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 2003-06-23

00:00:06jzossdos prompt
00:00:13jzossI thought you "knew computers" ??
00:00:45Zanei do it brings up a prompt
00:00:48Zaneand than closes it
00:02:31Zaneim just dling the installer
00:02:49Zane8 meg but im dlin at 150
00:03:20Zanenobody talks much here
00:03:23***Saving seen data "./dancer.seen"
00:03:46jzossMany of the ppl live in Europe, and are asleep now. =)
00:04:03Zanewell wake them up
00:04:09Zanei dont live in europ
00:04:50tracktheripperi live in europe!
00:04:53tracktheripperin London!!!
00:05:03Zanei live in usa
00:05:27Stevie-Otracktheripper: what a coincidence!
00:05:34Stevie-Othat's exactly where I don't live
00:07:49Zaneahh ahh
00:10:06 Quit Zane ("Leaving")
00:13:45 Quit Jet8810 (Connection reset by peer)
00:21:00 Join Jet8810 [0] (
00:24:58 Quit Jet8810 (Read error: 54 (Connection reset by peer))
00:24:58 Quit elinenbe (Read error: 54 (Connection reset by peer))
00:24:58 Join Josh_ [0] (
00:25:01 Join elinenbe [0] (
00:26:47 Quit tracktheripper ("Leaving")
00:27:39Josh_join gentoo
00:27:49 Nick Josh_ is now known as Jet8810 (
00:29:10Stevie-Oyou join jentoo
00:34:20 Join OliverKlozoff [0] (
00:34:21 Quit Stevie-O (Read error: 104 (Connection reset by peer))
00:37:49jzossjoin genthree
00:38:54OliverKlozoff* Now talking in #genthree
00:38:54OliverKlozoff* sets mode: +n
00:54:18 Join Bluechip [0] (
00:54:51Bluechipany programmers in?
01:00:04OliverKlozoffaccording to certain theories
01:00:18OliverKlozofftell me −− did you try menutest yet?
01:07:20Bluechipno, still working on my own code - which bit of the menu system does it test
01:08:41Bluechipanyway in button_get_w_tmo long is a "tick"
01:08:53Bluechipi presume it is some factor of "HZ"??
01:11:23jzosskernel.h: #define HZ 100 /* number of ticks per second */
01:12:19 Join Yeft [0] (
01:12:23 Quit _aLF ("bye")
01:12:36Bluechipso long is a "tick" then?
01:12:49Bluechipoh shit - sorry, two conversation
01:13:03Bluechipnote to self: STOP SCAN READING
01:13:12Yeft...can you guys help with the original firmware at all..(fm recorder)
01:13:12jzossif HZ is 100 ticks/sec, tick is 1/100 sec
01:13:26Bluechipyeah, sorry, just me being dumb!
01:13:53jzosswhat's your Q, Yeft? (I don't have an FM)
01:14:55Yeftwhen i turn it oh, everything loads fine, but its all jibberis (i.e. $#^DSF is a folder) and if i try to enter an of the folders it says hard drive error, and if i try to turn on the radio it crashes...
01:15:19jzossHmmm. Have you tried to scandisk the drive?
01:15:42Bluechiprun chkdsk before scandisk
01:15:48jzossthat'd be the first thing I tried.
01:15:54Bluechipchkdsk chkdsk /f scandisk
01:16:51earHurtsok, I just opened my fm and removed the hdd
01:17:32earHurtsI put it back in, re-assembled, and it is till functioning
01:18:01Yefti think it worked
01:18:07Yeftlol it said i bunch of stuff
01:18:13earHurtsthe case, however, is bent in acurve at the back top and bottom
01:18:17Yeftconvert lost chains to files?????
01:18:25earHurtsyeft: no
01:18:25Bluechipyes please
01:18:35BluechipConvert = YES *always*
01:18:53Yefti typed n
01:18:56Bluechipthis is your lost data - the alternative is lose it forever
01:18:58Yeftbrfore you chimed in...
01:19:07Bluechipbut it will not convert unless you use /f - stage 2
01:19:12earHurtswhat,s he going to do with partial mp3 filesa/
01:19:22jzossmost ppl have their JB synced w/ PC anyway. Easy to recover. =)
01:19:47Yeft<−− not most people
01:19:55earHurtsoh, dear
01:20:29Yeftit works!
01:20:35Yeftthank you
01:24:19Yefti love all my new ultimixes (ac/dc you shook me all night)
01:24:27Yefti love that one the most
01:25:37Yeftthank you all!
01:25:44Yeftseeya later
01:25:56 Quit Yeft ("i'd rather be rich than stupid - jack handy")
01:28:55 Part Bluechip
01:36:19earHurtsthe bent case means that the battery contact isn't secured as tightly
01:36:34earHurtsany ideas on straightening that bend?
01:36:56jzossIs the piece flat, or does it have the sides on too?
01:37:30earHurtssides, which are soldered to the mainboard
01:37:45earHurtsthat's why i bwnt, to avoid soldering
01:37:54jzossSo you can't exactly press it with anything, since it's got circuit stuff mounted to it
01:38:34earHurtsI was able to hold the pieces together
01:38:56jzossIf you can't pull the circuit stuff away from the case (to bend it back into shape), I'd probably just try and shim the battery contact to fit better
01:39:14earHurtsonce it was all reassembled
01:39:44jzossshim == stick small pieces of metal inbetween two parts to do fine-adjustment on spacing/offset/angle/whatever
01:40:13jzossAs in: can you just wad up some aluminum foil to fill whatever gap you're seeing?
01:40:30earHurtsthat's essentially what the archos battery cover does
01:41:10earHurtsno: the battery cover is the contact ( is attached to, anyway)
01:41:31earHurtsand the cover is flexed like a spring
01:42:06earHurtsit's held down becuase it's wwdged between the front cover and the back ( now bent) case
01:42:13jzossyou'd prolly be better off asking Linus, or someone who's actually disassebled the battery cover. =)
01:42:14 Join Bluechip [0] (
01:42:29earHurtswith the bend, it's not wedged as much.
01:42:34earHurtsgood suggestion
01:42:35Bluechipcan i retrive the current tick?
01:42:37jzossso shim under battery
01:43:16jzossBluechip: lots o references in mpeg.c for "current_tick". You could check there...
01:43:43Bluechipthanks :))
01:45:40ricIIfirmware 1.17i, must be old..
01:52:29ricIIinterface pcb also differs (v1.2)
02:03:25***Saving seen data "./dancer.seen"
02:07:46 Join Asmotaku-Neko [0] (
02:07:51Asmotaku-NekoHi there !
02:08:23Asmotaku-Nekono ones.
02:08:43Asmotaku-NekoAnybody there ?
02:09:05 Quit Asmotaku-Neko (Client Quit)
02:14:29 Join LinusN [200] (
02:14:56jzosshiya, LinusN =)
02:17:45Bluechipahh, is that a man who can tell me how to access a timer with units smaller than a second? ;)
02:21:02LinusNprobably, yes
02:21:17LinusNone tick is 10ms
02:21:36Bluechipjust can't work out how to read the current value into a variable!
02:22:24Bluechipalso - does it wrap at say, 2^65
02:22:25LinusNit *is* in a variable
02:22:44LinusNit is an unsigned 32-bit
02:23:22Bluechipand is it free running or reset every X seconds?
02:23:31LinusNif you use the TIME_AFTER() and TIME_BEFORE() macros, you won't have to think about the wrap issues
02:23:42LinusNit is free running
02:24:06Bluechipbrilliant - thank you - i will grep for these macros :) <- happy again
02:24:20LinusNsee the implementation of sleep() in kernel.c
02:24:32LinusNthe macros are defined in kernel.h
02:24:59Bluechipbrilliant - I will try to work out how to create a patch for my new game tonight :)
02:25:05Bluechipthis was my last problem
02:25:08LinusNa game! nice!
02:25:28LinusNwith subsecond timing?
02:25:32Bluechipsimple, but i just wanted to get used to accessing the display/keys/timers/etc
02:25:50Bluechipi want to slow the oponent down - he moves too quickly
02:26:34LinusNwouldn't sleep() do fine then?
02:27:12Bluechipi also tried button_get_w_tmo(), but they did not give me the flexibility i required over the keyboard
02:27:40LinusNi see
02:30:39Bluechipit will make sense when you see the code
02:30:49LinusNlooking forward to it
03:00:46 Part LinusN
03:33:44OliverKlozoffI missed Linus
03:34:32OliverKlozoff0009 nop 4041 9481550 2346.3
03:35:11OliverKlozoffnop takes an an average of 2346 instructions to execute
03:35:36OliverKlozoffthat's actually MORE than many instructions
03:37:17jzossStevie: Scan the IRC logs for today. Someone pointed out a link to a company that's offering a disassembly package of the Archos for like $2k a pop. =) I wonder if they got any of their info from rockbox?
03:37:34OliverKlozoffdisassembly package?
03:37:38OliverKlozoffwhat do they mean?
03:37:51jzossI think it's hardware: component lists, circuit diagrams, etc.
03:38:02OliverKlozofffuck that, we already have all those
03:38:04jzossNot sure if there's any software disassembly/analysis
03:38:06OliverKlozoffhowever incomplete
03:38:08Bluechipcould write a simple disassembler in an hour or so :( ...$2000 - LOL
03:38:17OliverKlozoffwe HAVE one
03:38:20jzossI know. That was my thought
03:38:26OliverKlozoffcomes with the source
03:38:48Bluechipstill working my way though it all :)
03:38:49CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
03:38:49*OliverKlozoff has used it extensively
03:40:47Bluechipanyone know if the preprocessor will handle this okay:
03:41:31Bluechipi'm expecting 70
03:42:16Bluechip...given HZ = 100 and REP_ST = 6
03:42:30jzossinteger division?
03:42:42Bluechipthere is a 0.7 in the middle of it all
03:42:56Bluechip(6+1)/10 = 0.7
03:43:27jzossor (6+1)/10 = 7/10 = 1 (or whatever int division gives you...I forget)
03:43:46Bluechipround down
03:44:25OliverKlozofffractional portion is dropped
03:44:37Bluechipis that a no then?
03:44:40jzossright. I usually avoid at all costs. =)
03:44:52jzossI think it'll always give you zero. Not 70
03:45:02Bluechipwhat i feared
03:45:12jzossJust multiply first, then divide
03:46:31*Bluechip cowers in a pool of missing the obvious
03:46:54jzoss=) it's 'cause your mind is involved in "higher things" ;)
03:46:59jzossEinstein couldn't do math, either
03:47:12Bluechipty ;)
03:55:01 Quit BoD[] ("hhh")
03:57:54 Join nu [0] (
03:58:03nugood evening
03:58:21nuhi Bluechip
03:58:33nuis LinusN ever around ?
03:58:52Bluechiphe popped on earlier, solved my problem and went away again
03:58:57Bluechip...such a nice man :)
04:03:27***Saving seen data "./dancer.seen"
04:04:36nui was browsing the internet looking for VBRI header data when i came across a log of this channel where LinusN stated he had figured it out
04:05:17nusince its 4am where i live and i need this done by yesterday, it would've been so very nice of him to share his findings with a perfect stranger ;)
04:05:33Bluechipsorry, all alien to me - i have a friend who may know - but he is not onlne atm :(
04:05:45Bluechipif he had been online - he surely would have helped
04:05:59nuthanks anyway...i'm probably too tired to even understand anything
04:06:13nui'll pop in tomorrow if i cant figure out this stuff by myself
04:06:14nuthanks again!
04:06:24 Quit nu ("smash is the way you deal with your life, like an outcast you're smashing your strife")
04:09:44 Join Stevie-O [0] (
04:15:05 Quit adi|home (Read error: 54 (Connection reset by peer))
04:16:42Stevie-Oso i need a 27 GHz CPU
04:16:59Stevie-Oto emulate the archos at full speed
04:17:04jzosslol. You can have mine. I just upgraded
04:19:09Stevie-Oyou have a 28GHz now?
04:21:23Stevie-Onow I only need a 25GHz CPU
04:22:27jzossdamn, I need a beowulf cluster of ..... nm
04:22:59earHurtsyou've been reading too much slashdot
04:23:20jzossnope. I was channeling. ;)
04:23:46jzosseither of ya'll know anything about the sim?
04:24:24earHurtsI know it doesn't work well for me, so i stopped trying to use it.
04:25:28jzossIt's working okay for me, but it looks like it's showing the mp3buf size as: 135M (in the code). The Info display shows 6.6M (maybe a var limit).
04:25:43jzossMaybe it's not limiting ram size to 2M? hmmm
04:26:15Stevie-Owhen I run the optimized Release build
04:26:35jzossYah, I think I must be trashing the memory somewhere.
04:26:40Stevie-OOnly a 9GHz CPU is required to execute full-speed
04:26:52jzoss'Cause when I run a fresh build, it shows 0.928M ram
04:28:25 Quit OliverKlozoff (Read error: 110 (Connection timed out))
04:29:47*Stevie-O notes that the clock almost runs full speed in Release mode
04:29:52earHurtsyou,re trashing mem
04:30:05earHurtsgood work stevie
04:30:27jzossdon't I get a "good work" (for trashin memory)?? ;) j/k
04:31:33earHurtstrashing memory's easi in C
04:31:43jzosstrue dat
04:31:51earHurtstomorrow I get my new 60gb hdd
04:33:12earHurtsso, stevie, when do the rest of us get to play with the emulator?
04:43:24earHurtsjzoss, are you running the win or x sim?
04:43:29jzossx sim
04:43:36jzossI've narrowed it down to:
04:43:45jzossmemset(mp3buf, 0, mp3end-mp3buf)
04:44:37earHurtsassuming end-buf is <= ptrdiff
04:44:50jzossBefore that line, mp3end-mp3buf = 973380. Afterwords, =135770052
04:45:46jzossI know. It's not rocket science. Hmmm.
04:46:01jzossI'm sure it's something trivial. :S
04:47:09earHurtsunless you are overwriting the stack space
04:47:27jzossyah, but I tried shortening the size by a byte and it didn't help
04:48:10earHurtsdivide it by 2
04:48:47 Part Bluechip
04:48:57jzossYah, I was goin there next. /2 works okay
04:50:14earHurtsapparently you are overwrting mp3end
04:52:48earHurtsdon't do that! ;)
04:53:06jzossoh! thx
04:54:49 Join adi|home [0] (~adi|
04:55:41jzossokay, so here's the scoop: memset(mp3buf, 0, mp3end-mp3buf-4) works, but memset(mp3buf, 0, mp3end-mp3buf-3) does not
04:55:57jzossI'm thinking that maybe memset is writing in units of (int) or something instead of units of (char)
04:56:40earHurtsshouldn't be
04:56:54earHurtsmemset works on bytes
04:57:49jzossthe prototype (in firmware/common) shows memset(void*, int, size_t)
04:58:15earHurtsis mp3end < mp3buf?
04:58:44earHurtssize_t, sure. but it's a byte count.
04:59:04earHurtsbut fine, test it.
05:01:00jzossHmmmm... it's actually mp3buf (the ptr value) that's getting overwritten.
05:01:59jzossw/ memset(mp3buf, 0, mp3end-mp3buf-3), before call mp3buf is 0x808d5a0 and after is 0x808d500
05:02:30jzossthat doesn't make any sense. It must be late or something. :S
05:03:12earHurtsis it taking the address of the first arg?
05:04:21earHurtsor is another thread doing it?
05:05:49earHurtsyou're hitting mpebuf.
05:08:05jzossI don't see an mpebuf anywhere...
05:08:53jzossI wonder if I need to be doing a make clean everytime....
05:09:56earHurtsmake clean's probably a good idea
05:13:34 Part jzoss ("Client exiting")
06:03:30***Saving seen data "./dancer.seen"
06:36:28 Join Bluechip [0] (
07:01:47Bluechipthe doc on creating patches is a little contradictory - is there a simple answer?
07:02:17Bluechipplease bear in mind that THIS is all completely new to me
07:03:26Bluechipmaybe there is a preferred doc?
07:03:55Bluechipi think everybody had gone to bed :(
07:07:41 Join Josh_ [0] (
07:20:07 Join thu [0] (
07:24:55 Quit Jet8810 (Read error: 110 (Connection timed out))
07:52:30 Quit earHurts (Remote closed the connection)
07:56:53 Nick dw|gone is now known as dwihno (
08:03:33***Saving seen data "./dancer.seen"
08:12:02 Quit thu ("Client exiting")
08:44:02 Join matsl [0] (
09:03:35 Part Bluechip
09:31:07 Quit hardeep ("BitchX: faster than a speeding bullet, more powerful than a locomotive")
10:03:34***Saving seen data "./dancer.seen"
11:05:21 Join Kuji [0] (
11:05:37 Part Kuji
11:33:47 Join Kuji [0] (
11:33:57Kujilogbot seen linusn
11:34:14>>>"help" by Kuji (
11:34:26>>>"cmd" by Kuji (
11:34:33>>>"seen" used by Kuji ( [snoop prevented]
11:34:44>>>"SEEN" used by Kuji ( [snoop prevented]
11:49:42 Part Kuji
11:55:01 Join LinusN [200] (
12:02:15 Part LinusN
12:03:35***Saving seen data "./dancer.seen"
12:06:27 Join Kuji [0] (
12:10:37 Join olivier [0] (
13:35:39 Quit Josh_ ("Client exiting")
13:59:57 Part Kuji
14:03:37***Saving seen data "./dancer.seen"
14:12:35 Quit olivier ("Aller, tous au ciné !")
14:31:05 Join Bagder [241] (
14:53:53 Join LinusN [200] (
14:54:01Bagderhey Linus
15:00:49dwihnomamma mu!
15:00:53dwihno<−− = kråkan
15:01:19LinusNmed flax!
15:02:28LinusNso, do you guys hear any difference between the soundcheck firmware and the standard? :-)
15:02:47dwihnosoundcheck firmware?
15:02:48BagderI never tried it
15:03:04LinusNdwihno: read the mailing list
15:05:25*dwihno has 2300 mails in the inbox
15:05:30dwihnoIt's summer for crying out loud! :)
15:07:04dwihnoIt's neato! :)
15:07:23dwihnoBagder: Will you and Björn go to Skellefteå this weekend?
15:07:46Bagderwhy would we?
15:07:47dwihnoI guess it's not worth it if not E-type is not around ;)
15:07:54dwihnoBjörn went there last year
15:10:02Bagderyou mean there's some kind of festival or something?
15:11:10dwihnoBjörn was there last summer. He watched E-type.
15:21:40LinusNif i add clip-protection code in rockbox, how would we want it to behave?
15:21:51dwihnoclip protection?
15:21:52dwihnowhat's that?
15:22:28LinusNvolume + loudness + max(bass, treble) must be <= 0dB
15:23:04LinusNit means that if we want lots of loudness, we must lower the volume to avoid clipping
15:23:59dwihnoMake it behave ... well :)
15:24:42 Quit Stevie-O (Read error: 110 (Connection timed out))
15:25:07LinusNso, do we lower the volume automatically when raising the loudness, or do we prevent the setting of high loudness values?
15:28:21LinusNhelpful comment...
15:28:34BagderI really don't know
15:28:43BagderI wouldn't care much for any version myself
15:29:10BagderI think
15:29:16BagderI vote for lowering the volume
15:29:35LinusNthat's what we do on the player
15:29:49dwihnoI always fine tune my settings
15:29:57dwihnoSo limitations are not really an issue.
15:31:14LinusNi will of course make it optional
15:31:52Bagder"We, Mitsubishi Electric Corporation, are planning to use cURL to our software products."
15:31:57Bagder<= from a mail I got
15:32:24LinusN24-tmmarsmyndigheten uses curl
15:32:45Bagderthey do? neato
15:33:15LinusNthey called me a few months ago asking for help
15:33:40Bagderaha ;-)
15:37:13LinusNgotta go, cu
15:37:15 Part LinusN
15:55:23 Join Stevie[FP] [0] (~whatsit2u@
16:03:41***Saving seen data "./dancer.seen"
16:17:46 Nick dwihno is now known as dw|gone (
16:51:20 Join hardeep [0] (1098@
17:15:43 Join mecraw [0] (~mecraw@
17:16:51 Quit Bagder ("")
18:03:43***Saving seen data "./dancer.seen"
18:15:49 Join Zagor [242] (
18:16:14Zagorhi guys
18:23:56Stevie[FP]hey Z
18:27:52*Stevie[FP] pokes Z
18:28:07Zagorhi stevie :)
18:28:16Stevie[FP]cheq eet
18:28:50Stevie[FP]gets better
18:29:04Stevie[FP]it wasn't *supposed* to be in USB mode :D
18:29:12Stevie[FP]but I had a glitch in the ADC emulation
18:29:45Zagorwow, i've never seen that message :)
18:29:59Stevie[FP]One of the things I find interesting about that message
18:30:06Stevie[FP]is that it implies Archos has support for FAT16
18:30:25Zagorthat was exactly my though
18:30:28Stevie[FP]... except FAT16 filesystems can only go up to 2GB
18:30:37Stevie[FP]and the smallest JB I know if is 6GB
18:30:37Zagorexactly. so rather pointless.
18:32:08Stevie[FP]It took me a fscking WEEK to get this far
18:32:10Zagori'm impressed
18:32:25Stevie[FP]to get around the problem I was having
18:32:31Zagorit's a bit more than an 1815 lcd emulator now :-)
18:32:40Stevie[FP]just a bit :D
18:32:42Stevie[FP]you know why it wasn't working?
18:32:58Stevie[FP]cuz the damn thing was trying to record
18:33:10Stevie[FP]as part of the startup, the recorder (or at least the FM) puts the MAS in record mode
18:33:19Stevie[FP]and waits for like 8KB or so of data
18:34:01Stevie[FP]so I had to emulate recording
18:34:15Stevie[FP]which meant emulating PA11, PB15, PB14, IRQ3, IRQ6
18:34:23Stevie[FP]and address 0x04000000
18:34:32Stevie[FP]I cheated on one thing
18:34:46Stevie[FP]I just hold the 'data ready' pin low
18:35:25Stevie[FP]maybe you can help me
18:35:36Stevie[FP]I can't get the device to restart properly after exiting USB mode
18:36:18Stevie[FP]it apparently sets up a watchdog timer
18:37:11Stevie[FP]according to the schematics, the watchdog output pin is connected to the watchdog input pin on the RTC
18:37:23Stevie[FP]and that the IRQ output on the RTC is connected to PA12
18:37:29Stevie[FP]but I don't understand what that does overall
18:38:15Zagori don't know if anyone's every messed with rebooting.
18:38:43Stevie[FP]are the schematics even correct about those pins?
18:39:42Stevie[FP]oh, I also need some empirical information on the Recorder
18:40:02Stevie[FP]runtime values of the AN0-AN7
18:41:20Stevie[FP](when not plugged in and no buttons pressed)
18:43:07 Quit matsl ("Client Exiting")
18:43:10 Join Norrin [0] (
18:43:30Stevie[FP]I see
18:43:35Stevie[FP]it causes a manual reset
18:44:32Zagorit's a pretty safe bet to not trust the schematics :)
18:44:33NorrinHello all. Question: The daily builds change in size even though the changelog reflects no changes. Also, the notes on the homepage reflect changes being made when don't appear in the CVS. Comments?
18:44:56Stevie[FP]cvs is kaput
18:45:11ZagorNorrin: a) the date changes, and compresses differently == different size b) sourceforge CVS is having baaaaad problems, causing erratic daily status display
18:45:53NorrinThanks for the update. Excellent product by the way. Look forward to see what Rockbox comes up with next.
18:46:46Stevie[FP]see, now I need some damn buttons!
18:51:44 Quit Norrin ()
19:42:48 Quit hardeep ("BitchX: better than a penis enlargement!")
19:45:04ricIIStevie[FP] an0 om the rec?
19:46:14Stevie[FP]that one I could have guessed ;)
19:46:48Stevie[FP]who wants to know the trigger thresholds for charging/not charging an FM
19:47:06ricIIgot an old one with 1.17i frimware///
19:51:39 Join Tiberious [0] (
19:52:24TiberiousQuestion, I can no longer turn on my recorder 15 gig, as with the power and usb on it freezes and turns off, without the usb is loads a little bit,then the screen flickers and a strange sound comes from the unit. ANy idea?
19:52:50ricIIdrive dead?
19:53:04TiberiousNo idea,this just happened.
19:53:35Tiberiousjesus im a moron...
19:53:45TiberiousPlugging the power in the wrong hole doesn't help. :P
19:54:44ricIIyou will also have to charge it then for a while..
19:59:21ricIIStevie[FP] also need other an values ?
20:03:14ricIIan2,an4,an5 typically float around 000-002
20:03:44***Saving seen data "./dancer.seen"
20:03:56 Join matsl [0] (
20:04:06ricIIpressing play raises an5 to ~2dc
20:06:53 Join _aLF [0] (
20:24:33 Join hardeep [0] (1098@
20:24:35ricIIthe v1.2 interface uses a different power regulater.
20:29:23 Join Jet8810 [0] (
20:44:52 Join Guest [0] (
20:46:22Guesthelp! I have just dissected my jbr20 to connect it directly to my computer, but the IDE-cable does'nt fit the hard-drive!!
20:48:42hardeepGuest: you need an adapter
20:48:53hardeepGuest: you can find them at most every electronics store
20:50:03GuestOK, thanx....
20:50:07 Quit Guest ("Leaving")
20:59:57ricIIsigh unless 'my computer' is a laptop that isn't verry surprising...
21:02:58 Join mecraw_ [0] (~mecraw@
21:06:46 Quit hardeep (
21:06:46 Quit Tiberious (
21:06:46 Quit mecraw (
21:06:46 Quit MT (
21:08:59NJoinMT [0] (
21:17:05Stevie[FP]wtf is Z doing
21:17:58Zagorgot questions?
21:25:31Stevie[FP]no, I got answers
21:26:06Stevie[FP]as in, when the Archos firmware considers the FM charged
21:26:18Stevie[FP]note the value of AN7
21:27:51Stevie[FP]that's at 100 (decimal)
21:28:11Stevie[FP]so 0x64
21:28:42Stevie[FP]also note that those values are shifted left 6 bits before being fed to the program being run
21:29:04Stevie[FP](which will then shift RIGHT 6 bits..)
21:29:45Stevie[FP]at 101, it still says charging
21:29:59Stevie[FP]once it's <= 100, however, it latches on 'charged'
21:30:43Stevie[FP]until it hits the threshold value of 150
21:32:22Zagorinteresting, but not really useful. it is generally acknowledged that delta-V is the right way to determine charging completed
21:33:27Zagora dip in voltage can be seen just as the battery becomes full
21:33:54Zagorit has to do with increased resistance, leading to higher temperature
21:34:57Stevie[FP]that might work for NiCd or NiMH
21:35:03Stevie[FP]but does it work for Li ion?
21:35:05Zagorright, that's what it's for
21:35:22Zagori don't know much at all about the liion charging cycle
21:35:23Stevie[FP]do keep in mind that the FM uses a lithium ion battery
21:35:38Zagoryes, but it also handles the charging completely in hardware
21:35:38Stevie[FP]li ion batteries don't have much of a 'cycle'
21:35:50Stevie[FP]all you do is keep feeding it current
21:36:24Stevie[FP]the LTC1734 just lowers the output current as the battery becomes more full
21:36:41*Stevie[FP] read the spec sheet a couple of weeks ago
21:36:58Zagorright, but there's no way we can affect any of this from software is it?
21:37:06Zagornot that i'd like to... :-)
21:37:47Stevie[FP]when the battery hits the 'full mark'
21:38:05Stevie[FP]which is decided by the voltage being >= 4.1V or 4.2V (depending on which LTC1734 version used)
21:38:54Stevie[FP]the value read by AN7 will drop to nearly zero
21:39:12Zagorah, right. we currently don't ever show the fm recorder as finished charging
21:39:51Stevie[FP]actually we do (prematurely)
21:40:13Stevie[FP]the battery looks to be at 100% long before it's really completely charging
21:40:14Zagorok. i'm a bit behind on the fm details...
21:40:35Stevie[FP]unless the Recorder has another way of indicating charging vs not charging
21:43:04Zagorno, it has the same problem on the regular recorder. we show the voltage as if supplied by the batteries, which means it indicates full long before actually completed
21:44:08Stevie[FP]does the recorder have some sort of 'we are currently charging' indicator?
21:44:11Zagormike holden is looking into various ways to fix/clarify that
21:44:24ZagorStevie[FP]: yes, it animates the graphic battery symbol
21:44:33Stevie[FP]like the original archos fw?
21:44:42Zagoryes, something like it
21:44:50Stevie[FP]is there a generic 'are we charging' function?
21:45:43Zagorthere's a global 'charge_state' variable
21:46:10Zagor0 = not charging, 1 = charging, 2 = trickle
21:46:33Zagoractually, 2 = top-off, 3 = trickle
21:47:15Zagorthat one should really be an enum...
21:47:50Zagori'm off a bit, bbl
22:03:48***Saving seen data "./dancer.seen"
22:08:30*Stevie[FP] wonders wtf these transistor-type things are
22:22:40Stevie[FP]I regret that I don't understand at all how this damn thing powers on/off
22:33:26 Join LinusN [200] (
22:33:36Stevie[FP]sup Linus
22:33:51LinusNStevie[FP]: can you do me a favour?
22:34:01Stevie[FP]depends on the favour
22:34:11*Stevie[FP] wonders if it'll be 'stop bugging you' :P
22:34:45Stevie[FP]did you know the original archos fw supports fat16?
22:34:51LinusNno, can you see what the orig firmware writes to the direct config registers of the MAS?
22:35:09Stevie[FP]yeah I have that info
22:35:24LinusNStevie[FP]: yeah, i saw that message when i shrunk my jb partition to debug my "disk full" code
22:35:39Stevie[FP]so its' the paritition size?
22:35:50LinusNthat is one factor i think
22:35:53Stevie[FP]what's the minimum partition, you know?
22:36:05LinusNdidn't care that much to find out, actually
22:36:17Stevie[FP]how bout this
22:36:18Stevie[FP]did you know
22:36:27Stevie[FP]the original FW records from the built-in mic on startup?
22:36:44LinusNno i didn't, was surprised when you told us
22:36:48Stevie[FP]that's the problem I've been trying to figure out for a week
22:36:56Stevie[FP]It waits for like 8kb of data
22:37:08LinusNself test, perhaps?
22:37:08ricIIwhy would it do that...
22:37:17Stevie[FP]that's the only thing I can think of
22:37:20Stevie[FP]but it doesn't actaully validate the data
22:37:27LinusNhow could it?
22:37:32Stevie[FP]mp3 frames?
22:37:42LinusNthat, maybe, yes
22:37:43Stevie[FP]I just have it "read" garbage data
22:37:59Stevie[FP]look at this:
22:38:12LinusNwell, if the handshake signals behave ok, there is little chance of garbage data
22:38:39LinusNcharged, yes
22:38:42Stevie[FP]when the value read by AN7 falls to <= 100, it considers charging complete
22:39:02LinusNi guess it only uses a batt level threshold...?
22:39:11Stevie[FP]it's not the battery level
22:39:15LinusNoh, its the fm
22:39:16Stevie[FP]it's the 'external power' level
22:39:34Stevie[FP]it's a way to determine if we're still charging on the FM
22:40:06LinusNlooks like they have added the extpower with the charging current
22:40:24Stevie[FP]oh, nm
22:40:36Stevie[FP]you knew that already, it just isn't mentioned in adc.h
22:40:43*LinusN thinks the emulator is a kickass reverse engineering tool
22:41:00Stevie[FP]it'd be more kickass if it was a little faster :(
22:41:04Stevie[FP]MAS stuff
22:41:08Stevie[FP]coming right up...
22:41:26LinusNnice, i'll go clean up the kitchen while you dig
22:42:38Stevie[FP]actually it's already logged (had to get the damn record mode emulationg working)
22:42:45Stevie[FP]I just need to filter out the writes to PBDR/PBIOR)
22:45:20 Quit Zagor ("Client exiting")
22:45:47LinusNStevie[FP]: if you don't mind, i'd like the complete log (if it isn't toooooo big)
22:45:58Stevie[FP]I can log a LOT of things
22:46:29Stevie[FP]I can even log individual instructions (which I will quickly rack up a few hundred megabytes if I do that)
22:47:22LinusNok, i'd settle with MAS accesses then
22:47:52Stevie[FP]I'll give you all the I2C stuff
22:47:56Stevie[FP]which includes the RTC
22:56:54Stevie[FP]that's the raw log output
22:56:59Stevie[FP]not very pretty
22:57:14Stevie[FP]but it's fairly explicit
22:58:24LinusNjust like i want it
22:58:31Stevie[FP]ok (^:
22:59:31LinusNhow come you write "MAS I2C ..." and then (write RTC 0x13 <- 0x10)?
22:59:46Stevie[FP]because I didn't realize at first that it was a shared bus
23:00:03Stevie[FP]and the LCD is an i2c bus too
23:00:09 Join Tiberious [0] (
23:00:11Stevie[FP]so it was LCD I2C and MAS I2C
23:00:12LinusNno, i believe it is SPI
23:00:33Stevie[FP]that's for the MP3 data transfer
23:00:42LinusNno, no, the LCD is not I2C
23:00:52Stevie[FP]well it's close enough to I2C
23:01:02Stevie[FP]it's like I2C that's output only
23:01:04LinusNI2C has start/stop conditions and the LCD hasn't
23:01:18Stevie[FP]well anyway
23:01:31Stevie[FP]I didn't realize initially that the MAS shared the line with the RTC
23:01:51Stevie[FP](which is why I was *VERY* confused at first when I saw address D0 −− I was like 'wtf? that's not in the MAS docs...')
23:05:46Stevie[FP]what we *really* need is the schematics for that M3Po device >:D
23:06:08LinusNi guess they look pretty much like the archos
23:06:16LinusNthe archos is not a unique design
23:06:58Stevie[FP]but the M3Po supposedly has speed change without pitch change (or something like it that implies that it programs the MAS it has)
23:07:26LinusNwould be nice to steal that MAS code
23:07:34*Stevie[FP] grins
23:07:50Stevie[FP]and if we were to emulate the HW well enough to be able to acquire that code
23:08:03 Join hardeep [0] (1098@
23:09:08LinusNhi hardeep
23:09:17elinenbehello everyone
23:09:22HesNow, that would be stealing, right? 8-)
23:09:32elinenbeanyone here live in or been to Croatia?
23:09:32LinusNHes: naaaaaaah
23:11:59 Quit Jet8810 ("Client exiting")
23:13:07LinusNStevie[FP]: thanks, the log has the info i was looking for
23:13:32Stevie[FP]what info was that?
23:13:52LinusNi wanted to know what was written to I2C address 6A
23:14:00LinusNthe Control Register
23:14:12elinenbeStevie[FP]: how is the emulator coming?
23:14:26Stevie[FP]elinenbe: it's great at emulating an unformatted hard drive
23:15:07ricIII can't wait to try it :)
23:15:19Stevie[FP]you can watch it in action here:
23:16:12ricIIit's a very enlightening tool..
23:16:32LinusNand it's friggin' cool as well
23:16:49Stevie[FP]it's only enlightening if you want to know what it does if it doesn't like the filesystem on the HD :P
23:17:15LinusNor if you want to trace the MAS communication...
23:18:06ricIIwhich could give us some goodies...
23:18:53Stevie[FP]nothing useful
23:19:48ricIII just read a rumor about getting some stuff from the M3Po
23:21:26ricIIhi hardeep, anything new on the dynamic-patch ?
23:22:20 Quit mecraw_ ("Trillian (")
23:22:25hardeepricII: nope, just waiting for it to be approved/rejected now
23:24:05ricIII will keep my fingers crossed than :)
23:24:37 Join [IDC]Dragon [0] (
23:25:07 Part Tiberious
23:25:23[IDC]DragonRemember the UART boot? It's working now.
23:25:53LinusNyou downloaded and executed some code?
23:26:29[IDC]DragonI made a liitle monitor to download, it can read+write memory.
23:27:18[IDC]DragonThe PC controls the momory of the box, in a way.
23:27:28[IDC]Dragonmemory, i mean.
23:28:31[IDC]DragonUsing mem read, I was able to download the flash content (once again).
23:28:45[IDC]DragonNo upload yet ;-)
23:29:29[IDC]DragonI have the "bad" flash, not in circuit programmable.
23:29:54[IDC]DragonA chip is on order, I'm desperately waiting for it.
23:30:53[IDC]DragonThere's one more experiment i'd like to try: downloading the rockbox image using my monitor, then executing it.
23:31:14[IDC]DragonI'm curious what that'll do.
23:31:36[IDC]DragonFirst I need to init the DRAM controller, though.
23:31:48[IDC]DragonStill with me?
23:32:23 Join diddystar5 [0] (
23:32:33diddystar5hey linus
23:33:58diddystar5i've been trying to get cvs to work for about 30 min and it won't
23:35:37Hescool things are happening, it seems.
23:36:31diddystar5sourceforge sayed they should have had it fixed on the 19th
23:37:39LinusNthe sf cvs sux atm
23:37:56*LinusN is in abbreviation hell
23:38:00LinusNat the moment
23:38:40diddystar5i wish i was a better programmer and i could have ssh access
23:38:53LinusNme too
23:39:04LinusN(wish i was a better programmer) :-)
23:40:06[IDC]DragonI made a new hardware mod: the UART boot mod. Requires pulling LCD lines low instead of high. I used a tiny switch.
23:40:08LinusNthe daily builds haven't had a successful cvs checkout for ages
23:40:26LinusN[IDC]Dragon: so you need a hw mpd to download code
23:40:40HesLinusN: I think the hard disk read error handling code is quite ok at the moment, I did some extensive motorcycle driving suit testing during the last couple weeks
23:40:53LinusNHes: nice to hear
23:41:10[IDC]DragonYes, unfortunately. That's just the way the ROM code does it.
23:41:15 Quit matsl ("Client Exiting")
23:41:25diddystar5i also wish i was old enough to drive :) i have 3 more years
23:42:12HesWent to barcelona and back in a couple weeks, I think I had two playback stops while driving
23:42:18Heswhich were not due to dead batteries.
23:42:40[IDC]DragonLinus, I'm thinking of a way to "authorize" flash images, avoiding anybody to flash the bleeding edge build. Do you have any ideas?
23:42:55HesOne recovered after some 10-20 seconds, the other was probably because of no keylock and the pause button got pressed, and before I stopped the idle timer shut it off.
23:44:06LinusN[IDC]Dragon: why?
23:44:23LinusNHes: that sounds good
23:44:37[IDC]DragonBecause you can lock dead your box.
23:44:55LinusNyou can do that with released builds too, no?
23:45:14LinusNjust interrupt the flashing and boom
23:45:17[IDC]DragonThe only way out is a UART boot, requiring that mod and serial mod.
23:45:52LinusNmy point is, there is a serious risk regardless of which version you flash
23:45:58[IDC]DragonReleased like somebody checking it, then publishing it?
23:46:16HesThought it might be my new driving suit has a better pocket 8-)
23:46:25LinusNHes: must be that
23:47:10[IDC]DragonIf the image is OK, I don't see a significant risk.
23:47:32[IDC]DragonSome sanity checks are in order, though.
23:47:57[IDC]DragonWhat's the meaning of the mask value?
23:48:01 Join Jet8810 [0] (
23:48:18LinusNyou mean the HW mask?
23:48:55LinusNi'm not even sure it is a bit mask, but it tells you which version of the hw you are running on
23:48:59ricIIpeople modding there box, sould be aware of the riks when flashing..
23:49:00LinusNso you can adapt
23:49:13LinusNricII: we can flash it without modding
23:49:32[IDC]DragonI'd like to tell Plaer, Recorder, FM apart to avoid the worst.
23:49:39ricIIO I didn't know that..
23:50:11LinusN[IDC]Dragon: you can probably use the hw mask to do that, but there are other ways too
23:50:27 Join tracktheripper [0] (
23:50:35[IDC]DragonLike what?
23:51:03ricIIlots of things changed while I was away..
23:51:46diddystar5have you built a flashed rockbox yet?
23:52:00[IDC]DragonIs it safe to fail if the mask value is different?
23:52:25LinusN[IDC]Dragon: different from what?
23:53:04[IDC]DragonDiffering between the current flash content and the image about to be flashed.
23:53:24LinusNthe image shouldn't have mask should it?
23:53:38[IDC]Dragondiddystar5: not yet, but very close.
23:53:43LinusNit should copy the mask from the current flash
23:54:05ricIIwhy not just flash rolo ?
23:54:51[IDC]DragonNo. I see 2 phases: composing an image from the own box content (to preserve the mask value), and second flashing it.
23:55:06LinusNah, then a mask check is in order
23:55:20[IDC]DragonYes, but is it enough?
23:55:24 Join BoD[] [0] (
23:55:37BoD[]hello !
23:55:53BoD[]so what's new in rockbox world
23:56:00[IDC]DragonWhich/how many mask values are out there?
23:56:05diddystar5hewwo Bod[]
23:57:49tracktheripperhello BoD
23:57:56 Part diddystar5
23:57:58[IDC]DragonFor the flash administration: We would need one responsible person per mask value/hardware with the UART boot mod, who can validate images.
23:58:04tracktheripperim the only Rockboxer in the whole wide world who can't play Sokoban :-(
23:58:40elinenbe[IDC]Dragon: have you been able to successfully flash anything at all?

Previous day | Next day