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 2004-04-06

00:01:13amiconnOn Saturday I discovered a rockbox bug :(, for which I filed both the bug report and the fixing patch today :)
00:01:52amiconnI did also complete my recording loop optimization and filed another patch for that :))
00:02:50LinusNi just committed your patch
00:03:17amiconnWhich one?
00:03:30LinusNthe postpone bug
00:04:24LinusNi don't see a recording path in the tracker
00:04:58amiconnHmm, may that have any ill side effect? I didn't observe one, but I do not have such nice equipment available as you (logic analyzer)
00:05:11LinusNhehe, i'll examin ethat
00:05:19LinusNexamine that later this week i hope
00:05:32amiconnRecording: patch #929702
00:05:33LinusNi am on vacation (with my laptop)
00:05:40stevenmoo coder people are here
00:05:49stevenmthere is an uber bug in the current cvs version
00:06:03LinusNstevenm: oh?
00:06:11stevenmfiled bug report, but still. look at the last menu item in General Settings, as well as the DEbug menu
00:06:24stevenmyou will be very unpleasantly surprised
00:06:37stevenmand the credits are gone- replaced by a counter and fragments of the speech menu.
00:07:10stevenmI'm guessing something went askew with the string table, if there is one.. these all seem related bugs
00:08:15amiconnHmm, I did build my personal Rockbox from current cvs and did observe no such problems (although my personal build incorporates various optimizaions that are not (yet) in the official build, but none of them have to do with menus & strings)
00:08:45stevenmis it just me, or is the whole debug menu shifted down 4 lines, and the speech menu gone, and the last option in general settings is blank ?
00:09:08stevenmthis is version 04052004-CVS I'm talking about
00:09:30LinusNwoo, cpuadrerr at 090120c0 :-)
00:09:54stevenmI downloaded file several times, reased from memory entirely, and reflashed, and the unit is behaving, well, erratically
00:10:05stevenmLinusN, hmm ??
00:10:12amiconnLinusN: ???
00:10:26LinusNbuilt from cvs and boom!
00:10:39LinusNi guess i'll have to rename the (old) voice file
00:11:02LinusNnope, already did that...
00:11:12LinusNtime to fire up gdb
00:11:29stevenmso I aint the only one seeing weirdness ?
00:11:51stevenmall right.. I figured this whole unit was dead.
00:12:04stevenmand theres a battery issue also, but I submitted that, and its far less important right now
00:12:21LinusNthe percentage while charging?
00:12:35stevenmno, not that
00:12:41amiconnstevenm, LinusN: on which Archos model(s)? I have recorder v1, no probs with current cvs?!?
00:12:57stevenmamiconn, Recorder V1. I used daily build.
00:13:10LinusNstevenm: 100% does not mean that it's fully charged
00:13:20stevenmLinusN, I mean, it switches to top-off charge
00:13:34amiconnHmm, will try the daily (although not flashing it)
00:13:44stevenmIt goes into top-off charge, but then runs for like 2 minutes on those cells, and dies
00:14:18LinusNare you using the original charger?
00:14:24stevenmLinusN, I am charging with archos firmware right now (yea yea, shame on me, but) and its STILl charging
00:14:33LinusNare you using the original charger?
00:14:48stevenmLinusN, I am using my own- never had the original. feeding it 9V, max charger output is 800mA
00:15:12stevenmit's worked before, with this same charger, worked for a while.. no problems. this started happening suddenl;y
00:15:18LinusNi'm wondering if it sees a negative delta prematurely
00:15:58stevenmand I am now charging by the method of holding F1 and then plugging it in, and it's still going. been charging for 3.5 hrs
00:16:38stevenmwell forget the deltas for now, what's up with the menu options? it's as if they had a mind of their own..
00:16:55LinusNthe original firmware doesn't use the negative delta method as far as i know
00:17:11amiconnI downloaded CVS-040405, put it on my box, rolo'ed - no problems at all! I'm puzzled.
00:18:09amiconnstevenm: Did you update your .lng file along with the firmware?
00:18:27stevenmamiconn, I unzipped the whole zip file
00:18:53amiconnWhich language do you use? (So I can try that as well)
00:19:01stevenmI use English
00:19:43stevenmI flashed, though, if that makes that big a difference
00:21:04LinusNi think i have a stack overflow
00:22:08amiconnBingo! The english.lng file somehow produces an overflow. I usually use german and have no probs like described. If I switch to english, I get them as well. They stay even if I switch back to german until re-rolo.
00:24:07amiconnIt also occurs with my personal build if I switch to english.
00:25:44amiconnBtw: There is still the bug that if you switch languages from the menu, the menu texts are garbled completely until you leave and re-enter the menu.
00:27:32LinusNstill? i don't even know about it
00:27:34stevenmamiconn, I confirm. Selected French, roloed, works fine. Selected English, and it died.
00:28:07 Quit matsl (Remote closed the connection)
00:28:10amiconnLinusN: bug report #918063 of 2004-03-17
00:31:43 Join matsl [0] (
00:37:18 Quit matsl (Remote closed the connection)
00:39:18LinusNi did "make clean", and the problem went away
00:41:36LinusNah, now i see it
00:42:54 Join AciD [0] (
00:45:42LinusNstevenm: a workaround is to reset the settings, making it use the internal string table instead of english.lng
00:45:53LinusNor delete english.lng
00:46:21LinusNbinlang hasn't been updated for the new string lable format
00:46:27stevenmthanks, I'll do that right now.. do you suppose there's gonna be a fix within the next few days?
00:46:39LinusNsure, give me a few more minutes
00:46:59 Join scott666_ [0] (
00:47:13stevenmoh and also, for some reason, if charging from the original firmware charge screen, pressing On will start Rockbox, then give ATA error -31
00:47:48stevenmyou have to disconect charger, hold Off, then press On for it to work properly
00:47:52 Quit scott666 (Read error: 104 (Connection reset by peer))
00:50:38stevenmand theres htis other annoying thing I've been meaning to bring up
00:51:20stevenmthe buttons aren't always read right. I press one button, and it reads it as another. I assume this has something to do with the ADC it uses for the buttons.. I spotted a bunch of resistors there
00:53:12LinusNgo to debug->view i/o ports
00:54:02LinusNpress F1, F2, F3 and Up and tell me the AN4 readings
00:54:20LinusNthen press Left, Play, Right and Down and tell me the AN5 readings
00:54:43stevenmF1 - bouncing from 18F to 191
00:55:06stevenmF2 - 262 - 264
00:55:22stevenmF3 - 3Fe - 3FF
00:55:38stevenmUp - 331 - 335
00:55:52stevenmLeft - 262-265
00:56:13stevenmPlay - 32F - 332
00:56:26stevenmRight- 18F - 192
00:56:41stevenmDown- 3FE - 3FF
00:56:50LinusNno problem there
00:57:25LinusNnow try to jiggle the buttons a little, or apply some force, and see if you can make it produce a different value for a button
00:57:25stevenmso.. what do you think is causing it?
00:57:55LinusNi think you have a hardware problem
00:58:27 Join diddystar5 [0] (
00:58:33stevenmI tried, Down is always 3FF
00:58:43stevenmand it most ofen mistakes Down for Right
00:59:10stevenmI guess maybe my pads are worn- I need to apply more force, or something
00:59:11scott666_i have that problem too
00:59:16 Nick scott666_ is now known as scott666 (
00:59:42LinusNtry to press it very lightly, or very fast
01:00:05stevenmLinusN, does the button driver look for one nonzero value, or does it sample several times and pick the most commonly occuring value, or what?
01:00:47LinusNit takes several samples
01:01:18scott666my debug values are completely different though
01:01:29LinusNor rather, it accepts the button if it reads the same value twice in a row
01:01:38LinusNscott666: gimme those
01:02:03scott666F1: 133-135
01:02:29 Join Nibbler [0] (
01:02:36scott666F2 1DF-1CF
01:02:59scott666F3: 309-30D
01:03:15scott666up: 26D-26F
01:03:52scott666play: 132-12F
01:04:09scott666right: 26B-26E
01:04:12LinusNscott666: do you have an fm recorder?
01:04:29LinusNthat explains the different levels
01:04:38scott666ok then
01:05:19LinusNyou have problems with f2?
01:05:43scott666down and right and up and f2 i think
01:06:05scott666i dont actually _use_ my fmr much lately as the headphone jack is broken
01:06:11scott666but i remember having those problems
01:06:41LinusNi did a survey long ago to find the best thresholds for the plain recorder, but we never did that for the fm
01:07:19scott666that could explain it
01:07:43LinusNyour readings look good, though
01:07:57scott666i dunno, its a strage bug
01:08:22LinusNyou press down and it is interpreted as right?
01:08:32scott666some of the time
01:08:36scott666its fairly rare
01:08:47LinusNlike when you scroll fast?
01:08:50scott666but its happened enough for me to know im not insane
01:09:30LinusNi think i know what might be the problem
01:09:46LinusNbut i'm not sure how to fix it in the firmware
01:10:06LinusNi think a very brief press gives a lower reading
01:10:13amiconnLinusN: my grayscale framework for plugins takes shape...
01:10:31LinusNgrayscale framework?
01:10:44LinusNfor generic grayscale graphics?
01:10:52scott666amiconn: yay
01:10:57amiconnYup, for use e.g. in games etc.
01:11:26diddystar5i have no dev enviroment atm :(
01:11:32diddystar5(or linux)
01:11:32stevenmLinusN, the way to fix the buttons would have been to put a TRANSISTOR on every button
01:11:57stevenmbut oh well
01:12:18amiconnBasically this will provide an additional framebuffer for grayscale (within the plugin buffer limits) that can be displayed as overlay.
01:13:07stevenmamiconn, can we get a nice grayscale breakout game?
01:13:14stevenma la ipod
01:13:35amiconnIt can even be used while the music is playing. The drawback is that it provides a fixed number of grayscales, no continuous rendering like in the video plugin by Jörg.
01:15:11amiconnThe number of available grayscales will be dynamic, the smaller the overlay gets, the more grayscales will be possible. If there are too many grayscales, it will probably flicker a lot. I'm just testing that.
01:20:01LinusNi have now committed a fix for the language file corruption
01:21:11diddystar5LinusN, how do i get raw mp3 data to use the mp3_xx functions with?
01:21:25diddystar5(i'm updating my battery low bepp patch with that)
01:22:00LinusNfirst of all you need an mp3-encoded beep, do you have that?
01:22:27diddystar5LinusN, no, but i will make on just a minute
01:23:41LinusNthen you will want to bitswap it and produce a nice C-array with the data
01:23:53LinusNand there are no published tools for that
01:23:57diddystar5how do i bitswap it?
01:24:33diddystar5how did you do it?
01:24:39diddystar5did you write a prog?
01:24:59LinusNi haven't, but if i did, i would write a prog to do it
01:25:25diddystar5ok, then how do it anyway possible?
01:25:41LinusNplease rephrase that
01:25:52amiconndiddystar5: If you're looking for a bitswap in C, look at the source of Jörg's rvf_mux
01:26:13diddystar5amiconn, thanks
01:27:02diddystar5i didnt have goldwave installed to make the sound.... will take me a few mintes to download
01:31:33diddystar5its a cpp, guess ill have to go to the library and get g++ along with my cygwin
01:31:42diddystar5or i could install linux....
01:33:38LinusNrename it to .c and you'll be fine
01:34:41diddystar5yeah, the only c++ ism's ther are i wont need
01:35:27 Quit diddystar5 (Read error: 54 (Connection reset by peer))
01:39:01amiconnAhh, my grayscales look really good now. I was able to get rid of that dreaded moire pattern.
01:41:00amiconnLinusN: Did you find my recording patch in the meantime?
01:43:40LinusNwill try it soon, just not today
01:43:46amiconnI would be _very_ interested if you could manage to do some logic analyzing with my new loop.
01:43:52LinusNwill do
01:45:02LinusNamiconn: do you care a lot about the contents of your jukebox?
01:45:23amiconnHow do you mean that?
01:45:26LinusNi mean, are you willing to reformat it?
01:45:35***Saving seen data "./dancer.seen"
01:46:12amiconnCould do that, since it has USB2, so transfer of the data should not take that long. But why should I reformat?
01:47:02LinusNi'd like you to reformat with a very small cluster size, to see how the seek times affect the recording
01:47:31amiconn(Most of the data is already backed up on my PC, since I tested my writing with it and did'nt want to put all data at risk)
01:48:01LinusNi always have a complete mirror of my jukebox on my pc
01:49:22amiconnWth has the seek time to do with cluster size? Seek time only depends on the disk model, or do I miss a point here?
01:51:00 Join joshn [0] (~joshn@
01:51:51 Quit joshn (Remote closed the connection)
01:53:41LinusNthe file seek needs to wade through the FAT chain to find the position in the file
01:54:03LinusNsmall clusters == more FAT entries to search
01:54:47amiconnGrr, I think I'm silly. Of course you spoke about the file seek and not the disk seek.
01:55:07LinusNhave a Red Bull
01:55:42amiconnHmm, not really. I happen to be on vacation also.
01:56:43amiconnWhy should the file seek affect recording? Is there a seek operation executed _while_ recording?
01:57:30LinusNit appends to the file, and i think we still do append by seeking to the end
01:57:39 Join joshn [0] (~joshn@
02:01:03amiconnAs far as I call tell from the source, the file is held open between consecutive writes, so no seek here.
02:01:35LinusNthe recording code does fsync(), which calls flush_cache(), which performs a seek()
02:02:21LinusNthis is done to make sure that the file is updated on disk in case of a crash
02:02:53amiconnAhh, I get your point.
02:03:26LinusNi suspect that the seek takes too long when the file grows, and the recording buffer overflows
02:03:39LinusN(nice rhyme)
02:04:31amiconnDid you get overflows? I didn't manage to get incomplete frames (except the last one) even in my 2:09 h test recording.
02:05:31LinusNi have never had any overflows, or corrupt files
02:05:49LinusNbut other people have, that's why i want you to test my theory
02:06:19amiconnStill I have to figure out how to format with a different cluster size (on Windows).
02:06:46LinusNi think you can select that in the formatting dialog box...
02:07:43amiconnTrue for NTFS, but for FAT i only get "default size" which is not very helpful.
02:08:07LinusNsilly windows
02:08:58LinusNi think win98 lets you choose
02:15:05amiconnI _can_ choose the cluster size _if_ I do the formatting from the device manager in WinXP :)
02:18:01amiconnBut.. Windows complains that it cannot format a 4 GB test volume with 512 byte/cluster. 1024 byte/cluster works. So I will have to choose 8192 bytes/cluster or higher for my 20 GB box.
02:19:05amiconnThis is strange, shouldn't FAT32 allow for 2^32 clusters equalling 2 TB with 512 bytes/cluster?
02:21:36LinusN4 bits are used for other stuff
02:22:09amiconnThis would still be good for 128 GB with 512 bytes/cluster...
02:23:03LinusNthere may be more limits
02:23:48LinusNlike a maximum fat size for example
02:25:48amiconnSome days ago I did take a look at the beginning of my jukebox hd with a hex editor and found my fat being ~10 MB large. This implies that I'm currently using 16 KB clusters.
02:27:06amiconn(I tried to find a way of getting only the root dir extracted, to put together a small program that does this. This could help in getting the cause for the file system corruption, since apparently the root directory is affected)
02:31:44LinusNnot necessarily, some people report that .rockbox is affected
02:53:34amiconnI guess I should try to get some sleep now. Can do more tomorrow.
02:53:34 Quit Nibbler (Read error: 54 (Connection reset by peer))
02:54:45amiconnLinus, please tell me if you think it's worth going from 16K to 8K clusters and test recording. Or would a _very_ long recording also help tracking down the seek problem?
02:55:06amiconnIn this case I would leave my Archos recording overnight.
02:56:26amiconnBtw: I have write optimization compiled in (of course).
02:57:34-->"*** addictirc!logbot_: sYk0_frEak kicked #guitar sYk0_frEak" received from Rhino (zachary@
02:58:03LinusNi have 4k clusters and have no problems
02:59:16LinusNhaven't done any _huge_ recordings lately, though
02:59:22amiconnWhat disk size do you have?
03:00:58amiconnWhy would that &§%$ Windows not let me format 20 GB with less than 8 K/cluster then?
03:01:42LinusNb/c it is silly?
03:01:44amiconnI think I will have to download knoppix and give it a try on my laptop.
03:02:44MTiirc if you used 512 byte sectors, you would quickly run out of entries in the FAT table (i could be very wrong)
03:03:16amiconn(This will have to wait till tomorrow, though. Will do a "overnight-recording" with my current cluster size a a first test).
03:03:30MTi tend to sleep through too many of my lectures :/
03:04:53amiconnNite all.
03:06:32 Part amiconn
03:12:17 Quit AciD (Read error: 54 (Connection reset by peer))
03:42:26ILuvithi guys
03:42:50ILuvitAnyone notice when you go from the fm radio to play a mp3 the Peak Meter is still tracking the fm radio, and if you stop the radio then go to play a mp3 file the Peak Meter doesn't work ?… you got to reboot first
03:45:37***Saving seen data "./dancer.seen"
03:50:25 Quit joshn (Remote closed the connection)
03:50:59 Join joshn [0] (~joshn@
03:51:14 Join AciD [0] (
03:51:34 Quit joshn (Remote closed the connection)
03:51:54 Join joshn [0] (~joshn@
04:01:49LinusNILuvit: oops, i'll have a look. please file a report
04:02:45 Join diddystar5 [0] (
04:07:47 Join elinenbe [0] (
04:07:59elinenbelet's port rockbox to the ipod!
04:09:16diddystar5go4it :)
04:11:23 Quit joshn (Remote closed the connection)
04:15:10diddystar5i would never get a ipod
04:15:23diddystar5infact i dont even look for just mp3 players anymore
04:15:36diddystar5i look for ones with video etc, like the av320
04:15:57 Join joshn [0] (~joshn@
04:16:28 Join [1]c0utta [0] (
04:20:06LinusNILuvit: fixed in cvs
04:20:25LinusNtime to sleep, nite guys
04:20:29 Part LinusN
04:20:41diddystar5LinusN, what do you think about my simplified vu meter patch?
04:31:50 Join Nibbler [0] (
04:34:01 Quit c0utta (Read error: 110 (Connection timed out))
04:34:05 Nick [1]c0utta is now known as c0utta (
04:45:29 Quit AciD (Read error: 104 (Connection reset by peer))
04:46:45 Join sagegoku [0] (
04:54:47 Quit diddystar5 ("Leaving")
04:57:31 Quit scott666 ("i'll be back...eventually...")
05:45:41***Saving seen data "./dancer.seen"
06:04:19 Quit sagegoku (Read error: 54 (Connection reset by peer))
06:04:20 Quit Nibbler (Read error: 54 (Connection reset by peer))
06:09:05 Quit stevenm ("Leaving")
07:21:09 Join Nibbler [0] (
07:22:27 Quit ILuvit (Read error: 60 (Operation timed out))
07:36:05 Join midknight2k3 [0] (
07:45:45***Saving seen data "./dancer.seen"
08:12:27 Quit Nibbler (Read error: 54 (Connection reset by peer))
08:24:31 Quit midknight2k3 (Remote closed the connection)
08:26:25 Join midknight2k3 [0] (
08:32:59 Quit midknight2k3 (Read error: 104 (Connection reset by peer))
08:35:04 Join midknight2k3 [0] (
09:01:48 Join monkey666 [0] (
09:07:49monkey666maybe somebody here knows if it is possible to bookmark text files?
09:08:06monkey666i checked the manual and site and came up with nothing
09:08:32 Join matsl [0] (
09:09:10midknight2k3bookmark text files?
09:11:37monkey666darn! thanks tho midknight2k3
09:12:28monkey666guess i'll to convert wolves of calla to audio
09:12:29midknight2k3no problem
09:12:48midknight2k3although i suppose it could be done rather easily
09:14:09monkey666i'm afraid to ask how
09:14:25midknight2k3lol i mean by adding it as a feature
09:15:52midknight2k3i could give it a shot tomorrow
09:15:59monkey666yah i would prefer reading it to hearing MS Mary speak it
09:16:03midknight2k3but i'm not promising anything :)
09:16:22monkey666well i would appreciate it either at any cost
09:16:31monkey666i assume others would too
09:16:44midknight2k3i'll try
09:16:50 Nick midknight2k3 is now known as midk (
09:16:54monkey666not that rockbox isnt great already thanks to all you coders
09:17:46monkey666thanks i think dark tower 5 can wait one more night
09:18:14CtcpIgnored 2 channel CTCP requests in 29 minutes and 35 seconds at the last flood
09:18:14*midk feels the pressure
09:18:18midki think i can do it
09:18:22midki think i can i think i can
09:18:35monkey666no no i would just break it up into smaller text files if not
09:18:59monkey666no biggee just a thing of me being extremely lazy
09:19:41monkey666i appreciate the effort tho
09:20:43monkey666night :)
09:20:50 Quit monkey666 ()
09:33:52 Quit matsl (
09:33:52 Quit Ka__ (
09:33:52 Quit midk (
09:33:52 Quit elinenbe (
09:33:52 Quit adi|home (
09:33:52 Quit Ka_ (
09:35:38NJoinmatsl [0] (
09:36:46NJoinmidk [0] (
09:36:46NJoinelinenbe [0] (
09:36:46NJoinKa_ [0] (
09:36:46NJoinKa__ [0] (~tkirk@
09:36:46NJoinadi|home [0] (~adi|
09:41:49 Join Nibbler [0] (
09:45:48***Saving seen data "./dancer.seen"
10:50:57 Quit midk ("yo yo yo yo yo")
11:08:18 Quit joshn (Read error: 110 (Connection timed out))
11:45:49***Saving seen data "./dancer.seen"
12:11:38 Quit Nibbler (Read error: 54 (Connection reset by peer))
12:36:53 Join track [0] (
12:36:56 Quit track (Read error: 54 (Connection reset by peer))
13:17:42 Join AciD [0] (
13:38:21 Join amiconn [0] (
13:45:53***Saving seen data "./dancer.seen"
13:55:06 Join Nibbler [0] (
14:13:22 Quit AciD (Read error: 54 (Connection reset by peer))
14:13:22 Quit Nibbler (Read error: 54 (Connection reset by peer))
14:31:20 Join Sixer [0] (
14:32:15SixerI've had a problem with RockBoxX since 2.1, where when I play an mp3, I get "Error loading playlist" (Just "Error" under 2.2). Since 2.2, it even refuses to play the mp3 after that. Have got a Jukebox 6000.
14:32:17SixerAny clues?
14:32:22Sixer(FAQ does not seem to cover this)
14:32:42SixerThe HD is accessible fine, no probs there
14:32:58SixerI have a feeling that the jukebox is looking for a playlist I've never created
14:32:59c0uttaare you playing a plylist or an individual mp3 ?
14:33:10SixerAn individual mp3, I don't have any playlists
14:33:42c0uttabehind the scenes you are actually playing a playlist
14:34:00c0uttayou have a .rockbox folder ?
14:34:00SixerI've done an rm -rf .rockbox, to no avail
14:34:42c0uttarm - that's linux huh ?
14:35:08c0uttaahh, anyway..
14:35:11c0uttayou have a .rockbox folder ?
14:35:18Sixerrm -rf == remove recursively by force
14:35:23SixerSo nope I have deleted that folder
14:35:35c0uttayou must have .rockbox
14:35:47c0uttaotherwise bad things might happen
14:36:16c0uttareinstall the .rockbox folder from the ZIP file
14:36:34SixerIs there originally any content in that folder?
14:36:53c0uttayes. stuff like fonts etc
14:37:18c0uttawhen you play an mp3 it'll try and create a file in .rockbox to keep track of the playlist
14:37:34c0uttano .rockbox = no play
14:39:58SixerYou've been very helpful, cheers
14:40:24c0uttaworking ?
14:41:14c0uttagood o
14:44:14Sixerheh funny, my jukebox6000 appears to have a newer firmware than mentioned on the archos website
14:44:36c0uttahave you flashed ?
14:44:49 Join guest [0] (
14:44:54c0uttahmm - 6000 is old now
14:44:58Sixersure is
14:45:01*c0utta has a 6000
14:45:14SixerTheir site says 5.07a, my jukebox says 5.08 on boot
14:45:16c0uttanow a 60000
14:45:44Sixerwhoa, they did an iTunes plugin
14:46:11c0uttai think archos stopped updating their site for the 6000
14:46:30c0uttayou mac guys love itunes, don't you ?
14:46:35 Part guest
14:46:38SixerYeah, sure do
14:46:51c0uttathought you'd have an ipod
14:47:07SixerWell I got this thing as a xmas present from where I work
14:47:18c0uttagood present!
14:47:22Sixeryeh definitely
14:47:27Sixergot an xbox this year ;)
14:47:35Sixer( )
14:51:25c0uttagood network
14:51:30Sixerheh cheers
14:55:21c0uttainternet connection
14:56:59Sixeroh, yah
14:58:02Sixer8192 later this year
14:58:57Sixerwhoa, Archos' iTunes plugin for this is bloody excellent
15:02:30c0uttawhat page is this on ?
15:03:04Sixerbottom of
15:04:43 Join cjnr11 [0] (
15:06:32 Part cjnr11
15:17:59 Join mecraw_ [0] (~mecraw@
15:18:24 Part mecraw_
15:20:45 Join sethians [0] (~jirc@
15:21:36sethiansHello, my HD has die
15:22:25sethianswhat HD can I put in my jukebox (recorder
15:22:42c0uttaany 9.5mm laptop drive
15:23:27sethiansmmm and wha about RPN ?
15:39:22 Nick c0utta is now known as c0utta{zzZZ} (
15:40:11silencersethians: there's no requirement i think
15:43:43 Part amiconn
15:43:44 Quit sethians (Read error: 104 (Connection reset by peer))
15:45:57***Saving seen data "./dancer.seen"
15:55:18 Join Nibbler [0] (
16:13:29 Join AciD [0] (
16:21:56 Join methangas [0] (
16:36:08 Join amiconn [0] (
17:34:07 Quit Nibbler (Read error: 54 (Connection reset by peer))
17:34:22 Join mecraw__ [0] (~mecraw@
17:36:24 Join mecraw___ [0] (~mecraw@
17:36:44 Part mecraw___
17:46:01***Saving seen data "./dancer.seen"
17:52:03 Join pfavr [0] (
17:53:19 Quit mecraw__ (Read error: 110 (Connection timed out))
18:14:49 Join stevenm [0] (
18:26:17 Join KnivesOut [0] (
18:26:31KnivesOuthey guys
18:27:06KnivesOuti just read about the NEC battery that charges in 30 seconds
18:27:38KnivesOuti'm anxious to see this in my archos
18:29:41 Quit KnivesOut (Client Quit)
18:39:43 Join Dogger [0] (
19:17:33 Join Nibbler [0] (
19:29:04 Join mattzz [0] (
19:42:10 Join midknight2k3 [0] (
19:46:04***Saving seen data "./dancer.seen"
19:59:36 Quit stevenm (Remote closed the connection)
20:00:07 Quit MT ("changing servers")
20:12:36 Quit mattzz ("Client exiting")
20:13:19 Join _aLF [0] (
20:18:29 Quit matsl (Remote closed the connection)
20:20:55 Join cjnr11 [0] (
20:20:57 Part cjnr11
20:25:00 Quit pfavr (Read error: 110 (Connection timed out))
20:34:00 Join cjnr11 [0] (
20:34:04 Part cjnr11
21:27:35 Part Sixer
21:44:35 Join scott666 [0] (
21:46:05***Saving seen data "./dancer.seen"
21:52:39 Join Frosh [0] (
21:54:19FroshOn the Archos website it says that the USB cable used with their players is specific to the device. Is it true that I can't just use any old USB cable?
21:54:55Nibbleri think ne cable should be fine....
21:55:06Nibbler(as long as its usb 2.0 for usb 2.0 devices...)
21:55:17Nibblerthats why it is "usb"
21:55:27FroshThat's what I was thinking, too. UNIVERSAL serial bus, you know?
21:56:21FroshI'd much rather get a spare 99-cent generic cable than purchase one of those overpriced "proprietary" ones from Archos. :P
21:58:39midknight2k3Frosh: model?
21:59:57midknight2k3any will work
22:00:27FroshGood. Thanks. :)
22:03:55 Quit Frosh ("Yeah, it's like that.")
22:08:36 Quit midknight2k3 (Read error: 104 (Connection reset by peer))
22:11:17 Join midknight2k3 [0] (
22:12:56 Join ricII [0] (
22:12:57 Quit Dogger (Read error: 104 (Connection reset by peer))
22:13:25 Join Dogger [0] (
22:20:29 Quit Dogger (Read error: 104 (Connection reset by peer))
22:20:41 Join Dogger [0] (
22:22:02 Quit Dogger (Read error: 54 (Connection reset by peer))
22:22:38 Join Dogger [0] (
22:33:28 Quit AciD (Read error: 104 (Connection reset by peer))
22:33:46 Join AciD [0] (
22:34:31 Quit methangas (" Like's GUI? Then try HydraIRC -> <-")
22:52:54 Join [IDC]Dragon [0] (
22:53:09amiconnhi Jörg!
22:53:39[IDC]DragonGood evening Jens, for a short one today.
22:53:58[IDC]Dragon(Have to leave again soon)
22:54:06amiconnI'm almost done with my grayscale framework :-))
22:54:15[IDC]Dragonhow does it work?
22:54:48amiconn(only one missing graphic primitive) Looks nice now, although it was harder than I thought at first.
22:54:48 Quit Dogger (Read error: 104 (Connection reset by peer))
22:55:12 Join Dogger [0] (
22:55:20[IDC]Dragonpurely plugin, or some IRAM code for dithering?
22:55:43midknight2k3idc: by the way, what sort of data does lcd_blit accept?
22:56:53midknight2k3proprietary format
22:56:53amiconn[IDC]Dragon: 100% pure plugin.
22:56:53[IDC]Dragonthe physicaldisplay format
22:56:57amiconnThe biggest problem was, that if you have a sequence of frames that are cycled, every pixel is represented by a bit sequence within these.
22:57:14midknight2k3idc: the same type of data that comes out of bmp2rb?
22:57:23[IDC]Dragonmidknight2k3: yes
22:57:33midknight2k3but then how would it know what was gray and not?
22:57:35amiconnIf the "phase" of these sequences is eual for all pixels, you get evil flickering.
22:57:47midknight2k3or is that integrated into the output too?
22:58:04[IDC]Dragonmidknight2k3: this is pure b/w
22:58:04 Quit Dogger (Read error: 104 (Connection reset by peer))
22:58:11midknight2k3OH right
22:58:19midknight2k3flickers between them
22:58:30[IDC]Dragonamiconn: you have to seed with a random phase
22:58:32*midknight2k3 is somewhat confused
22:58:34amiconnIf you shift the phases by a regular pattern, you get moire.
22:58:46midknight2k3you can flicker between two frames right?
22:58:56 Join Dogger [0] (
22:58:58[IDC]Dragonmidknight2k3: or more, yes
22:59:17amiconnSo I have to set _every_ pixel individually with a random phase. This is what I do now.
22:59:42[IDC]Dragonamiconn: I had the same pronlem with the stills
22:59:57[IDC]Dragonlooks bumpy on the boundary
23:00:27midknight2k3idc: so then, blit accepts one array of data. how would you give it two frames?
23:00:57[IDC]Dragonmidknight2k3: you have to swich them
23:01:07amiconnmidknight2k3: If you want grayscale for a plugin, wait for my grayscale framework. I will post a plugin to the patch tracker which will contain that and some simple example calls.
23:01:08midknight2k3so keep changing what's in the array?
23:01:22midknight2k3amiconn: grest, thanks
23:01:24midknight2k3it's a possibilty
23:01:44[IDC]Dragonamiconn: do you switch pixels as often as possible?
23:01:54amiconnThe framework allows 2..33 scales of gray
23:02:13amiconn[IDC]Dragon: I switch frames at 67 Hz.
23:02:31[IDC]Dragone.g. 50% gray is not 4 times black, then 4 times white
23:03:16[IDC]Dragonbut alternating every time
23:04:31amiconnYes, 50% gray alternates every time. The patterns are calculated in a way that the bit values that is needed less does never appear twice in a row.
23:05:26[IDC]DragonI have an inteper implementationof my halftoning algorithm, as I've "advertized" to you before.
23:05:46amiconnE.g 25% gray = 11101110..., 50% gray = 1010..., 75% gray = 10001000... etc.
23:05:57midknight2k3is it ever going to be possible, do you know, to be able to simply play bitmap files?
23:06:08[IDC]Dragonhow do you calculate these?
23:06:36[IDC]Dragonmidknight2k3: I have a half-done JPEG viewer
23:06:49midknight2k3that could work fine! :)
23:06:51amiconnunsigned long pattern = 0;
23:06:51amiconnint value = 0;
23:06:51amiconnfor (j = 0; j < graybuf->depth; j++)
23:06:51DBUGEnqueued KICK amiconn
23:06:51amiconnpattern <<= 1;
23:06:52***Alert Mode level 1
23:06:52amiconnvalue += i;
23:06:54amiconnif (value >= graybuf->depth)
23:06:56amiconnvalue -= graybuf->depth;
23:07:00amiconnpattern |= 1;
23:08:13amiconnBtw: i runs from 0 to graybuf_depth and is the desired brightness.
23:09:16[IDC]DragonI'll wait for the final version, don't understand this.
23:09:49amiconnmidknight2k3: With my grayscale framework it should be possible to write a bmp viewer. It would even be possible to view pictures while the music is playing if it is done clever.
23:10:08midknight2k3go idc and amiconn
23:10:08[IDC]Dragonyou have i and j ?
23:10:51[IDC]Dragonguess I have to implement that timer function soon, to prevent hardware banging in plugins
23:11:11amiconnYes, this routine is for precalculation of the patterns for all possible brightnesses. It is executed when you call get_graybuf().
23:11:49amiconnI'm using timer 4 (the same as you in your video plugin).
23:11:58[IDC]Dragonlater you need to collect 8 pixels again to get a screen byte
23:11:58 Quit Dogger (Read error: 54 (Connection reset by peer))
23:12:10 Join Dogger [0] (
23:12:35[IDC]Dragon(afk for a minute)
23:12:44amiconnThe problem is, that all bits that correspond to a screen pixel are scattered across different bytes.
23:13:26amiconnSo there is a lot of bit masking and wild addressing of the different planes.
23:14:08amiconnTo get this reasonably fast, the core function graypixel() is assembler optimized (what did you expect from me?).
23:15:25amiconnIt takes ~1.3 sec to set 3/4 of the screen pixel by pixel.
23:16:53***Alert Mode OFF
23:19:35[IDC]Dragona strenuous operation
23:19:51[IDC]Dragonso 1.3 sec for haltoning ang re-arranging?
23:21:57amiconnThis is for calling gray_drawpixel(int x, int y, int brightness) for a total of 5376 pixels (3/4 of the screen).
23:22:46amiconn(When using the maximum number of 33 scales of gray. When using less grayscales it get faster)
23:23:21[IDC]Dragona function call for each pixel?
23:24:12amiconnYes, necessary even for rectangles, becauses the phase of the bit pattern has to be shifted randomly for every pixel.
23:24:38[IDC]Dragonare you using rand() for that?
23:24:51[IDC]DragonI heared it's slow
23:25:15[IDC]Dragonbut a high-end random generator
23:25:22amiconnMy observation is that rand() cosumes about 1/2 of the total time it takes.
23:25:26[IDC]Dragonvery random ;-)
23:25:47[IDC]Dragonmaybe you should do your own
23:26:17amiconnBtw: The timing figures I gave were takes while the interrupt was running. This consumes ~50 % CPU time.
23:26:29[IDC]DragonI know, yes
23:27:02[IDC]Dragonyou can use one rand() to get bits for several pixels
23:27:15amiconnSo if you have to draw a lot and don't care about the display, you can switch it off while drawing.
23:28:04amiconnNo, I can't. This way the phase of these pixels would be equal -> flicker
23:28:47[IDC]Dragonyou didn't get me, i think.
23:28:53[IDC]Dragonif you do rand(), you get 16 (?) bits
23:29:04amiconnrand() gives 31 bits.
23:29:09[IDC]Dragonfor each pixel, you need up to 5
23:29:57[IDC]Dragonso you can split the result and use fractions for several pixel.
23:30:39amiconnBut this way I would have to calculate the bit pattern for the pixel from these bits, which is slower than precalculating the patterns and using the rand() result only for shifting.
23:32:11[IDC]Dragonall I mean is buffering the Rand() result, since you don't need all the digits, and instead call it only every 4 times or so.
23:32:35amiconnOops, perhaps this is nevertheless a good idea: I could have a "random bits reservoir" which I shift by 5 bits to get my random value and refill only every 5 calls.
23:32:50[IDC]DragonIn the meantime, shift the results to get your next bits, until consumed.
23:33:05[IDC]DragonThat's what I meant.
23:33:05 Quit Dogger (Read error: 104 (Connection reset by peer))
23:33:19amiconnOk, will try this.
23:33:22 Join Dogger [0] (
23:33:54[IDC]Dragonyou rotate your bit pattern, right?
23:34:18amiconnNo, I don't rotate beforehand.
23:34:41[IDC]DragonI mean rotate by the random phase
23:35:09amiconnI decide from the random value in which bit plane I shoud start, then go up to the last and finally do from the first to the one before my starting point.
23:35:36[IDC]Dragonso the "seam" will be on different planes for each pixel?
23:36:08 Join LinusN [200] (
23:36:18amiconnhi LinusN!
23:36:25[IDC]Dragonyou can as well rotate your pattern by random, then walk from plane 0 to n.
23:36:35[IDC]DragonHi LinusN
23:36:49[IDC]DragonI was just about to leave...
23:37:26[IDC]DragonYou do this on purpose, so I'm tired at work, and you've weakened germanies economy
23:37:37LinusNof course
23:37:51[IDC]Dragonvery subtle
23:38:03amiconn[IDC]Dragon: Rotating would take longer (this was what I considered first) because you have to shift out while setting the planes anyway.
23:38:19LinusNi work at the DOEP, Department Of Evil Plans
23:38:42[IDC]Dragontogether with Dogbert?
23:39:02 Quit Dogger (Read error: 104 (Connection reset by peer))
23:39:02LinusNhe's my boss
23:39:24 Join Dogger [0] (
23:39:29[IDC]Dragonyou're on vacation?
23:40:09LinusNlaptop + shitty v.90 modem
23:40:29[IDC]Dragonon some remote place, with a notebook and a satellite dish?
23:41:06[IDC]Dragonfast enough for IRC
23:42:05LinusNi'm on an island on the west coast of Sweden
23:42:45[IDC]Dragonwhat are you doing in here then?
23:42:51[IDC]DragonGo fishing!
23:43:25LinusNin the middle of the night? it's damn dark and cold :-)
23:44:17[IDC]DragonDon't they bite better then, easy to attract with light? (I'm no expert)
23:44:43LinusNwe usually catch salmon in the night, with a net
23:44:53LinusNit's not really season yet
23:44:59[IDC]DragonIf we can't persuade you, you'll have to anwer stupid questions:
23:45:19CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:45:19*LinusN goes to the boat
23:45:33midknight2k3LinusN, for some rason bmp2rb is seeing my bitmap inverted
23:45:35midknight2k3any ideas?
23:45:46[IDC]Dragon1) any particular reason against recursuve directory deletion?
23:45:57LinusN[IDC]Dragon: 1) no
23:46:05[IDC]DragonI miss that
23:46:05midknight2k3LinusN, do you know why?
23:46:08***Saving seen data "./dancer.seen"
23:46:14midknight2k3"yes" to me or to idc?
23:46:19LinusNmidknight2k3: your pallette is inversed in the bitmap
23:46:38midknight2k3the bitmap itself is in the correct way
23:46:55[IDC]Dragon2) how about sorting dirs by date?
23:47:05LinusN2) Sure, why not?
23:47:15LinusNmidknight2k3: you mean it looks correct in Paint?
23:47:21[IDC]Dragonwe don't store it yet, right?
23:47:23midknight2k3well gimp at this point
23:47:32midknight2k3but i did save it in paint yes
23:47:40LinusN[IDC]Dragon: i don't think so
23:47:46[IDC]Dragonso the dircache will grow
23:48:06LinusNmidknight2k3: just invert the bitmap before you convert
23:48:13LinusN[IDC]Dragon: probably
23:48:15midknight2k3do you know why it's doing this? a bug?
23:48:34LinusNmidknight2k3: no, it's not a bug
23:48:54LinusNthe pain application can define the palette in any way it wants
23:48:57midknight2k3just made tha way?
23:49:17midknight2k3um, do you how know i can invert it in the gimp?
23:49:25LinusNthe bmp2rb tool doesn't care about the palette
23:49:32LinusNmidknight2k3: nope
23:49:36midknight2k3got it
23:49:40midknight2k3well anyhow
23:49:42[IDC]DragonLinusN: can your delete function wipe (empty?) directories?
23:49:46midknight2k3i've got a huge clock update
23:50:16LinusN[IDC]Dragon: it won't free the fat entries for the files in it
23:50:48[IDC]Dragonso the dir has to be empty, that's what I meant
23:50:52LinusNyou'll have to recursively delete each file
23:51:13LinusNwould be great if you implemented that feature
23:51:18LinusNbeen missing for so long
23:51:43[IDC]Dragonyes' I'd like to delete unworthy albums
23:52:59[IDC]Dragonsoon Rockbox will hit the 200k barrier again
23:53:43[IDC]Dragonmaybe we can do an option cleaning day
23:54:54LinusNi'm about to move the VBR fixer to a plugin
23:54:54 Quit Dogger (Read error: 54 (Connection reset by peer))
23:54:58[IDC]Dragonwhen I did the voice menu and saw all the options and their granularity, I couldn't believe it
23:55:40[IDC]Dragonis VBR fix large?
23:55:52 Join Dogger [0] (
23:55:53LinusNnot really
23:56:29[IDC]Dragona table-driven config load/save could save substantially, I guess
23:56:57amiconnLinusN: I did my overnight test recording (a whopping 742 MB recorded in over 10 hours at q=7, f=44.1kHz, stereo). I did not find a single error in the frame structure.
23:57:14LinusNamiconn: good for you :-)
23:58:14LinusNamiconn: you want some logic analyzer traces?
23:58:28amiconnBtw: mp3fixer doesn't work correctly! It reports errors where there are none! Verified myself by looking into the given position with a hex editor.
23:58:49[IDC]DragonWell, at 200k the game is as follows: anybody who wants to implement a feature has to optimize the rest such that the new stuff fits.

Previous day | Next day