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 2006-03-06

00:01:07 Quit virtualball2 ("virtualball2 has no reason")
00:02:38 Quit bataman ("CGI:IRC (EOF)")
00:03:09Mark__thats an uppercase i
00:03:36Sando H120's play when they are plugged into the mains?
00:03:43 Quit Thus0 ("Leaving")
00:06:50midkayhaha. Mark__ - you're sick.
00:08:27 Quit damaki__ (Read error: 110 (Connection timed out))
00:09:01 Quit San ()
00:11:05 Join quobl [0] (n=quobl@tor/session/x-bd02e3dd8ba1ae6e)
00:12:05 Join damaki_ [0] (
00:12:33***Saving seen data "./dancer.seen"
00:14:34preglowpeakmeter is still wildly unresponsive for heavier files
00:14:45preglowcpuboosting still seems to work on nano
00:17:41 Quit aegray (Read error: 104 (Connection reset by peer))
00:18:11preglowlostlogic: am i just imagining things or was stuff way more responsive before your latest commit?
00:19:26 Join aegray [0] (
00:28:31 Quit damaki (Read error: 110 (Connection timed out))
00:31:28 Quit aegray (Read error: 104 (Connection reset by peer))
00:31:41 Join aegray [0] (
00:32:36lostlogicpreglow: I didn't notice a difference aside from the no longer skipping
00:33:33lostlogicpreglow: buttons unresponsive, or the peakmeter specifically?
00:33:56herz42ah, you're still there...
00:34:31herz42I did some research what causes all the i2c traffic
00:34:39herz42and its...
00:34:47herz42the rtc
00:35:11lostlogiceven with my limitting it to 4 calls per second?
00:35:31herz42should be read only HZ/4, but there's a small bug, so it's read 20 times per second
00:35:51herz42hope this works:
00:35:55herz42−−- firmware/common/timefuncs.c 5 Mar 2006 18:34:27 -0000 1.15
00:35:55herz42+++ firmware/common/timefuncs.c 5 Mar 2006 23:33:10 -0000
00:35:55herz42@@ -51,6 +51,8 @@
00:35:55DBUGEnqueued KICK herz42
00:35:55herz42 /* Don't read the RTC more than 4 times per second */
00:35:56***Alert Mode level 1
00:35:56herz42 if (last_tick + HZ/4 < current_tick) {
00:35:58herz42+ last_tick = current_tick;
00:35:59lostlogicherz42: eek... I didn't set the last_tick
00:36:01lostlogichaha yeah
00:36:02herz42 char rtcbuf[7];
00:36:02lostlogicI rule.
00:36:04herz42 rtc_read_datetime(rtcbuf);
00:36:28lostlogichey, that problem led me to improve the i2c driver :)
00:36:32herz42that fixed skipping before I updated cvs with your changes
00:36:34preglowlostlogic: peak meter specially
00:36:39preglowlostlogic: but it might be placebo
00:36:44herz42but now it's even better :)
00:38:03 Quit pyro ()
00:38:33 Quit aegray (Read error: 104 (Connection reset by peer))
00:39:03amiconnlostlogic: Why is the rtc read that often on iPod?
00:40:31 Join aegray [0] (
00:42:34 Join fiftyfour123 [0] (
00:42:38 Quit fiftyfour123 (Client Quit)
00:43:07herz42amiconn: that's a bug in the rtc code. It's present on all devices with rtc...
00:43:58lostlogicamiconn: it's read that often on all targets −− but not any more... they all call get_time that often, which calls rtc_read
00:44:08lostlogicyeah, what he said
00:44:19amiconnHuh? I thought the rtc is read once at startup, and written when the time is changed via the time setting screen
00:44:26amiconnThat's... odd
00:44:41amiconnIt's in fact a waste, since rtc means i2c
00:45:57***Alert Mode OFF
00:46:02lostlogicamiconn: do that and depend on the tick timer to maintain the clock during operation? How accurate is the tick timer? (More accurate than RTC on H3x0)
00:46:16lostlogicactually, I bet that this change makes H3x0 performance a lot better
00:46:19 Join ashridah [0] (
00:46:52amiconnThe tick timer is as acurate as the cpu frequency xtal, as long as no interrupts are lost due to too long disabling of interrupts
00:47:26amiconnA few ms error might be added everytime the cpu frequency is switched,
00:47:28 Quit ender` (" Consultation, n.: Medical term meaning "to share the wealth."")
00:47:55herz42hmm, now I get data aborts again. maybe I should try a clean build
00:48:16amiconnI performed some measurements a while ago. Deviation was a few seconds per day
00:48:26amiconn...i.e. comparable to the rtc chip
00:48:29Paul_The_NerdSo why not just check the RTC like, once per 30 minutes or something, long before anything close to "real" error can be introduced? Just in case one of those hypothetical situations does occur, and everything...
00:49:21 Quit webguest41 ("CGI:IRC (EOF)")
00:50:33 Quit aegray (Read error: 104 (Connection reset by peer))
00:50:44 Join aegray [0] (
00:58:46midkayi just helped a friend install rockbox on his nano and he seems to be having some freezing problems.. maybe it's due to some recent commits?
00:58:58preglowhow's the freeze look?
00:59:01preglowno error message?
00:59:07midkayeverything stops working..
00:59:21preglowi think my commit should be pretty safe
00:59:31herz42I also get data aborts/undefined instructions/freezes with the latest cvs code
00:59:35preglowhow old is his build?
00:59:37midkayyeah, wondering if it's a recent commit or something - i never got anythhing like that
00:59:44midkaylatest CVS (within half hour or so)
00:59:55lostlogicit's possible some of my stuff could cause problems on ipods
01:00:13preglowseems to work fine on mine
01:00:28midkayhe's had two freezes in the past 10m or so..
01:00:30lostlogicmidkay: does it freeze after the same length of time from bootup each time?
01:00:54nudeldon't know if it's the hardware or not but there's a nasty, loud click when shutting down rockbox (maybe it does it with the official firmware too? never used it :))
01:01:01midkaylostlogic, hm - let me see
01:01:06nudel(in the headphones that is)
01:01:22nudel(er, on 5G ipod... sorry, tired)
01:01:22 Quit aegray (Read error: 104 (Connection reset by peer))
01:01:57midkaylostlogic, no - it happened first about 25 minutes after he started using rockbox
01:02:07midkaythen he reset - happened again about 5 minutes later
01:02:23lostlogicmidkay: playing the same playlist both times?
01:02:29lostlogicmidkay: if so try recreating the playlist
01:02:46lostlogicI've seen similar behavior on my H3x0 with a playlist bug that I haven't tracked down yet.
01:02:50midkaylostlogic, no - first time directory browsing, second playing some music..
01:02:57lostlogicoh well
01:02:57midkayi believe
01:03:38 Join aegray [0] (
01:07:30midkaylostlogic, oh, so it appears both times he got the freeze was playing music (skipping around..)
01:07:38midkayjust random files in ipod_control/music
01:08:20lostlogiccould be one of those ghost swcodec skipping bugs then
01:08:27midkayhm :)
01:10:33lostlogicherz42: how do you know you are gettin' data aborts? Still having problems?
01:11:06preglowat least i don't think my fiq stuff should break anything, but it might
01:11:45herz42yes, still not working
01:12:37Paul_The_NerdWhen do they happen?
01:12:57lostlogicherz42: define data aborts? what identifies such a thing?
01:13:05herz42when playing music. 1/10 sec after play or after 2 mins...
01:13:28 Join BHSPitMonkey [0] (
01:13:29herz42data aborts are when the panic screen says: data abort at ....
01:13:38lostlogicahh, yeah, I haven't had that.
01:13:49herz42same with invalid instruction
01:14:09lostlogicpreglow: no problems here with your or my stuff
01:14:31preglowherz42: please supply us with the address
01:15:01preglowand a map file, if you've got one
01:15:16herz42address varies. I had eg. d.a. at 3f49c and just now at 15a
01:15:17preglowthat'd be to midkay
01:15:24preglowi had one in my fiq handler
01:15:29preglowbut i pruned that away
01:15:37preglowsince that i haven't had any
01:15:52 Quit Mark__ (Read error: 104 (Connection reset by peer))
01:16:11midkaypreglow, hm?
01:16:32 Join stripwax [0] (
01:17:09linuxstbSorry to interrupt this fine ipod hacking, but I'm trying to set up Grip to encode to FLAC and add replaygain. Am I right in thinking that I just need two tags for replaygain - REPLAYGAIN_TRACK_GAIN and REPLAYGAIN_ALBUM_GAIN ? Are the peak values useful for anything?
01:17:49Paul_The_NerdDoes "prevent clipping" use them?
01:17:58stripwaxSo I'm debugging the auto-change-directory-not-working-when-running-a-plugin bug and it's because get_tree_context returns a context that has a filter set for Rocks only, rather than the original filter (i.e. music files!). So all the music files are ignored, so the playlist doesn't get built, so it just plays (repeatedly) the last track on your playlist.
01:18:24stripwaxCould someone tell me how that's supposed to work?
01:19:33preglowlinuxstb: well, they're nice to keep around to make sure you never clip
01:19:45stripwaxThis bug is *really* annoying by the way :)
01:19:47amiconnstripwax: That happens when functions are used for things they weren't meant to be used for...
01:20:01stripwaxwhich functions in particular?
01:20:11 Quit antoine ("Client exiting")
01:20:38amiconnThe tree context was never meant to be used from anything else than the gui thread
01:21:04stripwaxamiconn - hrm, understood. the auto-change-directory code uses ft_load.
01:22:42 Quit ghode (Read error: 110 (Connection timed out))
01:22:43 Quit needleboy (Read error: 104 (Connection reset by peer))
01:22:46 Quit Kohlrabi ("Fast alle Menschen sind Regenwürmer")
01:28:31lostlogicpreglow: any chance of the CPU throttling working on ipod video? Want to show me the code that works with the nano?
01:29:51stripwaxmaybe a fix would be to make/update a local copy of the current tree context whenever an in-memory playlist is created/modified. So that way the get_next_directory code in playlist.c would be able to use the same tree_context as when the playlist was created (and will therefore look for music files when auto-change-directory kicks in). amiconn - makes sense?
01:30:30stripwaxhrm - although not sure how that would work over restart/resume ..
01:31:13amiconnI don't know how this code works internally
01:31:24stripwaxwho does
01:31:42amiconn<opinion> I would rather drop auto-change-directory </opinion>
01:32:04stripwax<opinion>I wouldn't since I use it every day</opinion>
01:32:26amiconnI never used it so far...
01:33:10amiconnBut if we want to keep it, a workable method shouldn't add too much complexity
01:34:25stripwaxiRiver owners are used to the behaviour of automatically playing the files in the filesystem (in directory order), as a convenient way to listen to music without having to set up / update any playlists. I think auto-change-directory is the only way to activate the same behaviour in rockbox; let me know if there's some other way to do the same. I agree that a fix shouldn't add too much complexity
01:34:48linuxstblostlogic: Yes, preglow's cpu frequency patch works on the 5g. It's just my 4g Color that dies.
01:34:49lostlogicstripwax: you can just make a playlist of the whole filesystem and play it
01:35:03stripwax^^^ "without having to set up / update any playlists"
01:35:04amiconnThere is no other way to mimic the iriver firmware in this respect.
01:35:17lostlogiclinuxstb: how's the battery life with the patch?
01:35:35lostlogicstripwax: it takes 5? button presses to make a playlist of the whole device...
01:35:43linuxstbI only tried it briefly - enough to test that it worked. So I don't know.
01:35:52lostlogiclinuxstb: ok, thanks
01:35:53amiconnHowever, the general opinion is that rockbox shouldn't try to mimic the stock firmware on any target
01:35:57stripwaxlostlogic - yes and the playlist needs to be updated *every* *single* *time* new music is added.
01:36:09amiconnIt should rather provide whatever we agree upon is convenient
01:36:18 Quit Moos ("Glory to Rockbox!")
01:36:28amiconnI for one tried the iriver firmware once- I couldn't bear it more than a few minutes
01:36:37stripwaxamiconn - of course. maybe having this feature already available in rockbox has spoiled me; I don't think removing features is a meaningful way to debug however
01:37:22preglowlostlogic: works fine on nano and 5g
01:37:24lostlogicpreglow: rockin' thanks
01:37:29amiconnI also don't think auto-change directory is a convenient way of listening music - I prefer to select the order *I* want
01:37:50stripwaxamiconn - oddly enough, I tried the iriver firmware for over a year, before rockbox was a reality, and I think this is probably the single feature I couldn't live without. the rest of the iriver firmware is, of course, trash.
01:38:01amiconnIf I just want arbitrary music, I could switch on the radio...
01:38:09preglowi was annoyed as hell with rockbox' way
01:38:35stripwaxamiconn - really? I don't think this is necessarily a productive conversation .. who is familiar with the internals of this code
01:38:35preglowauto change directory being off is one of the few features defaults i can't stand
01:38:47stripwaxpreglow - hear, hear
01:39:17amiconnI have quite a few defaults I can't stand, but auto-change directory being off is just fine
01:39:34amiconnThat's what .cfg files are for...
01:40:02amiconnstripwax: Anyway, you're right that removing features isn't the right way to debug,
01:40:29amiconnbut a feature cleanup will be unavoidable sooner or later
01:40:46preglowhell yes, and i can think of zillions of features to take out before that one
01:41:06preglowprimarily because i can't live without it :>
01:41:11amiconnThat doesn't necessarily mean to remove features, but also to look what features could be handled together
01:42:40preglowi'd really like it if there were two modes of using rockbox, and one being purely directrory based, with no playlist stuff
01:42:46preglowbut i realise that's probably never going to happen
01:43:05stripwaxok, so I need to find the function to call, that is used to set up the tree context when the user goes from the WPS to the File Browser. where can I find that function? once I've found it, I can just call that at the start of get_next_directory in playlist.c
01:43:41amiconnstripwax: You would need to reset it to what it was before
01:44:18CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
01:44:18*amiconn wonders what will happen if auto-change directory is enabled and the browser is in tagdb mode...
01:44:40preglowprobably something funky
01:44:42preglowit's a hack as it is
01:44:50preglowand only half works
01:44:53stripwaxamiconn - no, I just want to call the function that sets up the tree context (and pass that tree context around, like what the code currently does). I don't want to change the current tree context
01:45:03stripwaxtagdb is overdue for a feature cleanup :)
01:45:24amiconnstripwax: You want to use more than one context?
01:45:46*amiconn expects lots of reentrancy problems in the tree code
01:45:48stripwaxamiconn - huh? no, I want to pass the correct tree context to ft_build_playlist
01:46:16amiconnAs mentioned, the tree code was never written to be used from more than one thread
01:46:58amiconnA lot of code makes heavy use of global vars
01:47:28amiconnthat helps keeping code size down
01:48:32 Quit crashd (Read error: 104 (Connection reset by peer))
01:49:20stripwaxfair enough. so I ask again, who knows best how this is supposed to work? who'd have good ideas on how to fix simply
01:49:43stripwaxI'm assuming the solution is "not to use the tree code for auto change directory" by the way.
01:50:56amiconnHmm, I wonder how auto-change dir works
01:51:30amiconnAfaics this means reloading of the directory buffer, i.e. the browser will be switched to another dir
01:51:51 Join ScootScat72 [0] (
01:52:50stripwaxit calls playlist_create( directory, NULL ) .. it doesn't switch the browser to another dir, however, as far as I know (browser still seems functional)
01:55:02amiconnHmm, and where does it get the directory from?
01:55:04stripwaxbut the action of getting the next directory to use in the call to playlist_create is what's causing the problem. that part uses the tree stuff
01:55:34stripwaxamiconn - code in playlist.c, start at get_next_directory
01:56:32 Join Strath [0] (
01:57:06stripwaxit's pretty sick. I can see it backs up / modifies global_settings.dirfilter (though that flag seems to be duplicated in the tree context so doesn't behave as expected) and it does call reload_directory.
01:57:30 Quit aegray (Read error: 104 (Connection reset by peer))
01:57:43 Join aegray [0] (
01:58:22stripwaxok, well, i'm out of here
01:58:24 Part stripwax
01:58:56herz42just to wrap up the ipod freezing problem - I get freezes on both, the last version in cvs (23:39) and the one from 20:38. So it seems not to be related to the latest changes.
01:59:18 Quit BHSPitMonkey (Remote closed the connection)
01:59:50herz42With the lates cvs version I get the problem when I play certain songs.
02:00:02amiconnThis thing messes with global config flags + tries to splash() occasionally
02:00:26amiconnNasty hack when used from background threads
02:00:49amiconnI.e. it might crash an archos or an iriver h1x0 when a grayscale plugin is running...
02:01:01herz42If the song is directly selected, I get data abort. When the song is played as 'next-song', the ipod freezes.
02:01:30herz42so long - gn8
02:01:43 Part herz42
02:02:28 Join BHSPitMonkey [0] (
02:04:43lostlogicyay, it's nice to have my old friend the boost percentage back on ipod, thanks preglow
02:04:54lostlogics/back //
02:06:30preglowcan't say i missed it too much myself, but we just had to reunite
02:07:11preglowwonder what makes the 4g hate it so much, though
02:07:50preglowlostlogic: i bet the video is just lovely at 30mhz
02:09:14 Quit ScootScat (Read error: 110 (Connection timed out))
02:10:00lostlogicpreglow: hehe, I don't care about the UI responsiveness, just about battery life, and I'm sure that 58% boost is still better than 100% for that :)
02:10:32preglowshould be
02:10:56preglowPaul_The_Nerd did a test for me that didn't exactly make it look like it had too much to say, but then again, he didn't catch the exact new runtime
02:12:36***Saving seen data "./dancer.seen"
02:13:29Paul_The_Nerdpreglow: My test was at about 75-80% boost though
02:13:51preglowfor flac??????+
02:13:52Paul_The_NerdOh, no, wait..
02:13:58Paul_The_NerdI was thinking of a more recent one.
02:14:10lostlogicman, disk access seems slower on the ipod than H340.
02:14:20Paul_The_NerdYeah, for FLAC, almost no boost at all. Sorry
02:14:26preglowi'm working on a flac optimisation now that might speed stuff up a bit
02:14:45*Paul_The_Nerd forgot about the flac test.
02:15:09Paul_The_NerdSadly, the TSR plugin for battery testing doesn't seem to work on iPods either
02:22:24 Join harbel [0] (
02:23:45harbelIs there a way to set rockboy controls without actually being in rockboy? Like from windows, for iPod 5g
02:28:28lostlogicHey anyone know if the iPod 4g 40gig drive is teh same as the H340 drive?
02:29:09linuxstbpreglow: Do you think your CPU patch can be committed for the Nano and the 5g?
02:29:30preglowlinuxstb: well, i've never had any trouble with it apart from rockbox being slow as hell at 30mhz
02:30:12preglowand i could enable it for only those, i guess, but it won't be pretty
02:32:43preglowharbel: i don't even understand what you mean
02:32:48preglowharbel: set controls?
02:32:59preglowcan you even configure rockboys controls?
02:33:38harbelNope, it will only let the 5g use menu, select, and play
02:33:47preglowthen that's what there is
02:33:59preglowipod and rockboy doesn't exactly go well together yet, i think
02:34:18harbelSo there's no way to configure them?
02:34:25preglownot that i know of
02:34:41lostlogicgah, I totally don't read arm asm.
02:34:53preglowthat's exactly what i do now :/
02:35:21linuxstbThere's a handy "ARM instruction set quick reference card" - I'm sure google can find it.
02:36:07 Quit YouCeyE (Remote closed the connection)
02:36:41 Quit harbel ()
02:36:42lostlogiclinuxstb: yeah, I'll have to go get one, but I've been avoiding using a GUI for the entire weekend and I don't want to start.
02:37:30 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
02:37:42XavierGrPaul_The_Nerd: Here?
02:37:59*amiconn has display flip working on h300
02:38:16XavierGrWhat will happen if you load the battery_bench plugin on iPod?
02:38:21Paul_The_NerdXavierGr: The battery plugin just creates the .txt file, but never fills in a single entry.
02:38:26amiconnJust needs some more polishing
02:38:43 Join pyro [0] (
02:38:48XavierGrah yes, that is because battery_voltage is not used on iPods
02:39:00XavierGror so I remember from linuxstb comments
02:39:08Paul_The_NerdYeah, not yet.
02:39:45XavierGrmeasurements are taken when the voltage changes so....
02:40:48lostlogicarm -> all registers are data / address?
02:40:48Paul_The_NerdWell, there's also the fact that I have a Nano, so the disk never spins up/down, and I seem to remember the text file mentioning something being dependent upon that.
02:41:13XavierGrthe same applies on ondio
02:41:15XavierGrbut it works there
02:41:43preglowwe never sleep the "disk"
02:41:47preglownano hangs if we do
02:41:59preglowit's probably got something to do with that
02:42:23XavierGrhmm then maybe this will matter too.
02:42:33preglowi have just disabled the entire ata sleep code
02:42:54amiconnOndio fakes ata sleep
02:43:15amiconnIt reports the disk sleeping all the time, except if the mutex is locked
02:44:01amiconnIt was probably easier to implement faked sleep for Ondio that it will be for nano
02:44:17preglowideally we shouldn't have to
02:44:24preglowwe just need to figure out what the hell goes wrong now
02:44:48preglowthe data sheet for the ata->flash bridge chip says it supports sleep commands, even though they are redundant, since it autosleep after a few ms with no activity
02:44:51XavierGrcan someone see if the plugin works on a HD based iPod?
02:45:01XavierGrIf not I will have to update the wiki
02:47:44XavierGrSomeone is an AD&D (or D&D or just RPG games) fun that programmed the latest dice plugin.
02:48:33 Quit ashridah ("Leaving")
02:59:03preglowwoo, flac-opt works
02:59:13preglowturns out all i was missing was an 's'...
03:00:32linuxstbDoes it make much difference to your boost ratio? Or wasn't it boosting?
03:00:35 Join omghaxzs [0] (
03:00:35 Quit omghaxzs (Client Quit)
03:00:45preglowdunno, i'll test that afterwards
03:00:45ts-xlostlogic: if you get a chance, I *think* this bug may have come with the pcm buffer changes a few weeks back:
03:00:53preglowi started on this code before i had boosting
03:00:54 Join omghaxzs [0] (
03:04:10 Join ghode|afk [0] (
03:05:08 Quit youngcerea1 ("Download Gaim:")
03:10:24 Quit muesli__ (Read error: 110 (Connection timed out))
03:11:21preglowplus, there's still a bug in the generic routine, i see
03:11:58linuxstbWhat's that?
03:15:46preglowwell, i do like with the coldfire one
03:15:57preglowtons of unrolled loops, plus one generic one for higher orders
03:16:03preglowin the case of arm, all orders above nine
03:17:22linuxstbOK. I thought by "generic" you meant the C routine...
03:17:27preglowheh, no
03:17:30preglownothing wrong with that
03:17:36 Join perl|wtf [0] (
03:17:41preglowwill at least be easy to add 64 bit wide routines on arm
03:19:02linuxstbyep. Have you looked at the code gcc produces for flac?
03:19:17preglowwell, it's pretty standard fare
03:19:21preglowdoing reloads all the time
03:19:45preglowmy routines do block loads as often as possible and never reload the coef array
03:20:49preglowahh, so this is not the vim window
03:21:05 Quit perldiver (Read error: 110 (Connection timed out))
03:21:32linuxstbTime for bed. :wq and goodnight.
03:21:47midkayhaha. later linuxstb
03:22:33 Join quobl_ [0] (
03:25:49 Join Shadowarrior13 [0] (
03:25:55 Quit omghaxzs ()
03:33:39 Quit quobl (Read error: 110 (Connection timed out))
03:35:57 Join Bluechip [0] (
03:36:43 Quit Shadowarrior13 ("( :: NoNameScript 4.01 :: )")
03:44:43lostlogicts-x: not pcmbuf related... file that with playlist bugs incl. the one that makes it freeze when it tries to read songs at certain positions in certain playlists
03:45:08preglownow let's see
03:46:38preglowwell, it never boosts at 30mhz now
03:46:41ts-xlostlogic: Ok. It just started recently, I was just trying to figure out what triggered it...
03:47:55*preglow wonders if he should do a driveby commit...
03:48:54 Quit ghode|afk (Read error: 110 (Connection timed out))
03:49:08preglowoh well, it never boosted before either, it seems
03:49:42 Part Bluechip
03:49:46 Join Bluechip [0] (
03:51:48*BHSPitMonkey tips his hat to Rockbox
03:53:10preglowmost definitely faster now, yes
03:55:15preglowwent from 20% to 4% boost at 21/75mhz for a -8 flac file
03:56:04preglowthen there's another -8 file that decodes 100% realtime at 21mhz :>
03:57:28preglowthink i'll consider myself satisfied at that
03:57:51lostlogicman, so many more instructions involved in mips compared to m68k with the advanced addressing modes of m68k.
03:59:31preglowarm at least hasn't got that much
03:59:40lostlogicerm, not mips, I meant arm
03:59:42preglowthe ones you've got are pretty flexible, though
04:00:06 Join Shadowarrior13 [0] (
04:00:32lostlogicpreglow: the difference between m68k and arm is that you can use good addressing modes in loads, but not in other instructions, so you end upw ith more total lines of asm
04:01:46preglowthere are no addressing modes for instructions other than load/store
04:01:52preglowsince those are the only ones that can access memory
04:02:05preglowalmost all risc architectures are like that
04:02:15preglowi'm not too fond of that myself, but i'm starting to get used to it
04:03:08preglowbut that fact also makes the regular instructions more flexible
04:03:23preglowlike parallell shifting/rotating
04:04:50preglowthink i'll just commit this asm blob right now
04:09:29preglowbut in other places you get along with less asm on arm
04:09:42preglowlike the function prelude for the fucntion i just commited
04:09:49preglow7 instructions as opposed to 15 on coldfire
04:10:14lostlogicyeah, I see how it can be good... obviously have much to learn.
04:11:17preglowooh, i see another possibility for optimisation
04:12:38***Saving seen data "./dancer.seen"
04:12:38preglowsaving that for tomorrow, good night
04:14:52 Quit ts-x ("CGI:IRC")
04:36:15 Quit wizatcomputer (Read error: 104 (Connection reset by peer))
04:37:08 Join wizatcomputer [0] (
04:42:36 Quit wizatcomputer (Connection reset by peer)
04:43:28 Join wizatcomputer [0] (
04:55:30 Quit imphasing ("Lost terminal")
04:59:11 Join imphasing [0] (
05:01:02 Quit Rob2222_ (Read error: 104 (Connection reset by peer))
05:01:24 Join Rob2222 [0] (
05:02:58 Quit kernelsensei (Read error: 104 (Connection reset by peer))
05:03:04 Join kernelsensei [0] (n=boris@gentoo/developer/kernelsensei)
05:04:23 Quit godzirra (Client Quit)
05:07:13 Join egotrippen [0] (
05:17:38 Join virtualball2 [0] (
05:20:21 Quit BHSPitMonkey (Remote closed the connection)
05:20:57 Quit egotrippen ("CGI:IRC (EOF)")
05:22:21 Join BHSPitMonkey [0] (
05:42:50 Quit imphasing (Read error: 104 (Connection reset by peer))
05:45:07 Quit virtualball2 (Read error: 110 (Connection timed out))
05:45:07 Quit midkay (Read error: 104 (Connection reset by peer))
05:45:17 Join midkay_ [0] (n=midkay@
05:52:54 Quit BHSPitMonkey (Remote closed the connection)
05:53:47 Quit Shadowarrior13 ("( :: NoNameScript 4.01 :: )")
05:57:22 Join BHSPitLappy [0] (
05:58:59 Quit RotAtoR ("zzzzzzzz")
06:06:26 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
06:10:23 Part paugh ("Leaving")
06:12:40***Saving seen data "./dancer.seen"
06:15:48 Quit midkay_ ("Leaving")
06:23:38 Join nave7693 [0] (
06:34:32 Part nave7693
06:52:25 Join yeahx [0] (
06:53:31 Join midkay [0] (n=midkay@
06:53:58yeahxmidkay :)
06:54:07midkayhey yeahx :)
06:54:31yeahxhow goes it?
06:54:32yeahxI havent been here in a while
06:54:39yeahxhow about this awesome weather?
06:54:45*BHSPitLappy looks outside
06:54:47midkayyeahx, haha, pretty good, you?
06:55:22midkayBHSPitLappy, rejoice - clock update is in cvs + there's no more key combos :)
06:55:47midkayyou complained about the key setup i had, remember?
06:55:48yeahxI wore flip flops earlier thinking I was doing a quick out of the car back in the car trip, but I wasnt.... so my toes froze
06:55:56BHSPitLappynot really
06:56:03midkayBHSPitLappy, ouch, awful memory
06:56:24BHSPitLappyI remember it being overcomplicated
06:56:27midkayyeahx, haha - that sounds fun :)
06:56:32midkayBHSPitLappy, yes, that..
06:56:33BHSPitLappyno combos come to mind
06:58:06BHSPitLappyand slow-scrolling I-think
06:58:32midkayi sped it up considerably..
06:58:42midkayand simplified it a lot
06:58:53BHSPitLappyit's still slow-scrolling...
06:59:04BHSPitLappynow make a patch that adds slow-scroll as a Display Setting for all of rockbox :P
06:59:15yeahxmy friend is bugging me to put on ipl and I was wondering if I could put that on my nano as well with rockbox
06:59:21BHSPitLappyyou could.
06:59:28midkayBHSPitLappy, you mean the animations?
06:59:31*BHSPitLappy has a nano with three OS's :D
06:59:34yeahxis it a pain?
06:59:38yeahxoh cool
06:59:41BHSPitLappynot really
06:59:41midkayi don't know how slow the nano LCD driver is, but it's certainly quick enough on the 5g..
06:59:44BHSPitLappydo you already have rockbox?
07:00:00BHSPitLappyall you need to do then is repartition
07:00:08BHSPitLappyshrink the music partition by 48MB or so
07:00:15BHSPitLappyand throw an ext partition at the end
07:01:17yeahxaah ok
07:02:36BHSPitLappyI used qtparted
07:03:03BHSPitLappy(diskdrake actually, but I think it's based on qtparted)
07:03:11yeahxthen screw around in terminal to install ipl?
07:03:38yeahxhow do you switch between them?
07:05:41 Join B4gder [0] (
07:13:41yeahxso I would need to install rockbox again if I partition, which means I need windows access
07:13:46yeahxguess it will have to wait
07:14:01midkayyou don't need to reinstall rockbox
07:14:18midkayto load linux, you just hold play at boot (like holding menu for ipod firmware)
07:14:27yeahxyou can partition without wiping it?
07:15:15midkayyou need an app to partition, like partitionmagic.. or if you got a linux live cd, you could use that
07:16:34yeahxIm using Mac OS X
07:16:44midkayif osx has some partitioning tool..?
07:16:53yeahxdisk tools
07:17:15midkayyou need to create an ext3 partition, i'm not sure if it will let you - check
07:17:55yeahxoh crap I dont have my nano's cable
07:18:06midkaywhat did you do to it!!
07:18:09yeahxI was switching it with my friend
07:18:25yeahxand forgot to get the other one, its next door but he will be sleeping :)
07:19:07midkayhaha :)
07:19:44yeahxhe needs the usb one for the Dull I just sold to him, so Im taking the firewire cause I cant find my adapter, the Dull has a little firewire port on it
07:20:26yeahxDull laptop
07:25:54 Join ghode|afk [0] (
07:30:10yeahxmaybe another time
07:30:47yeahxmy friend wants me to test videos on it
07:31:02midkaycool :)
07:32:28Paul_The_Nerdyeahx: The nano won't transfer over firewire.
07:32:44yeahxIve read that vids cant be converted on mac though
07:32:50yeahxreally? how wierd
07:32:58yeahxI better find his attachment then
07:33:02Paul_The_NerdNeither will the 5G
07:33:18Paul_The_NerdiPods have moved to exclusively USB2. They'll charge on FW, but no transfers, if I understand correctly
07:33:21yeahxwhy did they change to usb and still keep that fucking stupid connector on the other side?
07:33:40yeahxI already have to jiggle the thing around to connect
07:33:51Paul_The_NerdBecause that's not a USB port on the bottom of the ipod. It has way too many pins.
07:33:59yeahxI cant believe they kept that shit a few generations already
07:34:08yeahxyeah it sucks
07:34:14yeahxits stone age crap
07:34:29Paul_The_NerdDo you have any idea what you're talking about? "Stone age"?
07:34:39Paul_The_NerdI mean, some of those pins are a line out
07:34:48Paul_The_NerdOther ones are there for the remote features.
07:35:17Paul_The_NerdIt's essentially a unified "extension" port. Much like the irivers have their little port the remote uses, etc.
07:35:43yeahxthere was a remote port next to the headphones before
07:35:59midkayPaul_The_Nerd, do you know - the irivers have chip8 emulator?
07:36:07midkayi don't see it in their /rocks folder, but on the wiki it's labeled as compatible
07:36:13yeahxits just so SCSI and the thing piece in the middle breaks easily
07:36:17Paul_The_Nerdmidkay: I think it's a viewer. Not sure
07:36:27midkayon my video it's in my rocks folder
07:36:57Paul_The_Nerdmidkay: Odd. I have an h120 sitting around, one sec
07:37:17midkayPaul_The_Nerd, ah, for the latest CVS builds i see it in viewers..
07:37:23midkayweird, but.. alright
07:37:27Paul_The_Nerdmidkay: Yeah, it's in viewers here.
07:38:05Paul_The_Nerdyeahx: I'm curious at your use of SCSI as an adjective. And what's bad about SCSI anyway?
07:39:03 Join Bg3r [0] (n=Bager@
07:39:37yeahxI had to disconnect and reconnect SCSI devices a lot back in the day and it was a pain in the ass
07:40:07Paul_The_NerdAh, so it's just the manufacturing of the port, rather than any actual aspects of SCSI itself.
07:40:54midkayum. the cube plugin runs on the archos player? that's not right
07:41:08B4gdermidkay: yes it is, iirc
07:41:25B4gderthanks to amiconn magic
07:41:27midkayB4gder, i don't understand how that could possibly be.. but i'll go with it :)
07:41:50B4gdermidkay: amiconn gfxlib uses the reconfigurable characters
07:42:01yeahxyeah scsi connectors
07:42:22yeahxit seems that soon my ipods connector will be shot and I'll have to make a trip to best buy
07:42:26midkayB4gder, still - with the amount of pixels on the player LCD.. i can only imagine you seeing about 5 pixels in the center occasionally morph into 6, then back to 5.. ;)
07:42:36yeahxis it only meant for occasional use?
07:42:41yeahxonce a week maybe?
07:42:42B4gdermidkay: yes, I bet it doesn't look very cool ;-)
07:42:51Paul_The_Nerdyeahx: How long have you had yours?
07:42:51yeahxthen how should you charge it?
07:42:59yeahxa few weeks
07:43:01midkayB4gder, apparently blobs of pixels are better than no pixels at all ;)
07:43:23Paul_The_Nerdyeahx: Man, what do you do to it? I've had mine for months without any problems, and I plug it in a minimum of twice a day for testing various things.
07:44:08yeahxdunno, just plug it in to charge and put on rockbox
07:44:53Paul_The_NerdWell, either yours was defective, or you're mistreating it, if it's broken already.
07:45:59 Part Paul_The_Nerd
07:46:08 Quit ghode|afk (Read error: 110 (Connection timed out))
07:46:48 Join LinusN [0] (
07:49:52yeahxyeah I have no idea
07:50:05yeahxI plug in my nano and lay it down gently
07:50:33yeahxId like a dock, but apple are cheap now and Im too broke to spend $30
08:01:44midkaymorning amiconn
08:02:59amiconnmidkay: You're right - cube on player doesn't exactly look smooth. You need a bit of imagination to recognise a cube
08:03:16midkayamiconn, haha - a "bit" of imagination? :)
08:03:20amiconnIt was more of an experiment, and testing the playergfx library
08:03:32midkayamiconn, cool that you got it to run though - bravi ;)
08:03:34amiconnYOu can check it out yourself
08:03:45amiconnJust build a player sim
08:03:59midkayi just may :)
08:04:03amiconnThere are some more playergfx'ed sims
08:04:31midkaywee-hee, i just updated the entire PluginIndex page.
08:04:38midkayi hope it wasn't a bad idea. :)
08:04:49B4gderI see you got a highscore (624) tonight midkay ;-)
08:04:55midkayB4gder, :D
08:05:24amiconnmidkay: logo, mosaique and snow...
08:05:44midkayamiconn, i was wondering how in the hell you managed logo :)
08:05:57amiconnlogo and snow use some additional trickery in order to use the whole screen
08:06:13midkayyeah, some limit on the number of custom cells, right?
08:06:40amiconnWith the 8 reconfigurable characters it is possible to show 14x20 or 7x40 pixels
08:06:45midkayJust About Anything's Possible When You've Got Amiconn (tm)
08:07:19midkayamiconn, ah, right
08:08:30 Join amiconn_ [0] (
08:08:58yeahxevery once in a while I try to turn off backlight but it still doesnt work :)
08:09:07 Quit amiconn (Nick collision from services.)
08:09:07 Nick amiconn_ is now known as amiconn (
08:09:11midkayyeahx, i know the exact solution: stop trying. ;)
08:09:13amiconnmidkay: logo.rock on player uses a 7x16 pixel logo, moving across the whole (!) screen
08:09:32yeahxI sort of did for a while
08:09:43midkayamiconn, that makes sense - what about snow? do you randomly tile the 8 character cells across the screen?
08:09:44yeahxjust tested it again recently thinking it might work
08:10:10amiconnmidkay: No, I just replicate the whole 14x20 image 3 times
08:10:41amiconnWow, that's a really big clock.rock update
08:10:42midkayyeahx, no - not yet - i was looking at it a few nights ago, and paul_the_nerd was looking at it too - we got "on" to turn the backlight off, but we have no idea why it works the way it does
08:10:55midkayamiconn, yes - huge ;)
08:11:04midkayworked on that for a few weeks
08:11:25yeahxoh so on turns it off?
08:11:53midkaycompiled size dropped by something like 7kb, that was nice :) more on my ipod..
08:12:00midkayyeahx, no..
08:12:10midkayyeahx, we changed some code and got "on" to turn it off
08:12:21midkaybut we couldn't figure out why changing that code in such a way would give that result..
08:12:27amiconnImho you should use the bmp build system integration for the logo. This will make it way easier to adapt clock to different LCDs
08:12:42***Saving seen data "./dancer.seen"
08:12:43midkayamiconn, that's my next update - bmp2rb during build for all the bitmaps
08:13:05DBUGEnqueued KICK Bluechip
08:13:05Bluechip has anybody any experience in removing the hard-drive spinup/write/spindown noise from recordings?
08:13:14midkaybtw - is there any objection to splitting it up into multiple files?.. i'm worried about how big it will get (even how big it is right now) trying to scale it up for a few different LCD sizes
08:13:21LinusNBluechip: is it possible at all?
08:13:50Bluechip Yes, I've had a reasonable level of success, but I'm looking to add any further improvements from other people's experiements
08:14:03LinusNi thought that would require some really advanced dsp
08:14:28***Alert Mode level 1
08:14:28Bluechip In post I have a 1.6GHz PC at my disposal
08:15:05amiconnmidkay: You could also try to just re-use the iPod 4G keypad for iPod 3G. This should work with no other change
08:15:48midkayamiconn, i checked the build errors - i don't know if it was related to the keypad thing or not, but it looked like the 3g build didn't like the way i used some defines..
08:15:49amiconn*keypad defines
08:15:51JdGordonwhere is the dircache thread code?
08:16:10LinusNJdGordon: dircache.c
08:16:17LinusNin firmware/common/
08:16:28JdGordonah, thanx
08:16:36***Alert Mode level 2
08:16:36Bluechip LinusN: I'm trying to get it as clean as I can before I try and script it
08:16:48 Join ender` [0] (i=ychat@
08:17:01LinusNBluechip: cooledit?
08:17:13midkayamiconn, ie #define BLAH "Blah." and then lcd_puts(0, 0, "Blah, Blah, " BLAH);
08:17:16***Alert Mode level 3
08:17:16Bluechip LinusN: Sound Forge is my app of choice
08:17:48amiconnmidkay: Yes, but BLAH wasn't defined for 3G...
08:17:54BHSPitLappyseriously... what's with the complete formatting abuse, Bluechip ?
08:18:13***Alert Mode level 4
08:18:13Bluechip formatting abuse?
08:18:25midkayamiconn, oh - i didn't realize that - whoops ;)
08:18:48 Join needleboy [0] (
08:19:22amiconnFor some unknown reason the 3G has its own keypad #define ...
08:19:40midkayamiconn, hm? the key layout is different..
08:20:02midkayi'd bet that it'd be desireable to have different button mappings when you've got the same buttons but in different locations, in some cases (ie games)
08:22:01 Join DJ_Dooms_Day [0] (
08:22:12amiconnAfaik the button layout of all iPods is identical, except the 1G
08:22:19BHSPitLappyBluechip: why do you need to have a space prepending each message, and then type everything in blue?
08:22:51ScoTTie3G and earlier have the button under the screen
08:23:05ScoTTielater models including the mini have the on the actually dial thing
08:23:51midkayamiconn, 3gs have the same number of buttons and button definitions.. but the buttons themselves are under the LCD in a row, instead of around or integrated in the touchwheel
08:24:00amiconnAh, ok
08:24:12***Alert Mode level 5
08:24:12Bluechip BHSPitLappy: I guess they are both client issues - I'm not sure about the 1st, but the second may be changeable if it is distressing to others
08:24:34CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
08:24:34*B4gder strips off IRC colors
08:24:40amiconnI see. 1G and 2G seem to have the buttons around the wheel, instead of integrated into the wheel
08:24:47BHSPitLappy+c would be good :P
08:24:58midkayamiconn, right - 3g = row under LCD, 4g+ = integrated in the wheel
08:25:19amiconnStill, they have the same buttons available
08:25:31 Join Siku [0] (
08:25:35***Alert Mode level 6
08:25:35Bluechip "B4gder strips off IRC colors" ...does that mean I don't have to go hunting through the config to work out if I am able to fix it here?
08:25:45midkayamiconn, right.. it'd be rare, but i can think of cases where different setups would be desireable..
08:25:45JdGordonyour purple now!
08:25:55midkayamiconn, btw - thanks for pointing out the 3g fix
08:26:03B4gderit means I'm saved from people throwing around colors in IRC
08:26:03***Alert Mode level 7
08:26:03Bluechip hm, guess it doesn't
08:26:13*Bluechip is confused
08:26:25***Alert Mode level 8
08:26:25Bluechip if there is something I can do to help, LMK
08:26:29BHSPitLappyJdGordon: "you're" ;)
08:26:40Bg3rBluechip remove your colors :)
08:26:50B4gderBluechip: it is your client that adds the colors
08:27:18BHSPitLappyhe said the client was responsible for the space, not the colors
08:27:24***Alert Mode level 9
08:27:24Bluechip B4gder: OK, I will go look see
08:27:54***Alert Mode level 10
08:27:54Bluechip so the client does the space, not the colour. OK
08:28:01B4gderIRC colors is the invention of the devil
08:28:08B4gder== mirc
08:28:16***Alert Mode level 11
08:28:16Bluechip I shall seek "prepend space to all messages" and remove the tick next to it
08:28:18yeahxso is irc scripting and bots
08:28:42*midkay defends logbot
08:28:49*B4gder pats logbot
08:28:50Bg3rB4gder why don't just add +c to the # mode?
08:29:03midkayB4gder, op him and make him kill yeahx :(
08:29:13B4gderbecause I have no idea what +c does
08:29:13amiconnmidkay: I just tried it. With using IPOD_3G_PAD as an alternative where IPOD_4G_PAD was used before, clock builds without error & warning
08:29:25B4gdera freenode extension I guess
08:29:28midkayamiconn, i just did too - i'll commit a fix now
08:29:45BHSPitLappyBluechip: I doubt it says that :P
08:30:03Bg3rB4gder it should disable using of colour in the channel
08:30:06***Alert Mode level 12
08:30:06Bluechip BHSPitLappy: me too, but I know of nothing else to do
08:30:07midkayamiconn, it's in cvs now
08:30:15BHSPitLappyBluechip: in all seriousness, stop using Trillian and start using a real IRC client
08:30:19JdGordonany1 wanna have a guess as to why plugging in the ac adapter kills the dircache thread on the hxx0 ?
08:30:22BHSPitLappywho has ops in here?
08:30:26BHSPitLappyjust set +c
08:30:45B4gderor get a client that can filter it
08:30:51JdGordonBluechip: get miranda
08:31:16*B4gder loves the build speeds we have these days
08:31:25***Alert Mode level 13
08:31:25Bluechip BHSPitLappy: please excuse me eschewing the "my <x> is better than your <x>" ...trillian serves me well
08:31:33B4gderI made bz2 versions of the sources available now
08:31:56B4gdershaves off more than one megabyte from the package
08:32:05BHSPitLappyBluechip: some of us around here believe in the open source philosophy... you can't deny that our license is better than your license :D
08:32:29***Alert Mode level 14
08:32:29Bluechip BHSPitLappy: I can, but this is neither the time nor the place
08:33:03BHSPitLappyB4gder: remember to clean the floor this time, it was like a carpet yesterday...
08:36:00JdGordonwhats autostart.rock?
08:37:35JdGordonold code rhats not used anymore?
08:37:37midkaybtw - why isn't dircache on the archos models?
08:37:48B4gderit uses lots of ram
08:37:48BHSPitLappybecause you left it out!
08:38:05LinusNJdGordon: well, it's actually code that was never used
08:38:08midkayB4gder, iirc it took like 8kb for my entire ipod disk, i've got quite a few files on it..
08:38:17JdGordonLinusN: what was it gonna be for?
08:38:22LinusNJdGordon: it's supposed to allow a plugin to be run at boot
08:39:11midkayoh. it takes 128kb currently. :E
08:39:15midkaymove along, nothing to see here..
08:39:19midkay*slips back into shadows*
08:39:43BHSPitLappy*where you belong*
08:40:19midkayBHSPitLappy, i know you do, but what do i? (oooooooooooooo)
08:40:23*LinusN just loves 8051...NOT
08:40:50B4gder8051? woooo
08:40:54B4gdersuper powers!
08:41:22midkayBHSPitLappy, drugs are cool man..
08:42:30***Alert Mode OFF
08:50:57markunLinusN: trying to read a book on my iriver yesterday motivated me to work on the viewer again :)
08:51:24LinusNmarkun: good
08:52:11markunLinusN: had to get the paperclip out a few times even :(
08:53:38amiconnJdGordon, LinusN: The autostart feature is used by the one who added it
08:54:27amiconn[IDC]Dragon uses (used?) it to autostart alpine_cdc.rock on his recorder
08:55:21amiconnThe principle is simple: If the feature is enabled in the build, rockbox looks for a plugin called autostart.rock at boot, and loads it if it finds one
08:55:56amiconnMost useful with tsr plugins
08:57:29amiconnmidkay: Dircache can take a lot more than your 128KB, and even setting 128KB aside is a significant waste on archos
08:57:56midkayamiconn, right - i was under the impression it was much smaller
08:58:16amiconnIt takes >200KB here - if I enable it, which I don't do for normal use
09:01:48nudelanyone know if the tag database supports various-artists albums?
09:02:38amiconnLinusN: stripwax wanted to fix the issue with auto-change directory not working while running a plugin.
09:03:04LinusNamiconn: how good for him :-)
09:03:12amiconnTurned out that this feature is a nasty hack - it fiddles with the tree context to find the next music directory
09:03:45amiconnThat doesn't work when running a plugin, because then the tree context is set to browse plugins...
09:04:03amiconnI'm not sure what to do about this
09:04:19LinusNme neither
09:04:29amiconnThere's another problem which can lead to a crash - this code tries to splash()
09:04:39amiconn..when something fails
09:04:52LinusNi believe the playback engine does that too
09:05:09amiconnImagine this happening when a grayscale plugin is running: it will confuse the LCD controller
09:09:04midkaydoes anyone know if there are any docs in the wiki pertaining to bmp2rb during compile?
09:10:45linuxstbmidkay: No, I don't think there are.
09:10:51 Quit quobl_ (Remote closed the connection)
09:10:58midkaymight be a nice addition
09:11:07 Join quobl_ [0] (i=id@tor/session/x-0c27ae77712638e5)
09:11:32midkaydoesn't seem overly complicated, but i did take a look a few nights ago and wasn't sure what exactly needed to be changed/added to put in some new bitmaps
09:12:16 Join petur [0] (
09:13:20 Join Zagor [0] (n=bjst@pdpc/supporter/sustaining/Zagor)
09:13:21amiconnmidkay: If you still have your recorder - cube now runs on it including the solid greyscale mode :)
09:14:46midkayamiconn, yes i do :) it's been like that for a while, hasn't it? or did you just update something? (very cool in any case) :)
09:17:24amiconnIt's like that for a while
09:18:28midkayamiconn, add dynamic lighting ;)
09:18:33amiconn..since 2006-01-23 to be precise
09:19:07midkayhell, i'd barely be surprised if you could manage to get the recorder LCD to show a 24-bit color range.. ;)
09:19:30amiconnI think about adding the other solids...
09:19:42amiconn*platonic solids...
09:20:16amiconnThe plugin name would need to be changed then
09:20:25midkayamiconn, that'd be very cool..
09:22:51midkayamiconn, how about a plugin that lets the user design his own geometry and have rockbox display it? ;)
09:23:18amiconnThe rather simple 3d algorithm should work for any convex shape
09:23:34midkayloadable from files would be cool..
09:24:11amiconni.e. there must not be an indented edge or corner
09:30:27 Join Farpenoodle [0] (
09:31:19yeahxisnt there a doom for rb now too?
09:31:34JdGordondooom???? no.. what gave u that crazy idea
09:31:50yeahxsaw it meantioned a while ago
09:32:19yeahxnever saw it in the plugins, you have to put this in yourself Im guessing
09:32:32yeahxif it even works on the nano
09:32:38 Join safetydan [0] (n=dan@
09:32:57JdGordonit works on the h300... i think ipod as well?
09:33:00BHSPitLappyjust install iPL, dude :)
09:33:11BHSPitLappyiDoom ftw
09:33:28B4gderthe rockbox doom works on ipod too
09:33:53B4gderI'm pretty sure on nano too
09:34:49BHSPitLappyhey, does anyone here have a black 5G iPod?
09:36:05yeahxyeah I might put ipl on it if I can
09:36:15yeahxjust need my cable back
09:36:37midkayBHSPitLappy, i do..
09:36:51BHSPitLappymidkay: does it show scratches like the nano does?
09:37:00midkayBHSPitLappy, yes
09:37:15BHSPitLappymaybe white is the way to go...
09:37:21BHSPitLappyalso, is the battery life crap?
09:37:26midkaythe nano and the 5g both seem to use the same Apple proprietary scratch-inducing coating on their plastic covers..
09:37:42BHSPitLappymy friends' white ones look fine
09:37:45midkayBHSPitLappy, audio playback is fine, and i've managed to watch a lot of video as well - i'd say it's fine for what it does
09:37:55midkaywhite probably shows less than black, i'd imagine..
09:38:04midkayeither way i'd strongly suggest getting a case asap..
09:38:12midkayone that you can leave it in all the time
09:38:27BHSPitLappyyeah, I know
09:38:49midkayconsidering getting one?
09:38:55BHSPitLappyyeah. tomorrow
09:39:08midkayexcellent - i don't think you'll be disappointed
09:39:16BHSPitLappyselling my nano to my mom :)
09:39:16midkayexcept perhaps with iPL..
09:39:25BHSPitLappyso I can keep it around for iPL and rockbox play, yay!
09:39:50BHSPitLappyiPL is the least disappointing venture I've involved myself with :P
09:39:53midkayhaha, "i'm selling it to my mom" aka "i'm getting money for it and then using it whenever i want".. :)
09:40:04BHSPitLappyexaaaactly :D
09:40:04midkayi'm saying it runs rather slow..
09:40:15BHSPitLappyI know :)
09:40:24midkayrockbox runs very well, though..
09:40:26BHSPitLappybut I get to witness all the growth it's experiencing
09:40:32BHSPitLappyhotdog's going to be amazing
09:40:36midkaywhat is it?
09:40:46BHSPitLappythe new rendering and compositing engine
09:40:52BHSPitLappyit runs wicked fast on the 5G
09:40:57midkayanywhere i can read about it?
09:41:05BHSPitLappyit's just a matter of its completion, and applications starting to use it
09:41:31BHSPitLappythe plan is for hd to become integrated into ttk
09:41:33midkaydon't see it at the wiki anyhwere
09:42:21BHSPitLappywell, you can grab the svn yourself
09:42:26BHSPitLappyand look at it from that point of view :P
09:42:30midkay.. something runs realtime?
09:42:38BHSPitLappythe hotdog demo runs
09:42:53midkayi don't have svn installed :( if it's something small, any chance you could fling a copy my way?
09:44:17BHSPitLappyinstall svn then.
09:44:23BHSPitLappyyou on winders or something?
09:44:58midkayi am
09:45:41midkayno builds of it anywhere? :(
09:46:29amiconnsvn is available as cygwin package
09:46:45midkayyes, but that means i need to _install_ it. ;)
09:47:25 Quit needleboy ()
09:48:11nudelif tortoise-svn is like toroise-cvs you can just install a setup.exe you're done. not sure though as i've never used svn either.
09:48:14BHSPitLappycygwin is the easiest to install for, when there's a package :P
09:48:24midkayah, a mere 1mb.. alright :)
09:48:35BHSPitLappycygwin setup.exe is pretty nice
09:48:47 Join ender1 [0] (i=ychat@
09:49:03nudelyeah i was impressed by the cygwin setup when i installed teh rockbox stuff yesterday
09:49:19nudelonly thing i'd change is it's stupid to put the package name miles away from the skip/install thing :)
09:49:32nudelbut when the setup dialog resized i almost hugged my PC hehe
09:49:40BHSPitLappythat's true
09:50:53LinusNi despise the cygwin installer
09:50:55nudelsince it was for a bunch of command-line tools i was expecting a rather hellish installer made of shell scripts and other stuff that would be broken all over the place, but it all just worked
09:51:18BHSPitLappyLinusN: why? it makes a P.O.S. easy to install :)
09:51:48BHSPitLappyof course it just worked, it's windows... everything's compiled for one system
09:51:55LinusNwell, it's partly the installer software itself, and partly their lousy mirroring
09:51:56BHSPitLappyno opts. ever. :D
09:52:06B4gderyes, their mirroring is less than fine
09:52:08BHSPitLappyI have a mirror I like
09:52:16BHSPitLappylucky, huh
09:52:19B4gderyou might need to chase down a mirror that works fine
09:52:22BHSPitLappywhen you just try at random, and find a nice one
09:52:25B4gdersince many are broken
09:52:40LinusNand sometimes the network installation can hang for no reason, and you have to restart the installation from scratch
09:52:41BHSPitLappyI've installed cygwin on both the computers i use at school :P
09:52:43midkayi just tried at random to get subversion.. first one worked great
09:52:47BHSPitLappyjust for X server...
09:52:50BHSPitLappyso I could ssh/X
09:53:12amiconnLinusN: I wonder why you do. Imho the cygwin installer works really similar to hwo a linux package manager works
09:53:14JdGordonhelp.. trying to build gcc for the h300 on linux.. i get this... (in make for gcc)
09:53:14JdGordonconfigure: error: No support for this host/target combination.
09:53:15JdGordonmake: *** [configure-target-libstdc++-v3] Error 1
09:53:41midkayJdGordon, oh, after you left last night, i think it was preglow who said "he wasn't supposed to build for c++" or something.
09:54:05midkayBHSPitLappy, svn is installed - now what? :)
09:54:07JdGordonhmm... i followed the steps from the wiki...
09:54:23BHSPitLappymidkay: /msg iplbot `svn
09:54:25yeahxbur bye
09:54:27 Quit ender` (Read error: 110 (Connection timed out))
09:54:36LinusNamiconn: yes, i guess the only real flaw in the installer is the shaky network handling
09:54:45midkayBHSPitLappy, he's retarded when it comes to `svn.
09:55:00LinusNthe rest of the problems are merely about mirroring and packaging
09:55:16JdGordonthis is the command i used before make.. "../../../gcc-3.4.5/configure −−target=m68k-elf −−prefix=/home/jonno/rockbox/build −−enable-language=c"
09:55:40LinusNmidkay: yup
09:55:56BHSPitLappymidkay: check the wiki then.
09:56:09midkayBHSPitLappy, i searched the wiki for "hotdog". two pages, 0 relevant.. iirc..
09:56:14JdGordonLinusN: have i done something wrong? any ideas?
09:56:17BHSPitLappysearch for subversion
09:56:45LinusNJdGordon: like midkay said, you misspelled −−enable-languages
09:57:05JdGordonoh fuck... bloody missing s
09:57:12JdGordontook 2 hours to compile last night :'(
09:57:13midkayand i think you need an underscore?
09:57:21midkayenable_ instead of enable-
09:57:23midkaymaybe i'm wrong though..
09:57:24*Mikachu hugs tabcomplete
09:57:24safetydanmidkay, no it's a -
09:57:24LinusNmidkay: nope
09:57:25midkayi can't remember
09:57:29midkayah, alright
09:58:26midkayBHSPitLappy, what am i checking out? podzilla?
09:58:58JdGordonok, here we go again.. start the clock!
09:59:04BHSPitLappymidkay: the whole thing
09:59:10safetydanamiconn, I looked at doing oscilloscope first but it used an ISR for timing which is a little out of my area of understanding right now
09:59:13midkayBHSPitLappy, tools?
09:59:20midkaygotcha, ty
10:03:36JdGordonthe list wiget only scrolls the selected line doesnt it?
10:04:40JdGordonand is it available to plugins?
10:06:28LinusNno, it is currently not exported
10:06:34JdGordongrrr... there goes that idea :p
10:06:39LinusNand it only scrolls the selected one
10:06:42midkayscrew this.
10:06:47LinusNJdGordon: why?
10:06:52LinusNexport it then
10:06:54 Join needleboy [0] (
10:06:57midkayno "arm-uclinux-elf-gcc". :(
10:06:58JdGordoncoz im lazy, i wanted to use it ina plugin
10:07:14midkayw/e, need to get to bed - tomorrow i'll try again
10:07:15midkaynite all
10:07:29LinusNwe export functions when they are needed
10:07:40LinusNso if you need it, you export it
10:08:53 Nick ender1 is now known as ender` (i=ychat@
10:09:16 Join _FireFly_ [0] (
10:09:32t0mas[09:57:10] <JdGordon> oh fuck... bloody missing s
10:09:32t0mas[09:57:16] <JdGordon> took 2 hours to compile last night :'(
10:09:40JdGordonno, linux
10:09:46JdGordonmy crappy lappy (TM)
10:09:46t0masall targets?
10:09:53JdGordonjust m68k
10:10:05JdGordon2 hours could be an overstatement
10:10:07B4gderhe's building gcc
10:10:14 Join Xerion [0] (
10:10:19t0masoh ok
10:10:31t0mas2 hours for rockbox builds would be a new world record, even for cygwin :)
10:12:44***Saving seen data "./dancer.seen"
10:14:30peturbuilding the crosscompilers in cygwin is a major pita...
10:17:05LinusNthat's why i made the cygwin packages
10:17:55peturand I thank you for that! now I can test build ipod and archos before committing ;)
10:18:55LinusNi once had the idea to build debian packages as well, but i haven't yet found the time
10:22:20peturLinus, now that I remember: what's up with that H3xx GPIO29 port connected to USB Host ID? What is it?
10:22:53LinusNi have no idea :-) i just got it from the disassembly of the iriver firmware
10:23:23peturwould that be the isp1362 host controller?
10:23:45peturmaybe that's the hardware enable? was it related to usbhost?
10:28:49 Join damaki__ [0] (
10:35:18 Quit yeahx (Read error: 110 (Connection timed out))
10:36:20LinusNpetur: i don't remember
10:36:57peturok... I hope I get the time to look at the disassembly myself
10:39:51 Join youngcereal [0] (
10:42:07youngcerealhas anyone implementad the dock line out g4 patch ?
10:42:37youngcerealon cvs ?
10:43:04 Quit damaki_ (Read error: 110 (Connection timed out))
10:43:08linuxstbWhat is the dock line out g4 patch?
10:43:43linuxstbLine-out should work fine in Rockbox on the ipods without a patch. Are you saying it doesn't?
10:43:59youngcerealno in the dock not
10:44:31linuxstbWhat is your patch?
10:44:37youngcereali think i have sent or not sorry
10:44:48youngcereali have no sleep
10:44:58youngcerealadc driver kill my brain ;)
10:46:18youngcerealone moment i send a mail with patch its easy if you put the ipod to dock with the menu button
10:46:35youngcerealit enabled it
10:47:20youngcereal5 min
10:52:35 Join muesli__ [0] (n=muesli_t@
10:58:20 Join Moos [0] (
11:01:51youngcerealhelp where is the cvs patch command
11:02:16B4gder'cvs diff -u' ?
11:05:43youngcereal linuxstb
11:06:01youngcereal linuxstb: i have send the patch to your email
11:08:12markunpreglow: complaints about the sound quality of the EQ:
11:09:15B4gderI don't know about you, but I never take those MR posts seriously
11:09:39BHSPitLappysomeone know of a way to convert Real audio to mp3 in linux?
11:10:01markunB4gder: Not too seriously, no :)
11:10:15webmindBHSPitLappy, mencoder?
11:10:45 Quit XavierGr (Read error: 110 (Connection timed out))
11:10:51BHSPitLappyI'm not too great with mencoder
11:11:11markunBHSPitLappy: "man mencoder"
11:11:13linuxstbBHSPitLappy: mplayer -vo null -ao wav:file.wav file.rm - and then encode the wav as normal.
11:11:34BHSPitLappyI guess that makes linuxstb the "man" ;)
11:14:25markunHm, didn't know realmedia used so many different audio codecs:
11:14:34BHSPitLappyhmm, I don't know how to save that damn pnm:// protocol :S
11:15:22markunBHSPitLappy: mencoder can
11:15:29markun't handle pnm?
11:15:58linuxstbyoungcereal: I'm sorry, I don't understand your patch. I have used the line-out with a dock in the past (but that dock is now broken), and it worked fine in Rockbox. Also, no-one has complained that the line-out doesn't work, and I know lots of people are using it...
11:16:05BHSPitLappydon't think so
11:16:18BHSPitLappyI got a nice message from real networks while trying to resolve the address
11:17:01markunBHSPitLappy: what's the URL?
11:17:06amiconnLinusN: I have a working display flip for H300 :)
11:17:14youngcereallinuxstb: not on my i have the new dock with ir
11:17:29BHSPitLappy(school assignment...)
11:17:30amiconnI'll lok into some other issues with the H300 lcd driver tonight, then commit
11:17:31LinusNamiconn: nice
11:18:14youngcereal linuxstb: without that i have no sound
11:18:28markunBHSPitLappy: I also get a message that I should upgrade my player..
11:18:46merbananBHSPitLappy: try -dumpstream
11:19:33BHSPitLappymarkun: yeah, message from Real, not mplayer
11:19:59merbanantry this instead rtsp://
11:20:00 Quit qwm_ (
11:20:20markunmerbanan: doesn't work
11:21:07BHSPitLappytried it in VLC
11:21:09markunmerbanan: ah, with rtsp it does work
11:21:15merbananmarkun: works for me, I'm getting a stream.dump file
11:21:22NJoinqwm_ [0] (
11:21:53linuxstbmerbanan: Yes, but it's downloading _very_ slowly for me.
11:22:14merbananlinuxstb: It downloads in realtime
11:22:18linuxstbAnd mplayer refuses to play it.
11:23:07linuxstbmerbanan: I know that - but it's coming in at about 2 kilobytes per second for me.
11:24:36amiconnLinusN:The current lcd handling is a bit strange. E.g. 'display on' uses lcd_init_device(), going through the whole init including port setup, controller reset and all
11:24:38merbananhmm, true, I'll go and bug the mplayer devs with this file
11:24:50amiconnI also have some questions regarding the init sequence
11:24:53LinusNamiconn: yes, it can be made a lot simpler
11:25:08LinusNamiconn: ask iriver
11:25:22amiconnIt looks a bit strange too. Is this just duplication of what iriver does?
11:25:35Bg3rLinusN ah ?
11:25:36warewolfwait for it
11:25:36Bg3rit isn't
11:25:38warewolfwait for it
11:25:45amiconnSome statements seem to have no effect at all...
11:26:04LinusNwell, it was a duplication in my first version of the driver
11:26:15 Join ghode|afk [0] (
11:26:15LinusNi haven't kept track of later changes
11:27:01warewolfis there any possibility of init'ing the remote-lcd (besides just the backlight) when the h300 boots?
11:27:02LinusNi may have made a mistake when i duplicated the settings, though
11:27:17LinusNwarewolf: yes
11:27:51warewolfcool, because that's one of the minor irks I have with rb
11:28:09warewolfanything −− including the boot progress messages would be killer.
11:28:22Bg3ramiconn :
11:28:41Bg3rdidn't have time to try this here
11:28:54Bg3rmore precisely
11:29:06amiconnI'd rather use the datasheet...
11:29:45amiconnWorked fairly well for display flip once I started to understand the meaning of the various settings
11:29:49 Join damaki_ [0] (
11:31:50*t0mas had a speaking test English today
11:31:55t0mas"Defenately the best I've heard untill now."
11:32:07t0massomewhere between 7 and 10 out of 10 :D
11:32:11t0masdepending on the others
11:33:01muesli__t0mas nerd ;)
11:33:08amiconnLinusN: Please compare lcd-h300.c, lines 151 & 152 and lines 145 and 150. The second line of each pair should have no effect...
11:33:27t0masmuesli__: the only language I'm good at :P
11:34:07muesli__what about dutch?
11:34:11amiconnBg3r: These on and off sequences apparently don't uses display standby or sleep
11:34:32t0masmuesli__: for English I have an 8,6 out of 10 now
11:34:37LinusNamiconn: yes, i'd rather use the data sheet too, but my first version of the driver was only a hack to get it running, and copying the iriver setup was the quickest and easiest way to get there
11:34:38t0masfor dutch just a 7
11:36:30amiconnLinusN: Okay. I'll compare the iriver init with what the datasheet says, and try to simplify
11:37:01amiconnRelated Q: Do we want inverse mode on the colour LCDs?
11:37:14amiconnIt's possible, but would look rather strange...
11:37:19amiconnColour negative...
11:38:14amiconnAnother question is whether we want a contrast setting
11:38:32Bg3ri vote for contrast
11:39:04amiconnI can't imagine where this could be useful
11:39:54Bg3rimho it's a nice to have (but definitely not important)
11:40:35t0masmost displays support contract by hardware right?
11:41:56amiconnThe colour lcd controller allows to manipulate the gamma curve
11:44:02 Quit damaki__ (Read error: 110 (Connection timed out))
11:45:03amiconnLinusN: Any news concerning the power issue?
11:45:21LinusNnope, i haven't had time to investigate further
11:45:43amiconnI get around 10 hours out of my H340
11:46:36amiconnDifferent target. I did some experiments with my pullup idea for the OFF button on v1.
11:47:20amiconnEven when forcing PB4 high for a couple of insns and the directly reading after switching back to input, I still get OFF events when the battery is low
11:47:37amiconnI begin to think that this is a hardware feature instead of just a glicth...
11:49:51 Join [IDC]Dragon [0] (
11:51:16*[IDC]Dragon peeks Archos talk
11:51:42amiconnmorning to the peeker ;)
11:52:03[IDC]DragonbutI'm off for lunch soon
11:52:49[IDC]Dragonplease don't remove the autorock, I use it every day
11:53:19[IDC]Dragon(should comment it more clearly, as extra define)
11:55:26amiconnI didn't plan to remove it
11:55:34t0mashm... how hard is it to play a beep from a plugin?
11:55:37webmindhmm was rockbox able to do video btw?
11:55:38amiconnJdGordon asked what it is, and I explained
11:55:47webmindand ifso could it do this on the ipod nano?
11:56:19t0masdoes the ipod nano play video in the apple firmware?
11:56:24amiconnt0mas: check out how metronome does it
11:56:28linuxstbwebmind: No, there is no video in Rockbox for the iPod, only for the old Archos players.
11:56:31linuxstbt0mas: No.
11:57:05webmindlinuxstb, any chance there will be ?
11:57:10amiconnlinuxstb: For the archos recorders and ondios to be precise. Please avoid the word 'player' for anything else than the archos player/studio
11:57:24linuxstbThat's the Player.
11:57:31linuxstbBut I understand your point.
11:57:43webminduhm my archos player didn't have video support
11:57:51webminddue to the crappy display
11:57:55 Quit DJ_Dooms_Day ("Trillian (")
12:00:34 Join BHSPitLappy_ [0] (
12:02:09 Quit BHSPitLappy (Nick collision from services.)
12:02:14 Nick BHSPitLappy_ is now known as BHSPitLappy (
12:07:38JdGordonoh bloody hell.... got errors with gcc again :'(
12:07:45JdGordonmake[2]: m68k-elf-ar: Command not found
12:07:46JdGordonmake[2]: *** [libgcc.a] Error 127
12:07:46JdGordonmake[2]: Leaving directory `/home/jonno/rockbox/build/gcc/gcc'
12:07:46DBUGEnqueued KICK JdGordon
12:07:46JdGordonmake[1]: *** [stmp-multilib] Error 2
12:07:46JdGordonmake[1]: Leaving directory `/home/jonno/rockbox/build/gcc/gcc'
12:07:46***Alert Mode level 1
12:07:46JdGordonmake: *** [all-gcc] Error 2
12:08:13ZagorJdGordon: "Command not found" is hardly a gcc error :-)
12:08:25JdGordonyes, but its while making gcc...
12:08:34LinusNJdGordon: you forgot to set the path to the m68k binutils
12:08:44amiconnLooks like you don't have the m68k binutils in your path
12:08:46LinusN(as the instructions told you to)
12:09:01JdGordongah, hahah, i actually did, but im a dill... i had your path not my path :p
12:09:11CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
12:09:11*JdGordon feels stupid
12:09:19*LinusN giggles
12:09:32*t0mas slaps LinusN
12:09:35t0masI did that too :P
12:09:51*JdGordon doesnt feel so bad now
12:11:15JdGordondoes rockword actually work? i want to make a text editor for rb but wont bother if it does
12:11:19JdGordoni cant get it to work
12:12:10preglowwhy do these people nagging about sound quality refuse to do blind tests?
12:12:42t0maspreglow: because they point out the wrong file as high quality 8/10 times :P
12:12:47***Saving seen data "./dancer.seen"
12:12:58preglowmarkun: there's nothing in that thread that seems to target the eq specifically
12:13:23preglowapart from the occasional "the eq sounds like a bag of shit", which i ignore until they actually give me some information i can use to fix it
12:14:41markunpreglow: "but once i turned it off it was fine."
12:14:48LinusNsomehow people are very trigger happy when it comes to complaining, but are less helpful when you ask them to help you find the problem
12:15:07preglowmarkun: well, that falls into that category
12:15:29preglowfixing an eq based on a subjective statement is all but impossible
12:16:04preglownow, if someone with more flac files than me would try out the new optimisation, i'd be grateful
12:16:17preglowi don't think it should blow up, but like i said, i don't have a lot of flacs
12:16:21markunpreglow: your EQ was very nice this weekend to compensate for the terrible speakers of my gf's mother this weekend
12:16:36preglowlinuxstb: some -8 files actually decode realtime at 21mhz here now
12:17:17preglowmarkun: someone should put in a preamp, though
12:17:24preglowso boosting is more practical
12:17:28preglowand i should finish my grapher
12:17:47***Alert Mode OFF
12:18:23LinusNpreglow: have you tried to feed a clean sinus into your eq and tried to suppress it?
12:18:32 Join linuxstb_ [0] (
12:19:41preglowLinusN: nah, but i guess that should be a no-brainer
12:19:45preglowi'll just have a shower first
12:19:54preglowi based myself primarily on white noise
12:19:59LinusNi wa sthinking about that winamp comparison
12:20:24preglowfor some reason i don't really trust the winamp eq too much...
12:21:51LinusNbut i believe the guy claimed that your eq failed to suppress the 60Hz
12:22:45preglowyeah, he did
12:22:48t0maswe have something to get a plugin in the background right?
12:22:55BHSPitLappythe eq is fun :P
12:23:00BHSPitLappythough it makes the music skip on my nano
12:23:05t0masand also some way to make it take back the focus?
12:23:06youngcerealhi anyone know about the pcf50605/6 ?
12:23:16LinusNt0mas: no i don't think so
12:23:20youngcerealand has a moment time
12:23:33LinusNyoungcereal: what do you want to know?
12:23:43t0masLinusN: no way to make a plugin go in the background for 2 minutes, and then make it pop back up?
12:23:52LinusNt0mas: nope
12:24:03youngcereal LinusN: g4 has 505 right
12:24:11LinusNi believe so
12:24:18youngcerealLinusN: and adc
12:24:56 Join Zoide777 [0] (
12:25:02LinusNwell, adc is a 506 feature iirc
12:25:21youngcerealPCF50605HN/xB/N1 HVQFN56 (SOT684-2) PCF50606, bondout version B
12:25:31LinusNah, no my bad
12:25:44LinusNtouch-screen ctrl is a 506 feature
12:25:49 Join Farpnut [0] (
12:26:24youngcerealipod g4 has i think a pcf50606
12:26:25youngcerealnot 5
12:26:57youngcerealin the datasheet of 05/06
12:27:02t0masLinusN: do you think it will be difficult to make a plugin go in the background, and get it to come back on some event?
12:27:13LinusNt0mas: yes i do
12:27:20t0masI want some sort of timer
12:27:28youngcerealpart manual ist the PCF50605HN/xA/N1 only a 505
12:27:30t0maswhat can a plugin being in the background do atm?
12:27:33youngcerealall other 506
12:27:43t0masI assume it can't catch key presses?
12:27:56LinusNt0mas: no it can't
12:28:02JdGordonyay i have a working setup on linux :D
12:28:20 Join Kohlrabi [0] (
12:28:52preglowBHSPitLappy: it's the combination of the rapid screen updates and the eq calculation that kills it, the eq itself runs quite happily once it's been set and allowed to run without interruption
12:28:59preglowthe filter calculation is pretty hardcore
12:29:25*BHSPitLappy upgrades to a 5G tomorrow
12:29:33LinusNyoungcereal: the ID register may give a clue
12:29:39preglowbut showernow
12:32:20 Quit linuxstb (Read error: 110 (Connection timed out))
12:33:35JdGordonyay, my linux box is slower than cygwin :))
12:33:41JdGordoni mean :'(
12:36:01t0masthat must be a 16 core super machine running cygwin... or a _really_ slow linux box...
12:36:18 Join ashridah [0] (
12:37:08JdGordonthe 2nd
12:37:18JdGordonup to the plugins...
12:37:49JdGordon12min for the whole thing...
12:38:14B4gderthat a 486? ;-)
12:38:23JdGordon120mhz p1
12:38:28B4gderJdGordon: ccache should help a bit
12:38:41B4gderfor rebuilds
12:39:41JdGordonwhats firmware_flash.rock?
12:40:12BHSPitLappyon-the-fly firmware flashing plugin!
12:40:26JdGordonis it?
12:40:44Moosfor archoses only
12:40:45LinusNJdGordon: it is for flashing rockbox on the archos
12:40:45B4gderbut for Archos only
12:41:06JdGordon.. so why did it compile for my h300 ?
12:41:26LinusNJdGordon: beats me
12:41:28BHSPitLappyseems like bad measure, to allow the user to flash without a power source
12:43:22MoosLinusN: any news on the X5 front? :-)
12:43:35JdGordon..ignoring it for the time being
12:43:37t0masLinusN: how do I get a plugin into the background?
12:43:38B4gderBHSPitLappy: without a power source, I doubt Rockbox runs ;-P
12:43:44safetydanpreglow, ep0ch did some EQ measurements the other day (yesterday?) comparing to an Adobe product... 60Hz for peak filter doesn't work basically.
12:44:00preglowdid you measure with white noise+
12:44:04LinusNt0mas: check the battery plugin
12:44:24LinusNMoos: not really, i haven't had much time lately
12:44:33LinusNMoos: the ATA driver works, more or less
12:44:43Moosw00t good to hear
12:47:01 Quit Farpenoodle (Read error: 110 (Connection timed out))
12:48:27 Join IcyStorM [0] (
12:49:05preglowsafetydan: did you get anywhere with the cordic?
12:49:36t0masLinusN: so there's no way for a running thread to do rb->button_get(); ?
12:49:50t0masor can I fetch 1 button press that way?
12:51:03preglowhmm, no, coef calc should survive just fine at 60hz
12:51:37Bg3rt0mas if you do it, u'll eat some button from the UI ...
12:51:50t0masok, that's what I want :)
12:52:06Bg3rand this will confuse the user...
12:52:22t0masin this case, it will not
12:52:27JdGordonshould we be carefull about the size of the plugin struct? im adding the list funcions but wondering if i should mlimit it to only the nescacery funcs to keep the struct smaller?
12:52:36JdGordon.. also coz im lazy :p
12:52:46Bg3rJdGordon add only these you need
12:52:54t0masI want a plugin with 2 settings, sleep interval, button press time
12:53:03t0masthe first is the time between "pings to the user"
12:53:13t0maslike making 2 beeps... and wait for a key to be pressed
12:53:23t0masyou get X (like 30) seconds to press a button if it beeps
12:53:28t0masif you don't, it will shut down
12:53:41t0masthat way you can listen to an audiobook, with the timer on 10 minutes
12:53:44Bg3ri'm not sure that this will work ...
12:54:00safetydanpreglow, I don't know how he measured and I don't have a cable I can use for recording output unfortunately
12:54:00t0masif you fall asleep (as I always do) you will still know in which 10 minute blok you were
12:54:10t0masso you don't have to search for the right position
12:54:11safetydanif only there was a wav writer for the PCM layer
12:54:18Zoide777what happened to the chess game? It's no longer in Browse Plugins. (I'm running a 4g grayscale)
12:54:58t0masBg3r: why not? :)
12:55:17safetydanpreglow, also, I only started looking at it last night so I haven't got anywhere yet. Maybe tonight I might make some tentative attempst. Though could you just use the same trick for (a)sinh as you do for sin, i.e. lookup table and interpolation?
12:55:27safetydanre: cordic that is
12:55:31t0masThe running in the background thing is possible, and you told me I can rb->button_get()
12:55:55preglowsafetydan: problem is that lookup table and interpolatioon isn't sufficient for all tasks, as i've found out
12:56:05preglowit's quite nice for audio, but not for coefficient calculation and such
12:56:39safetydanWould Rockbox benefit from a common set of "fixed-point" math routines? Or would that limit people too much to a particular representation?
12:56:48preglowdepends how you code them
12:56:55preglowmy sqrt function, for example, would be quite sufficient for most tasks
12:57:02preglowyou can set the precision yourself
12:57:08IcyStorMHow to leave Dice on iPod video
12:57:09safetydanI suppose cordic can be variable precision too
12:57:20preglowwell, for sin and cos, it doesn't really need to
12:57:29preglowjust use all the precision you can, and let the user scale to precision of his choice
12:57:33preglowthat's what i do now
12:57:45preglowsin and cos returns a long where all bits are used, LONG_MAX being +1 and LONG_MIN being -1
12:58:05JdGordoncan any1 give me a regexp to get the funciton name from void gui_list_select_item(struct ??
12:58:39Bg3rJdGordon ??
12:58:58JdGordoni gotta replace it with (*func_nbame) and im to lazy to do it manually
13:00:09Bg3ror something similar :)
13:00:17Zoide777so... does anyone know what happened to the chess game?
13:00:28JdGordonBg3r: ye, but hte func name changes which is the problem
13:00:40ashridahthere was a chess game? thought it was just a chess timer
13:00:48safetydanpreglow, well it'll be a good excuse for me to relearn my trig :)
13:00:54Bg3rashridah there is a chessbox too ...
13:01:12Zoide777yeah, chessbox.. where did it go? can't find it in Browse Plugins anymore
13:02:05 Part IcyStorM
13:02:31Bg3rs/.*\([a-z_]*\)\((.*\)/(* \1)\2/
13:02:50amiconnt0mas: Reading button events from more than one thread will cause strange side effects
13:03:05JdGordonta, too slow... lots of mouse clicks later its done :p
13:03:08amiconnMost button loops rely on button sequences
13:03:17t0masso there's no way to make a plugin sleep for X seconds, and then beep and check if the users hits a button?
13:03:30Bg3rJdGordon :) then u're not lazy enough :)
13:04:57preglowsafetydan: the eq seems to have a huge precision problem at 60hz
13:05:09*preglow misses floating point...
13:05:24safetydanpreglow, is it related to the A*Q vs Q for peak filters? Or are you just running out of bits?
13:05:34preglowwait, i've made a 64 bit mistake...
13:05:44preglowi'm just running my eq coef calc on my computer right now
13:05:53preglowand it doesn't exactly look like my code is 64 bit safe, heh
13:06:21amiconnt0mas: Get more interesting audio books ;)
13:06:51t0masdan brown's deception point wasn't interesting enough to keep me awake :)
13:07:09t0masor maybe I just can't listen for longer than a few hours :P
13:08:57 Join Matze41 [0] (
13:13:03t0masbut do you agree it would be a really nice feature for audio books in trains... and planes?
13:13:26amiconnI don't have any audio books, so I don't know
13:14:29 Join zhilik [0] (
13:15:33preglowsafetydan: as far as i can see, the eq coef calculation is fine and dandy even lower than 60hz
13:15:44preglowsafetydan: i'll just have me some food, then i'll check it out
13:16:19*preglow then chooses the correct window
13:16:20JdGordonwhats faster.. working with pointers or indicies for an array?
13:16:29preglowdepends on the compiler and what you do
13:16:39preglowfor modern compilers, the last might be fastest
13:16:40JdGordon.. in general for rb..?
13:16:46 Quit ashridah ("Leaving")
13:17:15 Join Mong [0] (
13:17:17 Quit Mong (Remote closed the connection)
13:17:48t0masamiconn: more low level idea...
13:18:00amiconnMy experience is that using pointers is faster, and more compact
13:18:04t0masI can wakeup my thread, pause audio playback
13:18:21t0masthen wait for a queue change, without the sleep_thread() the kernel makes in between
13:18:51t0masthen I'm sure I'll get the next button event...
13:19:08t0masafter that, also eat the release event
13:19:17t0masand then sleep my thread again
13:19:44t0masor, if I don't get it in X seconds, shutdown.
13:21:27t0masor would that cause major trouble somewhere else? :)
13:24:54 Join tucoz [0] (
13:25:08tucozI am playing around with screenshots for the manual
13:25:20amiconnYou can't wait for the event without yielding
13:25:34amiconnThis will interrupt playback within milliseconds
13:25:46t0masthat's not a problem, I'll pause the playback anyway
13:25:55t0masbut can the button driver post the event w/o yielding?
13:26:06t0mas(read: is it done in the interrupt handler?)
13:26:33tucozand I wonder what you think of the following setup. 1) In each chaper directory, we have a images directory
13:27:16tucoz2) the naming of the images is ss-cube-160x128x4.png
13:27:57tucoz(where ss=screenshot, cube is the name of the screenshot, and 160x128x4 is the resolutionxcolor-depth
13:28:33 Part Bluechip
13:28:44tucozAnd in the includegraphics statement, this will be called as \includegraphics{plugins/images/ss-cube-\genericimg.png}
13:29:31tucozwhere \genericimg is defined in the platform file, and point to the resolution of the actual target
13:29:53JdGordondoes this look right? it should loop through 0->(MAX_LINES-1) starting from somehwere in the middle but checking every item once?
13:30:21tucozOr should we call the \genericimg something else?
13:30:59 Quit perl|wtf (Read error: 110 (Connection timed out))
13:31:39tucozIf it is a screenshot or a picture of the target, we could simply use the \playertype macro for that.
13:34:32 Join perldiver [0] (
13:36:29preglowsafetydan: eq works fine at 60hz, i can even hear it with my ears
13:36:53LinusNJdGordon: the i=last+1 assignment must use the modulo as well: i=(last+1)%MAX_LINES
13:37:50JdGordoni realsied it wouldnt work if i==-1, so ive changed it to do while which i tihnk works
13:40:13tucozlinuxstb_, are you around?
13:42:05tucozanyway, the \genericimg i used because of the fact that several players share screenshots (same resolution, bitdepth)
13:42:28t0masJdGordon: that if (start == -1) shouldn't that be if (start < 0) ?
13:42:33t0masor is -2 a good value?
13:42:34tucozBtw, is it 18 or 16 bits that is used by the h300 and ipodphoto?
13:42:45JdGordont0mas: oops, your right
13:42:57t0masand if you define a start > MAX_LINES
13:43:05t0masthis will be an endless loop right?
13:43:36JdGordonif ((start <0) || (start >=MAX_LINES))
13:43:44B4gdertucoz: we have only 16 bit targets
13:43:52B4gderI mean, no more than 16 bits
13:43:56t0masthat would be a solution
13:44:10t0masbut what are you tying to do? loop through all lines starting at start?
13:44:16tucozB4gder, ok. great. I was wondering because of the data in the devicechart
13:44:26B4gdertucoz: yes, that date is misleading
13:44:47safetydanpreglow, I think ep0ch said that it's not cutting/boosting as much as the gain value would indicate
13:44:56JdGordont0mas: ye, but starting in the middle somehwere
13:45:08safetydanBut that may be the A*Q thing actually now that I think about it
13:45:13t0masand you want it to go through the first ones if it doesn't find a new one after start?
13:45:18t0masor just go from start to the end?
13:45:29tucozB4gder, Ok. Do you think I should just get working on the concept I mentioned ^^, or wait?
13:45:42JdGordonfirst one
13:45:52t0masah ok
13:46:00B4gdertucoz: sounds fine to me, I think you could go with that
13:46:11tucozB4gder, ok. cool.
13:46:29JdGordont0mas: most of the time it should hit the first time...
13:48:27t0masI don't know if this is any faster... or better...
13:48:30t0masbut I would write: if ((start <0) || (start >=MAX_LINES))
13:48:37t0masthat one
13:48:42t0maswrong paste command :)
13:50:02JdGordonpretty much the same as my fixed version.. cept i get to say i finnaly got to use do {} while () in a real program :p
13:50:43t0masmaybe that's why I would do it another way :)
13:52:39tucozhmm. Is it possible to take screenshots in the sim?
13:53:26B4gderor F5 (iirc)
13:53:27 Join ep0ch [0] (
13:58:01 Quit BHSPitLappy (Read error: 110 (Connection timed out))
13:58:32tucozand what do you use to convert bmp to png? command line tool would be nice ;-)
13:59:03B4gder'convert' from imagemagick
13:59:37tucozB4gder, ah. Thanks
14:00:43B4gderit really is a magic tool
14:01:45 Join banan_ [0] (
14:04:40tucozB4gder, hehe. That you can say.
14:10:28 Join damaki__ [0] (
14:12:49***Saving seen data "./dancer.seen"
14:15:48amiconnLinusN: Did you see my optimised archos button check?
14:16:02LinusNlooked ok
14:16:16amiconnI used a similar binary-tree if() structure as on the newer targets
14:16:38LinusNbtw, the iriver remote control detection must be revised
14:16:44amiconnI'm thinking about rewriting the other target's if(9 trees as well
14:17:06amiconnThe binary tree should be balanced, with one important exception
14:17:08LinusNit turns out that the a/d value is way off if the hold button is on
14:17:38Bg3rLinusN ah, i noticed this when i plugged my non-lcd remote ...
14:17:39amiconn..which I realised when doing it for archos: The most common case (no button pressed) should be catched by the first if
14:18:04 Quit merbanan (Read error: 110 (Connection timed out))
14:18:20lostlogicI notice that we use 8 bit rather than 10 bit ADC mode on the iRiver, could this be related?
14:18:20amiconnBefore, no button pressed needed 4 if()s on archos. Now it needs just one
14:18:28LinusNlostlogic: no
14:18:55amiconnLinusN: How do you mean? If hold is non, the values don't need to be checked..
14:19:09LinusNamiconn: not entirely correct
14:19:25LinusNthe a/d is only read when inserting the remote
14:19:57LinusNthus the detection fails
14:20:13amiconnAh, you're talking about the type detection...
14:20:14LinusNand it isn't re-detected when the hold is switched off
14:20:33amiconnI can't do that. I only have a h100 remote
14:21:10LinusNthe iriver firmware reads the remote type all the time in their button handler
14:21:25LinusNamiconn: can't do what?
14:21:35amiconnRework the remote type check...
14:22:17LinusNah, no, but you can see for yourself how it fails
14:22:42LinusNi'm not entirely sure how to handle the problem
14:22:56amiconnAnother related thing: I've implemented a better debounce check (for irivers atm), that avoids triggering on transients
14:23:04LinusNi guess we'd have to monitor the hold button changes and redetect
14:23:06amiconnYou know, my H340 has mushy buttons...
14:23:19LinusNamiconn: good
14:24:03amiconnI implemented a last-value tracking directly in button_get(), and only accept the adc value if it's within +/- 1 range of the previous reading
14:24:15amiconnWhat do you think?
14:24:54preglowep0ch: here?
14:26:40LinusNamiconn: sounds good
14:27:05LinusNbtw, it seems like the pcf50606 i2c timing might be a tad too aggressive
14:27:30LinusNso the adc button reading sometimes fails
14:27:52 Quit damaki_ (Read error: 110 (Connection timed out))
14:28:00amiconnThat might also explain the problems I'm having...
14:28:13LinusNit only happens when boosting
14:28:21amiconnCould you verify this?
14:28:38 Quit ScoTTie ()
14:28:58amiconnAlso: did you catch my comment about OFF on recv1?
14:30:06LinusNyes, that the pseudo-pullup didn't work
14:33:54*JdGordon is brilliant :D
14:36:27 Join Jungti1234 [0] (n=jungti12@
14:37:01Jungti1234What is 'Prerecord'?
14:37:44 Join ScoTTie [0] (n=scott@unaffiliated/scottie)
14:37:57preglowsafetydan: i think all this is just a misconception, really, bands begin to get so narrow at lower frequencies that the cut might be observed to be lower than it should be
14:38:33peturJungti1234: it lets your recording start a certain amount of time before the moment you press the record key
14:39:57preglowsafetydan: thanks to the low freq bin granularity of most spectrum analysers at low frequencies
14:40:19preglowi can't find anything wrong with the eq in any case
14:40:34safetydanpreglow, quite likely to be something like that as there's a lot misconceptions about what the EQ does and doesn't do
14:40:44Jungti1234petur: sorry, I don't understand.
14:41:22 Quit JdGordon (Read error: 104 (Connection reset by peer))
14:42:26peturJungti1234: rockbox is constantly buffering the signal that comes in, and when you press 'record' it also saves the 30 seconds before. the 30 seconds is configurable
14:42:56preglowsafetydan: i think a proper bandwidth parameter instead of q might clear a lot of it up
14:43:16preglownow if i only could think of a nice way to handle the gigantic dynamic range of sinh
14:43:51Jungti1234Why is it 'Prerecord'?
14:44:16B4gderbecause pre means before
14:44:47Jungti1234Meaning of 'Prerecord' is 'Record beforehand'.
14:45:00preglowbut i think i'll try to finish the eq plotter first
14:48:20tucozOk, I have made some changes to the manual. But, I have updated my ubuntu and the CVS info seems to be lost. I get a cvs add: cannot open CVS/Entries for reading: No such file or directory cvs [add aborted]: no repository
14:48:20 Quit Zoide777 ("CGI:IRC (EOF)")
14:48:43 Join Zoide777 [0] (
14:49:44tucozgrr, is it possible to revert to the local changes if I made a cvs up -dP ?
14:50:17 Quit Zoide777 (Client Quit)
14:52:10Bg3rtucoz are you sure u're in the right directory ?
14:52:33tucozI am not in the rockbox-devel, but in rockbox-devel/manual
14:54:43tucoznevermind, I needed to add the new directory :)
14:57:54 Quit perldiver (Read error: 110 (Connection timed out))
15:00:09markunJungti1234: yes
15:01:50Jungti1234markun, are you busy?
15:02:07markunnot very much, why?
15:02:40Jungti1234I updated korean language file.
15:02:48*tucoz realizes that additon of screenshots will be a tedious task
15:03:40tucozhopefully, the concept to do it is not that complicated and that patches will be submitted :)
15:04:48Jungti1234marknu: Do I upload it to Flyspray?
15:04:55Jungti1234oh, markun. :)
15:06:39 Join muesli- [0] (n=muesli_t@
15:09:27markunJungti1234: just send it over IRC
15:11:27 Part tucoz ("Leaving")
15:12:50 Nick banan_ is now known as merbanan (
15:13:18^BeN^how can i fill in white in archos?
15:13:54 Quit Kohlrabi ("Fast alle Menschen sind Regenwürmer")
15:14:27lostlogicAny reason not to make generic the pcf50606 and pcf50605 adc driver code?
15:16:45LinusNlostlogic: because 50605 doesn't have an adc?
15:17:38lostlogicLinusN: ... interesting that you say that considering the 50605 adc is how we read battery on the ipod :-P
15:18:06LinusNthe datasheet clearly says that 50605 doesn't have an adc
15:18:24LinusNpage 8
15:18:28lostlogicI was pretty sure it was only missing the MBC battery and therefore short 1 adc
15:18:32lostlogic*looks at page 8*
15:19:05lostlogicah, mine's not in color, so the blue vs. green doesn't help :(
15:19:32youngcerealand in the same register were get the batvolt
15:19:49youngcerealare touchscreen elements
15:19:55lostlogicbut regardless ... I'm quite certain that the battery readings are coming from the pcf50605 on ipod, so someone's lying
15:21:28youngcerealhow has sayd that there an pcf50605 is in ipod
15:23:48 Join larc [0] (
15:25:46 Quit Jungti1234 (Read error: 104 (Connection reset by peer))
15:26:04 Join Jungti1234 [0] (n=jungti12@
15:30:11 Quit needleboy ()
15:30:49Bg3rbye, Jungti1234
15:31:02 Quit Jungti1234 ()
15:39:42muesli-just ouf of curiosity: what improves the new clock code?
15:40:27amiconnLinusN: (recv1 OFF) I begin to think this is a hardware feature rather than a glitch
15:40:39amiconnAny idea how to check this?
15:41:03amiconnYes, to prevent the batteries from deep discharge
15:42:52LinusNamiconn: wouldn't it be better to turn off the player instead of sending fake OFF events (which the original firmware ignores)
15:43:20amiconnMaybe the feature itself is buggy.
15:43:33amiconn...or it's because of flaky battery connections
15:47:05LinusNamiconn: my recorder doesn't have that problem
15:50:52youngcerealhas anyone the speedstep patch or is it on flayspray
16:04:15 Quit larc (Read error: 113 (No route to host))
16:07:07markunyoungcereal: what is "the speedstep patch" ?
16:07:46youngcerealfor g5 cpu speed
16:08:30youngcerealcpu frequency scaling
16:12:50***Saving seen data "./dancer.seen"
16:17:27webmindcan anyone tell me why rockboy doesn't work on the nano?
16:21:28B4gderI believe because nobody adjusted it
16:22:17 Join SereR0kR [0] (
16:24:27 Quit DreamTactix291 (Read error: 104 (Connection reset by peer))
16:24:48 Join DreamTactix291 [0] (
16:25:12 Join fairway [0] (
16:25:49peturhow's the snow?
16:26:03muesli-got a great line ;)
16:26:04fairwaynot bad
16:26:07 Quit _FireFly_ ("Leaving")
16:28:43 Quit fairway (Client Quit)
16:30:39preglowlinuxstb_: do we need endian swapping lots on the ipod?
16:33:59 Join _FireFly_ [0] (
16:34:29amiconnLinusN: Do you think it might be helpful to connect a scope to PB4 and watch the signal when running on low bat?
16:34:44 Join Kohlrabi [0] (
16:35:14 Join needleboy [0] (
16:35:23 Quit needleboy (Client Quit)
16:35:29 Join needleboy [0] (
16:35:38Bg3rbtw, nobody added the Lang v2 to the discussion list for the devconf ?
16:35:56B4gderits not that much to discuss about it, is there?
16:36:19Bg3rso it just needs to be done ? :)
16:36:40peturare there plans to have a battery low event? Would be nice to prevent issues like power loss during recording causing file corruption
16:37:41 Join youngcerea1 [0] (
16:37:51B4gderpetur: sounds a like a good idea to me
16:38:24 Quit youngcereal (Read error: 110 (Connection timed out))
16:38:30peturthe hardest part of course is defining the voltage limit...
16:40:36[IDC]Dragonpetur: doesn't the Archos have that?
16:40:47peturno idea
16:41:03[IDC]DragonI vaguely remember
16:50:41 Join webguest96 [0] (
16:52:16webguest96lostlogic, I get pops/artificats when skipping tracks and when playing from pause, this occurs with optical out as well. also on powerup I get a pop
16:52:45webguest96I have tried several recent builds, all exhibit same artifiacts
16:53:14webguest96err forgot to mention I have a H140
16:55:14amiconn[IDC]Dragon: It doesn't have an event, but it has a function to check for low bat
16:55:37amiconnIt's even exported to the plugin api, and your flash plugins use it as a safeguard
16:57:48[IDC]Dragonah, yes ;-)
16:58:10[IDC]DragonI did remember it's not an event
16:59:07[IDC]Dragonpetur: so just implement that, if not already done
17:04:29 Quit quobl_ ("Leaving")
17:05:06 Quit OneFix_Work (Remote closed the connection)
17:05:16 Part webguest96
17:05:25 Part LinusN
17:08:00 Quit B4gder ("time to say moo")
17:10:17 Join chitin [0] (n=20080218@
17:10:23chitinmake it for x5
17:10:24 Quit chitin (Client Quit)
17:10:30preglowyes, yes
17:10:31ep0chpreglow: yo
17:10:40preglowep0ch: did you do any eq measurements?
17:10:56ep0chyeah i did briefly yesterday with some sine waes
17:11:16ep0chPF filters don't seem accurate at low frequencies
17:11:22preglowi think they are
17:11:51ep0chdid you see the results i posted yesterday?
17:12:05preglowmeasurements on white noise is also nice
17:12:10preglowbut i've gotta go and meet someone now
17:12:13preglowi'll be back later
17:12:42ep0ch15.01.32 # <ep0ch> safetydan: 60 Hz sine wave can't filter with a peak filter, 200 Hz sine wave filters with PF@110 Hz, 440 Hz sine wave filter with PF@410 Hz
17:13:28ep0chpreglow: ok
17:18:40 Quit Rob2222 (Read error: 104 (Connection reset by peer))
17:19:26 Join Rob2222 [0] (
17:19:43 Quit YouCeyE (Read error: 110 (Connection timed out))
17:26:16 Quit Shani}{Coder (Read error: 110 (Connection timed out))
17:29:59 Join damaki_ [0] (
17:30:49 Quit safetydan ("Leaving")
17:40:14 Join virtualball2 [0] (
17:42:14 Quit _FireFly_ ("Leaving")
17:43:51 Quit damaki__ (Read error: 110 (Connection timed out))
17:46:05 Join harbel [0] (
17:46:11 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
17:47:48 Join webguest77 [0] (
17:51:59webguest77i've got a question
17:52:08webguest77concerning rockbox on my ipod mini
17:54:16muesli-will rockbox ever support that subwoofer`? :o
18:00:25 Quit webguest77 ("CGI:IRC")
18:03:40preglowep0ch: what q factor did you use?
18:03:48preglowep0ch: and what gain/cut?
18:03:53ep0chwell i tried various
18:04:09preglowfilters get really narrow at lower frequencies
18:04:19pregloweven at pretty low q
18:04:31ep0chgain was -24db, Q from 0.5 to 6.4, used 6.4 most of the time to get the accuracy
18:04:48preglowwell, it's hard to say
18:05:01preglowmy filters are pretty much exactly what they should be
18:05:26preglowi need to find that recording cable again
18:06:26ep0chwell if you generate lots of sine waves and loop them you can pretty much tell just by listening if its affected by the eq
18:07:25preglowi did lots of tests with peaking filters and white noise
18:07:34preglowand i always got a correct center frequency in the frequency plot
18:08:12ep0chi think its just for < 500 hz that its not too accurate
18:08:25preglowit should be completely accurate there as well :/
18:08:41ep0chmy 1000, 2000 and 4000 hz test seemed spot on
18:09:15ep0ch(ps 4000 hz is not good to listen too)
18:09:36lostlogicHmm... so we should setup a tick task that updates a time_t seconds value rather than using the RTC to get the time during runtime, with perhaps an occasional sync with the RTC?
18:10:17 Quit petur ("dada")
18:11:20preglowisn't current_tick enough?
18:12:31lostlogicpreglow: have to do lots math on every get_time call if basing the time on current tick, that happens ~20hz, seems better to do one if check 100hz and less math 1hz than more math 20hz
18:12:54***Saving seen data "./dancer.seen"
18:13:55 Quit mikearthur ("Konversation terminated!")
18:14:50lostlogicI was thinking if (! (current_tick % HZ) ) { current_time++; tm = localtime(current_time); }
18:15:08preglowep0ch: hmok, you might have something here yes
18:16:28 Join mikearthur [0] (
18:17:13preglowi tried making a filter at 60hz, and it ends up at 45hz
18:17:33preglowno, 50
18:18:11 Join Paul_The_Nerd [0] (
18:19:20preglowoh, no, the error is far larger
18:19:32preglowok, i'm willing to admit there's a problem at lower freqs now
18:19:43 Part harbel ("Leaving")
18:20:09preglowbut luckily, i think i know the problem
18:20:49preglowman, you'd really be surprised how sensitive these coefficients are
18:20:50 Quit virtualball2 ("virtualball2 has no reason")
18:22:39ep0chbah i had my sine waves on loop, now rockbox has crashed for the first time in ages, it just says "loading", disk is spinning, and i cant shut it down, grrr
18:23:19 Quit muesli- (Read error: 104 (Connection reset by peer))
18:23:32preglowyes, suspicion confirmed, the problem is i need a better sin() function
18:25:25preglowthe eq coef calc will be even slower, though :/
18:25:42preglowi could always just use a bigger sin table...
18:27:13 Part [IDC]Dragon
18:29:01preglowmaybe i should just try a quickndirty taylor series
18:29:22 Join xmixahlx [0] (n=xmixahlx@
18:30:47 Join actionshrimp [0] (
18:31:33preglowep0ch: but yeah, on a lighter note, when i do have a better sin, the eq graph plotting will also be closer
18:31:54lostlogicImporting the offtime function from glibc, good idea bad idea?
18:32:39preglowamiconn: are you completely certain the memset16 test in your memtest plugins is correct?
18:38:40 Join Nico_P [0] (
18:41:15 Join webguest46 [0] (
18:45:05webguest46ach man, leute... ich frag einfach mal so rein: funzt rockbox aufm ipod mini, oder nicht???
18:45:58markunwebguest46: nein, but it's better to ask in English
18:47:04webguest46so, is there any possibility that rockbox will be released for the ipod mini?
18:47:20markunSure, but I don't know if anyone is working on it
18:48:33webguest46i'm looking for an alternative os for such a long time, but everything i find is not compatible with my ipod mini...
18:49:20markunipodlinux didn't work?
18:49:28webguest46not for my mini
18:49:42Mikachuweren't there some commits for mini recently?
18:49:45webguest46and now rockbox doesn't work either
18:49:55webguest46i actually don't know
18:50:31preglowwebguest46: it's on the way
18:51:00 Join damaki__ [0] (
18:51:10webguest46so it will be released some time?
18:51:27webguest46lets say within the next 2 months?
18:52:58webguest46and how can i use this "firmware/export/config-ipodmini.h"-file??
18:53:01preglowreleased, no
18:53:05preglowbut working, yes
18:53:21webguest46that's what i meant
18:53:39Mikachui just wanted to show it's in the source tree, the file alone isn't a lot of fun
18:53:43 Quit actionshrimp ("( :: NoNameScript 4.01 :: )")
18:54:25^BeN^how can i fill in white in archos?
18:54:39webguest46so it's in the source tree
18:55:03webguest46and how can i make use of this file?
18:55:24webguest46are there other files or even a intruction to install this file?
18:55:39markunwebguest46: you can't install the file, just be patient
18:56:42markunwebguest46: why doesn't your ipod mini work with ipl btw? Do you have a 2nd generation?
18:57:20markunwebguest46: there are a lof of screenshots for a device that doesn't work...
18:57:25Mikachuwow, the clock really feels overkill now, it has its own menu system why exactly?
18:58:24webguest46it really works?
18:59:12webguest46i guess i should have looked it up before writing stupid things...
18:59:34webguest46but let me resume
19:00:22webguest46rockbox for the ipod mini isn't developed yet, but it will be the next time
19:00:49webguest46and now i just have to wait and be patient... is this correct?
19:06:06 Nick qwm_ is now known as qwm (
19:07:51 Quit damaki_ (Read error: 110 (Connection timed out))
19:10:02lostlogicblurgh, i see why we were just reading from the RTC... converting RTC->secondssince1970 is significantly untrivial.
19:10:33webguest46what do they mean with "ipod Mini support" (
19:13:38Mikachuit means bejeweled should work on ipod mini
19:15:13 Join Shani}{Coder [0] (i=Coder@Is.A.TclCoder.Org)
19:15:46amiconnpreglow: What about the memset16() test?
19:17:58^BeN^amiconn, how can i fill in white in archos?
19:18:54^BeN^i want to fill place in white, not in black
19:19:02 Nick ^BeN^ is now known as Paprica (
19:19:07 Join Benacool [0] (
19:19:20amiconnIt's simple, there's a lot of places where this is done
19:19:44amiconnSet drawmode to (DRMODE_SOLID|DRMODE_INVERSEVID), then use lcd_fillrect()
19:19:45Papricaok i will take a look
19:19:51Benacoolpyro: are you there ?
19:20:13 Join safetydan [0] (
19:21:12 Join damaki [0] (
19:22:03*amiconn forgot to ask something when Linus was around :(
19:27:01preglowamiconn: just wondering if you're sure it's correct
19:27:12 Join Zoide777 [0] (
19:27:15safetydanCan we actually display 18-bit colour on the H300 screen?
19:27:15preglowit insists there's a bug in my memset16
19:27:18preglowon arm
19:27:44amiconnpreglow: I'm quite sure it is. Did you run it against C memset16() ?
19:27:49Zoide777sorry to ask again, but does anyone know why he chess game isn't in Browse Plugins anymore?
19:28:20amiconnsafetydan: I admit that adapting oscilloscope for swcodec *cleanly* isn't as simple as I thought. There's a quick-n-dirty SH1 asm leftscroll routine in it.
19:28:33lostlogicamiconn: you wouldn't by any chance be interested in helping out with importing struct tm -> time_t -> struct tm conversion functions from somewhere into rockbox, would you?
19:28:39amiconnWe need proper scrolling for 1 and 2 bit targets in xlcd first
19:29:45amiconnlostlogic: I think these functions would be fairly complex. We might be better off with simply throttling RTC access
19:29:52amiconn(like you already did)
19:30:09amiconnPerhaps we could go down to once per second?
19:30:37lostlogicamiconn: yeah, I'm trying to figure out hwo to do the throttling without making the seconds flow inconsistently on the WPS (if you have my rtc in wps patch and show seconds)
19:31:32Zoide777is there a way of recording in the ipod? I saw that ipodlinux allows recording by plugging in a mic into the headphone jack. Plus it seems from the latest CVS updates that Rockbox has ADC support for the ipod.
19:31:33safetydanZoide777, do you have a .rockbox/rocks/chessbox.rock file?
19:32:01lostlogicwas thinking that using the tick timer to flag when the tm struct needs updatin gmight be a workable solution
19:32:03Zoide777safetydan: no. The daily builds used to have it, but apparently not anymore. (4g grayscale)
19:34:13 Quit damaki__ (Read error: 110 (Connection timed out))
19:34:30safetydanamiconn, yeah the scrolling was what stopped me. I'm sure I could have bodged together something but I was just doing the easy things at the time.
19:35:12 Quit webguest46 ("CGI:IRC (EOF)")
19:35:24amiconnI'll look at that. I've done practically all scrolling routines in rockbox anyway (xlcd for 16 bit, grayscale library)
19:35:38preglowamiconn: i did, and that seems to work ok
19:35:39amiconnA moderate challenge is to provide a C alternative for the sims
19:36:12preglowsafetydan: seems we need a good and accurate sin() for the eq itself too
19:37:14safetydanpreglow, good, accurate, and fast... pick two? :)
19:37:38safetydanZoide777, I just compiled the ipod 4g sim and it includes the chessbox plugin... weird that the daily's miss it
19:37:54amiconnpreglow: Then I'm pretty sure your memset16() is wrong.
19:38:35amiconnI had a similar effect with my coldfire version. Turned out I just overlooked to change a move.b into a move.w (I based memset16 on memset of course)
19:38:47Zoide777safetydan: That's very strange. I swear it's not in the .zip build! (
19:39:53preglowsafetydan: in this context accurate will have to do
19:39:54 Join IcyStorM [0] (
19:40:20preglowsafetydan: a bigger 32 bit lookup table would do the job nicely, but i'm not too fond of huge lookup tables
19:40:21amiconnlostlogic: What about this idea:
19:40:45safetydanZoide777, I just compiled an ipod4g build here and it was included in that. So... I've no idea then. Try a CVS build maybe?
19:41:53Zoide777safetydan: Well, isn't the file a cvs build? (the "bleeding edge" build from the CVS Builds page)
19:42:17Zoide777b/c doesn't have the chess game
19:42:43amiconnChange the test in get_time to if (last_tick + HZ < current_tick), and then don't set last_tick = current_tick in the body, but
19:42:58amiconnlast_tick = 100 * (current_tick / 100)
19:43:20amiconnbetter: last_tick = HZ * (current_tick / HZ)
19:44:22amiconnThe tradeoff is one division...
19:44:35amiconn...but only once per second
19:44:45safetydanZoide777, yes that's the one. I am now officially out of ideas. Bagder might know more.
19:48:09Zoide777safetydan: ok, thanks anyway.
19:48:25 Quit Farpnut (Read error: 104 (Connection reset by peer))
19:48:42 Quit Kohlrabi (Nick collision from services.)
19:48:52 Join Kohlrabi [0] (
19:49:24 Join jocke1s [0] (
19:50:17 Join mozetti [0] (
19:50:28Zoide777safetydan: it's a bit scary though... i wonder if there are other differences between the compiled-at-home CVS build and the automatized .zip that get puts on the website
19:51:53jocke1shi, I am running linux and have a 30gb ipod (bought it today). Can someone point me to install instructions ?
19:53:48jocke1sthanks. Will it work since I am not using windows at all?
19:54:52ghode|afkah files look like windows only, but there must be a way
19:55:43 Part IcyStorM
20:00:15 Join charkins [0] (n=casey@pdpc/supporter/active/charkins)
20:00:41charkinsYa'll rock
20:00:45Paul_The_Nerdjocke1s: The ipodpatcher program can be replaced with dd if you're familiar with using it for the ipodlinux installation. The ipod_fw utility can be compiled for linux, and its source is available in the rockbox source package I believe.
20:00:59amiconnt0mas: Something is fishy in the distributed build system...
20:01:24Mikachujocke1s: it's a lot easier under linux than under windows
20:01:47t0masamiconn: what's wrong?
20:02:00jocke1sit is. Think you can point me to how to get started :)
20:02:02amiconnt0mas: chessbox.rock is zero-byte in the automated build dirs (ipod), so it's not included in the zip.
20:02:25amiconnIf I try the very same build on the very same machine just from the commandline, everything is ok
20:02:41charkinsjockels: from the rockbox/tools/ directory, run 'make ipod_fw'
20:02:53amiconnThat's not only on my machine, but also on others
20:03:12amiconnE.g. the last ipod4gray build was done elsewhere, but has the same problem
20:03:34t0mascan you try to build it with make -j2 on your machine?
20:03:53amiconnI already tried make -j
20:04:01amiconnSame result; I'll just try -j2
20:04:19Slasherihmm, make -j just means make -j1
20:04:36Mikachuum, no
20:04:42Mikachumake -j means unlimited subprocesses iirc
20:04:45amiconn-j means "use as many jobs as make sense"
20:04:51lostlogicamiconn: yes.
20:05:02Slasherioh, interesting
20:05:12jocke1scharkins: What should I download to get the tools dir? (CVS?)
20:05:15amiconnt0mas: make -j2 triggers the problem...
20:05:28*amiconn wonders why...
20:05:33t0masthen something is wrong with the makefile
20:05:39charkinsjockels: you can grab a cvs snapshot (, lemme find a link
20:05:46t0massome depency is wrong
20:06:01t0masthe chessbox thing...
20:06:09t0masdoesn't that use it's own makefile?
20:06:10amiconnAhahaha, yes indeed
20:06:18t0masmaybe that one is not 'make -j' safe
20:06:21amiconnIt tries to objcopy before linking...
20:07:00jocke1scharkins: Ok on its way
20:07:19 Join bluey [0] (
20:07:36charkinsjocke1s: new to rockbox myself (just bought an ipod and installed rockbox this morning)
20:07:36 Quit Zoide777 ("CGI:IRC (EOF)")
20:07:44 Join Zoide777_ [0] (
20:07:45charkinsbut it was pretty painless to get going
20:08:06linuxstb_lostlogic: I read in the logs that you've been talking about the pcf50605, but do you know that the 5g ipod has a pcf50607?
20:08:13 Nick linuxstb_ is now known as linuxstb (
20:08:17jocke1swell rockbox was what made it possible to get
20:08:25amiconnt0mas: I even know why...
20:08:49t0masok, then fix it :)
20:08:50charkinsjocke1s: agreed, everything is ogg/flac here
20:09:07jocke1scharkins make worked. What next :)
20:09:09lostlogiclinuxstb: no, I didn't know that. so ... we have a pcf50606 driver for the pcf50606 in H3x0 and X5 and a pcf50605 driver for the 05 and 07... *sigh*
20:09:22lostlogiclinuxstb: either way, we can unify the ADC controllers between them.
20:09:42charkinsjocke1s: follow these (from step 2)
20:09:43amiconnThe archos loader plugin must not even tried to build for iriver/ipod. Otherwise there'll already be an (empty) chessbox.elf which is then used to make chessbox.rock
20:09:52linuxstbI haven't tested your code on my 4G yet. I'll need to make similar modifications to the audio driver (something else that needs unifying...)
20:09:53amiconn...before the correct chessbox.elf is created
20:09:53jocke1sok. Thanks!
20:10:14charkinsjocke1s: then follow from step 2 here:
20:10:15lostlogiclinuxstb: I think the audio driver modifications are no longer needed
20:10:28jocke1scharkins. Ok got it
20:10:46lostlogiclinuxstb: because I moved the interrupt protection and locking to a per-byte level inside the i2c driver
20:10:50 Part jocke1s
20:10:55amiconnHmm, but plugins/chessbox.c is properly disabled for ipod in SOURCES...
20:11:46charkinsjocke1s: the first instructions refer to a bootloader.bin file which you need to grab from the bottom of the second link i gave ya
20:12:26amiconnCC chessbox.c ... wtf???
20:12:31linuxstblostlogic: You might like to bookmark this page -
20:12:56***Saving seen data "./dancer.seen"
20:13:04Zoide777_Btw, the Sourceforge sponsor image is still present in many of website's subsections
20:13:13lostlogicahh, yes, I was looking for that yesterday.
20:13:32 Quit Matze41 (Read error: 104 (Connection reset by peer))
20:14:15amiconnBagder: Is it possible that MEM isn't defined when processing SOURCES ?
20:14:31vcawould someone familiar with the ipod nano's boot process be able to explain the boot process to me a bit? or direct me to some docs explaining it? in particular i'm interested where the ipod's hardware hands over control to the bootloader.
20:15:29amiconnD'oh! it's called MEMORYSIZE for SOURCES...
20:16:30linuxstblostlogic: So what should I change if I wanted to test your battery code on the 4G? Anything more than just enabling it in the config file?
20:16:58lostlogiclinuxstb: I think you just need to define the battery as LIPOL1300
20:17:29linuxstbOK, I'll see what happens...
20:18:01youngcerea1i work on it
20:18:21youngcerea1lostlogic i send you my patch
20:18:22linuxstbyoungcerea1: Have you looked at lostlogic's code?
20:18:23youngcerea1ther is
20:18:37youngcerea1yes a moment
20:18:49youngcerea1there is nothing in powermanagment
20:18:58youngcerea1for ipod
20:19:15youngcerea1there where need a new battery
20:20:46youngcerea1Index: firmware/powermgmt.c
20:20:47youngcerea1RCS file: /cvsroot/rockbox/firmware/powermgmt.c,v
20:20:47DBUGEnqueued KICK youngcerea1
20:20:47youngcerea1retrieving revision 1.103
20:20:47youngcerea1diff -u -r1.103 powermgmt.c
20:20:47***Alert Mode level 1
20:20:47youngcerea1−−- firmware/powermgmt.c4 Mar 2006 09:48:53 -00001.103
20:20:49youngcerea1+++ firmware/powermgmt.c5 Mar 2006 14:53:53 -0000
20:20:51youngcerea1@@ -123,6 +123,8 @@
20:20:53youngcerea1 310, 345 /* alkaline, NiHM */
20:20:55youngcerea1 #elif CONFIG_BATTERY == BATT_LIPOL1300 /* iRiver H1x0 */
20:20:57youngcerea1 339
20:20:59youngcerea1+#elif CONFIG_BATTERY == BATT_IPODPHOTO
20:21:03youngcerea1 #else /* Player/recorder, NiMH */
20:21:05youngcerea1 475
20:21:06 Join jocke1s [0] (
20:21:07youngcerea1 #endif
20:21:09youngcerea1@@ -141,6 +143,8 @@
20:21:11youngcerea1 #elif CONFIG_BATTERY == BATT_LIPOL1300
20:21:13youngcerea1 /* Below 337 the backlight starts flickering during HD access */
20:21:15youngcerea1 { 337, 358, 365, 369, 372, 377, 383, 389, 397, 406, 413 }
20:21:17youngcerea1+#elif CONFIG_BATTERY == BATT_IPODPHOTO
20:21:19youngcerea1+{ 120, 150, 170, 190, 220, 250, 270, 290, 300, 320, 345 }
20:21:21youngcerea1 #else /* NiMH */
20:21:23youngcerea1 /* original values were taken directly after charging, but it should show
20:21:25youngcerea1 100% after turning off the device for some hours, too */
20:21:27youngcerea1oh shit
20:21:29youngcerea1i send a mail
20:22:05lostlogicyeah, don't paste patches to the channel man.
20:22:22lostlogicthose battery readings look ... wrong.
20:24:00 Quit charkins ("Leaving")
20:24:14youngcerea1it works
20:24:26amiconnt0mas, Zoide777_: Fixed. Next bleeding edge should contain chessbox.
20:24:38Zoide777_amiconn: Thanks
20:24:41youngcerea1i have send out
20:24:44t0masamiconn: I can force a new build
20:24:54amiconnIt should rebuild anyway
20:24:57t0masDBG: Creating temp. buildpage.
20:24:59t0masDBG: Building now...
20:25:05t0masit's already building
20:25:23markunyoungcerea1: the numbers should represent voltage*100
20:25:47amiconnWho would expect the very same #define to be called different in SOURCES than in .c/.h files ...
20:25:58youngcerea1not on ipod
20:26:08markunyoungcerea1: why not?
20:26:11youngcerea1if i do this never have power
20:26:35youngcerea1ipod battery have 700 mah of 3.7v
20:26:38amiconnThe number should always represent centivolts
20:27:17youngcerea1yes i know but the reading option where use at the moment is not korrekt
20:27:24jocke1scharkins: bootloader.bin is missing and when looking at compiling I need arm-elf-gcc How do I get that?
20:27:30amiconnMaybe the iPod does some kind of range spreading, then this has to be accounted for by the voltage reading function
20:27:44lostlogicyoungcerea1: then fix the reading, don't make up false values
20:27:59Mikachujocke1s: there are instructions for compiling a cross compiler somewhere
20:28:03youngcerea1i work on it
20:28:13lostlogiclinuxstb: do you get battery voltages similar to youngcerea1's on your ipod 4g?
20:28:39youngcerea1lostlogic: this where for you too see wath you need
20:29:27linuxstblostlogic: It's currently reading 4.13
20:29:37blueyjocke1s: do you use gentoo? if yes try crossdev
20:29:39linuxstb(this is a 60GB 4G Color)
20:30:21jocke1sbluey: ubuntu and yoper
20:30:27youngcerea1i go to lunch back in 15 min
20:30:48***Alert Mode OFF
20:30:57Zoide777_wow that's a fast lunch
20:32:48lostlogiclinuxstb: sounds right to me
20:33:01 Join hydrahead [0] (
20:33:27Zoide777_Does Rockbox support iPod recording? iPodLinux lets you do it through the headphone jack, and I noticed that Rockbox has an adc driver for iPods...
20:35:03blueyis a C++ compiler needed for rockbox or only C ?
20:35:41Mikachuonly c
20:36:40 Join linuxstb_ [0] (
20:36:55youngcerea1Zoide777_: if i programming it is so
20:37:57Zoide777_youngcereal: :D
20:38:24Zoide777_Has anyone heard of any tagcache updates from Slasheri?
20:39:08 Nick Zoide777_ is now known as Zoide777 (
20:39:42 Part jocke1s
20:44:02amiconnZoide777: Grab your chessbox :)
20:44:23Zoide777amiconn: I did, it's working :)
20:44:32Zoide777amiconn: thanks a lot
20:44:43Zoide777gotta go
20:44:49Adityathank god
20:44:53 Quit Zoide777 ("CGI:IRC")
20:46:16 Join LinusN [0] (
20:47:48 Join Bger [0] (n=Bager@
20:49:54amiconnLinusN: It looks like lcd_roll() is impossible to implement for H300.
20:50:41amiconn(1) The LCD has no roll, only scroll. (2) This is vertical scroll as the controller sees it, but the panel is mounted sideways
20:50:49amiconnDid I overlook something?
20:52:29 Quit preglow (Read error: 110 (Connection timed out))
20:54:38 Join Rondom [0] (
20:55:43 Quit bluey ("Leaving")
20:56:04 Part commanderx ("Leaving")
20:58:32 Part mozetti
21:00:45 Join bluey [0] (
21:01:36hydraheadhey can i set the rockbox so that the backlight doesnt turn on everty time i do something?
21:03:47 Join preglow [0] (
21:03:57preglowhow to run a good server, point 1: reboot every ten minutes
21:05:06Benacoolpreglow: humor ? lol
21:05:39preglowhumour would be if i actually meant that
21:06:34hydraheadis there any way to fix the apple firmware showing from right to left? i just installed rockbox and booted into original os
21:10:13 Quit SereR0kR (Remote closed the connection)
21:11:15 Join SereR0kR [0] (
21:11:20 Quit SereR0kR (Remote closed the connection)
21:11:43 Join damaki_ [0] (
21:12:06hydraheadanyone have an idea? please?
21:12:49Bgerlinuxstb ? preglow ?
21:12:51Paul_The_Nerdhydrahead: Assuming you have a grayscale iPod, that's just the way it works right now.
21:13:25Paul_The_Nerdhydrahead: Since it's still in development, Rockbox doesn't 100% peacefully coexist yet.
21:14:20Paul_The_Nerdhydrahead: If I understand correctly, there is a fix if you feel like doing a little coding. It's not the desired solution overall, but as a hack it works. I *think* it was said that if you remove the LCD initialization portion of the bootloader, the problem disappears.
21:14:34CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:14:34*Paul_The_Nerd could be remembering wrong
21:15:00hydraheadyes I'm on a greyscale ipod... :( I have 2 more questions: 1. Can I stop the backlight from turning on every time i do something? 2) Do I have to enter retail os diskmode in order to add songs?
21:15:29hydraheadi would gladly use the hack but outside of BASIC i dont know anything 'bput codin
21:16:09Paul_The_Nerdhydrahead: At the moment (assuming it wasn't fixed since last night and I missed it), the backlight currently can't be set to "Off" because it's the same as "On" (at least on color iPods. Check the setting in the menus and try it yourself, it may be working on grayscale for all I know)
21:16:25Paul_The_Nerdhydrahead: And there's no USB mode in rockbox itself yet, so yeah, it reboots into retail to do USB
21:18:03 Join webguest06 [0] (
21:18:10hydraheadyeah i just switched it to off and it wont shut down
21:18:19hydraheadthe best i can do is set it to 1s
21:18:31 Join bluebrother^ [0] (
21:18:50webguest06lostlogic, I get artifacts/pops when skipping tracks and when I select play from pause
21:19:25 Quit linuxstb_ ("CGI:IRC")
21:20:25hydraheadlinuxstb: paul mentioned some hacked solution for retailos showing upside down. do you know if there is a compiled version of it?
21:21:59linuxstbhydrahead: Maybe this one:
21:23:13linuxstbPaul_The_Nerd: Have you tried enabling the battery status on the Nano?
21:23:41Mikachulast time i turned on my nano i got a fancy graph in the debug screen and the battery meter is full instead of empty
21:24:04Paul_The_Nerdlinuxstb: No, I haven't. Is there a patch for it?
21:24:26 Nick bluey is now known as bluey^afk (
21:24:33linuxstbJust look at the last changes to config-ipodvideo.h and apply them to config-ipodnano.h
21:24:50linuxstbI've just done that for config-ipodcolor.h and it seems to be fine.
21:24:51hydraheadI just paste this into the ipod?
21:25:22hydraheadoh yeah BTW i see the battery shows as full on my grayscale, way to go! :)
21:26:00linuxstbhydrahead: That's a replacement bootloader - so you need to follow the bootloader installation instructions (but starting at step e, and using the original bootpartition.bin)
21:26:01Paul_The_Nerdlinuxstb: Should the battery scale be the same as 'video, or do you have a guess for what I should use?
21:26:22linuxstbPaul_The_Nerd: I've no idea.
21:26:27Paul_The_NerdI'll try that then
21:26:40Paul_The_NerdWhat does the scale represent?
21:26:41 Quit damaki (Connection timed out)
21:26:57linuxstbNo idea... :)
21:27:19*linuxstb is fresh out of ideas today.
21:27:23Paul_The_NerdIt's a vague term. =/
21:28:05 Join tinodeleste [0] (
21:29:39linuxstbpreglow: I tried arm-elf-gcc-4.1 on my 32-bit Debian system, and it didn't compile either. This is using a native gcc-4.0.3
21:29:44hydraheadLinuxstb: so i just do step e?
21:29:56linuxstbhydrahead: And step f)...
21:30:46 Join jocke1s [0] (
21:30:58 Join imphasing [0] (
21:31:06jocke1sIs it sh-elf or arm-elf for the ipod video ?
21:32:18preglowlinuxstb: someone in ipodlinux says it's working
21:32:23preglowahh, that'd be you
21:32:50jocke1sI get a arm-elf-ar command not found when trying to compile gcc :(
21:33:07Mikachudid you forget to set the path after you installed binutils?
21:33:30Paul_The_Nerdlinuxstb: Well, the battery scale is off, but it displays a value. It's sorta fluctuating between 88 and 90%
21:33:44jocke1sI might have missed that. I'll check
21:34:05linuxstbbluey^afk: How did you compile arm-elf-gcc 4.1?
21:34:25bluey^afkused gentoo's crossdev script and installed it with binutils-
21:34:28 Join crashd [0] (
21:35:04linuxstbWhich native version of gcc are you using?
21:35:14bluey^afkboth are 4.1 now
21:36:13webguest06lostlogic, you around ?
21:37:17hydraheadlinuxstb: i did it but now it doesent seem to want to boot into original os
21:37:31hydraheadim holding down menu but it always gives me rockbox
21:37:50linuxstbThat's one way to fix it.... :)
21:38:12jocke1slooked like I only got the rot dir of the binutils I am adding the bin dir too and Iäll try again
21:38:52linuxstbhydrahead: Can you try this file?
21:39:04hydraheadi will, one moment :)
21:39:47 Join SereR0kR [0] (
21:40:19 Quit SereR0kR (Remote closed the connection)
21:41:03 Quit bluey^afk ("Leaving")
21:41:30hydraheadlinuxstb: it just booted into retail os without that secon bootloader, dont know how i did it. the display is fine :))
21:42:10hydraheadi turned it of, pluged the firewire cable and held the menu button...
21:43:02 Quit Nico_P ()
21:43:17hydraheaddo you still want me to try test1 or should i leave it alone?
21:43:40 Part jocke1s
21:43:41stevebjust in case anyone is interested (and uses it) i started a channel for discussion of the H300 Experimental build #rockbox-h300
21:44:02linuxstbhydrahead: If it's working for you, then just leave it. Someone with a greyscale LCD needs to try and debug the problem and fix it properly - but I don't have such an ipod.
21:44:14stevebyes, there is no one there yet and it might be a waste of time but i thought it was a good idea :-/
21:44:54hydraheadok ill try to boot it without the cable, if that works ill leave it alone
21:47:30hydraheadok its working, thanks a lot!
21:47:56linuxstbDoes anyone know any Linux scripts/apps to download album art based on a CDDB ID?
21:49:19steveblinuxstb: there are very few album art downloaders. the best i have seen is
21:50:02 Join tucoz [0] (
21:51:21linuxstbsteveb: That looks promising. Any idea if it supports FLAC files?
21:51:39steveblinuxstb: hmmm no idea sorry
21:52:08linuxstbsteveb: n.p. I'll download it and have a look.
21:52:23stevebits pretty good. gives you a nice selection of sites to search from
21:52:31steveband then you can pick which image you want
21:52:37hydraheadi think it supports id3 tags only if you want to embed
21:55:13 Join herz42___ [0] (
21:56:09 Nick herz42___ is now known as herz42 (
22:02:52Paul_The_Nerdlinuxstb: Good news is that with the patch, battery_bench now seems to be recording voltages.
22:03:50 Join solexx [0] (
22:04:21linuxstbsteveb: That's a very nice program. It's going to need a little configuration (to get the album/track names from my directory structure), but it's working.
22:04:47stevebi cant get mine working.... i had to boot in to vmware an windows to use it!
22:04:54stevebsome problems with utf8 i think
22:05:07stevebit works for individual tracks, but not for recursive directories
22:05:48linuxstbIt seemed to work OK for me recursively.
22:08:16steveband everyone else it seems...
22:08:22 Quit imphasing (Read error: 104 (Connection reset by peer))
22:08:34 Quit Paul_The_Nerd ("Leaving.")
22:08:36 Quit Rondom ("I'm leaving on a jetplane, don't know when I'll be back again...")
22:08:38 Join SereR0kR [0] (
22:11:43 Part webguest06
22:12:06 Join imphasing [0] (
22:12:59***Saving seen data "./dancer.seen"
22:13:25linuxstbsteveb: What directory structure are you using?
22:13:43steveblinuxstb: i use /music/genre/album
22:14:46 Quit solexx_ (Read error: 104 (Connection reset by peer))
22:14:59 Quit needleboy ()
22:15:41linuxstbsteveb: Ah, that may be the problem. Do you have the artist name anywhere?
22:16:07 Join vektek [0] (
22:16:08steveblinuxstb: it works fine on windows
22:16:30vektekHi I'm looking at getting an archos 202s..... Any good?
22:16:33 Join XavierGr [0] (
22:17:17linuxstbvektek: It doesn't run Rockbox...
22:17:34vektekWill it in the future?
22:18:48preglownot likely
22:18:59preglowneed someone to port it, at least
22:19:10stevebyou never know. in the future we may have 10 feet and be able to fly
22:19:17*steveb has dreams of this
22:19:36stevebok that was a little random
22:19:42vektekTrue... Steve you're scaring me
22:19:59stevebits been a long day
22:20:44vektekWhat is the machine like even though it doesn't run on rockbox?
22:21:34Bagderthis is #rockbox, we mostly know and speak Rockbox...
22:22:04*safetydan 's head explodes from fixed point trig overload
22:22:10BagderI doubt many people here have or uses a gmini
22:23:23vektekok badger. i was asking for a friend who is not technical.... it was a long shot.... thanks anyway
22:24:04tucozI hope it's okay that I go havoc on the manual once in a while (without asking you guys before I commit). It's in a rather sorry state as it is, and needs lot's of fixes.
22:24:27Bagdertucoz: we consider you mr manual anyway, feel free
22:24:43MikachuBagder: do many people call you badger?
22:24:56Bagderall the time
22:25:06preglowsafetydan: got some eq plotting magic going on here now
22:25:14Mikachuwhere does "bagder" come from?
22:25:14 Quit vektek ("CGI:IRC (EOF)")
22:25:23preglowsafetydan: and i'm really sceptical as to how informative this will be on a tiny screen
22:25:26tucozOk. good. But if anyone else gets tempted (not that I would imagine that) do change something, please do :)
22:25:47BagderMikachu: I picked a nick ~20 years ago, and misspelled it but then I thought it was cool like this and kept it
22:25:54Mikachuah, okay
22:25:57preglowtucoz: go ahead
22:26:04preglowtucoz: i haven't had time to look at it yet, but i'm planning to
22:26:32tucozpreglow, ok good.
22:26:56safetydanpreglow, really? What about those lucky iPod Video owners, do they have enough res?
22:27:10safetydanOr am I going to have to force you to implement zoom? :)
22:27:45preglowsafetydan: they might, we'll see
22:27:48hydraheadis there some throttling patch that doesent crash 4g grey?
22:27:52preglowi've still got a ton of bugs to iron out
22:27:56preglowhydrahead: no
22:28:23linuxstbWell, no-one's tested it on a greyscale 4g afaik. But it doesn't work on my 4g Color.
22:28:30hydraheadaw bugger, its getting really hot :)
22:28:40 Join Daishi [0] (
22:28:51hydraheadwell since the processors are the same i dont figure there is a difference
22:28:53Mikachuare both oscillo* ported or only oscillograph?
22:29:05linuxstbhydrahead: No, I would be surprised if there was any difference.
22:29:18hydraheadunless the lcd causes it to chrash :)
22:29:21safetydanMikachu, only oscillograph, oscilloscope had a few too many SH1 dependencies for me to do a quick port
22:29:41Mikachui got an oscilloscope.rock too but it seems to be 0-length
22:30:04linuxstbpreglow: I get two types of crashes with the cpu speed patch. It either just freezes, or I get a prefetch abort at somewhere like 0x60005ffe
22:30:15safetydanMikachu, yeah that's normal. It's got a big #ifdef (CODEC != SWCODEC) around it so you get 0 byts on compile
22:30:25crashdevening guys
22:30:31crashdi ended up getting an ipod 5g
22:30:44Mikachui would have thought that would be considered a small bug in the SOURCES file
22:30:49preglowlinuxstb: still not much i can do without an actual 4g, i'm afraid
22:31:06safetydanMikachu, 0 byte .rock files aren't included in the actual .zip file in the end
22:31:18 Join Paul_The_Nerd [0] (
22:31:23Mikachui've never run that make zip thing
22:32:07Mikachuwhich you could say is my fault then :)
22:33:17Papricaaha! rockcalendar is ready for archos recorder!
22:33:23 Join Farpenoodle [0] (
22:40:45*safetydan gives up on cordic for the night
22:41:21hydraheadpreglow: so unless you get your hands on a 4g, no chance of variable cpu speed ?
22:44:35preglowi'm not the only one capable of fixing it
22:45:15preglowlinuxstb: but yeah, a prefetch error most certainly suggests something goes wrong with the memory controller
22:45:47preglowlinuxstb: you could try making the delay after the reclock largen than 2ms
22:46:01linuxstbYes, I tried adjusting that delay, but it didn't seem to help.
22:46:18preglowno, can't say i'd expect it to
22:51:08 Quit SereR0kR ("XChat Aqua")
22:51:26 Join SereR0kR [0] (
22:52:51 Quit Bger ("BitchX: often imitated, never duplicated!")
22:53:59 Quit zhilik (Read error: 104 (Connection reset by peer))
22:55:30linuxstbpreglow: Any idea what the write to 0x70000020 does in set_cpu_frequency() ?
22:57:31preglowgimme a sec
22:58:19preglowi'm guessing it enables the pll
22:58:22preglowpowers it up
22:58:30preglowi think 0x7000xxxx is device control
22:58:42 Part tucoz ("Leaving")
22:58:45linuxstbSounds reasonable.
22:59:46preglowthe 60006020 writes just select the clock source, i'd imagine
23:00:05preglowthe first selects xtal, then you set the scalersr, wait for relock, the select the pll
23:02:01preglowyou think yielding within the set_cpu_freq is possible? :>
23:03:22 Join jocke1s [0] (
23:04:34preglowi guess i should just accept there's no point in fretting over 2ms and stop thinking about it
23:04:40jocke1sI have installed the bootloader.bin but how do I install .rockbox folder. I am doing this from linux
23:04:55preglowmake zip
23:04:59preglowthen unzip that zip to the root of the ipod
23:05:21preglowor if you've just downloaded a build, just unzip it to the root of the ipod
23:05:35jocke1sok i'll try it
23:06:10linuxstbpreglow: Have you tried decreasing that delay?
23:07:38preglowi figured 2ms was already pretty small for a relock
23:09:09 Join antoine [0] (
23:10:40 Quit SereR0kR ("XChat Aqua")
23:13:36 Join quobl [0] (n=quobl@tor/session/x-2ecff1375e601095)
23:15:54 Quit safetydan ("Leaving")
23:16:05 Part ep0ch ("Kopete 0.11.1 :")
23:17:14 Join RaTG [0] (i=RaTG@
23:18:36jocke1sso where do I put my music? In the root or under for example a mp3 folder?
23:18:54Bagderjocke1s: whereever you want ;-)
23:19:11nudelargh :( my FLAC albums just finished converting to MP3, after about 2 days of calculation, and now i realise/remember that fresh installs of Foobar are set to write archaic ID3v1 tags that can't fit most song titles andit's all screwed.. . AFASKJHDLSAKJDH!"!!! KILL
23:19:16*nudel stabs foobar
23:19:44*preglow patches foobar up again
23:20:26*Aditya gives foobar a cookie
23:20:34nudelsecond time i've done this as well :-/
23:20:48*Aditya then slashes foobar's hand off as he reaches for the cookie
23:21:06*preglow gives foobar an appreciation hat
23:21:20Paul_The_Nerdnudel: I thought foobar did both ID3v1 and ApeV2 by default. Do the ape tags get trimmed as well?
23:21:35nudeloh don't know, my file manager doesn't support ApeV2
23:21:47 Join BHSPitLappy [0] (
23:21:53nudeli'll load them into TGF and see if i can move the tags
23:22:37nudeli don't agree with ApeV2 when ID3v2 has always worked fine and is less supported... i'm not aware of anything (worth using, made post 1968) that supports id3v1 but not id3v2 either... seems a silly default
23:24:10 Part jocke1s
23:24:39 Join ashridah [0] (
23:25:18 Quit tinodeleste (Read error: 110 (Connection timed out))
23:33:12BHSPitLappyNEW 5G! YAY!
23:34:20 Join jocke1s [0] (
23:34:53crashdBHSPitLappy: ;)
23:34:59crashdis there a tutorial for using loader2 with rockbox?
23:35:05crashdnot because i want one, but ive just been doing it
23:35:16crashdand i thought i may as well share the experience ;
23:35:35Paul_The_NerdI imagine that since loader2 is their project, it'd be better to ask around the ipodlinux channel?
23:35:49jocke1swhat theme would be nice for a beginner :)
23:35:50crashdPaul_The_Nerd: well, i was going to, but loader2 is handy for us rockbox users
23:35:59crashdand as im not actually using ipl atm
23:36:20Paul_The_NerdYeah, but I'm just saying that if there were a tutorial, you'd expect to find it where you get the software from.
23:37:07crashdyeah, loader2 is still in dev tho, so the tutorial is thin on the ground (and not really for beginners)
23:37:12crashdfairplay for the answers : )
23:37:16Paul_The_NerdI mean, it's just a multi-purpose bootloader right? Or does it offer specific advantages to rockbox over those that it offers in general?
23:37:30crashdwell, it is just a multi purpose bootloader, sure
23:37:40lostlogicanyone know how to make itunes not open when I connect ipod to computer?
23:37:46crashdi just hear a lot of people moaning on the forums about not being able to default to apple os, or whatever
23:37:47linuxstbcrashd: I don't see any problems with putting it on the Rockbox wiki - those kind of documents don't really have a home.
23:37:55Paul_The_Nerdlostlogic: Right click on your ipod in itunes, choose properties, and there's a check box somewhere in there
23:38:04crashdlostlogic: tell it not too in the config, or stop the 'ituneshelper' running on startup
23:38:11crashdlinuxstb: cheers, ill get to it inabit
23:38:11linuxstblostlogic: Add/Remove Programs, uninstall iTunes.
23:38:21crashditunes is a real bane
23:38:34Paul_The_Nerdlostlogic: My vote goes for linuxstb's suggestion really, unless you *want* iTunes for some reason.
23:38:49lostlogichahaha thanks for all the help guys
23:39:09 Join TCK- [0] (
23:40:01 Join BHSPitMonkey [0] (
23:40:48Paul_The_NerdHehehe. I forgot that the iFP-790 port still uses the WMA trick to load rockbox. After building, I was staring at the build dir looking and rockbox.wma and thinking "Well, that's new..."
23:42:37cooblostlogic: there's an option in the itunes ipod prefs not to open itunes when the ipod is connected
23:42:58coobyou need to set this pref whilst the ipod is connected, as the ipod prefs are kept on the ipod
23:43:12nudeli've got that off and itunes doesn't open when i plug the ipod in, but one weird thing is it takes *ages* for the drive letter to appear and be ready... is that normal for ipods?
23:44:27 Quit antoine ("Client exiting")
23:48:42 Part jocke1s
23:48:48 Join quobl_ [0] (n=quobl@tor/session/x-76a1302b839e6152)
23:49:10 Quit quobl (Remote closed the connection)
23:49:19Mikachuhow can it read the prefs from the ipod to know it shouldn't start, without actually starting first?
23:49:34linuxstbIsn't there an ipodservice.exe running?
23:49:51Mikachuright, windows - the land of random processes
23:51:27nudeli don't think it's windows' fault that apple write crap software
23:51:54Paul_The_NerdNah. Windows is at fault merely for being Windows.
23:52:08 Join JdGordon [0] (
23:52:09nudelyeah :) no need to blame it for extra stuff
23:52:11Adityaand you my friend are a zealot
23:53:19nudeli like windows... last time i tried out linux my monitor exploded shortly afterwards (okay, i was playing quakeworld back in windows at the time, but i still blame linux!)
23:53:22Paul_The_NerdTo be a zealot I'd have to be a proponent of something else, technically
23:53:26markunAditya: hey, you here to spoil our fun again? ;)
23:53:39Paul_The_NerdSince I'm not frothing-at-the-mouth pro-Linux, I don't really qualify as a Zealot.
23:53:57Adityayou seem to be doing the frothing for apple just fine
23:54:08markunPaul_The_Nerd: you're a rockbox militant, isn't that bad enough?
23:54:32Paul_The_Nerdmarkun: You have a point.
23:55:27Paul_The_NerdAditya: I'm not pro Apple. In fact, as a whole I'm generally not a fan of them at all. But I tend to see a lot of baseless negative statements, and while I have no problem with people pointing out flaws in things, I do expect those flaws to at least be supported clearly and logically with fairly solid facts.
23:55:50markunWhat should the viewer do with tabs? Just insert a few spaces?
23:56:01 Part LinusN
23:56:15Paul_The_Nerdmarkun: Text viewer?
23:56:18AdityaPaul_The_Nerd: I am kind of just kidding
23:56:28nudelthe question is how many spaces, and whether to make it configurable for different text file extensions :)
23:56:29AdityaI like to just poke people
23:56:29Paul_The_Nerdmarkun: Yeah, a few spaces.
23:56:48markunPaul_The_Nerd: I could also do real tabs, but I don't know if that's overkill.
23:57:15Adityawell its just a viewer
23:57:16Paul_The_NerdAditya: Ah, sorry. I'm just kinda used to some extremely anti-this and pro-that people who don't really back it up. I have a tendency to pick at arguments, even when I agree with them.
23:57:20Adityajust do what s easy
23:57:48markunAditya: yes, I know. But I'm fixing the viewer anyway right now..
23:57:58AdityaI know.. its still a viewer
23:58:08Paul_The_Nerdmarkun: Hrm. Real tabs may be better, just for consistency then.
23:58:11Adityaso if spaces is easier then tabs would just be a waste of brain cells =P
23:58:35*Paul_The_Nerd is all in favor of "If you're gonna do it, may as well do it _right_"

Previous day | Next day