's unfortunately nothing we can fix
11:08:50ashridahheh. it's livable until rockbox becomes the only thing i want to start :)
11:08:51LinusNwhen rockbox matures, we will remove the start-original-with-remote feature
11:09:16 Join Dave99 [0] (
11:10:43amiconnpreglow: [10:48:48] <preglow> i'm insane, so i sometimes like to write the channel name in mixed case to see if anyone notices
11:10:58amiconn??? I don't see the channel name anywhere ???
11:11:10Bgeramiconn on join ?
11:11:38preglow10:47 -!- Irssi: Join to #rOCKBOX was synced in 1 secs
11:12:01*Bger guessed ;)
11:12:22ashridahHCl: i've got a mountain of oggs
11:13:00ashridah725 of them, infact
11:13:13crwli've got several thousand oggs, but not java
11:13:35crwlogg/mp3/flac ratio seems to be 3117/1379/518 right now
11:13:48ashridahand yes, i've got java
11:13:51preglowi rip all my own cvs to oggs, so i've got a bunch
11:13:51*ashridah pats his eclipse install
11:14:25amiconnBger, preglow: This is the only thing I see on join:
11:14:27amiconn[10:47:43] *** preglow ( joined
11:15:34Bgeramiconn it depends on client
11:16:05Bgerbut the IRC server sends channel name in the way it was typed by the join-er ...
11:16:13amiconnSeems so. So obviously I can't notice the weird case :-P
11:16:47 Part ashridah ("Leaving")
11:16:49 Join ashridah [0] (
11:16:53Bger[11:47:18] * preglow ( has joined #rOCKBOX
11:17:01Bger[12:16:50] * ashridah ( has joined #RoCkBoX
11:17:08ashridahcase insensitivty, oh yeah!
11:17:32Bgerashridah could you part again?
11:17:58 Part ashridah ("Leaving")
11:18:04 Join ashridah [0] (
11:18:25***Saving seen data "./dancer.seen"
11:18:25Bgeron leave it's the same case ...
11:18:32Bgerbut it's logical
11:22:29preglowsuppose so
11:27:47 Join martijn [0] (
11:37:35HClashridah: got java?
11:37:44HClokay, nice
11:37:48HCli was just adding some more features
11:37:51HCllike the strip, add
11:38:03HCllet me finish up and try to make a proper .jar out of it.
11:41:19*tvelocity is alergic to java
11:41:22ashridahwhat's wrong with java?
11:42:06niobosit's slow...
11:42:28ashridahwhat're you trying to do with it that you consider slow?
11:42:39HCli haven't run a benchmark yet. but as long as you don't use the java graphical interface, its quite fast
11:42:41tvelocityjava is REALLY slow... even on modern mashines...
11:43:04preglowjava isn't THAT slow
11:43:09preglowlike hcl says, the gui part is slow
11:43:14tvelocityand it's not as portable as sun is trying to convince people that it is
11:43:15niobosa friend was planning to make a EXIF-library in it, to extract jpeg-info, and it was soooo slow...
11:43:32nioboswell, exif lib + interface, actualy
11:43:48ashridahtvelocity: nothing new there, but then, it does a crapload better than a majority of things.
11:43:49tvelocityjava's portability actually SUCKS
11:43:52preglowmono is faster than java now, they say
11:44:12tvelocityi'll have to agree with them
11:45:30ashridahnow, i'll grant you that Swing is the arse end of computing.
11:45:41ashridahbut there's stuff you can use instead which mitigates that.
11:45:43HClashridah: i got a .jar for you..
11:45:48HClhold on..
11:46:28HCli know the .jar is pretty big, but thats only cause of my ogg/mp3 tag reading backend
11:46:37HCli'll strip it someday.
11:46:39ashridahhow big are we talking here?
11:46:58ashridahah. that's not really all that big :)
11:47:02HClnah, its okish
11:47:06HClbut still relatively big
11:47:48HClrun with java -cp SongDB.jar SongDB
11:48:21ashridahhrm, no manifest with the class name set in it?
11:48:24ashridahtsk tsk
11:48:31HClit does support strip, add, dirisalbum, dirisalbumname, but the commandline interface doesn't handle options yet
11:48:35HClso its hardcoded at the moment
11:48:40HClhehe. i don't know how to do manifest yet ;p
11:49:03HClits set to create a v2 database at the moment
11:49:11HCli'm really pleased with java using big endian as default
11:49:35preglowinternally as well?
11:49:40HClit also tried to output song names in unicode, but i turned that off ;p
11:49:47HClpreglow: nah, i meant when writing binary files
11:49:52ashridahFailed reading tag for /home/ashridah/atec/cliche-01-phoebus_apollo_cox_edit.ogg, tried mp3 and vorbis.
11:50:10ashridah$ file /home/ashridah/atec/cliche-01-phoebus_apollo_cox_edit.ogg
11:50:12ashridah/home/ashridah/atec/cliche-01-phoebus_apollo_cox_edit.ogg: Ogg data, Vorbis audio, stereo, 44100 Hz, ~128031 bps, created by: Xiph.Org libVorbis I (1.0.1)
11:50:14HClas long as you don't get too many of those..
11:50:20HClor do you get it for every one?
11:50:23ashridahi got it for ALL of the ones i tried :)
11:50:26HCli might've forgotten a library
11:50:31HClcan you send me some oggs?
11:50:31ashridahwhich, admittedly, was just a small sample :)
11:50:40ashridahalso got a NullPointerException at the end
11:50:43ashridahHCl: dude, i'm on dialup
11:50:44HCl :p
11:50:49HCli'll get an ogg and test :p
11:50:54preglowi can get you oggs
11:51:03HClnah, its ok
11:51:19preglowbut i need to get some work done, alter
11:51:21preglowlate r<-
11:51:22 Part preglow
11:51:33ashridahHCl: yeah, if i run it on a bigger directory, it immediately dies with a NullPointerException
11:51:41HClgot a trace for me?
11:51:49ashridahException in thread "main" java.lang.NullPointerException
11:51:51ashridah at TagDatabase.calcLimits(
11:51:51ashridah at TagDatabase.writeDatabase(
11:51:51ashridah at SongDB.main(
11:52:45HClthats caused by it not getting any tags at all.
11:53:32ashridahwell, all the oggs were in subdirectories
11:53:35HClits not handling that well yet
11:53:39HClyea, it handles subdirectories
11:53:43HClthe backend is failing
11:53:50HCli might've forgotten a library that ogg needs
11:55:50bobTHCme too :
11:56:03HClno, it seems like something went wrong when i tried to merge the .jars i had and the .classes into one .jar
11:57:02HCl :X how do you properly merge those? :/
11:57:13bobTHCi have a directory with only Mp3 files and i had the same error than ashridah
11:58:13amiconnHCl: Just implement it in perl...
11:58:35ashridahcpan install humungousmessofunfinishedtosh
11:59:01amiconnThe current doesn't need any modules
11:59:34*HCl bites his capslock
11:59:47*ashridah huggles his capslockless keyboard
12:00:48Bgerashridah hahaha
12:04:49 Join niobos_ [0] (
12:05:11 Quit niobos ("using sirc version 2.211+KSIRC/1.3.11")
12:05:19 Nick niobos_ is now known as niobos (
12:06:09*HCl found one of his bugs.
12:06:52BgerHCl: maybe the bug is in the language u're using ;)
12:07:09HCljava is *really* bug free, actually
12:07:30HClits just that my filename filter is searching for .ogg and .mp3
12:07:33BgerHCl there's no such thing as bug free
12:07:35 Quit tvelocity ("Leaving")
12:07:37HClbut ends up stripping directories
12:07:44HClBger: in other languages, there isn't, no ;p
12:07:51BgerHCl i was just joking
12:13:02HClthere we go
12:13:10HCli was missing a couple of libraries...
12:13:44HCli really need to strip it, heh.
12:19:47gromit`< HCl> +java is *really* bug free, actually
12:19:52HClSystem.out.println(f.getAbsolutePath()+" directory: "+f.isDirectory());
12:20:00gromit`there were many security holes reveales quite recently
12:20:00*HCl prods putty
12:20:17HCli more meant code stability
12:28:41 Join Mirfle_ [0] (
12:29:44*HCl prods ashridah
12:30:14HCljust checking if you're still awake
12:30:16HCllet me upload this
12:30:19HCland you can give it a spin
12:30:30Bgerwhy is this banlist ?
12:31:05HClyou can run it with java -jar SongDB now
12:31:12HClSongDB.jar *
12:31:36HClit should do vorbis, emphasis on should
12:32:05 Quit Mirfle_ (Read error: 104 (Connection reset by peer))
12:32:52HClhaven't tested it
12:32:59HCli fixed the directory stuff too.
12:33:10HClit was stupidly filtering them out because they didn't end on .mp3 and .ogg
12:33:12HCl ;p
12:33:45ashridah$ java -jar SongDB.jar atec/
12:33:49ashridahArtist with longest name (8) :Carl Cox
12:33:49ashridahArtist with most albums (1) :Carl Cox
12:33:49ashridahAlbum with longest name (24) :At The End Of The Cliché
12:33:49DBUGEnqueued KICK ashridah
12:33:49ashridahAlbum with most songs (12) :At The End Of The Cliché
12:34:28HClseems to work nicely, here at least
12:34:41ashridahheh. needs a progress counter
12:34:55HClyea, it seems slow :/
12:34:58HCli'll look at that later
12:35:32ashridahand then it just wrote "ioexception" to the screen a few zillion times :)
12:35:35ashridahreal descriptive there, btw :)
12:35:42HClwhat happened there?
12:35:44HCl :)
12:35:49CtcpIgnored 2 channel CTCP requests in 9 seconds at the last flood
12:35:49*ashridah gives HCl the cookie of crap error handling
12:35:51HCl :P
12:36:02HCli didn't really expect to get ioexceptions
12:36:17HClread only directory? faulty files?
12:36:19 Quit Mirfle (Read error: 60 (Operation timed out))
12:36:22HClbad sectors? o_o
12:36:39ashridahHCl: it's on a fat filesystem, so it's not going to be a read-only directory.
12:36:51ashridahand hard to say, since it doesn't tell me What it was trying to do at the time
12:36:58HCllet me check the code
12:37:02*ashridah hands HCl another instance of the cookie of crap error handling
12:37:05HCl :P
12:37:21HCli found the crap error handling
12:37:23HCllet me adjust it :P
12:37:52ashridahof course, since many of my tags are dodgy, a lot of the stats it tosses at the end suck :)
12:38:05HClhow so?
12:38:07ashridahbut it does finish
12:38:13HClthe errors you got
12:38:18HClwere ioexceptions while trying to read a tag.
12:38:31HClbit odd, i didn't really expect that to happen
12:40:15HClokay, reget it off my ftp
12:40:20HClit should poop out proper error handling now
12:40:24HClbut you'll probably get flooded more
12:40:37HClcause now it prints a stacktrace for every one of those
12:41:51HClashridah: you should be able to use the database browser in rockbox now, with your oggs..
12:42:31ashridahwhoa. just dialed up again and got exactly the same ip
12:42:39HCl :p
12:42:44ashridahso, download again?
12:42:55HCli made it print a proper stacktrace
12:43:16BgerHCl URL ?
12:44:09Bgerhow to run it ?
12:44:17HCljava -jar SongDB.jar
12:44:22ashridahjava -jar SongDB /directory/
12:44:31ashridahSongDB.jar rather
12:44:44HCli'll add the interface for the options soon, but after this works semi decently, i'm gonna shower and make breakfast first
12:45:23ashridahahahaha JOrbis: VorbisFile: /usr/local/archives/common/Music/SpineShank/The Height of Callousness/09 - Seamless.ogg (Too many open files)
12:45:30HCl :P
12:45:41ashridahyeah, close them :)
12:45:46HClyea, i wonder how :P
12:45:56HCl*digs into the code for readogginfo*
12:46:27ashridahdon't just toss away the file descriptor, close it, THEN toss it away
12:46:30HClyea :P
12:46:41ashridahyou'll find often stuff doesn't get flushed to disk if you forget to close it
12:46:41HClits supposed to do that
12:46:44HClits a bug in my backend
12:46:46ashridah(when writing, anyway)
12:47:07niobosshouldn't tho oh-so-wonderful java-garbagde-collecotr destroy the object and close the file himself???
12:47:50niobosok, drop the "oh-so-wonderfull"
12:48:23ashridahniobos: there's no garuntee in any part of java that the destructor will run
12:48:39ashridahclosing files yourself is a) good manners and b) smart programming.
12:48:50niobosI know...
12:49:06niobosbut it would be nice to catch this in the garbadge collecotr...
12:49:07ashridahbecause even if the garbage collector kicks in properly, there's still a filesystem limit, and you might beat the garbage collector to the punch
12:49:16 Join Chamois [0] (
12:50:33ashridahi don't see how closing the file is the garbage collector's job. as i say, garbage collection isn't necessarily going to kick in.
12:51:15niobosok, I agree
12:51:30HClyea, and its my backend not closing the file..
12:51:39amiconnHCl: Didn't you say java is inherently bug-free? :-P
12:51:41HClfortunately it can also read from an inputstream
12:51:45HClamiconn: it is.
12:51:56HClits a bug from the coders who made the backend
12:52:06HClits part of a music player written in java
12:52:08HClnot java itself
12:53:11 Quit ehntoo ("Leaving")
12:53:29rasherGuess there'll be no id3-db browsing for me then :-<
12:53:41rasheror whatever the name is now
12:53:46HClwhy? :/
12:54:00rasherWell you're writing this in Java, which makes me wonder about your sanity
12:54:28BgerHCl what about GNU Java compiler ?
12:54:46 Join zeeeeeeee [0] (~Miranda@
12:54:53rasherWhat about just extending the existing script to reading vorbiscomments instead.. I honestly don't get the point of this exercise
12:55:04amiconn[12:35:23] <HCl> yea, it seems slow :/ <== didn't you also say java isn't slow?
12:55:29HClBger: should work
12:55:46HClamiconn: yea, it shouldn't be.
12:56:33rasher <−− this should do it
12:56:44Bger (pasv)
12:57:50HClrasher: the problem isn't reading ogg tags.
12:58:00 Join _DangerousDan [0] (
12:58:00rasherWhat is it then..?
12:58:07HClthe problem is the code getting *completely* unreadable and unmaintainable
12:58:37rasherassuming you code unreadable and unmaintainable code
12:58:45HClhave you seen
12:59:04HCltry implementing reading an old database in before manipulating it
12:59:06HClgood luck
12:59:06HCl :/
12:59:10*Bger thinks that the link given by rasher is very maintable ...
12:59:24Bgerthe example in the link i mean
12:59:34ashridahBger: that makes it highly unusual for a cpan module
13:00:01HClashridah: can you give the one on my ftp a spin..?
13:00:06HCli should've fixed the open file errors
13:00:30HClthe bug bger reported should also be fixed
13:00:43Bgeri'll try it again ;)
13:01:08BgerHCl: nope ...
13:01:22Bgerbtw, are you searching only in files .mp3 & .ogg ?
13:01:25rasherI still think using Java for such a functionality is shooting yourself in the foot
13:01:42HClBger: are you sure it redownloaded? windows has this nasty habit of caching ftp files
13:01:47HCleven though they have actually been updated
13:01:52Bgergrr w8
13:02:01HCli hate windows for that
13:02:03Bgeri'm making bul....ts :)
13:02:11BgerHCl i'm using wget ...
13:02:13HCli use a windows port of lftp...
13:02:25Bgerbut i haven't downloaded it again (don't ask me why)....
13:02:49*Bger is wandering somewhere
13:03:00HClthe backend i got isn't entirely clean
13:03:09HClat some point i'll probably want to rewrite it to c++
13:03:26HClonce everything works, including runtimedb and stuff.
13:03:32Bgergrr... rm SongDB.jar under cmd.exe ....
13:04:11rasherdoing it in C(++) sounds a lot more sane
13:04:53ashridah$ java -jar SongDB.jar /usr/local/archives/common/Music/ mark/reset not supported
13:04:55ashridahalmost immediately
13:05:05HCleasy fix there..
13:05:07HClhold on.
13:05:22 Join preglow [0] (
13:05:26preglowdoing it in perl sounds even saner :P
13:05:30BgerHCl worse ...
13:05:34Bgerit's showing this non-stop
13:05:42HCl :p
13:05:57HClsorry, should've tested it myself
13:06:04preglowwould anyone care massively if the module were taken out of
13:06:05rasherpreglow: well doing it in C would eliminate the dependency oon perl, which isn't that great for windows users
13:06:05HClapparently the backend wants a buffered inputstream
13:06:19*ashridah urinates on perl
13:06:21preglowpasting all the modules into one file is not very maintainable
13:06:30ashridahthe sad part is, perl's so masochistic, it likes it
13:06:31preglowrasher: doing it in c would also be a major bother :P
13:07:04preglowrasher: but of course, if it were written carefully in c, we could also use the same code for generating the db on the player
13:07:20rashernow that's good thinking
13:07:26Bgerreally good. ..
13:07:37HCloh for crying out loud.
13:07:45preglowit wouldn't be THAT much of a bother, really, but we'd need good libs
13:07:47HCli don't like my backend o.o
13:08:48*HCl scratches his head.
13:09:00preglowoh pretty, pretty please
13:09:04HClwell. i think this error is semi-okish, and just caused by a not proper mp3
13:09:05preglowsomeone make a good audio editor for linux
13:09:20preglowyou can't count on mp3s being proper
13:09:21BgerHCl which one Resetting to invalid mark
13:09:34HCl at Source)
13:09:34HCl at org.tritonus.share.sampled.file.TAudioFileReader.getAudioFileFormat(T
13:09:45amiconnpreglow: The module is pasted for a reason, which is also stated within the script
13:09:53HClmy backend isn't good with handling broken files, but yea.
13:10:07HClit might poop out some exceptions on bad files, but let me toss this one online for testing
13:10:50BgerHCl the last version writes out this on nearly every file ...
13:10:59HClthats correct
13:11:11HClredownload now
13:11:13HCland try
13:11:23HClit might give some "resetting to invalid mark" exceptions
13:11:25HClbut it should work
13:11:26preglowyou distribute your own perl.exe anyway, and perl can be compiled to look in a default dir for modules
13:11:59 Quit edx (Read error: 110 (Connection timed out))
13:12:18preglowsomeone do this in c, please :V
13:12:24HClpreglow, sure.
13:12:42pregloweverything else is either a flaming hack or a major bother
13:12:45HCli can do this in C, but i like to get it straight in java first
13:12:55preglowthen hooray
13:12:58HClit shouldn't be hard to port it to C, as long as someone can provide me with a tag reading backend
13:13:13preglowwhy do you call it a backend? it's a tag reading library you want :>
13:13:17amiconnpreglow: (1) This perl.exe isn't a special build. (2) What default dir do you think should be selected?
13:13:21HClsame thing
13:13:27preglowamiconn: \.rockbox\perl !
13:13:27Bgerpreglow maybe Java-ish style ;)
13:13:33niobosHCl: shouldn't be too hard.. I wrote a complete ID3v2 edit-lib in about a week (3 years ago...)
13:13:36HClashridah / Bger: test latest version?
13:13:58BgerHCl there are also errors, but not as much as in previous "release"
13:13:59 Quit DangerousDan (Read error: 110 (Connection timed out))
13:14:00HClniobos: mmm, but before anything, i want to get it completely straight in OO
13:14:05HClBger: which? invalid mark?
13:14:17HClyea, i got one of those too.
13:14:23niobosHCl: C++ IS OO
13:14:25HClbut it was on a file without a proper tag anyways
13:14:30HClniobos: i'm well aware of that.
13:14:36HClbut java is *much* nicer to code in
13:14:48preglowi dunno
13:14:50 Join oxygen77 [0] (
13:14:51HCli highly prefer debugging in java
13:14:52amiconnpreglow: The point is that the way it is now, you can put the .pl (and the .exe for windows) on your player, and execute it on any machine, regardless whether linux or windows and regardless how the box is mounted (on linux) or which drive letter it gets (in windows)
13:14:54HClthan c or c++
13:14:58preglowc++ is a bastard, but i like prefer it to java
13:14:58niobosanyway, If you need help in that library coding, let me know
13:15:29preglowtag reading libs should be out there somewhere
13:15:35Bgerid3lib ?
13:15:40preglowwe need more than id3
13:15:42HClthere are.
13:15:46HClthere was one written in c++
13:15:49Bgeryeah, i know
13:15:54rasherI'm sure there are like a gazillion for each format
13:15:54HClsupporting mp3, ogg, flac
13:15:57preglowwe'll need id3, vorbis tags, ape tags, flac tags, etc
13:16:03HClyea, there was one like that..
13:16:09BgerFLAC is vorbis afaik ?
13:16:11rasherdoesn't flac just use vorbistags?
13:16:15Bgeris with vorbis tags
13:16:16preglowBger: think you can choose
13:16:22preglowbut might be wrong
13:16:23rashervorbiscomment, I believe actuallly the name is
13:16:28*HCl prods ashridah
13:16:39Dave99are these other tag formats similar to id3?
13:16:47amiconnpreglow: This would not work if it would rely on external modules, regardless of a possible default path for modules
13:17:36preglowno, that's true
13:17:36preglowapart from the default ones, of course
13:17:38preglowbut we'd still need to have the module sitting right next to perl.exe or pasted into
13:17:50preglowANYWHO, someone should do it in c :]
13:17:58HCli'll do it in c
13:18:03HClafter its completely finished in java.
13:18:06preglowthen again, hooray!
13:18:26***Saving seen data "./dancer.seen"
13:18:42preglowjust keep in mind that you'll be severly punished if you lose interest after writing the java version
13:18:47HCl :P :P
13:19:36HClthat chance isn't non-existant though ;x
13:19:51preglowi know
13:20:00preglowbut if you love your knee caps it should be
13:20:09*HCl continues prodding ashridah with a stick
13:21:48 Quit _DangerousDan (Read error: 104 (Connection reset by peer))
13:24:36HClBger: can i toss you a new songdb.jar, and can you try to use the database on rockbox?
13:24:46preglowSlasheri: what up, yo, commit! commit!
13:24:55BgerHCl i
13:25:00Bger'm H340 owner ....
13:25:06HClokay :X
13:25:40Bgerbut if Linus give me a bootloader ... :P
13:25:58HCl :p
13:26:07Lynx_HCl: you switched from perl to java for the db generation?
13:26:47HClfor now, yes
13:27:03preglowdon't tar and feather him quite yet
13:27:05HCli need an OO language to keep it maintainable
13:27:43HClhas anyone tested slasheri's patch with the database browser? does that work?
13:27:49Bgerthere's one phrase ... the temporary things are most permanent ...
13:27:51preglowyou really can't write maintainable code in c ?
13:28:07HClnot with the ease of java, no.
13:28:19HClthe only reason i managed to write this overnight
13:28:25HClis because java is so nice to debug
13:30:31LinusNeveryone can be efficient in their favourite language
13:30:42LinusNjava happens to be yours
13:30:59LinusNperl happens to be mine
13:31:10LinusNor c
13:31:11LinusNor asm
13:31:20HCldo you know perl oo ?
13:31:32LinusNi haven't dealt with it much
13:31:41HClk ;/
13:32:17LinusNi wonder how programmers managed before oo was popular
13:32:27HClcruddily :)
13:32:56LinusNback then the silver bullet was "structured programming"
13:34:07HClwell, using structures can be very much like oo
13:34:16HClwhich is exactly how i plan to make it in C
13:34:25HClor perl. depending on the readability of perl OO (for me)
13:34:53LinusNmy point is that anyone can write a crappy program, regardless of oo
13:35:01LinusNit will just be a crappy oo program
13:35:34LinusNi can go on and on about this'
13:35:47LinusNi'd better stop and do something useful
13:36:57HCl :p
13:37:01HCllike prodding slasheri
13:37:02HClto give his patch
13:37:11LinusNfor example
13:37:43*ashridah suspects many a crappy programmer is going to come out of his year when it graduates :(
13:40:37HClits awake!
13:40:39*HCl prods ashridah
13:40:48HClashridah: get the last version, and try it on your rockbox?
13:40:53HClsee if database browsing works with oggs
13:44:13ashridahsorry, been distracted by a CMS system i'm poking at out of boredom (you know, instead of studying for my exams)
13:44:16 Quit lostlogic ("Going to the moon")
13:44:39ashridahhm, CMS system. ATM machine. oops
13:45:07HClokay, test it? :p
13:45:36ashridahstill leeching
13:45:44HClk :)
13:45:47preglowperl is most definitely my favourite language
13:45:59ashridahas CMS's go, xoops isn't too bad.
13:46:01HCl :)
13:46:05*HCl adjusts code a little..
13:46:11LinusNdid markun finish his margin-scroll patch?
13:46:14ashridahperl can die. along with perl's creator, and every single one of perl's users.
13:46:21ashridahwith the possible exception of LinusN
13:46:24ashridahbecause we need him :)
13:46:35*LinusN runs for his life
13:46:46HCl ;p
13:46:50*preglow draws strength from perl and prepares to counterattack
13:47:00Bgerrun, LinusN, run, run to the hills ... ;)
13:47:05LinusNuse the source luke
13:47:29LinusNrun to the hills, run for your liiiife
13:47:33Bgeryep :)
13:47:47*HCl asks markun
13:47:54ashridahHCl: still getting exceptions
13:47:58*preglow goes back to his jethro tull binge
13:48:00HClashridah: invalid marks?
13:48:19HClthose are "normal", its a bug in the backend, means tag reading failed
13:48:23ashridahresetting to invalid mark
13:48:28HClLinusN: no, not yet
13:48:40HClit should poop out a correct database regardless
13:48:43HClcan you try it on your rockbox?
13:48:54ashridahwhere do i put it?
13:49:06HClin .rockbox
13:49:14HCli hope the paths are okay and stuff
13:49:24HClcan you toss it in as well?
13:49:33HCljust want to analyse it briefly
13:49:35*niobos is off to lunch
13:49:45ashridahhow do i use it?
13:49:56HClLinusN: there's a small bug in changing fontsize, but its finished aside from that.
13:50:05HClashridah: file view -> id3 database
13:51:22ashridahi don't think i have that option
13:51:35*HCl checks his iriver
13:51:37HCldid you reboot it?
13:51:59HClgeneral settings -> file view -> show files -> id3 database
13:52:08CoCoLUSrockbox.id3db is ~6mb
13:52:13CoCoLUSwith -a-lot- of zeros in it :)
13:52:21HClthat can happen, yea..
13:52:34ashridahit's only 760kB here
13:52:40HCllike i have an album with 715 songs
13:52:42HClcausing all other albums
13:52:49HClto take up at least 715*4 space
13:53:22CoCoLUSis that the last line of the songdb output? :)
13:53:24CoCoLUS"Album with most songs (1276) :neu" ?
13:53:32HCl1276 :P
13:53:35HClmore than me :P
13:53:39ashridahviewing the "Songs" option seems kinda crappy
13:53:48HClashridah: mm? what happens?
13:53:54ashridahshows lots of music icons, but only about 1 in 20 or so actually have text, and it's the END of the text
13:54:03HClthats a bug. hm..
13:54:03CoCoLUSyeah the folder only contains single songs
13:54:20CoCoLUSdunno why that is considered as an al
13:54:24CoCoLUSbum anyway :P
13:54:24HClashridah: can you send your database to me..?
13:54:58ashridahyeah, hang
13:56:21*HCl goes to upload his test database to his iriver to check..
13:56:56ashridahdone. incoming/
13:56:58HClmine checked out on my database checker, at least
13:58:20HClyea, your database is definately broken
13:58:26HClthats rather odd.
13:58:39HCli don't see how that could've happened :/
13:58:47*HCl checks his own
14:00:59HClnope, i got it too. hm.
14:04:57HClokay, i got the bug, not sure how to fix it yet..
14:06:05*HCl fixes..
14:06:15 Join webguest12 [0] (
14:06:35 Quit webguest12 (Client Quit)
14:12:21Bgerhey ... have you ever heard about coLinux ?
14:12:29HClokay, i'm starting to get pretty annoyed at java :P
14:12:38BgerHCl :P
14:12:43HClits utterly fucking up when trying to get a songcount
14:12:51HCland i have no idea why
14:13:47HCleverything works, but its saying it has more songs than it actually has
14:14:36HCli even did a manual count, and it still manages to fail, i have no idea how it manages that
14:16:13*HCl knows.
14:16:40 Join tvelocity [0] (
14:17:28HClmy song comparer says that if two songs have the same title they're the same song, which is obviously not the case.
14:18:51ashridahi'm pretty sure that's not the case with my music
14:22:16 Join ripnetuk [0] (
14:25:52HClit was my sloppy coding that caused it :)
14:26:15HCllet me just fix one more thing..
14:27:45HCli'm going against java a bit :/
14:28:33HCllet me upload it
14:29:07HClon my ftp
14:29:09HClgive it a try..
14:29:34 Quit ripnetuk ()
14:30:35*HCl fixes bitrate..
14:31:27 Part Mr_Wik_ ("Leaving")
14:31:34HClmental note; get a decent sort algorhythm, instead of putting it in a treeset
14:32:11ashridahi agree
14:33:57HClthe set property is throwing out "duplicates" which are equal according to the comparator, which defines the way it sorts
14:34:10HClwhich was causing the bug
14:35:26ashridahstill broken
14:35:33HClugh :/
14:35:39HCltoss me your database?
14:35:46*HCl is close to giving up the java version o.o;
14:35:56HClnot that its any easier
14:35:58HClin any other language
14:36:58 Join Aramil [0] (
14:37:38 Quit tvelocity (Read error: 110 (Connection timed out))
14:40:09HClcan you put your database in incoming?
14:40:17 Nick Aramil is now known as tvelocity (
14:42:50HClgot it..
14:42:52HClsame bug
14:42:55HCli need a proper sorter..
14:46:12ashridahjust rm -r'ed my build tree
14:46:49*Slasheri catches digital flow with the new ring buffer :)
14:47:07Slasheriand got illinstr after that ;)
14:47:20 Join webguest94 [0] (
14:47:47ashridahyay! even
14:48:59bobTHCHCl > this algo is not bad for sorting :
14:49:17SlasheriLinusN: exactly :)
14:50:37HCl :P
14:51:21tvelocityso.... (/whinner mode on) are you gona commit soon?:P
14:51:48Slasheritvelocity: i will release the patch as soon as i got this issue fixed :)
14:52:05tvelocitywhich means "when it's done" right? :D
14:52:33Slasheriit's not done after that but should be ready for commit
14:52:57tvelocityi think I should donate
14:53:22 Join edx [0] (
14:53:27HClashridah: new version up, with new sorting algorhythms :/
14:53:35HCli do hope this one will work now :/
14:55:31HCloh wait
14:55:43ashridahmight take me a while to get it. checking out a new copy of rockbox-devel
14:55:45HClthere, now its up
14:55:58HClthats okay, i had forgotten to copy the correct version from incoming
14:56:02ashridahheh. my system hadn't finished logging into your ftp :)
14:57:30webguest94Can anyone confir/deny: I'm upgrading HDD on recorder10 to 40mb
14:58:12webguest94I formatted fat32 on w98, then copy all files to new hd from old hd. Anything else?
14:59:33LinusNsounds about right, provided you partitioned it correctly
15:00:23webguest94? iI just made one big fat32
15:00:43ashridahHCl: seems to work now
15:00:58HCli was stressing there for a bit
15:01:09HClcan you look around everything see if there are any bugs left?
15:01:12HClany glitches
15:01:32HCli'm never using a treeset for sorting again ;p
15:01:42HClit was removing "duplicates"
15:02:04ashridahwell. no. since i just made the mistake of selecting a song in the Songs list
15:02:12ashridahand now the hard drive is grinding non-stop
15:02:17HCl :x
15:02:30ashridahhm. no. drive stopped. player's still hung tho
15:02:34LinusNwebguest94: does your hd have a partition table with one primary FAT32 partition?
15:03:18webguest94yes, exactly
15:03:35LinusNwebguest94: that should be it
15:03:42LinusNis there a problem?
15:04:30 Part LinusN
15:05:02webguest94don't know yet, I'm still copying mp3s off of cdroms to take advantage of the faster xfer speed while it's on the ide adapter. The recorder I have is only USB1.1
15:05:13ashridahHCl: yeah. trying to play a song definently doesn't work that well :)
15:05:16ashridahbut the list seems to look okay
15:05:22HClok :)
15:05:23ashridahgeneration seems to take a fair whack less time too
15:05:30HClit does?
15:05:32HClcompared to perl?
15:05:45HClwait o.o perl didn't do ogg.
15:05:48HClwhat? o.o
15:05:52ashridahcompared to earlier versions :)
15:06:02HCli wouldn't know, i didn't change that much
15:06:05ashridahthe new sorter sucks less
15:06:11ashridahreal 0m9.595s
15:06:11ashridahuser 0m7.795s
15:06:11ashridahsys 0m1.616s
15:06:14ashridahbefore it was taking ages
15:06:32ashridah(of course, i foolishly don't have a copy of an older version)
15:07:14ashridahbut anyway, my collection has 2054 files, of mp3 and ogg
15:07:19HClnot bad
15:07:30HCllet me try to run it on my iriver
15:07:37HCli don't have ogg
15:07:42HClit should fix all the bugs of the old songdb
15:08:11ashridahit definently displays entries for oggs
15:08:20ashridahof course, still tosses exceptions for the files that don't have tags
15:08:28HCli'll get that debugging out soon
15:09:09HCland i do need a progress thing like the old one :p
15:12:04HClah, its done
15:12:17HClsorting :X
15:13:47HClbah, it still has that yPW bug
15:14:00*HCl whacks himself upside the head XD
15:14:07HCli'm dumb o.o
15:14:20HClnot copying the new database to my iriver :p
15:14:57ashridahif i change the volume a chunk in rockbox, why does it keep slowly raising or lowering after i STOP ?
15:15:28HClno idea..
15:15:33*HCl bites java
15:15:41HClits failing to sort songs..
15:16:30 Join RED_M_CHIU [0] (
15:17:57ashridahyou're right. it's just output them in read order
15:18:27***Saving seen data "./dancer.seen"
15:18:58HClyea, let me fix stuff..
15:25:58HCli think its faster than the perl one, not 100% sure yet
15:26:03HClits pretty fast
15:26:06HCli added progress
15:26:23HClits going throug it pretty fast
15:27:44HClhm, thats odd.
15:27:52HClit sorted really quickly, but it stalls at the end
15:27:56HCli thought it stalled cause it was sorting
15:28:11 Join spiral [0] (
15:28:25 Join asdsd____ [0] (
15:28:44ashridahwhich sorting algorithm are you using?
15:28:45*HCl will have to find out why it stalls in the end
15:28:50HClno idea. the java default one.
15:28:54HClits not the sorting thats slow
15:29:04HClit just. almost skipped through the sorting as if it didn't sort
15:29:08HClmaybe it didn't.
15:30:14*niobos is back
15:32:54HClit sorted
15:32:55HCl :)
15:33:22HClno idea why it didn't before, but i rewrote it and it should be faster too.
15:33:42HCllet me upload it
15:33:53bobTHCyou use the tuboBM sorter ?
15:34:00HCljust the java default one
15:34:03HCland its taking near 0 time
15:34:07HClso its more than fast enough
15:34:14HClthe default java one is *not* bubble :P
15:34:23HClits probably mergesort or something like that
15:34:37ashridahyeah, bubble sort doesn't scale too well :)
15:34:51HClmergesort is better than quicksort
15:35:05niobosdonno... I learned in school that quicksort was the quickest...
15:35:07HClashridah: can you give the current one a try?
15:35:09bobTHCbut the best is Turbo BM
15:35:13niobosdidn't learn mergesort...
15:35:16HClniobos: yea, but then mergesort came along :)
15:35:21HClmergesort is very much like quicksort
15:35:27HClonly, rather than choosing a random pivot point
15:35:30HClyou choose the middle
15:35:39HClso it has an average complexity of n*log(n)
15:35:42niobosmiddle element? or middle value?
15:36:12HClquicksort takes an element, and splits the list into elements that are lower than that element and higher than that element
15:36:15ashridahyeah. you keep merging with larger and larger chunks
15:36:21HClit has an average complexity of n*log(n)
15:36:27HClbut a worst case complexity of n^2
15:36:37HClmergesort has both average and worst case n*log(n)
15:36:46HClashridah: check out the last version?
15:36:47niobosi'll google for it
15:36:54ashridahHCl: leeching atm
15:36:56HClk :)
15:37:01HClit should be much userfriendl
15:37:05HClmostly due to printing progress
15:37:10ashridahniobos: wikipedia probably has a decent description
15:37:31niobosashridah: k, i'll look there... after this chapter of electromagnetism :-(
15:38:35ashridah12 seconds total
15:39:06ashridah7 seconds user
15:39:13ashridahnot too shabby
15:41:13preglowthere is no "best" sorting algo
15:41:21preglowit all depends on your data
15:41:34ashridahyeah, it's a time v memory tradeoff
15:46:35HClashridah: check that its sorted now?
15:47:00 Part asdsd____
15:48:14ashridahit is
15:49:05HClsorting songs within albums still needs to be done
15:49:12HClbut i think its pretty ok
15:49:20HCli need to add the options..
15:55:31*HCl goes to fix sorting songs within albums.
15:57:08 Nick ashridah is now known as Lost-ash (
15:57:08DBUGEnqueued KICK Lost-ash
15:58:47HClsongs in an album need to be sorted by track, correct?
15:58:58HCland if there are no track numbers.. filename?
15:59:03Lost-ashimho, yes
16:00:36HClreturn s1.getFile().getFile().getName().compareTo(s2.getFile().getFile().getName());
16:00:39HCl :P
16:00:50preglownow THATS clarity
16:00:59preglowjust because fucking sun doesn't like operator overloading
16:01:22HClcompareTo is actually a function that returns -1, 0 or 1
16:01:29HClso it wouldn't be an operator anyways
16:01:34preglowwhy not?
16:01:45preglowcould have made the operator return those as well
16:01:57HClbecause it returns a value saying its less, equal, or greater
16:02:14HClall thats left is adding an commandline interface for the options
16:02:22 Quit webguest94 ("CGI:IRC (Ping timeout)")
16:02:23preglowno matter, i still hate java for not having operator overloading
16:02:24HCland then i can move on to extending it
16:06:02Bgerpreglow, yes, that's silly
16:06:25preglowespecially when they claim not having it helps readability....
16:06:44preglowMatrix lol1 = lol2.Mult(lol3); is a lot clearer than Matrix lol1 = lol2*lol3; surely
16:07:37HCli dunno.
16:07:50HClcoming from the java corner, i'd prefer the first
16:07:55HClcause it shows that a function is getting called
16:07:59HCland its not a small operation
16:08:20Bgerbut that's one of OO's advantages ...
16:08:27Bagdercoming from a corner "don't hide what's happening", I prefer the first one too
16:08:43HClBagder: yea, that was the idea behind not having operator overloading in java
16:08:44BgerBagder i don't think that it's hidden ...
16:08:47Bagderbut then I dislike most OO stuff that hides
16:08:58preglowunless you're braindead, you know something's happening there
16:09:06Bagderin fact, OO is pretty much made to hide where things happen
16:09:12Bgerit's obvious that it's not normal multiplication
16:09:19BgerBagder that's what i mean...
16:09:46niobosOO is made to make it easyer... if I want to multiply 2 things I use *; I don't care HOW that multiplication is done, just DO it
16:09:49Bgerif you don't like hiding things, you don't like and OO
16:10:11preglowhaving to spell out in bloody detail everything you do is why i hate languages like python
16:10:31preglownothing is hidden, alright, but it makes me want to kill myself
16:10:46Bgerand why like languages like asm (especially coldfire's one)
16:11:02 Quit oxygen77 ("CGI:IRC")
16:11:08 Join t0mas [0] (
16:11:30tvelocitypython is my preciouussssss:P
16:11:51crwlwhat about python interpreter as a rockbox plugin :PP
16:12:17tvelocitythat would truly ROCK... would make writing plugins extremely easy!
16:12:27preglowover my dead body!
16:12:31preglownow a perl interpreter...
16:12:34 Quit t0mas (Client Quit)
16:12:36tvelocitybut propably won't make any sense perfomance wise:P
16:12:41niobos< preglow> nothing is hidden, alright, but it makes me want to kill myself
16:12:44niobos< preglow> over my dead body!
16:12:49nioboshmm... coincidence?
16:13:46tvelocityi think a custom BASIC language would make sense... more expierienced coders could use pure C anyway...
16:14:04HCldo we want unexperienced coders making plugins?
16:14:10tvelocityok, i'm just a n00b, now i'll stfu :)
16:14:26Bagderfeel free to make one
16:14:31Bagderwe can have all those languages
16:15:32nioboskeep your voice down... or we'll be having LISP-plugins soon...
16:15:40CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
16:15:40*preglow gets rope
16:15:44tvelocityhey how about MONO?
16:15:46niobos((I) (can't) (stand) (all) (those) ()())
16:15:51Lost-ashlua lua lua lua lua lua lua lua!
16:15:54 Nick Lost-ash is now known as ashridah (
16:15:54DBUGEnqueued KICK ashridah
16:16:07tvelocitywe can't do without bash, can we?:P
16:16:08CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
16:16:08*ashridah nods
16:16:12crwlbrainfuck or whitespace?
16:16:13*niobos just missend an oportunity to stfu
16:16:16*ashridah does the grep/awk/sed
16:16:16preglowand then everyone went mad
16:16:21Lynx_well, a simple scripting system for all rockbox functions would be nice!
16:16:32tvelocitymirc scripting!
16:16:41crwlvisual basic macros!
16:17:09Bgerjavascript! :P
16:17:15nioboscrwl: hmm... might be difficutlt to get all the security holes in it...
16:17:35tvelocitynow Bger you have to name a language that is lamer than VB, that how the game works... right?
16:18:14tvelocityok i found it
16:18:19tvelocityDOS batch files
16:18:29preglowvisual basic makes baby jesus want to vomit with rage
16:18:30nioboswell, why not Windows Scripting HOst?
16:18:31ashridahz80 assembly
16:18:34tvelocityrockbox would be like crippled without .bat file support :)
16:18:42preglowashridah: rockboy already does that :P
16:18:45Bagdernow, time to go looong weekend
16:18:49*Bger suggests something more perverse like Scheme or LISP
16:18:54Bagdersee ya monday night
16:19:01preglowlet's hope so
16:19:03ashridahpreglow: fine. arm.
16:19:05preglowhave a nice weekend
16:19:06HCl :3
16:19:17Bgernice weekend, Bagder
16:19:18crwli said brainfuck already ;P
16:19:24niobosnice weekend
16:19:25HCldidn't see that
16:19:25HCl :)
16:19:32Lynx_ook then ;)
16:19:51 Join XavierGr [0] (
16:20:31XavierGrHi all!
16:20:33tvelocitywow a fellow geek, er, i meant greek
16:20:38*HCl prods Slasheri
16:21:20XavierGryou caught that from the "Gr" extension?
16:22:01 Quit Chamois (" HydraIRC -> <- The dawn of a new IRC era")
16:22:03niobosactualy he fetched your IP and remembered that it belongs to a greek ISP...
16:22:13XavierGrahh! but the same extension aplies to my nick
16:22:23tvelocityyeah obviously
16:22:31XavierGrXavier as a nick is always registered..
16:23:06XavierGrwhat about yours: T velocity or TV elocity? ;p
16:23:30tvelocityterminal velocity
16:23:38tvelocityit's old PC game
16:23:45tvelocityit's a old PC game*
16:24:09tvelocitywhen PC's still sucked and SNES rulled the world :P
16:24:21XavierGrYeah I agree
16:24:21preglowsnes STILL rules the world
16:24:31HClyes :P
16:24:34XavierGrthough I dont remember terminal velocity as a game
16:24:34HCli have snes on my xbox
16:24:53tvelocityi often play snes games on zsnes
16:25:19tvelocityi think zelda: a link to the past, is the best video game in known history :P
16:26:25preglowfinal fantasy 6
16:26:32HClspace invaders!
16:27:11tvelocityis FF6 better than FF2?
16:27:32preglowby far
16:27:47tvelocitythen I MUST check it ou :)
16:27:56preglowit's called FF3 in the us
16:29:09tvelocityin the EU too?
16:31:34preglownever was released there
16:31:57XavierGroh yes Zelda best ever!
16:31:58*ashridah wants more gb and gbc roms :(
16:32:24tvelocityspeaking of games, is anyone playing WoW?
16:32:31ashridahpity i've got to crash now
16:32:32 Quit ashridah ("sleep")
16:32:48HClflatmate of me plays it
16:33:40tvelocityit rulez, but it's a real time burner...
16:34:04HClultima online is probably worse
16:34:09tvelocityi left designing of my LUG website when i started it
16:34:12HClhaving to water the plants you own every day
16:34:49HClWoW is nice, but i'm still waiting for a game to have the flexibility and possibilities of ultima online
16:35:03HClwith better graphics
16:35:09HClcause ultima online is 7 years old by now :/
16:35:15tvelocityi don't care about graphics
16:35:21HClthen you should try uo, heh.
16:35:30HClit gives loads more freedom than wow
16:35:35HCland its free-ish
16:35:38tvelocityreally?i thought it doesdnt exist a
16:35:46HClit does.
16:35:57HCla new update got out last november
16:36:09tvelocitybut i'm allready hooked on wow
16:36:12thegeekGuild Wars
16:36:23thegeekit's very nice
16:36:29thegeekbest game I've ever played so far
16:36:37thegeekand the only one to keep my attention for more than 2 weeks
16:36:38tvelocityfrom what i've heard, GW is not really a RPG
16:36:57thegeekwhy do you think that?
16:37:17thegeekthere are so many people leaving wow and joining gw these days
16:37:25thegeekit really is a bit better
16:37:34HClwow graphics are great though
16:37:37tvelocitywell i havent seen it, but i've heard that combat is more action-like
16:37:40HClwish uo had that
16:37:48thegeektvelocity : that is true
16:37:56thegeekbut it is still very much a rpg
16:38:07thegeekis that "an rpg"
16:38:10thegeekor "a rpg"
16:38:14thegeekI figured "a"
16:38:17XavierGris WoW an rpg?
16:38:20thegeeksince rpg = role playing game
16:38:20tvelocitywell, i dont give the title RPG easily to a game
16:38:27thegeekrole playing game
16:38:41thegeekI'd say that covers both wow and gw
16:38:44tvelocityi think WoW BARELY is allowed to be called a RPG
16:39:01tvelocitywell, it's played on a computer
16:39:08tvelocitythat's enough
16:39:14thegeekthat's lame
16:39:28XavierGrI had a fried that said to me that Diablo 2 was concidered an rpg I laughed at him
16:39:29tvelocitycomputer RPGs are not really RPGs
16:39:38tvelocitydiablo an RPG? LOL!
16:39:38thegeekthe medium in which the game is presented should not matter imho
16:39:41XavierGrI dont think that Diablo2 can be called an rpg
16:39:48thegeekI agree that d2 is perhaps on the border of an rpg
16:39:52tvelocitythegeek, ir's not a matter of medium
16:39:56thegeekbut gw is much more advanced than d2
16:40:03thegeekI see your point
16:40:08tvelocityif diablo is anRPG
16:40:15tvelocitythen pokemon is an RPG too...
16:40:20thegeekI agree that it barely is
16:40:27tvelocityactually pokemon is a lot more RPG like than diablo
16:40:29thegeekand I do understand what your are saying
16:40:39tvelocityit has a lot more RP elements than diablo
16:40:42XavierGrI havent seen WoW so I cant really say if it is or not
16:40:54thegeekdiable is hack'n'slash
16:41:04thegeekwith a few rpg parts thrown in
16:41:07XavierGrI agree
16:41:22tvelocitythat's what video gamers say
16:41:29tvelocitybut RP-gamers dissagry
16:41:31thegeekold school rpg-ing is ofcourse not compareable to anything played on a computer
16:41:44XavierGrBut it is the same friend that now plays WoW and he is whining that WoW is so rpg.
16:42:00HClfrom what i saw
16:42:01thegeekimho it is;)
16:42:04HClWoW is mostly fighting
16:42:04tvelocitywell WoW is just RPG like...
16:42:13thegeekbut it's not the same as an real life rpg
16:42:19tvelocitya computer RPG really can't be a rpg
16:42:30thegeekat this point
16:42:40tvelocitya computer can;t generate a realistic role playing expierience
16:42:42thegeekwhat we disagree on is just what "rpg" should apply to
16:42:49thegeekno point in discussing that any further imho
16:42:56tvelocityyou're right
16:42:57thegeekit's just a difference of opinion
16:43:04XavierGrwell FF series I think are rpg (classic ones)
16:43:25tvelocityit's really a personal matter, since there is no exact definition of RPG
16:43:49XavierGrso what internet says? Is WoW entitled RPG or not?
16:43:58XavierGrBecause Diablo was an action-rpg
16:44:35tvelocitythe good thing about WoW is that is has 2 RP servers
16:44:53tvelocitywhere people actually are roleplaying to each other
16:45:11tvelocitythey can't role play with NPC;s, but at least they role play with other people
16:45:32tvelocityso wow at least has *some* role playing involved
16:45:54tvelocitypropably other mmo's have that too
16:47:51XavierGrHooray I managed to compile rockbox alone!!!
16:48:15tvelocitycongrats, u just earned 1600 XP
16:53:06XavierGrHey this Rockbox Development Kit is unbelievable! The cygwin was way too complicated and big to install.
16:53:33tvelocitywhy don't you just get linux? :P
16:53:49XavierGrThe DevKit has it all for Rockbox development and it is very small. Someone must update the wiki for the starters as me.
16:54:09XavierGroh and thanks those that pointed it out!
16:54:31XavierGrwell I have knoppix installed on my iHP-140! ;P
16:54:54XavierGrcrashed Winodoze PC's be aware...
17:00:46 Part martijn
17:14:08HClare there any objections against adding the new songdb (in java) to cvs?
17:18:29***Saving seen data "./dancer.seen"
17:22:41HClits silent here o.o.
17:22:44*HCl prods Slasheri
17:23:05niobosi can make some noise... but I doubt if it'll be constructive...
17:24:22HClwhere'd linus go anyways
17:24:26 Join theebag [0] (
17:24:46 Quit XavierGr ()
17:27:52 Part theebag
17:37:20 Join DangerousDan [0] (
17:38:09SlasheriHCl: now doing some final testing :)
17:38:12HClk :)
17:38:23preglowvery goodie
17:38:27HCli got everything ready to implement the runtime database
17:38:35HClwell, a lot, anyways :p
17:39:07HClits almost trivial to add runtime database to the generation o.o..
17:39:18thegeekcheck if your java code compiles using gcc
17:39:24thegeekgcj or whatever it is
17:39:37thegeekI think that would help convince some..;)
17:39:46HCl ;p
17:40:00HClwell, my primary goal is to get a maintainable tool that will do the job
17:40:01preglowi don't even have gcj
17:40:21thegeekI dont mind java all that much
17:40:28thegeekbut I can see why other don't want it
17:40:38*HCl installs gcj
17:40:39thegeekI run windows on my main desktop so it's not a problem
17:42:10thegeekbut I'm not that keen on installing it on my server
17:42:16thegeekI'd do it
17:42:23niobosIf you give me a Java-source, I could port it to C++, given some time and maybe support on non-commented code
17:42:35preglowif you're going to port it to anything, please make it c
17:42:41preglowthen rockbox can use it as well
17:42:47tvelocityC pwnz :)
17:42:51thegeekthat'd be nice
17:42:56thegeekdoing it on the iriver itself
17:42:56preglowbeing able to regenerate the db from rockbox would be nice
17:42:58nioboshmm, wasn't the point of using Java to get OO-code?
17:43:12preglowthat's just a coding aid for hcl
17:43:16preglownot a point in itself
17:43:22HClyou can do OO in c
17:43:23niobosWell, I'll try C then...
17:43:24HClwith structs
17:43:25preglowwe'd rather it is c for rockbox, it being embedded and all
17:43:42niobosshouldn't be too hard, I guess...
17:43:45HCli'll just pack the .java in with the .jar
17:43:55tvelocityyou can do OO in any language, it's a matter of the language being designed for OOP actually..
17:44:07HClOO in brainfuck...
17:44:11HCli think you need at least pointers
17:44:11niobossure, but I don't like to try OO in eg BASIC...
17:44:12HCl ;p
17:44:35niobosThe only thing that might be a problem is the tag-reading lib's
17:44:44tvelocitybut there are object systems for C
17:44:50tvelocitylike Gobject
17:44:52nioboswhich might be C++
17:44:53thegeekthat'd be fun HCl
17:45:17thegeekI can't imagine the _true_ brainfuck that would be
17:45:33thegeeknot to mention it would do wonders for maintainability;)
17:45:37HCl :P
17:46:06thegeekhas anyone got any experience with air shipping from hongkong?
17:46:12thegeekmy package left yesterday
17:46:16thegeekhow long would it take?
17:46:16HCli do, actually
17:46:20HClit took 11 days
17:46:21HClfor me
17:46:22thegeekI've only used ground before
17:46:31thegeekmy new dac:)
17:46:42HCli got my new battery from hongkong :)
17:46:47HCl13 euro including shipping :P
17:46:53thegeekI bought an xbox psu once
17:47:01thegeekI blew up the original
17:47:07HCl yes :p
17:47:20tvelocityare they making any profit?:P
17:47:20niobosincluding shipping? what would that be WITHOUT shipping?
17:47:26HCl6 euro
17:47:32HCl :p
17:47:43thegeekwas it not for the long delay in shipping
17:47:52thegeekI would buy a lot more from hongkong
17:48:05HClniobos: latest download has the source in the .jar
17:48:13nioboswhat was the url again?
17:48:44HClalso. the backend is *bloated*
17:48:48HClit could use some stripping :/
17:49:18HCli haven't quite tested this version yet, but it should work great
17:49:23HCli haven't added comments though.
17:49:32preglowi don't take it, what about a tag reading library makes it a backend?
17:50:09HClsomething that provides core functionality that you depend on yet haven't written = backend o.o
17:50:29preglowmakes...... sense...
17:52:01niobosHCl: is there a plact that I can find some schematics or somthing about the file format?
17:52:15niobosor should your code be sufficient?
17:53:23*niobos should notice that I'm supposed to study for my exams... so the coding might take a while...
17:55:07niobosfor the record: for command-line options: what is your prefered format:
17:55:08HClniobos: there is
17:55:14niobos-a -b -c?
17:55:16HClniobos: on the wiki, DatabaseV2
17:55:54niobosonly use long options? or also provide -a shorthands for −−append (or whatever)?
17:56:10HClonly long
17:56:19HCljust the same like the old
17:56:29HClkeep in mind the java version isn't finished yet
17:56:37HCland it still needs stuff added
17:56:39nioboshaven't tried that one...
17:56:57niobosk, but the C version won't be finished today either...
17:56:57HCltitania:/home/hcl/rockbox/rockbox-devel/tools# ./
17:56:57HClsongdb −−path <dir> [−−dirisalbum] [−−dirisalbumname] [−−db <file>] [−−strip <path>] [−−add <path>] [−−verbose] [−−help]
17:57:11HClyea, just saying that you should keep account of it in your design
17:57:32HClpretty much the only things that are gonna get added is more taginfo stored in the song structure, and reading in an old run/tagdatabase
17:58:10HClthat reminds me.
17:58:20HCli need to get hashing to work before i'm able to read in the old runtime db
17:58:24*HCl prods rasher
17:58:30HClwhat do you got for me?
17:58:57nioboswell, I'll be checking that DBv2 page and see if I have any more questions
18:00:24HClit'll have to be renamed to database soon, since "v3" is imminent
18:01:00nioboswe're talking about the TAGdb, right? not about the runtimedb?
18:01:41 Join DomZ [0] (
18:01:57HClwell, pretty much. they're connected at least somewhat.
18:02:13HClthe tagdatabase storing offsets of the entries in the runtimedb
18:02:30DomZi have tried the SongDB.jar and i have the following error
18:02:35DomZC:\Documents and Settings\Administrateur\Bureau>java -cp SongDB.jar SongDB
18:02:35DomZException in thread "main" java.lang.UnsupportedClassVersionError: SongDB (Unsup
18:02:35DomZported major.minor version 49.0)
18:02:35DBUGEnqueued KICK DomZ
18:02:35DomZ at java.lang.ClassLoader.defineClass0(Native Method)
18:02:35DomZ at java.lang.ClassLoader.defineClass(Unknown Source)
18:02:35***Alert Mode level 1
18:02:35DomZ at Source)
18:02:37***Alert Mode level 2
18:02:37DomZ at Source)
18:02:39DomZ at$100(Unknown Source)
18:02:41DomZ at$ Source)
18:02:43DomZ at Method)
18:02:45DomZ at Source)
18:02:47DomZ at java.lang.ClassLoader.loadClass(Unknown Source)
18:02:47HClmeans you're using an old version of java.
18:02:49DomZ at sun.misc.Launcher$AppClassLoader.loadClass(Unknown Source)
18:02:51***Alert Mode level 3
18:02:51DomZ at java.lang.ClassLoader.loadClass(Unknown Source)
18:02:53DomZ at java.lang.ClassLoader.loadClassInternal(Unknown Source)
18:02:55DomZsorry ;)
18:02:59HClits not really official yet either
18:03:03HClbut its a good point
18:03:06HCli'll recompile with 1.3
18:03:10DomZhumm ok
18:03:13HClto support older versions
18:03:20DomZthanks a lot
18:03:22HClcode should be 1.3 compatible
18:03:38DomZbut i think i have at least 1.4
18:04:15DomZthis is the version i have
18:04:16HClyea, i used 1.5.0..
18:04:18 Quit spiral (Nick collision from services.)
18:04:20HCllet me get an older sdk
18:04:24DomZah ok
18:04:47niobosHCl: just to check if I get it right: the album-part contains (per album) a fixed-length array of songs?
18:05:09niobosso if I have 1 album with 100 songs and another with 1 song, It'll take up 99 0x00000000's?
18:05:33nioboshmm... seems a bit stupid to me, but it'll be my small brain
18:05:46HClits the only way we can still search quickly through the database
18:06:07nioboswell, my sister's player has another nice DB-structure...
18:06:13SlasheriHmm, about 5 bugs fixed now.. Should be stable soon :)
18:06:16nioboswell, it's not worthy the name DB:
18:06:29niobosit just generates a m3u list for eacht album
18:06:37HClthat won't do ;p
18:06:38DomZthanks a lot Slasheri !
18:06:39niobosand let's the file-system-driver do the "lookup"
18:06:47SlasheriDomZ :)
18:06:53HClhave you looked at the searchengine that works on the database?
18:06:59niobosnot yet
18:07:03HClit can do stuff like
18:07:32HCl(title starts with "Moon" or artist equals "air") and year > 1980
18:07:41HCljust simple m3u playlists won't do ;p
18:07:47DomZi think the space is not the problem for H series
18:08:02DomZthe disk space
18:08:12DomZihmo :)
18:08:14niobosno, but it IS a waste of space to have 1000 songs with album "unknown" and loose much space
18:08:25HClyea, largest database i've seen so far is 6mb
18:08:36HClwhich isn't that much compared to 20 or 40 gb
18:08:41DomZyes but if this is the faster way
18:09:03niobosk, I agree... let's waste space for speed... it's always a tradeoff
18:10:08niobos(i might check the code first, but I'm lazy): what to do with various-artist-albums?
18:10:11HCli'll just get 1.4.2
18:10:26HClniobos: they should work with current design...
18:10:34HClalbum keys are based on albumname + directory name
18:10:39HClthere's an option to make it directory name only
18:11:03niobosbut the album-entry has a pointer to an artist-entry, which, obviously isn't defined
18:11:11HClthats correct
18:11:40HCli'm planning to add an artist "<multiple artists>"
18:11:40 Join Stryke` [0] (
18:11:49 Nick Lynx_ is now known as Lynx_awy (HydraIRC@
18:11:51HCland use that for multiple artist albums
18:11:55HClbut i haven't gotten around to that yet.
18:12:05niobosk, I think I'm getting the struct...
18:12:42niobosI'll try to get some C-code together, but, as said before, I "should" be studying for my exams, so I "shouldn't" have much time to do it
18:12:48HCl :p
18:12:52***Alert Mode OFF
18:12:57CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
18:12:57*HCl starts a small todo list
18:13:27 Quit bobTHC ("Smoke Weed Every Day !")
18:25:51HClthere. multiple artists should be labeled multiple artists now.
18:26:19tvelocityisn't various artists more common?
18:26:33HClokay, consider it fixed :)
18:26:59tvelocityanyway i didn't sleep tonight... so... goodnight :P
18:29:02thegeekremind me to enver start a replaygin reading for 40gb of music _ever_ again
18:29:11thegeekI can't abort without loosing all the work done so far
18:29:35 Quit tvelocity ("Leaving")
18:30:31HClheh o.o
18:30:33HClwhats replay gain?
18:30:51Stryke`an intelligent algorithm for volume "normalization"
18:31:35HClyou might want to write a tool to enter those values into the runtime database
18:31:42HClit has support for that
18:31:42niobosHCl: are you CapitalVariableName or underscore_variable_name or nothinginbetweenname minded?
18:32:06niobostypicaly Java...
18:32:10HClunless its classes
18:32:11HCl :p
18:32:24niobosI 'm used to underscore_names, but I'll use firstNotButRestYes
18:32:32HClnah, go ahead
18:32:33HCli don't really mind
18:32:53*HCl is just staring at his various artist code to see whether he sees any bugs
18:33:39 Join MoosCamaro [0] (
18:33:41niobosthe pointers in the file, are bigEndian? or LittleEndian?
18:33:46HClbigendian file offsets
18:33:48MoosCamaroHi all
18:34:31*HCl throws his latest version online
18:34:52HClhaven't recompiled with 1.4 yet :/
18:34:57HClit refused to install it, for some reason
18:35:03HCli'll look at it again in a bit
18:35:26DomZok no pb ;)
18:35:44HClbut this version has better various artists support, and some code restructurization
18:35:51HCltagdatabase is a nice singleton now, as it should be :)
18:37:59 Join webguest34 [0] (
18:38:22niobosHCl: is there a reason to reserve 32bits to flag the runtime-db dirty?
18:38:33HClarchos needs 32bit alignment
18:38:54preglowmany things do
18:39:00niobosk, can I make it <31bit reserved><1bit synced> ?
18:39:11HClum, yea, i think so o.o
18:39:16niobosor should it remain DWORD synced?
18:39:32preglowdoes it matter?
18:39:45niobosnot realy, but it's more extensible in the bit-thing
18:40:09nioboselse you'd have to define sync=1 means this, sync=2 means that, sync=4 means that...
18:40:17 Join bobTHC [0] (
18:41:10 Join Sucka [0] (
18:43:32HClits just a boolean, saying whether the runtime db is synced or not (whether it was available during database creation)
18:50:02HClhm. odd.
18:50:11HCl1.4 fails to compile on my assertions
18:50:13HCl*gets them out*
18:54:45HClDomZ: try now?
18:54:50HClafter redownloading, ofcourse ;p
18:56:13DomZsame error
18:56:38HClusing windows?
18:56:47HClwindows has a nasty habit of caching ftp files with internet explorer
18:56:50HCleven though they have changed
18:57:23HClworks here, with 1.4.2_08
18:57:30HClC:\programming\songdbv3j>"\Program Files\Java\jdk1.4.2_08\bin\java.exe" -jar Son
18:57:36HClUsage: SongDB <directory>
18:58:42HCli'll bbiab
18:59:47 Nick Sucka is now known as Sucka`away (
19:01:14DomZi'm using a ftp client and this is the same error
19:03:30DomZmy exact version is :
19:03:31DomZC:\Documents and Settings\Administrateur\Bureau>java -version
19:03:32DomZjava version "1.4.2_04"
19:03:32DomZJava(TM) 2 Runtime Environment, Standard Edition (build 1.4.2_04-b05)
19:03:32DomZJava HotSpot(TM) Client VM (build 1.4.2_04-b05, mixed mode)
19:04:14DomZhumm your version is more recent that mine
19:05:04DomZnovermind i will updated my jdk with the latest
19:08:53preglowSlasheri: so, how's progress?
19:09:28 Quit DomZ ("CGI:IRC (EOF)")
19:10:08 Join Chamois [0] (
19:13:00Slasheripreglow: Now i will take a patch
19:13:11SlasheriIt should be stable enough for cvs now :)
19:13:21 Join Shagnar [0] (
19:13:47*preglow summons the commiting authorities
19:13:52MoosCamarocongratulations super Slasheri :)
19:14:08preglowSlasheri: i'd like to take it for a spin, so lemme know when you've got a patch up
19:14:23Slasheriit's still not done but basic wps works for mp3 playback
19:14:34Slasheripreglow: ok :)
19:18:30***Saving seen data "./dancer.seen"
19:19:17 Join DomZ [0] (
19:19:35amiconnniobos: rockbox coding style is underscore_names
19:19:40DomZHCl this is working like a charm now ;)
19:20:23HCl :)
19:20:59HClsorry, i wanted 1.4.0, but they only allowed me to download 1.4.2_08 :/
19:22:07DomZthis due to big security flaws with older versions
19:22:11DomZthis is
19:22:12Slasherinow the patch has been updated:
19:22:18Slasheri(sorry, still separate files)
19:22:25DomZyeahhhhhhhhhhhhhhh :))
19:22:32HCltoo bad linus isn't around
19:23:17DomZChamois !! :)
19:26:01preglowSlasheri: can't you get the new files into the patch?
19:26:51 Join mrlala [0] (
19:29:10crwli was about to say that i got illinstr when opening a mp3 file, but then i remembered that i didn't upgrade codec*.rock files
19:30:03HCl :p
19:30:54DomZthat could be great to have the file compiled files ;)
19:32:20niobosamiconn: k, that's my style too... can I find an overview on the coding style?
19:32:45HClyea, there is one on the wiki
19:32:55niobosk, i'll find it
19:34:50nioboshmm... no match on "code style" (without quotes)...
19:36:23IsmoSlasheri: this is great!
19:36:24HCltry looking for submitting patches or so..
19:36:30Slasheripreglow: i can but it's very difficult and will take at least about 30 minutes to do.. I don't want to do that
19:36:33SlasheriIsmo :)
19:37:22Ismobtw, why the volume must be touched once to get sound out?
19:37:26preglowwps :PP
19:37:30Ismodo you know the reason already?
19:37:35SlasheriIsmo: yep, that's still a bug
19:37:41preglowSlasheri: it only pretends to be seeking, yes?
19:38:04*HCl has never even seen the wps
19:38:06Slasheripreglow: real seeking doen't work yet :)
19:38:16IsmoI also noticed that the volume wraps around at 85 or so
19:38:22preglowwhats the thing beneath the seek bar?
19:38:23preglowa level meter?
19:38:28IsmoI noticed that by accident of course :)
19:38:34SlasheriIsmo: Hmm.. it shouldn't with the new patch
19:38:35preglowIsmo: no longer
19:38:50preglowthat's been fixed ages ago
19:38:50IsmoSlasheri: ok, I haven't tried with this patch yet, it was the previous
19:39:17Slasheriah :)
19:39:27Ismoyep, seems to be working :)
19:39:29Slasherithis patch has many improvements
19:39:43Chamoissometimes sound stop et restart
19:39:45Chamoiswith the new patch
19:39:45preglowdo i still have to reboot after a track?
19:39:54Slasheripreglow: no, you haven't :)
19:39:59preglowyes i do :/
19:40:00Ismostop and play works :)
19:40:11preglowit chokes vehemently on this file
19:40:21HClwe should probably get it in cvs
19:40:23HCland work on it there.
19:40:28preglowyes indeed
19:40:31preglowagreed completely
19:40:42DomZhumm i have to click on stop 2 times for listen a new music
19:40:49HCllinus wanted to commit it right?
19:40:55preglowDomZ: hey, that's the trick
19:41:01preglowHCl: yup
19:41:04DomZoki :p
19:41:08HClwe'll wait for him
19:41:13preglowof course
19:41:30preglowSlasheri: two stop presses does it :P
19:41:37preglowdamn, sounds like i've got a couple of bugs going here
19:41:45preglowin the mp3 decoder
19:42:14Slasheripreglow: Hmm, there might be many bugs out there :) What kind of issues?
19:42:27niobosHCl: the counters are unsigned, I asume?
19:42:27preglowi can hear slight distortion
19:42:47preglowSlasheri: nothing you need to worry about
19:42:52preglowSlasheri: that's probably my fault
19:43:02preglowsince i did the libmad optimising
19:43:19HClniobos: yes
19:43:23DomZyes i know distortion too
19:43:43DomZyes i hear distortion too
19:43:48preglowdo you? i can only hear it on one track
19:44:02preglowsome kind of sample and hold-ish sound on a bassdrum
19:44:03Slasheripreglow: try to reboot and play the same song again. Is there still the distortion?
19:44:09DomZthat depends on the tracks
19:44:26preglowi actually can't hear it any longer
19:44:31SlasheriIf there is, it's probably due to the decoder
19:44:39SlasheriHmm :/
19:44:39preglowyes i can
19:44:42preglowit's very subtle
19:44:47preglowand it's a libmad bug
19:45:02IsmoSlasheri: you've most probably noticed that when you stop a song and start another one you can hear a very short piece of the previous song
19:45:04preglowi heard distortion like this all the time while i debugged libmad
19:45:10Ismolike that the buffer is not cleared
19:45:13preglowit's not
19:45:21SlasheriIsmo: yes, i will fix that soon
19:45:31Ismook :)
19:45:40Ismothis is so nice :)
19:46:03preglowwhy is there a loud click at the end of the fade out? :P
19:46:41amiconnpreglow: Did you get to measure the dc offset (== did you find a suitable cable?)
19:46:42DomZah i haven't this one :)
19:46:52Slasheripreglow: i think that the wps integration is still not complete and that will cause it
19:46:52preglowamiconn: still haven't found the extension cable
19:46:54 Join DrRickDaglessMD [0] (
19:47:08preglowSlasheri: i rather think it's connected to a uda driver bug
19:47:33 Join TROGDORRR [0] (
19:47:55Slasheripreglow: Hmm.. Might be. Anyway, the wps (or something) fades the volume out when you press stop and after that the volume is set back to the previous level
19:48:06SlasheriI think there should be fade in too but there is not..
19:48:09preglowahh, so it's the wps doing it
19:48:54preglowthen i can probably turn it off as well :P
19:48:59webguest34wow... really like to congratulate you slash... I just patched, and it looks awesome!
19:49:02webguest34bye bye iriver firmware :-D
19:49:16amiconnSlasheri: On archos we don't need a fade in because the MAS employs an internal lowpass to avoid pops/ clicks
19:49:20preglowthe odds are even lower of me ever booting that again now, yes
19:49:20Slasheriwebguest34: nice if you liked it :D
19:49:33preglowas a matter of fact, i have sworn i wont :P
19:49:35DomZwhat is the second horizontal bar ?
19:49:41preglowlevel meter, i guess
19:49:43Slasheriamiconn: ah, ok.
19:49:55amiconnIt seems the UDA does not...
19:50:02Slasheriamiconn: it would be still nice to have such a feature
19:50:13preglowi can code up a lowpass in a jiffy
19:50:19SlasheriDomZ: no idea about that
19:50:35amiconnpreglow: I mean a lowpass filter for the volume setting, not for the actual audio
19:50:40preglowamiconn: of course
19:50:42DomZah ? but so who made it ?
19:50:43webguest34I think the second bar is actually a peak meter...
19:50:44preglowamiconn: that would be braindead
19:50:45amiconnThis is in fact trivial
19:51:16webguest34btw, is it just me, or does anyone else notice that the firmware seems to take maybe .5 seconds longer to boot? Not really a complaint, just wondering...
19:51:59DomZhumm haven't notice that
19:52:02preglowme neither
19:52:13preglowand i don't usually sit watching it booting anyway
19:52:21webguest34hmm... maybe it's my weird compile then... I'll try out the CVS edition when that's released
19:52:29preglowbut yeah, this works just fine
19:52:34preglowhasn't crashed/locked once
19:52:56Slasheriwebguest34: That might be true, because pcm driver will start dma to play a short sample of zeros (that will remove dc voltage from the headphone output)
19:53:02DomZyes as you ! :)
19:53:14webguest34ahhh.... that would explain... thanks!
19:53:36preglowso that's where the dc offset comes from
19:55:07HClwhere is linussss :p
19:55:38preglowSlasheri: damn, the boost counter is at 24, i think you've need to call clock down the cpu in places
19:56:07amiconnAnyone knowing a nice little coldfire disassembler?
19:56:07HClat some point we'll probably won't want to boost for playback but just keep it at 48 mhz or so?
19:56:14HClamiconn: ida pro
19:56:18preglowamiconn: objdump? :>
19:56:31webguest34sorry for so many questions, but is there any way to access the file browser while a song is playing?
19:56:43amiconnNot a big fat package, but something like tools/sh2d, only for coldfire...
19:56:46Slasheripreglow: eventually the boost counter will reset to zero, just watch it for a moment :)
19:56:55Slasheriit will not reset before buffering is done
19:57:28preglowSlasheri: there's still dc on the headphone jack, btw, i can hear it crackling when i connect/disconnect
19:57:38 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:58:13Slasheripreglow: Hmm, are you sure it's dc and not static voltage?
19:58:27preglowof course not, but it's never there with the iriver firmware
19:58:35preglowstatic also tends to discharge
19:58:36amiconnpreglow: objdump probably won't work on the bare binary, or do I miss something?
19:58:51SlasheriBefore the sort sample at boot, there was substantial crack when headphones was plugged in
19:58:53preglowamiconn: well, it should, somehow, i could never make it though
19:59:34preglowforget it
19:59:39preglowit's just the same with the iriver now
19:59:45preglowbut after playing a track it's worse, though
19:59:46Slasheriokay :)
19:59:56preglowbut that's just nitpicking
19:59:57preglowforget it for now
20:00:55HCldoes the database browser work?
20:01:00HClwhen trying to play songs from it?
20:01:19crwli take it using the filebrowser while playing is still not supported very well
20:01:45 Join XavierGr [0] (
20:03:07 Quit RED_M_CHIU (Read error: 110 (Connection timed out))
20:03:25Slashericrwl: i don't know yet how this should be done. Maybe some iriver related keybinding for the wps has to be added
20:03:45crwlSlasheri, to get from wps to filebrowser and back?
20:04:12crwlwell, maybe someday ;)
20:04:51XavierGrhi all
20:04:53Slasheriyeah. now it's exactly a week ago when i started working for this playback stuff :)
20:05:24HCl :p
20:05:43MoosCamaroyou are very eficiant ;)
20:05:51HClwe just need linus
20:10:05HCli need rasher too.
20:10:07*HCl pokes rasher
20:10:20thegeekwhen have started one song, then stop it, and start another song
20:10:23XavierGrwhat pokes mean?
20:10:27thegeeka short part of the first song will play
20:10:32MoosCamaroSlasheri: to this rhytme this more should not be much long
20:10:37preglowthegeek: known bug
20:10:48thegeekgreat work Slasheri:)
20:11:17thegeekI'm listening to coldplay's newest right now:)
20:11:44preglowsomeone should persuade rasher to do some battery lifetime tests with these codecs :P
20:12:32 Join webguest67 [0] (
20:12:33thegeekbut I'm running it through a cmoy, there is no way I'll risk running dc into my headphones;)
20:12:35preglowi expect vorbis gobbles cpu power right now
20:12:49HClmp3 too..
20:13:12preglowHCl: not by far as much as vorbis
20:13:27HClmore than iriver
20:13:36HCli think thats a pretty good standard to go by
20:13:36preglowhow'd you know that?
20:13:43HCli don't, i assume :p
20:13:51HClit still boosts for mp3, right?
20:14:05preglowa few more small opts and it doesn't need to
20:14:08HCl :)
20:14:39 Join FingerSoup [0] (
20:16:48amiconnA quicker method than doing extensive runtime tests would be to monitor the current consumption of the box
20:17:12amiconnIs it easy to disconnect the battery and put the ampmeter in between?
20:17:28HCli did a battery replacement
20:17:30HClit just has a connector
20:17:34HClthat you can detach
20:17:38XavierGrI am using rockbox Devikit and I get this error when I try to make the simulator
20:17:41XavierGrmake[1]: *** No rule to make target `@make'. Stop
20:17:42XavierGrmake: *** [all] Error 2
20:17:53XavierGranyone knows what it is?
20:18:53FingerSoupI've got a question regarding Slasheri's recent work on playback.... Is this the "Multi-Codec Architecture" he's implementing, or is this a "Quick(TM) Hack" to enable playback while we wait?
20:18:55amiconnThe version of 'sed' included in the devkit is probably too old
20:19:11XavierGranything I can do?
20:19:59XavierGrI have downloaded cygwin too but know it cant recognise the "make" function
20:20:23HClFingerSoup: first
20:21:05FingerSoupW00T! good news! Thanks!
20:23:10HClincidentally, can we play wavs with it?
20:23:20XavierGramiconn:nevermind I replaced the sed files with those that came with cygwin it is now fixed!
20:23:34preglowbut a wav codec should be quick to whip up
20:23:35HClshould be easy to add, no?
20:23:46HClthats something to do when its in cvs
20:23:58HClflac was realtime too, right?
20:24:08preglowwavpack too
20:24:09HClnice, nice.
20:24:16HClwe really need linus :/
20:24:33preglowhe'll come on some time tonight with a bit of luck
20:24:38HCl :)
20:27:33 Quit zeeeeeeee (Remote closed the connection)
20:27:42XavierGrwhat a nick
20:28:19preglowwhy does sound.c have cr lf?
20:28:21HCli've seen worse
20:29:43FingerSoupAs have I - Nicks that even mentioning will get you /kicked or /banned
20:31:38FingerSoupANyways, thanks for the good news. Just updated the thread on the Mistic River forums with that tidbit of info. Later all!
20:31:49 Part FingerSoup
20:33:11webguest67argh.. can't understand how to apply Slasheri's "snapshot.patch"
20:33:22XavierGrme too :(
20:33:31webguest67cant find file bla bla
20:33:32HClpatch -p1 < snapshot.patch?
20:33:35HClhow hard can it be?
20:33:52XavierGrah I typed patch < snapshot.diff
20:34:00*HCl gets the patch
20:34:07XavierGrurl again?
20:34:26webguest67user@debian-unstable:~/src/rockbox$ patch -p1 < snapshot.diff
20:34:31webguest67can't find file to patch at input line 21
20:34:46 Join Musicmad [0] (
20:34:50webguest67|−−- apps/SOURCES 30 May 2005 13:43:14 -0000 1.9
20:35:34webguest67oki.. thanks
20:36:47amiconnmrf :(
20:36:58HClsup amiconn ?
20:37:12amiconnCan you spell 'sillymas'?
20:37:20amiconnStill those recording issues...
20:38:14*HCl is gonna try Slasheri's patch
20:38:39amiconnWhatever optimisation I try for the transfer loop, it ends in corrupted frames sooner or later
20:39:14 Join t0mas [0] (
20:39:15 Quit t0mas (Read error: 54 (Connection reset by peer))
20:39:18 Join t0mas [0] (
20:39:21HClhi tomas
20:40:11t0masnew screen :D
20:40:17HClwhat did you get? :)
20:40:19t0masIiyama E435S-B
20:40:45HClmy next monitor will most likely be an lcd tv
20:41:19HClhey tomas, do you have any oggs?
20:41:29t0masyes I have
20:41:32t0mastested them with iriver
20:41:34t0masit works :)
20:41:45HClcare to give the java version of songdb a run?
20:41:52t0masyes I do
20:42:00t0masbut I have to reboot first
20:42:45 Quit t0mas (Client Quit)
20:44:54 Join t0mas [0] (
20:45:08t0masback :)
20:45:15HCl :)
20:45:19HCli need more bugtesters
20:45:31HClit feels like it went too easy
20:48:56XavierGrwhat is the url of Slasheri's latest patch? And do you remember where to put each file?
20:49:26HClif you look at the top of the patch
20:49:29HClit says where to put the files
20:49:55XavierGryes but is quite brief to understand
20:50:08XavierGrit says codec* libs
20:50:21XavierGrdoes it mean allcodec files to libs
20:50:28HClthe top of the patch.
20:50:35XavierGrahhh okay thanks
20:50:38HCllook at snapshot.patch with a text editor
20:51:07XavierGrnow that I have set Rockbox Devkit I will patch it in no time!
20:51:45XavierGreh...actually it's snapshot.diff am I right?
20:53:22 Quit Bger (Read error: 104 (Connection reset by peer))
20:53:37 Join Bger [0] (~Bager@
20:55:36 Nick Sucka`away is now known as Sucka (
20:57:35XavierGryeah but codec.c isnt in the ftp so does codecflac !?
20:57:59HClyou don't need those
20:58:59XavierGrand where do i put the diff file?
20:59:05XavierGrand run it
20:59:06HClyou don't. you use it to patch it.
20:59:22XavierGranywhere I like? ok
20:59:26XavierGrlets see
20:59:53webguest67hm.. and how to you play songs? i've applied slaheris patch, but when i press play on a song it jumps to the WPS screen and then 0.5 seconds later back to the file browser?
21:00:30HClyou probably didn't install codecs?
21:00:45webguest67aha.. .rockbox/codecs ?
21:01:25HCli think so
21:02:19XavierGri got the same error. couldnt find file. then it asks: Which file to patch
21:03:21XavierGrthat did it thanks!
21:04:26amiconnHmm :/
21:04:37*amiconn scratches head
21:04:57XavierGrwhat is bothering you amiconn?
21:05:50 Quit DrRickDaglessMD ()
21:05:50amiconnRecording issues on archos recorder - it seems I found the 'cpu power eater'
21:06:35XavierGrarchos?after all this time and still you are developing for it eh?
21:06:59amiconnWhen recording, there is a part of code (turned out it's the peakmeter) which eats almost all cpu power... some operations (like the scan for a free cluster when there is no hint in fsinfo) take much longer than they should
21:07:38webguest67hcl: so now i've got the codec rocks in .rockbox/rocks, but it still doesn't work
21:07:57HClno. you need them in /codecs
21:08:02*XavierGr rushes to listen to Rockbox Playing!!!!
21:08:18webguest67you mean /.rockbox/codecs ?
21:08:26amiconnXavierGr: I have more than one rockbox'able device
21:09:27XavierGrwebguest67: we are twins I got the same wps and back
21:10:19*HCl uploads an updated grayrockbox
21:10:48preglownew grayscale patch????
21:11:03HClbut i keep a build on my ftp around
21:11:07webguest67yea.. mp3 playing
21:11:58XavierGrdo you have one with Slasheri's Patch?
21:12:38webguest67XavierGr: copy build/plugins/codec*.rock to to .rockbox/codecs/
21:15:16XavierGrmpa is mp3?
21:15:31webguest67xavier: yes
21:16:01XavierGrso I will create a folder in the .rockbox directory codecs
21:16:59XavierGryeah men thats it!! Woohoo.
21:17:06XavierGrThanks all :)
21:17:24 Join Tangleding [0] (
21:17:44Tangledingcongratulation for last patch update
21:17:48Tangledingit's way cool
21:18:11Tangleding(incidentaly, i've checked about my paypal donation seems it worked)
21:18:12MoosCamaroHi crazy Tang
21:18:33***Saving seen data "./dancer.seen"
21:18:35TangledingHi Momoos :)
21:18:52TangledingHi XavierGr too
21:19:10TangledingMm i'm hareing a live in mp3 lame
21:19:14Tangledingseems gapless
21:19:19XavierGrhi I have to go now see ya all later!
21:19:25XavierGrits perfect!!!!
21:19:26Tangledingbye :)
21:19:46 Quit XavierGr ()
21:22:33Tangledingmm gonna update the fornoobs section if no contrincation
21:22:52preglowcontradictions? :)
21:23:02Tangledingsorry i mispelled
21:23:07preglowi'll survive
21:23:11Tangledingmaybe the word isn't correct
21:23:19preglowi get what you mean
21:23:21preglowupdate away
21:23:50amiconnImho Tangleding meant contraindications
21:23:56Tangledingokay i male it
21:24:07Tangledingthanks amiconn
21:24:56Tangledinghum maybe i should wait that the patch is included in CVS?
21:26:36amiconnNow I am getting somewhere... I think I've nailed the high cpu consumption of the peakmeter
21:27:26amiconnExecution time of the first write() (causing the fsinfo scan) down to ~10 sec (from ~30 sec, and with ~5 sec when no other activity takes place)
22:00:07 Quit DomZ ("CGI:IRC (EOF)")
22:00:45Tangledingarf i didn't noticed him before
22:00:52HCli'm bored...
22:01:10TangledingSome problems HCl?
22:01:14 Join DomZ [0] (
22:01:16amiconnHCl: YOu could work on dynarec? ;)
22:01:19HClwaiting for linus
22:01:34Tangledingah for CVS?
22:01:35HClamiconn: i don't think dynarec would be much faster. there are a few routines that could use asm versions
22:01:37Musicmadany idea when he'll be back? :)
22:01:53HCland rockboy is no priority at the moment
22:01:57 Quit Stryke` (Read error: 110 (Connection timed out))
22:02:20Tangledinglol what's the priority Database?
22:02:30Tangledingor sthing else?
22:05:31Tangledingnot a bad priority
22:06:14TangledingLinus is waited for comit the patch to CVS or another reason?
22:07:07TangledingI see you need the build in CVS to work from it that's it?
22:07:39TangledingNot sure i understand well in fact
22:08:48 Quit mrlala (Read error: 110 (Connection timed out))
22:08:56HClwell, i just want stuff in cvs
22:08:58HClbefore i work on it
22:14:08Tangledingok finalyy i understood well
22:15:33Tangledingguys i've to leave
22:15:50Tangledingi'will make the wiki 4noobs update tomorrow
22:16:01Tangleding or sunday at last
22:17:41 Quit webguest67 ("CGI:IRC")
22:18:08 Quit Tangleding ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
22:21:31 Join webguest93 [0] (
22:21:58 Join web [0] (
22:21:58 Quit webguest93 (Client Quit)
22:33:27 Join einhirn [0] (
22:35:17 Join bg_ [0] (
22:35:47bg_so do i have to build rockbox myself to put it on my player are are binaries posted somewhere?
22:35:59bg_for iriver that is
22:36:16bg_thank you sir
22:36:32Bagderand bootloader info here =>
22:37:20bg_should i just download the latest binary then or is there a stable version out that i should get?
22:37:36Bagderthere's no stable release for iriver
22:37:43Bagderit is in heavy development
22:38:41bg_right, i guess what i meant was stable to the point where it wont just crash upon loading, if its a fresh cvs build which is always a possibility
22:39:24Bagderwe only have releases and daily builds, no in-betweens
22:39:39Bagderand we have no release for iriver
22:39:42HClhowever, most daily builds won't crash upon loading
22:40:17bg_gotta upgrade my firmware first i guess, no bootloader for 1.60u
22:40:30Bagdereh, no
22:40:39Bagderyou'll get it done at the same time
22:40:58Bagderinstalling the rockbox bootloader implies installing a iriver firmware too
22:41:22preglowyou need to patch an existing firmware file, might as well get the newest you can find
22:41:26preglowwhich happens to be 1.65
22:41:34bg_yeah, thats what im currently doing
22:41:50Bagderbut there's no need to separately upgrade to that first
22:41:54Bagderthat's what I meant
22:42:12bg_i get it
22:42:22bg_ok, so im patching the ihp_120.hex
22:42:30bg_and then installing it through the iriver firmware
22:42:34bg_got it
22:42:43Bagderthen install the daily build
22:42:51Bagderwhich is a zip to unzip on the disk
22:43:05preglowit's better to do that first
22:44:22bg_so i downloaded the bootloader, im guessing its not a patch, but an already patched firmware?
22:44:37 Join [IDC]Dragon [0] (
22:44:47Bagderthere is no bootloader to download
22:44:53preglowof course there is
22:44:59Bagderah the bin
22:45:01preglowbootloader.bin qualifies
22:45:12preglowbg_: if you use windows, i recommend fwpatcher
22:45:13amiconnhi [IDC]Dragon...
22:45:21[IDC]Dragonhit there
22:45:23amiconn[IDC]Dragon: ltnirc...
22:45:28Bagderbg_: no, that's not an already patched firmware
22:45:40Bagderbg_: we don't distribute any such things
22:45:46[IDC]DragonI have less than 0 time these days
22:45:48bg_ok, so how do i patch it then?
22:45:56preglowbg_: it's in the wiki
22:45:59bg_yeah, sorry
22:46:12Musicmadbg_ : yeah it's all there. Really easy to follow
22:46:15[IDC]Dragon"Richtfest" next week
22:46:23preglowwithout the apostrophe...
22:47:05amiconn[IDC]Dragon: Do you have at least a little bit f time now?
22:47:07[IDC]DragonI see amiconn is progressing well on all fronts
22:47:19[IDC]Dragonan hour perhaps
22:47:35amiconnI have an Ondio FM build with my new recording code, but my sis can't test atm
22:47:39[IDC]Dragonif I don't fall asleep
22:47:55amiconn(no charged batteries)
22:48:01bg_will be nice when I dont have to wait 5 minutes for my iriver to boot...
22:48:06[IDC]Dragonmine should be OK
22:48:08preglowoh yes
22:48:09preglowit's quite nice
22:48:17Bagderipodlinux seems not so picky about licenses
22:48:36Bagderseeing that real codec in there
22:49:01*[IDC]Dragon has 3 MBit/s now
22:49:09 Quit bg_ ("Chatzilla [Firefox 1.0.4/20050511]")
22:49:28amiconn[IDC]Dragon: Please try recording with that. Choose a high bitrate.
22:49:48amiconnWatch the MMC icon, i.e. how often it comes on and for how long
22:50:11amiconnFinally, check the recording (or make it available somewhere for me to check)
22:50:24amiconnI'd prefer independent frames as well
22:50:49*[IDC]Dragon needs to setup an ftp
22:50:55 Join zezayer [0] (~chatzilla@
22:51:13 Quit web ("CGI:IRC")
22:51:23amiconn[IDC]Dragon: ftp for what?
22:52:20preglowBagder: most people aren't
22:52:44preglowjust tend to ignore them
22:53:13[IDC]Dragonwill we find immediately what you're looking for?
22:53:13preglowBagder: they also claim the real codec license is gpl compatible
22:53:43BagderI didn't find any such claim, did you?
22:53:48amiconn[IDC]Dragon: I'd just like to know whether it performs as it should
22:53:56preglowBagder: no, but some people claim the claim it :P
22:54:07[IDC]Dragonhow long should I record?
22:54:11Bagderpreglow: yeah, I noticed that and tried to find one ;-)
22:54:40amiconnThere might still be that disk-full problem, but since it also happens with current cvs, I'll ignore that for now and look at it later
22:54:51amiconnIt may also be gone with my changes
22:55:04amiconnYou should record for several minutes
22:55:05[IDC]Dragondisk-full problem?
22:55:39preglowBagder: i still think faad is a better codec
22:55:45preglowBagder: though my research is pretty limited
22:55:54amiconnYes, when you are recording and the disk gets full, current cvs _sometimes_ overwrites the beginning of the file with the last minute recorded or so
22:56:26amiconnObviously it is very unlikely to observe this on the jbr; on the Ondio this might happen much more often
22:56:28Bagderpreglow: I wouldn't know, I've just read the license and discussed that with rasher, and we're pretty convinced it is not gpl compatible
22:56:32[IDC]Dragonyour build has the backlight on all the time
22:56:46amiconnMy build doesn't have backlight control
22:57:10[IDC]Dragonyes, but somehow permanently enabled (port setup, I guess)
22:57:21amiconnShould I make one including it?
22:57:28amiconnI didn't change port setups
22:57:39[IDC]Dragonthat would be nice
22:57:47preglowBagder: then there's even greater reason to just go with faad
22:57:50[IDC]Dragonhmm, it should be off by default
22:58:05Bagderpreglow: indeed
22:58:06[IDC]Dragonit was for the other dailies
22:58:08preglowBagder: besides, i've still not forgiven real for realplayer
22:58:20[IDC]Dragonbut I haven't tried since a while
22:58:27amiconn[IDC]Dragon: Did you rolo from your build which has backlight control?
22:58:44[IDC]Dragonrolo and back-boot, both
22:59:05preglowthat program surely convinved me real hater mankind and seek our extinction
22:59:09amiconnAfter rolo, I'd expect this, but not on cold boot
22:59:38preglowmaking aac run well on rockbox will be a bit of a challenge, i believe
22:59:40HClyea, realplayer is awful
22:59:49preglowit would be _really_ cool if we could support he-aac
22:59:51HCli don't understand why people ever used it to generate realmedia files
22:59:58preglowi believe no other player on the market has enough cpu power for it
23:00:09preglowHCl: plain ol' stupidity
23:00:16 Quit zezayer (Remote closed the connection)
23:02:36 Join spiralout [0] (
23:03:01HClwhats he-aac?
23:03:16 Quit spiralout (Client Quit)
23:03:28 Join spiralout [0] (
23:03:33preglowaac for lower bitrates
23:03:40 Part Musicmad
23:03:40preglowuses parametric stereo and spectral band replication
23:03:51preglowpretty nice quality
23:04:24 Join mrlala [0] (
23:06:38amiconn+1032 bytes
23:07:25 Join Simplet_Onasme [0] (
23:08:16[IDC]Dragonamiconn: afk
23:09:37 Quit Nibbler ("blubber")
23:09:37 Quit Chamois (Read error: 104 (Connection reset by peer))
23:11:13 Join Chamois [0] (
23:13:02HClno linus :
23:13:04HCl :/
23:16:26preglowtry summoning him
23:16:59 Quit spiralout ("ChatZilla [Firefox 1.0.4/20050511]")
23:18:34***Saving seen data "./dancer.seen"
23:19:15*HCl plucks some preglow hair and burns it in a bowl, and casts an evil enchantment to summon forth linus!
23:20:13HClit didn't work
23:20:13HCl :/
23:20:48amiconn[IDC]Dragon: I'll soon update again - found the probable cause for the overwritten beginning
23:20:55 Join spiralout [0] (
23:20:57[IDC]DragonI'm back
23:21:11[IDC]Dragonrecorded ~4 min now, seems ok
23:21:24amiconnHow does the MMC icon behave?
23:21:26[IDC]Dragonthe MMC is active for 6 sec all 50 sec
23:21:36amiconnSounds really good
23:21:48[IDC]Dragonfaster than loading
23:21:55amiconn...that's like it should have been for ages, but wasn't
23:22:33[IDC]Dragonwhat did you do? / what was wrong?
23:22:52amiconnA number of things
23:23:21amiconn(1) When the buffer got full, it didn't always flush everything, depending on the ring buffer positions
23:24:18 Join bg_ [0] (
23:24:30amiconn(2) The peakmeter drew so much cpu power that the first write() on fat32 sometimes took so long that the buffer wrapped meanwhile
23:24:49amiconn(when the fsinfo hint was invalid)
23:25:28[IDC]DragonI saw tidbits of the discussion in the logs
23:25:41[IDC]Dragonwhy is peakmeter so busy?
23:25:50 Join xen` [0] (
23:26:13amiconn(3) The overwriting of the beginning with the last recorded data happened because the mpeg thread reacted to stray save events (after stop)
23:26:48amiconnPeakmeter used to poll in a tight loop, only yield() and peak_meter_peek()
23:27:10amiconnpeak_meter_peek() needs quite some time due to i2c communication
23:27:19[IDC]Dragonwe should do a v2.5 release when this is resolved
23:27:34amiconnI fixed this by only peeking once per redraw, sleeping in the meantime
23:27:48 Join ghode|afk [0] (
23:28:15amiconnIn fact, peeking once per redraw is much better than peeking all the time. We don't increase the chance to miss a peak by doing that
23:28:43amiconnThe mas saves the highest peak until it is read
23:29:04[IDC]Dragondoes it really?
23:29:20[IDC]DragonI remember Kurzhaarrocker cursing
23:30:08bg_so i installed the bootloader and iriver files, and when it starts up i just get a blank screen with a little hard drive spinup. then ill get a backlight if i press buttons and thats it. Reset it and original firmware loads
23:31:11[IDC]Dragonamiconn: should I check the recorded file somehow?
23:31:28amiconnListen to it...
23:31:33[IDC]DragonI did
23:31:49amiconn...and maybe check it with an mp3 check tool
23:31:53amiconnI use mp3utility
23:32:00Bagderbg_: completely blank?
23:32:43 Join Stryke` [0] (
23:33:05Bagderthat's not expected
23:33:29bg_didnt think so
23:33:44bg_but a reset starts the original firmware, so at least its not toast
23:34:12ghode|afkhi, are you still waiting for linisn to commit Slasheri's patch?
23:34:20ghode|afk+linusn ><
23:34:32Bagderbg_: so a blank screen, just a disk spinup and then nothing more?
23:34:43amiconn[IDC]Dragon: I remember having read somewhere that the mas quasi-peak stores the highest value until it is read, but I can't find it atm
23:34:57Bagderghode|afk: rather simplified, but yes
23:35:29bg_just held record and started it, loaded the original firmware fine, some text went up on the screen for < 1 sec
23:36:25bg_got rockbox.iriver in root dir as well as .rockbox folder...
23:37:00[IDC]Dragonamiconn: last frame is tuncated, otherwise ok
23:37:59amiconnLast frame truncated happens almost always
23:38:25amiconnRockbox doesn't search for frame start when stopping
23:38:41amiconnHowever, it *shoudl* work on split, but it doesn't
23:38:56amiconnI'll investigate further after committing my thread changes
23:40:15[IDC]Dragondo you have a new test upcoming?
23:40:31amiconnI'm just testing on my jbr
23:40:58amiconnThe new version should fix the overwritten beginning, but that doesn't happen always anyway
23:41:23amiconnObviously you didn't get that
23:41:44[IDC]Dragonnot with this try
23:42:09amiconnOtherwise there would have been a sync error, and an audible 'jump' on playback
23:42:22[IDC]Dragonbefore, I did another one, which couse a disk full
23:42:39[IDC]Dragonthen I deleted an album
23:43:03[IDC]Dragon_the_ album, internal MMC holds about one
23:43:32[IDC]Dragonthe first recording played shorter than I expected
23:43:33amiconnIf I am right, this was caused by the file save reacting on stray save events which arrived after the stop event.
23:44:05amiconnI have some albums that don't even fit completely on Ondio internal flash
23:44:18amiconnFortunately I have that 2 GB MMC :)
23:44:34*[IDC]Dragon blushes in envy
23:45:07amiconnGet a 3 GB ;)
23:45:35[IDC]Dragonis that avail?
23:45:37amiconnBtw, my sis now owns an 1 GB card (Transcend MMC plus == MMC 4.0)
23:45:56amiconn[IDC]Dragon: I already told you about that taiwan company...
23:46:15[IDC]Dragonbut not that you can actually buy it
23:46:33[IDC]Dragondoes the MMC 4.0 have new tricks?
23:46:52[IDC]Dragonlike, addressing commands > 4 GB?
23:47:17amiconnI didn't find any specs so far :(
23:47:48[IDC]DragonI asked, we don't have them in-house
23:48:01amiconnMMC4.0 has high speed transfer (4-bit and 8-bit modes in addition to the old compatible 1-bit, and higher clocks as well iirc)
23:48:12[IDC]Dragoncould order them, but don't find this ethical
23:48:34amiconnThose MMC4 cards have more contacts than MMC3.x (in a second row)
23:49:19[IDC]Dragonmaybe Rocker would like that
23:49:40[IDC]Dragonif you can get a 4.0 reader
23:50:17*[IDC]Dragon is sleepy
23:52:05amiconnThere is a datasheet
23:52:18amiconnHowever it does not describe the command set
23:52:57preglowmmc uses 1 bit transfer?
23:55:18preglowdoesn't exactly sound blazingly fast
23:55:57amiconn20 MBit/s maximum (MMC3.x standard)
23:56:08preglowwell, that isn't bad
23:56:09amiconnFaster than USB1.1 :-p
23:56:29amiconnIn rockbox, we can only read at 3 MBit/s
23:56:37preglowthanks to low clockrate?
23:56:53preglowor is there a separate mmc controller?
23:57:02[IDC]Dragonthankis to stupid USART
23:57:05amiconnThe SH1 SPI can only run at cpuclk/4
23:57:41preglowwell, which is it, spi or usart? :P
23:58:08[IDC]Dragona USART in SPI-ish mode
23:58:22preglowhmm, never seen that before
23:58:30[IDC]Dragonwith wrong bit order

Previous day | Next day