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 2005-03-09

00:00:00preglowPatr3ck: rockbox is corrupt
00:00:15Patr3ckoh my
00:00:24RickPatr3ck: well... if rockbox were in the flash, yeah, but it's not ;P
00:00:27shxi'm kinda f...d i think, to stay polite....
00:01:00preglowbattery should be fine around now
00:01:27Patr3ckshx: i had that same error booting iriver and then freeze on reading filesystem a lot
00:01:31shxi just try... nothing more....
00:02:09shxwell... no good news...
00:02:13preglowwell, yes, unless it suddenly starts working on its own, you're screwed
00:02:45Patr3ckshx: it always went away at least after hold button, reset (and power off), hold button, boot (iriver fw)
00:03:07shxi just did that twice
00:03:19shxand, well, nothing...
00:03:22preglowcomplete with pushin the reset switch, yes?
00:03:35shxall the sequence
00:03:59preglowwell, i have no further advice
00:04:22Patr3ckme too, I am scared now
00:04:22RickI have to say shx, that really blows
00:04:29*Rick would hate it if he lost his H1x0 harddrive
00:04:34shxanyway, thanks guys
00:04:50shxThanks rick for support :-)
00:05:04Patr3ckis it possible to disable write caching on linux?
00:05:13Bagderyou would think that it would be able to shortcut that "read the file system" thing
00:05:15shxwould it help to ship it to a hardware dev ?
00:05:25preglowiriver firmware is strangely programmed
00:05:33shxu're right
00:05:34Rickwhat happened if you removed the harddrive?
00:05:34preglowit scans the entire fs on boot
00:05:36Rickwould the iriver boot?
00:05:41preglowhell no
00:05:45preglowit would panic violently
00:06:08Bagderyes, but a file system screwup is possible entirely without rockbox so it must've happened to people before
00:06:09shxthink so too
00:06:20RickBagder: just more rare ;P
00:06:34preglowshouldn't be
00:06:40Bagderlike what if you reformat the disk and reboot
00:06:40preglowrockbox has nothing to do with usb
00:06:49 Join midk [0] (
00:06:51RickBagder: uh, I don't think that's a problem
00:06:56 Join LinusN [0] (
00:06:57RickBagder: the new firmware lets you format and reboot
00:07:07Bagderif you use a bad format
00:07:11Rickoh, true that
00:07:14RickI dunno
00:07:20Rickbug Linus to test it on his? :P
00:07:36shxi should do this ?
00:07:40Rickshx: do what?
00:07:51 Join Camilo [0] (
00:07:53shxsend it to Linus for him to ply with
00:08:07RickI wasn't suggesting that if you thought I meant that
00:08:08LinusNwhat's wrong with it?
00:08:18preglowdo a bit of googling first, this is bound to have happened to other people as well
00:08:19RickLinusN: filesystem is fubar'd
00:08:23shxits kinda dead....
00:08:26RickLinusN: freezes at read filesystem on boot
00:08:29shxi'm afraid
00:08:47Rickdunno, what shx said
00:09:05shxchecksum fails with current cvs firmware
00:09:24shxoriginal firware starts and feezes on FS read
00:09:26HClyou could screw it open, plug the 1.8" hdd into a normal computer with a converter..
00:09:28shxto be short
00:09:34HCltry charging a bit more
00:09:35preglowHCl: you know where to get those?
00:09:37HCli had that once
00:09:44shxconverter ?
00:09:48HClpreglow: its a compactflash interface, as far as i know
00:09:49shxteel me plase
00:09:58preglowi didn't know that
00:10:01LinusNis it fully charged?
00:10:02HCli'm not even sure if there are any on the market, but there should be.
00:10:06shxok, fairly easy to find then
00:10:07HClthere are converters for 2" disks
00:10:12HCl(3:2) at position (14,15) on the map,
00:10:12HCl (5:16) move any furre present to (14,16) if there's nobody already there.
00:10:12HCl (3:2) at position (14,16) on the map,
00:10:12DBUGEnqueued KICK HCl
00:10:12HCl (5:16) move any furre present to (16,17) if there's nobody already there.
00:10:12HCl (3:2) at position (16,17) on the map,
00:10:14HCl (5:16) move any furre present to (16,18) if there's nobody already there.
00:10:17HCl (3:2) at position (16,18) on the map,
00:10:19 Quit HCl ("leaving")
00:10:19preglowHCl: what gives
00:10:33 Join hcl [0] (
00:10:41hclignore that.
00:10:42preglowi daresay :P
00:10:43 Nick hcl is now known as HCl (
00:10:49HClwhy the *fuck* did irssi's accidental paste protection
00:10:50HClnot work
00:10:52shxi'll take a look
00:11:11RickHCl: drangspeak?
00:11:16Rickor whatever it's called
00:11:21preglowHCl: it occasionally doesn't, i should know, i'm the master of pasting compromising information to large irc channels
00:11:33LinusNshx: what happens if you start wit the usb cable inserted?
00:11:35HCl :P
00:11:45Ricknot really comprimising imho
00:11:58shxLinusN :nothing more
00:12:15 Join amiconn_ [0] (
00:12:24shxit tries to load original firware and freezes
00:12:38LinusNthat sucks
00:13:01shxyeah, you're right
00:13:06shxthat really sucks
00:13:15HClif i were you, i'd try to get my hands on a 1.8" -> ide converter
00:13:23HCland repair the filesystem that way
00:13:29LinusNyou learned a lesson though, always "safely remove"
00:13:30shxit looks like i've lost a couple of 100s euros....
00:13:48shxyes, daddy; i won't do that again ;-)
00:13:57LinusNshx: it's fixable
00:13:58shxexpensive trial and error...
00:14:20preglowperhaps i should reconsider my development habits
00:14:25shxLinusN : you mean with HDD adapter ?
00:14:38preglowi must have plugged it in out without 'safely remove' about, ohh, five hundred times
00:14:41LinusNlooks like i have to add usb support to the boot loader
00:14:42Rickpreglow: hehe
00:14:54HClthat does sound like a good idea.
00:15:02shxLinusN : good idea !
00:15:04HClthat way you can always recover.
00:15:13shxa little to late for me though....
00:15:14HCland its even faster with usb plugins too.
00:15:27HClum. plugging it into the usb port, i mean.
00:15:32HClnot the.. software module kind o.o
00:16:02HClshx: a 1.8" converter should be findable.. and they tend to be cheap because its literally just rewiring, no chips or anything
00:16:27shxHCl : an address maybe ?
00:16:29CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
00:16:29*HCl has a converter for laptop hdds -> ide..
00:16:45HClmrf, i dunno
00:16:59shxok, i'll dig a little for it
00:17:01CamiloLinusN, I've seen a USB bootloader, it works great. Beats xmodem
00:17:30preglowpfaw, zmodem, please
00:17:46HClshx: ebay has ide cases for 1.8" drives
00:17:48HClthose should work fine
00:17:53HClthey're at the price of 10$
00:18:00HClusb2 ide cases*
00:18:17*Rick considers buying one ahead of time to be safe
00:18:41HCljust buy one of those, screw your iriver's hdd in after you got it out of your iriver
00:18:41shxRick : need one too ?
00:18:47HClrepair it, and put it back in your iriver
00:18:48HClshould work
00:19:02 Quit joe1 (Remote closed the connection)
00:19:03shxthanx a lot
00:19:05HClnp :)
00:19:13preglowis the h120 driver 1.8" ?
00:19:13Rickshx: no, but it'd be nice to be equipped to handle it if I have the problem in the future
00:19:16Camiloare 1.8" drives the same as 2.5" ones then?
00:19:26shxand the hex screw thing
00:19:29HClpreglow: yes
00:19:36shxcan that be found easily ?
00:19:38LinusNshx: torx
00:19:40Ricki need one of those too
00:19:55preglowyou can find whole sets of torx screew drivers pretty easily
00:20:21shxok, kinda 'classical' thing i presume
00:20:31*LinusN has to sleep
00:20:39LinusNnite all
00:20:41shxthanks for all, guys, i wish that won't happen to you all !
00:20:46preglowsize t-6 should do
00:20:59preglownite, linus
00:22:19 Part LinusN
00:22:20preglowam i still the only one knowing anything about the emac instructions?
00:22:28preglowif not, i could use some help :>
00:22:43HCli've only glanced past them while looking for the normal ones :P
00:22:47HClwould that do? :P
00:22:53shx<preglow> wanna know what ?
00:22:55HCldarn :P
00:23:18preglowi need to bloody negate an accumulator register, but i just can't do it
00:23:22Camilois it a binutils problem preglow ?
00:23:26preglowCamilo: no
00:23:37preglowbinutils does its stuff nicely
00:24:13HClpreglow: times -1 ?
00:24:19HClor am i thinking too easy?
00:24:25preglowyes, taht's it
00:24:36HClit is? :3
00:24:37preglowbut it's harder than it sounds like
00:24:49preglowi can't multiply the accumulator registers by anything
00:24:51preglowjust add and sub
00:25:13preglowand it has more precision than the ordinary registers
00:25:19preglowso i can't just sub itself with itself two times
00:25:41HClset a reg to 0, then sub it with the other one?
00:25:53HClor can't you do that either?
00:25:53preglowcan't add or sub accumulators with eachother
00:25:59HClthey sound annoying.
00:26:01preglowthey are
00:26:42preglowi thought of a way of doing it, it required fifteen asm instructions
00:26:45preglowbut it didn't work
00:27:15HCl :X
00:28:18shxpreglow : isn't ther a NEGATE instruction in Moto 56K assembly that does accumulator negation ?
00:28:35preglowshx: no
00:28:43shxNEG i meant
00:28:50preglowshx: there's neg for ordinary registers
00:28:56preglowshx: but not for accumulator registers
00:29:12 Quit amiconn (Read error: 110 (Connection timed out))
00:29:12 Nick amiconn_ is now known as amiconn (
00:29:20preglowthe only thing you can do with accumulators is add a product of two ordinary registers to them, sub a product of two ordinary registers from them, or move them
00:29:48preglowthat's 56k
00:29:52preglowwe're on coldfire
00:30:02preglowhad this been a 56k, all would have been peachy :>
00:30:21Camilopreglow, could you move the negate to where you are unloading the acc anyway?
00:30:57preglowCamilo: what i tried was unloading but parts of the accumulator, doing a standard two part 0 - accumlator, and then storing the result back
00:31:00preglowbut it doesn't work
00:31:03preglowbut = both
00:31:23preglowthe accumulator is wider than a 32 bit register, so you have to use sub, not neg
00:31:45 Quit matsl ("Leaving")
00:32:29preglowto top it off, i also have to switch accumulator modes
00:32:40preglowi'll have to find a way of avoiding it, plain and simple
00:32:47shxprglow :
00:33:08shxpreglow :
00:33:20preglowshx: shx again, again, that's 56k asm, that won't help me on a coldfire
00:33:58 Quit shx ("CGI:IRC")
00:33:59Camiloyou have got the programmers ref manual?
00:34:03preglowCamilo: yes
00:34:06 Join shx [0] (
00:34:06HClits on the wiki
00:36:50Camilopreglow, what's the overall maths that is going on - there might be a way to move the negate
00:37:57preglowCamilo: yes, that's what i'll end up doing
00:38:11preglowCamilo: i'm quite simply just translating imdct_l_asm.S in libmad to coldfire asm
00:38:41*Camilo reaches for apps/codecs
00:38:46***Saving seen data "./dancer.seen"
00:39:24Bagdera greener cvs build table
00:39:42preglowline 241 to see how simple it is in arm asm
00:39:45HCla greener world = good
00:39:45HCl :)
00:40:22*preglow looks towards greener arm pastures
00:41:00preglowseems like such a nice cpu
00:41:13Camiloarm is so different... didn't I see a coldfire mac dct white paper on the motorola site?
00:41:32Bagdergee, the FLAC code is *wide*
00:41:32preglowthere is a paper on the mac unit on the site, yes
00:42:05preglowthe flac guy loves his variable names
00:42:35Bagderand he must have his editor fullscreen
00:42:50Bagdersome lines are almost 200 cols
00:43:02preglowprobably just doesn't worry about wrap
00:43:14CamiloASCII TEXT FOR FFT, DCT Source Codes
00:43:17preglowCamilo: dead link
00:45:45 Quit shx ("CGI:IRC")
00:45:49 Join shx [0] (
00:49:50preglowdoesn't help me much anyway, they've only got ordinary mac unit code, and that's very much simpler to manage
00:49:54preglowsince it's got no extension bytes
00:51:09Camiloshame. The freescale site sucks. I tried to DL something big from there today @ work, and failed
00:51:22preglowworks fineish from here
00:51:27preglowi've got everything i need downloaded anyway
00:53:43 Join Naked [0] (
00:53:54 Nick Naked is now known as Hadaka (
00:58:49 Quit shx ("CGI:IRC (EOF)")
00:59:58preglowsounds painful
01:00:09preglowin the morning, that is
01:00:10Bagderdaughter with early habits
01:00:25preglowgood night
01:06:24HClsounds like a good idea.
01:06:42preglowyes, i actually have to get up semi early tomorrow :/
01:06:55HClme too.
01:07:00HCl :/
01:07:02preglow10:00 :///
01:07:11HClwell, i usually get up at 10
01:07:20HClcause i get late if i get up at 10:30
01:07:39preglownah, i'll probably have to go for 09:00 tomorrow
01:09:24 Quit lolo-laptop ("Client exiting")
01:09:27 Quit Patr3ck ()
01:10:20 Quit jyp ("poof!")
01:14:00 Quit Digital007 ("CGI:IRC (Ping timeout)")
01:17:13 Quit Camilo ("Chatzilla 0.9.67 [Mozilla rv:1.8a6/20050111]")
01:21:17 Quit skav (Read error: 104 (Connection reset by peer))
01:23:55 Quit Sucka ("a bird in the bush is worth two in your house")
01:39:21 Quit SoulEata (Read error: 110 (Connection timed out))
01:45:53HCl"Come to the darkside - We have cookies" *nods at that*
01:53:47 Quit preglow ("quiet")
02:02:58 Quit sofaSpud ("There are lies, damn lies, and MS facts")
02:15:33 Quit Aison ("( :: NoNameScript 3.72 :: )")
02:16:15 Join StrathAFK [0] (
02:21:59 Join Soul_Eater [0] (
02:29:03 Join YouCeyE [0] (foobar@youceye.user)
02:34:10 Quit Strath (Read error: 110 (Connection timed out))
02:35:27 Quit mecraw ("Trillian (")
02:38:47***Saving seen data "./dancer.seen"
02:44:27 Quit cYmen ("zZz")
02:48:44 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
02:51:53 Join Mong0 [0] (
02:52:38 Join Bonkers [0] (
02:53:03Mong0should rockbox.iriver be in the root or in a rockbox folder?
02:53:23geoff_oAre you talking about the wiki?
02:53:52Mong0nope the actual compiled code?
02:56:48Mong0anyone buller buller
03:31:07 Join DMJC [0] (
03:37:42 Quit DMJC ("Leaving")
03:56:31Mong0you awake Badger?
04:05:39 Join QT_ [0] (as@area51.users.madwifi)
04:08:56Mong0anyone know the command to make all to include the plugins?
04:09:35 Quit QT (Read error: 60 (Operation timed out))
04:10:40 Quit XShocK (" HydraIRC -> <- 100,000+ downloads can't be wrong")
04:13:51 Quit Mong0 ("Chatzilla 0.9.67 [Firefox 1.0.1/20050225]")
04:22:35 Join ze__ [0] (
04:26:42 Join Mong0 [0] (
04:28:21 Nick StrathAFK is now known as Strath (
04:34:31 Quit ze (Read error: 110 (Connection timed out))
04:34:31 Nick ze__ is now known as ze (
04:38:50***Saving seen data "./dancer.seen"
04:57:44 Quit Mong0 (" HydraIRC -> <- Go on, try it!")
05:01:44 Quit amiconn (
05:01:44 Quit [Zmaj] (
05:01:44 Quit MrMuffin (
05:01:57NJoinamiconn [0] (
05:01:57NJoin[Zmaj] [0] (
05:01:57NJoinMrMuffin [0] (
05:04:21 Part geoff_o ("Kopete 0.9.1 :")
05:52:56 Join DMJC [0] (
06:15:55 Join lostlogic [0] (
06:38:52***Saving seen data "./dancer.seen"
07:15:28*Rick registered out of sheer boredom
07:25:40 Join stevenm [0] (
07:27:39 Quit stevenm (Client Quit)
07:30:19 Join ashridah [0] (
07:30:20 Join LinusN [0] (
07:31:50 Quit Hadaka (Read error: 54 (Connection reset by peer))
07:31:51 Join Naked [0] (
07:32:07 Nick Naked is now known as Hadaka (
07:45:02 Join Terminal-Velocit [0] (
07:59:03 Quit einhirn (Read error: 104 (Connection reset by peer))
08:12:31 Join hile__ [0] (
08:14:51 Quit Ka (
08:14:51 Quit Tipi^ (
08:14:51 Quit nozomiyume (
08:14:51 Quit hile (
08:14:51 Quit rob- (
08:16:17 Nick QT_ is now known as QT (as@area51.users.madwifi)
08:16:56NJoinTipi^ [0] (
08:23:40Bagdergooood morning
08:28:07BagderI greenified more of the cvs table yday
08:28:16LinusNyou are a hero :-)
08:28:43NJoinKa [0] (
08:33:54Bagderthe FLAC code is... special
08:38:53***Saving seen data "./dancer.seen"
08:41:15DMJChow's tremor going?
08:42:25LinusNdon't know
08:42:47DMJCI'm tempted to boot back into linux to try it out
08:44:56amiconnBagder: I thought you'd want to fix the win32 warnings by building a less oldish cross compiler...
08:45:06Bagderyes, one day
08:48:19BagderI still haven't seen any docs on how to do it for a modern compiler version
08:48:34Bagderwhich indicates it is not well supported
08:50:09amiconnI just checked on MSDN... it's strange that _chsize() is declared, but _commit() isn't. They're part of the same header...
08:50:31Bagderthe same MS header, mingw does not use them
08:52:02 Join einhirn [0] (
08:53:18amiconnBagder: From searching my cygwin installation, it's the same header in mingw (io.h)
08:53:38Bagderwhat mean is that the ming guys have rewritten the headers
08:53:42Bagderthey're not copied from MS
08:54:46amiconnYes... so simply there's a line missing in the old mingw io.h
08:59:12BonkersMSDN doesn't necessarily complain to ANSI standards
08:59:57Bagderfor this case that doesn't matter
09:00:08Bonkersoh, you're just warning hunting
09:03:28amiconnAnother thing... I wonder why filesize() is implemented in such a strange way, with all that seeking...
09:05:13Bonkersfor FAT? is there any way to see hwo big a file is without seeking to the end?
09:05:34amiconnNo, I mean in the simulators
09:06:02RickBonkers: ftell() doesn't exist?
09:06:02Bagdershould use stat() instead
09:06:06Rickor tell()
09:06:21amiconnOn unix one could use fstat(), and there is _filelength() on windows
09:06:28Bagderor fstat() rather, tes
09:06:51Bagderbut it doesn't matter much
09:06:52Rickoh wait, ftell() tells you your current position in the file
09:07:09*Bagder rejects a few random feature-requests
09:07:30RickI saw a request to put the rockbox on the h120
09:07:34Rickhehe... from last year
09:08:00amiconnBagder: stat() takes a path, while fstat() takes a file handle, right?
09:08:34Rickthat reminds me, why arn't the file apis prefixed with f?
09:08:41amiconnSo fstat() would be the right thing to use, since filesize() in rockbox also uses the handle
09:09:28BagderRick: because we don't support those functions, only the non-f ones
09:09:31amiconnHow handy, _filelength() on windows also uses the handle... I think I should do that then...
09:09:51RickBagder: I was under the assumption that they were essentially the same?
09:09:59Rick(other than the fact that it doesn't use FILE for handles)
09:10:18Bagdermore or less, yes
09:10:27amiconnAnother thing, regarding an improved button handling on X11: Could I take the x11 multithreading extension for granted? It's an X11R6 addition...
09:10:52Bagderamiconn: I think you can, we all do this on fairly modern X versions
09:11:27amiconnIf not, the button driver will look very ugly...
09:13:29 Quit DMJC ("Leaving")
09:14:47 Quit Seed (Nick collision from services.)
09:14:54 Join Seed [0] (
09:23:26 Join preglow [0] (
09:29:52 Join kurzhaarrocker [0] (
09:30:36kurzhaarrocker:( Bagder closed my feature request for a feature implementor
09:30:48*Bagder cleans mercilessly
09:34:28 Quit kurzhaarrocker (Remote closed the connection)
09:41:40 Join Sando [0] (
09:44:28*preglow yawns furilously
09:46:16BagderI suggest we close all fixed and rejected feature requests and bugs
09:53:36 Join Schnueff [0] (~mah@
09:55:03LinusNany ARM ninjas around?
09:55:29dwihnoI'm a one-armed ninja, is that good enough? ;)
09:55:43Bagdera dwihno!
09:56:07kergothheh, i write linux kernel code for ARM all day every day, but i doubt i'm ninja caliber
09:56:09dwihnoWhere? Who!?
09:57:14 Nick hile__ is now known as hile (
10:01:33 Join rob- [0] (
10:07:00LinusNi'd like to know how the exception vector table looks like
10:08:09LinusNfrom what i have gathered, the vectors are in fact jump instructions, right?
10:08:54preglowhow risc of them
10:09:07LinusN(trying to get a nice crib for the iFP cryptanalysis)
10:09:58preglowsomeone should scramble stripwax, heh
10:10:01dwihnoLinusN: you're considering ifp a new target?
10:13:51 Join DMJC [0] (
10:14:14dwihnoah, that would be cool!
10:14:24dwihnoConsidering the fact I'm a owner of such a thingie ;)
10:15:48LinusNthe sad part is that i've seen source code for ifp decryption on the net, but i can't find it again...
10:17:53Rickifp... the flash irivers?
10:18:39RickLinusN: Couldn't you check on that one project? I remember there already being a new firmware project for ifp, or somesuch
10:18:47*Rick searches sourceforge
10:19:06Bagderok, feature-request cleaning done for today
10:19:19LinusNthere is a project for a linux device driver for the ifp
10:19:49RickOh well
10:20:10LinusNhm, maybe we don't need cryptanalysis after all
10:21:24LinusNthe SAA7750 has an ARM7 core, and it has a jtag debug port
10:21:34LinusNand i have a wiggler
10:22:13LinusNand plenty of spare time for reverse engineering...not!
10:22:31Rickhow much memory does the ifp have?
10:22:50LinusNthere are so many different models
10:23:07LinusNup to 512mbytes i think
10:23:49LinusNphilips has sent us plenty of documentation on the SAA7750
10:23:58LinusNbut not on the DSP
10:24:30BagderSAA7750 seems to be commonly used
10:24:34LinusNand we don't know if we are allowed to distribute the docs
10:24:46Rickask? ;P
10:24:53LinusNi have asked them for the dsp docs as well
10:25:13LinusNi won't ask them about redistribution until they have sent me the docs :-)
10:25:46*Bagder detects experience talking ;-)
10:26:42LinusNtexas instruments were very helpful with identifying the "mysterious" AZU chip on the H1x0
10:27:12LinusNit's a li-ion charger
10:27:20 Join Zagor [0] (
10:27:28Bagderdr Zivago!
10:27:36Bagderah, no it was just Zagor ;-)
10:27:41LinusNwhich (and this makes me angry) is made for USB charging
10:28:03LinusNbut iriver doesn't use the usb charging feature!
10:28:19Rickthe 3x0 does doesn't it?
10:28:22Rickalso, can it be enabled?
10:28:36LinusNthe 300 uses a different version of the chip
10:28:48LinusNand it seems like it is hard to enable it
10:29:04Rickwhat would be hard about it?
10:29:10Rickwould it require hardware mod?
10:29:18 Nick Lynx_awy is now known as Lynx_ (HydraIRC@
10:29:24LinusNthe +5V USB voltage doesn't seem to be routed to the main board
10:29:51LinusNi'll have to investigate more, but it doesn't seem to be an easy mod
10:30:21BagderZagor/LinusN any objections on closing fixed/rejected bugs and features?
10:30:22LinusNespecially since they connected both USB and AC inputs to the external AC connector
10:30:31Bagderwe need to get the numbers down
10:30:34LinusNBagder: kill them all!!!
10:30:40Zagorfine with me
10:30:55LinusNwe desperately need a feature request moderator
10:31:22RickHehe :P
10:31:23Bagderok, so embrace for massive impact
10:31:25LinusNRick: we get a lot of those :-)
10:31:33RickHow often is the wiki paged updated?
10:31:44LinusNwhen we feel like it :-)
10:31:55Rickdone manually? god I hope not
10:32:07LinusNwiki is manual
10:32:08Bagderthe feature-request list is not wiki
10:32:23*Rick didn't realize that
10:32:35LinusNit's updated every night
10:32:46LinusN(swedish time)
10:32:47Bagdertwice per day in fact
10:33:03Bagder6 and 18 our time
10:33:44Bagderand now I'll force an update
10:38:56***Saving seen data "./dancer.seen"
10:40:47Bagderfeel free to join in and close a few ;-)
10:41:05Terminal-Velocith, i would prefer python:P
10:41:16RickI have the power to close them?
10:41:29*Rick didn't know that.
10:41:29BagderI got give you the power
10:41:34BagderI could
10:41:45RickIf you want, sure
10:41:56Bagderyou have a sf account user name?
10:44:43Bagderokay, now you're added
10:45:17Bagderall fixed/invalid/rejected entries can be closed
10:45:41Bagderthere's no good way except manually doing it
10:46:18 Quit Nibbler (Read error: 110 (Connection timed out))
10:49:52 Quit Terminal-Velocit (Read error: 104 (Connection reset by peer))
10:51:28 Quit izzy (Remote closed the connection)
10:51:30 Join izzy [0] (
10:53:23 Join jyp [0] (
10:57:28 Join Terminal-Velocit [0] (
11:02:00Terminal-Velocithow exactly do i build the firmware? tools/configure; make ?
11:02:20 Quit mirak (Read error: 60 (Operation timed out))
11:02:29Bagderthere are descriptions in the wiki
11:02:41Terminal-Velociti get an endless loop of make[207]: Entering directory `/home/tony/Projects/iriver/rockbox/firmware'
11:03:26LinusNin which dir did you run configure?
11:04:10Terminal-Velociti tried in rockbox/firmware first, then in rockbox/
11:04:21Bagderread the docs
11:04:22LinusNi suggest you read the docs
11:04:33Terminal-Velocitok thanx:P
11:04:42Bagderyou've ruined one Makefile now
11:04:47Bagderyou need to fix that
11:05:10Terminal-Velociti'll figure it out somehow
11:05:30Terminal-Velocitfor some reason i didnt notice that particular page while browsing the site:P
11:08:02Terminal-Velocithow dangerous is the bootloader, btw?
11:08:07Terminal-Velocitshould i risk it?
11:08:57Bagdergo go go
11:09:28RickTerminal-Velocit: it's pretty safe
11:09:38Terminal-Velocitcan it fry my precious?
11:09:45Terminal-Velocitoh that's good
11:09:47Rickit's possible, just not a common thing
11:09:53Rick(not 'fry' it though)
11:10:25Terminal-Velocitwell the md5sum i got was the same as the one on the wiki, so i guess i will be ok
11:11:33Rickthat's why that list exists ;)
11:13:11 Join mirak [0] (
11:14:15Zagorcurry time
11:14:42*Bagder wants curry too
11:15:00Zagordrive fast ;)
11:15:12Bagderno car here
11:15:13RickOr adbuct him with a spaceship
11:15:35Zagori'll think of you ;)
11:15:45amiconnBagder: Ask Scotty...
11:21:39*preglow has dry bread with brown cheese instead
11:22:47Terminal-Velocitit's not my day today
11:22:48Terminal-Velocitcalculator.c:1072: error: insn does not satisfy its constraints:
11:22:49Terminal-Velocit(insn 356 138 140 24 (nil) (set (reg:QI 9 %a1)
11:22:49DBUGEnqueued KICK Terminal-Velocit
11:22:49Terminal-Velocit (mem/f:QI (const:SI (plus:SI (symbol_ref:SI ("n"))
11:22:49Terminal-Velocit (const_int 3 [0x3]))) [0 n+3 S1 A8])) 37 {*} (nil)
11:22:49***Alert Mode level 1
11:22:49Terminal-Velocit (nil))
11:25:11Bagder? what tool versions are you using?
11:26:10LinusNlunch time
11:26:33 Quit jyp (Read error: 110 (Connection timed out))
11:26:58Bagder Feature Requests ( 493 open / 1143 total )
11:27:16 Join jyp [0] (
11:27:32preglowTerminal-Velocit: use gcc 3.4.x
11:27:34Terminal-Velocitby tool you mean gcc, binutils etc?
11:28:15Terminal-Velocithow lame i was using gcc-core-3.3.4
11:28:49Bagderit should work too
11:29:34Terminal-Velocitwell i'll try gcc 3.4.x anyway
11:29:58Terminal-Velocitcould it be binutils? i was using a cvs build from yesterday
11:30:26BagderI don't know
11:30:30dwihnoHow many rockboxers are there out there? 10 000? 20k? 50k?
11:30:44Bagderdwihno: good question, nobody knows
11:31:10RickIt's too bad we can't do an anonymous polling system like winamp ;P
11:32:16Terminal-Velocityou couldprogram rockbox to send pidgeons carrying stats to you
11:32:50***Alert Mode OFF
11:33:04Bagder"when you download Rockbox, you agree to also order these nice fully anonymous pigeons" :-)
11:33:05dwihnoBagder: how about tracking the number of ip's, downloading binary builds?
11:33:10preglowBagder: plenty problems with gcc 3.3.x on calculator.c
11:33:50Bagdernot for i386 though
11:34:01preglowahh, no, that's true
11:34:25preglowbut still, try 3.4.x, worth a shot
11:34:41Bagderindeed the simplest fix
11:34:55Bagderfor us
11:39:03preglowi don't even know what triggers it
11:39:27preglowi don't want to know, looking at the gcc source makes me queasy
11:43:08 Quit DMJC (Read error: 131 (Connection reset by peer))
11:43:23CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
11:43:23*Bagder cleans up
11:43:58Bagdermassive bug report closing
11:55:04preglowi don't bloody believe there's not easy way to negate an accumulator
11:58:56Bagdercall me mr janitor
11:59:42*Bagder sweeped away 40-something bug reports
11:59:53Bagdertime to eat!
12:07:09 Join cYmen [0] (
12:07:29 Quit Heidel (Read error: 60 (Operation timed out))
12:11:35preglowanyone want to refactor an imdct for me?
12:11:44*preglow almost forgets an appointment
12:11:48 Quit preglow ("off")
12:30:46 Join Aison [0] (
12:37:29Terminal-VelocitOMG i did it
12:37:36Terminal-Velociti built rockbox :D
12:38:58***Saving seen data "./dancer.seen"
12:39:18 Quit Bagder (Read error: 60 (Operation timed out))
12:39:25Terminal-Velocitlet's hope it wont send my ihp to the unknown
12:42:06ashridahdon't use your own built bootloader.bin
12:42:10ashridahuse the one off the website
12:42:25Terminal-Velocityeah that's the one i use
12:42:32ashridahthen you shouldn't have to worry about the iriver blowing up, since rockbox will be loaded off the hd, not from flash
12:42:37Terminal-Velocitand i also checked the md5sum
12:42:59HClwell. yesterday someone managed to brick his
12:43:22HClby loading rockbox, rockbox did something bad with the filesystem, and then neither rockbox nor original firmware would load
12:43:23Terminal-Velocitsaw that on the logs
12:43:41LinusNHCl: what makes you think rockbox had anything to do with it?
12:43:42HClwell, either rockbox or his OS
12:44:02HClLinusN: yea, you're right, bit sloppy of me, it might very well have been his OS
12:44:04Terminal-Velocitwell it didn't seem to me that it had anything to do with rockbox
12:44:26LinusNhe did unplug the usb without unounting the file system
12:44:38HCli do that all the time though..
12:44:48Terminal-Velociti never do that
12:44:56LinusNi always safely remove
12:44:59Terminal-Velocitfat32 is SOOO easy to break
12:45:58Lynx_what did they actually change that xp does not require you to 'safely remove' anymore?
12:46:44Lynx_or did they just remove the warning 2k gives and assume people will unplug only after writing is done...
12:46:59LinusNmy xp protests vigorously when i remove with out unmounting
12:47:24Lynx_hmm, mine doesn't and i don't remember turning anything off
12:47:40HClwhich versions?
12:47:43Zagorsounds like caching is turned off
12:47:43HCli have xp pro..
12:47:45LinusNmaybe it wasn't my xp after all...
12:47:51Lynx_and i remember this being a 'freature' of xp
12:49:46Terminal-Velociti have 2003 but haven't tried to see what happens when not cleanly unplugging
12:50:01Terminal-Velocitbut i guess they will complain
12:50:39Lynx_just try with some usb stick
12:51:10Terminal-Velocitnah. i don't plan to boot to windows soon :P
12:53:09DeadManthis is one reason I am waiting for Rockbox to be sorta finished
12:53:25DeadManno point flashing right now for me. not interested in games
12:53:28LinusNDeadMan: this will happen even when rockbox is "finished"
12:53:30Lynx_Terminal-Velocit: it does not reboot either way
12:53:40HCli think that if/once the bootloader can do usb, it should be fine
12:53:53LinusNHCl: that's a think i will look into
12:54:01HCl :)
12:54:03Lynx_Zagor: ah, you ar right about caching, that's the difference
12:54:20DeadManI want MP3 playback with gapless, true random playback and otf playlists and then I am happy. Nothing else. Everything else to me is just trimmings.
12:56:38Terminal-Velocitis low-level knowledge needed for writting plugins, or doeas rockbox provide something like a basic lib to abstract things?
12:57:03DeadManwell there is no API for audio yet
12:57:10DeadMandunno if anyone is working on it either
12:57:18HClTerminal-Velocit: abstract
12:57:22DeadManbut without audio
12:57:49Terminal-Velocitso it shouldn't be too hard writing a simple plugin right?
12:58:21Lynx_Ah, in the device manager win xp let's you "opitmize for quick removal" or "optimize device for performance" as caching options.
12:59:00HClnot bad
12:59:06Lynx_maybe that also changes the error message behavior, but i'm also too lazy to reboot ;)
12:59:24HCli must say, though i don't like the company, microsoft made a huge step forward userfriendlyness and stability wise with windows xp
12:59:49Terminal-Velocityeah but gnome is still better ;) :P
13:00:07HCli'd use linux, but linux simply doesn't support some things yet..
13:00:08Lynx_Terminal-Velocit: gnome also runs on xp with cygwin ;)
13:00:17HCllinux is better for wireless cracking though..
13:00:32Terminal-Velociteh yeah, but why use cygwin when you can have the real thing?
13:00:51Terminal-Velociti use cugwin only for it's x server
13:01:18Lynx_Terminal-Velocit: because then you can have windows also...
13:01:38Terminal-Velocitdo you need windows that desperately?:P
13:01:42Lynx_but gnome or kde on cygwin is not a good idea, it's to slow to be usable
13:01:58 Join deflate [0] (~vampire@
13:02:32 Quit deflate (Client Quit)
13:02:36Lynx_Terminal-Velocit: nah, i just used it very long and i am more comfortable with it. and there is nothing i can't do on it so far...
13:03:12Terminal-Velociti can understand that
13:03:40Terminal-Velociti had to use windows for one and a half week, and i was constantly opening command prompts and typing "ls" :P
13:04:19Lynx_Terminal-Velocit: acually since i installed cygwin some months ago i never used dir, ls is one letter less to type ;)
13:05:16 Join kevinlai [0] (
13:05:46kevinlaihi... i got a question regarding firmware 2.4...
13:06:30 Quit Terminal-Velocit ("BRB")
13:06:30kevinlaiwhen a player is charging, will the status on the LCD change when the battery is full??
13:07:54kevinlaii charged the player for almost 10 hour but the LCD keep showing charging and wont show if the battery is full or now
13:08:42kevinlainormally my player will show "battery charged" after 6 hour
13:10:13kevinlaibut somehow after i load the firmware into flash, the player keep charging and wont change the status on the LCD
13:12:11kevinlaiany1 here??
13:12:26amiconnThe player/studio models always charge as long as the charger is connected.
13:12:49amiconnThe archos firmware simply shows 'charged' after a defined amount of time
13:13:16amiconnThe charging is done completely in hardware, no chance of controlling it
13:13:17kevinlaiyes but not after i load the firmware into flash rom
13:13:57 Join Nibbler [0] (
13:14:09kevinlaib4 i load the firmware into rom, it shows charged
13:15:08kevinlaibut after i load into rom, it wont show "charged", it just keep showing "charging"
13:16:09amiconnYes, the archos firmware shows 'charged' after some time, but that's meaningless, it keeps charging anyway as long as you keep the charger connected
13:16:55kevinlaibut the problem is that it wont shows "charged"
13:17:19 Join Patr3ck [0] (
13:17:45kevinlaiit never shows "charged" even after charger connected for 10 hour
13:18:57kevinlaibut later i try to charge it using the original firmware (5.08), it shows "battery charged" after 6 hour of connecting to charger
13:19:16 Join Terminal-Velocit [0] (
13:20:07 Quit Aison (Read error: 131 (Connection reset by peer))
13:20:34 Join F1^Aison [0] (
13:25:12 Join [IDC]Dragon [0] (
13:25:46 Quit Terminal-Velocit ("Error loading new keyboard description")
13:26:06*[IDC]Dragon noticed amiconn
13:27:19 Join Renko [0] (
13:33:10 Quit kevinlai ("CGI:IRC")
13:33:25[IDC]Dragonbut amiconn doesn't notice [IDC]Dragon :-(
13:40:40Lynx_[IDC]Dragon: did you ever receive all the broken archos boxes?
13:47:52 Join DMJC [0] (
13:55:43 Join Heidel [0] (
13:57:56 Quit Schee4 (Read error: 110 (Connection timed out))
13:58:28 Join z63706c [0] (
14:00:35 Join Schee4 [0] (
14:08:38[IDC]DragonLynx_: yes, they're here now, but I only briefly checked
14:12:17 Quit fuzzie (Remote closed the connection)
14:14:03 Join R3nTiL [0] (~zorroz@
14:16:14 Quit Sando (Read error: 113 (No route to host))
14:16:22 Join zzz [0] (
14:16:22Lynx_[IDC]Dragon: they all came without hd, so they're kinda hard to check i guess?
14:20:49 Quit z63706c (Read error: 60 (Operation timed out))
14:22:23 Join fuzzie [0] (
14:23:44[IDC]DragonLynx_: yes, and that's what I haven't done yet
14:24:02 Quit lostlogic ("Going to the moon")
14:24:02[IDC]Dragonjust plugged some in to check for any sign of life
14:27:58 Join bobTHC [0] (
14:28:25 Join preglow [0] (
14:39:00***Saving seen data "./dancer.seen"
14:44:20 Quit ashridah ("sleep")
14:44:48jypAnyone minds if change SOFTWARECODECS to be the list of wanted codecs instead of a mere yes/no?
14:45:16jypBecause I don't plan to support all codecs on gmini, at least all at once
14:46:40jyppreglow? LinusN?
14:53:14preglowdon't ask me, i don't deal with the build system
14:58:13 Nick DMJC is now known as DMJC-sleep (
14:58:33jypSo here, a question for you preglow ...
14:59:04jyphow come the part inside # if CONFIG_CPU==MCF5249 gets compiled ?
14:59:41jypOr is it considered build-system too ?
15:00:15preglowit gets compiled on gmini?
15:00:30preglowwell, then CONFIG_CPU is quite clearly set wrongly
15:00:33jypI mean, it tries to ;)
15:01:14jypIs there another config.h ?
15:01:31preglowi really don't know
15:01:36preglowi haven't dealt with that
15:01:46preglowif CONFIG_CPU is what everyone else uses
15:01:48preglowso i used it as well
15:02:06jypits usage is fine...
15:02:18 Join lolo-laptop [0] (~lostlogic@
15:02:19jypyet I wonder why it's defined wrongly
15:02:34LinusNjyp: where is that code?
15:03:26jyprockbox/apps/codecs/libmad/layer3.c:2166: error: unknown register name `a5' in `asm'
15:03:35 Join Bagder [0] (
15:04:38Bonkersnot running in the simluator?
15:04:59Bonkerswell compiling for
15:05:08jypno; gmini target
15:06:45jypwhy do you insist on "silent" compilation btw ? It makes quite hard to track these problems
15:07:08Bonkersand ARCHOS_GMINI(120/SP) is definitely defined?
15:07:27jypApparently not
15:07:30LinusNjyp: because it is hard to see the warnings with all the output
15:07:36jypI'm trying to understand why
15:07:38preglowit's badger! no wait
15:08:09preglowi love silent compilation, really easy to turn it off anyway
15:08:16jyphow so ?
15:08:30preglowjust remove a couple of @'s here and there, no?
15:08:41preglowthat's have i removed silent objcopy and objdump, at least
15:08:47preglowahh, just objcopy
15:08:51jypfor each makefile ...
15:08:58preglowno, believe has it
15:09:15jypnah, each codec makefile has its own
15:09:19 Join kurzhaarrocker [0] (
15:09:22jypjust found it
15:10:24jyparg! calmrisc16-unknown-elf-gcc ... -DARCHOS_GMINISP ...
15:11:28*Bagder comes in in the middle of a discussion
15:11:30jypproblem is CONFIG_CPU=MCF
15:11:47jypdespite the target being gmini
15:12:31jypAlso, maybe you can answer my earlier question; do you mind if change SOFTWARECODECS to be the list of wanted codecs instead of a mere yes/no?
15:12:40jyprockbox/apps/codecs/libmad/layer3.c:2166: error: unknown register name `a5' in `asm'
15:13:45kurzhaarrockerscandisk reports problems for each file rockbox has written. Is that a known problem? (I run quite an old version of rockbox)
15:13:47Bagderdoes the codecs include rockbox's config.h properly?
15:14:14jyp# ifdef HAVE_CONFIG_H
15:14:14jyp# include "config.h"
15:14:14jyp# endif
15:14:25Bagderand is that defined?
15:14:29jypSo, no ;)
15:14:34LinusNkurzhaarrocker: weird
15:14:42LinusNnot a known problem
15:15:02jypShall I drop the ifdef HAVE_CONFIG_H ?
15:15:12Bagderor add the include in global.h
15:15:29Bagderas I believe that is included by all codecs
15:15:34kurzhaarrockerOk then I'll report back with a more precise problem description tomorrow, LinusN
15:16:03jypAin't it ../codec.h ?
15:16:24LinusNjyp: it must have included config.h, else CONFIG_CPU wouldn't have been MCF5249
15:16:31Bagderjyp: right, it is
15:16:52jypLinusN: all I can tell you is adding the include manually fixes the problem
15:17:05LinusN"the include"?
15:17:39jypadding #include "config.h" to codec.h
15:18:01BagderI believe that is a correct fix
15:18:28LinusNbut how did it include config-h100.h in the first place?
15:18:46Bagderwell, possibly both were undefined and treated as zero?
15:19:23jypmust be
15:19:25LinusNaha, so #if VAL == num is true if VAL is undefined?
15:19:45 Quit kurzhaarrocker (Remote closed the connection)
15:21:00jypLinusN: yes
15:21:08 Quit R3nTiL ()
15:21:14jypjust tested:
15:21:15jyp#if A == B
15:21:15jyp#error ouch
15:21:15DBUGEnqueued KICK jyp
15:21:19LinusNman, and i thought i knew C...
15:21:53Bagderbut gcc warns for that, doesn't it?
15:22:03jypboth must be undefined though
15:22:17jypno warning
15:22:32Bagderit might need some -W option to make it warn then...
15:22:42jypeven with -Wall
15:23:37jypAlso; the codec makefile overrides the optimization level
15:23:49jypis this wanted behaviour ?
15:24:14BagderI doubt that
15:24:47jypI'm not sure with option takes precedence;
15:24:59jypie. gcc -Os -O2 ...
15:25:10jypwhich will prevail
15:25:12Bagderme neither
15:25:39Bagderimho, the codecs should not set those options but use the globally set ones
15:25:56jypmy thinking too; I'm removing it
15:26:47jypYou're ok with the codec list thing ?
15:27:06LinusNi can imagine that the codecs might want different optimization flags than the rest of the system
15:27:12Bagderjyp: I am
15:27:30preglowLinusN: yes, so can i
15:27:45preglowinlining might be pretty critical for some codecs
15:28:38LinusNwhile the rest of rockbox can be optimized for size, you don't want the codecs to be too slow
15:29:23Bagderwon't we then add specific opts with -f instead ?
15:29:46Bagderif not, we should set the -O thing in a separate variable in the root Makefile so that the codecs can override it easier
15:29:47jypok, I leave as is for the time being; committing the compilation of libmad for gmini
15:33:12 Nick Strath is now known as StrathAFK (
15:35:20preglowthis emac stuff would be pretty cool if it was a bit easier to manipulate the acc registers
15:38:31LinusNwhine, whine... :-)
15:41:26preglowi read the acc register and the extension word, i do a double sub from zero with carry,then store it back, and bam, accumulator now contains garbage instead of negated value
15:42:33jypsub zero wins!
15:43:46preglowdouble fatality
15:48:38[IDC]Dragonpreglow: can't you multiply it with -1 ?
15:50:40Bagderjyp: the libmad Makefile commit looks like a mistkae
15:51:17Bagdersince it just removes the @
15:52:09jypyou're right
15:54:18preglow[IDC]Dragon: no, i can only add or subtract to the accumulator
15:55:43LinusNpreglow: show me the code
15:59:30jyppreglow: I had to slightly change libmad to implement a workaround a gcc bug; (we talked about this some time ago)
15:59:49jypperhaps you'll want to double check it ...
16:00:01jypeven if it is of trivial nature
16:00:09preglowLinusN: hold on, i think i've figured out the bug,
16:00:55amiconn[IDC]Dragon: Now I'm here
16:00:59Bagderjyp: you checked any decoding performance with libmad on target yet?
16:01:37jypnot yet
16:01:50jypI expect it to be very bad though ;)
16:01:55preglowLinusN: it's the negacc function
16:02:21preglowi think i need to sign extend the top extension byte
16:04:14 Part jyp ("poof!")
16:04:19preglowoh, and the clobber list is on crack
16:04:20 Join jyp [0] (
16:04:30preglowbut that's just me forgetting to revert it
16:04:52LinusNoops, gotta run
16:04:58 Part LinusN
16:22:08[IDC]Dragonamiconn: got my email?
16:22:28amiconnSent today? Then not (yet)
16:22:45[IDC]Dragontoday, late morning
16:23:00amiconnI saw there's another aucktion on ebay, 5 defective archos'es
16:23:09amiconn(same vendor iirc)
16:23:25[IDC]Dragonoh, he withheld some?
16:23:30amiconn3 Studios, 1x V1, 1x FMR
16:23:34amiconnall without hd
16:24:00amiconnI consider bidding...
16:25:00[IDC]Dragonthere's another auction, V1 with bad display
16:27:37[IDC]DragonOndioSP for 42 EUR, BuyNow
16:28:07amiconnAbout the same I paid...
16:28:16amiconnThey usually go cheaper now
16:30:08[IDC]DragonI can't find the broken batch
16:32:41[IDC]Dragonfeel free, I'm not competing :-)
16:33:11[IDC]Dragonthe V1 I got seem more or less alive
16:33:35[IDC]Dragonthe older recorders (USB1) are all dead
16:33:38 Join mecraw [0] (~mecraw@
16:34:51[IDC]DragonI even got the inner part of a V2 and an FM, but both dead
16:36:18[IDC]Dragontoo bad, else I could have revived my bursted FM
16:39:04***Saving seen data "./dancer.seen"
16:46:00 Join davae [0] (
16:46:51 Join DrRick [0] (
16:52:21amiconn[IDC]Dragon: Did you already try logo.rock on the player?
17:02:30 Quit davae ("Lost terminal")
17:10:36 Join Terminal-Velocit [0] (
17:26:55 Quit Renko (Remote closed the connection)
17:29:01 Join Sucka [0] (
17:29:04 Join sofaSpud [0] (~john@
17:32:16Zagoranyone feeling like playing with the m68k gcc might want to have a look at this patch that adds the -mregparm option, passing function parameters in registers instead of on the stack:
17:32:39Zagorcould speed things up for us with our slow ram
17:32:54amiconnZagor: That's not default with m68k? Strange...
17:33:14amiconnSH1 gcc passes parameters in regs for functions with <= 4 parameters
17:33:29amiconn(using r4..r7)
17:33:44Bagder68k has a different ABI convension iirc
17:33:57preglowamiconn: 68k has stack based parameters by default
17:34:02Zagorthe mregparm option "voilates" the ABI
17:34:14Bagderbut could speed things up for us
17:34:16preglowwho cares, i say we use it if we can
17:34:30preglowbefore i write any more assembly :P
17:34:46amiconn(the 'urgs' was related to the default, not to the abi violation)
17:36:04preglowdo you know how they're passed? pointers go in d registers ?
17:37:32Zagori don't know
17:39:09preglowwell, we should use it anyway, it'll be faster and make the code a bit smaller
17:40:00preglowseems like the patch was ignored pretty thoroughly, though, would be nice having it in the official gcc
17:41:17BagderI believe Linus' patches were received pretty much in silence (too)
17:41:50amiconnpreglow: From a quick look at the patch, it seems that the number is settable (default = 2, max = 4). Then <number> registers of each type (all of address, data and fp) are used for regparms
17:43:24amiconnOf course coldfire has no fp registers
17:43:25preglowmax four?
17:43:28preglowwhy not more?
17:44:05amiconnThat's more than on sh1... because 4 means 4 data registers _and_ 4 address registers (and 4 fp registers on 68k with fpu)
17:44:20preglowahh, great
17:44:35amiconnLook at the patch file, near the end
17:53:38 Join Lmcmann [0] (
18:05:33jyppreglow, anyone else, what would be a decent size for the mp3 buffer?
18:05:34jyp1mb ?
18:07:15preglowas large as you can make it?
18:07:25preglowor no, gmini is a flash player?
18:07:48jypno, hd
18:07:59preglowthen use as much ram as you can
18:08:08jypthe problem is that I'm rather ignorant of the memory mapping
18:08:39jypI'm asking what would be the minimum size for a test
18:09:06 Quit kergoth (Remote closed the connection)
18:09:15 Join kergoth [0] (
18:09:28preglowwell, any size will do
18:09:34preglow1 meg should be ample
18:10:42 Nick Terminal-Velocit is now known as tvelocity[away] (
18:10:46jypthank you
18:11:41jypAny alignment requirements ?
18:11:55preglownone that i can think of
18:20:20 Join kergoth_ [0] (
18:21:02 Quit kergoth ("leaving")
18:21:03 Nick kergoth_ is now known as kergoth (
18:22:50 Quit Patr3ck ()
18:25:26 Quit Lmcmann (Read error: 110 (Connection timed out))
18:34:30jyptime to go
18:34:35 Quit jyp ("poof!")
18:39:05***Saving seen data "./dancer.seen"
18:44:12amiconn[IDC]Dragon: G6t y64r 0a53 n6w,,,
18:44:25amiconnOops, meaning 'Got your mail now'
18:44:32amiconn(dreaded numlock)
18:46:11 Join Stryke` [0] (
18:48:29preglowand i'm off
18:48:31 Quit preglow ("laul")
18:50:40 Quit bobTHC ("( :: NoNameScript 3.81 :: )")
19:00:15 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:01:29[IDC]Dragonamiconn: I thought you're practicing this funny speak, what's it called?
19:03:23amiconnNo, not practising 1337
19:04:07[IDC]DragonI have lo clue, does it translate to leet?
19:04:27[IDC]Dragonand what's the story?
19:08:10[IDC]Dragonok, google told me
19:10:00[IDC]DragonOnce used by the hacker community as sort of a sub-culture language, ANYONE using this type of speak in the present day in age is either: A) A clueless net newbie who somehow thinks its still "cool" to speak this way. B) A veteren net user who uses it to parody the type of people mentioned in example A.
19:15:59DeadManl33t $p3a|<?
19:16:05 Join lImbus [0] (
19:16:19lImbuswow. we got ops :-)
19:16:23DeadManl33t $p34|< hehe
19:16:36DeadMananyone who types like that is a dumbass
19:17:00CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
19:17:00*lImbus bows before ops and, of course, logbot
19:17:11DeadManI will h4x0r j00 ;)
19:17:39DeadManwas amusing for a while. amazing how some still type that way on some channels
19:17:43*amiconn doesn't see any ops...
19:18:54 Join DrRickDee [0] (
19:19:35DeadManwow! in China they are building the worlds biggest dam. it can generate electricity the equivalent of 10 nuclear power stations
19:21:06DeadManbut they crap all over their people by building these dams :P
19:21:48lImbusamiconn: lol. It's hydrairc's buddy group :-S
19:22:37amiconnMuhahaha. In fact, ops are pretty easy to spot in Hydra, even if they are also in a buddy group, because they're coloured red
19:24:58lImbusyes, I remember. I should practice more irc :-)
19:27:20 Quit DrRick (Read error: 110 (Connection timed out))
19:31:15 Join R3nTiL [0] (~zorroz@
19:33:35Lynx_DeadMan: how many hours of mp3 playback is equivalent of 10 nuclear power stations?
19:37:08Suckahow long do the nuclear power stations run for
19:38:08 Quit R3nTiL ()
19:39:31Lynx_Sucka: that depends entirely on which country they are located in ;)
19:43:36mirakI would like to buy the iRiver h320
19:43:49mirakhowever I have seen it can't record in wav
19:44:04mirakI mean the normal firmware
19:44:17mirakshould rock box be able to record and read wav ?
19:44:54Bagderit will, yes
19:46:21mirakok I buy it then :)
19:46:31miraka bit expensive but well
19:50:53mirakBagder: I was wondering something about the host to host usb
19:51:05BagderI know nothing about that
19:51:13mirakdoes this mean that a usb device vcan be pluggued on the iRiver
19:51:23miraklike if it's just a usb port
19:51:31Suckaeuropean h320s i believe but not american
19:51:39miraks ?
19:51:42mirakwhat means the s
19:51:44mirakah ok
19:51:47mirakit's plural
19:52:10mirakso we could be able to plug a keyboard on it ?
19:54:05Suckano i think its more for cameras etc
19:54:11Suckato get pictures off them
19:54:38Soul_Eateryeah i was wondering that myself.
19:54:55Soul_EaterIf hardware components could be made like a keyboard to work with the iriver usb port, with different firmware
19:55:02mirakbut it means there is a UMS driver
19:55:18mirakon the h3xx side
19:55:52miraktechnicaly it makes sens
19:55:54mirakI think
19:55:54 Quit tvelocity[away] ("Error loading new keyboard description")
19:56:19mirakexept maybe about the power
19:56:32mirakwith a auto alimitented usb hub it could be ok
19:57:39 Nick Lynx_ is now known as Lynx_awy (HydraIRC@
20:16:06 Join loki1228 [0] (
20:17:47loki1228does rockbox support Ogg?
20:21:06loki1228cool - I know it runs on the iRiver h120. Waht about the h3XX?
20:25:02lImbusneither :-)
20:25:23lImbusrockbox does not yet run in a "install-and-go" manner on either iRiver-device
20:25:37 Join nozomiyume [0] (
20:25:48loki1228ok - i guess it takes quite a bit of work
20:29:36Soul_Eaterit hasnt even been started for the h3xx
20:30:53 Quit [IDC]Dragon ("CGI:IRC")
20:32:32 Quit lImbus (" brb, reboot")
20:35:35Soul_Eaterso anyone alive i have an incredibly easy math program
20:35:38Soul_Eaternot program
20:35:42Soul_Eatermath question
20:39:08***Saving seen data "./dancer.seen"
20:41:24 Join lImbus [0] (
20:44:22 Join Lmcmann [0] (
20:52:34 Join otsego [0] (
20:55:53 Join bg_ [0] (~chatzilla@
20:59:21Soul_Eaterwhats the percent of 66/23
21:00:06 Join stevenm [0] (
21:00:55Lmcmannthe percent of 66/23's?
21:01:14Lmcmanndivide 66 by 23 and read off the number
21:01:41Lmcmanntwo hundred and 87 percent
21:01:49 Part otsego
21:02:06Soul_Eateri mean if 66 = 100%, what percent does 23 make up of that
21:02:12LmcmannI think we need a bit more info on what you mean by percent
21:02:20Lmcmanndifferent question
21:04:41Soul_Eaterhow did you come to that
21:05:17CoCoLUSor (100/66)*23 :)
21:06:09Lmcmannlook at it like this. iF 100% of something is 66, and you have 23 of them, that's about one third, so about 33%
21:06:54stevenmAre you looking for what percent of 23 is 66, or vice versa ?
21:07:19Lmcmanndivide how many you have out of the total number to get the percent
21:07:46Lmcmannmove the decimal point two places to the right, and read off the number.
21:07:54Schnueff2/3= .666 the number of the beast −−- scary
21:08:33LmcmannFollow what we're saying?
21:09:44LmcmannActually, I had a RockBox question. Anyone here know about the display settings?
21:10:05stevenmanyone know when Linus or stripwax are coming back ?
21:10:19stevenmOr, anyone know anything about Gravis Envelope Bytes for that matter ?
21:10:21LmcmannI have an FM, just put todays build on, and can't change the volume and batt display back to numeric.
21:10:33LmcmannHmmm, never heard of them.
21:10:42LmcmannDoesn't help, I know.
21:11:07stevenmYea I even looked at the Timidity++ source code, but from their comments, it doesn't look like they are too sure about how to process that data
21:11:36LmcmannWhat are you using Timidity for?
21:13:23 Quit bg_ ("Chatzilla 0.9.67 [Firefox 1.0+/20050201]")
21:14:49Soul_Eateroh ok
21:14:53Soul_Eateri was dividing wrong
21:21:29 Nick StrathAFK is now known as Strath (
21:23:27 Part Lmcmann
21:27:58 Quit Sucka ("a bird in the bush is worth two in your house")
21:35:35 Join jyp [0] (
21:38:55 Quit edx (Read error: 145 (Connection timed out))
21:46:23 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
21:47:06 Quit YouCeyE ("Leaving")
21:48:15 Join Stryke` [0] (
21:50:20 Join Acoc [0] (
21:51:18Acochey guys- I just took apart my jukebox 6000 and that little piece of metal that holds one of the springs down broke
21:51:28Acocis it absolutely necessary
21:59:18 Join preglow [0] (
22:01:21CoCoLUSdoes is work now? if yes, then no, its not necessary :P
22:01:31preglowthe aligned malloc was very clever
22:01:37preglowwhy the hell didn't i think of that
22:02:41CoCoLUSbecause you're not clever enough? :)
22:04:05 Join LinusN [0] (
22:08:57 Quit stevenm (Read error: 110 (Connection timed out))
22:09:34AcocI tried it once without solder and it started smoking
22:10:01Acocfigured something was touching something bad
22:10:21DeadManSo it's time for my daily 'is anyone working on audio out on the iRiver?' question :)
22:10:57Acoclinus do you happen to know if that little metal piece holding the spring is important
22:13:06pabsAcoc: just send me the jukebox 600 and i'll tell you whether or not it's necessary :D :D :D
22:14:17Acocwhen I catch it on fire your welcome to it
22:16:40Acocalright time for a quick go of it now that it's soldered
22:17:25preglowDeadMan: you can work on it!
22:17:56DeadManI wish I knew how
22:18:02preglownever to late to teach that
22:18:05preglowlearn, i mean
22:18:23DeadManby the time I learn it will be done lol
22:18:27preglowyes, sure :)
22:19:57DeadManY HELO THAR
22:20:46Acocalright no smoke this time and the poor connection I was having was fixed
22:22:33LinusNAcoc: it isn't important, but it helps retracting the spring to ease the pressure on the pcb
22:28:34 Quit Schnueff ("leaving")
22:30:29 Join skav [0] (
22:31:57 Join stevenm [0] (
22:33:13 Nick DMJC-sleep is now known as DMJC (
22:37:41Soul_Eaterfuck i forgot what i was doing
22:39:09***Saving seen data "./dancer.seen"
22:40:37miraklImbus: hey, should it be possible to use usb device via the usb 1.1 port of the Hxx ?
22:40:58preglowwhat the hell? is .word 32 bits on arm gas?
22:41:01mirakdoesn't the iRiver just use a UMS driver to acces to cameras ?
22:41:15mirakword is always 32 bits
22:41:24mirakhum no
22:41:28mirakthat's 16 :)
22:41:49preglowyes, that i know
22:42:44preglowbut some guy declares a list of 32 bit numberrs with .word here
22:43:56 Join [IDC]Dragon [0] (
22:44:26lImbusmirak, what do you call use ?
22:45:07lImbusI've seen working a archos 20 rec as external hard drive for a iRiver H340 :-)
22:45:34lImbusi'm not sure if it's solely "usb-to-go" or full usb host adapter
22:46:00mirakI eman, plug a keyboard usb or a mouse
22:46:32 Quit [IDC]Dragon (Client Quit)
22:47:11lImbuswith usb host adapter yes (there would've been the driver for the keyboard to be written though) with usb-to-go not
22:47:22 Join [IDC]Dragon [0] (
22:47:42[IDC]Dragonhi again
22:48:00miraklImbus: what is the difference ?
22:48:18[IDC]Dragonamiconn: do you have any opinion about my MMC panic?
22:48:27lImbuswell, usb-to-go is just for external usb-mass-storage devices
22:48:40[IDC]Dragonelse I'll plaster the code with splashes
22:49:00BagderlImbus: I don't think that is true
22:49:03miraklImbus: hum yes, but the control is not hardware isn't it ?
22:49:17miraklImbus: isn't it the firware that control this ?
22:49:29mirakwhy wouldn't it work ?
22:49:46 Join stripwax [0] (
22:49:52lImbusBagder, you think usb-to-go can have all usb-stuff ?
22:50:00BagderOTG is not restricted to usb-storage
22:50:21lImbusthen the only difference would be the maxcount of devices (1 vs. 254 I suppose)
22:50:31BagderOTG simply means it can serve in both ends of a USB communication
22:50:58mirakcool we will be able to run gnome on it then
22:51:21Bagderwell, in this particular case on this chip, it might very well be limited to usb-storage of course
22:51:28lImbuswhy does the iRiver H340 then have a usb2.0 (usb-mini-b) and a separate usb1.1 (usb-mini-a) connector ?
22:51:46BagderI don't know
22:52:00 Quit Stryke` (Read error: 104 (Connection reset by peer))
22:52:10Bagderbut I'm not a USB wizard, I might be wrong
22:52:58 Part LinusN
22:53:15 Join stripwax_ [0] (
22:56:55miraklImbus: well maybe because it gives two way then
22:57:06miraklImbus: 1.1 incomming
22:57:10mirakand 2.0 outgoing
22:57:19miraknot worth having 2.0 incomming
22:57:28mirakthat's my theory
22:57:59Bagderi586-mingw32msvc-gcc −−version
22:57:59Bagderi586-mingw32msvc-gcc (GCC) 3.4.2 (mingw-special)
22:58:18 Quit stripwax (Read error: 60 (Operation timed out))
22:59:11 Quit Acoc ("ChatZilla 0.9.61 [Mozilla rv:1.7.2/20040803]")
23:00:58stripwax_is irc playing up or is there really no one here
23:01:02 Part stripwax_
23:01:09 Join stripwax_ [0] (
23:01:17Bagderthere is people here
23:01:34*lImbus waves
23:02:05stripwax_hmm, I see nothing in the user list, hi anyways!
23:02:23[IDC]Dragon57 people here
23:03:14 Part stripwax_
23:03:15[IDC]Dragonamiconn, are you there?
23:03:46[IDC]Dragon(stripwax_ got scared by the masses)
23:04:38BagderI found out there is a win32 cross-compiler already done as a debian package
23:05:00Bagderso the win32 sim cross-compiles will soon be made with gcc 3.4.2
23:05:17 Join [IDC]Dragon2 [0] (
23:05:48[IDC]Dragon2my WLAN got scared, too
23:06:19 Join stripwax [0] (
23:06:25stripwaxahh, better. ello again
23:07:02 Join edx [0] (
23:07:46 Part stripwax
23:08:00 Quit Zagor ("Client exiting")
23:13:44 Join [IDC]Dragon3 [0] (
23:14:03[IDC]Dragon3this isn't funny
23:14:33[IDC]Dragon3since about a week, my WLAN is very intermittant
23:14:47[IDC]Dragon3I have no clue why
23:14:52lImbussure it's not t-online ?
23:14:58Bonkersmaking lots of microwave popcorn?
23:15:06[IDC]Dragon3no, the wired PC is ok
23:15:30[IDC]Dragon3reception can't be better, the acess point is in the same room
23:19:31 Join silencer [0] (
23:24:33Rickwohoo! ;)
23:24:39Rickmy h120 remote just got delivered
23:26:41 Quit [IDC]Dragon (Read error: 110 (Connection timed out))
23:27:21Rickthe wire is backward
23:28:17Rickcoulda sworn the wire
23:28:21Rickwent out on the left side
23:28:22Ricknot the right
23:28:28Rickdid they update the remote design?
23:28:33Bagdermine goes out to the left
23:28:56Rickyeah, the one I just got goes out the right
23:28:58Rickguess they updated it
23:29:16 Join matsl [0] (
23:31:01mirakdo you plan to support pmp ?
23:31:28Ricklet's support h1x0 first
23:32:24Bagdermirak: we'll support all platforms people make rockbox work on
23:32:33Bagderwithout exceptions
23:32:46mirakpmp is a bit to expensive
23:32:51mirakwell they are all to expensive anyway
23:33:04mirakI plan to buy a h320
23:33:11mirakbut when I see the pmp I say to myself
23:33:27mirakwell for 130 more euro I can got full video
23:33:41mirakbut that's about twice the price of the h320 more
23:34:46 Quit [IDC]Dragon2 (Read error: 110 (Connection timed out))
23:36:34 Nick [IDC]Dragon3 is now known as [IDC]Dragon (
23:38:52stevenmpreglow, Do you think it would be fine for the MIDI codec to just ignore all the envelope data (for now at least) ?
23:39:14 Quit sofaSpud ("There are lies, damn lies, and MS facts")
23:42:01Bonkersno, prefection is the only option
23:42:47preglowstevenm: of course
23:42:50preglowstevenm: it can always be added
23:43:08preglowi can code it if i get time some day
23:44:00 Join Patr3ck [0] (
23:44:04stevenmpreglow, I have spent the past three days pouring over the numbers and encodings for the env rate/offset
23:44:31stevenmmy roommate even helped me, and he is a math genius.. and he couldn't see any pattern
23:45:11stevenmI am seeing a small correlation between some things but not others.. ie, some patches follow a general ratio but others do not. I think it so far sounds well enough without release and junk
23:45:39 Quit lolo-laptop ("Client exiting")
23:46:11stevenmpreglow, now all that is left is pitch bend, and it doesn't look like that much of a pain. First parameter doesnt seem to do anything, and the second is 32 points for each MIDI note
23:46:38preglowi there's probably a lot of midi cc messages to implement as well
23:46:45stevenmcc ?
23:46:51preglowcontrol change
23:46:57preglowvibrato, reverb, chorus, etc
23:47:10stevenmoh.. at the moment I have volume and panning. don't really know what the other ones do
23:47:35stevenmand I dont even have files that I've seen those in.
23:47:35preglowvibrato is probably needed
23:47:37preglowwe'll see
23:47:54stevenmI have a whole lot of files I've been testing it on
23:47:59preglowhmm, mod wheel is vibrato, i think
23:48:03preglowhave you implemented that?
23:48:08stevenmmod wheel ?
23:48:19preglowi'll take that as a no :P
23:48:37stevenmno.. is that the same as pitch wheel ?
23:48:54preglowwell, on a midi keyboard, it is usually placed next to it
23:49:46stevenmah... never owned any midi equipment at all :(
23:49:58preglowhaven't got much myself
23:50:05preglowa midi keyboard and a sid station
23:50:14stevenmi never even used any either... other than a computer
23:51:04stevenmnow, I don't suppose there are plans to interface a MIDI keyboard to the thing, and have it record ?
23:51:06preglowwish i could afford more synths to play with
23:51:10preglowi'm a certified synth freak
23:51:43stevenmthat's good.. this stuff must not be new for you then
23:51:49preglowno, not really
23:52:35preglowi've done a lot of synth programming as well, so i'm quite familiar with it
23:52:58stevenmI'm about to resort to opening each and every intsrument file with this windows program, write down the env data and hard code it into the code.. or in a config file
23:53:34preglowcan't you get any sense out of the timidity parser?
23:54:05stevenmI used their routines to convert the GUS encoded bytes into actual values
23:54:16stevenmthe values themselves make little sense
23:54:22preglowgreat, then the rest should be easy
23:54:29stevenmoh... just you wait
23:54:36stevenmcomments such as this occur throughout the file :
23:55:44stevenmstatic int32 convert_envelope_offset(uint8 offset)
23:55:44stevenm /* This is not too good... Can anyone tell me what these values mean?
23:55:44DBUGEnqueued KICK stevenm
23:55:44stevenm Are they GUS-style "exponential" volumes? And what does that mean? */
23:56:17stevenmnot too encouraging
23:56:53stevenmI took the converted rate values from 2 instruments and divided their release times by them, and in both cases, the ratio is 1.818
23:57:05stevenma third instrument, did the same thing.. not even CLOSE
23:57:46stevenmthere is also the scale_freq and scale_factor parameter in the patchfile that no docs explain at all.
23:58:11stevenmthere appears to be only one or two GUS docs floating around, all written by one person and just put in different format

Previous day | Next day