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 2002-07-25

00:01:17Learevery time. don't know if it is size, or the fact that the last entry begins with "ZZ"...
00:02:30Learhe, played around in the directory browser (to see if I could see it in other directories), but it seems it crashed then...
00:02:38Linushow many files
00:03:23Linus#define MAX_FILES_IN_DIR 200
00:03:47Learcrash again; both times when trying to back up from a directory ("stop" to go to parent)
00:03:56Linus#define MAX_FILES_IN_DIR 200
00:04:25Learlargest should be about 50, the next largets one (that also shows the problem) is about 30 or so.
00:05:26Linushow long is the path+name?
00:06:17Lear43 and 35 respectively, to be exact.
00:06:37Learpath+name to what, the file having the cursor?
00:06:52Linusfor example
00:07:46Lear13 (/Music/ZZ Top)
00:08:46Linusand no dir has more than 200 entries?
00:08:57Linusnot even /Music ?
00:09:12Learnope, not that I'm aware of... /music has 43.
00:11:27Linuswell, i haven't heard of a bug like that before...
00:11:40Linusreally strange
00:13:47 Quit Zagor ("Client Exiting")
00:30:48elinenbewho rocks the house? Linus rocks the house. And when Linus rocks the house, he rocks it all the way down!
00:31:04Linuswow! what did i do?
00:33:43 Quit datazone ("Client Exiting")
00:39:56Learhm... "defined(PREFER_SIZE_OVER_SPEED) || defined(__OPTIMIZE_SIZE__)" isn't true during build, is it?
00:40:50Linusit shouldn't be
00:41:13adi|homethats it.. i give up...
00:41:22Linusadi|home: ???
00:41:23adi|home2.95, 3.03 and 3.04 of gcc
00:41:27adi|homenone of them compile
00:41:34*adi|home beats his head against the wall.
00:41:42Linussame error?
00:41:53Learthought I should make it true, to see if that works better...
00:42:27adi|homeits bitching about my includes
00:42:33adi|homeIn file included from ../../compiler/gcc-3.0.4/gcc/gengenrtl.c:23:
00:42:33adi|home../../compiler/gcc-3.0.4/gcc/system.h:330: conflicting types for `getrlimit'
00:42:33adi|home/usr/include/sys/resource.h:52: previous declaration of `getrlimit'
00:42:33DBUGEnqueued KICK adi|home
00:42:33adi|home../../compiler/gcc-3.0.4/gcc/system.h:341: conflicting types for `setrlimit'
00:42:33adi|home/usr/include/sys/resource.h:72: previous declaration of `setrlimit'
00:42:34***Alert Mode level 1
00:42:34adi|homemake[1]: *** [gengenrtl.o] Error 1
00:42:36adi|homemake[1]: Leaving directory `/home/rhak/build/gcc/gcc'
00:42:38adi|homemake: *** [all-gcc] Error 2
00:43:31Linuswait a minute
00:43:50Linusdo you have the INCLUDE environment variable set?
00:47:17adi|homedoes soemthing outside the norm need to be set?
00:47:39Linusi am a little puzzled about the error
00:47:47CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
00:47:47*adi|home nods
00:48:06adi|homesame here...
00:48:12Linusit seems that it includes files from both your system includes and the includes in your compile tree
00:48:22adi|homethats why Im confused...
00:48:32Linusdo you have the INCLUDE environment variable set?
00:48:49Linusif that is the case, unset it
00:49:00adi|homeits not set...
00:49:08adi|homemaybe I should set it for the binutils dir?
00:49:29 Quit mecraw (Read error: 110 (Connection timed out))
00:49:52Linusso you build binutils, install it, add it to the path, and then build gcc?
00:50:32Linusso weird
00:50:44adi|homei agree..
00:50:48adi|homethus my frustration
00:50:52***Saving seen data "./dancer.seen"
00:50:52adi|homestupid fucking 7.1
00:51:36Linusi built the compiler with my RedHat 7.1
00:51:49adi|homehmmm odd...
00:52:35***Alert Mode OFF
00:53:30Linuswhat prefix did you choose?
00:54:04wodokmlinus: just to know, did you get all those wave files? gotta delete them soon, i'm over disk-quota ;(
00:54:54Linusdelete them
00:55:46wodokmlinus: okay
00:56:02adi|homenope.. it don't like my #includes...
00:56:27adi|homelinus.. can you do me a favor and see if setrlimit is defined in your /usr/include/sys/resource.h?
00:58:02Linusseems so
00:58:58adi|homethen wtf is it complaining about ...
00:59:03*adi|home goes away to play
01:02:53wodokmlinus: seen the waves? it's strange, it's not eally clipping (with the 100% volume) any idea?
01:03:13Linusi have looked at them
01:03:45Linusit looks almost like a 2's complement overflow
01:03:59Linusbut it probably isn't
01:05:24wodokmlinus: so we gotta see how "loud" we can go max, and let that volume be 100% ! ?
01:05:33Linusgood idea
01:05:36wodokmmax without distorion...
01:06:04wodokmi maybe can show that pink, our musician, maybe he knows what's going on...
01:06:45wodokmyeah i'll do that.. how's the processing done? stream into MAS, then into DAC, then the amplifier?
01:07:12wodokmand signal goes to line-out just before it's amplified, directly after DAC?
01:07:46Linusthe amplifier is in the DAC itself
01:09:14wodokmah okay, so it's an othe rdac than i have in the c't mp3player project (that line-out is really good quality - compared to the achos output... )
01:11:33elinenbehas anyone here set up a php based email system? I am trying to set one up for a domain I have for the users of the domain.
01:15:59 Join mecraw [0] (
01:17:11mecrawLinus: still there?
01:17:13wodokmlinus: how does the volume setting work exactly? can the stock f/w be disASMd to see if there's wherre's the diff to your volume setting, maybe there's something else wrong - or it's really just that it's too much for the DAC (maybe these are not
01:17:33wodokmreally allowed values), or not in combination with something else that's set wrong...
01:17:52Linusyou are welcome to disassemble the stock firmware
01:18:16PsycoXulwhat distortion are you getting?
01:18:24Linuswodokm: the problem is that not all models have these problems
01:18:24 Quit mecraw (Client Quit)
01:19:10Linusbut yes, we probably set the volume higher than the hardware allows
01:19:21Linusbut not higher than the data sheets say
01:19:37PsycoXulthere's 2 volume controls right, one for the mas and one for the dac?
01:19:50PsycoXuland as it is rockbox's settings control the mas?
01:20:01PsycoXulwhats the dac set to through all this?
01:20:01Linusno, the dac
01:20:26PsycoXulthe bass/trebble's on the mas though isn't it?
01:20:36Linusthe volume control multiply factor is set to 1.0 in the MAS
01:22:03Linusbass/treble is on the mas
01:24:51wodokmhmmm you can tell me what the registers/ports are for DAC/MAS volume setting, so i could check the disasmd archos for that?
01:25:20Linusoooh. check the data sheets
01:25:51wodokmi mean where they are connected to the cpu, but that's on the h/p i guess..
01:26:11Linusit is an I2C serial connection
01:26:30PsycoXulcould there maybe be an "advanced" settings area that controls things like mas' volume and dac's other options like channel select etc?
01:26:40wodokmsometimes when playing the rockbox skips to the next song just when in the mid of some mo3.. i then got back to that mp3 (with |<) and it then played the full 100% lengt..
01:26:54PsycoXulwodokm: i get that too
01:27:24wodokmah yes that I2C stuff.. therre's all standard stuff int here, cool=9
01:28:18 Join WetFlax [0] (
01:28:47elinenbeLinus: get to sleep!
01:28:58Linusmaybe i should... :-)
01:29:15elinenbeLinus: you are going to get sick!
01:29:29PsycoXuli just can't figure out what would cause the kind of distortion i get, with the bass being any more than several dB lower than the treble... and irrespective of volume
01:29:38wodokmi think archos limited the volume for ear/law whateverr reasons (does someone understand the text on the back in french?), so it's no wonder why we can/could go louder...
01:29:48PsycoXulit sounds like something's being overdriven, but why turning treble up would help i dunno
01:30:48LinusPsycoXul: raising the treble causes the mas driver code to increase the scale factor for compensation
01:31:05Linusto avoid clipping
01:32:05PsycoXulwell then either the default values for that should be tweaked, or maybe a dedicated control for it? or something...
01:32:21Linusstill, the clipping is not in the MAS
01:32:29Linusit's in the DAC
01:33:34wodokmyes, we saw that on the sampled wavews.. maybe it's eally just a not allowed value, even if the specs say it is... who knows.. but it's an interesting clipping, i'll check that out with our musician as soon as i can...
01:41:10Linuswell, time to sleep
01:41:14Linusnight all!
01:41:20wodokmg/n linus
01:41:42 Part Linus
01:49:10 Quit Lear ("ChatZilla 0.8.8 [Mozilla rv:1.1b/20020721]")
01:54:53 Quit wodokm ("Leaving")
02:14:02 Quit WetFlax (Remote closed the connection)
02:15:23 Join WetFlax [0] (
02:17:41 Quit WetFlax (Remote closed the connection)
02:18:08 Join WetFlax [0] (
02:20:03 Quit WetFlax (Client Quit)
02:25:23 Join datazone_ [0] ([NNijq0pN8@
02:40:29 Nick datazone_ is now known as datazone ([NNijq0pN8@
02:42:43 Join MeRWiN [0] (
02:50:53***Saving seen data "./dancer.seen"
03:17:49 Quit MeRWiN ()
03:25:49 Join nsauzede [0] (
03:26:28nsauzedeanybody there ?
03:27:20 Quit nsauzede (Client Quit)
03:48:18geoffegjust us chickens
03:48:31datazoneand the bats
04:50:57***Saving seen data "./dancer.seen"
05:34:22 Join wettoad_ [0] (
05:34:59 Quit datazone ("Client Exiting")
06:42:30 Join Bagder [0] (
06:43:00*Bagder walks in
06:47:04Bagderthings been good in rockbox land the last month?
06:48:12Bagdergood! ;-)
06:48:55PsycoXuli haven't used archos' firmware for a while now
06:49:30BagderI only have some 3500 mails to go throw first, then I'll jump in and get going on rockbox again
06:49:36Bagderjust got back last night
06:51:01***Saving seen data "./dancer.seen"
07:09:41 Part Bagder
07:42:07 Quit wettoad_ (Remote closed the connection)
08:32:31Hesgood morning.
08:37:10 Join WetFlax [0] (
08:39:26 Join WetFlax_ [0] (
08:39:38 Quit WetFlax_ (Read error: 104 (Connection reset by peer))
08:39:41 Quit WetFlax (Client Quit)
08:39:47 Join WetFlax [0] (
08:51:04***Saving seen data "./dancer.seen"
09:23:10 Join Lear [0] (
10:28:52 Join lohap [0] (
10:30:51matp|workis it me, or is the random become less random again?
10:32:38lohapi have a question which GNUSH tool chain should i downbload v0202 or v0203
10:38:12 Quit lohap ()
10:42:29 Join Linus [0] (~linus@
10:42:43HesGood morning, Linus
10:43:50HesHow about applying the AVC patch, I am quite confident it works properly?
10:44:06HesAnd thanks, Tetris works now again. 8-)
10:44:06 Quit ironi_ (Read error: 104 (Connection reset by peer))
10:44:29LinusHes: i forgot! :-)
10:45:06Hesah 8-)
10:45:32HesThe battery stats seems to work too without problems, not that it would be of much use.
10:46:01HesBoth seem to apply cleanly to current CVS code, updated just a couple hours ago.
10:46:51HesAny specific targets to be done before an 1.2 ?
10:47:32 Join ironi__ [0] (~ironi@
10:49:19LinusHes: yes, battery charging
10:49:31HesI could attack that next week
10:49:43Linusi think i can do it too
10:49:45Heshave to work with the asm '02 network quite a lot too
10:50:05Hesjust set up the web proxies
10:50:10Linusi haven't been to a finnish party
10:51:06***Saving seen data "./dancer.seen"
10:51:49Linusi worked a little on the Recorder LCD updates last night
10:52:05Linusthey take a considerable amount of time
10:53:29Linushehe. the dir browser does 3 lcd_update() calls for each cursor movement!
11:01:06HesNo wonder it is a little slow.
11:01:28HesWe don't have the lcd_update_rect() yet? Would be nice for the peak display
11:02:08Linusno, we don't
11:02:43 Join notch [0] (
11:02:54notchHey all,
11:03:09HesI wrote the peak display code earlier but if I made it update fast enough to give a good visual effect, I got playback problems
11:03:22Hesdid it in wps.c
11:03:34Heslike the battery line
11:03:45Linusi don't think it's the lcd_update
11:03:58Linusit could also be the i2c
11:04:13HesHm, of course
11:04:24notchIf I wanted to start writing the routines for the remote control, should I add a source file or append an existing one
11:04:46Linuswhatever you think is best
11:05:12Linuswill you generate button events?
11:05:19notchokay I'll have a look :-) more questions to follow :-)
11:05:21HesI'm off for the weekend now, bye!
11:05:27adi|homeLinus.. you see that suggestion on the sleep timer?
11:05:38adi|homewhat do you think?
11:05:42notchLinus : that seems the easyiest way
11:05:49adi|homei figure we don't have a clock...
11:05:56adi|homebut the idea of a song count is cool...
11:06:01adi|homeie: play 10 songs..
11:06:13Linuswe don't have any code for the clock yet
11:06:27Linusbut it is trivial to add it
11:06:48*adi|home nods
11:07:05Linusand a countdown timer for the player is also trivial
11:07:39adi|homewell.. i was thinking of working on it a bit during next week...
11:07:45adi|homeboth time and song counts
11:33:56Learhrm... default colors in uisimulator are _very_ hard to read, imho
11:34:51Linusi have never used it
11:35:07Learbacklight doesn't kick in, I suspect... looks like that anyway
11:35:45Linusi don't think the uisimulator has a backlight
11:36:12Learit does; tries to change the background color. doesn't work though...
11:36:24Linuscan you fix it?
11:36:43Leardon't know why it doesn't work yet. :)
11:38:55 Join Zagor [0] (
11:39:11Linusyo Zagor!
11:39:15Zagorhi there
11:39:42Zagori was thinking, we need something of a feature freeze soon, so we can get a release out
11:39:51Zagor1.2 is looong overdue
11:40:15Linusand this has struck you *now*?
11:40:44Linusi don't think we need more that the battery charge
11:40:55Zagorno, last week. but I was going away, so it wasn't a good time
11:41:07Linus1) charging
11:41:09Linus2) AVC
11:41:40Linusautomatic volume control. the patch is ready but not merged
11:42:20Zagordoes that work?
11:42:25Linus3) possible volume decrease on player, to attack distortion problems
11:42:26Zagoravc in general, I mean
11:42:49Linusyes, it gives you the nice NRJ "pump" :-)
11:43:06Linustotally useless if you ask me, but many people ask for it
11:43:29Linusand since Archos doesn't have it, it gives us an edge
11:43:55Zagorthere's lots of things Archos doesn't have... :)
11:44:10Zagoryou haven't merged the font-load patch, right?
11:44:26Learwhat is avc anyway? something like replaygain or what?
11:44:31Zagorgood. let's hold it until after 1.2
11:44:32Linusi wanted to give Alex a chance to brush it up
11:44:47LinusLear: what is replaygain? :-)
11:45:10ZagorI was thinking we should enable proportional fonts by defaults. I use it all the time and it's really a great advantage
11:45:22Linusso it works that good?
11:45:48Zagoryeah, I never get any problems with it. I haven't tested the games much, but I don't think there's a problem.
11:45:51Leara way to measure the percieved sound level of a song, and it then suggests a volume change so that all songs get about the same sound level
11:46:00Zagorbreakfast time. see you later
11:46:05 Nick Zagor is now known as Zagor|breakfast (
11:46:05Linuscu Zagor
11:46:34LinusLear: it analyzes the sound level continuously, and adapts in realtime
11:46:53Learanalyzes how?
11:47:20Linuswell, that algorithm is in the MP3 decoder
11:47:43Linusi don't know how it works
11:47:57Linusit is a built-in feature of the MAS
11:48:34LearI see. Doubt it is as advanced as replaygain though, but hopefully more than check the peak value or something like that. :)
11:48:48HesNot much than that. It's a compressor of a sort.
11:49:12HesYou can tune the decay time (after which it will slide the volume up again) to 2s, 4s or 8s
11:49:23Hesquite dumb, but probably useful for some...
11:49:29Hes=> gone
11:50:19Learanyone knows x86 assembler? trying to figure out a minor problem in the win32 uisimulator...
11:50:28Linusah, finally second-exact track info changes...
11:50:45LinusLear: my x86 skills are terribly limited
11:50:53 Join yro [0] (
11:51:25Leartrying to make sense of "mov byte ptr [ebp-7],43h". looks a bit strange to me
12:03:59Learhe, seems like lcd_backlight isn't called anywhere. Shouldn't it be?
12:04:36Linusthat one is obsolete
12:04:41 Join Electrocut [0] (
12:05:10Learanyway, explains why I didn't get any backlight in the simulator, I guess...
12:05:51Learbtw, that browsing bug (enter on last entry rather than first) doesn't appear in simulator, only on my player
12:06:13Linusseems likely
12:27:52Learlinus: almost suspected this, but if I change the name of the last entry from "ZZ top" to "YZZ top", the bug doesn't appear. Seems to be a sorting bug, but if it is the compare function or the actual sorter is another question...
12:28:20Learmaybe that also explains the random crashes I get...
12:30:53Linusis it different if you change the case sensitivity in the settings?
12:33:32 Quit Electrocut ("offile")
12:35:05Learbtw, usb detection isn't included in tree.c, is it?
12:36:04Linusline 502
12:38:03Learstrange, didn't work when I tried it a little while ago; worked now though.
12:38:16Learmaybe I didn't wait long enough or something...
12:38:47 Join naduir [0] (
12:39:04 Part naduir
12:40:25Learwhat the... the ctype.h used here (I think) contains this: ((_ctype_+1)[(unsigned)(c)]&_L). But the ctype array in ctype.c doesn't seem to be made for that...
12:40:32Linuso my god! The proportional font is hideous!
12:41:08 Join Lovechild [0] (
12:41:12Lovechildanyone up ?
12:41:18Linusi'm here
12:41:25Lovechildcool, the man himself
12:42:01LovechildI just got my Recorder 10 in the door, and I decided to load RockBox on it, how safe are the daily builds about now ?
12:42:19Linusabout now....very safe
12:42:33Linusbut don't ask Lear. :-)
12:43:09Lovechildcool... then I just copy the downloaded files ontop the .ajz file on the box right (even though WinXP (yuk) claims the the dailybuild is a text file... )
12:43:18LearI've only mentioned some bugs, not that it is unsafe... :/
12:43:50*Lovechild is looking forward to getting of this hellish Windows machine and play with his rockbox on Linux
12:43:53Learlovechild: keep a copy of the old one though
12:44:06LovechildLear: it's backed up... no worries botha
12:44:36LovechildI wouldn't be so dumb would I.... come to think of it I might
12:45:16LovechildLinus: how's the charging' bug progressing ?
12:45:36Leargah, windows 2000 and a USB keyboard is NOT a good combination... Back to PS/2 I guess...
12:45:53LovechildLear: honestly Windows + anything is a bad idea... :)
12:47:15Learlinus: sort bug fixed, I think. Just add a zero first in the array in ctype.c...
12:49:03Learlinus: but check ctype.h first...
12:49:10LinusLovechild: the charging issue is technically not a bug
12:49:30Linuswe just haven't written the driver to do the chargning
12:49:47Linusbut we are quite close now
12:50:03LovechildLinus: I refrase, how's the Recharging' feature :) - I kinda like that feature because I'm using my recorder as an MP3 player for my stereo
12:50:17Learlinus: so it is actively worked on then? that's good news.
12:50:27LovechildLinus: you're the man
12:50:37Linusi have worked on it for weeks now
12:51:08***Saving seen data "./dancer.seen"
12:51:23Linusi just don't want to rush it, because if there is something that Rockbox doesn't want, it is a reputation as a battery killer
12:51:55Leargood point. though I guess a first version could be conservative (as in charging slowly)...
12:52:11LovechildLinus: Hell, I like my recorder but I would like to offer mine for testing
12:52:45LinusLear: it is not a matter of slow or fast charging
12:52:58Linusthe problem is to know when to stop charging
12:53:24Linusactually, it is quite hard to kill the batteries
12:53:41Linusit is more like wearing them out faster than necessary
12:54:18LinusI can have a charging test out later tonight (Stockholm time)
12:55:22LovechildHere goes nothing....
12:55:30LinusLear: inserting a 0 first in ctype.c is not a good idea
12:55:44Learwhy not?
12:56:17Linusyou will shift all character types
12:56:52LovechildCool I'm a RockBox User now
12:57:19Learnot if the is* macros contains "ctype+1". then you need it to NOT shift the types.
12:57:36LearI tried it. It fixed the sorting bug.
12:57:36Linusaha, now i see
12:57:43Linusgood catch
12:58:09Linusi wonder why it adds 1?
12:58:23Leargood question :)
12:58:31LovechildRockBox is written in C right?.... then since I know quite a bit of C++ it should be no problem for me to read the code... I hope
12:59:06LinusLovechild: the C code in Rockbox is vear easy
12:59:33LinusLear: what is your name?
12:59:50LovechildLinus: yup, I hope so... Downloading it tonight...
13:01:49Lovechildthe song selection feature is a bit strange... sometimes it picks the song below the one the "arrowthingy" points at
13:02:03Linusit does?
13:02:25LovechildLinus: well on my archos anyways.. I'll go see if I can figure it out...
13:02:41Linusdo you have any non-mp3 files?
13:03:05LovechildLinus: not in this dir.
13:03:35Lovechildother than that my recorder holds the lastest Mandy beta and a version of LoTR... :)
13:04:20Lovechildoh I love the lost feature... it's so cool that it filters the non MP3s
13:04:46Lovechilds/lost/list... doh
13:05:01Lovechildmaybe sleep would have been good
13:07:05LovechildI kinda miss the battery indicator though...
13:07:35Linusyeah, that is in the pipe
13:07:44Linusit works for the Player
13:07:58*Lovechild should look into writing a users manual for this stuff...
13:08:16Linusif you are curious about the battery status. go to main menu->debug->i/o ports
13:08:51Lovechilddamn... thanks... I was holding out on debug untill I had read the code
13:09:25Linusdon't push anything but OFF in that menu
13:09:29Lovechildlol it even comes with games.. :)
13:09:42Linusof course
13:10:42Lovechildnow I'll never have to be bored on the bus again
13:11:19Learlinus: Magnus Holmgren
13:15:09Linuscoffee time
13:19:40*Zagor|breakfast sneaks in, downloads the latest build, and resumes packing...
13:19:46 Nick Zagor|breakfast is now known as Zagor|packing (
13:24:44 Quit ironi__ (Read error: 110 (Connection timed out))
13:24:51Learat tree.c:399, shouldn't stuff like dircursor, start and dirlevel be reset?
13:25:36Lovechilddumbass question, RockBox no likes Winamp created .m3u files?
13:26:10Learlovechild: sure does. It took a while before it was supported though. :)
13:26:33Learlovechild: the daily builds, that is. Not 1.1, AFAIK.
13:26:51Lovechildbut I can't seem to select my playlist anywhere.... this sucks (using CVS as of 25-07-02
13:27:57LinusLear: it sure wouldn't hurt to reset them
13:28:01Learjust browse to it and select it, like any other entry in the browser
13:28:16LovechildLear: it's simply not there... and it's on the HD..
13:28:39Linusis it really called .m3u?
13:28:45LovechildLinus: yup
13:29:39Lovechildhrmm deleted the old file and resaved.. and now it's here... problem solved
13:34:47 Quit Lovechild ()
13:36:25Learand at tree.c:485, maybe dircursor and start should be reset or limited (the directory can come to contain fewer entries, making the display empty, AFAICS)
13:37:08Learand same fix at CASE_BUTTON_ON as for tree.c:399...
13:37:23Lears/CASE_BUTTON_ON/case BUTTON_ON:/
13:38:14Linuscheck the latest commit i did
13:39:42Learlooks fine
13:41:21 Join ironi [0] (~ironi@
13:45:47yroHi there
13:47:58yroLinus: I've had the I09CPUAdrE problem this morning (yesterday afternoon cvs)...
13:48:42Linusoh no!
13:48:52Linustrack change?
13:48:54yrojust after powering up, I choose a playlist, no sound, direct freeze...
13:49:17Linusand no way of repeating it?
13:49:22Linusbattery status?
13:49:40yroNo way. But I was in the subway, not the better place for testing...
13:50:13yrobattery status with 2 bar on (JB 20 Player)
13:50:56Linusdid you happen to write down the address?
13:52:08yroNo... There was too many people around... I'll try to reproduce it here, at work, and write everything down
13:52:26Linusdo so. thanks.
13:53:02yroOn another subject, I have a question for the disk writting setting:
13:53:54yroArchos use sect 62, we can use another one, but how to be sure it's not in use with a file ? (or overwritten)
13:54:16yroI'm new to disk-ATA, so the question may be trivial ;-)
13:54:45Linusthe Player models use the ISD-200 chip for USB-ATA
13:55:20Linusthe (windows) drivers can't handle partitionless disks with that driver
13:55:47Linuseach partition must start at a cylinder boundary
13:56:07Linushence, the first partition starts at cylinder 63
13:56:26Linussorry, sector 63
13:56:39yroOk, I see. So, we can use sect 61 for rockbox, isn't it ?
13:56:51Linusthe MBR and the extended partition info is on sector 0 and a few sectors ahead
13:57:02Linusyes, we can probably use 61
13:57:29Linusbut we need to implement a "pending setting write"
13:57:58Linuswe must not try to save the settings immediately, if the music is playing
13:58:23Linuswe should instead wait until the disk spins up to read mp3 data, or a dir read
13:58:32Linusand then save the settings
13:59:15Linuswe could save the settings immediately is the music isn't playing
13:59:26Linusotherwise the disk might never spin up
13:59:52yrowhy ?
14:00:13Linusbecause the user might turn off the unit before doing anything that causes the disk to spin up
14:00:41Linusstill, that can happen even when the music is playing, but is less likely
14:01:01yroOh, yes, not a technical constraint, but user related
14:01:06Linusall in the name of battery lifetime
14:01:27 Join liam__ [0] (
14:01:43Linusand we should of course not save anything if the battery is too low for safe disk writes
14:01:49 Nick liam__ is now known as jedix` (
14:02:15yroNot as simple as I first though :(
14:02:25yroI will try to do it anyway
14:04:20Linusyou can skip the pending stuff at first
14:05:09yroSure, the first step is to write and read the sect.
14:05:25yroThen, the usability...
14:07:23yroI will first port the rtc code to the sect 61
14:08:32Linusyeah, just save the same blob to disk instead of rtc
14:17:42yroBingo !!
14:17:50Linusyou made it?
14:18:07yroI09CPUAdrE error at 657765F7
14:18:34Linusnot very good
14:18:47Linuslooks like a buffer overflow
14:19:08yroPower on -> - key (to go to my playlist) -> Play key -> Shuffling appears, hdd seems to be hold -> I09blah
14:19:35Linusis the playlist long?
14:20:10yroI haven't seen the # of songs in the playlist appearring. Yes, playlist is exactly, wait, 1733 files
14:21:34Linusso it crashes during the shuffle?
14:21:42yroI think so
14:22:26yroUsually, I see 1 then 1733 files. There, I just see shuffling, that's all.
14:22:39yro(I've never saw shuffling before...)
14:25:36Linusbut it doesn't always happen, right?
14:26:14yroNo. I have tried at least 20 power on/off cycle before reproducing it...
14:28:12yrothe randomise code seems correct to me...
14:28:33yroAre you sure it's a buffer overflow ?
14:28:53Linusno, but it seems like it
14:29:06Linus657765F7 isn't a valid address
14:29:11Linuslooks like ASCII
14:29:36yroI try to reproduce it again (to see if the address is the same )
14:30:20Linusgood luck :-)
14:30:54Linusstill, i am pleased that it probably isn't the same bug as the other CPUAdr error
14:31:29yroAn unidentified bug instead ;-)
14:31:44Linusfresh, new bugs, just what i needed!
14:32:36yrobtw even with shuffle mode, the same songs are often the firsts played...
14:33:02Linuseven when reshuffling?
14:34:00yroreshuffling ? I mean, when I power on -> choose my playlist and play it (I have set the shuffle true by default in the code)
14:34:20Linusi mean selecting the same list again
14:35:06yroI was trying to see, and bang, new freeze...
14:35:18yroI04:IllInst at 0000000C
14:35:24yroA new one to me...
14:35:59Linuscan you try it without shuffling?
14:36:13yroSo, I was on and selecting my playlist -> off -> play -> off -> play -> badaboom
14:37:24Linuswe're getting closer
14:38:19yroA precision, last error when I say off, i meant stop, sorry for the mistake
14:40:23yroHum, another comment: I have pressed stop too long, so it's powered off. I turned it on, and forget to set shuffle to off (just right now)
14:40:39yroand I09:CPUAdrE at 657765F7
14:40:55yroIt's exactly the same address
14:41:22Linusare the file paths long?
14:41:31yroAt least, it's stable...
14:43:41yroNot a lot. Around 2 or 3 dir
14:43:51yroI can send you my m3u, if you want
14:44:18Linusplease do, linus at
14:46:11yroAhum... I have accentuated char in filenames... (like )
14:46:19yroIt may be that, no ?
14:46:31Linusit could be anything
14:47:12yroI try to set up a playlist with special chars, to see...
14:49:30yroNo, all playlist runs ok...
14:51:10***Saving seen data "./dancer.seen"
14:53:41yrono. It doesn't seem to be this. I don't know...
14:53:54yroWell, I have to work a little now...
14:54:07 Nick yro is now known as yro|work (
15:33:26 Join Electrocut [0] (
15:46:57 Quit Lear ("ChatZilla 0.8.8 [Mozilla rv:1.1b/20020721]")
16:03:06mbr_Linus: Can you please include Nate Nystroms patch to fix the elapsed time display?
16:04:14mbr_Linus: The patch was on the mailing list yesterday
16:10:25mbr_thanks :)
16:11:15 Join mecraw [0] (
16:18:55 Quit Electrocut ()
16:47:22 Quit yro|work ("ircII EPIC4-1.1.5 -- Are we there yet?")
16:51:14***Saving seen data "./dancer.seen"
17:08:34 Quit notch ()
17:15:18 Join notch [0] (
17:15:33 Join Electrocut [0] (
17:16:16notchanyone know which uP pin is used for the remote data function? Is it 'RCON' on the interface diagram?
17:17:21 Quit Electrocut (Client Quit)
17:18:45Linusit is the rx pin for serial port 1
17:19:17Linuspin 116
17:29:04geoffegso the microprocessor in the jukebox is equivalent to what? 386?
17:33:11Linushard to compare
17:33:33Linusbut it is a ~12MHz 32-bit RISC
17:33:40Linuswithout cache
17:34:02geoffegso 386 sounds about right, how many MIPS?
17:34:51Linusin internal RAM, i would say about 12
17:35:17geoffegi keep wondering about OGG support
17:35:29geoffeglunch time
17:35:31*geoffeg &
17:35:49mecrawLinus: any new "must-have" before i start on my 3 day road trip?
17:36:09Linusnot really
17:37:04Linustake note of the battery life
17:37:20mecrawwill do, i think i'm getting about 6 hours right now
17:37:59Linusplayer or recorder?
17:38:11Linusthen it's bad
17:38:37mecraware you getting more?
17:39:03Linuswith archos firmware i got 8-10 with my 1800mAh batteries
17:39:22Linusas far as i remember, i haven't used it for months
17:40:38mecrawi got 8-10 hours with archos firmware and stock batteries
17:41:18Linuswe still have a long way to go before we reach their battery life...
17:41:33Linusactually, i cant really figure out what more we can do...
17:43:14mecrawdo they use less threads?
17:44:13Linusit doesn't matter to the cpu how many threads it uses
17:44:37Linusbut maybe they go to low-power mode once in a while
17:46:54*mecraw hits the road
17:47:14 Part mecraw
17:57:57Linuswell, time to go home!
17:58:33 Part Linus
18:10:18 Quit notch ()
18:51:17***Saving seen data "./dancer.seen"
18:57:18 Join g003y [0] (
18:57:38g003yhello all.
18:58:40 Quit WetFlax ("I like core dumps")
19:19:13 Part g003y
20:51:19***Saving seen data "./dancer.seen"
21:53:51 Join pyvasene [0] (
22:14:59 Quit pyvasene (Remote closed the connection)
22:21:32 Join Electrocut [0] (
22:25:10 Quit Zagor|packing ("Client Exiting")
22:32:07 Join bmidy [0] (
22:32:14bmidyhi all
22:32:46bmidyI've just installed the gnush for win, does it there a cvs for windows ??
22:38:51 Quit jedix` ("going home")
22:45:06 Quit bmidy ()
22:51:22***Saving seen data "./dancer.seen"
22:51:26 Quit Electrocut (Read error: 110 (Connection timed out))
23:18:43 Join Linus [0] (~linus@

Previous day | Next day