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 2005-02-26

00:00:16 Join amiconn_ [0] (
00:00:49preglowoutput/input constraints, i mean, "+whatever" ones
00:02:43 Quit [IDC]Dragon ()
00:06:24jypIt's not a bug
00:06:24rasherpreglow: no shit, it's pimpalicious!
00:06:24rasherI especially like the leopard
00:06:24preglowwhy not? gcc never discards asm blocks that have regular output statements
00:07:20jypMaybe I didn't get what you meant...
00:07:47jypI mean, you need to have a non empty output list
00:08:21preglowit is none empty
00:08:22preglow: [X] "+a" (X), [yptr] "+a" (yptr) : : "d0", "d1", "d2", "d3", "d4", "d5", "d6", "d7", "a0", "a1", "a5");
00:08:48 Quit amiconn (Read error: 60 (Operation timed out))
00:08:48 Nick amiconn_ is now known as amiconn (
00:09:01HClwhats the [X]
00:09:04HClhey amiconn :)
00:09:08HClguess what
00:09:20preglowHCl: a way to not have to write %0 %1, i can write %[X]
00:09:36preglowHCl: that way i don't have to rewrite all code if i need more asm arguments
00:11:36preglowso, now imdct_s has the tightest bleeding code seen on this side of the planet
00:11:43preglowtoo bad imdct_l will be a hundred times worse
00:25:28 Quit midk ("Leaving")
00:35:19 Quit methangas (" Like's GUI? Then try HydraIRC -> <-")
00:36:12preglowyes, remember to increment loop iterator,yes
00:40:49 Quit gromit``` (Read error: 104 (Connection reset by peer))
00:41:50 Join gromit` [0] (~gromit`
00:49:06 Quit skav (Read error: 104 (Connection reset by peer))
00:51:50 Join markun [0] (
00:53:10 Quit bg_ (Read error: 110 (Connection timed out))
00:53:17markunpreglow: libfaad2 (aac decoder) implements mdct using fft. It'a also fixed point:
00:54:07markunI wonder if it's faster than the mdct used by Tremor.
00:56:52preglowshoulnd't be too hard to find out
00:57:13preglowif we only had a good ifft routine provided to us by motorola, i suspect the answer would most certainly be yes
00:58:24preglowxiph themselves say the imdct algo in tremor has plenty room for improvement
01:00:28markunYes, that was also proven by the huge speed improvements by the guys who wrote their master thesis about Tremor on a DSP.
01:00:53***Saving seen data "./dancer.seen"
01:02:32preglowyes, as long as you bear in mind the fact that they had near perfect optimized fft routines for their platform at hand
01:03:05 Nick Strath is now known as Priestess (
01:03:45 Nick Priestess is now known as Strath (
01:04:23 Quit jyp ("poof!")
01:06:04markunI'm also still looking into unicodifying rockbox. It works ok, but I still need to think of a way to load only part of a font.
01:06:43preglowfonts getting big on you? :P
01:07:39markunWell, for iriver it doesn't matter too much I think. I now have a font with about 1500 glyphs 6x13 pixels which is 20kB.
01:09:20 Join ashridah [0] (
01:09:39markunThe textviewer also works fine as long as you only use utf-8. I should make some character set convertors.
01:10:03markunI now only have utf16 and latin1 to utf-8
01:10:59markunhi ashridah
01:11:55 Quit shx ("CGI:IRC (EOF)")
01:12:56preglowutf8 is what counts anyway :)
01:15:04*ashridah nods
01:15:06markunYes, a lot of tags are in UTF-8 (all of the Ogg tags anyway) and converting filenames from UTF-16 to UTF-8 was easy.
01:18:38markunThe language files also work fine after converting them to UTF-8. Do you think unicode could/should become a part of rockbox or do you think many people have no need for it?
01:23:14*markun should go to sleep
01:23:19 Quit markun ("good night")
01:27:59preglowif you happen to be reading logs, i think rockbox should have unicode unless it bogs things wayyy to much down
01:31:10 Join prpplague [0] (
01:32:39 Join kergoth [0] (
01:32:46HClhey kergoth.
01:32:52HClwhat brings you here?
01:32:53kergothheh, hey
01:33:00kergothpicked up an iriver h320 the other day
01:33:03prpplagueconversation on the h320
01:33:17HClhows oz going?
01:33:49kergothmoving along well. 2.6 on the clamshells is now to the point where its more usable than 2.4 (sd/mmc is more stable for example)
01:34:01HClsounds like i really need to upgrade
01:34:03kergothfamiliar and oz are both putting out incremental releases on a regular basis
01:34:03HCl :P
01:34:10kergothi'd wait for 3.5.3, it releases in march here
01:34:13HClok :)
01:34:15kergothwill be the best release to date
01:34:17HClgreat news :)
01:34:38HCli still need to get open1x to run on my zaurus.. our entire wireless network is 802.1x...
01:34:46HClits annoying cause i really need wireless on my z :x
01:35:29kergothHCl: wpa-supplicant is in oe.
01:35:32kergothHCl: and works with the hostap drivers
01:35:43HClsweet :)
01:35:54HCli used to work on an open1x port, but i hadn't gotten it to work lately
01:35:58HCli'll wait for the march release :)
01:36:09prpplaguekergoth: hehe, thats what i have to work on next week
01:36:18prpplaguekergoth: gotta get wpa running on the m7100
01:36:43kergothheh, fun
01:36:58HClwhat kind of network do you have..?
01:37:15HCli need someone with linux that has wpa running on an ttls-pap network...
01:37:31kergothi should play with wpa at work, see if i can get the sl6000 on ti's wireless
01:37:56prpplagueHCl: initially we are gonna just support wpa-psk with tkip
01:38:15prpplagueHCl: after thats running we are suppose to add some more options
01:38:19HCl :x ok.
01:38:31HCli'm kind of hoping securew2 will be usable
01:38:36HClthey announced to go opensource a while ago.
01:39:30HClah, they went open source by now :)
01:39:39HCli mean, .com
01:50:12*HCl goes to sleep
01:51:32 Join muesli- [0] (
02:00:00preglowgood night ;)
02:00:07 Quit preglow ("woot")
02:03:53rasherrockbox should totally be all-unicode if possible
02:05:52ashridahneed a decent font caching system then. unicode fonts tend to be overly large
02:07:45kergothusing utf-8 doesnt require that you use a font that has glyphs for every codepoint in unicode.
02:22:05 Quit mecraw ()
02:23:56 Quit muesli- (Read error: 110 (Connection timed out))
02:40:07 Quit elinenbe (Read error: 104 (Connection reset by peer))
03:00:54***Saving seen data "./dancer.seen"
03:01:01 Join AC [0] (
03:06:22AChow's the state of audioplayback?
03:07:24XShocKsound is playing nicely
03:07:30ashridahsomeone had straight pcm playing before
03:07:41ashridahdoesn't mean it can play mp3s/ogg/flac yet tho
03:08:04ACa nice..
03:08:06XShocKyes, only pcm
03:08:21*AC looks at m68k asm to speedup libwavpack
03:08:58ACwhen will the changes be commited to cvs?
03:09:58XShocKi gave it to Linus, so they will be commited soon
03:14:46 Quit AC ("CGI:IRC (EOF)")
03:14:47 Join AC [0] (
03:20:47XShocKby the way no SRAM needed
03:20:53XShocKfor sound buffer.
03:26:06ACso will go to bed now... have a nice time
03:27:16 Quit AC ("CGI:IRC (EOF)")
03:32:23 Quit cYmen ("leaving")
03:37:28XShocKI want to add a chip to i2c bus in h120. as i figured out, i don't have to do anything besides connecting it to bus lines. am i right?
03:37:53XShocKselected fm-radio bus lines for this purpose.
03:39:38XShocKas i understand the lift up resistor is 10k. will it be ok if i connect the chip through resistor without changing the 10k lift-up resistor?
03:39:46XShocKplease, somebody.. :)
04:03:17 Quit QT (Read error: 54 (Connection reset by peer))
04:05:59 Join QT [0] (as@area51.users.madwifi)
05:00:56***Saving seen data "./dancer.seen"
05:57:45 Quit pill (Read error: 60 (Operation timed out))
06:06:07 Join pill [0] (
06:39:37 Quit Lynx_awy (Read error: 104 (Connection reset by peer))
06:40:04 Join Lynx_awy [0] (HydraIRC@
06:57:57 Quit XShocK (Read error: 110 (Connection timed out))
07:00:57***Saving seen data "./dancer.seen"
07:14:47 Join StrathAFK [0] (
07:23:17 Join midk [0] (
07:30:28 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
07:33:20 Quit Strath (Read error: 110 (Connection timed out))
07:52:59 Nick kergoth is now known as kergoth`zzz (
08:08:11 Join methangas [0] (
08:40:00 Nick StrathAFK is now known as Strath (
08:46:54 Nick kergoth`zzz is now known as kergoth (
08:57:35 Quit Diway (Read error: 104 (Connection reset by peer))
09:00:59***Saving seen data "./dancer.seen"
09:01:07 Quit midk ("Leaving")
09:18:33 Join jyp [0] (
09:27:22 Nick kergoth is now known as kergoth`zzz (
09:46:39 Join midk [0] (
09:51:19 Join ze__ [0] (
09:52:20 Quit ze (Nick collision from services.)
09:52:25 Nick ze__ is now known as ze (
10:06:51 Quit midk (Read error: 104 (Connection reset by peer))
10:18:39 Join pillo [0] (
10:27:54 Join Patr3ck [0] (
10:52:40 Join ze__ [0] (
11:01:01***Saving seen data "./dancer.seen"
11:05:55 Join Quelsaruk [0] (~kvirc@
11:06:45amiconnI have a Makefile problem I don't understand :(
11:07:13amiconnI added a codec.h file for all codec libraries in order to fix the Win32 build
11:07:42amiconnWhile including this works fine for libmad and libwavpack, I have a problem including it for liba52
11:08:13amiconn(did not yet look how to handle flac and tremor; they don't seem to have a central include file)
11:09:00amiconnWhen I add #include "codec.h" to liba52's config.h , then make, I get a strange message. "make: *** No rule to make target `typedef', needed by `/dep-liba52'. Stop."
11:09:20amiconnUnfortunately the depfile is not preserved, so I can't take a look at it :(
11:10:09 Join ze___ [0] (
11:11:46Quelsarukmaybe a pretty and young goat?
11:12:33ashridahamiconn: look at the SRC definition in the makefile in the directory in question
11:12:35 Quit ze (Read error: 110 (Connection timed out))
11:12:35 Nick ze___ is now known as ze (
11:12:43ashridahit should have a command you can rerun that'll dump to stdout
11:13:31ashridahthe first line will probably end up being 'typedef' for some odd reason you'll need to pick at
11:13:46ashridahthis is similar to the issue some people were having on macosX
11:14:35amiconnSRC is defined the same way as for the other codecs - preprocessed from the SOURCES file
11:14:56amiconnHmm, it seems that one is to blame :(
11:15:20 Join Chamois [0] (
11:15:25amiconnliba52 has a config.h, but this preprocessing is supposed to use rockbox' config.h....
11:16:41ashridahamiconn: yeah, but my point is that something appears to be adding something odd into the SRC definition.
11:17:47amiconnYes - the preprocessing done to get the actual sources uses config.h, which uses liba52's config.h because of the same name - and that one now includes my codec.h, which in turn includes sys/types.h, which contains some typedefs....
11:18:12amiconnImho the only solution is to rename liba52's config.h
11:20:49amiconnThat's it! Thanks!
11:21:59 Quit ze__ (Read error: 110 (Connection timed out))
11:37:15 Join preglow [0] (
11:42:49 Join ze__ [0] (
11:49:14 Join muesli- [0] (
12:07:04 Quit ze (Read error: 110 (Connection timed out))
12:07:07 Nick ze__ is now known as ze (
12:14:51 Quit Patr3ck ()
12:17:49*HCl stares at the unreadable gnuboy code of death.
12:18:52preglowi wonder if i should try optimizing imdct_l today
12:19:10preglowlibmad should be well above realtime then
12:20:46amiconnpreglow: I hope your optimisations are selected via a #define. Otherwise building libmad in the simulator will get tricky...
12:21:23preglowamiconn: yes, i suppose i'll do that eventually
12:21:30preglowthey aren't in cvs yet anyway
12:21:30jypalso porting to gmini is a concern
12:21:58jypat least for me :P
12:22:26preglowmy optimizations are huge asm blocks
12:22:28preglowso good luck
12:22:34muesli-hi btw
12:22:46ashridahhope they're usefully documented
12:22:58preglowi don't think theres any way of utilizing the parallel load any other way
12:22:59jypI wasn't thinking of porting your asm code;
12:23:13preglowwithout writing a gcc intrinsic, and looking at the gcc source code gives me headaches
12:23:15HClisn't A=A-A; the same as A=0 ?
12:23:30preglowHCl: yes
12:23:39jypbut ability to rely on old C code
12:23:41preglowHCl: that's the way to zero a address register fast
12:23:42HCljust making sure whether i'm not really stupid
12:23:43HClor so.
12:23:48preglowHCl: since clr doesn't do address regs
12:23:59HClbut sub does address regs?
12:24:15preglowi use that trick myself
12:24:20preglowin imdct_s opt
12:24:38preglowashridah: they're adequately documented :P
12:25:02ashridahholy sweet bejesus i'm lagged
12:25:20preglowIII_imdct_s in for a nice example
12:25:47preglowit's not pretty, but should be very fast
12:26:44HClyea, okay
12:26:52HCli was mostly asking cause the gameboy code is doing it
12:27:00HCland gnuboy is actually. well. computing it.
12:27:05HClwhen it really doesn't have to.
12:27:41 Quit Chamois ("CGI:IRC")
12:27:44 Join ze__ [0] (
12:28:23amiconnHmm. With my malloc -> codec_malloc etc changes, all of mad, tremor, a52 and vorbis work on the cygwin-built simulator
12:28:47 Quit preglow ("Changing server")
12:28:47 Join preglow [0] (
12:29:05amiconn(apart from the non-working file seek, so I had to move the correct .wav header to the front)
12:29:13amiconnFor some reason I couldn't make .wv work from viewers.config
12:29:28amiconnThe native win32 simulator still has problems...
12:30:16amiconnAH, I
12:30:26ashridahis it just me, or does the coldfire programming reference manual not state the cycle count of instructions?
12:30:29amiconn'm silly. Wrote .vw instead of .wv
12:33:23preglowashridah: it does
12:33:44amiconnHmm. Could someone please try wavpack in the simulator? I get a "crc errors detected" message...
12:33:45preglowthat is, not that one
12:33:57preglowyou need the mcf5249 manual
12:34:09amiconn(With t1.wv provided by AC that is)
12:34:25preglowi mean the coldfire2um.pdf one
12:34:29preglowchapter nine
12:34:59ashridahi see
12:35:37 Join Chamois [0] (
12:36:03 Quit Chamois (Client Quit)
12:36:55preglowdamn, the mac instructions are blessedly fast
12:37:21ashridahprobably a good thing
12:37:46preglow32x32 bit multiply, accumulate AND parallel move in 4 cycles isn't too shabby
12:38:13ashridahwell. how fast is that in non-mac asm?
12:38:34preglowextremely slow, you'd need to muls, two adds, probably a couple of moves and then another move
12:38:40preglowtwo muls
12:38:50preglowa mul alone is three cycles
12:41:33*ashridah wanders off to record another 2 hours of music off satellite tv. >:)
12:44:22*HCl sighs.
12:44:31HCli can't stand undocumented code
12:44:40HCl ^ is xor in C, right?
12:45:01amiconnHCl: I read in yday's log that dynarec is working now?
12:45:11amiconnYes, ^ == xor
12:46:19 Quit ze (Read error: 110 (Connection timed out))
12:46:19 Nick ze__ is now known as ze (
12:51:22HClamiconn: it is, i just need to encode the z80 opcodes properly
12:51:35HCland some gnuboy code is absolutely horrid.
12:53:07amiconn(z80 opcodes) Ah, yes. (gnuboy code) What code are you talking about, and why is it horrid?
12:53:30HCllet me paste.
12:53:38HCl#define CP(n) { \
12:53:39HClW(acc) = (un16)A - (un16)(n); \
12:53:39HClF = FN | FZ
12:53:39HCl| (ZFLAG(LB(acc))) \
12:53:39HCl| (FH & ( (A ^ ( n ) ^ LB(acc) ) << 1)) \
12:53:41HCl| ((un8)(-(n8)HB(acc)) << 4); }
12:53:52ashridaho.O even
12:53:58HClsee what i mean :X
12:54:03amiconnAh, the flag stuff.
12:54:10ashridahi might suggest i've seen worse. barely.
12:54:27HClyea, i'm reading about flags on my datasheet thing.
12:54:37ashridahsome weird recursive expanding macros in the linux kernel sourcecode a while back that were mindnumbing to consider
12:55:03HClit seems that the BCD flags are only used if the next instruction is a DAA instruction.
12:55:16HClso i can get away with just setting the zero and carry flag
12:55:23HCland detecting whether the next instruction is a DAA
12:55:27HCli think.
12:55:43amiconnYes, the bcd flags are only used by DAA, but it may happen that DAA is not the next instructions
12:55:52amiconn*next instruction
12:56:18amiconnThere may be instructions in between that don't touch the flags
12:56:43*HCl scratches his head.
12:56:57HCl :(
12:57:45HCli don't suppose you could explain how i should set FH
12:58:07HClwell meh, nevermind, i'll come across this later.
12:58:14HClfor now, i know what i need to set flags to..
12:58:17HClthanks :X
12:58:51amiconnThere is a table that defines the BCD flag states depending on the register value
12:59:07amiconnI think using that table is the fastest solution
12:59:56HClthere is? where?
13:01:04***Saving seen data "./dancer.seen"
13:02:29amiconncpucore.h: daa_table
13:02:31 Join muesli_ [0] (
13:02:48HClodd, i wonder why its not being used in the interpreter
13:02:54amiconn..and daa_carry_table
13:03:28amiconnI'm not sure how these are to be used correctly, but it seems the i386 asm implementation uses them
13:04:01amiconnAnother think I wonder is how dynarec is supposed to handle conditional jumps
13:04:57 Quit muesli- (Read error: 110 (Connection timed out))
13:05:37HClbut i'm not there yet.
13:09:07 Join markun [0] (
13:20:34 Join muesli- [0] (
13:22:52 Join ze__ [0] (
13:29:06 Quit muesli_ (Read error: 60 (Operation timed out))
13:37:50 Join AC [0] (
13:38:09markunAC: hi
13:38:49markunI'm having problems with my cross-compiler on FreeBSD :( It works fine on a remote linux account.
13:39:02AChas somebody a good starting point for m68k asm - i want to optimize libwavpack
13:40:43preglowAC: do you know any other assembly languages?
13:41:10 Quit ze (Read error: 110 (Connection timed out))
13:41:12 Nick ze__ is now known as ze (
13:42:14ACpreglow: i have done somemmx and sse asm for x86
13:44:12preglowi know x86 and 6502 asm, i picked up 68k asm pretty quickly from the reference manuals alone
13:44:41preglowi can't see any obvious opt canidates in libwavpack
13:44:45preglowhave you profiled it?
13:45:36ACnot yet, but i want to do it today
13:45:50 Quit muesli- (Read error: 104 (Connection reset by peer))
13:45:52preglowyes, that would be wise
13:45:55 Quit rasher (Read error: 104 (Connection reset by peer))
13:45:57preglowreplace as little as possible with asm
13:46:19preglowif you find any hotspots, let me know, i might be able to give you some pointers
13:49:15amiconnHCl: I just had an idea for further dynarec optimisation (of course after your first version is working):
13:49:49 Join Renko [0] (
13:50:07amiconnInstead of loading & saving the whole z80 context before/after every block, the dynamic compiler could track which z80 registers are actually used during compilation of the block.
13:50:43amiconnThen it would add the appropriate context load/store at the beginning/end of the block
13:51:10HClthe problem is that then you can't do the optimization of not having to recompile blocks while in the middle of a block
13:51:21HClwhen jumping into the middle of a block
13:51:32amiconnAh, yes.
13:51:36HClif you keep all registers consistant with the gameboy regs, thats possible to do
13:52:09amiconnBut if you want to do this, you have to find the address within the block that corresponds with the z80 address
13:52:11HCleventually, i'm gonna have to want to swap context only when needing to recompile
13:52:26HClthats not too hard to do. the dynarec can do that along the way
13:53:02 Join webguest32 [0] (
13:53:09HClbut yea, it'd require an array of 0x7fff pointers
13:53:23amiconnHmm. You would need an array that stores which z80 address corresponds to which block address. I guess you'll need that anyway though
13:53:54amiconnThat array may get very large - or do you have another method how to deal with rom bank switching?
13:54:26HClfor now, i'm just focusing on getting it to work
13:54:32preglowbut linux
13:54:33 Quit preglow ("wo")
13:55:59 Quit webguest32 (Client Quit)
13:56:47amiconnIiuc the problem is that you must consider rom bank switching even without allowing jumps into a block. You need to keep track for each compiled block to with z80 address *in which bank* it corresponds.
13:56:52 Join preglow [0] (
13:56:57amiconnHow do you intend to handle this?
13:58:17HClnot, at the moment o.o
13:58:33HCli haven't really looked at rom bank switching yet
13:58:55amiconnHmm, okay.
13:59:01amiconn(away now)
13:59:03*AC gets hand dirty with libmusepack
13:59:32HClbut you're right.
13:59:54 Nick Strath is now known as StrathAFK (
14:00:46HCli need to keep account of different rom banks from addresses 4000-7fff
14:01:34ashridahhurray for 16 bit addressing
14:02:01pillocan somebody explain me why memset() expects an "int c" as the filler value, even though it uses it as a byte? why not declare it "unsigned char"?
14:02:52preglowit's not important
14:03:00preglowall memsets do that
14:03:09ashridahpillo: it's 'standard'[tm]
14:03:28preglowon most machines, an int has a lot less latency involved than a byte
14:03:41preglowso you should only use bytes if space is an issue
14:04:08pillopreglow: yep, but that was confusing me- i thought it was used as 4 bytes (too used to x86;)
14:05:04HClamiconn: yea, you're right, i need to get PC through the address/rombank translation first before dynareccing
14:05:22preglowi certainly hope dynarec ends up being a lot faster, heh
14:05:52 Quit pill (Read error: 54 (Connection reset by peer))
14:09:04HClto be honest, i have no clue
14:10:10preglowcoldfire might be a special case
14:10:19preglowthe ram is shit slow, and you need to use it _a lot_
14:12:30 Join cYmen [0] (
14:14:05 Quit DMJC ("Leaving")
14:14:44 Quit AC ("CGI:IRC (EOF)")
14:16:55 Join ze__ [0] (
14:17:25 Join elinenbe [0] (
14:18:05HCli think that moving the selected bank into that fast ram would help, if it wasn't for the problem of how to switch them back and forth
14:27:16HClits appalling how badly readable the lcd of the iriver is without backlight
14:27:37preglowit isn't very good however you look at it
14:27:49preglowit handles motion poorly
14:30:55 Join DMJC [0] (
14:32:46 Join R3nTiL_ [0] (~zorroz@
14:33:04CoCoLUSturn on the backlight? :)
14:34:37*HCl 's linked list implementation is failing for some unknown reason
14:35:52 Join nobby [0] (
14:36:15 Quit ze (Read error: 110 (Connection timed out))
14:36:18 Nick ze__ is now known as ze (
14:36:32ashridahdoesn't rockbox already have utility data structures like that?
14:37:12ashridahlinked lists.
14:37:28HClyou mean a struct ll { void *data; struct ll *next; } thing
14:38:04*HCl knows his flaw.
14:38:56nobbyany progress on the audio driver since the sine wave tests?
14:38:58HClbug in my malloc thing.
14:40:16preglownobby: we've got sound, that's thta
14:40:19HClnobby: as far as i know it now plays wav
14:40:41nobbyand the mp3 and ogg codecs?
14:40:45nobbyhow are they?
14:41:20HCltoo slow.
14:41:29HClhmm. how fast is the flash ram compared to the normal ram?
14:41:40preglowHCl: it's faster
14:41:43HClcould it be our codecs are simply too slow cause their code is in ram rather than flash?
14:41:48HCl :/
14:41:49preglowHCl: and it's flash rom, not ram
14:41:55HClyea. well.
14:42:02 Join rasher [0] (
14:42:04HCli don't actually call flash read only memory
14:42:10HClsince its far from read only
14:42:18ashridahnobby: the codecs need many weeks of optimisation before they're >100% realtime. then the buffering and streaming needs to be designed adequately. it's not going to happen overnight.
14:42:19preglowwell, but calling it ram is misleading
14:42:30HClits not random access either.
14:42:35preglowbut yes
14:42:35HClat least
14:42:36 Join webguest23 [0] (
14:42:39ashridahparticularly when the people doing development have lives and jobs
14:42:40nobbyok, fair enough
14:42:42preglowsome mp3 and flac are realtime now, i guess
14:42:44preglowbut just barely
14:42:46*nobby goes back to idling
14:43:00nobbythanks for the heads uop
14:43:07ashridahnobby: best way to keep track is to peruse the irc logs and the cvs logs (if you're technically inclined) and watch the wiki
14:43:18ashridahbut yeah, lots of work to do :)
14:43:36webguest23hello, I got a error during install CVS binutils -> "msgfmt: not found" - could someone help me ?
14:43:55nobbyi keep an eye on the main page of the site for cvs updates, but they dont exactly say "ok, [thing] works but its crap because of [non-techy explaination]"
14:44:19nobbyi'll post this in the iriver subforum to keep the others updated
14:44:35ashridahi'm not exactly authoritative.
14:44:41ashridahthere's still lots of technical issues to work out
14:45:35ashridahlots of things that rockbox has never had to deal with before
14:46:38nobbyi understand
14:46:52nobbywhat do the chips that currently lack drivers do?
14:47:02nobbyI2C driver 50% Writing to I2C done
14:47:02nobbyI2S driver 10% Basic I2S output, but not interrupt-based yet
14:47:02DBUGEnqueued KICK nobby
14:47:02nobbySPI driver 0%
14:47:02nobbyRemote LCD driver (SPI) 0%
14:47:03***Alert Mode level 1
14:47:03nobbyAudio ADC/DAC driver 10% UDA1380 Initialization and basic control done (volume control, muting etc)
14:48:51HClthats out of date.
14:48:51 Quit webguest23 ("CGI:IRC (EOF)")
14:48:54HClcheck the wiki.
14:49:04ashridahnobby: i2c and i2s are command and streaming busses (as i understand them). you use them for talking to other devices and streaming data (such as audio). i'm guessing SPI is serial peripheral interface (iirc, the remote is connected via this).
14:49:06ashridahafaik, audio is 100%
14:49:13HClno, not 100% at all.
14:49:15HClno spdif
14:49:16HClno recording
14:49:17rasherno recording
14:49:24ashridahhm. true
14:49:35nobbybut headphone out is 100%? :)
14:49:41HClpretty much, yes
14:49:46HClas far as i know
14:49:57nobbygood to hear
14:50:24nobbyand is the amount of optimization needed definitly possible?
14:50:42ashridahnobby: well, the iriver does it. means it can be done
14:50:56ashridahanother device with the same cpu can do flac too, so that's clearly possible
14:51:18nobbywhat about .SID? :D
14:51:33ashridahsomeone's workign on that (or was it midi? i forget)
14:51:38ashridahbut there are other challenges there
14:51:45HCllike the gameboy emu :P
14:52:06nobbyi would if i could figure out how to compile the bootloader
14:52:08preglowsid is going to be tricky
14:52:20preglowyou need a very high precision 6510 emulator for it to work properly
14:52:26preglowthe resid core can be used as it is
14:52:43nobbyanyone here want to put a flash together and send me it? i'll host it for anyonw wanting it after that (i know youre all scared of legal stuff)
14:52:55HClsure. hold on.
14:53:13nobbyi want to show off playing rockboy
14:53:42 Join ze__ [0] (
14:55:12preglowyou can't play it yet :PP
14:55:18nobbywhy not?
14:55:23preglowit's incredibly slow
14:55:39nobbyyeah, but i can still show it off in slow motion, cant i?
14:55:59preglowit's just very, very slow motion
14:56:09HClits on your own risk, no clue whether it works.
14:56:36nobbyi assume thats a 120/140 flash (120 and 140 use the same flash)
14:56:41HClcurrent bootloader built from cvs against iriver eu 1.63
14:57:04***Alert Mode OFF
14:57:09nobbyi take it i need the .rockbox folder etc as well?
14:57:22nobbyhow do i get all that stuff from the cvs under windows?
14:57:23rasheror it'll just boot iRiver
14:57:29 Join Sando [0] (
14:57:32preglowHCl: you really should use the wiki bootloader
14:57:37rasherjust grab a snapshot
14:57:49preglowi don't think anyone has tested the cvs bootloader in a while
14:57:57rasherI did yesterday :)
14:58:24nobbyyou mean this?
14:58:30rashersomething like that, yes
14:58:43nobbysomething like? youre confusing me....
14:59:13rasherSorry, yes, that's right
14:59:20nobbyok, thanks
15:00:04ashridahrasher: that's actually a little dangerous.
15:00:06rasherHCl: you have the same md5sum as me, fwiw
15:00:17rasherashridah: what is?
15:00:17nobbyif this kills my iriver, do you recon linus would rescue it it with his bdm thing, if i sent it to him (and payed to ship it back)
15:00:19ashridahrasher: the rockbox code for the loaded firmware and the bootloader share a lot of code.
15:00:27nobbywhats dangerous? O_O
15:00:30rasherHCl: for what it's worth
15:00:38ashridahusing current cvs for the bootloader.
15:00:53 Quit edx (Read error: 60 (Operation timed out))
15:00:55rasherI don't follow
15:01:01ashridahnobby: dangerous is using untested code
15:01:05***Saving seen data "./dancer.seen"
15:01:16HClpff, you people have no feel for adventure :P
15:01:16nobbythen could someone zip up a known working .rockbox directory and send me it?
15:01:31HClnobby: daily build should be fine
15:01:39HCl :P
15:01:39ashridahnobby: the .rockbox directory isn't the problem
15:01:43nobbyok, daily build it is...
15:01:50HClfine. i'll get the bootloader off the site then.
15:01:52ashridahyou can replace that using the original firmware if you have to
15:01:55rasherashridah: Why would using current cvs for the bootloader be particularly dangerous?
15:02:16ashridahrasher: okay. we've been posting changes to the code, which aren't being verified to work with the bootloader.
15:02:37ashridahthe firmware and the bootloader share quite a bit of the basic rockbox library
15:02:47rasheroh, like that
15:03:13rasherGuess I got lucky
15:03:18ashridahyeah. so using cvs bootloader when it's using code that hasn't been verfied to work with the bootloader could lead to a brick.
15:03:35ashridahif the firmware works, it's not likely, but it's also not certain
15:03:44CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:03:44*HCl yawns
15:03:46nobbyHCl, have you tried the build you made for me?
15:03:52ashridahit's probably safer to let linus sanction a release of bootloader.bin occasionally.
15:04:18HClnobby: nop, let me build one of the bootloader on the wiki
15:04:23nobbyok, thanks
15:04:36pillook guys, I have prepared a patch that adds proportional text support to the viewer plugin. Does somebody want to test it? Or should I submit it to the sf tracker?
15:04:48preglowHCl: is the bootloader in wiki capable of loading rockboy?
15:05:06preglowpillo: try buggin amiconn
15:05:21ashridahpreglow: would that matter?
15:05:37preglowashridah: well, yes, if he wants to try rockboy :)
15:06:48nobbygtg, /msg me with whatever you decide and i'll be back in a few
15:07:09ashridahpreglow: well, i was under the assumption that the firmware would be able to do whatever it likes, regardless of what state the bootloader is in
15:07:22ashridahsince the firmware would probably make no assumptions about the state of anything
15:07:32HClpreglow: yes
15:07:42HClbootloader has nothing to do with rockboy
15:08:08preglowashridah: and you would be correct in assuming that, disregard me
15:08:13HClbut current rockbox won't load rockboy since rockboy requires mkdir to be added to the plugin api
15:13:03 Quit ze (Connection timed out)
15:13:03 Nick ze__ is now known as ze (
15:19:12pilloI submitted the text viewer patch to the SourceForge tracker.
15:19:29pilloLet me know if you like it, gotta get back to business...
15:22:33*HCl flashes the new bootloader and gets reminded of how much he hates to flash things
15:22:49*nobby is back
15:22:53nobbyis it working?
15:23:12HClits a heck of a lot faster too
15:23:21preglowi'll go and catch some sunlight
15:23:22nobbyso i should use the first one you uploaded?
15:23:30HClno, redownload
15:23:36HClthis one is the confirmed to be working one
15:31:58 Quit elinenbe (" HydraIRC -> <- Go on, try it!")
15:32:06nobbyflashing now....
15:32:19nobbyjust checked the hd5, and im going to disconnect and flash
15:32:31 Part amiconn
15:33:55nobbyflash in progress...
15:34:15*ashridah waits for the unearthly scream
15:34:34nobbyi'll have to run crying to linus if it dies
15:34:35 Join ze__ [0] (
15:34:42nobbypowering back on after flash...
15:35:20nobbysomeone;s been fiddling with the images :P
15:35:29nobbyin the regular firmware
15:35:45nobbythat works anwaya
15:35:52nobbynow to try to get rockbox to load
15:36:00nobbyi hold record to do that, yes?
15:36:22nobbyrockbox doesnt want to load...
15:36:35nobbykeeps loading official fiormware
15:36:47ashridahrockbox loads by default
15:36:52ashridahhold record to load the original firmware
15:37:03nobbyit wont load rockbox
15:37:13nobbymy first attempt at booking i held nothing
15:37:46ashridahdid you uncompress the nightly build into the root, creating a .rockbox directory?
15:38:22 Join Sucka [0] (
15:38:29nobbyit said something like "loading rockbox" and on the next line something like "result:-t"
15:38:38nobbybut for like .5 of a second
15:38:47nobbythen it loads original
15:39:04 Nick R3nTiL_ is now known as R3nTiL (~zorroz@
15:39:07nobbyah, its "result:-1"
15:39:28Renkosounds like it cant find rockbox.iriver (?)
15:39:37nobbyok, i'm a dumbass
15:39:39ashridahodd. not an error message i've seen. and i'm not in a position to poke through the bootloading code to see where it coems from
15:39:47nobbyi thought that was what was combined to make the flash
15:39:55nobbyi dont have rockbox.iriver on it >_<
15:40:13nobbyno problem
15:40:15*nobby adds
15:40:15Renkoshould be in the zip that you unzipped in the root
15:40:18nobbyit is
15:40:22ashridahdidn't i just tell you you need to unpack a nightly into it?
15:40:22Renkoahh ;)
15:40:33nobbylook, im a dumbass, ok? :P
15:40:44 Nick nobby is now known as nobby_isadumbass (
15:40:44DBUGEnqueued KICK nobby_isadumbass
15:41:04ashridahi think it needs to be in capitals.
15:41:14 Nick nobby_isadumbass is now known as nobby_ISADUMBASS (
15:41:14DBUGEnqueued KICK nobby_ISADUMBASS
15:41:55ashridahso, it's booting now?
15:42:11nobby_ISADUMBASSim trying now
15:42:22nobby_ISADUMBASSmy pc takes ages to safely remove hardware
15:42:42nobby_ISADUMBASSit booted :D
15:42:57ashridahin the rare instances where i use windows, i just left click on the usb tray icon, select the drive to remove, and it's done
15:43:41nobby_ISADUMBASSmine takes a few seconds to work
15:43:46nobby_ISADUMBASSit need reformatted
15:44:20nobby_ISADUMBASSits working great
15:44:51ashridahyou have an interesting definition of 'work'
15:44:52nobby_ISADUMBASSi'll post a mini-tutorial for other dipshit windows users on the fourm (with plenty of warnings)
15:45:06ashridahnobby_ISADUMBASS: there's really not that much point.
15:45:08nobby_ISADUMBASSok, the bouncing text plugin works :P
15:45:09ashridahnot yet anyway
15:45:50nobby_ISADUMBASSyeah, but they have been asking, might as well let them have their fun
15:46:05 Join webguest62 [0] (
15:46:16ashridahno, their fun's likely to just create the first brick
15:46:26 Part webguest62
15:47:00nobby_ISADUMBASSok, ok
15:47:08nobby_ISADUMBASSno bricks
15:47:26CtcpIgnored 2 channel CTCP requests in 30 seconds at the last flood
15:47:26*nobby_ISADUMBASS sits and stares at bouncing rockbox logo
15:47:46 Nick nobby_ISADUMBASS is now known as nobby_ISEASILYAM (
15:47:46DBUGEnqueued KICK nobby_ISEASILYAM
15:47:46***Alert Mode level 1
15:47:54nobby_ISEASILYAMcharacter limit
15:47:56 Nick nobby_ISEASILYAM is now known as nobby (
15:47:56DBUGEnqueued KICK nobby
15:47:56***Alert Mode level 2
15:47:56***Alert Mode *RED* level 102
15:48:01nobbyback to the idling
15:48:20ashridahplay stars and rocktris
15:48:52nobbywhat does battery test do?
15:49:07nobbyand how do i make it stop!
15:49:30ashridahi don't remember, and there's a reset hole on the bottom of the device :)
15:49:43nobbyits accessing the hard drive though
15:49:51nobbyi think its trying to waste battery and time it
15:49:54ashridahit's a battery test
15:50:00ashridahit tests the battery
15:50:07HCli'm stupid.
15:50:14ashridahnot all of the plugins are finished yet
15:50:16nobbyi dont want to test the battery any more!
15:50:21nobbyah, it stopped
15:50:39 Quit R3nTiL (Excess Flood)
15:51:27nobbysokoban works, but is tiny
15:51:44ashridahyeah. not all of them have been modified to scale
15:52:08nobbyits impressive how much does work, though
15:52:12HClyea, its on the todo list
15:54:13nobbydo any games support saving progress? sokkoban doesnt seem to
15:55:57 Quit ze (Read error: 110 (Connection timed out))
15:55:58 Nick ze__ is now known as ze (
15:56:15ashridahthey could in theory
15:57:57***Alert Mode OFF
16:00:07nobbyis it jsut me or is cube distorted?
16:00:26preglowit's not
16:00:34preglowit's perspective corrected, if that's what you mean
16:00:42ashridahit's not like the lines are anti-aliased :)
16:00:56preglowwith the slow display on the h1x0, it looks like it is :)
16:01:09nobbyyeah, looking at it from a distance its cool
16:01:29nobbyyou just need to work out which side is which
16:01:35ashridahalthough when we move the display to 2bit greyscale, we'll be able to antialias it :)
16:02:12CtcpIgnored 2 channel CTCP requests in 10 seconds at the last flood
16:02:12*preglow ponders on porting doom
16:02:27nobbyit better by better than Mdoom on my phone :P
16:02:46preglowahh, think of the possibilites with the colour screen of h3x0
16:02:58preglowif it can play xvids, then hell, it can do doom
16:03:05nobbyport milkdrop and avs from winamp :P
16:03:10preglowthat would rock
16:03:14nobbyrock the box
16:03:16preglowi miss milkdrop
16:03:17preglowisntant acid trip
16:03:23nobbymiss it?
16:03:29nobbytheres a linux port
16:03:30preglowi don't use winamp any more
16:03:35nobbywell, reimplimentaion
16:03:46ashridahplease. someone ported a version of doom1 to the TI 90 series of graphing calculator
16:03:52ashridahif that can play doom, anything can
16:04:06nobbyI need a TI90 to play in maths...
16:04:26preglowwhat ti90 ?
16:04:29*preglow strokes his ti89
16:05:14nobbydoes the iriver hardware allow a real EQ or is it bass and treble ajustment for life?
16:05:57ashridahnobby: assuming one can cram it into what will probably be 96MHz max, sure.
16:06:26nobby"debug (keep out)"
16:06:45nobbyi assume i should not under any circumstances click the button on that
16:07:10rasherit's just useless (to most) information
16:07:13ashridahyou can poke, but it's not really going to give you useful info
16:07:14preglownobby: i'll fix up an eq once the audio is up and going
16:07:27preglownobby: feel free to suggest features for this eq, i don't use them myself
16:07:38nobby10 band or so
16:07:42pregloware you mad?
16:07:48preglowwell, ok
16:07:54nobbywhy not?
16:08:02nobbywhats parametric?
16:08:14preglowyou decide frequency and bandwidth of each band
16:08:21preglowand gain, of course
16:08:24nobbyooh, sounds good
16:08:50preglowwell, i'm just wondering, seeing as how i don't use eqs when i listen to music myself, i don't know what people prefer
16:08:57nobbyis anyone currently working on supporting a grapihical WPS?
16:09:05preglowwhat _is_ a graphical wps?
16:09:34nobbyone with a bitmapped progressbar and some nifty peakmeters :D
16:10:10ashridahnobby: again, someone probably will, if you can cram it into 96MHz
16:10:48nobbycould we try to overclock it? >:D
16:11:00preglowit can't even run at full clock
16:11:05nobbyhow come?
16:11:12preglowit gets too hot during hd accesses
16:11:30preglowat least, that's our current theory
16:11:39preglowlinus is working on it now
16:11:58preglowwe could replace the disk with a heatsink
16:12:18preglowit'd be killer for rockboy
16:15:38ashridaharghl. 2am. sleep
16:15:41 Quit ashridah ("sleep")
16:17:25 Join ze__ [0] (
16:23:05 Quit jyp ("poof!")
16:28:16 Nick kergoth`zzz is now known as kergoth (
16:35:44*rasher . o ( with someone wanting to port EGL to rockbox, projectm might just be possible on h300 )
16:36:27nobbyi am a genius
16:36:31 Quit ze (Read error: 110 (Connection timed out))
16:36:31 Nick ze__ is now known as ze (
16:39:04 Join _aLF [0] (
16:49:03preglowanyone here sitting on a heap of spare h120 parts? :P
16:49:08HClwhy? :P
16:49:41nobbyi have a working 140 sitting on me right now :P
16:49:52nobbymy remote is dead
16:49:53nobbywant it?
16:49:57preglowthe plastic parts on the side of it, around the screw holes
16:50:04preglowthey both show signs of stress
16:50:05nobbyit works, but the record and volume buttons are fucked
16:50:11preglowhell, one of them's flat out broken
16:50:39preglowso i must've done something wrong while screwing it together once
16:51:05preglowpoint is, i don't like having them broken :P
16:52:15 Join edx [0] (
16:52:57nobbyanyone know where to get chip8 games?
16:55:49nobbyfound a few just now
16:57:05nobbyand more...
16:57:09nobbylegal ones
16:57:12nobbytheyre tiny
16:57:23nobbyshould i add links to the wiki?
16:57:47*nobby is confused
17:01:08***Saving seen data "./dancer.seen"
17:02:18 Join ze__ [0] (
17:03:31preglowdoes that site still exist?
17:03:36nobbyzophar? sure
17:04:00nobbyyeah, i remember getting snes roms off it when i was about 10
17:04:18preglowi stopped visiting it after it went down the drain, and the new site admin turned out to be a gun nut
17:04:20*HCl sighs.
17:04:33nobbyi never visited it regularly anyway
17:05:20preglowi did visit it regularly, until the day i was greeted with a bloody 9/11 lament, a "WE WILL NEVER FORGET" logo and instructions on how to get firearms
17:05:45nobbyi dont know wheter to lol or be sad
17:05:51preglowi chose the latter
17:06:23preglowthen i saw that the forum avatar of the site admin was him pointing a gun at the screen, i just stopped visiting it
17:06:42nobbyits ironic that "we will never forget" is the attitude that the hijackers on 9/11 towards america
17:07:21nobbyits true, the middle east still hates the west because of the crusades!
17:07:26preglowi just think it's funny that usa refuses to forget, when it has no problem forgetting shit it itself has done to other parts of the world for decades
17:07:30preglowbut enough of that
17:07:40nobbysave it for #politics
17:08:04preglowi've got to go code other stuff :/
17:11:43 Quit Sando (Read error: 104 (Connection reset by peer))
17:21:27 Quit ze (Read error: 110 (Connection timed out))
17:21:29 Nick ze__ is now known as ze (
17:28:55 Join ze__ [0] (
17:36:37 Nick nobby is now known as n[o]bby (
17:36:37DBUGEnqueued KICK n[o]bby
17:48:10 Quit ze (Read error: 110 (Connection timed out))
17:48:10 Nick ze__ is now known as ze (
17:48:28 Join R3nTiL [0] (~zorroz@
17:54:22CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:54:22*HCl smacks computers with a hammer.
17:54:31 Join Phoenix983 [0] (
17:55:40HClno caps please.
17:55:49HClisn't the default password empty?
17:56:15Phoenix983is the master user password
17:56:30HClgoogle is your friend
17:57:09QuelsarukPhoenix983: try with all empty
17:57:17Phoenix983i tried that, but i not find nading, give some idea for the good words for search
17:57:31Quelsaruki tried all 0's with an IBM drive
17:57:36Quelsarukand worked
17:58:10HCl"western digital" default harddisk password lock
17:58:13HCltry that in google.
17:58:18Phoenix983me too, but nading... :( this program of atapwd is great
17:58:49 Quit R3nTiL ()
17:58:55*HCl knows too many people from the xbox scene who locked their hdd with an unknown key and ruined their hdd
17:58:58Quelsaruktried all spaces?
17:59:36Phoenix983exist any program for reveal the code??
17:59:47Phoenix983but let me try again
18:00:44Quelsarukthere's a special tool for unlocking the drive... using linux
18:01:56HClactually. there isn't.
18:02:06HClthe key is stored in a piece of ram only accessable by the firmware
18:02:14HClso even a firmware upgrade won't help
18:02:41HClharddisk keys are designed to be unrecoverable thus secure
18:03:30Phoenix983i tried the 0's, the spaces, and the empty password
18:03:43HClhow did you lock it in the first place?
18:03:49Phoenix983any other idea?? :'(
18:04:17HCloh, putting an password on a disk without knowing the password also voids its warranty, as far as i know.
18:04:48Phoenix983it is locked power on and power off the system in repeat ocations, is how a electric down.
18:05:05HClno, harddisk locks are designed to be secure, you need a general hdd unlocking tool that unlocks it with the correct key
18:06:16Phoenix983ok i have the atapwd, this is not a hdd unlocking tool???
18:06:28HClwell, the name sounds right
18:07:14Phoenix983this is the program in the web page of rockbox for unlock hard disk
18:07:23 Quit Phoenix983 ("CGI:IRC")
18:07:30HClthen its probably right. i haven't used it myself... bye o.o;
18:07:46 Join Phoenix983 [0] (
18:07:53Phoenix983sorry lol
18:07:58Phoenix983what did you said?
18:07:59HClits ok.
18:08:07HClthat its probably right and i haven't used it myself
18:08:28CoCoLUSyou'll still need the pw to use it
18:08:35Phoenix983mmm what hdd unlocking tool use???
18:08:51CoCoLUSyou don't even need to search for one if you don't have the pw
18:09:18Phoenix983what is the pw??
18:09:24CoCoLUSthe one you set :)
18:09:39Phoenix983aaah password
18:09:57CoCoLUSdid you set it manually?
18:10:05Phoenix983lol i never set any password
18:10:17Phoenix983this hdd is locked auto
18:10:25HClby what?
18:10:27CoCoLUSwell you know who/what locket it?
18:10:39Phoenix983for a repeately power on and power down process
18:10:52HClby what?
18:11:59 Quit Nibbler ("blubber")
18:12:42*HCl goes to eat dinner.
18:13:48Phoenix983for repeat shut down and power on proces
18:14:50CoCoLUStheres no hard drive on this earth that locks itself just because you power it up and down rapidly
18:16:32Phoenix983man, this is the first case
18:16:47CoCoLUSsorry but i somehow just doubt it :)
18:17:00Phoenix983only next to this process, that is locked
18:17:10CoCoLUStell me which device/pc/abacus did that, and i'll believe you
18:17:15 Join Nibbler [0] (
18:17:25Phoenix983i dont need lie man
18:17:55Phoenix983it is how a range of electric down current
18:18:17CoCoLUSok so i believe you, but then, i can't help you, because i don't know what happened to your drive
18:18:40CoCoLUSits a moot point anyway, if you don't know the pw, you're screwed
18:18:52Phoenix983exist million of western digital hard disk locked auto in the world, specific my disk WD800BB
18:19:11 Quit courtc (Read error: 113 (No route to host))
18:19:20Phoenix983this model
18:20:41Phoenix983i know that if i not know the pw, i need to go to put the default factory password, but i dont know where find them for my model
18:22:10Phoenix983or for the mark western digital
18:22:24 Join XShocK [0] (
18:22:32CoCoLUSwell if it is such a ultra high-security auto-locking model like you claim don't you think the master pw would be set as well?
18:27:07HClwhat exactly do you mean with locked auto?
18:27:11*HCl returns
18:27:29Phoenix983the same of the words
18:27:52HCli'm unfamiliar with the concept of autolocking.
18:27:59HClthere are no disks that do that.
18:28:17HCli only know one system that runs off a locked drive and unlocks it during boot
18:28:31CoCoLUSthe xbox.
18:29:24Phoenix983this mark this signs−−−−−−> S= E= L= F=- X= F=H
18:29:30HClso even if you need a secure system, there's absolutely no reason for a computer system to keep locking their drive on boot.
18:29:52HClunlocks are temporary unless the computer system states otherwise
18:30:23Phoenix983no the auto lock was for the shut-down and power on
18:30:36HClharddisks don't lock on computer shut down.
18:30:53HClas soon as a locked harddisk loses power, when it boots up again, it needs to unlock.
18:30:55Phoenix983you dont satnd me
18:31:11HClunless the computer states that it needs to unlock permanently.
18:31:37 Quit Zagor (Read error: 104 (Connection reset by peer))
18:31:54 Join Zagor [242] (
18:32:00Phoenix983my hard disk is not lock for power on the computer
18:32:15HCleither way. i can't help you if you can't tell me what locked it.
18:32:17Phoenix983it was for ON an OFF operation
18:33:06Phoenix983this WD hard disk WD800BB first version have this problem of auto lock
18:33:32HClyou mean its a bug in the firmware of the disk?
18:33:49CoCoLUSyou have any info on this?
18:33:51Phoenix983i know of a great quanty of reports for auto lockñ of this modelsç
18:33:57CoCoLUScause google finds absolutely nothing on that model
18:34:25Phoenix983mmm let me show you the marks of the atapwd
18:35:05Phoenix983S= sume E=sume L=sume F=- X=sume F=H
18:35:41Phoenix983S=( ) E=( ) L=( ) F=- X=( ) F=H
18:35:59HCli don't know atapwd. that literally says nothing to me.
18:36:01Phoenix983damn this chat dont show the adition sign
18:36:22CoCoLUS+ ?
18:36:43Phoenix983but here not
18:37:18CoCoLUSi guess this is because your hard drive's locked :P
18:37:31Phoenix983ok but if you dont use the atapwd, what program use to??
18:37:53Phoenix983h in f indicate high secure mode
18:38:09CoCoLUSwhy are you so fixated on the program
18:38:19CoCoLUSit doesn't matter what tool you're using if you don't know that friggin pw
18:39:05Phoenix983dont sayme that :'( my hope is the default pw of the factory
18:39:21HClthe only default password is all 0's / all spaces
18:39:40CoCoLUSits not like different unlocking tools "know" default factory pws
18:39:46Phoenix983but this is too for the Western Digital Hard Drives???
18:40:10Phoenix983because this is for fujitsu and hitachi hard disks
18:40:20HCli suggest you email WD
18:40:22HCland ask them yourself.
18:41:07 Quit Quelsaruk ("KVIrc 3.2.0 'Realia'")
18:41:16Phoenix983mmm i had think that, but the response is slow, but if i cannot find the solution... i ll do it
18:41:46HClunlike bios passwords and the like, harddisk passwords are designed to be secure.
18:41:54Phoenix983ok thank you for try help me :)
18:42:21Phoenix983ALGUIEN HABLA ESPAÑOL???
18:42:36CoCoLUSdon't think so, amigo :P
18:42:47Phoenix983ok byeç
18:42:51 Quit Phoenix983 ("CGI:IRC")
18:43:30HClstrange person.
18:44:01CoCoLUSaren't we too? :)=
18:44:19HClyes, but thats not the point!
18:44:21HCl :P
18:44:47CoCoLUSok :P
18:44:55CoCoLUSi'm getting myself something to eat
18:47:23*HCl goes to try to fix his thing.
19:01:10***Saving seen data "./dancer.seen"
19:09:19 Quit XShocK (" HydraIRC -> <- The dawn of a new IRC era")
19:23:47 Join jyp [0] (
19:23:47 Quit Zagor (Read error: 104 (Connection reset by peer))
19:49:36 Join zezayer [0] (
19:55:52preglowcoffee in the keyboard! i win!
19:56:19preglowif i suddenly go silent, you know why
19:56:31 Quit zezayer ("ChatZilla 0.9.61 [Mozilla rv:1.7.2/20040804]")
20:03:44 Join Stryke` [0] (
20:05:00 Join Zagor [242] (
20:05:05 Quit Zagor (Read error: 54 (Connection reset by peer))
20:26:10 Nick StrathAFK is now known as Strath (
20:37:51 Join jyp_ [0] (
20:43:15 Quit jyp (Read error: 110 (Connection timed out))
20:43:52 Nick jyp_ is now known as jyp (
20:51:13 Join Chamois [0] (
21:01:12***Saving seen data "./dancer.seen"
21:07:33 Quit n[o]bby ()
21:09:19 Quit Renko ("Leaving")
21:15:16jypAnyone knows the system call to seek in a file past 4Mo ?
21:15:57jypon linux/x86
21:18:32 Join hubble [0] (
21:22:25jypthanks ;)
21:23:02jypI can't believe this isn't in the 'see also' of lseek ... ,(
21:23:15preglowwell, it is actually
21:23:17preglowthat's how i found it
21:23:35jypnot in my local manpages
21:23:37preglowi just vaguely remebered it was something64
21:23:57preglowwell, then you've got old man pages
21:24:06preglowit's on both of my boxes
21:24:19preglowconsider upgrading
21:24:27jypdebian testing
21:24:51preglowthat's plain poor work from their side
21:24:56preglowtesting should have newer pages than that
21:25:00jypI'd say
21:25:20pregloweven stable should have newer pages than that
21:25:52jypKnowing debian, there must be a subtle political reason :(
21:27:15 Join joecool [0] (
21:27:38 Part joecool ("Leaving")
21:30:27HCli swear, debugging through the means of a filesystem is nasty.
21:31:02jypWhat do you say of _debugging a filesystem_ ? ;)
21:31:42jypActually, I love to debug by analysing gigabytes of traces ;P
21:32:28preglowi want a gdb stub :/
21:34:33HClwell, write one :/
21:34:45*HCl goes to add it to the todolist
21:34:57rasherit's already on the iriver pages
21:34:57 Part hubble
21:34:59rasherisn't it
21:35:00 Join hubble [0] (
21:35:37HClyea, i guess so.
21:35:45HClbut nobody's actively working on it nor sees it as a priority
21:35:57preglowwell, i have no idea how to write one
21:36:07preglowand i'd rather use the little time i have available on dsp stuff
21:36:40HClwell, aside from using the time i have on rockboy stuff.
21:36:54preglowif i could devote all my time to rockbox, i'd leap at it
21:37:03preglowbut i've got this pesky masters thing to write
21:39:09jypYou know Donald Knuth?
21:39:23preglowi know _of_ him
21:39:31HClis there a flush() command to flush a filedescriptor to disk or is write guaranteed to be flushed?
21:40:03preglowsync() ?
21:40:10jypWell, the lesson he teaches is that you should take care of building good tools if you don't have them
21:40:21preglowmight not be in rockbox, fo course
21:40:29HClata_flush() ?
21:40:33HClits not in the plugin api...
21:40:48preglowdoes rockbox actually do buffering?
21:41:03jypHe wanted to write books; he started to write character rendering tools...
21:41:12HCli don't know.
21:41:14preglowjyp: yes, and semi nice ones at that
21:41:54jypThe best of their times
21:41:55preglowunfortunately for the entire scientific community, he did not make good fonts
21:42:29preglowand now we all have to torture our eyes be reading papers typeset with computer modern roman
21:43:32preglow(la)tex is still pretty good, but its age is starting to show in the font support department
21:45:33jypWhatever you say; the point is few people know of his books; but alot use the tools he did in the process
21:46:17preglowhis books are quite boring, but his tools are mostly good
21:56:05 Quit Chamois (" Try HydraIRC -> <-")
22:14:40*HCl is fixing sokoban
22:17:45preglowdynarec acting up on you? :P
22:18:28HClsomewhat :P
22:18:42HClnothing bad, really.
22:18:46HClits working as expected
22:18:52HClits just that i don't have my assembly right
22:24:05preglowi'm tiring early tonight, it seems
22:24:13*preglow puts on drum&bass
22:34:08HClokay, i got a patch for sokoban
22:35:24HClcan anyone test/try that and/or commit it?
22:36:26preglowi'll give it a go
22:36:43HClmade things dependant on model and lcd width
22:39:34preglowwhy isn't my player automounted
22:40:58preglowfills the entire screen
22:41:46HClwork ok?
22:41:59preglowi really think it should buffer several levels in one go, though
22:42:05preglowspinning up the disk on level change is not very clever
22:42:17preglowi haven't played this before, it's fun :P
22:46:01 Join XShocK [0] (
22:46:45hubbleXShocK: yo.. maybe time to upload the code so someone can add it to the CVS?
22:46:53hubbleXShocK: I've cleaned and commented the code
22:46:56 Join Chamois [0] (
22:47:28preglowhave you put the i2c stuff in i2c.c, or do you use the controls regs directly?
22:48:15*HCl fixes snake
22:50:24Chamoisso anyone can put sound into CVS ?
22:50:38preglowas you see, they're working on it
22:51:08Chamoisyes but Xshock said yesterday he gives his work to LInusN but Linus didn't come here of the day
22:51:19hubblepreglow: it's a seperate file for i2c stuff, i2c-h100.c
22:51:43preglowthis will be a fine commit
22:51:53preglowChamois: then we'll wait another day
22:52:04preglowChamois: it's not like you can use rockbox to play sound now anyway
22:52:16preglowwe still need to make the codec layer
22:52:21Chamoisii know
22:52:27Chamoisno problem
22:52:59hubblepreglow: do you know of the status of the codec interface?
22:56:49 Part pillo ("Kopete 0.9.2 :")
22:57:30preglowhubble: hmm, not really
22:58:26*HCl goes back to dynarec after fixing snake and sokoban, and concluding that snake2 is darn hard to fix
22:58:36HCli think, anyway.s
22:58:43preglowwe need a couple of core coders around to discuss it
22:58:49preglowand they've been mostly absent lately
22:59:21preglowbut dave's done a bit of work and has things more or less worked out
23:00:30 Part Chamois
23:01:14***Saving seen data "./dancer.seen"
23:04:35HClenough coding for me today
23:09:41XShocKhubble: i came back.
23:10:37XShocKyes, i gave the code to LInus
23:38:44 Join asdsd [0] (
23:38:50asdsdhey guys
23:40:09asdsdi wanted to gather some information
23:40:25asdsdi noticed you guys have been able to load rockbox onto the h1xx
23:40:59asdsdwell i just wanted to know if u guys know the entry point address for the decompiled firmware
23:41:05asdsdsee im trying to load up on IDA
23:41:20asdsdbut its in binary and its telling me i have to press C on the entry point address
23:41:26HClhm, thats a question for linus, really.
23:41:31asdsdinstead of figuring it out myself, i decided to find u guys
23:41:39asdsdis linus here today?
23:42:02hubbleasdsd: entry point in .bin is 512+32 = 546
23:42:10hubbleasdsd: see
23:42:14asdsdthanks hubble
23:42:19asdsdomg thankyou
23:43:05hubbleasdsd: I've created a IDA project that contains all code at the correct virtual adresses
23:43:38asdsdit would be a huge help if u cud just email it to me, or send it to me via xdcc
23:44:23preglowasdsd: what are you planning on doing?
23:44:34asdsdwow that memorymap is really gonna help, u guys really ARE working on it
23:44:48asdsdwell don't laugh but im trying to fix the drm issue on the US firmware
23:45:05preglowstandard cracking job, heh
23:45:15HCldoes h1x0 have drm?
23:45:28preglowh3x0 does
23:46:10 Quit jyp ("poof!")
23:46:15asdsdim working on the h320
23:46:22asdsdbut i figure they both use the same entry point
23:47:53hubbleasdsd: yes, and if you follow it to the section which copies from flash to sram and sdram you can probable see the h3x0 memory layout aswell
23:48:16preglowthey copy flash code to sdram?
23:48:52hubblepreglow: sure.. see 0x1528
23:48:55preglowwhy do they do that? according to linus, executing from flash is actually faster
23:49:16asdsdok guys
23:49:33asdsdso im loading the h300.hex file right, its decoded using ihpfirm
23:49:42asdsdthen i set the cpu to moto:coldfire
23:49:44hubblepreglow: most of the irivers firmware is at 0x31000000 (sdram) and a few routines are in sram (0x10000000)
23:50:05asdsdin the input file, i set the loading address to 0x544?
23:50:07preglowand by the way, do anyone know how stripwax managed to figure out the encryption?
23:50:35hubblepreglow: nope, i'm a little curious too.. we should ask him :)
23:50:48preglowi was planning to, but haven't seen him lately
23:50:54preglowlooks like it's a block xor code with feedback
23:51:03preglownot something i can reverse without prior knowledge, at least
23:51:25hubblepreglow: maybee he did some statistical analysis?
23:51:28asdsdstripwax is a genius
23:51:54preglowhubble: perhaps, i'll be the first to admit i really don't know much about that

Previous day | Next day