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 2004-09-11

00:00:10Dma-Scbother you i mean ;)
00:00:14ZagorDma-Sc: yes, a motorola coldfire (modified 68k)
00:00:40Dma-Scoh really
00:00:50Dma-Sclet's have a st emu on it! ;)
00:01:24Dma-Scor at least play sndh tunes ;p
00:02:07amiconnZagor: There is a sidplay port to the Amiga. The author says an 68030 (standard 68030 clock on Amiga is 25 MHz) is sufficient, though he recommends 68040
00:03:20Zagorthe very first SID emulator is even faster:
00:04:08amiconnYeah, but playsid _is_ assembler
00:04:11Dma-Sczagor : and it got an additional dsp for the main depacking job then?
00:04:31Zagorah the port you're talking about is c? got a link?
00:04:51ZagorDma-Sc: no, just an emac for the multiplications
00:05:21amiconn...and playsid uses custom chips of the amiga, namely the paula and the cia
00:05:21Dma-Scah ok
00:05:46Zagorplaysid uses all the tricks in the book, and then some... :)
00:09:15amiconnThe sidplay lib port to amiga is on aminet, but without sources...
00:09:36Zagorthe "old" sidplay lib is apparently faster than the newer sidplay2, but it too is written in c++.
00:09:41Zagorah, badness
00:10:22Dma-Sczagor : btw do you think such a cpu could handle mod/xm/s3m playback routs?
00:10:31ZagorDma-Sc: definitely
00:10:47Zagormod is a very low-cpu format
00:11:21amiconnLow-cpu, low-memory and low-diskspace, yes
00:11:21Dma-Sczagor : depends on the number of voices and effects though
00:11:54ZagorDma-Sc: even with lots of voices and effects, it's a lot cheaper than mp3 or ogg
00:12:02Dma-Scah yes
00:12:40Dma-Scanyway this would be cool when noticing there's still a long way back home and you've already reached the last bits of batteries ;p
00:13:03 Quit lImbus (" HydraIRC -> <- The dawn of a new age")
00:13:24 Quit methangas (" HydraIRC rocks! -> <-")
00:13:31Zagorwell we don't know yet how much adaptive power saving we can do.
00:13:39Zagorbut one can hope
00:13:55Dma-Sconly things we, users, can do yes ;)
00:15:21Dma-Scok, good luck on this people then and see you :)
00:16:26 Quit Dma-Sc ("YM2149")
00:23:55 Quit zeekoe (Read error: 110 (Connection timed out))
00:28:27Zagor"Ogg/Vorbis in embedded systems. Master thesis by Johannes Sandvall"
00:32:22_Lucretia_Zagor, yup, they have those here as well ;-)
00:32:43 Join vadin2003 [0] (
00:33:56 Quit vadin2003 (Client Quit)
00:38:29 Quit midk (Read error: 104 (Connection reset by peer))
00:38:52 Join midk [0] (
00:41:00 Join bagawk [0] (
00:46:01***Saving seen data "./dancer.seen"
01:09:20 Quit Bagder ("Leaving")
01:21:35Zagorbed time. night all!
01:21:36 Quit Zagor ("Client exiting")
01:38:48 Part amiconn
01:49:53 Quit bagawk ("umount /dev/brain")
01:54:37 Quit scott666 ("i'll be back...eventually...")
02:12:42 Quit mecraw ("Trillian (")
02:21:47 Quit midk (Read error: 104 (Connection reset by peer))
02:24:01 Join midk [0] (
02:34:17 Join hardeep [0] (1098@
02:46:03***Saving seen data "./dancer.seen"
03:15:15 Quit hardeep ("[BX] Reserve your copy of BitchX-1.0c19 for the PalmPilot today!")
03:16:03 Join BC [0] (
03:26:39 Join BC_ [0] (
03:26:39 Quit BC (Read error: 54 (Connection reset by peer))
03:33:26 Join ashridah [0] (
03:40:33 Part BC_
04:10:01 Quit ashridah ("Client exiting")
04:10:37 Join ashridah [0] (
04:20:32 Join scott666 [0] (
04:39:12 Join ze__ [0] (
04:41:06 Quit ze (Read error: 104 (Connection reset by peer))
04:41:06 Nick ze__ is now known as ze (
04:46:05***Saving seen data "./dancer.seen"
05:06:27 Join Synthe [0] (synthe@
05:06:41 Part Synthe
05:24:29 Join maikeul [0] (
05:42:30 Quit gromit`` (Read error: 110 (Connection timed out))
05:55:33 Join [av]bani [0] (
05:55:41[av]banianyone have a good picture of the disassembled lcd?
05:56:19midkyou can't really disassemble an lcd.. unless you shatter it with a hammer >:)
05:56:57 Join Ka [0] (
05:57:09[av]banii mean, separate from the ihp
05:57:21[av]banithats not terribly useful
05:57:28[av]banieg, the ribbon cable, cog etc
05:57:43midkwell, if it did exist it would be there
06:26:30pikemust be something useful on the flipside fo the lcd
06:46:08***Saving seen data "./dancer.seen"
07:10:13 Part [av]bani
07:55:50 Join methangas [0] (
08:08:27 Quit scott666 ("i'll be back...eventually...")
08:32:32 Quit midk ("just STOP it arspy")
08:36:52 Join midk [0] (
08:46:09***Saving seen data "./dancer.seen"
08:59:29midkbed, nite
09:18:32 Join funkymonkey [0] (
09:18:33 Quit methangas (Read error: 54 (Connection reset by peer))
09:28:47 Join amiconn [0] (
09:33:21 Join Bagder [0] (
09:38:41 Quit pike (Read error: 104 (Connection reset by peer))
09:59:48 Join [IDC]Dragon [0] (
10:00:59amiconnmorning Jörg
10:01:33[IDC]Dragondo you expect your Ondio taday?
10:02:02amiconnDunno whether dpd delivers on Saturdays too, if so ->yes
10:03:03amiconnWhile were very productive the last days (hey, you cover the cvs activity log almost completely!) I still can't be of much help
10:03:32amiconn(other than I looked into you changes a bit, and why the plugins don't compile)
10:03:49amiconn*While yo were...
10:04:14[IDC]Dragondid you find it out?
10:04:41[IDC]Dragonright now I'm almost happy they don't compile, keeping the make small
10:06:08 Join pike [0] (
10:06:39amiconn(1) The Makefile in apps/plugins/lib does need the same extension as that in apps/plugins (including the recorder path for the Ondio too
10:07:12amiconn(2) Then _many_ plugins fail with the "same case value" error, so need fixing
10:07:42dwihnoMorning Jens & Jörg
10:07:58[IDC]Dragonhmm, we may better leave (2) until we have a better button driver
10:08:12[IDC]Dragonhi dwihno
10:08:47[IDC]Dragoni'm not worried about plugins right now, without disk I/O
10:08:57amiconnOf course
10:09:35amiconnThe source tree needs some major restructuring imho.
10:09:42dwihno[IDC]Dragon: how do you boot rockbox on the ondio?
10:10:24amiconn(1) Until now the recording feature was connected with having a graphic display. This is no longer true
10:11:18[IDC]Dragonis it that dirty?
10:11:33[IDC]Dragonwe have a config define for that
10:12:00amiconnYes, as e.g. the widgets are in the recorder sub-dir
10:12:07[IDC]Dragondwihno: the usual way: hook it to USB, copy ajbrec.ajz onto it
10:12:40[IDC]Dragonamiconn: the recorder subdir is in the make
10:12:54[IDC]Dragonah, you're talking plugin make?
10:13:24amiconn(2) Same goes for the number of buttons and their layout. Imho this requires adding some pre-processing in the button driver, i.e. emitting events (e.g. "go to the menu") instead of or in addition to raw button events
10:13:54[IDC]DragonI woudn't touch that yet
10:13:55dwihno[IDC]Dragon: ah, okay... I started thinking there was a need of some kind of special flashing
10:14:33[IDC]Dragonamiconn: let's try to walk before we run
10:15:01[IDC]Dragonperhaps you could work on the MMC driver
10:15:20[IDC]DragonI found some nice code (for Atmel uC) to start with
10:15:38[IDC]Dragonsince you're good about datasheets
10:16:16[IDC]DragonI'd carry on the reverse engineering then
10:17:01amiconnI'll try that, as soon as my Ondio arrives. Of course I will also take a look into the hardware
10:17:30[IDC]Dragonlook under AVR Projekte
10:17:49[IDC]Dragonthen MMC-SD
10:18:35amiconnFound it.
10:18:40[IDC]Dragonsourcecode has german comments
10:18:50[IDC]Dragonso no problem :-)
10:19:07[IDC]Dragonthe MMC driver won't need much code
10:19:42amiconn..and not much optimization either, since it uses the serial i/f, which does dma
10:19:56amiconnBut perhaps we need byte swapping... lets see
10:20:59[IDC]Dragonit's "just" establishing the syncronous serial, working with the few commands
10:22:31[IDC]Dragonand working with the clock/select gates of the quad AND
10:22:44[IDC]DragonI don't fully understand that part yet
10:23:15[IDC]DragonThought I'm close to do a USB enable within Rockbox, but it doesn't work
10:25:11amiconnA propos and gate: I looked a bit at your table, and found that there is a contradiction in it:
10:26:49amiconn(1) It says that pin 13 is connected to pin 11, but pin 11 is not connected to pin 13
10:27:26amiconn(2) it says pin 4 is connected to pin 13, but pin 13 is not connected to pin 4
10:27:48amiconnI *guess* that means pins 4 and 13 are in fact connected
10:28:37amiconnUnfortunately, there are still too many connections missing to completely understand what it does
10:28:55[IDC]DragonI know a bitmore, updating now
10:30:00amiconnBtw: I have an idea what the connection of SIBI to the cpu may be good for:
10:30:42amiconnThis must be held high for mp3 (bitstream) playback. However, for word oriented transfers this is the word sync
10:31:10amiconnThis may be useful e.g. for the pcm pass through codec...
10:32:34[IDC]Dragonbut the parallel is read only, grr :-(
10:37:02 Quit _Lucretia_ (Read error: 110 (Connection timed out))
10:38:25 Join _Lucretia_ [0] (
10:39:36amiconnI really would like to go into mas hacking... I have some dreams... Super-fast jpeg decoding... .mod playback...
10:40:53[IDC]Dragonwe could mod the Ondio for bidirectional
10:46:13***Saving seen data "./dancer.seen"
10:47:13 Join [av]bani [0] (
10:47:25[av]banimystery chip identified
10:47:39[av]baniAZU 3AW ZD8L
10:47:48[av]banivoltage regulator
10:48:49amiconn[IDC]Dragon: I think this is not really necessary. The only advantage would be to use the pcm pass through codec unmodified. I doubt it would be much faster than the serial with out boxes, since the parallel is completely software controlled
10:51:45[IDC]Dragontwiki updated
10:56:00amiconnTwo pin functions are swapped: presuming your input pin names, pin 8 is O2 and pin 11 is O3
10:56:25Bagder[av]bani: cool
10:57:45[IDC]Dragonamiconn:fixed, thanks
11:00:19amiconnPin 3 (O0) does go nowhere???
11:01:05[IDC]Dragonit goes somewhere, doing it's job
11:01:18[IDC]Dragonbut I don't know exactly how
11:04:11amiconnGate 3 apparently "mixes" the mmc clock from the bridge and the cpu; that of course requires that the passive devices has to hold the clock line high
11:04:57amiconnGate 0 simply buffers the usb enable signal
11:05:22amiconnIt would be really interesting what O0 is connected to
11:06:25amiconnDid you measure levels while the unit is running (with archos fw)?
11:08:13[IDC]Dragonone thing I know about the output: it is connected to the USB line itself, via a resistor
11:08:36[IDC]Dragonprobably messing up the bus while inactive
11:09:58[IDC]Dragonfirst I'd like to work out those selects
11:12:10*amiconn also thinks that deciphering the functions of this quad and is most important, since wrong setup could also prevent an mmc driver from working
11:20:32[IDC]DragonI made a mistake with my USB mode
11:21:43[IDC]Dragonthe bridge will use MMC mode, not SPI mode
11:21:52 Part [av]bani
11:22:01[IDC]Dragonso no /CS involved, that's only for SPI
11:22:29[IDC]Dragonbut still the card has to be switched
11:26:37[IDC]Dragonneed to measure the levels, in USB mode
11:41:02amiconnAlready read a bit about mmcs: (1) The driver shouldn't be hard, mmc is much less complex than ata (2) mmc and sd are identical in spi mode, so the Ondio should also support sd cards. Could/ did you verify this?
11:41:38amiconnThis is in fact interesting for me, because I already have a 256 MB sd card...
11:45:24[IDC]Dragonit won't fit, mechanically
11:45:30[IDC]DragonI tried
11:45:47[IDC]DragonI'm away now, mostly
11:46:29[IDC]Dragonperhaps you can mount a different socket
11:52:13amiconn[IDC]Dragon: r u there?
11:52:41amiconnOops, was way scrolled up..
11:53:28amiconnI have an idea what PA12 _might_ be good for
11:55:12amiconnYou need a way to switch the card back from spi mode to mmc mode. The docs say "The only way to return to the MultiMedia Card mode is by entering the power cycle."
11:56:44[IDC]Dragonbut PA12 only influences the clock gating...
11:57:19amiconnHmm, there must be some logic to do this...
11:58:07[IDC]Dragoniirc, Vcc is hard wired
11:59:12[IDC]Dragonmaybe the bridge uses SPI mode, too
11:59:27[IDC]Dragon(haven't studied the datasheet)
12:14:40 Quit Bagder ("Leaving")
12:23:44amiconn[IDC]Dragon: Argh! MMC serial communication is msb first, while the SH serial does lsb first. Sounds familiar?
12:46:16***Saving seen data "./dancer.seen"
12:46:47[IDC]Dragon<irony>Great! Then we need to do nothing for mp3</irony>
12:47:23[IDC]Dragonwell, at least the CPU gets something to do while the DMA runs
12:47:46amiconnThe serial comm itself should be possible up to 1.5 MBit/s, assuming cpuclk == 12 MHz
12:48:03[IDC]Dragonbut the mp3 payload gets bitswapped twice
12:48:11amiconnyup :(
12:48:48[IDC]Dragonif I ever build a USART, I'll make this programmable, promised!
12:49:41amiconnAt least our bitswap routines are faster than the serial xfer itself :/
12:56:40amiconnArgh! I missed the end of a Studio 10 auction!
13:00:03[IDC]Dragonhow much?
13:02:59amiconnIt went for € 94,- which would have been within my limit...
13:24:42[IDC]Dragonoh, that's a lot
13:25:10[IDC]Dragonbut the price is the limit of the 2nd bidder
13:25:33[IDC]Dragonyou'd have to be above the limit of the 1st guy
13:26:02[IDC]Dragonwho knows what that would've been
13:26:27[IDC]Dragon"Mittagsschläfchen" now
13:27:42dwihnosounds like a good idea
13:27:57dwihnoIch habe doch zuviel geschlafen :)
13:33:22kaboofaIch liebe apfels!
13:33:58kaboofaSie sind Orangen!
13:34:09kaboofaAs usual, I have nothing useful to contribute to #rockbox :(
14:09:12 Join R3nTiL [0] (
14:09:24 Join Bagder [0] (
14:26:25 Quit [IDC]Dragon (Read error: 110 (Connection timed out))
14:40:29 Quit R3nTiL ()
14:46:17***Saving seen data "./dancer.seen"
15:05:27 Join Zagor [0] (
15:06:55Bagderhey bro, see ya in a while. Agnes just woke up
15:06:58Zagorargh, [av]bani said he found out the chip but doesn't updated the wiki with the information...
15:07:04 Quit Bagder ("Leaving")
15:28:43elinenbehow is progress on the iriver coming?
15:32:10Zagorwell most of the chips are documented now and I incorporated the MAD mp3 decoder in rockbox yesterday
15:32:23Zagorit's really slow on the 12MHz sh1 :)
15:33:14Zagorhopefully we'll have a BDM cable on monday
15:41:57ashridahzagor: where's the BDM port on the pcb of the iriver? the 5 solder pads on the righthand side of the front of the board, or the set of a lot more pads on the left hand side?
16:19:00 Quit funkymonkey (" HydraIRC -> <- irc client ownage!")
16:23:51 Join methangas [0] (
16:24:20 Join LinusN [0] (
16:25:27LinusNashridah: the bdm port is the 20-pin connector on the left side
16:27:38ashridahhm. so any idea what the block of 5 pads on the right is?
16:28:57LinusNtwo of the pins go to one of the serial ports, so i guess we can use it for debugging as well
16:29:19LinusNcheck the back side
16:29:48 Join _aLF [0] (
16:31:50ashridahah, yeah, so i see
16:38:54LinusNthe main "problem" right now is the lcd
16:40:12ashridahyeah. no details on the connection. i assume there's a PIC or something buried into it?
16:40:39ashridahmaybe you'd be better off decoding the signals sent to the remote LCD first, since it looks like it might just be VCC,GND, RX/TX ...
16:41:02ashridahhrm. scratch that. it has two jacks as well as the data pins
16:43:24LinusNall lcd modules have a chip on them, Chip On Glass
16:43:42*ashridah never really had to deal with LCD units at uni
16:44:52LinusNwe can disassemble the lcd driver in the original firmware, and we can analyze the signals to the lcd
16:44:58LinusNi intend to do both
16:46:19***Saving seen data "./dancer.seen"
16:46:33LinusNgotta go, cu around
16:46:36 Part LinusN
16:49:44 Join scott666 [0] (
16:58:08 Quit ashridah ("gone even")
17:12:38 Join zeekoe [0] (
17:16:20 Part scott666
18:25:36 Join _Lucretia [0] (
18:25:56 Quit _Lucretia_ (Read error: 110 (Connection timed out))
18:32:02kaboofaI just got a weird idea....
18:32:42midkgo on!
18:33:04kaboofaWhat if I wrote drivers for rockbox
18:33:16kaboofathat uh..
18:33:18kaboofaugh, wtf
18:33:34midkfeel free to go on... :)
18:34:08 Quit zeekoe (Read error: 60 (Operation timed out))
18:34:30kaboofaSorry, got ap hone call
18:34:32kaboofa*a phone call
18:34:45midkoh. np ;)
18:35:12kaboofaOk, make some hacks to the usb driver, so if an option was enabled, then when the usb wire was plugged in the recorder would show up as a game controller
18:35:15kaboofaor a joystick or something
18:35:26kaboofaI know that sounds stupid
18:35:48kaboofabut what if you had like.. an american football game or something, you could select the plays on the rockbox screen
18:35:51kaboofaon the lcd
18:35:53kaboofawith special drivers
18:35:55midksounds awesome until you tell someone about it :)
18:36:13midkhmm. very doubtful.. but i'm not sure of much about it
18:36:16kaboofaBut it still sounds awesome to me, so i'm going to investigate ring0 for windows xp and hacking the kernel
18:36:23kaboofathe kernel meaning rockbox
18:36:41kaboofai could probably make a unix port too
18:37:05midkmhm, sure, and i'll ..uh.... cool
18:37:45kaboofaThat would be cool though
18:38:33kaboofaHmm, need to find my msdn cds...
18:38:43midkwhat ever happened to that polish notation calculator or whatever?
18:38:47kaboofaOr wait until oct.
18:38:50kaboofamidk: it died.
18:38:59kaboofaI didn't need it anymore, because I found one for my ti89
18:39:00midkghetto_sprintf didn't work?
18:39:04kaboofathat too
18:39:07midkwtf.. we could still use it.
18:39:20midkthat doesn't mean i have one for my.. uh... yeah
18:39:20kaboofaok =(
18:39:31midkwhat's so hard about ghetto sprintf anywyas
18:39:35midkwhy do you need it?
18:39:51kaboofaNeed what?
18:39:54kaboofaRPN or my MSDN cds?
18:40:06 Join zeekoe [0] (
18:40:44dwihnoghetto code! \o/
18:41:38midkbut, why do you need it kaboofa?
18:43:05kaboofabecause you can't do float->string conversions in rockbox
18:44:12Zagorkaboofa: the usb is hardwired to the isd300 chip. the processor has no control.
18:44:34kaboofathere goes my plans.
18:44:40midksee, that's exactly what i told him
18:44:46midkisn't it? isn't it?! ;)
18:44:55kaboofathanks Zagor.
18:46:21***Saving seen data "./dancer.seen"
18:49:38kaboofaWhen did you say that?
18:50:00kaboofaAnd it would be s/Zagor/midk/; ;p
18:50:19kaboofaregex == leetsauce;
18:52:58 Join R3nTiL [0] (
19:03:53 Quit R3nTiL ()
19:11:59midkkaboofa, not really
19:12:04midk-/+ works too
19:20:31 Quit zeekoe (Read error: 60 (Operation timed out))
19:29:18 Quit _aLF (Remote closed the connection)
19:32:09 Join zeekoe [0] (
19:32:53 Quit midk (Remote closed the connection)
20:00:45 Join R3nTiL [0] (
20:03:28 Quit R3nTiL (Client Quit)
20:12:06 Join Bagder [0] (
20:24:02 Join bagawk [0] (
20:26:07bagawkhi ze
20:26:10bagawki mean zeekoe
20:30:10 Quit zeekoe (Read error: 110 (Connection timed out))
20:30:42Bagderyou scared him away! ;-)
20:46:23***Saving seen data "./dancer.seen"
21:19:17 Quit bagawk ("umount /dev/brain")
21:50:05 Join scott666_ [0] (
22:04:31 Part scott666_
22:13:21 Quit Bagder ("Leaving")
22:21:09 Quit silencer (Nick collision from services.)
22:21:11 Join silencer [0] (
22:21:22 Quit silencer (Nick collision from services.)
22:26:10 Join silencer_ [0] (
22:46:27***Saving seen data "./dancer.seen"
22:59:37 Join bagawk [0] (
23:18:22 Join zeekoe [0] (
23:20:38 Quit zeekoe (Read error: 104 (Connection reset by peer))
23:21:18amiconnGrr, this is annoying!
23:22:08amiconnDid some test recordings, with varying level including clipping, archos fw & rockbox
23:22:26amiconnVery same CD all the time, 3 different level settings each
23:22:28bagawkAnd the result?
23:22:47amiconnarchos fw: no frame sync errors
23:23:28amiconnrockbox: 2 of 3 recordings have sync errors, the 3rd merely has the last frame truncated
23:24:25bagawkany ideas where the errors are coming from?
23:24:37amiconnNot really :(
23:25:19bagawkcould you do some debug, and find out where the sync errors occur, and what rockbox was doing at the time?
23:25:21amiconnNext test: rockbox modified to not use m/s stereo encoding. Perhaps the mas has a bug with m/s stereo (although I don't really believe this)
23:25:57amiconnIn parallel I'll write some more thorough mpeg frame analyzer tool
23:32:14 Join Fusen [0] (~Fusen@fusen.user)
23:32:23Fusenwhere can I get descramble from?
23:32:44bagawkFusen, you need to compile it yourself
23:32:51bagawkFusen, why do you want it?
23:33:06Fusennew av300 firmware was released
23:33:18Fusenthere were numerous copies of it on the av300 yahoo groups
23:33:22Fusenbut they have all gone
23:33:35Fusenand I'm sure rockbox used to have it as a download
23:33:37Zagorgrab the source tarball or take it from cvs
23:33:58 Quit Fusen ("( #SG1.ET Perform and Idle )")
23:34:45 Join Fusen [0] (~Fusen@fusen.user)
23:34:51Zagor...or download from
23:35:15bagawkcan the tool really descramble the av300 firmware?
23:36:28Fusenall done, thanks for that
23:36:33 Part Fusen
23:38:06 Join CHubaca3256984 [0] (

Previous day | Next day