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 2006-03-30

00:00:14lostlogicand then 481 should go bye bye
00:00:37preglowlostlogic: yup, just trying to cut down on gcc 4.1 warnings in case we go 4.1 some day
00:00:42 Join sharpe [0] (
00:01:22lostlogicpreglow: actually on coldfire it's faster (on older GCC) to do it with *buf = asdflkjas; buf++ so do it that way if you're changing, I probably made that change for that reason but accidentally left the * in there for no good reason
00:01:29sharpethat was quite strange...
00:02:08sharpeyay for power being out for two hours when you get home...
00:02:10goffa__curious... what does the endian defines do for us?
00:02:11BarretJlinuxstb: doesnt work
00:02:25preglowlostlogic: why would that be faster? both should be optimised down to a post-increment
00:02:26Papricapreglow, thanks fo the help, i'll try again tomorrow, when i'll less nervous
00:02:31BarretJlinuxstb: and i dont think the ipod is frozen, since the screen dims after some inactivity
00:02:34amiconnlostlogic: Really? I'd expect *buf++ = something to be faster. Would use the post-increment indirect addressing.
00:02:34preglowPaprica: doens't work out?
00:02:52preglowamiconn: both cases should use that, unless gcc is truly braindead beyond all hope
00:02:59BarretJfsck it, i'll just hard reboot
00:03:02Papricai'll try again tomorrow
00:03:05preglowPaprica: do so
00:03:21amiconnpreglow: gcc someties *is* braindead. Remember my move-around-registers sequence?
00:03:29preglowyeah, i know gcc is all too well
00:03:39preglowi've seen some pretty foolish things
00:03:43preglowit's _far_ too free with the stack
00:03:57preglowbut anyway
00:04:04preglowi don't think optimising like that is a good idea
00:04:29preglowwhat is a good idea on one platform is a bad one on another one
00:04:30amiconnI'd think *buf++ = something is a common thing
00:04:37preglowbetter to just write c code like you ordinarily would
00:05:00*amiconn usually prefers pointers over arrays anyway
00:05:10preglowdepends on what i do
00:05:20amiconnYes of course
00:05:32preglowi usually prefer pointers because i think they're easier to use
00:05:33lostlogicamiconn: gcc is brainded
00:05:51 Quit iJason ("CGI:IRC (Ping timeout)")
00:06:28lostlogicfor *buf++ = asdlfkjasdf; it does r0 = r1; r1++; *r0 = asdlfjasdflj;
00:06:36lostlogicI shit you not.
00:06:48lostlogicermh make those 'a' not 'r' in this case
00:07:24preglowit's a miracle the codecs run at all
00:07:26preglowit truly is
00:07:39lostlogicpreglow: my local vorbis is optimized around that behavior :-P
00:07:59 Join JdGordon [0] (
00:08:30preglowthe gcc people told me they'd work on coldfire
00:08:37preglowbut so far i haven't seen anything
00:08:43lostlogicso, what does it produce on 4.1 for same code? :)
00:08:53preglowi don't know, perhaps i should give it a spin
00:09:00preglowwhen it even agrees to compile, that is
00:09:09preglowwhich is another thing the gcc people are telling me they'll fix
00:09:33goffa__wow... looks like the maytag man shouldn't have been slacking on the job
00:09:40goffa__whirlpool bought maytag
00:10:39amiconnlostlogic: Remember this one: at 19:59? A jewel from devcon...
00:12:07preglowthey even tell me 4.1 has seen some coldfire work
00:12:26 Join Daishi [0] (
00:12:45*amiconn would like to be able to even build m68k-elf-gcc on amd64 :(
00:12:53 Join Genre9mp3 [0] (
00:14:31preglowgod, me too
00:14:38preglowi'd really enjoy that
00:14:43preglowbargh, they've only put in fpu support
00:14:47preglowthat's all they've done performance-wise
00:15:05preglowlostlogic: you wouldn't happen to have a list of stupid shit gcc has done? :>
00:17:11 Quit webguest15 ("CGI:IRC (EOF)")
00:17:45lostlogicpreglow: not using register indirect post inc addressing and not using addq are my two favorites
00:18:20lostlogicoh, and this applies to other arches too, I think, moving data from register to register for fun, and ignoring practical scope limitations causing major register waste
00:19:20preglowi've seen it moving values to and from the stack for fun
00:19:27preglowseeing that has almost always left me in a high rage
00:19:44lostlogicamiconn: ah yes, <3 gcc
00:20:38preglowthis _almost_ makes me want to dig into gccs backend internals
00:20:43preglowhowever, i am not completely insane yet
00:20:46 Quit Rondom (Read error: 113 (No route to host))
00:21:01lostlogicI think it may be a hobby of mine to pick on other people's software... I spent the past 2 days of my life hunting down a bug at work... that turned out to be in the apache weblogic module... involving a failure to use the _clearly defined_ cookie parsing logic and thereby breaking certain cases of cookies vs. session cookies
00:21:04preglowi don't expect to reach that level for a couple of weeks time yet
00:21:08 Quit petur ("here today, gone tomorrow")
00:21:38goffa__so where's every one from?
00:21:44 Quit [IDC]Dragon (" I love my HydraIRC -> <-")
00:21:47preglowvaries wildly
00:21:48preglowi'm from norway
00:21:55goffa__montana here
00:21:56lostlogicChicago, USA
00:22:05preglowmostly europeans here, though
00:22:10preglowwith the odd american now lately
00:22:11goffa__dragon must be overseas too
00:22:27lostlogicgoffa__: the project was founded in Sweden
00:22:32lostlogicso we are overseas, not them
00:26:23 Join goffa_work [0] (
00:26:33goffa_workhmm.. connection at home seems to have died
00:26:53 Join RotAtoR [0] (n=e@rockbox/developer/rotator)
00:27:05 Quit BHSPitLappy (Connection timed out)
00:27:53*Febs tries unsuccessfully to think of something funny to say about "odd americans."
00:27:54goffa_workbut i guess you could say i'm overseas :)
00:28:16goffa_worki know a lot of odd americans
00:28:17lostlogicgoffa__: 16:22 < lostlogic> goffa__: the project was founded in Sweden
00:28:17lostlogic16:22 < lostlogic> so we are overseas, not them
00:28:31goffa_worksaw that.. then i died
00:28:35lostlogicah, gotcha
00:28:53preglowsounds traumatic
00:29:03goffa_workhope its not :)
00:31:07*preglow looks at m68k.c
00:31:14preglowthe first thing i notice is mixed tabs and spaces
00:32:21 Quit Genre9mp3 ()
00:32:34 Join ender1 [0] (i=ychat@
00:33:50 Join BHSPitLappy2 [0] (
00:34:45 Join [1]ender [0] (i=ychat@
00:38:29 Join XavierGr [0] (
00:39:55 Join ashridah [0] (
00:40:05 Quit obo ("bye")
00:43:18amiconnHmm, the 'no rockbox dir' message makes no sense in .lang
00:43:33amiconnIf there is no .rockbox directory, the message can't be localised...
00:45:22*amiconn cuts that
00:46:51 Quit bluebrother^ ("Leaving")
00:46:54 Quit ender` (Read error: 110 (Connection timed out))
00:47:12YouCeyEhow to save the config files before upgrade?
00:47:22YouCeyEhow to restore the old config?
00:48:32goffa_workis there only 1 central config file to back up?
00:48:54YouCeyEjust the general settings
00:49:06YouCeyElike song resume on
00:49:17FebsMain menu −−> Manage Settings −−> Write .cfg file
00:50:06FebsTo restore: Main menu −−> Manage settings −−> Browse .cfg files. Or, navigate to the config file and "play" it.
00:50:50 Join webguest05 [0] (
00:51:23webguest05YouCeyE, manual here
00:51:27preglowtime for bed
00:51:45 Part webguest05
00:52:09 Quit ender1 (Read error: 110 (Connection timed out))
00:56:10 Quit gursikh ()
00:57:52 Join BHSPitLappy [0] (
00:58:49 Join warthawg [0] (
00:59:18warthawgon a nano, how do i boot into apple os after installing rockbox?
00:59:48 Quit [1]ender (Read error: 104 (Connection reset by peer))
01:01:11vcawarthawg: i'm pretty sure you hold MENU during boot, but i'm going from memory here
01:01:35vcait should be answered in the nano rockbox FAQ
01:01:42warthawgok, thanks
01:02:59 Quit BarretJ ("leaving")
01:05:10 Quit BHSPitLappy2 (Success)
01:06:44MoosGood night all !
01:06:48 Quit Moos ("Glory to Rockbox !!!")
01:11:09 Join midkay [0] (n=midkay@rockbox/developer/midkay)
01:11:19 Join Sinbios [0] (
01:12:43***Saving seen data "./dancer.seen"
01:13:59 Quit muesli__ (Read error: 110 (Connection timed out))
01:19:37amiconnOkay. Now I need the perl magic...
01:24:14 Join BHSPitLappy2 [0] (
01:24:15dpassen1When did the quickmenu change?
01:24:38amiconnYOu mean when it was uglified?
01:24:43amiconnAsk midkay...
01:24:50*amiconn goes to bed...
01:24:53midkayamiconn, fu. ;)
01:24:59dpassen1Uglified as in arranged vertically with absolutely no obvious button for each option?
01:25:09midkaydpassen1, that was a few weeks ago.
01:25:20dpassen1Just updated my build for the first time in quite a while
01:25:26midkaydpassen1, can you not see the >> << and downwards icons?
01:25:46XavierGrmidkay why did you changed it anyway?
01:26:24midkayXavierGr, to fix a bug with overlapping text in some languages. and it's far from finished. i have an 80% done much-better-looking version sitting here, just haven't got around to finishing it quite yet..
01:26:38XavierGrah ok nice to know
01:27:16midkaydpassen1, just for my information - do you or do you not see the icons in the quickscreen?
01:27:19dpassen1midkay: Nope, don't see
01:27:27dpassen1This isn't the latest build
01:27:32dpassen1Plus I have two patches
01:27:38midkaydpassen1, ah, kind of weird. seems like a common-ish bug.. but i have a way around it anyways.
01:27:55dpassen1I remember the buttons, so its not a large problem for me
01:28:29midkaydpassen1, probably is for a number of people.. i happen to see the icons which is rather weird though...
01:28:38amiconnmidkay: I hope to get around throwing together that perl script. Then I'll commit my .lang rework. I'm done with english.lang now :)
01:28:46midkayanyways, i'll put an update in CVS maybe today or very soon..
01:28:54dpassen1midkay: Ill quickly try a bleeding edge build?
01:29:06midkaydpassen1, no need - it hasn't changed since i committed it a few weeks ago..
01:29:10dpassen1Oh, ok
01:29:18dpassen1This is on a 120 for the record.
01:29:19midkayamiconn, great. what script?
01:29:35midkaydpassen1, i'm only able to test it on an archos recorder, so.. possibly that's it..
01:29:37amiconnmidkay: The quickscreen look can be vastly improved which viewports (post-3.0 stuff)
01:30:00midkayamiconn, yeah, but i'd like to "fix" them as much as possible before that..
01:30:12dpassen1Be interesting if anyone else can verify other models
01:30:15amiconnmidkay: The script that adapts all non-english .lang files to the reworked english.lang
01:30:29amiconnOr rather, has to adapt
01:30:44midkayamiconn, oh, cool.. sounds a bit tricky, but i'm sure you can handle it :)
01:30:51amiconnIt's not yet written, and I would prefer a perl expert around which I can ask
01:31:10amiconnBut I have to sleep now...
01:31:45XavierGrgoodnight amiconn
01:31:49midkaynight amiconn :)
01:31:54midkayamiconn, sounds cool though :)
01:31:59midkaygood luck
01:32:54 Quit BHSPitLappy (Connection timed out)
01:36:23 Join luigi_ [0] (
01:42:07 Join TCK [0] (
01:43:23 Join maeck_ [0] (
01:43:46 Quit maeck_ (Client Quit)
01:50:27 Join maeck_ [0] (
01:52:11 Join Shadowarrior13 [0] (
01:52:28 Nick Doomed_zZz is now known as Doomed9 (
01:52:59 Quit maeck_ (Client Quit)
01:53:06 Quit luigi (Read error: 110 (Connection timed out))
01:54:20 Quit goffa_work ("leaving")
02:00:00 Quit maeck (Read error: 113 (No route to host))
02:02:22scottderweird, sometimes when I plug in to my car charger/fm transmitter in my ipod comes on in rockbox and says "cleared" and I lose all my settings :(
02:03:53 Join damaki [0] (
02:08:01lostlogicscottder: save a config file in manage settings so you can restore, I've seen the same, don't know the cause yet
02:10:12scottderugh hard to manuver around the filename entry
02:12:29scottdergot it, thanks lostlogic
02:12:35linuxstbscottder: Do you have the hold switch on?
02:12:46midkayscottder, it's the hold switch..
02:14:12scottderNo....just figure out what is what :)
02:14:55*scottder hopes they get doom workin on the Nano :)
02:15:25 Quit luigi_ (Connection timed out)
02:16:15*linuxstb hopes they do as well, whoever they are.
02:17:34scottderthe neboulous "they"
02:18:46 Quit damaki__ (Read error: 110 (Connection timed out))
02:19:58 Quit midkay ("Leaving")
02:27:38 Quit ashridah ("Leaving")
02:28:59*lostlogic gets distracted from track skipping and messes with crossfade code instead
02:38:52 Join fiftyfour123 [0] (
02:40:34scottderIs there a way to adjust the amount of crossfeed?
02:40:45scottder(not crossFADE) :)
02:42:16 Nick BHSPitLappy2 is now known as BHSPitLappy (
02:43:36lostlogicI think preglow was working on that... but I don't think so at this time
02:44:17scottderThat'd be cool...the default setting cuts into the volume a bit much
02:44:22scottderat least for my taste :)
02:44:37scottdercourse I can just use the crossfeed in my headphone amp
02:44:42 Quit fiftyfour123 ("Chatzilla 0.9.71 [Firefox]")
02:46:01 Join quobl [0] (n=quobl@tor/session/x-72d96a4b73e43995)
02:46:15 Quit Mark_ (Read error: 104 (Connection reset by peer))
02:54:57 Quit Kohlrabi (Nick collision from services.)
02:55:06 Join Kohlrabi [0] (
02:55:31sharpehow about... they're 'odd'?
02:55:41sharpei'm really bored.
02:55:47sharpeback to work.
02:56:10DBordellocrossfeed, intresting
02:59:40scottderesepcially good for early stereo recordings with really hard left/right seperation
03:00:06DBordelloah, i was just reading about it
03:00:50DBordelloAnybody around that i can talk to about becoming a build server?
03:02:09sharpeteehee. you said build. and server.
03:03:16DBordelloyou lost me
03:04:14 Join jd|uni [0] (
03:04:23jd|unihey guys
03:04:54sharpei'm out of it today.
03:06:28 Quit Febs ("CGI:IRC (EOF)")
03:06:49sharpei must still say the same things that get no response.
03:10:40 Quit Xerion (Read error: 104 (Connection reset by peer))
03:12:47***Saving seen data "./dancer.seen"
03:22:06 Join DBLaptop [0] (
03:22:44 Quit DBordello (Nick collision from services.)
03:22:57 Nick DBLaptop is now known as DBordello (
03:23:17jd|uniperl q.. why does this keep printfing "Page 23" print "\nPage ",($line_count%23)+1 ($line_count does get incrememneted correclty)
03:23:35DBordelloDoes anybody have foobar2000 0.83?
03:23:41jd|unioops.. im a dill
03:24:33dpassen1DBordello: I do
03:25:31DBordellodpassen1, wanna help me out :)
03:26:16scottder0.9 is quite nice
03:26:36dpassen10.9 will be nice when there's an ASIO output
03:28:34DBordelloi run linux, i just want to give foo_pod a shot
03:28:46DBordellosee if it can overcome this wierd 46 character limit
03:28:53dpassen1Audio Stream Input Output
03:31:12dpassen1I knew I had that link somewhere
03:31:32 Join Paul_The_Nerd [0] (
03:32:14scottderI never heard enough difference with ASIO to make a difference
03:32:28dpassen1It gives me that warm fuzzy placebo feeling ;-)
03:32:30jd|unihow do u do integer division in perl? i dont want the decimal places..?
03:32:44scottderTried kernel streaming?
03:32:54 Part Paul_The_Nerd
03:32:55scottderLike ASIO with half the calories
03:33:44dpassen1Isn't Kernel Streaming a hack providing ASIO-like bypassing of the kernel mixer?
03:34:34dpassen1I don't see any advantage to using Kernel Streaming
03:34:47dpassen1Either way, I'm out.
03:37:41DBordellowhy would ASIO be useful to anybody but people recording?
03:44:16scottdermainly for when one wants to bypass the windows mixer...some feel it degrades sound
03:44:22scottderme I have never really heard a difference
03:45:50 Quit jd|uni ("CGI:IRC (EOF)")
03:52:21 Join unhcr [0] (
03:52:21 Quit warthawg (Read error: 104 (Connection reset by peer))
03:52:29unhcrhey deos anyone here work on the x5?
03:53:03 Quit unhcr (Client Quit)
03:54:10 Quit XavierGr (Read error: 104 (Connection reset by peer))
03:54:13 Join Vertigo_t [0] (
04:00:39 Quit Shadowarrior13 ("( :: NoNameScript 4.01 :: )")
04:01:00 Quit Vertigo_tdl (Read error: 110 (Connection timed out))
04:08:35 Join Arrogant [0] (
04:24:09 Quit Vertigo_t (Read error: 110 (Connection timed out))
04:36:25 Join midkay [0] (n=midkay@rockbox/developer/midkay)
04:40:35 Join maeck [0] (
04:46:09biffhero_jd|uni: example, please. Are you trying to round, to truncate, or to be fast?
04:46:44sharpethat was like so thirty minutes ago.
04:46:56biffhero_oh, sorry
04:47:04biffhero_is he gone?
04:47:25biffhero_oh, yeah, I missed that line. the bold escaped my brain
04:48:10biffhero_well, just the same, use "man perlfunc" /integer you lazybones!
04:48:17 Quit midkay ("Leaving")
04:49:35Doomed9anything new with tag cache?
04:50:40biffhero_you can't do tag cache in the ether. it has to be done either on disk or in ram. ether is on track for next week
04:51:16 Quit Arrogant ("Leaving")
04:51:21sharpeoh... my rubix cube will fall apart now...
04:52:49 Quit hardeep ("BitchX: now with wings")
04:52:52 Quit Kohlrabi ("Fast alle Menschen sind Regenwürmer")
04:55:58 Join Aditya [0] (
04:57:20 Quit Aditya (Client Quit)
05:02:15 Join Aditya [0] (
05:07:08 Part Aditya
05:08:57 Join ipod5G60Gb [0] (
05:09:00 Quit TCK (Read error: 110 (Connection timed out))
05:12:12ipod5G60Gbis there a program that converts your music so that both appleos and rockbox reconzige it? called foobar2000 or something?
05:12:41 Quit Daishi ("Client exiting...")
05:12:51***Saving seen data "./dancer.seen"
05:13:04ipod5G60Gbis that it? i did a search on google and it pulled up a windows mp3 player
05:13:27DBordelloit has some plugin that can generate a ipod database for the apple os based on what you have on your ipod
05:13:30DBordellopeople ahve had mixed results
05:13:36DBordellothe plugin is foo_pod
05:15:13ipod5G60Gbok, so d/l the foobar and look into the plugins, is it a seperate plugin or included the package?
05:15:45 Join Vertigo_t [0] (
05:16:35 Join midkay [0] (n=midkay@rockbox/developer/midkay)
05:16:43 Quit JoeBorn (Nick collision from services.)
05:17:27 Join jborn_ [0] (
05:21:43BHSPitLappyI've used foo_pod before
05:22:02BHSPitLappybut how does it make the music compatible with both retailos and rockbox?
05:22:17ipod5G60Gbdont know
05:22:33BHSPitLappyI'm not sure I understand what it does...
05:22:34midkayBHSPitLappy, a word of warning - didn't work for me.
05:22:45DBordellorockbox just browses the directory structure
05:22:54BHSPitLappyanyway, yeah, you install foobar and then extract the separate foo_pod files to the plugins folder
05:22:56DBordelloretail needs a database relating to the directory structure
05:22:56ipod5G60Gbof the ipod
05:23:05DBordellofoo_pod creates the structure
05:23:14BHSPitLappyDBordello: so, foo_pod uses your existing folder setup?
05:23:23DBordelloit just builds the db, nothing else
05:23:28BHSPitLappythat's interesting...
05:23:41BHSPitLappyI've only used foo_pod to pull music off
05:23:56BHSPitLappydoes it freak out when you plug the ipod back into itunes?
05:25:26DBordellohmmm, don't :)
05:25:33DBordellobut, i think the easiest way to do this now
05:25:45DBordellois to put your music on using itunes, and then use tag cache in rockbox
05:25:49DBordello both will behave the same
05:26:02ipod5G60Gbwait so explain to me
05:26:49DBordellodo you perfer the directory structure or the database
05:26:53DBordelloto browse your music
05:27:48ipod5G60Gbi want both worlds if possible
05:28:18DBordelloin rockbox, which do you perfer
05:28:20ipod5G60Gbthats what i like about RB other than that i see few reasons for running RB
05:28:43midkayipod5G60Gb, i haven't heard any success stories about foo_pod, and i tried it myself and failed..
05:28:52midkayipod5G60Gb, at least for the 5G.
05:29:17ipod5G60Gbok, thanks, i read that same statement in a furom while looking to D/l
05:29:26midkaydon't waste your time
05:30:29ipod5G60Gbwell is it possile to have both worlds boot retail like normail and use my music and boot RB and file direct browse through it that way? what is tag chache?
05:30:54midkayipod5G60Gb, the apple firmware is bootable, sure.
05:31:01midkayjust hold MENU right after you turn on your ipod.
05:31:07midkayshould load to default firmware.
05:31:35midkaytagcache is available in the menus.. scans your disk for music and lets you browse it similar to the apple firmware (artists, albums, etc).
05:33:45ipod5G60Gbdo any of you run podzilla
05:34:05midkayrarely, but yes..
05:36:11ipod5G60Gbcool, so i nuked my ipod and a linux partion and am running all three programs. If i put the music back on via itunes will i be able to view the music through RB and PZ2 or is this possible
05:36:22ipod5G60Gbsorry, installed a linux par
05:36:51midkayboth ipodlinux and rockbox can play the music loaded via itunes, in a way.
05:37:03midkayonly through rockbox's tagcache can you play what you want easily, though.
05:37:16midkaysince itunes renames them all to random characters and stuff.
05:38:29ipod5G60Gbyeah, thats what i heard, ok well i didnt want to transfer all 35 gigs of music back on before i knew i was doing it right
05:40:54BHSPitLappyDBordello: I don't think the tag viewer works...
05:41:33ipod5G60Gbi loaded PZ2, and RB this morning so i am a newbie, but am I right in saying PZ2 doesnt not play music but plays games and apps?
05:41:42DBordelloBHSPitLappy, huh?
05:41:53midkayBHSPitLappy, the 'tag viewer'?
05:41:57BHSPitLappy"[DBordello] is to put your music on using itunes, and then use tag cache in rockbox"
05:42:08midkayindeed it works.
05:42:31midkayipod5G60Gb, PZ2 only really seems to have games/crap at the moment. i haven't figured out audio playback, nor do i really want to, it's so slow..
05:43:11ipod5G60Gbhaha, i wanted to play around with the ines, and idoom. havent figured out either so far
05:43:23midkayidoom isn't 5g-compatible, 'not yet'..
05:43:29midkaynor do i believe is ines.
05:44:53ipod5G60Gbhaha great, so im SOL i had thought i had seen idoom on a 5G on google video
05:45:24midkaygoogle video has a ton of losers playing doom on their nano like they're special, but no, last i heard (maybe a week ago), idoom isn't quite working yet..
05:46:12ipod5G60Gbisnes seems like it would be impossbile to navigate buttons would be redicoulus, but a cool idea, and the fact the programers want to take there time is no concern of mine
05:46:50ipod5G60Gbim bummed about idoom however
05:46:57midkayipod5G60Gb, well, iirc ines/whatever emulates buttons with the wheel also..
05:47:19midkaye.g. touching it at a 45 degree angle from the right/next track button would be read as "A" or something.
05:47:33midkayand 135 degrees, B.. stuff like that.
05:47:36ipod5G60Gbyeah, i read that the ipod 5g has 8 buttons
05:48:11midkayyou could emulate about a hundred, but it wouldn't be very fun. :) that's how it works, as far as i know..
05:50:22ipod5G60Gbno that would be more work than it worth
05:51:22midkaywell, it's not much coding work.. just a pain in the ass for users.. :)
06:00:00 Nick Sinbios is now known as Sin|showah (
06:00:42 Join RoC_MM [0] (
06:01:17 Join Rob2222_ [0] (
06:07:10 Quit Vertigo_t (Read error: 110 (Connection timed out))
06:17:05 Quit maeck ("Chatzilla 0.9.65 [Mozilla rv:1.7.12/20060202]")
06:18:14 Quit Rob2222 (Read error: 110 (Connection timed out))
06:37:59 Join unhcr [0] (
06:38:10unhcrhey is anyone here working on the iaudio x5 port?
06:38:29 Quit ipod5G60Gb ("CGI:IRC (EOF)")
06:38:29 Quit HuntermicAtWork ("CGI:IRC (EOF)")
06:38:57JdGordonunhcr: u gota wait around for LinusN
06:39:11midkayunhcr, not atm, but if you have any questions, maybe you could ask..
06:39:36unhcri watned just to know
06:39:40unhcrhow to go about
06:39:49unhcrcompiling the album art patch
06:40:05JdGordondownload the source, patch the source, compile
06:40:16JdGordonnothignspecial in that resspect for the x5
06:40:34unhcrdont i have to do something
06:40:38unhcrcause of the new feature
06:40:42unhcrthat had screwed it up
06:40:44unhcrand it wasnt wokring
06:40:59midkaydo you need to press enter every three words or so? :)
06:41:08unhcryah sorry
06:41:17JdGordonum.. ye, but its not gonna work on the x5 if its not working on other targets
06:41:22midkaywe're not going to run away if there's not a constant stream of words or anything, btw..
06:41:23JdGordoni mena, if its not working in general
06:41:32*JdGordon might!
06:41:41midkaywell, JdGordon's weird like that.
06:42:27unhcrhmm i cant download
06:42:33 Join muesli__ [0] (n=muesli_t@
06:42:36unhcrthe daily build of the iaudio so i cant do it lol
06:44:13unhcrcan you help me
06:44:18unhcrlike compile it and stuff
06:44:29unhcr do i download the latest patch all the way on the bottom
06:51:15 Join TCK [0] (
07:01:37preglowmidkay: yo, you broke the pitch screen behaviour
07:03:15midkaypreglow, yo, in what way?
07:04:53preglowmidkay: it doesn't snap back when you press the sideways keys, then release them
07:05:00preglowmidkay: it used to be that way before for a reason
07:05:09midkaypreglow, that's intended.. what reason was that?
07:05:14preglownamely dj use
07:05:33midkayadd two percent temporarily? i see absolutely no point..
07:05:35preglowthe whole screen looks like it does for dj use
07:06:18midkaywith the irivers and ipods, anyways, it takes a few seconds for the changes to take effect due to the buffer size..
07:06:30preglowwell, yeah, but you broke it for archos as well
07:06:49midkayso.. it should function differently on archos? makes little sense..
07:06:56preglowwe should make another pitch setting for the settings screen anyway and not rely on that screen
07:07:12midkaypreglow, 'settings screen'?
07:07:13preglowthis isn't even saved
07:07:23midkaythat is also intended, it never was before..
07:07:33midkayapparently saving of that setting is unwanted.
07:07:46lostlogicpreglow: are you back up again?!
07:07:55preglowlostlogic: unfortunately :///
07:08:00midkaythe problem is that 0.1% steps are mandatory and that takes unbelievably forever to adjust in a standard settings menu.
07:08:11lostlogicdamn, that sucks dude... thought you hit the hay a bit early last night.
07:08:26preglowmidkay: yup, some settings acceleration would be sweet
07:08:31midkaywhich is why i made left/right actually add/subtract 2% instead of just temporarily..
07:08:40preglowmidkay: but yeah, the old screen was the way it was for a reason, it's unusable for that purpose now
07:08:46lostlogicand now, I sleep.
07:08:50preglowlostlogic: gnight
07:08:53midkayi think quick adjustment is way more important than simply adding or subtracting 2%. why just only 2%?
07:08:57midkaynight lostlogic
07:09:11 Quit bluey ("Leaving")
07:09:28preglowmidkay: because you need small adjustments fast for mixing
07:09:47preglowget two units and try it to find out :)
07:09:58preglowand btw, there is a way to make the latency of those settings way less
07:10:13midkaywhat kind of mixing? it's such a small step.. besides, you can do the same thing about as easy - just tap right to add 2% and instead of "releasing" to go back 2, press left.
07:10:36midkaypreglow, i don't doubt it, but i didn't want to dig around for any of that..
07:11:07preglowmidkay: it's pretty easy, we do it for the eq screen, gimme a sec
07:11:27midkaypreglow, ah, "way less" then as in still a few seconds.. hmm..
07:12:15preglowi don't like that solution at all, but it's all we've got until we rework the playback system to something a bit more low latency
07:12:27midkayah well, i guess..
07:12:41 Join P1mP [0] (n=muesli_t@
07:12:54***Saving seen data "./dancer.seen"
07:13:02unhcr do i download the latest patch all the way on the bottom
07:14:03midkayi just don't think that the number of archos users, running rockbox, who even USE the pitch screen, less for DJ capabilities, less who need INSTA-change of a small set amount warrants changing the functionality - it's just about as easy to press left as to release right..
07:14:05 Quit mikearthur ("Konversation terminated!")
07:14:18midkayunhcr, why not?
07:14:24midkayit's the latest..
07:14:27 Join B4gder [0] (
07:15:17unhcrhow do i compile it though
07:15:18preglowmidkay: but ok, that screen is unusable for that purpose now anyway, why just not remove it and make a regular setting to cut down some code bloat?
07:15:22unhcrwith the x5 firmware im new im soryr
07:15:37midkaypreglow, because it takes forever to adjust with a standard menu. believe me, i wanted to do it that way.
07:15:47midkaycode bloat? the pitch screen is like.. 40 lines..
07:15:57preglowwhich is less than ordinary settings code
07:16:04midkayversus probably 15-20 overall to use a standard menu.
07:16:18midkayoooh, my god, 20 more lines to be able to adjust it considerably easier. :)
07:16:20preglowthe right way to do that would be to add some button acceleration code
07:16:26preglowa ton of other settings would love that as well
07:16:27midkaythat'd be way more than 20 lines..
07:16:29preglowso please do it :>
07:16:38midkaypreglow, feel free - i don't know how. :)
07:16:41preglownah, not much
07:16:42unhcrhow do i compile it though
07:16:55B4gderunhcr: what docs have you tried reading?
07:17:13midkayB4gder, none, i think. he just wants the album art patch on his X5..
07:17:14preglowyeah, but please consider asking people before you change behaviour that's been static for a year or two
07:17:25preglowsome people might easily have been depending on the old behaviour
07:17:30preglowand i for one see now reason to break it
07:17:41B4gderunhcr: 1) build a full rockbox 2) apply the patch and rebuild 3) profit
07:17:54midkaypreglow, the number of archos users who use the pitch screen, much less for DJing, much less who will complain about releasing right as opposed to tapping left... i haven't heard of any complaints yet, i didn't realize it was such a big deal.
07:18:57preglowtapping left isn't as fast
07:19:04unhcrokay b4gder
07:19:07unhcrcan you tell me how to do that
07:19:09midkaybasically, now instead of releasing right we tap left, but we can also adjust the actual value way quicker, and we don't need to have our finger on RIGHT for however long it takes.
07:19:10unhcri need an FAQ
07:19:11unhcror something
07:19:27unhcrno one on the forums wants to do it and i want the album art and i know itll work
07:19:31midkaypreglow, and how useful is a hardlocked +/- 2%?
07:19:36preglowmidkay: very, actually
07:19:37midkayi just don't see it..
07:19:43preglowmidkay: like i said, you should try beatmatching some times
07:19:53B4gderunhcr: I told you. 1) build rockbox from source 2) apply patch
07:19:53midkayi don't know what that is or why you'd try to do it.. :)
07:19:57 Quit imphasing (Connection timed out)
07:20:04unhcrhow do i do that
07:20:07unhcri dont know what to do lol
07:20:09midkaynor why adding 2% temporarily would be useful..
07:20:16B4gderunhcr: you read the docs, we have PLENTY
07:20:19preglowif the mix is off by just a smidgeon, you wnat to nudge it forward just a little bit
07:20:33preglowand for that it's nice to just tap a button gently instead of pressing right than left
07:20:39preglowwhich isn't exactly a precision operation
07:21:10preglowbut anywho, if i'm the only one who cares, it doesn't matter
07:21:11unhcrwhere do i read them
07:21:16midkaypreglow, maybe there's another way, let me think about it..
07:21:24unhcris there any way someone can do this for me
07:21:29preglowunhcr: no
07:21:31B4gderunhcr: we have a little site:
07:21:45B4gderstat here for example =>
07:22:08unhcrokay thank you
07:22:15midkaypreglow, i do think it's rather pointless and i don't think i/we'd get any more complaints ;) but i'm very willing to consider yours, and i'm sorry for changing that, i wasn't aware that it was so important. i can try it on some different buttons or using a combo, or something.
07:22:39preglowmidkay: that code's from before my time, i just adapteted it for swcodec
07:23:00midkaypreglow, okay...?
07:23:28unhcrgod ill do this tomrorw
07:23:31unhcrim going to fuck it all up
07:23:44 Quit muesli__ (Connection timed out)
07:23:54unhcrhow long should it take me to patch my firmware with album art
07:24:03B4gderunhcr: the patch takes 1 second
07:24:09unhcryes but i mean like
07:24:13unhcrsetting up this stuff
07:24:15B4gdergetting everything else up will take you hours
07:24:27preglowOH NOES
07:24:29midkayany objections to making "first keypress enables backlight" to default OFF for all targets instead of just the non-color LCDs? who decided that color LCDs should have that on by default+why?
07:24:46preglowmidkay: zagor, i believe
07:24:53midkayhe's nutso!
07:24:57unhcrwhen is the guy who does the x5 around
07:25:03preglowbut yeah, sounds like a good idea to me
07:25:06*B4gder confirms Zagor is a "nutso" ;-)
07:25:16midkayunhcr, he's not going to compile a patch for you, i'm quite sure..
07:25:17preglowi've got that behaviour on my clickwheel anyway :PP
07:25:24B4gderbut hey, I'm stuck with him!
07:25:27midkaypreglow, haha, yeah.
07:25:33midkayB4gder, :)
07:25:35unhcrwhy woulndt he if he has his stuff set up
07:25:39unhcrdoesnt he just have to add the patch
07:25:42preglowunhcr: because it's not his job
07:25:45preglowunhcr: and it'd waste his time
07:25:49preglowunhcr: go on, learning is fun
07:25:54unhcrhed do it in 20 second
07:26:01preglowunhcr: which is besides the point
07:26:06unhcrokay god bye
07:26:08midkayunhcr, all of us "have our stuff set up". if you want a patch, feel free to do it.. we're working, not compiling for users who want patches...
07:26:36B4gderand it is quite easy and fun to have it going anyway
07:26:39lostlogicmidkay: I think it should default to on on h3x0 and x5 only.
07:26:46midkaylostlogic, why?
07:26:48 Quit unhcr ()
07:26:58preglowi usually just give up whenever i want something done
07:27:02lostlogicmidkay: because you can't see their screens _at all_ withou tthe backlight (at least the H3x0 for sure)
07:27:04preglowway faster than the ardeous process of learning
07:27:13B4gderI figure the reason would be that they are the totally unreadable ones
07:27:27midkaylostlogic, well, for one i think users might be a bit weirded out that their initial keypresses aren't working..
07:27:40preglowoh, it's working
07:27:57lostlogicmidkay: that's what stock firmware does on H3x0
07:27:59amiconnmoorning :)
07:28:09lostlogic(not that we should base shit on stock firmware, and why am I awake?)
07:28:10 Join Matze [0] (
07:28:11preglowamiconn: at least you don't claim it's a good one
07:28:11 Quit RoC_MM (Read error: 104 (Connection reset by peer))
07:28:21midkayfor two, it seems like a rather small number of people want such functionality. i have no problem with adjusting the volume by 1 when i want the backlight on.. or moving the cursor in the browser..
07:28:24preglowlostlogic: i was wondering that...
07:28:29midkaylostlogic, hmm..
07:28:36midkaylostlogic, hahaha :)
07:28:42amiconns/moorning/good morning/ :)
07:28:44preglowi just figured it was because the time is 07:28
07:28:55preglowand going to bed now would be insane, but that's just me
07:29:01lostlogichmm... I don't care one way or the other... I still want "Backlight Hold-Off" functionality
07:29:28midkaylostlogic, the ipod 5g's screen to me is just about unreadable except in considerable light levels, without the backlight.. and yet with it on by default i kept going wtf..
07:29:29lostlogicpreglow: good point... I should get ready for work... :-\ Europe... who thought of that.
07:29:46midkayafter like 10 seeming dropped keypresses i realized the option was on for some reason..
07:30:07lostlogicmidkay: same on 5g here. I agree that for the 5g with the touch wheel interrupt turning on the backlight w/o any action the feature should be either nonexistant or at least disabled.
07:30:30midkaylostlogic, true dat :) hmm..
07:30:32lostlogicbut that's specific to the 5g.
07:31:23preglowlostlogic: oh, and btw, you planning to commit the absolute wheel patch some day?
07:31:38*amiconn finished the first round of english.lang rework last night.
07:31:40lostlogicpreglow: the one on patchtracker?
07:31:43preglowamiconn: wooo
07:31:47preglowlostlogic: yea
07:31:56amiconnNow all that's needed is a perl script that adjusts all other .lang files
07:32:05amiconnThen I can commit that stuff
07:32:52lostlogicpreglow: I'm not terribly happy to go away from completely event based wheel handling...
07:33:05preglowlostlogic: no, but it could be one possible use
07:33:15preglowlostlogic: it actually makes brickmania possible to play, you see :)
07:34:01lostlogicpreglow: yeah, I know, I keep meaning to experiment with some of my other ideas with the same goal, but I haven't gotten to it yet... work's been eating me lately.
07:34:43midkaybtw.. 2 questions. (1) objections to a crossfade mode for track skipping only? e.g. gapless except when you change tracks manually. very nice IMO. (2) putting it before "always" in the selector menu (e.g. Off, Shuffle, Track Skip Only, Always) - this means that anyone currently having it set to 'always' ends up with it on 'track skip only' on new builds - doesn't sound too good. doesn't make sense to have it after 'always' though..
07:38:34preglowi'd never use that on the simple grounds that it'll waste tons of battery just for the few times i skip
07:38:42preglowgod, how i hate the crossfade implementation
07:38:47 Quit JdGordon (Read error: 104 (Connection reset by peer))
07:38:56midkaypreglow, well, if you hate crossfade, your input is semi-useless. :)
07:39:10preglowi don't hate crossfade, i hate how we have to do it
07:39:17preglowwith the 10 second long buffers and the extreme latency
07:39:29lostlogicpreglow: suggestion for better impl.?
07:39:45preglowlostlogic: well, there can only be one other, and that's running two codecs in parallel
07:39:49preglowlostlogic: which probably isn't too feasible
07:39:53midkayi don't like crossfading normally.. annoying for gapless albums, just kind of ruins the flow altogether. but when i want to skip to a new track in the middle of another, i really like fading there - the music doesn't just stop and start on a new track...
07:40:18midkayi guess no input is good input..
07:40:26 Join Bg3r [0] (n=Bager@
07:40:38lostlogicpreglow: we have voice thread... but it's way less practical to constantly switch codecs for 15s than it is to just have the 15s buffer pre decoded
07:41:06preglowlostlogic: yes, i didn't claim the alternative was the way to go :)
07:41:09B4gderthe iaudiofile guys are offering a whopping 60USD bounty for an 'audio driver with BBE-like + "3D Surround" enhancements'
07:41:24B4gderPROFIT! ;-)
07:41:27lostlogicwow, that's huge.
07:41:36midkay3d surround? um.
07:41:37*lostlogic vomits on '3d sound'
07:41:37preglowthe long buffer wouldn't matter to me at all if it wasn't for the dsp latency
07:41:44midkayjust add in a little.. tinny effect..
07:41:49lostlogicpreglow: ??
07:41:50midkayvoila, 60usd.
07:42:05preglowlostlogic: try adjusting dsp settings at 10sec crossfade
07:42:12amiconnlostlogic: Colour sound then? ;)
07:42:22lostlogicpreglow: doesn't low latency mode work even with crossfade enabled?
07:42:23*amiconn dislikes audio fade in general
07:42:26midkayjust add a "3D Sound" option, when enabled overrides the user's EQ settings with some weird ones..
07:42:35lostlogicamiconn: please, color sound.
07:42:36preglowlostlogic: yeah, and defeats the entire purpose of the long buffer in the first place
07:42:50lostlogicpreglow: only while you're adjusting settings... seems a reasonable solution to me
07:43:07preglowlostlogic: sure, but what if the track changes while you're adjusting settings?
07:43:17lostlogicyou don't get crossfade, gee darn zomg death
07:43:27lostlogic(not that I use crossfade myself, I'm just sayin')
07:43:45preglowfeatures not suddenly working is supposed to be a big deal, you know
07:44:29preglowh120: "wtf, no time for crossfade!"
07:44:37preglowuser: "what the hell, rockbox sucks!"
07:44:39amiconnpreglow: There's a third option that allows crossfading, doesn't run codecs in parallel, and offers low latency for sound adjustment.
07:44:41preglowh120: "kekeke"
07:44:42lostlogicugh, can't have everything... and if I don't go to sleep, I'm not going to get up and if I don't get up, I'm not going to ever work on that track skipping synchronization related crap I've been saying I'd work on.
07:44:48preglowthere's a scenario for you!!!!!111
07:44:52lostlogicamiconn: ??
07:45:00midkayat the least, any input on the insertion of a setting? that changes what the user had selected before.
07:45:02amiconnIt means more buffering though
07:45:39amiconnDecode and crossfade into an intermediate buffer, large enough to hold data for the whole crossfade time.
07:45:56preglowisn't that what we're doing currently?
07:45:58amiconnThen apply dsp stuff shortly before actually playing the data
07:46:03preglowahh, yeah
07:46:09preglowi'd like that
07:46:16Bg3rmorning :)
07:46:22preglowbut it'd double the space requirements
07:46:27preglowwhich i don't really care about
07:47:10lostlogicI've got a better idea... apply the DSP as the last step before the pcm buffer chunks are sent to the DMA/FIQ?
07:47:14midkayamiconn, no biggie if i add a lang string to cvs is it?
07:47:21preglowlostlogic: problem is the samples are 16 bit by then
07:47:36preglowlostlogic: we'd need to let the pcm buffer be 32 bit
07:47:37lostlogicpreglow: not if we didn't DSP them before...
07:47:41amiconnpreglow: What *I+ would actually like would be to cut out crossfade completely. Substantially reduced complexity with no lost functionality (for me)
07:47:49lostlogicpreglow: which would be a simple case of moving where DSPing is done
07:47:57lostlogicadding complexity to the pcmbuffer but removing it from playback.c
07:47:58*amiconn always has all fades disabled
07:47:59lostlogic(which I like)
07:48:13preglowlostlogic: sure, sounds nice, but the space requirements will double, and wow, how i don't care
07:48:22lostlogicpreglow: why would space requirements double
07:48:29preglowlostlogic: 32 bits > 16 bits
07:48:50lostlogicpreglow: oh, right.
07:49:01amiconnmidkay: If you add just one string, no. Means I have to hand-insert it here (as I'll surely get a cvs conflict)
07:49:13amiconnpreglow: Why 32 bit?
07:49:14 Join Mantice [0] (
07:49:20midkayamiconn, so you don't mind?
07:49:22preglowamiconn: that's the current output format of our codecs
07:49:59preglowamiconn: if we truncate that to 16 bits in the pcm buffer, then the later dsp will see a reduction in quality
07:50:11amiconnHow much? Noticeable?
07:50:18lostlogicwell regardless, it's not happening now
07:50:24preglowdepends on the process in question
07:50:29preglowfor some, not much, for others, pretty much
07:50:34preglowlostlogic: sure ain't
07:50:45amiconnWe could use 24 bit...
07:50:52preglowamiconn: don't go there...
07:51:06preglowit'll slow everything down
07:51:12amiconnBetter quality than 16bit, and less space requirements than 32bit
07:51:21 Part Mantice
07:51:56amiconnI don't think 24bit is any slower than 32bit, when using dram (and optimising the access)
07:51:57Mikachuwouldn't you need to read in two longs and mask out parts and combine them in different ways depending on alignment?
07:52:20Mikachuunless you can read longs non-long aligned
07:52:27preglowyou can, some places
07:52:40amiconnMikachu: coldfire can, but there's a performance penalty
07:52:47preglowamiconn: of course it is slower, on arm, you'd need to bloody read three times for every sample
07:52:49amiconnSo I would avoid it
07:52:59preglowgood bye block reading
07:52:59amiconnpreglow: I don't think so
07:53:14amiconnYou can read 3 longwords and convert these to 4 samples
07:53:38 Join webguest79 [0] (
07:53:53preglowyes, and we're restricted to reading only three at a time
07:54:00preglowsome places i read two, other places four
07:54:04preglowother places again, eight
07:54:13preglowdepends on how many registers i have left
07:54:45preglowhah, my nano is still going
07:54:49amiconnWell, from all variants you mentioned, only reading 2 samples would be difficult
07:54:58amiconn4, 8 etc are perfectly possible
07:55:07preglowamiconn: but leaves the pointer in a non-aligned state
07:55:08 Quit webguest79 (Client Quit)
07:55:42amiconnIf I'm reading 3 longwords to retrieve 4 samples, the pointer is still longword aligned
07:56:08preglowahh, yeah
07:56:17preglowtrue, true, i'm messing up
07:56:26amiconnReading 4 samples as 32bit vs. reading 4 samples as 24 bit - no difference in pointer alignment, only that I read one longword less for 24bit
07:56:28preglowbut still
07:56:30preglowlet's not do this :>
07:56:42amiconnYou don't even need any extra register
07:57:07lostlogicpreglow: brilliant
07:57:16amiconn(but it means some bitshifting)
07:57:18preglowwe'd need to change the codec output format _again_, and change all dsp code
07:57:19lostlogiclet's put mention in the manual what settings temporarily disable crossfade
07:57:31lostlogicnow it's a featuer, not a but
07:57:49Mikachuthat's a good way of fixing bugs
07:58:05amiconn24bit pcm might end up faster on coldfire...
07:58:07lostlogicit's one of my favorites.
07:58:20preglowlostlogic: all sounds settings but volume and panning will, then
07:58:25preglowor at least should
07:58:40lostlogicpreglow: so let's just say "When changing sound settings, crossfade is temporarily disabled"
07:58:46lostlogicnow it's even a single sentence bug fix :)
07:58:52amiconnpreglow: ??
07:58:53preglowon ipod you might see skips, but i guess that wouldn't change even if we were to move dsp to a latter part in the chain
07:59:06amiconnI'd think all hw sound settings don't touch crossfade
07:59:15amiconn...i.e. volume, treble, bass
07:59:55preglowamiconn: yeah, like i said
08:00:00preglowvolume, panning, bass, etc
08:00:02lostlogicso we need a "Changing some sound settings disables the crossfade feature until the settings are complete"
08:00:18amiconnErm, but panning is software??
08:00:29preglowit is?
08:00:33amiconnAh, no, you mean balance...
08:00:43preglowbalance/pan, same thing :(
08:00:45preglow:) <
08:00:55preglownothing like just typing in the wrong emoticon
08:00:58amiconnYeah, /me was confused
08:05:33 Quit RotAtoR ("zzzzzzzzzz")
08:07:29Slasherithen we need 32bit pcm buffer
08:07:32amiconnB4gder: Just a thought on langv2, regarding yesterday's idea: We'll need n:m matching of target names
08:07:39Slasheriups, never mind
08:08:10amiconnI.e. both the target field in th .lang files and the filter needs to be able to process lists
08:08:12preglowSlasheri: yeah, and fast dsp :)
08:08:27Slasherii just had forgotten to scroll down the log and answered to some old logs :)
08:08:34Slasherihehe, yeah
08:08:40 Join ender` [0] (i=ychat@
08:08:45*amiconn votes for 24bit pcm
08:10:48midkayhmm.. when you update rockbox and there's a new settings version or whatever and it resets everything - is the new settings struct version detected on-target and reset on-target?
08:11:20preglowit's saved in the config block
08:11:30preglowit has to match the current version in the rockbox code
08:12:14midkayso.. the new version is loaded, checks the config block version/whatever, and if it's not the same, the settings are reset?
08:12:22preglowdefaults all around
08:12:33B4gderamiconn: n:m ?
08:12:49*preglow definitely votes for 32 bit pcm, on the grounds of him actually having to work on it :)
08:13:03amiconnB4gder: Yes, as explained afterwards.
08:13:26B4gderI don't get it
08:13:29midkayi really think the user should be prompted to save a config file before the settings are reset, if a different version is detected.. that'd be a great feature. losing your settings inadvertently sucks..
08:13:31amiconnWe will need something like target: recv1, recv2, fmr in the .lang file,
08:13:43preglowmidkay: we'll be moving to pure .cfg files
08:13:49B4gderamiconn: it's already supported
08:13:53amiconnand match that against e.g. recv1, alarmmod
08:14:06midkaypreglow, that adds a nice few seconds to boot time, doesn't it?
08:14:11midkayas well as shutdown time..
08:14:15preglowmidkay: then we're doing something really wrong...
08:14:18 Join mnhnhyouh [0] (
08:14:23preglowparsing a text file sure as hell shouldn't add seconds
08:14:29midkayif it's loaded from disk, it takes a couple seconds..
08:14:31B4gderyou can make a string like "h100, h300, archos*: text"
08:14:35amiconnWell, the .cfg read could need some optimisation
08:14:36midkayand saving a CFG file from the menu takes like 5 seconds for me..
08:14:38preglowmidkay: where do you think the config secvtor comes from?
08:14:45midkaypreglow, what?
08:14:56preglowmidkay: the current config is also stored on disk
08:15:11midkaypreglow, i thought it was in RAM..
08:15:15preglowhow can it be?
08:15:17preglowram is cleared
08:15:21midkayor. wait. no..
08:15:27midkayhmm. never really thought about it. :)
08:15:33midkaybut. wait.
08:15:42preglowsettings are stored in a special binary block after the partition table
08:15:53midkayso it's just the actual config file that takes a long time i think. you're not just writing settings, but a bunch of other stuff too.. text, etc.
08:16:18preglowwell, like i said, there's no reason it should be super-slow
08:16:28preglowso if it is, we need to work on it
08:16:29midkaypreglow, if it's used that way currently, you're right.
08:16:33amiconnmidkay: What takes most of the time is the disk spinup.
08:16:35midkayit's not slow at all as it is now..
08:16:43midkayamiconn, for what?
08:17:00amiconnOf your stated 5 seconds for saving a .cfg
08:17:26midkayamiconn, you're right..
08:17:40amiconnSwitching to .cfg files has multiple advantages:
08:17:45midkayalways seemed to take longer than i thought.. but i just checked, with the disk on already it's only like 2 seconds or a bit less..
08:18:06midkayamiconn, i agree that it's a nice idea :)
08:18:06amiconn- no more 'config' version which would reset settings occasionally
08:18:14midkayah, great..
08:18:38preglowi'm all for cfg files, but it probably will be very slightly slower
08:18:41amiconn- saving the settings in a user-visible place allows for easy clearing of settings from the PC
08:18:47amiconn(just delete the file)
08:19:17midkayright, and manual adjustment too..
08:19:22midkaye.g. if you screw up your color settings..
08:19:28amiconn- It would work always. The config sector approach needs two preconditions to work
08:19:29midkaybut don't want to lose everything else..
08:20:00amiconn(1) It won't work on superfloppy formatted media, as there is no partition table.
08:20:21B4gderthe .cfg might not be as fun for very frequent resume position storage
08:20:38amiconn(2) If the number of sectors per track is low, the config sector might get overwritten by the computer OS accessing it
08:21:04midkayamiconn, ah.. cool. :)
08:21:10midkaysounds like a great idea - so why isn't it done already? ;)
08:21:22amiconn(the latter is *Not* made up, we actually had one such occasion with 8 sectors per track and MacOS overwriting rockbox' settings at each connect!)
08:21:37midkaymac. could have guessed. ;)
08:21:54amiconnB4gder: Yes, and that's why *some* settings will stay binary (but in an extra file, not the config block)
08:22:11midkayisn't resume position in /.rockbox/.playlist_control?
08:22:35amiconnNo, that's the description how the playlist has to be rebuilt
08:23:26midkayoh, also - what about a bmp viewer plugin, any reason it's nonexistant? considering the WPS can load/display BMP files... just nobody's done it yet?
08:23:42preglowno one done it yet
08:23:46preglowbut it'll come
08:24:00SlasheriHmm, why not to store binary data as fixed length hex string in the configuration file?
08:24:01midkaymaybe i'll try it sometime..
08:24:01preglowwe might want to contemplate a multi-image viewer for the newer targets
08:24:05 Part mnhnhyouh
08:24:11preglowSlasheri: 'cause then we can't use the eeprom for it! :)
08:24:21Slasherihmm :D
08:24:57preglowdid you ever make that work, btw?
08:25:02preglowi remember you had some problems
08:25:06Slasheriyes, eeprom reading works fine
08:25:11 Join JdGordon [0] (
08:25:14Slasheriat least it worked when i tried it the last time
08:25:17preglowtried writing?
08:25:22Slasherinot tried that..
08:25:36Slasherii would need to reinstall the iriver fw somehow..
08:25:54preglowwhy would you need that?
08:25:55Slasheriand then modify the eeprom content to find out how iriver fw reacts
08:25:59preglowahh, yes
08:26:13 Join pinkzeppelin [0] (
08:26:14 Part biffhero_
08:26:17preglowright, you're the only guy in the world with no iriver firmware, forgot about that :)
08:26:24Slasheriit would be good to find a safe location in eeprom for use with the bootloader :)
08:26:29Slasherihehe :D
08:26:36preglowi don't think there's much space left there
08:26:38amiconnSlasheri: (1) Storing the whole .cfg just for the often-changed binary data would be overkill
08:26:47Slasheripreglow: i need only one bit
08:26:54preglowahh, right
08:27:16amiconn(2) Using a separate file matches better the platforms where we can use true separate storage
08:27:32Slasheriamiconn: hmm, true
08:27:32amiconn(like the rtc ram on archos recorders, or the serial eeprom on iriver)
08:27:33pinkzeppelinHello everybody. I own a Ipod 5G, nobody has commented on the "clicking" problem that the Rockbox firmware exhudes. I posted RMAA analysis results that show the problem in the forum here:
08:28:00B4gderGraphs Are Good (tm)
08:30:07 Quit pinkzeppelin (Client Quit)
08:30:20 Join pinkzeppelin [0] (
08:30:30 Quit JdGordon (Read error: 104 (Connection reset by peer))
08:31:11B4gderpinkzeppelin: it has been discussed in here before
08:31:24B4gderI don't think anyone has any real clue yet to why they happen
08:31:32preglowpinkzeppelin: that's a wav you've used there?
08:31:44pinkzeppelinyes but it is audible even when nothing is playing
08:31:50pinkzeppelinmeaning it isnt a codec, but the audio driver, apparently
08:31:59 Join JdGordon [0] (
08:32:39preglowyeah, sure, i was thinking about something else i'm seeing
08:32:48 Quit JdGordon (Read error: 104 (Connection reset by peer))
08:33:16preglowbut if it's a wav i'll just assume it's a minor variation in the measurement
08:33:53pinkzeppelinyes, they are waves, and i normalized them
08:33:58preglowB4gder: btw, read the new "conclusion" on the x5 highfreq issue on the forums? sounds really unlikely to me
08:34:00pinkzeppelinif that is what you were seeing...
08:34:10preglowi can't see how you can use the wrong sampling rate without hearing it
08:34:28B4gderI agree
08:34:36B4gderthey just throw out wild ideas
08:34:44preglowyou'd actually have to be skipping some samples when writing to the dac
08:34:52preglowwhich isn't really easy to do without wanting to, heh
08:36:02amiconnB4gder: Concerning voice, as you mentioned it on the ML: For easy processing, the converted output should look like today's .lang files, with just the id: and voice: lines
08:36:19preglowif someone would post a sample of how x5 audio sounds, i might be able to figure out what's wrong
08:36:30amiconnNote that it should have *all* IDs, even those without voice
08:36:32B4gderamiconn: ok, I'll fix a special option for that
08:37:12amiconn...and it would be really nice if it would convert to a selectable encoding, instead of writing utf-8
08:37:39 Join JdGordon [0] (
08:37:50preglowignore that, please, and force amiconn to get a proper editor
08:38:02amiconnpreglow: This has nothing to do with the editor
08:38:18amiconnThe sapi speech engines seem to need whatever the default encoding on the used windows installation is
08:38:35amiconn...which happens to be Win1252 (~ISO8859-1) here in europe
08:38:37preglowwell, perl has a nice code page translation module
08:38:46Galoiswindows? who needs that?
08:39:06amiconnIt's not *that* big a concern, if the perl script doesn't do that, the vbscript would have to
08:39:09 Quit JdGordon (Read error: 104 (Connection reset by peer))
08:39:11preglowperhaps i should try experimenting with festival to see if it's feasible to use as a speech synth
08:40:05amiconn..and we already have encoding conversion functions in rockbox, which shouldn't be hard to adapt to vbscript
08:40:57preglowit's a one-liner in perl
08:41:02preglowif i could just remember how...
08:43:36amiconnOh, and btw, since we are going for target-specific voice files anyway, there's no need to keep the bitswap for swcodec platforms...
08:43:49amiconn...and for arm, the tables could be little endian
08:44:02 Quit pinkzeppelin ("CGI:IRC")
08:50:05*amiconn plans to be able to just type 'make voices' in a build dir
08:50:25amiconnAt least on cygwin, as the first step
08:50:57B4gdera worthy quest!
08:52:57amiconnIt will require some configure.vbs in the voice tools folder, which scans for available voices and allows to select which ones to use
08:53:25 Join JdGordon [0] (
08:53:56amiconnThen, 'make voices' would build all selected voices for that target
08:54:35amiconnYou can run windows binaries and hence vbscripts from within cygwin bash, so there's no problem
09:04:32*amiconn now has an idea how this would have to look like
09:06:34amiconnBtw, imho the manual build should work in a similar way,
09:06:51amiconnby typing 'make manual' in a build dir, instead of requiring a separate one
09:07:07amiconnIt uses a sub-dir anyway...
09:08:04 Join biffhero_ [0] (
09:08:13B4gderthe manual build needs some attention yes
09:08:39preglowahh, yeah
09:08:42preglowthat's a good point
09:08:51preglowunless it depends on any configure magixc
09:08:59 Quit TCK ("well, if you say so.")
09:09:23B4gderI don't think it does
09:09:27B4gderor need to at least
09:10:27amiconnIn fact we could even bundle 'manual.pdf' with the full distribution .zip
09:12:05 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
09:12:07 Part LinusN
09:12:25amiconnmister master bouncer...
09:12:26 Join LinusN [0] (
09:12:33LinusNamiconn: that's me
09:12:56 Quit JdGordon (Read error: 104 (Connection reset by peer))
09:12:57***Saving seen data "./dancer.seen"
09:13:16Bg3ramiconn in fact i think se *should* bundle the manual with ver 3.0
09:13:48amiconnCassandra: are you around?
09:16:43 Join luigi [0] (
09:24:59 Quit P1mP (Read error: 110 (Connection timed out))
09:32:31 Join XavierGr [0] (
09:32:47Bg3rXavierGr morning :)
09:32:52Bg3rB4gder here ?:)
09:33:12B4gderat least a bit
09:43:15 Join ZDaWizzieZ [0] (
09:45:32ZDaWizzieZOh, you are that guy who made audio playback work on x5?
09:45:44XavierGryes he is!
09:45:48 Join DJ_Dooms_Day [0] (
09:46:07ZDaWizzieZYea, i'll flash mine x5 once I find my usb cable...
09:47:21 Join ZZZZZ [0] (
09:47:46ZZZZZAnyone have a color screen of the rockbox gui?
09:49:42amiconnWee, my mini is here now :)
09:50:31amiconnJudging from the fact that it has "4GB" written on the back, it should be a 2G
09:50:43amiconnBut no time for hacking right now... :/
09:50:48preglowthe battery benchmark obviously didn't work too well
09:50:51preglowthe file is almost empty
09:51:10preglowit says it's resuming, then nothing more
09:51:29XavierGrwow you ordered a 2GB model and they sent you a 4GB one?
09:51:43 Join Gibbed [0] (
09:51:47 Quit Rick (Nick collision from services.)
09:51:57amiconnXavierGr: 2*G* not 2GB
09:51:59 Nick Gibbed is now known as Rick (
09:52:03crashdXavierGr: he means the fact that it has the size winscribed on the back means it's a second gen
09:52:11amiconnMinis exist in 2 generations
09:52:11XavierGrpreglow: Which target
09:52:11crashdnot that he got a 4gb for a 2gb :0
09:52:15XavierGrah ok my bad
09:52:20preglowXavierGr: nano, but it's my fault, forget about it
09:52:38preglowit's the ata sleep hack i did to make nano rockbox work
09:52:40XavierGryou did a full benchmark?
09:52:49preglowbut nothing was written to the file
09:52:50amiconnThe backlight is nice
09:53:04preglowamiconn: now for trying rockbox on it!!
09:53:17amiconn[09:50:44] <amiconn> But no time for hacking right now... :/
09:53:21preglowmake time :)
09:53:50preglowi've gotta go soon anyway
09:53:53ZZZZZHow do you do a batt bench?
09:53:58preglownot much rockbox time for me today either
09:55:22 Quit Rick (Client Quit)
10:03:06 Join bobTHC [0] (n=bobTHC@
10:04:18 Quit ZDaWizzieZ (Connection timed out)
10:05:07preglowthe graph in the x5 thread seriously doesn't look like anything i'd expect my resampler to spit out
10:05:11preglowtoo high attenuation
10:05:44B4gderthe person doing the graph didn't say he used 22Khz
10:06:03preglowhe very certainly didn't
10:06:19B4gderI think LinusN misread
10:06:29preglowi'm pretty sure rmaa wouldn't measure up to 22khz if it used a 22khz wav file
10:10:06 Join Rick [0] (
10:11:46 Quit midkay (Read error: 110 (Connection timed out))
10:13:12 Join JdGordon [0] (
10:14:57LinusNthe tlv320 data sheet sure isn't the best i've read
10:16:46*preglow has a look
10:20:40preglowit sure isn't...
10:21:23 Quit ZZZZZ ()
10:24:48preglowLinusN: how is the tlv clocked?
10:25:05LinusNthe coldfire provides mclk
10:25:22preglowso 11.2?
10:26:18B4gderanyone figured out a way to rename a flyspray issue?
10:29:56 Join RRR [0] (
10:29:59 Part RRR
10:30:41preglowLinusN: why is the input clock divided by two?
10:31:00LinusNfunny that you asked, i was just checking that out
10:31:17preglowthat'll yield exactly the problem you have now
10:31:30preglowand why isn't a sample rate set?
10:31:48preglowthat is, you're currently running at 48khz
10:31:53LinusN1) the playback was 2x too fast if i didn't divide the clk
10:32:17LinusN2) is it clear that the sampling rate settings apply to the DAC?
10:32:58preglowapplies to both adc and dac, it seems like
10:33:29preglowis there any other way to set it anyway?
10:33:49LinusNthe sample rate is derived from the i2s clock
10:34:07LinusNor rather the word strobe
10:34:15preglowso, you've checked that setting a rate in the SRC reg doesn't have any effect?
10:34:29preglowi would have thought it had an effect on the oversampling filters
10:34:30LinusNi couldn't hear a difference when i set the sample rate to something else
10:34:48LinusNbut then my ears aren't that sensitive :-)
10:34:50 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
10:34:55preglowspectrum analysers are your friend
10:35:04preglowi can't hear anything above 15khz myself
10:35:39LinusNanyway, i'll have a look at the mclk issue
10:35:59preglowbut yeah, it sounds unlikely to me that the oversampling filter is selected based on the i2s clock
10:36:02XavierGr«Who the hell wants to hear actors talk?»
10:36:02XavierGrH. M. Warner, co-founder of Warner Brothers, 1927.
10:36:07preglowso better set the sampling rate correctly
10:36:36LinusNi bet the filter could be way off if mclk is wrong too
10:37:03preglowthe magnitude response in the forum plot looks very much like the start of an oversampling filter
10:37:18LinusNi couldn't tell :-)
10:38:02preglowthe stopband attenuation is just right
10:38:20 Quit JdGordon (Read error: 104 (Connection reset by peer))
10:38:59preglowthough the last part of it isn't exactly right, but heck
10:41:08 Quit ScoTTie ()
10:41:40 Join JdGordon [0] (
10:44:37 Join petur [0] (
10:44:54 Quit JdGordon (Read error: 104 (Connection reset by peer))
10:45:25preglowbut yeah, the i2s setup code is the same as on iriver, i bet, so that code should be right
10:45:32preglowwhich makes it a bit hard to explain how the sound can be too fast
10:45:53LinusNthe coldfire can't generate a clock faster than 11.2mhz
10:46:05 Join JdGordon [0] (
10:47:41preglowthe clock shouldn't be anywhere near that anyway, should it+
10:48:50LinusNit should be 11.2896MHz
10:49:53preglowdoesn't look like there's much point to using 48khz output on this thing, the corner frequency of the filter is the same as for 44.1khz anyway
10:51:13preglowhmm, no, it's not, but the passband looks worse
10:51:38 Quit amiconn (Read error: 110 (Connection timed out))
10:51:38 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
10:52:00LinusNafaics, it sets it to an invalid rate in cvs
10:52:28LinusN0000 isn't in the table
10:52:51 Join RoC_MM [0] (
10:53:03preglowno, not for 11.2
10:53:13preglowit is 48khz for some of the other clocks
10:53:23preglowso setting that to a legal value would be a nice start
10:53:42 Quit JdGordon (Read error: 104 (Connection reset by peer))
10:54:16RoC_MMI was reading on the forums that it's possible to enable the battery indicator for the iPod. I have a 4G and can use it possible to patch or modify one of the source files to enabled the battery indicator?
10:54:16LinusNi'll do that
10:54:36 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
10:54:41preglowRoC_MM: isn't enabled by default? it might not work too well, but should be anbled
10:54:44preglowenabled, even
10:54:50RoC_MMnot on 4G iPod.
10:54:52RoC_MM5G yes.
10:54:59RoC_MM4G it just displays as full always.
10:55:10preglowthen we haven't got calibration values for 4g
10:55:24 Join JdGordon [0] (
10:55:50RoC_MMis there a page that lists commits to iPod target?
10:56:03preglowthere's a page that lists all commits
10:57:08RoC_MMDo you think calibration values will be obtained Any Time Soon for 4G?
10:57:30RoC_MMor anything I could do?
10:58:19preglowRoC_MM: sure, put a full charge on your ipod, then use the battery_benchmark plugin while letting it play a large number of files in repeat mode until it shuts down due to no power
10:58:34preglowRoC_MM: afaik, that's what's needed to calculate what we need
10:58:47B4gderwireless mice are cool enough until you sit there without batteries...
10:59:01*B4gder sits here
10:59:20RoC_MMpreglow, would doing that make the indicator start working, or provide a file I could contribute?
11:00:19LinusNpreglow: got it
11:01:15preglowLinusN: so, what was up?
11:01:25LinusNusing mclk instead of mclk/2 gives more highs
11:01:39preglowhow'd you compensate for the speedup, then?
11:01:52XavierGrpreglow does battery_bench works for 4G Ipods?
11:01:56preglowXavierGr: no idea
11:02:03preglowlinuxstb is the only one who's got one
11:02:14RoC_MMI have one!
11:02:25RoC_MMWhat exactly do I get or do once battery_benchmark finishes?
11:02:41XavierGrRoC_MM: If you are goint to do a battery benchmark using the plugin, test it for a couple of minutes to see if you get any output.
11:02:45LinusNpreglow: the mclk input setting on the tlv320 is only for the filters
11:03:11preglowLinusN: but i thought you said you got speedy audio if you set it
11:03:28LinusNmy bad, that was the pll setting in the coldfire
11:03:35preglowwell, then all is well
11:03:44LinusNso the filter is still not right
11:04:08XavierGrRoC: first let it run for 10 minutes, then if you get a file that has measured time and voltage level you can do a full benchmark.
11:04:20preglowLinusN: oh?
11:04:52LinusNsince the mclk output from the coldfire is fin/2
11:05:15LinusNi'll have to fix that
11:05:22RoC_MMpreglow, It seems that doing this would give me an idea of the batteries life under normal conditions, but would it get the battery indicator working or could I contribute to that effort in any way?
11:05:23preglowbut what about the sfreq setting?
11:05:27preglowdid that have any impact?
11:05:38XavierGrI guess that will be the first good use of the plugin. Taking measurements to balance battery calibration values for future targets :)
11:05:39preglowRoC_MM: it would help us get the indicator working, yes
11:05:42LinusNpreglow: no
11:06:06RoC_MMWhere should I upload the file once I do a full benchmark (assuming it logs correctly)
11:06:39XavierGrRoC: If you do a full benchamark. And volt is measured with time then a developer can take those numbers and calibrate the battery thingy.
11:06:45XavierGrRoC: On the wiki
11:06:49preglowyou can upload it there
11:06:53preglowand tell one of us it's there
11:06:56preglowso we can use it
11:07:37preglow4gs with 60 disks exists? or should those "4g" be "5g" in that page
11:07:52preglow60 gig disks, that is
11:09:01XavierGrRoC: Please tell us if the plugin will manage to measure voltage and runtime.
11:09:39RoC_MMMost of my MP3's are mixed bitrates, between 128 and 256, with some 320 MP3s and a few q5 Oggs thrown I need to only play one select format and bitrate for my benchmark results to be useful?
11:10:23RoC_MMlike just one album or something?
11:10:31XavierGrwell usually benchmarks are usefull on strict rules, but you can do virtually anything you like.
11:10:49XavierGryeah, choose a specific album more than 50-60MB
11:10:56XavierGrand set to repeat
11:11:26XavierGr128 kbps would be good to compare with the default firmware values
11:11:37XavierGrso does it capture voltage?
11:11:48RoC_MMonly got one reading
11:11:52RoC_MMI will start it again
11:12:52XavierGrhmmm you will have to let it for some time. It will log when you ÇD is active (from idle) and the voltage has changed
11:13:00***Saving seen data "./dancer.seen"
11:13:27XavierGrso play a track let it buffer and then leave it alone for 30 secs, then play another track e.t.c
11:14:03XavierGrdo this for 5 times and then load another plugin (to end the benchmark)
11:14:13XavierGrand see if you got more measurements
11:14:45XavierGrif you got then I think that a full benchmark will work.
11:16:50RoC_MMdid as you said
11:16:52RoC_MMBattery type: 1500 mAh Buffer Entries: 1000
11:16:52RoC_MM Time:, Seconds:, Level:, Time Left:, Voltage[mV]:, M/DA:, U:
11:16:52RoC_MM00:00:31, 00031, 100%, 10:20, 11410, 0001, -
11:17:25RoC_MMthat's what it says
11:17:26preglowXavierGr: the scaler is dead wrong
11:17:39preglowXavierGr: but that shouldn't matter for a calibration run, i think
11:17:50RoC_MMit still only has one line
11:17:59XavierGrhmm this is what bothers me.
11:18:02RoC_MMshouldn't it have like a whole bunch of lines in the log?
11:18:09RoC_MMI played 5 tracks
11:18:13preglowXavierGr: btw, ata_disk_is_active() always returns false on nano, that'll break your plugin, yes? :P
11:18:20RoC_MM30 seconds or a little more each
11:18:25XavierGrbut on 4G
11:18:30XavierGris it active?
11:18:31preglowRoC_MM: each time your disk spins up, it should add another line
11:18:46preglowXavierGr: how often does it log?
11:19:00RoC_MMso I should pick songs that aren't in that 30meg cache and will make it spinup?
11:19:18XavierGrevery disk activity that occurs. Also it logs when the voltage level is changed
11:19:38B4gderRoC_MM: run a big playlist and let it play non-stop
11:19:56RoC_MMand I'll definately get more than one line right?
11:19:57preglowXavierGr: but i don't know how to do it, the nano flash _is_ always sleeping
11:20:11preglowit goes into sleep mode just a couple of msecs after the last access
11:20:12XavierGrhow is the ondio implementation?
11:20:29preglowi'll probably just have to fake that it's spinning
11:20:35Bg3rXavierGr in fact it'll log on every disk spinup *only* if there is anything for logging ...
11:20:47XavierGrBger: Yes, my bad
11:20:50preglowthat also always returns false
11:20:54XavierGrI mean that's what I meant
11:21:16preglowso battery bench probably doesn't work too good on ondio either
11:21:16XavierGrpreglow: amiconn told me that on ondio the plugin works well.
11:21:18RoC_MMIs the log needed to when I finish the benchmark?
11:21:37preglowXavierGr: depends on if he used an mmc or not
11:21:40preglowmmcs will work
11:21:49preglowthey don't always return false
11:21:50RoC_MMWhat should I do with the log?
11:21:52preglowfor the internal flash, it does
11:22:05XavierGrpost it in the wiki and tell us where to find it
11:23:26XavierGrpreglow any idea on this?
11:24:00preglowand i've gotta go for a little while
11:24:23XavierGrso you say that nano never returns true on reading/writing?
11:24:27XavierGrah okay later then
11:24:33RoC_MMwill do the benchmark later
11:24:51 Quit RoC_MM ("Leaving")
11:30:35XavierGrIn fact I think it is time to make an H300 benchmark with my 1900mAh battery.
11:30:55XavierGrRockbox managed to give H140 25 hours of playback with the same battery
11:33:54webminddoes anyone know the battery spec of the nano ?
11:34:58 Join bluebrother^ [0] (
11:35:05linuxstbpreglow: Yes, the 4G Color/Photo came in 60GB models - I have one.
11:35:39dwihno5g is video?
11:36:10 Join Moos [0] (
11:36:51dwihnoare they easily scratched?
11:37:07dwihno(i.e. prone to scratches)
11:38:03linuxstbProbably easier than other DAPs (e.g. the irivers), but my 5g is still looking OK after a few months use.
11:38:31linuxstbBut I think most people just keep them inside skins.
11:39:31Slasherilinuxstb: hehe, now i got tracknumbering to work :)
11:39:58Slasheristill have to deal somehow with the case where now tracknumber tags are available
11:39:58dwihnoThen I have to be extra careful once I get one. (when/if)
11:40:10linuxstbNice - that's the most common question about tagache that I've read so far.
11:40:10dwihnoMy first project will be a ffmpeg port ;D
11:40:32linuxstbThe whole of ffmpeg? :)
11:41:23dwihnoWhy not? :)
11:41:42dwihnoNah, to be honest, I don't think I'm good enough to pull it off.
11:41:44linuxstbSlasheri: Can you default to alphabetical order if there are no tags.
11:42:22 Join Jungti1234 [0] (n=jungti12@
11:42:39linuxstbdwihno: ffmpeg is not suited for embedded use - it's far too big and uses far too many mallocs and floating point operations. But porting individual codecs is possible (our FLAC and Shorten decoders are from ffmpeg).
11:43:05 Join ashridah [0] (
11:43:24dwihnolinuxstb: well, if anyone wants to initiate a video player, the gpl player "tcpmp" might be of use
11:44:21dwihnoShould be pretty non-floatized
11:44:31Slasherilinuxstb: hmm, yep. Probably we could generate the tracknumber tags somehow while generating the database
11:44:31linuxstbMaybe, but Windows CE and PalmOS devices are normally much more powerful than Rockbox targets. It wouldn't surprise me if it's C++ as well.
11:44:33 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
11:45:04Slasherilinuxstb: do you think i should add also composer, length, bitrate etc. tags to the tagcache? I have already added the year
11:45:19XavierGrcomposer for sure!
11:45:35XavierGrwhat will classical junkies like me will do without it.
11:45:43linuxstbYes, I think so. It may be worth canvassing views in general about which tags should be standard in Rockbox.
11:46:05Slasheriok, i will do that. But i will leave the UI design for some other guy then ;)
11:47:39linuxstbPersonally, I would also like a tag for "type" - e.g. studio album, official live recording, bootleg live recording, demo, TV/Radio broadcast... But maybe that's a bit specialised...
11:48:01Spidadoes USB-charging work for iriver h3x0 now?
11:48:46SpidaI only have the USB-cable, and battery voltage is to low to start the original firmware
11:48:51Spidaor any firmware.
11:49:09Slasheriwell, we have 10 tags now :)
11:49:20Spidausb the bootlaoder-usb mode works (which I don't understand. doesn't it check for battery-voltage?)
11:49:53 Join muesli__ [0] (n=muesli_t@
11:53:21linuxstbSlasheri: Will Track Rating be something external to tagcache?
11:54:11Slasherilinuxstb: i am not yet sure what is the best way to implement it
11:54:14LinusNSpida: it just displays the voltage, nothing else
11:54:47Slasheriit could be easily integrated with the tagcache (as separate numeric tag), but somehow it must be also exportable outside the tagcache
11:55:15peturyou don't want it lost when the cache is rebuild...
11:55:54Slasherilinuxstb: probably i will integrate it with tagcache, as i need to create a special changelog file for all tags in future when tags can be changed inside rockbox
11:55:56SpidaLinusN: so, how to I get it charged without the original charger?
11:56:08LinusNuse the original firmware
11:56:15Slasherithat way regenerating the tagcache wouldn't be a problem
11:56:57SpidaLinusN: I can't use the original firmware. It won't start, claiming voltage is to low
11:57:26LinusNSpida: then you must have turned off usb charging :-(
11:57:54Spidayeah. sometime I wan't to use the ueb-port.
11:58:35LinusNSpida: you can use it without turning off usb charging, can't you?
12:00:30Spidahm, not with the original firmware.
12:02:43tucozbluebrother^, around?
12:04:15MoosLinusN: already spoted and fixed the audio problem? :)
12:04:32bluebrother^tucoz: yes
12:04:51tucozbluebrother^, what do you think of ditching the use of floats for tables and figures?
12:05:35bluebrother^I was already thinking about changing all figures as the [!h] produces warnings.
12:05:51tucozchange to what?
12:06:23bluebrother^as a first step changing to [!ht] as latex uses this automatically.
12:06:53bluebrother^another idea would be using floating figures so the screenshots could placed aside of the text.
12:07:06bluebrother^I think this would be nice especially for the plugins.
12:08:04tucozlike having the text wrap around figures?
12:08:20tucozyes, I've also thought about that.
12:08:39tucozbut, some of the tables and figures should (imo) be placed where in the text they appear
12:08:40bluebrother^but I need to look more close to this, I only tried the h1xx manual for now.
12:09:32 Join imphasing [0] (
12:09:34bluebrother^don't know if this works good for targets with bigger displays
12:10:08tucozwell, we could (like we do now) have the images of the same size for all targets
12:10:29tucozjust define a width
12:10:32bluebrother^yeah, this would be good.
12:10:50Mikachulostlogic: i can't scroll in View ID3 Info without letting go of the wheel after every step, that must be your fault :)
12:10:53tucozall screenshots are now of a 4 cm width
12:11:13XavierGrtucoz: how is the manual progress going? Will it catch the 3.0 release?
12:11:30bluebrother^I think we should concentrate on the style guidelines.
12:11:32tucozXavierGr, I would say it's progressing nicely
12:11:57bluebrother^aside with this we could introduce a command for our pictures and use that.
12:12:13bluebrother^so we could set the size globally.
12:12:26tucozbluebrother^, I think I could write a mail to the dev-ml and ask the manual writers to follow the style guidelines
12:12:46bluebrother^and later surround it with a floating environment if we decide to do so.
12:12:56bluebrother^good point
12:12:59tucozsounds good
12:13:08bluebrother^but we need to work the guidelines out more.
12:13:16tucozyes, of course
12:13:19 Quit Seed (Nick collision from services.)
12:13:21bluebrother^I just moved some stufff and added my ideas ;-)
12:13:26 Join Seed [0] (
12:13:45 Quit petur ("CGI:IRC 0.5.7 (2005/06/19)")
12:13:56bluebrother^I was playing around with the preamble yesterday also.
12:14:10tucozWhat are the issues right now? Table headers, floats/non-floats
12:14:26tucozwhat did you find out then?
12:15:17bluebrother^I was playing with the headers. I think it would look great to have a heading saying "Rockbox Manual for(targetname)"
12:15:35 Join webguest78 [0] (
12:15:47bluebrother^do you think we should stick to a one-sided layout?
12:16:11tucozHmm, not sure. You seem to be the Latex guru around here :-)
12:16:16bluebrother^then I'll try to adjust the headers and submit it as patch later today.
12:16:30bluebrother^really? I haven't done latex for quite some time ;-)
12:16:54 Join ScoTTie [0] (i=skotty@unaffiliated/scottie)
12:17:15tucozwell, it seems you know what you're talking about at least. I am glad for that.
12:17:15bluebrother^tucoz: sorry, I need to leave for some time. cul
12:17:29bluebrother^(1h or so I guess)
12:17:36tucozok ,see you then
12:20:37Mikachulostlogic: probably the problem is near line 1100 in screens.c and a bit down
12:27:22 Quit quobl (Remote closed the connection)
12:27:32 Quit Bg3r (Read error: 110 (Connection timed out))
12:29:01 Join Bg3r [0] (n=Bager@rockbox/developer/Bger)
12:32:02JdGordonok, did some1 change something? pu need to press play twice for it to pause now
12:32:22LinusNwas the lcd off?
12:32:40LinusNthere you go
12:32:57JdGordon.. u finnaly add that and im used to the other way :p
12:33:05LinusNchange the setting then
12:33:17Spidadoesn't rockbox check the battery-voltage in bootloader-usb-mode?
12:33:25Bg3rJdGordon :)
12:33:38LinusNSpida: i told you
12:33:56LinusNSpida: it just displays the voltage, nothing else
12:38:14*B4gder pulls down the curtains
12:38:19B4gderdarned light!
12:39:18safetydanFear the daystar.
12:39:33preglowexceptionally nice weather today
12:39:51preglowabout time the damn snow started melting
12:39:56safetydanif you call windy and overcast nice weather... oh wait, not all the world is London
12:40:23preglowthanked be the gods, no
12:43:20XavierGrI dislike much light too!
12:44:14preglowtons of light and a blue sky
12:44:29 Quit Bg3r ("bbl")
12:44:55DJ_Dooms_DayHey does anyone know how to access the search feature on the H3xx series rockbox?
12:44:56preglowmakes me want to shirk work for an hour's time and go find some nice spot in the sun to have a beer it even IN the H3xx build yet?
12:47:24B4gderwhat search feature?
12:48:01tucozXavierGr, where in greece do you live?
12:48:10DJ_Dooms_Daywhen i had my H1xx i could bring up a search function and type in a few letters and it'd search the DB file and produce a list
12:48:21XavierGrpreglow I've got Sun and Blue Sky for granted so that's why i am not very fond of it.
12:48:39XavierGrtucoz: City of Heraklion in Crete
12:49:15tucozok. Lots of tourists there, eh?
12:49:21XavierGryes indeed
12:49:25preglowhow can you stop loving it?
12:49:38tucozI would almost assume you speak scandinavian then
12:49:47preglowi've seen my fair share of it, but it never fails to make me happier than i was
12:49:49DJ_Dooms_DayANyone know???
12:49:50XavierGrIt's like if you are eating the same food every day.
12:49:56B4gder"Apple, the company has released an update to the iPod nano and 5th generation iPod that allows the user to set the maximum volume level"
12:50:03preglowhaha, sure
12:50:14preglowbut yeah, having had a hard winter helps
12:50:17DJ_Dooms_Dayok nevermind, found it
12:50:22preglowi'm beginning to grow tired of snow now
12:50:42XavierGrBut I am considered "strange" for my habits. Most of the people here love Sun.
12:50:48DJ_Dooms_DayB4gder - Which means what? YOu couldn't set max vol level before?
12:50:52tucozXavierGr, that is the same here. It's like eating the same food every day by living in Bergen. But exchange the sun and blue sky with rain and gray sky.
12:51:13linuxstbDJ_Dooms_Day: I believe it's a user-defined cap on the volume.
12:51:27preglowtucoz: trondheim was just the same...
12:51:35*preglow hugs the southern parts of norway
12:51:49XavierGrtucoz: I know what you mean. But sometimes due to much sun I like cloudy days.
12:52:03tucozyou are lucky :)
12:52:05B4gderthere it is =>
12:52:22linuxstbAnd I saw some requested that feature for Rockbox this morning as well....
12:52:27DJ_Dooms_DayLinuxstb - Obviously, but why?
12:52:32B4gderI noticed that too
12:52:36Mikachuis there any particular reason the statusbar is hidden while inserting tracks?
12:52:52linuxstbDJ_Dooms_Day: I haven't a clue...
12:52:59B4gderDJ_Dooms_Day: there's a lawsuit against apple going on
12:53:04preglowi like the fact that disk mode hangs when it's finished charging
12:53:10B4gderclaiming hearing loss
12:53:19B4gderwhich could be one reason
12:53:27DJ_Dooms_DayAre you serious?? THats fucking stupid
12:53:41preglowDJ_Dooms_Day: oh, you mean just like people?
12:54:04B4gderthey like sueing over there
12:54:10DJ_Dooms_DayI would say predominantly Americans preglow, but yeah, sure ;)
12:54:28preglowoh, just can find stupid people everywhere
12:54:41preglowyou can find ones that aren't able to write a coherent sentence too
12:55:08B4gderthat's the worst kind!
12:55:19preglowyeah, i hate those bastards
12:55:23preglowscum of the earth
12:55:24DJ_Dooms_DayANyway, so people are actually requesting that you put THAT feature on rockbox?
12:55:38B4gderone user did at least
12:55:43preglowgood, plain old stupidity
12:55:46DJ_Dooms_DayBan him
12:55:49preglow<−−- rockbox political spokesman
12:56:12B4gderI bet we can reach 1 million feature requests within a few years ;-)
12:56:28preglowthat section will be closed long before then
12:57:11preglowLinusN: so it works fine now?
12:57:52*webguest78 wonders why this is still open
12:58:03 Join petur [0] (
12:58:11webguest78Seems like a case of rtfm
12:58:14 Join MarcoPolo [0] (
12:58:24LinusNpreglow: sounds really good
12:58:53preglowB4gder: how much for shipping a shirt to norway, you think?
12:58:57LinusNbut the boost ratio is still too high
12:59:05preglowLinusN: isn't it always?
12:59:26LinusN192kbit/s -> 10% on h300 and 30% on x5
12:59:41LinusNbeats me
12:59:45linuxstbLCD driver?
12:59:59preglowwell, not too many candidates left, then
13:00:31LinusNwell, i have to try less lcd updates, but i doubt that it would make a diff
13:01:41LinusNlast time i checked, we had numerous threads that didn't sleep properly, but it should be the same on h300...
13:01:58LinusNoops, i forgot to eat lunch
13:02:10LinusNcu l8r
13:06:28B4gderpreglow: I think it is the same "within europe" cost
13:07:49*preglow boos at the swedish postal service
13:08:36*petur hopes he was quick enough to reply ;)
13:11:49peturI wonder how long it will take before somebody files a feature request for doom on x5
13:12:29webguest78Isn't it already built?
13:12:51webguest78Not that built==works, but
13:13:01***Saving seen data "./dancer.seen"
13:14:12 Join Bg3r [0] (n=Bager@rockbox/developer/Bger)
13:14:41 Quit safetydan ("Leaving")
13:16:58linuxstbHas anyone tried adding button mappings for Doom and building it for the X5?
13:17:12 Quit petur ("CGI:IRC (Ping timeout)")
13:17:48webguest782006 28 Mar 17:22Christianapps/plugins/Makefile 1.74 x5: lets build doom...
13:17:50 Join b00st4 [0] (
13:18:38b00st4hm, i tried to change the bootscreen yesterday, compile worked fine, but it doesnt switch on my ipod...
13:19:04b00st4rockboxlogo.320x98x16.bmp replaced this 24-bmp for my ipod 5g
13:19:22muesli__any gmini 400 owner here?
13:21:24b00st4hmm, any idea?
13:21:47 Join quobl [0] (n=quobl@tor/session/x-d4eef964a82b9353)
13:24:56b00st4rockboxlogo.h do i have to edit this?
13:25:31linuxstbHave you tried "make clean", and then rebuilding Rockbox?
13:25:42linuxstbNo, you shouldn't have to edit that file.
13:26:06b00st4how ya mean make clean?
13:26:17b00st4del all on my ipod?
13:26:43linuxstbNo, just type "make clean" in your build directory, followed by "make", followed by "make zip"
13:27:01b00st4ah, ok
13:27:09b00st4but dir was empty before
13:27:25b00st4compiled totally new
13:27:29linuxstbIn which case, that's not going to be the problem...
13:28:10b00st4what im wondering is, that the file it shows doesnt even exist any more
13:28:16DJ_Dooms_Dayhey anyone know where the database builder is, off the top of their head
13:28:24b00st4cuz i replaced it
13:28:56 Join petur [0] (
13:29:01tucozgeneral settings -> playback
13:29:02LinusNDJ_Dooms_Day: yes, it's integrated in rockbox nowadays
13:29:14DJ_Dooms_DayAh, nifty
13:29:28b00st4anyways, im at lunch, try to fix that later on....
13:29:46linuxstbb00st4: Did you definitely copy the new rockbox.ipod correctly? If you go into the main menu, and then Info -> Version, does it display the data/time of your build at the bottom?
13:30:49DJ_Dooms_DayHey whats a 60/80GB HDD go for these days? I mean, one of the one small enough to fit in a iriver? (1.5 inch yes?)
13:32:24petur80GB have been announced but haven't seen 'm yet
13:33:13tucozwonder if the new (probably expensive as h**l) flash based hard disks would work in an iriver.
13:34:00b00st4<linuxstb> yap it does
13:34:07b00st4other patches applied
13:34:14*b00st4 restart
13:34:20 Quit b00st4 ()
13:34:37 Join b00st4 [0] (
13:34:48DJ_Dooms_DayWhy would you want flash based HDDs?
13:35:10tucozSilent operation for instance
13:35:33peturshocks won't harm it
13:35:38tucozthat also
13:35:42DJ_Dooms_DayAre you serious? You can't even hear HDDs
13:36:00tucozThe iriver spin-up sound is getting on my nerves
13:36:04Mikachutucoz: are you sure you don't mean a solid state disc?
13:36:23tucozMikachu, don't know. The new ones from Samsung
13:36:23 Quit Jungti1234 ()
13:36:36tucozprobably called solid state. I thought they were flash based though
13:37:08tucozI think the largest atm is a 32GB
13:37:11DJ_Dooms_DayTucoz? wtf do you do with your iriver? With mine, i listen to music. Generally, if your listening to music, you can't here the sound
13:37:47tucozhehe, I sometimes start it. And when I don't listen to music, but browse around it spins up.
13:37:48Mikachuram with a battery is a lot faster than flash roms...
13:38:04Mikachuand i think those are called solid state maybe
13:38:10tucozI see.
13:38:50tucozMikachu, the solid state is based on NAND flash
13:39:14DJ_Dooms_DayWow, i just munched through an entire battery in a few hours.
13:40:12tucozIt is faster than mechanical drives they say. 3 times the read, and 1.5 times the write. Less power consumption. I think it sounds sweet.
13:40:41 Join Shagnar [0] (
13:41:06Shagnarhello everybody!
13:41:53Shagnarrecently i got a bug, i can't browse the files anymore because there are no more files shown (just an empty screen when entering the file browse mode) rockbox (newest version) on H-140
13:42:00DJ_Dooms_DayThat one uses IDE. 7500 RPM drives can already read at around 100MBs, and with IDE cables that can only transfer at around 130MBs (or was it 150?), there isn't much room for improvement.
13:42:08Shagnardoes anybody have an idea?
13:42:17LinusNShagnar: what is your "Show files" setting?
13:42:44DJ_Dooms_DayShagnar - Hold down the AB button and change the 'show files' to all
13:43:23Shagnardj, did that already
13:43:43tucozMaybe you are in an empty directory?
13:43:47Shagnaruhm, very strange. i pressed the "play" button in this empy screen, and the iriver got frozen
13:43:57Shagnarafter resetting the files are shown again.
13:44:09Shagnartucoz: no, just nothing was shown anymore
13:44:15DJ_Dooms_DayI think i know what your issue was
13:44:29DJ_Dooms_DayIts to do with that ID3 tag database
13:44:44DJ_Dooms_DayI used to get that problem too if the database was out of date
13:44:51 Join Coldtoast [0] (
13:45:01Coldtoasthowdy all
13:45:11Shagnarah okay
13:45:17Shagnarthx to all^^
13:45:18Coldtoasthey. is there any likelihood at all that Rockbox might ever be ported to the PSP?
13:45:31DJ_Dooms_DayI doubt it
13:45:59Coldtoastall the homebrew media players as well as the Sony MP3 support are horrible
13:48:12B4gderit "just" takes a dedicated skilled person or team
13:48:45muesli__is the archos gmini 120 supported?
13:48:57muesli__oh :(
13:49:05B4gderwe don't have gmini support
13:49:28muesli__that model is quite old though..but useful as a data carriere
13:53:12linuxstbColdtoast: Is SDL ported to the PSP?
13:56:42Lynx_How do I change the font colour in rockbox?
13:57:26DJ_Dooms_DayGeneral settings>display
13:57:26JdGordonsettings > display >lcd >
13:57:30MarcoPoloI have the following problem compiling cvs : make[3]: *** No rule to make target `i_net.h', needed by `/home/marc/Install/rockbox/rockbox/build/apps/plugins/doom/d_net.o'. Stop.
13:57:31DJ_Dooms_Daythen foreground
13:57:46MarcoPolobut I can't find i_net.h in the files
13:57:48Coldtoastlinuxstb: not that I know of
13:58:01B4gderMarcoPolo: how did you get the sources?
13:58:02Coldtoastbut that doesn't mean much. heh
13:58:07MarcoPoloB4gder: cvs up
13:58:19 Join Kohlrabi [0] (
13:58:54Lynx_DJ_Dooms_Day: ah, thx
13:59:05webguest78linuxstb: casual googling suggests it may be
13:59:38B4gderI have no references to i_net in my doom sources
14:01:09MarcoPoloB4gder: so do I
14:01:19MarcoPoloB4gder: only at compilation I get this message
14:01:31MarcoPoloB4gder: well, make clean was the solution
14:01:38MarcoPoloB4gder: sorry for disturbing you
14:01:48B4gderwell, your makefile must be referencing it
14:01:55B4gderotherwise it wouldn't say that
14:02:05MarcoPolono, it said .o
14:02:44B4gderthat's what the makefile uses
14:02:57B4gderas it checks if the .c is newer than the .o
14:03:05 Join mikearthur [0] (
14:03:15webguest78linuxstb: yeah, there definately is an SDL port for psp.. can't find it though, only references to it
14:07:05bluebrother^tucoz: I'm back
14:10:51tucozbluebrother^, now I've got to go :)
14:11:11bluebrother^how nice ;-) cu later.
14:11:31tucozSee you later. I'll read the log later on. case you come up with something nifty
14:11:53 Part tucoz ("Leaving")
14:11:58MarcoPoloWhen tagcache says "updating in background", my ipod is frozen
14:19:45 Join _FireFly_ [0] (
14:19:47 Quit mikearthur ("Konversation terminated!")
14:20:31 Join mikearthur [0] (
14:21:41 Quit JdGordon (Read error: 104 (Connection reset by peer))
14:26:17linuxstbwebguest78: In which case, the Rockbox sim (once audio playback becomes stable) could be ported to the PSP.
14:30:43webguest78Creating a sim-only psp target?
14:30:55linuxstbMarcoPolo: Do you have any other patches installed? e.g. an old version of the album art patch?
14:35:11 Join voodoo_laptop [0] (
14:36:24voodoo_laptopanyone know why my LCD remote for my iriver H320 would stop working under rockbox? the remote is still getting power and can still control the remote but the remote screen has no text on it, the remote has worked in older version of rockbox
14:42:42MarcoPololinuxstb: indeed, i updated aa patch in now it works well :)
14:43:03MarcoPoloshould doom be working on ipod currently ?
14:43:42B4gderit should, but it doesn't
14:44:09MarcoPoloB4gder: I get some blank screen when I run it
14:44:26B4gderyes, it doesn't work on ipod
14:44:41MarcoPoloB4gder: ok
14:45:42linuxstbMarcoPolo: If you fix it, I'll send you a beer of your choice.
14:45:57 Join aliask [0] (
14:46:06MarcoPololinuxstb: i really have no idea of what it could be, sorry ;)
14:47:48 Nick Sin|showah is now known as Sinbios (
14:52:20*amiconn is back
14:52:22amiconnpreglow: battery_bench works perfectly on Ondio
14:53:06preglowamiconn: ondio doesn't use ata_flash?
14:53:20preglowahh, no, that's probably iriver
14:54:19 Join Jungti1234 [0] (n=jungti12@
14:54:21amiconnOndio uses ata_mmc
14:54:45preglowamiconn: right, i forgot the internal flash is really just mmc too
14:54:55Jungti1234I read some document.
14:55:18 Join SereR0KR [0] (
14:55:39linuxstbwebguest78: Sorry, I missed your question. Yes - the idea would be to create sim-only target.
14:55:39amiconnI also got a remote for my mini...
14:56:07Jungti1234bdf file does that unuse much in embedded system.
14:56:42linuxstbamiconn: So you're the person to reverse-engineer a serial port driver and the apple accessory protocol?
14:56:57 Quit aliask ("sleep")
14:57:18amiconnlinuxstb: Depends. Remote is low-prio for me
14:57:31webguest78But it'll be FUN
14:57:36 Quit B4gder ("time to say moo")
14:58:03Lynx_what does 'committing tagcache' at boot actually mean?
14:58:04amiconn(or should I say low-low prio?)
14:58:40amiconnI use my archos remote like 1..2 days per year, and never actually used my iriver remote except for development
14:58:49Jungti1234bdf file is big size, and the loading speed is slow
14:59:15webguest78Jungti1234: bdf files are not used on rockbox. They're converted to a special rockbox format
14:59:35webguest78the originals are just bdfs
14:59:46Jungti1234:) ah..
15:00:24amiconnBtw, is the remote protocol supposed to be serial?
15:01:07amiconnThe remote connector is a 4-pin conector next to the earphone jack, and I thought the serial port is connected to the dock connector at the bottom?
15:05:02linuxstbI think the serial port is connected to both the remote connector (which isn't present on the Nano and the 5g) and the dock. Or maybe there are two serial ports...
15:05:57linuxstbLike you, I'm not really interested in the accessories.
15:13:02***Saving seen data "./dancer.seen"
15:18:23 Join Febs [0] (
15:18:47 Quit Jungti1234 ()
15:22:58MarcoPolowhere can i see the new tags added to tagcache engine ? the browser doesn't show anything more...
15:23:15MarcoPoloand could it sort files by disk number too (not only by track number ?)
15:26:22amiconnNow, the interesting procedure of opening this thing...
15:32:20preglowgood luck
15:35:55 Join Toshik [0] (
15:39:52 Quit Coldtoast ()
15:40:53*webguest78 gets a segfault in convbdf with ridicolously large fonts
15:41:18webguest78Like this one:*checkout*/xorg/xc/fonts/bdf/100dpi/UTBI__24.bdf
15:44:51webguest78Might be related to the fact that it's italics, perhaps
15:45:29 Quit Toshik ("CGI:IRC (Ping timeout)")
15:45:34 Join Jungti1234 [0] (n=jungti12@
15:46:32webguest78Hm, no..*checkout*/xorg/xc/fonts/bdf/100dpi/UTB___24.bdf does it as well
15:47:05 Part XavierGr
15:47:39Jungti1234What is 'tag cache'?
15:48:23webguest78A replacemtn to "id3 database"
15:49:04Jungti1234What's that?
15:49:26webguest78Metadata browser, so you can browse all your files by artist, instead of the file structure
15:49:26bobTHCto avoid file reading to display, search,ID tags iirc
15:49:50Bg3rand not only by artist
15:50:08webguest78Yeah, artist, album, track etc. just an example
15:52:44Jungti1234Where can I saw it?
15:54:20amiconnHmm. It's almost open (major part is removing the top & bottom plastic covers, which is done), but now I have the problem that I can't find a small enough philips screwdriver...
15:54:46preglowif you did find one, it'd probably break anyway
15:54:53preglowif it's anywhere near as hard to dislodge as the nano
15:55:26amiconnThere are 2 tiny philips screws. One is already off, but the other is somewhat more tight.
15:55:35 Quit ashridah ("Leaving")
15:56:04amiconnUsing an 1.5mm flathead screwdriver worked for one of them, but not for the other
15:56:35preglowthere acrews?
15:56:42preglowthere are screws? <-
15:57:01amiconnYes, after removing the 2 plastic covers.
15:57:39preglowahh, of course, the mini is built very differently
15:58:20webguest78I've submitted a bug for the convbdf thing with a little detail:
15:58:56 Quit Kohlrabi ("Fast alle Menschen sind Regenwürmer")
16:00:08Papricapreglow? can you help me again with the HDD error?
16:01:05preglowPaprica: not much i can do
16:01:10 Quit bluebrother^ ("changing clients ...")
16:01:20preglowPaprica: if disconnect the hard drive, see if the connector looks fine and is more or less clean
16:01:34 Join bluebrother^ [0] (
16:01:35preglowPaprica: then stuff the harddisk firmly onto the connector again, if it doesn't work after that, something else is broken
16:02:13Papricapreglow, the hardisk seem to work, but the sound from it is strange
16:04:11preglowdropped it or anything?
16:06:19Papricathe player?, no..
16:06:20 Quit voodoo_laptop (Read error: 104 (Connection reset by peer))
16:06:22 Join voodoo_laptop [0] (
16:08:42preglowthen i have no idea
16:09:17Bg3rPaprica how does it sound ?
16:09:32LinusNlike a bag of shit?
16:09:40LinusNthen it's normal :-)
16:09:52Papricaits not regular sound
16:10:20 Join Nico_P [0] (
16:10:57amiconnHmm seems I have to buy such tiny philips screwdriver somewhere
16:11:21amiconnI need to find out what WM... codec is used in the mini 2g
16:11:30crashdamiconn: ipodlinux
16:11:34crashdhas it on there specs page iirc
16:11:35amiconnThe ipodlinux website doesn't tell
16:11:40crashdyou suer ?
16:11:46crashd!@"!£:!"@ argh. keyboard retardation
16:12:31 Part LinusN
16:12:32crashdamiconn: id assume it's go tthe same one as the ipod 1g, as they dont specify a difference
16:12:52 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
16:15:58bobTHCsomeone can test this font(bdf) to be sure that have been well generated ?
16:17:14bobTHCsorry no dev env here
16:18:50Jungti1234what is 'Precut'?
16:19:09webguest78bobTHC: convbdf has no problems with it
16:19:18safetydanJungti1234, it allows you to reduce the volume of the audio before the equalizer settings are applied. It's to avoid distortion.
16:20:02Jungti1234Volume regular control?
16:20:05MarcoPolosafetydan: what is the tagcache tracknumfix for ?
16:20:16bobTHCthx webguest78
16:20:38bobTHCand the .fnt iss usuable ?
16:20:44Jungti1234Does it take away noise of loud sound?
16:20:54webguest78bobTHC: let me check
16:22:06webguest78bobTHC: looks right
16:22:25bobTHCwhat the fnt file size plz ?
16:22:39webguest78260283 bytes
16:22:49safetydanJungti1234, no it's not an automatic volume control. It's for when you increase the gain of a filter in the equalizer (for example you increase the 200 Hz band by 6 dB), you will likely also need to decrease the overall volume of the audio (by 6dB) to avoid distortion.
16:22:52webguest78hrm, the rs seem liftet off the ground
16:23:14safetydanMarcoPolo, uhh... I've no idea. Slasheri is the one you want to talk to about that.
16:23:18Jungti1234Translation is not good...
16:23:27safetydanJungti1234, sorry, it's a little hard to explain
16:23:41bobTHCrs, liftet ?
16:23:50Jungti1234Is explanation impossible much more easily?
16:24:03MarcoPoloSlasheri: ping ?
16:24:17safetydanJungti1234, have you used the equalizer? Have you noticed that you get distortion (bad sound) when you increase the gain in a filter band?
16:24:45safetydanMarcoPolo, it sounds like the track listing in an album in the tagcache will be sorted by track number now
16:24:50Jungti1234bad sound....
16:25:03Jungti1234Noise of loud sound?
16:25:13webguest78bobTHC: the r characters seem 1px above the baseline
16:25:14safetydanJungti1234, clipping
16:25:32Jungti1234Cutting with scissor? :(
16:25:37Papricasomeone knows if there's a way to connect regular harddisk to the h300?
16:25:50webguest78Paprica: It's been done.. check MisticRiver
16:26:22 Join midkay [0] (n=midkay@rockbox/developer/midkay)
16:26:33safetydanJungti1234, no, audio clipping is something that happens when you amplify music too much. It causes distortion.
16:26:55Jungti1234I don't understand 'distortion', 'clipping'
16:27:07Papricawebguest78, can you give me a link?
16:27:18webguest78Paprica: not offhand, no.. in the mods section iirc
16:27:19Jungti1234Frequency adjustment?
16:28:39bobTHCwebguest78 > except that that(R), that looks good or crappy ?
16:28:52bobTHCon target
16:28:59safetydanJungti1234, hang on I'm trying to find a picture that might explaing clipping
16:29:02*amiconn will try ipl on his mini next
16:29:13webguest78bobTHC: looks decent.. I wouldn't use it, but it's certainly not bad
16:29:21amiconnToo bad weather today to go out and buy that tiny screwdriver
16:29:23Jungti1234safetydan, Thanks...
16:29:47Papricaoh, i cant do it =\
16:30:30bobTHCoki thx webguest78..
16:30:50 Quit Febs (Read error: 104 (Connection reset by peer))
16:32:07SlasheriMarcoPolo: yep, the track order should be correct now
16:32:17 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
16:32:41linuxstbSlasheri: I was happy to see that you are displaying the track number as well. I was going to request that... It seems to work well.
16:33:40safetydanJungti1234, maybe this will help
16:33:41Slasherilinuxstb: hehe, nice :)
16:34:01Slasherilinuxstb: in that way the sorting was also simple without any specical tricks
16:34:36Jungti1234volume down?
16:34:50safetydanJungti1234, precut essentially turns the volume down yes
16:35:45Jungti1234What abbreviated word is 'Precut'?
16:35:47tucozbluebrother^, are you here?
16:35:55*petur wonders if there exists a tool to restore a clipped waveform by lowering the volume and plotting a bezier through the points
16:36:08Slasherilinuxstb: if tracknum tag is not found, the engine tries to find the track number by searching for the filename while generating the database
16:36:12amiconnDoes ipl play plainfiles?
16:36:59amiconnBah, teh wheel is too sensitive in ipl
16:37:02Jungti1234hm ok, I have go to now..
16:37:04webguest78Jungti1234: It's just "pre-" which means before, and "cut" which just means lower volume here
16:37:07linuxstbamiconn: You mean files not transferred using itunes?
16:37:12safetydanJungti1234, precut is something like "reduce volume by this before using equalizer"
16:37:25linuxstbI think so, yes. But I never managed to get audio working...
16:37:36Jungti1234hmm.. precut.....
16:37:47linuxstbamiconn: Have you installed "podzilla" or "podzilla2" ?
16:38:06amiconnHmm, I don't get disk access from ipl
16:38:29tucozamiconn, you got your ipod?
16:38:30amiconnWhen I plug USB, it asks me whether I want to go to disk mode. But it just reboots...
16:39:06Jungti1234safetydan, sorry and thank you
16:39:16 Quit Jungti1234 ()
16:41:13 Quit webguest78 ("CGI:IRC (EOF)")
16:41:28 Quit MarcoPolo (Read error: 104 (Connection reset by peer))
16:41:51amiconnBah, now it doesn't do anything except showing the apple logo :(
16:42:09linuxstbamiconn: That will be a bug.. Diskmode is reached by writing a cookie to the end of IRAM and rebooting. The flash-based Apple bootloader will check for that cookie and enter disk mode. I'm guessing that podzilla doesn't know that your IRAM is 128KB, so it's writing the cookie in the wrong place.
16:42:58 Join Spida_ [0] (
16:44:08amiconnIs there a hardware trick to enter disk mode? I tried the reset (holding meu+select), but that does only restart with the apple logo...
16:45:24 Join Febs [0] (
16:45:57linuxstbYes, hold SELECT+PLAY immediately after it reboots.
16:46:35amiconnAh, thx
16:47:08linuxstbSlasheri: What do you mean by "searching for the filename" ? How does it determine the track number?
16:49:32amiconnFrom the one time that ipl ran, I am now sure it's a 2G
16:49:41Papricaaff my harddisk dieeeeee
16:49:45amiconnCouldn't try playback because ipl now refuses to boot
16:49:55Papricaonly 5 monthes!!
16:50:12amiconn...and the ipl loader doesn't manage to boot apple fw either
16:50:39peturPaprica: be quick to reflash it if possible...
16:50:47Papricanot possible
16:51:02Papricait's not recognize the harddrive
16:51:07 Quit quobl (Remote closed the connection)
16:51:27Papricaim very nervous :(
16:51:29Slasherilinuxstb: currently it tries to find the first digit from the filename, and if found, it also checks the next character for another digit
16:51:53Slasheriand then converts that string into a track number
16:52:03peturotoh, now you can install that 80GB disk....
16:52:39Papricawhich one?
16:53:31peturbloody long url
16:54:16 Quit Spida (Connection timed out)
16:54:41Mikachuif you have the doom patch applied and then update cvs and don't remove the extra vsnprintf in plugin.c, bad things happen
16:55:31petur300,000 MTTF for your disk, I guess they misread the , as .
16:56:12bluebrother^tucoz: you're in?
16:56:33preglowpetur: algorithms for restoring clipped audio exists
16:56:43preglowpetur: mostly based on the atan function, i believe
16:56:50preglowpetur: you never get a perfect reconstruction anyway
16:56:53 Nick Spida_ is now known as Spida (
16:57:06peturbetter than that horrible noise...
16:57:10bluebrother^I'm just at this moment preparing a new patch for the tracker.
16:57:50bluebrother^for screenshots.
16:58:11bluebrother^I also tried making a new environment "buttonmap" for all those button map tables.
16:58:28bluebrother^but this has still some problems I need to look after.
16:58:49tucozsounds goood. Does this patch use float?
16:59:00tucozfigure-environment that is
16:59:24Papricapetur, we dont have it here, in israel
16:59:48Papricaand i'm gusseing that it's very expencivw
17:00:04tucozI am not really sure we want floats. Imo, the screenshots should be placed exactly where we want them
17:00:31bluebrother^hmm. Using the [!ht] should be sufficient imo.
17:00:32peturPaprica: I'm not surprised, it's probably still a bit vaporware atm ;) but 60GB should be nice too
17:00:48bluebrother^but we could also remove the figure environment.
17:01:20bluebrother^with the new command it is only one place to change it ;-)
17:01:32tucozlet's try it out
17:01:46bluebrother^... which was the main reason I wanted a command for screenshots.
17:01:50Spidadoes anybody know an algorithm to find differences/similarity between different audio-signals?
17:02:22bluebrother^should I change all other files too?
17:02:28 Quit CoCoLUS (Read error: 110 (Connection timed out))
17:02:40Spidathat is, I have some wave-files that contain various "BEEP"s, and I want to find beeps with the same distribution of frequencies
17:02:52tucozI think floats will be hard to get placed right in all the different manuals, as it might be placed correctly for some targets, but move to the next page on others
17:03:24tucozbluebrother^, let me have a look at the patch first. If I like it, I commit it, and you could go from there.
17:03:27Mikachuthat's why they have numbers and you refer to them by their \label
17:03:56bluebrother^as the screenshot images are pretty small I don't think floats will be a problem.
17:04:11tucozBut they are. Look at the plugin-section for instance
17:04:34tucozmaybe it works better with [ht!], but I would not think so.
17:05:06bluebrother^ah, ok. I see what you mean.
17:05:25bluebrother^haven't noticed that shift before.
17:05:43bluebrother^maybe I'm too used to look after image descriptions than at an actual place.
17:05:44tucozfloats are too unpredictable imo
17:06:07tucozYes, that works for reports but maybe not for manuals
17:06:17Nico_Plinuxstb: now oggs work, but MP3 start playing and after a couple seconds, they skip to the next track in playlist...
17:06:24Nico_P(talking about album art)
17:06:30tucozwe would have to cross reference each image if we used floats
17:06:38tucozI think
17:06:45linuxstbSlasheri: It's common for bootleg albums to have filenames such as gd19790822d1t01.flac - which are in track number order if sorted alphabetically, but the track number is at the end, not the start...
17:06:47bluebrother^I agree with this.
17:06:59amiconnOkay, I've now restored my mini to factory defaults...
17:07:06bluebrother^removing the floats also makes it a lot easier ...
17:07:20Slasherilinuxstb: ah :/ hmm, maybe it would be better to look the track number from the end then..
17:07:26tucozyes, maybe we should remove the table environment as well.
17:07:34bluebrother^we don't need that cryptic code anymore I found on the net.
17:07:37Slasheriusing the alphabetical sorting would be a way more difficult
17:07:48tucozwhat code?
17:07:55bluebrother^I was thinking (and trying) about an environment "buttonmap"
17:08:00tucozah, I see.
17:08:21tucozMaybe just keep the tables, and define a tableheader macro then
17:08:57bluebrother^which is mainly a tabularx which automatically does the table header −− as far as I've seen its always the same.
17:09:02tucozsometimes the tables are key-action, and sometimes button-effect or something similar
17:09:23tucozwe should always use the same wording
17:09:59bluebrother^thats my point. So we can do the header with a new environment which also takes care of the table with etc.
17:10:33tucozyes, so that we can define the table width for in-text tables in the preamble
17:10:59tucozAs I said, I don't think textwidth tables look good in the text
17:11:08tucozbut for reference tables, they do
17:11:26bluebrother^I'd like to be able using it as \begin{buttonmap} first button & button action \\ ... \end{buttonmap}
17:11:32amiconnBtw, ipl has problems with the lcd driver. The rightmost pixels don't scroll and aren't always cleared
17:11:58bluebrother^we need to split between two different types of tables.
17:12:39amiconnlinuxstb: How would I proceed to making a rockbox bootloader?
17:12:41tucozin terms of width?
17:13:00tucozthere could be tables with more than two colums also
17:13:04***Saving seen data "./dancer.seen"
17:13:54tucozBut the width should be the same for all in-text tables I think
17:15:18amiconnNm, there is already a mini 1G/2G configure target...
17:15:53bluebrother^I think all button map tables should look the same, so this would make one table class.
17:16:10bluebrother^and reference tables would be another.
17:16:54tucozyes, I agree
17:17:00bluebrother^then we have to look if there are other tables left. I'm too new to the manual so I don't know atm.
17:17:44tucozThere are other tables, but the button tables are the most common ones
17:17:51bluebrother^So I'd perfer building a button map environment and a reference map environment.
17:18:04tucozsounds good
17:18:13linuxstbamiconn: I think it's mainly the button driver that is different between the two mini generations - so I would expect the existing mini build to work, but the buttons probably won't.
17:18:21bluebrother^all other tables could just used as tabularx environments. I guess they aren't too much.
17:18:51tucozno, and they do not differ as much between the targets as the button tables
17:19:22tucozOther tables could be for instance a bitrate table or something similar
17:20:07bluebrother^ok, sounds good :)
17:20:28tucozshould I simplify the screenshot command? That is, remove the figure environment
17:20:50bluebrother^as it it somewhat problematic I'd say yes.
17:21:17tucozEven in the rockbox interface, the images float too much
17:21:40bluebrother^maybe we'll find a better way later, but I think we should try reaching the 3.0 goal first.
17:21:53bluebrother^(but I want a *nice* manual for 3.0)
17:22:04 Join fiftyfour123 [0] (
17:22:07tucozyes, me too. But, I think we should remove the floats for now
17:22:18bluebrother^I agree with this.
17:22:29fiftyfour123is there any sort of hack or way around using rockbox on my hfs 5h without converting it to fat32?
17:22:31bluebrother^a better solution can wait for post-3.0
17:22:49bluebrother^fiftyfour123: no.
17:22:50tucozyes, then we should go for a really pro-looking manual
17:22:54FebsThere are some instances where, IMO, options for particular sub-menus would look better in tables than in lists.
17:23:08fiftyfour123i cant make a small fat32 partition on my ipod and put the rockbox files there?
17:23:15fiftyfour123and use the ipodlinux loader
17:23:26tucozMe too, the configure rockbox section is too itemized in my opinion
17:23:44amiconnfiftyfour123: Perhaps you could, but then you couldnt access your music
17:23:45bluebrother^fiftyfour123: you could, but rockbox can't read hfs.
17:24:02FebsSplitting it into two chapters will help somewhat, but tables will be more readable.
17:24:05fiftyfour123will it ever be compatible with hfs?
17:24:29safetydanfiftyfour123, unlikely as it's a huge amount of work
17:24:56tucozFebs, yes. I agree
17:25:29 Join scf [0] (
17:26:19tucozhmm, I wonder if we could fake a caption in some way.
17:26:34bluebrother^how do you mean this?
17:26:48bluebrother^I don't think really long tables look good.
17:27:11tucozWhen we work with a figure environment, the caption is placed underneath the image. But, that doesn't work without the figure env.
17:27:20 Quit fiftyfour123 ("Chatzilla 0.9.71 [Firefox]")
17:27:25 Quit prh (Read error: 104 (Connection reset by peer))
17:27:28tucozthinking about the config file reference?
17:27:32bluebrother^a caption for the screenshots?
17:27:56bluebrother^no, I meant the Configuring Rockbox.
17:28:12bluebrother^having only a huge table will look ... strange.
17:28:20tucozyes, that will look strange
17:28:29bluebrother^and the left column will eat up much space
17:28:32tucozbut I don't think that was what Febs had in mind
17:29:24tucozI think smaller parts, e.g the different values available for crossfade could fit in a table
17:29:33bluebrother^what I like about that section is the fact you can read it pretty straightforward.
17:29:36amiconnmidkay: You didn't make wormlet compatible to the mini lcd...
17:29:47 Join prh_ [0] (n=paul@
17:29:49bluebrother^ok, I agree with values.
17:30:17tucozSome places there are three levels of itemize
17:30:38FebsNo, one large table is not what I had in mind. For the most part, the itemize-d lists work. However, when one reaches the deepest level of a menu and one is describing options for a particular command, a table would look better than a list.
17:31:02tucozFebs, I agree with you. I think that will make that section look better.
17:31:13 Join tonz [0] (
17:33:31FebsI'll work on that next time I have a chance to spend some time on the manual. (Which will not be today. Too much PaidWork to do!) However, I have a long flight tomorrow and can spend some time then.
17:33:52bluebrother^sounds nice.
17:34:18tucozbluebrother^, do you think I should still allow the empty {}{} in screenshot?
17:34:38tucozor should we just use 1 argument?
17:34:43bluebrother^if we don't have a caption it isn't needed anymore.
17:34:45tucozthat is, the filename
17:34:53amiconnHmm, rockbox bootloader doesn't work either.
17:34:57bluebrother^so I'd say use one argument only.
17:35:03amiconnApple logo, nothing else :(
17:35:07tucozok, I'll see how that works
17:35:21bluebrother^If we'd need more later we could write a wrapper command.
17:35:45 Join qwm_ [0] (
17:35:49 Join webguest68 [0] (
17:35:54 Quit qwm (Nick collision from services.)
17:36:02 Nick qwm_ is now known as qwm (
17:37:22 Join quobl [0] (n=quobl@
17:38:52tucozbluebrother^, maybe we should use floats after all. The we could use captions, and just say [ht!]. The images aren't placed where we want them anyway
17:40:07bluebrother^I feared something like that ...
17:40:26tucozif they don't fit, they are placed on the next page anyway.
17:41:09bluebrother^I found a package some time ago (can't remember the name) that was able placing pictures aside from the text. Maybe this is something to consider later on.
17:42:02tucozYes, I've had a look at that package as well. It makes the text wrap around images. That is, you could place an image on the left or right side of the page, and the text will be written next to it
17:42:16 Quit Febs ("PaidWork calls.")
17:42:28bobTHCSomeone can test this other font(bdf) in a dev env to be sure that have been well generated ?
17:42:30tucozwhat did you say about the 1 page formatting earlier today?
17:42:51bluebrother^I was thinking about if we should use twoside or oneside.
17:43:10tucozok, for the headers?
17:43:11bluebrother^I'd say oneside, as I guess most users will read the manual online and not print it out.
17:43:35bluebrother^I want to make them nicer. As I said, I played around with it yesterday.
17:44:18bluebrother^atm on the header there is the section name, not the chapter name. This is strange for pages with small sections.
17:44:33 Part scf ("Leaving")
17:44:47webguest68Slasheri: are you her?
17:44:53tucozI agree there. The chapter name should be there. I haven't looked at the headers really
17:45:19bluebrother^I'll keep on working on them.
17:45:36bluebrother^So I'll assume we'll stay at oneside for now.
17:46:03webguest68Slasheri: is the tagcache extract track number from title? cause here all my tracks have the numbers at the begining of the title (like made by plenty of Tag software
17:46:07tucozdo that. It isn't that big of a change if we would decide otherwise, is it?
17:46:50bluebrother^not really, aside from the fact that I know fancyhdr nearly to nothing. But I consider this as a problem ;-)
17:47:15tucozok, you are free to fix it :)
17:47:50webguest68Slasheri: I have to go, please reply to my question at webguest68 I'll read the logs
17:48:10webguest68Rockbox is great
17:48:16 Quit webguest68 ("CGI:IRC")
17:49:07safetydanwebguest68, the tagcache will look for id3 tag with the track number first, then look from the beginning of the filename for a tracknumber
17:49:14safetydanat least that's how i read the code
17:49:39amiconnHmm, I have no idea what's happening :(
17:50:19tucozbluebrother^, I think we go for floats, but we should encourage people to use captions. It doesn't look too bad then.
17:50:36amiconnAt first reboot, the rockbox bootloader switches on the backlight. It doesn't manage to display anything (apple logo still shown), then it switches off.
17:51:11amiconnReboots from disk mode behave similar, just that it doesn't switch off, but stays on indefinitely
17:51:31amiconnBacklight goes off after a while, so the timer seems to be running
17:53:56bluebrother^tucoz: ok. I think we should write this down on the style guidelines with a small description how to use the command (link I added as comment)
17:54:24tucozyes, I am about to commit the macro now.
17:55:08tucozthere. comitted
17:55:45 Quit Paprica (Read error: 110 (Connection timed out))
17:56:44tucozyou or me?
17:56:58tucozI do it
17:57:40 Quit ohrn ()
17:57:58bluebrother^I was just playing with fancyhdr. Missed a bracket :(
18:00:57 Join Paul_The_Nerd [0] (
18:01:26 Part Paul_The_Nerd
18:02:07tucozoh, I didn't notice the other stuff in the preamble you added
18:03:31bluebrother^what other stuff do you mean?
18:04:05bluebrother^that stuff that defines the \blank?
18:07:09tucozok, I just saw that
18:08:14 Join BarretJ [0] (
18:08:23tucozBut I think I wait with the other patch. Should I perhaps close that?
18:08:30BarretJdoes rockbox play video files on the video ipod?
18:08:37tucozBagder, no
18:08:57tucozBarretJ, no
18:09:33BarretJah ok
18:10:13tucozBut you could use still use the apple fw for that
18:10:25linuxstbamiconn: I think it's time for you to delve into the IPL kernel source and compare the LCD driver with the one in Rockbox.....
18:14:02tucozbluebrother^, do you really think it's necessary with a 80-cols limit? Some places that is not practically possible. Like, when adding some opts in the text
18:15:26amiconnNow if I only knew where to look...
18:16:28amiconnHmm, probably fb
18:16:30 Quit petur ("CGI:IRC 0.5.7 (2005/06/19)")
18:16:38linuxstbYes - fb.c contains everything.
18:16:46amiconnThis linux source tree is confusing...
18:17:18bluebrother^tucoz: I think we should try when its not too complicated.
18:17:47linuxstbNearly all the IPL-specific files are in that directory (where fb.c is).
18:17:51bluebrother^for example vim breaks lines automatically which makes the indentation nearly unreadable in some cases.
18:18:16 Quit t0mas (Read error: 110 (Connection timed out))
18:18:23tucozsame with emacs, but I think it's somewhat ok with latex-mode anyway.
18:18:32amiconnWhere are these magic hardware versions defined??
18:19:02linuxstbThe IpodPort wiki page lists them.
18:19:04 Quit quobl (Remote closed the connection)
18:19:37linuxstbSorry, IpodHardwareInfo now...
18:20:11amiconnI mean ipod_hw_ver in the ipl sources...
18:20:23amiconnThese seem to be rather small numbers
18:20:55bluebrother^I don't think we should enforce the limit but at least try it. In most cases it isn't that problematic.
18:21:07tucozbluebrother^, ok.
18:21:25bluebrother^Tables for example can be indented similar to wiki style, so no need to adjust them lining up everything.
18:21:26amiconnSeems to be the upper part of the hw revision (?)
18:21:49linuxstbSo it will be 0x7 for your ipod.
18:22:24tucozbluebrother^, hmm. what is wiki style then?
18:22:48amiconnHmm, init is the same as mini1g, but data is sent differently
18:23:08amiconnDo we want one binary for 1g and 2g? That'd mean runtime detection...
18:23:34safetydanWhat font that Rockbox ships with is the largest vertically? xtal-14?
18:23:49linuxstbI originally thought that we could get away with one binary, but I'm now thinking that two would be easier. I guess it's up to you to decide.
18:24:16linuxstbI think the other differences will be in the button driver, and also the extra IRAM.
18:24:35 Join fergie [0] (
18:26:00amiconnThis inl() and outl() business is outright confusing :/
18:26:29amiconnipl checks the hw version at runtime
18:26:55bluebrother^tucoz: like indenting every new field a bit more. At least I've seen this on some wiki.
18:27:18tucozLike the tables are now?
18:27:45fergiethanks for adding the track numbers, now i can finally easily test the gapless playback
18:31:17linuxstbamiconn: Yes - that's the big difference with IPL, there is always just a single "run everywhere" binary.
18:31:26bluebrother^which tables do you mean?
18:31:46linuxstb(the kernel and all the applications)
18:32:11amiconnWhy the hell does ipl use numeric constants, even without any comments?? Grrr!!
18:33:43tucoznevermind. It's just that I think the current indentation works pretty well for tables. like every depth in the definition block gets indented by two spaces
18:34:32tucoz\begin{tabularx}..\begin{center}..key & action\\
18:39:19 Nick tonz is now known as tonz|away (
18:39:48 Quit ansivirus (Read error: 104 (Connection reset by peer))
18:41:16 Quit Nico_P (Remote closed the connection)
18:41:27bluebrother^tucoz: thats mainly the way I meant. But I found lines where the about 3rd depth was indented with something like 60 chars ...
18:41:44tucozthat is wrong then
18:41:44bluebrother^I just changed the plugins to use the screenshot macro.
18:42:13bluebrother^that is also a reason why I stated we should try not to write more than 80 chars per line.
18:42:37tucozI understand.
18:43:21bluebrother^I also have now some headers I like pretty much :)
18:44:33bluebrother^urgs. The "Game controls" table for the wormlet plugin ist pretty much broken.
18:45:03tucozno kidding. There is lots to do with some of the tables
18:46:07tucozI think nls said he was going to work on that plugin
18:47:18bluebrother^nice. I put the plugin changes to the tracker now.
18:47:21 Quit DJ_Dooms_Day ("Trillian (")
18:50:51 Quit Nibbler ("Think of someone of "average" intelligence. Then think half the world is dumber than that.")
18:50:59 Join Nibbler [0] (
18:52:45tucozthanks, comitted
18:54:36amiconnWee, lcd is working :)
18:54:45tucozcongratulations :)
18:54:52amiconnHangs at boot logo though
18:55:34amiconnlinuxstb: I went for a separate build
18:55:53amiconnProbably we should consolidate builds before an iPod release
18:56:00 Join ender1 [0] (i=ychat@
18:56:11goffa__amiconn: what player?
18:56:13linuxstbSounds sensible.
18:56:21amiconngodzirra: iPod mini 2G
18:59:39Slasherihmm, just ported dircache to work with simulators so it might be easier to find the tagcache bugs
19:02:02bobTHCtest2.fnt > plz try it on diff targets
19:02:10amiconnlinuxstb: I have defined a new model number and a new id string for scramble, bootloader etc. Is this the correct way? Any other places I should check?
19:02:28amiconn(Also a new number in configure, of course)
19:02:46amiconnAnd a new config-ipodmini2g.h
19:03:05 Join obo [0] (
19:03:16amiconnI'll test a mini2g bootloader next
19:03:36amiconn(For now I just cheated and loaded a 2g rockbox.ipod with the 1g bootloader
19:04:27 Join schwab [0] (
19:05:21linuxstbamiconn: That sounds like everything.
19:06:56 Quit safetydan (Read error: 110 (Connection timed out))
19:07:07 Quit ender` (Nick collision from services.)
19:07:09 Nick ender1 is now known as ender` (i=ychat@
19:08:01 Join xmixahlx [0] (n=xmixahlx@
19:09:37lostlogicamiconn: I think I mentioned, i have an eager ipod mini 2g tester here once you commit something that has buttons, screen and audio playback for it :)
19:09:48amiconnHehe :)
19:09:57amiconnMy first commit will just have lcd
19:10:07 Join petur [0] (
19:10:25 Quit tonz|away (Remote closed the connection)
19:10:42 Quit Shagnar ("( :: NoNameScript 3.81 :: )")
19:10:54 Quit bobTHC ("Smoke Weed Every Dayz !!!!!!!")
19:10:54 Join charkins [0] (n=casey@pdpc/supporter/active/charkins)
19:11:13 Join quobl [0] (n=quobl@tor/session/x-934399f0669d200b)
19:12:07amiconnlinuxstb: The cpu frequency definitions in the config-ipod*.h files are really odd...
19:12:47amiconnWow, my mini 2g rockbox just shut down itself. Must be the 10-minutes default poweroff timeout. So that's already wroking...
19:13:09***Saving seen data "./dancer.seen"
19:13:16schwabcan anyone help me create a config file with my current settings in rb (h300)?
19:13:33charkinsis there a flyspray task for the new on-the-fly database that's in development?
19:13:50 Join yeahx [0] (
19:13:54tucozschwab, just go to the menu and select manage settings
19:14:48schwabi just select write cfg file, and it'll have all my settings in it?
19:15:20 Quit _FireFly_ ("Leaving")
19:15:32tucozschwab, yes.
19:15:48linuxstbamiconn: They are probably just random...
19:16:02amiconnThey are coldfire...
19:16:11linuxstb:) The 5g and Nano should be correct I think.
19:16:14lostlogicamiconn: preglow tells me the settings for the 5g are wrong in CVS too *shrug*
19:16:39amiconnHmpf. Why did it build bootbox instead of a bootloader??
19:17:03amiconnNm, found it
19:17:09amiconnconfigure needs fixing...
19:19:00 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
19:19:02 Part voodoo_laptop ("Leaving")
19:19:18 Quit yeahx ()
19:22:21amiconnHmm, bootloader works. But now rockbox shuts itself down immediately after boot...
19:23:20tucozthe mini got a will of its own. It doesn't want rockbox. You have to convince it properly :-)
19:23:59amiconnNow back to usual (i.e. hang at boot
19:25:32 Join johnybyku [0] (
19:26:29schwabyou people are amazing, thanks for rockbox! bye
19:26:38johnybykuhi all. i'm back with my question. could one of the developers fix the battery monitoring problem on ipod nano?
19:27:10safetydanjohnybyku, it's possible to fix, just will take time. AFAIK only one dev has a nano
19:27:11 Quit schwab ("CGI:IRC (EOF)")
19:27:36johnybykuthere is a patch in the bugs section that is suposed to fix it
19:28:09johnybykucould anyone just include it in the daily build?
19:29:25johnybykurockbox is absolutely great, but the lack of battery monitoring is really annoing me at the moment
19:29:45johnybykui think it's quite an important thing to do, isn't it?
19:31:26lostlogicjohnybyku: patience, getting battery monitoring calibrated is a time consuming process. hopefully preglow will calibrate it soon... the patch on the tracker only activates the voltage monitoring, but does not properly calibrate it, IIRC.
19:32:12johnybykuok, i got it.
19:32:42johnybykuis there any way i could help him with that?
19:33:12 Join scf [0] (
19:37:46 Join CoCoLUS [0] (
19:38:07 Quit johnybyku ("BitchX: the cure for the common client")
19:40:05fergieDoe the load to ram work for tagcache? the icon of disk access still comes up when I click on artist, albums, etc.
19:40:15amiconnHmm, button handling should be identical to 4g/photo and higher. Probably just ifdefed wrong, as the mini 1G is different
19:40:30 Quit BarretJ ("leaving")
19:41:14lostlogicamiconn: yeah, the mini1g implementer wanted runtime 2g detection, sounded impractical to me, but I hadn't actually looked at it...
19:41:19 Join RotAtoR [0] (n=e@rockbox/developer/rotator)
19:41:43Slasheribtw, has anyone managed to get the sdl sound working on linux without continuous glitches? I get something like a two seconds of playback and three seconds of silence.
19:41:46amiconnlostlogic: We should probably consolidate targets before a release, but not right now
19:42:00preglowamiconn: so, how's it going?
19:42:00Slasherifergie: should work, just make sure dircache has been enabled
19:42:40 Join OPP [0] (
19:42:54OPPquick question, how do i set a backdrop. i put it in the backdrop folder but it still doesnt work
19:43:22lostlogicamiconn: makes sense... code duplication is going to eat us if we're not careful.
19:43:41lostlogicOPP: I assume it's the right file format?
19:43:46OPP24bit bmp
19:44:03OPPdoes it have to be named something special? like bg.bmp
19:44:14OPPor backdrop.bmp
19:44:25preglowSlasheri: sound works fine here
19:44:28amiconnlostlogic: Yes. The bad thing is that sometimes code is identical for units clearly distinguishable by the user, while there are differences when the units look similar
19:44:30 Quit scf (Read error: 104 (Connection reset by peer))
19:44:40Slasheripreglow: hmm, weird :/
19:44:42preglowSlasheri: although it doesn't take much to make it glitch
19:44:49preglowSlasheri: if i adjust the eq heavily, the buffer underruns
19:44:54amiconnE.g. button handling of 4g and mini-2g is identical, but mini-1g and mini-2g are different
19:45:01 Join scf [0] (
19:45:10Slasheripreglow: hmm.. even cpu is idle, so that shouldn't be a problem..
19:45:15preglowSlasheri: specs?
19:45:21Slasherimaybe there is a too big loop somewhere
19:45:29Slasheripreglow: intel P4 2.8 GHz
19:45:35OPPany idea lostlogic?
19:45:36preglowthen you should be in the clear
19:45:43preglowi'm on a 3400+ a64
19:46:30safetydanSlasheri, sdl sim sound works for me
19:46:46amiconnsdl sound works on my linux vm. That is on a Pentium M 1.5GHz
19:46:47safetydantry changing the buffer length in uisimulator/sdl/sound.c
19:47:01Slasherihmm, weird.. i will try that
19:47:04 Join linuxstb_ [0] (
19:47:13preglowsafetydan: how long is it now?
19:47:21safetydanI have a 2 GHz mobile athlon that's usually running around 800 MHz when the sim is playing sound
19:47:43safetydanpreglow, 2048 bytes... but it's only a suggestion to the audio card. 2048 is the recommended value apparently
19:47:57safetydanactually, maybe it's 2048 samples
19:48:11 Join PhR3aK [0] (
19:48:21 Quit biffhero_ (Read error: 110 (Connection timed out))
19:49:09preglowsafetydan: does it use only one buffer?
19:49:36PhR3aKi have a question, it normal that the buttons are only working when backlight is on? i have this issue on my x5
19:49:44safetydanpreglow, I believe it's magic and up to the sound card
19:50:10lostlogicOPP: sorry, haven't actually used backdrops, was just throwing out a thought.
19:50:10peturOPP: browse to it on your player, long-press navi and select 'set as backdrop'
19:50:12PhR3aKi have the newest cvs on my x5
19:50:23OPPthanks petur
19:50:33preglowsafetydan: most sound apis don't seem to feel that
19:51:08preglowsafetydan: with windows mme and dsound, it's pretty common to use multiple buffer, sometimes having something around a second's worth in total
19:51:15preglowbut i don't really expect it'll help us much
19:51:19lostlogicamiconn: yeah, hopefully / maybe we can improve our dealing with that with Linus' target tree
19:51:25preglowsdl might very well do something like that interally
19:51:29preglowinternally too
19:51:48preglowlostlogic: i was thinking of starting to port code over to target/ as soon as 3.0 is out
19:51:57preglowi'm sick of ifdefs
19:52:22 Join frederic_ [0] (
19:52:41safetydanpreglow, it's more likely to be which driver is being used, alsa, esd, oss, arts, etc
19:53:27preglowaudio in linux is a dream
19:53:33preglowa pleasant summernight's dream
19:54:03amiconnWee, buttons are working, and rockbox no longer hangs :-)
19:54:13 Join [TCK] [0] (
19:54:14safetydanIt might be easier if people let go of oss and just used alsa. Which would also eliminate the need for arts and esd (unless you want network audio).
19:54:35amiconnI can browse the filetree, use the menu
19:54:51linuxstb_amiconn: I'm expecting a "Gentlemen..." email before you sleep tonight.
19:54:54lostlogicamiconn: wow, so it wsa just a matter of activating the 4g code for the 2g mini then?
19:54:59 Quit OPP ()
19:55:05amiconnlostlogic: Yes
19:55:27amiconnIn fact changing some ifdefs from APPLE_IPODMINI (defined for all minis) to IPOD_MINI
19:55:28preglow2g mini uses 4g code for what?
19:56:20Slasherihaha, if i change in sound.c fmt.freq = 44100/2, then audio plays without any glitches but at the half speed. (and cpu idling now too)
19:57:30amiconnSomething is fishy with the lcd driver...
19:57:33Slasheriand if i remove SDL_Delay from yield() in thread-sdl.c, then playback goes crazy and jumpy. And progress bar in wps jumps much
19:57:56safetydanSlasheri, try setting SDL_AUDIODRIVER environment variable before running the sim
19:58:09Slasherisafetydan: hmm, how i should set that?
19:58:10 Quit muesli__ (Read error: 110 (Connection timed out))
19:58:15safetydanyou possible values are dma, oss, alsa, esd, and artsc
19:58:15preglowamiconn: afaik, the 2g mini lcd driver doesn't even work right in ipl, so you've got a bit of work there
19:58:26Slasheriah, i will try that :
19:58:39preglowsafetydan, Slasheri: note that setting that environment variable made all audio stop for me, even if i set it to alsa
19:59:02safetydanthe sdl docs are somewhat lacking when it comes to that area
19:59:05amiconnpreglow: I suspect lcd-2bit-horz.c
19:59:32amiconnThe access itself works fine, but some rectangles are off (e.g. when clearing things)
19:59:36Slasheripreglow: hmm, same here.. now i have no sound at all :/
19:59:58safetydanSlasheri, apparently SDL_DSP_NOSELECT might also help
20:00:02goffa__PhR3aK: are you still here?
20:00:13safetydan is a list of env variables, maybe one of them will help
20:00:13goffa__if so, i notice the same behavior on mine
20:00:28goffa__the first press activates the backlight
20:00:36goffa__2nd press does whatever you press
20:00:51 Quit linuxstb (Read error: 110 (Connection timed out))
20:00:52PhR3aKi thought its a feature of rockbox or something
20:00:55goffa__i think it was designed that way
20:01:00 Nick linuxstb_ is now known as linuxstb (
20:01:11goffa__which is fine, but kind of the opposite of what it used to do
20:01:52PhR3aKits sometimes not that practical because i swich songs with my x5 in my pocket
20:01:53amiconnWow, this microdrive really spins up fast
20:01:59amiconnSpinup time: 460 ms
20:02:14goffa__yeah... i know
20:02:20goffa__you have to remember to press it twice
20:02:21PhR3aKand then i have to presss the button 2 times...
20:02:44goffa__it'll be fine once i get used to it i think
20:03:01PhR3aKwith the newest cvs i didnt notice any sound issues anymore... did you?
20:03:19goffa__they seemed to have fixed that
20:03:22PhR3aKok... and what says the battery life?
20:03:25safetydangoffa__, PhR3aK, I think you can configure the button behaviour
20:03:48goffa__didn't do the battery test.. let me check
20:03:49amiconnNow the problem is how I commit my button fixes. Both button.c and system.c have unrelated changes I don't want to commit yet
20:04:02amiconnSeems I have to duplicate my changes in my second tree...
20:04:20safetydangoffa__, under General Settings -> Display -> LCD Settings -> First Key... something
20:04:28goffa__oh.. ok.. i'll look
20:04:31goffa__thanks for the tip
20:04:36PhR3aKno i mean do you own the x5l or x5? i have the x5l but i havent done any tests
20:04:46safetydannot sure if it controls the behaviour you're seeing though
20:04:51PhR3aKbut i heard that the battery life should not be that good
20:04:57amiconnAny1 here with a 4G grayscale?
20:05:09goffa__hmm... thing has run for like 2 hours.. says 61% and only shows 22 seconds left
20:05:18goffa__so i'd say that's not very accurate :)
20:05:26goffa__but i can live with that
20:05:38amiconnpreglow: I want wheel clicks...
20:05:59preglowamiconn: i've got code for it
20:06:06PhR3aKhm... why does no one know how much mAh the x5l battery got...
20:06:09 Part tucoz ("Leaving")
20:06:52goffa__PhR3aK: i set mine to 3200mah...
20:07:00goffa__it had battery warning flashing for 4 hours
20:07:07goffa__so i wouldn't worry about it
20:07:23goffa__cool.. safetydan that worked
20:07:31goffa__PhR3aK: you should try that too
20:08:01PhR3aKimo i have the cowon firmware flashed
20:08:16PhR3aKbut i think i will give it a try tonight
20:08:22goffa__oh ok
20:08:33goffa__yeah... rockbox for me is lightyears better
20:08:44PhR3aKfor me the cowon firmware sounds better...
20:08:50goffa__but it gives me access to 30% more music in my collection that i can play
20:08:52PhR3aKand the battery lasts linger
20:09:01goffa__i don't use the "sound enhancements" on the cowon
20:09:07goffa__battery does last a lot longer
20:09:29goffa__with cowon
20:09:38PhR3aKi dont use them too
20:09:49PhR3aKthey make the sound kind of unnatural
20:11:05PhR3aKwhich version do you own? x5 or x5l?
20:11:21goffa__i think i got 10 hours of playback yesterday
20:11:29goffa__maybe 12
20:11:38goffa__but a lot of turning off, turning on, high volume, and random
20:11:40PhR3aKand then it was empt?
20:12:00safetydanPhR3aK, there were some changes to the X5 sound code recently that should fix any issues with high frequency sounds on the X5
20:12:17goffa__safetydan: it helped a lot
20:12:37PhR3aKok... i will give the new cvs a try
20:12:40goffa__i haven't done listening tests with my ety er4-s
20:12:51goffa__sounds better on my ascend acoustic setup though
20:13:18PhR3aKimo im just using the mx 500, my cx 300 should arrive tomorrow
20:13:19 Quit charkins ("Leaving")
20:13:22goffa__haven't tried my energies in the living room yet either
20:13:39kkurbjundoes anyone know if it's possible to get the ipod mini's to read/write to CF cards that don't have a true IDE mode? I have a broken one and a few CF cards, but I can't mount any of them when they're in the ipod
20:14:18 Nick jborn_ is now known as JoeBorn (
20:14:27PhR3aKthe only bad thing for me is that the most wps's have no clock
20:14:49goffa__i always wear a watch
20:14:57goffa__so i seldom use the clock on the iaudio
20:15:00PhR3aKi havent got a watch
20:15:01goffa__i've been using engineer
20:15:07goffa__for my theme
20:15:13goffa__black background white foreground
20:15:30goffa__i like the what's next
20:15:54PhR3aKjep me black background an orange foreground... and i used the theme which is the last one seen from the top
20:16:04safetydanyou can always add the clock to the wps
20:16:35PhR3aKi cant compile... so...
20:16:42amiconnWhy does rolo not work on iPod? It does nothing...
20:16:58safetydanPhR3aK, wps's are just text files, you can edit them in notepad
20:17:48b00st4[20:28] <PhR3aK> the only bad thing for me is that the most wps's have no clock
20:17:53goffa__wow.. thanks again safetydan
20:18:00b00st4i would add one too if i knew how...
20:18:33PhR3aKhm... i will try that
20:18:44b00st4u know, what to edit?
20:19:07PhR3aKnot really... i just wanted to try it out
20:19:45b00st4mustn be too difficult
20:19:54PhR3aKimo im reading this
20:20:02b00st4but if u solve, could ya pls send me a pm?
20:20:23PhR3aKim not registered so i cant send pm's
20:20:51safetydanMaybe also read the mistic river forum on WPS creation
20:21:01safetydanthose guys have been creating them for a while
20:21:03 Quit [TCK] (Read error: 110 (Connection timed out))
20:22:02 Quit frederic_ ("Chatzilla 0.9.72 [Firefox]")
20:23:10b00st4safetydan where should i add the codes to appear in the headline?
20:23:18b00st4or status bar
20:23:22PhR3aKoh man, the iriver 3xx themes look very good
20:23:27PhR3aKvery very good
20:24:04safetydanb00st4, I haven't actually edited any WPS's myself and it would depend on the wps
20:24:42safetydanyou can always experiment to find out :)
20:25:27b00st4so i just have to edit the file on my ipod? no compile?
20:25:49safetydanno, no compile
20:25:54Mooson your computer and put it in .rockbox dir
20:26:33goffa__wow yeah
20:26:37goffa__i like joltamp
20:27:54Slasherihmm, changing SDL_Delay(counter++ % 2); in yield() in thread-sdl.c seems to have fixed the playback issue
20:28:14safetydanso sleep 1 or 2 ms?
20:28:15goffa__might have to try that tonight when i get home
20:28:42linuxstbamiconn: Implementing Rolo isn't very useful on the ipods - you have to reboot to enter/leave usb disk mode anyway. So no-one's looked at it.
20:28:42Slasherisafetydan: it was SDL_Delay(1); and now it's something like SDL_Delay(0.5);
20:28:58safetydaninteresting that would fix it for you
20:29:01amiconnlinuxstb: rolo is useful for rebooting...
20:29:06Slasherisafetydan: but changing it any lower, causes audio to jump, or higher and it will glitch
20:29:18amiconnlinuxstb: Is there a reason why config-ipodmini.h defines HAVE_WM8975 ?
20:29:19safetydanAFAIK SDL_Delay only guarantees 10ms accuracy
20:29:37linuxstbamiconn: No reason. Looks wrong to me.
20:29:44amiconnAccording to the ipl sources, both gen minis have WM8721, which is WM8721 without recording
20:29:58amiconnErr, the second should be WM8731
20:30:31preglowamiconn: if you want to fix rolo, you need to reset some hardware
20:30:42preglowamiconn: for one, you need to undo the memory mapping
20:30:43amiconnIs there working WM8731 support in rockbox?
20:31:07linuxstbI implemented WM8731 support, and it _should_ work, but doesn't.
20:31:41amiconnMini 1G didn't have audio working either, did it?
20:32:11amiconnOkay, so I'll change to HAVE_WM8721
20:32:30amiconnIn fact the minis should have WM8731 but are handled as WM8721
20:32:41amiconn1G and 2G have 'true' WM8721
20:32:49amiconn3G has 8731
20:33:17 Join damaki_ [0] (
20:33:43b00st4PhR3aK works for me
20:34:03b00st4just have to reload the wps
20:34:07PhR3aKyeah? how did you do it
20:34:16b00st4just edit
20:34:41PhR3aKok... but i have to look where there is a free space for the clock right?
20:34:43b00st4added %bs and a clock
20:35:01PhR3aKi mean on the wps
20:35:01 Join darkless_ [0] (n=darkless@
20:35:02b00st4do it somewhere between the bmps
20:35:08b00st4no free space
20:35:13b00st4just somewhere
20:35:16 Join TCK [0] (
20:35:20b00st4u will see, where it appears
20:35:27b00st4than u can chance
20:35:38 Quit darkless (Nick collision from services.)
20:36:53PhR3aKwill it be possible to make themes like this for the x5?
20:38:25PhR3aKbut i cant use for example iriver themes on my x5 right?
20:40:36MoosPhR3aK: sure but keep in mind the X5 resolution screen
20:40:50goffa__yeah... i forgot what that was
20:40:54goffa__pretty low
20:40:54Moosyou will need to resize and adapte
20:40:58PhR3aKyeah i know 160x128
20:41:10Moosthen go go
20:41:29PhR3aKok... is this very difficult? im not that good coder you know ^^
20:41:33goffa__yeah... i'd be using a smaller font.. thats for sure
20:41:45 Join RedBreva [0] (
20:41:45Moosno code knowledge needed
20:42:03Moosjust look at the link safetydan pointed to you
20:42:07 Join Shadowarrior13 [0] (
20:42:14goffa__yeah.. change a few variables, etc
20:42:16Moosit's the WPS "language"
20:43:19MoosPhR3aK: I'm currently reusing my iriver WPS :-) then no color and others yet :-)
20:44:35fergieQuestion how do you put dircache on? I put it on reboot and then it is back off?
20:44:57b00st4hmm, cant find the place to edit, that it will appear in statusbar
20:45:06Slasherifergie: put it on, exit the menus (go to the dir browser) and then reboot
20:45:09b00st4moos are u familiar with wps?
20:45:39Moosb00st4: not a lot sorry, but what is your problem?
20:46:16b00st4i cant get the clock into the statusbar
20:46:18 Quit damaki (Read error: 110 (Connection timed out))
20:46:24fergieThanks it works now
20:46:36b00st4and %bs is always hidden under other icons
20:46:44Moosboosts4: is it diplays somewhere in the screen?
20:46:54b00st4right in the middle
20:47:12Moosthen just find one good localisation to put it
20:47:28MoosI mean where they aren't text displayed
20:47:29b00st4i already put it on top and it still was tere
20:47:54Moosb00st4: do you use sim?
20:48:00 Join Kohlrabi [0] (
20:48:05b00st4whats this?
20:48:43b00st4just know it with compile
20:48:52b00st4what is i good for?
20:49:12Moosin your case try your WPS in ;-)
20:49:23b00st4how can i do that?
20:49:27 Join miner49er [0] (
20:49:38Mooscan you compile yourself?
20:49:58miner49erhow does one compile oneself?
20:50:00Moosthen choose just sim instead of normal build
20:50:10b00st4but not needed for wps i thought
20:50:14Moossorry for the english :)
20:50:22b00st4its ok
20:50:26Moosminer49er: but Hi btw
20:50:26b00st4im not native too
20:51:00b00st4what do i get after compile sim?
20:51:01miner49erMoos: I was 'joking', sorry english humour.
20:51:10Mooshehe :-)
20:51:25miner49erMoos et al: Hi there!
20:51:34Moosthe famous english humour ;)
20:51:52b00st4hi back
20:52:02b00st4so? moos
20:52:13miner49erMoos: yeah, generally just sarcasm...
20:52:15b00st4btw, ur german?
20:52:23PhR3aKim german, btw ;)
20:52:48b00st4u already succeeded with ur wps?
20:52:54PhR3aKbut i think were not allowed to speak german here right?
20:53:07PhR3aKim currently doing homework :D
20:53:13b00st4i just cant get it into the status bar, thats all
20:53:23b00st4id had to do so too
20:53:31 Join biffhero_ [0] (
20:53:33Moosb00st4: just build one sim and try on it !
20:53:49Moosmore fast and easy for test
20:53:57Moosthan on the device
20:54:30b00st4whats the difference?
20:54:36b00st4normal is easy too
20:54:39b00st4and fast
20:55:01b00st4but it wont help me with wps i think, uz its easier to just edit the wps
20:55:07miner49erCan anyone help me with a small coding issue?
20:55:46 Join Nico_P [0] (
20:55:59Moosb00st: can you make screen dump?
20:56:19Moosof the WPS screen you have
20:56:49miner49erI'll throw the question in anyway: What does the numerical value passed into button_get_w_tmo actually mean? Is it ms or what?
20:57:30miner49erHow does it relate to the value returned from current_tick
20:59:25DBUGEnqueued KICK PhR3aK
20:59:25PhR3aK(b00st4): got icq?
20:59:48miner49erActually, forget it. I think I just found my answer!
21:00:47Moosminer49er: you are coding one new game again?
21:00:52 Join Chamois [0] (
21:01:11miner49erlol, still on the same one. I'm trying to get it accepted into CVS!
21:01:34Moosgood luck then ;)
21:01:39miner49erBut, I want it to be as perfect as possible before I submi my final change.
21:02:02PhR3aKwas the "jap" to my answer?
21:02:29PhR3aKok... could i add you?
21:02:56b00st4moos: so with sim i dont have to put files on my target to see if it works?
21:03:05b00st4why, wanna get my ip?
21:03:57 Quit obo (Read error: 104 (Connection reset by peer))
21:03:58bam_any good news on the tagdb?
21:04:16 Join obo [0] (
21:04:21Moosb00st: that is the purpose of sim
21:05:04 Quit Chamois (Client Quit)
21:06:03linuxstbbam_: tagdb is dead. tagcache has been born.
21:06:27Mooshahaha XD
21:06:42b00st4coo, ill check that out
21:07:39b00st4im sick of unplugin my ipod all the time :)
21:08:48 Join XavierGr_ [0] (
21:10:55 Quit quobl ("Leaving")
21:11:06XavierGr_Open Proxy found on your host. Please visit for more information
21:11:06XavierGr_What the hell?
21:12:17preglowyou got a mailserver?
21:12:18amiconnlinuxstb: The WM8731 register definitions are all wrong...
21:13:13***Saving seen data "./dancer.seen"
21:15:03XavierGr_preglow: no
21:15:39XavierGr_I just tried to connect with trillian and it will not let me. It says I am banned. But why can I join from the web client?
21:15:40fergieIs there any reason why dircache is off by default?
21:16:04XavierGr_fergie: Buffer hungry feature
21:16:22fergieHow hungry is I may ask?
21:16:26XavierGr_(though I like it like mad)
21:16:39XavierGr_depends on your file structure
21:17:00XavierGr_for me that I have 4000 tracks it is around 450-500 kB
21:17:06fergiethere is 32 mb buffer in ipod 4G I think?
21:17:10peturXavierGr: using IE? there's a flaw in it that allows it to be used as a proxy. As a result, your IP can become blacklisted
21:17:24fergieso that not that much
21:17:27XavierGr_petur: no I use only firefox
21:17:30preglowXavierGr_: because the web client is based on another host than your own
21:17:46fergiehow can one check the size of the dircache?
21:17:48preglowfergie: it's also a bit buggy yet
21:17:57XavierGr_fegie: debug menu
21:18:19fergieis that the one with keep out?
21:20:34 Quit bam_ (Remote closed the connection)
21:21:12fergieThe dircache is 128010 Bytes on my ipod.
21:22:20preglowwe might have musepack seeking soon
21:23:28Mikachudoes tagcache need dircache?
21:23:55SlasheriMikachu: no, but it makes it work faster
21:23:58preglowit it does benefit from it
21:24:40Mikachui think i saw some places that looked like they waited for dircache to start but i didn't check for ifdefs higher up
21:25:11amiconnI have SOUND on mini 2G :-)
21:25:19Mikachuis that gentlemen material?
21:25:29amiconnI think so....
21:25:36SlasheriMikachu: to load tagcache in ram, then dircache is required
21:25:43Mikachuah, okay
21:25:50Slasheriamiconn: ooh =)
21:26:04fergieIt benefits a lot, also the disk isn't touched so it sould save battery time.
21:26:24amiconnlinuxstb: In fact I was wrong about the registers. It's just that mini2g uses the WM8731 driver, but has PP5020 instead of PP5002, which wasn't accounted for in rockbox
21:26:55preglowall the other ipod targets were gentlemen material, so...
21:27:05 Join Bger [0] (n=Bager@
21:27:12CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:27:12*amiconn is listening to Fiamma Fumana, "1.0"
21:27:15preglowyank that sentence out of context, and you've got a hit
21:27:43amiconnIs that a porting speed record, btw?
21:29:23preglowgot to be something close to it
21:29:26Mikachuwas mini1g already working?
21:29:30linuxstb:) It took me a similar amount of time to get Rockbox running on the 5g, but I didn't get sound.
21:29:31preglowMikachu: no
21:29:39Mikachuthen it's pretty impressive
21:29:54preglowMikachu: that is, lots of the code needed was commited, but not working
21:30:12linuxstbamiconn: So that means the wm8731 driver should be working - it's just the pp5002 parts of it that are wrong?
21:30:22 Join t0mas [0] (
21:30:38amiconnpreglow, linuxstb: Btw, I used CONFIG_CPU PP5020 for now. I think I should switch to (true) PP5022
21:30:55linuxstbWe've done the same for the PP5021 devices - Nano and 5g...
21:30:59amiconnI'll probably need that for the proper diskmode cookie, right?
21:31:02preglowamiconn: i don't really think i matters much, only difference i can think of is the iram
21:31:22amiconnDoes PP5021 also have 128KB of IRAM?
21:31:32linuxstbamiconn: I think I just use IPOD_NANO || IPOD_VIDEO when setting the cookie.
21:31:47amiconnOkay, so I have to add IPOD_MINI2G then
21:32:01linuxstbJust grep for "diskmode"
21:32:33 Join luigi_ [0] (
21:33:35 Quit sneakums ("wossname")
21:34:56webmindout of curiosity, will the ipod port get more optimalisations.. I noticed the osciliscope stops the music evey n-seconds
21:35:31amiconnlinuxstb, preglow: I don't see any mp3 performance problems.
21:35:36Mikachumaybe it will get optimifications, but probably not optimalisations
21:35:43 Join bam_ [0] (
21:35:45linuxstbamiconn: Try enabling the EQ
21:36:03amiconnHmm, is variable frequency enabled?
21:36:20Mikachuwebmind: try lowering the scroll speed from max
21:36:43linuxstbNo - but I think it's just a matter of enabling the #define in the config file. The code works on the Nano/5g but crashes on my 4g Color.
21:36:59 Join maeck [0] (n=chatzill@
21:37:51amiconnpreglow: With EQ enabled the buffer fills very slowly
21:38:05preglowamiconn: it's too slow
21:38:16preglowamiconn: with mp3 and all eq bands, it skips here
21:38:24amiconnThe UI doesn't lag at all though
21:38:51 Join sneakums [0] (
21:39:09preglownah, that has improved significantly at some point
21:39:12preglowi don't really know when
21:39:18preglowbut now even peak meters are pretty fluid
21:39:29amiconnYeah, all 5 bands make it skip here as well
21:40:25amiconnApproximately once every 10 seconds. That's with ~ 177kbps VBR mp3
21:40:43preglowmp3 is too slow
21:40:48preglowand i'm planning to look into it
21:40:53preglowi alreaedy know synth_full can be made faster
21:41:19preglowi need to sit down and focus on a couple of things, i've got too much stuff uncommited
21:41:36*amiconn too
21:41:37preglowand not too many other people seem interested in codecs
21:41:53amiconnI'll commit mini2g sound, then turn back to .lang
21:42:01preglowand i need to find out why the flaming hell iram is so slow
21:42:03preglowdamn, that annoys me
21:42:18Mikachucodecs are sort of harder than other stuff
21:43:12amiconnBah, I really need to look into lcd-2bit-horz.c
21:43:32amiconnMono bitmap drawing is fishy, scrolling text looking weird
21:43:41preglowMikachu: nah, not really, i very seldom actually have need to actually modify the codecs, i just speed them up
21:43:49preglowwhich doesn't even always require asm
21:46:55fergieMaybe stupid question, but will it ever be possible to beat the playtime of mp3's of the apple firmware?
21:46:58Nico_Pis there anyone here familiar with the playback system ?
21:47:24webmindMikachu, is that an osciliscope option, or just the generic rocbox scroll option?
21:47:41Nico_Pi'm having problems with the current read position...
21:47:48Nico_Pi never seem to get it right :(
21:48:26preglowlinuxstb: we got any mini gen1 people?
21:48:38webmindMikachu, nm, found it
21:49:56 Quit luigi (Connection timed out)
21:50:33 Join ipod5G60Gb [0] (
21:51:15ipod5G60Gbk so any one played around ines, i need some help
21:51:38preglowcome again?
21:53:30fergieanyone who can answer my question?
21:53:48amiconnI need to fix the port debug screen for iPod mini. It doesn't fit...
21:54:18amiconnHaha, battery voltage is way off. 11.61V ...
21:54:36preglowamiconn: yeah, i bet the scaler is completely off the bat
21:54:39preglowit was on nano too
21:54:45preglowi had a voltage of 10.3 at half gauge
21:54:49Nico_Pno one to help me ?
21:54:53preglowahh, right, another thing for me to do
21:54:57preglowfix the ata sleep hack
21:55:19preglowamiconn: how should i deal with that, really? just pretend that the disk spins for a couple of seconds after each access?
21:55:32amiconnI'd say just fake the sleep, i.e. let the rockbox code do its sleep magic, just don't send the actual command
21:55:33preglowideally the bloody thing should just work, but that seems to be impossible
21:55:45b00st4scripter of ipod_unplug here?
21:56:25bam_after searching for a song per say (tagdb) how does one create a playlist from the search?
21:57:02amiconnI love that microdrive spinup time. Now someone make larger microdrives, please ;)
21:57:11preglowhow fast is it?
21:58:01amiconnaround 500 ms
21:58:22amiconnpreglow: Are nanos always formatted FAT32?
21:58:22preglownot too shabby
21:58:26amiconn(disregarding HFS)
21:58:35preglowamiconn: i think so
21:58:37preglowi think all newer ipods are
21:58:47amiconnThe 1GB version might be FAT16?
21:59:05 Quit maeck ("Chatzilla 0.9.72 [Firefox]")
21:59:08preglowseriously doubt it
21:59:20preglowdoesn't fat16 only support 8.3 ?
21:59:29amiconnFor large files like music, FAT16 has less overhead
21:59:41amiconnNo, that has nothing to do with the FAT bit width
21:59:47preglowthen it might, i have no idea
21:59:49preglowi still doubt it
21:59:54preglowapple prefers to stick to their feature set
21:59:58amiconnThe LFN extension is called VFAT, and works on all FAT variants
22:00:00fergieWhat are the chances that newer rockbox versions will beat the total playtime of one battery cycle with the ipod firmware?
22:00:05preglowahh, right
22:00:12preglowfergie: very slim
22:00:26preglowfergie: the chances are slim we'll ever reach that level
22:00:47amiconnpreglow: I'm asking because we have to support both FAT16 and FAT32 on Ondio, so it would be simple to enable FAT16 support for nano
22:00:54fergiewath do apple so much better then?
22:01:15preglowamiconn: well, i've no idea
22:01:28preglowamiconn: i might find out soon, though, i'm going to mod an ipod 1gig for a client soon
22:02:32preglowyeah, mod and do a small custom rockbox
22:02:37preglowfor acoustics measurements
22:02:45preglowam i the first to be paid for rockbox work? :)
22:03:49 Quit Kohlrabi ("Fast alle Menschen sind Regenwürmer")
22:03:59 Join Kohlrabi [0] (
22:04:09lostlogicso fergie's question stands... why is our battery life so bad on ipod?
22:04:31peturand h3x0...
22:04:47preglowbecause our codecs aren't as optimised as apple's
22:04:52preglowor anything at all
22:04:55amiconnOur codecs are not as optimised as the commercial ones
22:05:11preglowapple almost certainly uses a decoder written in pure asm
22:05:11amiconn...and on H300 there's the mysterious power drain problem
22:05:23preglowby people who know what they're doing
22:05:42amiconnIriver most probably uses the motorola mp3 decoder
22:05:55fergiebut is it only the codecs where the time is lost?
22:05:56amiconn...which is well optimised
22:05:57preglowand there are plenty of mp3/aac decoders for arm floating around
22:06:04preglowfergie: no, not only, but very much
22:06:11preglowfergie: the codecs are by far the biggest cpu drain
22:06:48amiconnBut, I had the impression that the motorola codec doesn't sound as good as libmad
22:06:57preglowme too
22:07:01preglowi think they cut some corners
22:07:08preglowamiconn: btw, you got any idea how much work is left for 1g mini?
22:07:39fergiethis is probably illegal, but can you locate the mp3 decoder in the apple firmware?
22:07:56preglowfergie:of course
22:07:59preglowfergie: but it wouldn't help us
22:08:11amiconnpreglow: Probably *very* little, if at all
22:08:25preglowamiconn: there's a girl i know that wants rockbox on her ipod, and i think she's got a 1g mini. might be able to borrow it to do the port
22:08:31amiconnmini 1G uses PP5020, and the button driver is already in.
22:08:40preglowyeah, doesn't look like much is required
22:08:52amiconnWM8731 driver should work out of the box as soon as I commit it
22:09:08amiconnIt may even just work...
22:09:35 Join [IDC]Dragon [0] (
22:09:38amiconnOf course there are still some quirks left, but having it playing music is the milestone
22:10:18 Join pfavr [0] (
22:10:38amiconnGood evening to Hanover :)
22:10:52[IDC]Dragonand to Berlinn
22:11:11amiconnPls look up the english spelling of Hannover
22:11:24[IDC]DragonI see there's one more Ipod in berlin
22:11:35amiconnAnd it even plays music :)+
22:11:41[IDC]Dragonso you went to the dark side
22:11:44amiconn...from the HD, without iTunes
22:11:55 Join nave7693 [0] (
22:12:01*[IDC]Dragon goes reading backlog
22:14:32Nico_Pnoone to help me with the playback code ?
22:14:51[IDC]Dragonand me with recording?
22:15:33[IDC]DragonPCM, on Archos ;-)
22:15:52petursorry ;)
22:16:41safetydanargh, someone would post again to the real names thread
22:16:56peturit 'll never end :D
22:16:58[IDC]DragonI fail to see how to tell the recording it should use parallel port
22:16:59preglowargh, could someone just ignore the real names thread
22:17:05pfavr[IDC]Dragon: you're working on PCM recording for Archos?
22:17:19[IDC]Dragonworking is an exaggeration
22:17:29*safetydan adds thread to ignore list and never mentions it again
22:17:32peturor maybe do what the subject says :)
22:17:50pfavrif you get it working one day I might use it :-)
22:18:09safetydanOn the upside, I have an equalizer screen that scales with the resolution and user chosen font
22:18:38pfavrgoodnight everybody
22:18:42 Quit pfavr ("ChatZilla 0.9.61 [Mozilla rv:1.7.12/20060205]")
22:18:48 Quit linuxstb (Read error: 104 (Connection reset by peer))
22:18:51 Join linuxstb [0] (
22:21:22 Join Paul_The_Nerd [0] (
22:21:24amiconn[IDC]Dragon: Huh? The PCM codec only supports PIO for recording, so there's nothing to tell
22:21:32[IDC]Dragonah, ok
22:22:27*[IDC]Dragon has to read more thoroughly
22:23:19amiconnAll the audiocodec registers keep their function when the PCM codec is loaded. Only the DSP memory cells change
22:23:19*[IDC]Dragon waits for a "Ladies" email
22:25:17preglow"Fellow living beings, we have sound!"
22:30:31amiconnSlasheri: You caused a 64bit sim warning...
22:30:49amiconn(with enabling dircache)
22:31:24amiconnBagder, t0mas: Build system seems to hang
22:31:42*t0mas checks
22:31:57t0mastomas@labb:~/rockbox-auto$ Read from remote host Network is unreachable
22:32:00t0masConnection to closed.
22:32:03t0masthat doesn't look good :)
22:32:46Bagderits reachable to me
22:32:51t0masto me now too
22:32:54t0masjust had to reconnect
22:32:55obob00st4: patch has been updated
22:33:12t0masBagder: 11923 ? 00:00:00 find <defunct>
22:33:28Bagderyeps, looks weird
22:33:40Bagderthe log says 2 builds left
22:34:12t0mastry killing it as root?
22:34:16t0masI can't kill it :|
22:35:06 Quit RedBreva (Excess Flood)
22:35:29b00st4cool thx
22:35:35b00st4for unhold too?
22:35:50 Quit linuxstb (Read error: 104 (Connection reset by peer))
22:35:51obob00st4: no - new features need to be added as new patches
22:35:53 Join linuxstb [0] (
22:35:55 Join RedBreva [0] (
22:36:15b00st4but thx for this one
22:36:24Bagderits a zombie, its not easily killed
22:36:26b00st4when its gonna be embeddet to cvs?
22:36:31obob00st4: no worries
22:36:45obob00st4: erm, pass
22:37:12biffhero_obo, I am trying your headphone patch against a clean CVS checkout
22:37:22obob00st4: one of the devs needs to look at it, decide if it's a wanted patch, and if it's good code
22:37:26Bagdert0mas: kill the whole buildmaster script
22:37:27fergieAnother question: How does the mp3 decoder in Rockbox compares to the apple one quality wise? or is there on difference?
22:37:40biffhero_patching file apps/settings.c Hunk #1 succeeded at 578 (offset 1 line).
22:37:47t0masBagder: yeah, that works
22:37:50t0masit was the parent
22:38:09BagderI figured
22:38:10b00st4hope your patch will be in it soon, its really usefull
22:38:12obobiffhero_: yes, there was a recent commit that affected the patch
22:38:13Bagderweird though
22:38:26fergieIf've some tracks where the apple decoder always skips, the rockbox one never skips
22:38:30*Bagder listens to the improved x5 sound
22:38:38Bagderand hears no difference ;-)
22:38:41biffhero_building now
22:38:43obob00st4: yup, me too :)
22:38:59t0masBagder: I've fixed the buildlog... on the next rebuild it should just remove the failed line
22:39:13Paul_The_Nerdfergie: Do you have a Nano or Mini?
22:39:22fergieIpod Color
22:39:28Paul_The_Nerdfergie: Aah. VBR tracks though?
22:39:29MoosBagder: that was bugfix
22:39:30ipod5G60Gbi have noticed no difference in sound quality on a 5G
22:39:36fergieThe welknown? VBR issue
22:39:44BagderMoos: I know
22:39:58Paul_The_Nerdfergie: The VBR issue doesn't affect rockbox. CPU frequency adjustments are handled quite differently.
22:40:46MoosBagder: I happily listened sound for my musics need yesterday and today, that Rocks! :)
22:40:56fergieyep it has only improvements compared to apple
22:41:51fergieIf only Rockbox would have 80% of apple's playtime it would be perfect
22:44:17preglowit'll come
22:44:21preglowbut there's a long way to go
22:44:42preglowand the bloody lack of documentation isn't helping
22:44:49*preglow curses portalplayer
22:44:57linuxstbfergie: Have you done any comparisons of battery life?
22:45:37fergieNo complete battery tests
22:46:28fergieMaybe I will run a complete one tomorrow
22:46:38amiconnBagder: Could you setup a mini2g target build?
22:46:38Bagderamiconn: you think we should add a 2g mini to the cvs builds?
22:47:01amiconnSim is not yet possible - image missing
22:47:10*amiconn needs to scan his mini
22:47:14Bagderbut boot is?
22:47:47 Join muesli__ [0] (n=muesli_t@
22:48:17preglowdoom has _massive_ warnings in gcc 4.1
22:48:19preglowof the type:
22:48:21fergieBtw are there some risk at letting ipod run till it completely drained the vattery?
22:48:22preglowp_ceilng.c:94: warning: dereferencing type-punned pointer will break strict-aliasing rules
22:48:28preglowfergie: nah, don't think so
22:48:38preglowfergie: but i recommend avoiding it
22:48:58fergieBut how will I know it's empty?
22:49:18biffhero_obo: know what's funny? if you set your duration to rewind to 3 seconds, by don't set 'pause on headphone unplug' to 'yes', you might think it is broken.
22:49:31preglowfergie: it shuts off...
22:49:57fergieAnd why do you recommend avoiding it? ;)
22:50:16preglowbecause i'm not completely certain how good apple's hardware failsafes are
22:50:47amiconnThe failsafe has to be part of the battery controller for LiIon
22:50:52obobiffhero_: what was 'pause on headphone unplug' set to? off/pause/pause and resume?
22:51:09amiconnOtherwise these beast would be quite dangerous
22:51:29biffhero_obo: it was set to 'off'. ha!
22:51:56obobiffhero_: if it's set to off then it skips the entire thing
22:51:56preglowanyone with a nano currently here?
22:52:09Paul_The_NerdI am
22:52:12obobiffhero_: can you check it's still set to off?
22:52:16biffhero_obo: yeah, I figured that out, after 3 seconds of 'huh?'
22:52:22fergieIt's sooo nice I can finally listen to cd's without enervating myself about the pauses between tracks.
22:52:25 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
22:52:40biffhero_obo: yeah, it leaves it to 'off' even after you mess with the 'duration to rewind'
22:53:06 Quit Sinbios (Connection timed out)
22:53:08obobiffhero_: ah, okay, I see what you're getting at.
22:53:09 Join Sinbios [0] (
22:53:11tucozanyone know how to replace all occurrance of tab with a space
22:53:16tucozin a file
22:53:26biffhero_tucoz: i can do it in emacs
22:53:34Paul_The_Nerdtucoz: Many good editors can do it for you.
22:53:34tucozbiffhero_, oh. tell me
22:53:40tucozI use emacs
22:53:43 Join Genre9mp3 [0] (
22:53:44safetydantucoz, s/\t/ /g
22:53:59biffhero_^Q^I RET <sp> RET !
22:54:01safetydanassuming you actually want a single space
22:54:07 Join MrStaticVoid [0] (n=jlee@resnet-252-226.resnet.UMBC.EDU)
22:54:24tucozsure, I just want the tabs removed
22:54:34tucozand let emacs do the indenting
22:54:56tucozsafetydan, is that with sed or something?
22:55:21safetydantucoz, it's a regular expression so anything that supports those
22:55:21amiconnpreglow, linuxstb: Btw, why are there 2 types of ipod #defines?
22:55:26safetydanI'd be shocked if emacs couldn't do it
22:55:31tucozsafetydan, it can :)
22:55:46amiconnOne set with configure, IPOD_*
22:55:52Bagdertucoz: M-x untabify
22:55:59biffhero_badger: good one
22:56:00amiconnAnother one set in config-ipod*.h, APPLE_IPOD*
22:56:03tucozBagder, ah, thanks
22:56:06 Join ashridah [0] (
22:56:13biffhero_tucoz: you didn't see my emacs-way to do it?
22:56:20Bagderand we have 42 builds
22:56:32tucozbiffhero_, yes. But I didn't get the ^Q^I part
22:56:56Bagderthat's not a working approach
22:57:06Bagderuntabify replaces it with the proper number of spaces
22:57:25biffhero_he said he only wanted 1 space, so that's what I did. I see that untabify uses tab-width
22:57:25safetydanBagder's method will definitely be more what you want
22:57:36tucozbiffhero_, thanks anyway :)
22:58:10tucozbiffhero_, I din't know there was a untabify command, and I just wanted to get rid of the tabs and do the indenting afterwards
22:58:35Bagderhehe, the warnings wander ;-)
22:58:38safetydanweird, I get a divide by zero exception in eq.c when I start the ipod video sim
22:58:41 Part [IDC]Dragon
22:58:44Bagder64bit sims only
22:59:26 Quit Genre9mp3 ()
23:00:07preglowsafetydan: well, perhaps the cutoff is 0 some place, then
23:00:21obobiffhero_: if pause is off and rewind is set, there is no way to tell how long the phones were out for, so I'm not sure why rewinding a set amount would help....
23:00:40amiconnSlasheri: r u there?
23:00:53 Quit ipod5G60Gb ("CGI:IRC (EOF)")
23:01:30biffhero_obo: it was my mistake. I forgot to do the pause setting, only ran in and hit the rewind setting really quickly, and then went "D'oh!" when it didn't work. my bad
23:01:41linuxstbamiconn: That's a good question about the ipod #defines. I can't see a need for the APPLE_* set defined in the config files.
23:01:55amiconnyes, that's what I mean
23:02:12obobiffhero_: ah, okay
23:02:15amiconnI went for the configure ones, but having both might be misleading...
23:02:33 Join ipod5G60Gb [0] (
23:02:35*amiconn is going to test variable CPU frequency now
23:02:40linuxstbYes, I think we should get rid of the APPLE ones.
23:02:45safetydanpreglow, it was probably that, forgot to delete the sector3D.bin file
23:03:33amiconnlinuxstb: The APPLE_* ones are used in quite a number of places... Some search & replace work is needed.
23:03:46obois it possible to get some dev feedback on 4899??
23:03:55linuxstbYes. I'll do it, but not until tomorrow.
23:04:01linuxstb(unless you want to do it...)
23:04:21 Join herz42_ [0] (
23:04:43preglowamiconn: just disabling the sleep commands doesn't change anything
23:04:47 Part Paul_The_Nerd
23:04:48preglowi really wonder what this problem comes from
23:06:09safetydanComments on the potential new eq layout here?
23:06:19safetydanBasically vertical sliders with two lines of text at the bottom
23:06:21 Join luigi [0] (
23:06:27safetydanscales with user selected font and screen resolution
23:06:54 Join solexx [0] (
23:07:15amiconnsafetydan: Does it fit the mini's screen?
23:07:30amiconn(afaik the smallest across all devices supporting eq)
23:08:25amiconnpreglow: CPU boost is working just fine on mini 2g
23:08:43*linuxstb curses his 4g Color...
23:08:45Doomed9any tagcache updates?
23:08:58preglowamiconn: how nice
23:09:05amiconnlinuxstb: Are you sure it doesn't work?
23:09:14preglowsafetydan: if it scales, then it's just fine
23:09:24Doomed9oh yeah! gotta get the new cvs
23:09:30amiconnPerhaps you didn't replace everything when you tried. The api changes with it...
23:10:08amiconnpreglow: Hell, 68% boost ratio with mp3...
23:10:23*b00st4 kisses his ipod 5g :)
23:10:36linuxstbI haven't tried it since before preglow committed it, but it just caused random crashes when playing audio.
23:10:47linuxstbMaybe it's time to give it another go...
23:11:06amiconnCrashes how often?
23:11:09Doomed9linuxstb what are u talking about
23:11:21linuxstbamiconn: Did you just define HAVE_ADJUSTABLE_CPU_FREQ in the config file?
23:12:00preglowamiconn: i didn't joke when i said it was slow
23:12:03biffhero_I see the tagcache sorts by track number now
23:12:08preglowamiconn: even tremor is faster than mp3 at most bitrates
23:12:12linuxstbI think it was whenever it tried to change the speed.
23:12:24linuxstbDoomed9: Dynamic CPU frequency on the 4G.
23:12:55amiconnpreglow: Hmm, CPU frequency is working on nano and video (and now, mini2g)?
23:13:06preglowamiconn: is there anything that says i can't just comment out the sleep commands and have it work?
23:13:09preglowamiconn: everything but 4g
23:13:12preglowamiconn: and untested platforms
23:13:15***Saving seen data "./dancer.seen"
23:13:33preglowamiconn: so yeah, basically what you said :)
23:13:38amiconnSo perhaps there's a problem on PP5020. Iiuc nano and video are PP5021, and mini2g is PP5022
23:13:49linuxstbWe only have the 1g mini and 4g greyscale without developers now.
23:13:52preglowi bet we'll find out if i get my hands on that 1g mini
23:13:58miner49erI've added the final version of asteroids, can it be added into CVS now please?
23:14:06amiconnlinuxstb: and the 1g 2g and 3g iPods
23:14:14linuxstbamiconn: The odd thing is that the code is taken from IPL, and was written for the PP5020.
23:14:21miner49erwell, perhaps someone could test it on the irIver and Ipod for me?
23:14:39preglowlinuxstb: bet they never used it like we do, though
23:14:57Nico_Pstill nobody to help me with the playback engine ?
23:15:14biffhero_miner49er: test asteroids on the 5G?
23:15:26Nico_Pi'm almost there, i just have a few ogg failures left
23:15:38miner49erBueller. Bueller. Bueller.
23:15:40 Quit ipod5G60Gb ("CGI:IRC (Ping timeout)")
23:16:16amiconnBtw, the default font is quite large on the mini's LCD
23:16:39amiconnI'm using the same font as I do on archos now: rockfont-8
23:16:59miner49erbiffhero: It might work, if it doesn't let me know soon, whilst i'm in the mood for playing with it. I could try and get the sim working for it I 'spose
23:17:05 Join RoC_MM [0] (
23:17:22biffhero_miner49er: where do I find the code?
23:17:27preglowRoC_MM: yo
23:17:32RoC_MMDid the benchmark.
23:17:41preglowgot the file stuffed somewhere?
23:17:49miner49er −−the last entry has the latest version
23:18:00Bagdert0mas: happened again
23:18:03RoC_MMI'm curious if the informatiion between the last disk access and the battery failure is important...or is that data lost?
23:18:08linuxstbI enabled the adjustable CPU frequency, and my 4g Color crashed immediately after displaying the file browser.
23:18:11biffhero_miner49er: ok, going against latest cvs + 4899
23:18:22amiconnt0mas, Bagder: Builds hanging again...
23:18:40preglowRoC_MM: won't know until we see it
23:18:54RoC_MMi'll post when it's up
23:19:35biffhero_miner49er: do I just put it in rockbox/apps/plugins/ and then re-run make ?
23:20:08miner49erbiffhero: You'll have to it add asteroids.c to SOURCES as well
23:20:08linuxstbNow Rockbox won't even start - it just freezes whilst still displaying the bootloader screen...
23:20:32preglowlinuxstb: that's just weird
23:20:39preglowlinuxstb: at that point, it shouldn't have boosted much
23:20:39safetydanamiconn, it scales itself based in LCD_WIDTH, etc so it should work on "any" resolution
23:20:52preglowlinuxstb: perhaps the lower cpu freq sits really badly with it?
23:20:57preglowlinuxstb: care to tweak the prescalers a bit?
23:20:59safetydannow to see if I have time to finish it before disappearing for a month
23:21:02 Quit safetydan ("Leaving")
23:21:54linuxstbpreglow: Sure. I think I tried that before though. Any suggestions?
23:22:05preglowlinuxstb: no, not anything i can think of
23:22:11preglowtry using predominantely high freqs or something
23:22:25amiconnlinuxstb: Perhaps try different fixed frequencies?
23:22:59 Quit solexx_ (Read error: 110 (Connection timed out))
23:23:07 Quit luigi_ (Success)
23:23:13amiconn(preferably the ones that would be used in the adjustable system)
23:23:22 Join Xerion [0] (
23:23:48b00st4n8 @ all, im off
23:24:15*Bagder builds manuals... many manuals
23:24:41tucozis it Iaudio or IAudio, for the manual that is.
23:24:56 Quit b00st4 ()
23:24:57preglowyou know my stance on that question :-)
23:25:06tucozyes, Iaudio then
23:25:16preglowyeah, sure, but i'm not an authority
23:25:28preglowall i want is that we use the same thing throughout
23:25:43preglowwell, that's predominantely what i want, at least, heh
23:25:50ashridah$(iaudio) or whatever the latex equivalent to a macro is
23:26:04tucozashridah, yes. We will use a macro
23:26:04preglowcumbersome to do that everywhere, though
23:26:05Doomed9cvs update
23:26:15preglowyou need to use a \ after every macro if you want a space
23:27:14tucoznot an iaudio-specific macro though
23:27:29ashridahwell, true, platform, not 'iaudio'
23:27:38ashridahah, crap.
23:27:49*ashridah heads off to uni :S
23:27:51 Quit ashridah ("uni")
23:28:09RoC_MMpreglow, it's posted:
23:28:47preglowouch, five hours
23:28:55preglownot exactly the very epitome of impressive
23:29:05RoC_MMit's not exactly a new battery either
23:29:16RoC_MMand the other two aren't more than 30 minutes away from my time either
23:29:18preglowRoC_MM: but it was fully charged, yeah?
23:29:23RoC_MMyes fully charged
23:29:27 Quit nave7693 ("Leaving.")
23:29:37tucozhmm, I think we drop the Apple use in the manual. And refer to them as Ipod instead
23:29:37amiconnpreglow: Would your backlight dimmer stuff also work on mini 2g?
23:29:37preglowlinuxstb: you got any idea how to fix the battery gauge from the data in that file?
23:29:41preglowamiconn: try it
23:29:52RoC_MMhad headphones plugged in at reasonable listening volume with it just sitting there doing it's own thing playing from a playlist of all the songs on it
23:30:09amiconnAny special #define?
23:30:10 Join Paul_The_Nerd [0] (
23:30:20preglowamiconn: you just need to define BACKLIGHT_HAVE_PWM_FADING in the config
23:30:29 Join nave7693 [0] (
23:30:37preglowahh, no
23:30:45*preglow is tired
23:30:58Doomed9erm...can u view lyrics in rockbox
23:31:15Paul_The_NerdDoomed9: Yes and no. Not synced, but there's a text viewer.
23:31:34Doomed9so u cant look at the lyrics that are in the ID3 tage>
23:32:09amiconnThe mini 2g seems to have some sort of hardware dimming. The backlight fades out over a short period
23:32:21biffhero_miner49er: a couple of observations. 1. the ship seems small on 5G. 2. the evil spaceships seem to come out too quickly, in the first level, at the beginning 3. the shots from the flying saucer is the same size as the pieces of rock or saucer that are flying by. Could the bitmap for the ship have a thing out the back while thrusting?
23:32:27preglowamiconn: you mean by default?
23:32:37preglowi think all ipods have that
23:32:39preglow4g also has
23:32:44preglowbut i don't know how to enable for the other platforms
23:33:11preglowthen i seriously doubt pwm fading will look very pretty
23:33:14 Quit arf-arf (Connection timed out)
23:33:15 Quit Nico_P (Remote closed the connection)
23:33:20biffhero_miner49er: I have shot straight through the heart of the saucer a few times, and not scored a hit
23:33:42preglowlostlogic: omglolroflmao?
23:34:49preglowtucoz: my kind of commit! :-)
23:34:50 Join webguest26 [0] (
23:35:27RoC_MMpreglow, is there any way to know or track or get an idea of how the progress on 4G battery indicator is coming?
23:35:31lostlogicpreglow: ?
23:35:33miner49erbiffhero: I'm amazed it runs! I've copied your comments and will work on them. Thanks for testing. But bascially, you don't hytin it's ready for CVS entry then?
23:35:44amiconnpreglow: Fading backlight itself is working - but it breaks playback (!)
23:36:09amiconnYou can still play music, but it's all distorted, and skipping from time to time
23:36:13preglowlostlogic: how do i fix battery meter from a battery log file?
23:36:19biffhero_miner49er: no, quite the opposite. I think that it is ready for CVS, just fix some bugs as time goes by
23:36:25 Quit bam_ ("Leaving")
23:36:26miner49erbiffhero: Hmm, i've seen that too. Okay, comments appreciated. I'll look into it :-)
23:36:37preglowRoC_MM: just wait a sec and we'll see
23:36:56webguest26phaedrus961: are you here? I'm the submitter of .. I am running the latest version of convbdf. On Linux, 32-bit intel, gcc-4.0.3
23:37:08preglowamiconn: well, weird..
23:37:39 Quit XavierGr_ ("CGI:IRC (EOF)")
23:37:49preglowamiconn: i have no idea why it'd do that, HAVE_BACKLIGHT_PWM_FADING basically just enabled the old h1x0 specific backlight fading code which i adapted to be platform agnostic
23:37:54biffhero_miner49er: odd, I can't get out of it now
23:38:06*preglow falls over
23:38:27tucozBagder, yikes. Now it's out in the open.
23:38:28miner49erbiffhero: that's not good. Maybe the key mapping isn't set up properly for iPod.
23:38:30*tucoz hides
23:38:31lostlogicpreglow: use it to determine 0...10...20... ...90 ... 100% battery voltage in centivolt levels, then put an #ifdef IPOD_NANO those values #else in powermgmt.c near line 184
23:38:44biffhero_miner49er: I was able to exit before....
23:38:47Bagdertucoz: indeed it is!
23:38:49preglowBagder: got any fancy-ass auto-building going on for those?
23:38:59lostlogicpreglow: also cahnge the battery level dangerous for the nano in there in a similar fashion (just above the 0% level)
23:39:00Bagderthis is done with it
23:39:04miner49erbiffhero: I'm confused as to what keys are actually available - I need a diagram or something
23:39:19*linuxstb decides to read the manual and find out how Rockbox actually works
23:39:24biffhero_miner49er: select+menu is usually the exit key
23:39:37preglowlinuxstb: would you mind doing this? i can't test it in any way
23:39:55lostlogicpreglow: the values in that log seem to be with the scale factor wrong, so you'll have to rescale them to get real ones
23:39:55preglownot having a 4g ipod and all
23:39:56tucozI should write a mail to some mailing list to excuse the state the manual is in :)
23:39:58miner49erbiffhero: Is that what worked previously then?
23:40:03biffhero_miner49er: yes
23:40:04preglowlostlogic: well, i don't know the scale factor
23:40:19preglowlostlogic: and i expect i'll have to pop my player open to find that out
23:40:22lostlogicpreglow: it's probably whatever's in CVS, so just rescale from that to the correct one in ipod video cvs
23:40:27lostlogicpreglow: no, no need for that
23:40:35biffhero_miner49er: I think there is some sort of problem between the 'fire to start', and 'fire+menu' to exit
23:40:36lostlogicpreglow: we don't care about _real_ measured battery voltage
23:40:39 Quit PhR3aK ("get satisfied! • :: ««« (Gamers.IRC) »»» ::")
23:40:41lostlogicwe only care about the voltage the ADC reports
23:40:56biffhero_miner49er: maybe "hyperspace to start game" would be the right way
23:41:01Bagdertucoz: I think you can expect to start getting more feedback at least :-)
23:41:01lostlogicpreglow: so why not just use the scale factor the ADC datasheet tells us to?
23:41:12preglowlostlogic: didn't know it specified one
23:41:19miner49erbiffhero: pfuh, gawd knows what's causing that then. Can you start adn play and stuff. Is it only exiting that isn't working?
23:41:20tucozBagder, that is why I'll write a mail.
23:41:42lostlogicpreglow: it specifies the forumula to calculate voltage from ADC value, which I applied to determine the scale factor, which is the same for any ipod with a pcf5060x
23:41:42tucozBagder, will the manual be linked from some page?
23:41:52biffhero_miner49er: yes, I am stuck playing now (and listening to my podcasts) until I reboot
23:41:53bluebrother^I just saw the download links for the manual. Is it build regularly? Say, daily?
23:41:55Bagderyes, I'll make something like the daily build page
23:42:03Bagderbluebrother^: yeps
23:42:08miner49erbiffhero: Why is that preferable to "Fire to start"?
23:42:08bluebrother^nice :)
23:42:10Bagderdaily built manuals
23:42:14biffhero_miner49er: also, I think maybe a top end for the thrust speed is a good idea. I just flew through an asteroid
23:42:31biffhero_miner49er: use 'down', or 'hyperspace'
23:43:25bluebrother^tucoz: I have made some changes to the headings.
23:43:28preglowlostlogic: aight
23:43:32tucozbluebrother^, perfect :)
23:43:35 Quit Bger ("BitchX: nine out of ten doctors recommend it")
23:43:36miner49erbiffhero: I will work on getting the simulator working I think
23:43:45lostlogicpreglow: so from the log file, I'd probably create a graph of time vs. mv, pick out the 11 points I need, then rescale them, then stick them in the powermgmt.c with an ifdef and it's fairly likely to work
23:44:10preglowi just bloody need to make ata.c work properly without the sleep hack and i'll do it for nano
23:44:13miner49erbiffhero: Use down or hyperspace for what?
23:44:22biffhero_miner49er: to start the game
23:44:27bluebrother^still some minor glitches but I like it better.
23:44:53biffhero_miner49er: start a game, hold down 'up', and get going so fast that you only appear on the screen 3 times per time around. cool!
23:45:43preglowamiconn: i don't think you answered, so i'll just regurgitate the question. does anything completely obvious say that i can't just comment out the actual ata sleep commands and expect stuff to work?
23:46:23amiconnI'd expect it to work
23:46:47miner49erbiffhero: I still understand why hyperspace is preferable to fire to start a game. And that normal space physics isn't it? You keep thrusting, you get faster and faster! lol :-)
23:47:21preglowamiconn: it doesn't :///
23:47:32preglowso i'm really curious as to where the bug is
23:47:42amiconnWhat happens?
23:48:03Paul_The_Nerdminer49er: Somewhat, but there's also dust in space, especially around an asteroid belt or whatnot. In "real" space, going too fast would result in micropunctures and eventually death.
23:48:25preglowamiconn: when i try to access anything disk based, rockbox just hangs for a number of seconds, then pretends nothing happens
23:48:55biffhero_miner49er: yes, normal physics, that's true. :-) but not normal asteroids. ;-)
23:49:06miner49erPaul_the_nerd: Yeah, but my spaceship has a special shield that protects it from dust...but not asteroids!
23:49:15amiconnpreglow: Hmm, weird
23:49:26biffhero_miner49er: I also seem to recall that in asteroids there was a speed decay, but I don't see one in your version
23:49:39amiconnVery weird, just like that backlight fading disturbs playback...
23:50:07preglowamiconn: like loading a plugin, it just says 'loading' for around three seconds, then indicates failure
23:50:18Paul_The_Nerdbiffhero_: There's a GPLed asteroids out there, because they were distributing it at Gamestop as a J2ME app from this little kiosk. It's J2ME obviously, but the physics in it seemed pretty correct
23:51:34 Join Daishi [0] (
23:51:48 Join matsl [0] (
23:51:48webguest26Am I the only one who gets a segfault if I try to convbdf this file:*checkout*/xorg/xc/fonts/bdf/100dpi/UTB___24.bdf ?
23:51:56preglowamiconn: everything disk based just doesn't work, and the only change i have in the file now is not writing those commands to the ata controller, it makes no sense
23:53:34miner49erbigghero: I'm gonna see if if I can scale the asteroids up...what the resolution of the iPod?
23:53:58tucozthank you
23:54:11Paul_The_Nerdminer49er: 176x132, 220x176 and 320x240 for Nano, Photo, and Video I believe
23:54:13bluebrother^I hope I'll finish the buttonmap environment tomorrow.
23:54:25miner49erlol @ bigghero!
23:54:37miner49erI meant biffhero, hee hee...hic
23:54:48tucozbluebrother^, I am looking forward to that.
23:54:57miner49erPaul_the_nerd: Excelllent, cheers dude.
23:55:45amiconnpreglow: What I would try:
23:56:36amiconnCompletely comment out ata_perform_sleep() for nano, and comment out the call to it in the ata thread
23:56:46miner49erPaul_the_nerd: Whilst we're at it, what's the res. on the iRiver? archos is 112*64 Isn't it?
23:56:49amiconnRemoved all other nano adjustments
23:57:03dpassen1iRiver iHP-1x0 = 160*128
23:57:36amiconn(i.e. the one in ata_disk_is_active())
23:58:15 Join arf-arf [0] (
23:58:20Paul_The_Nerdminer49er: I think that's right for Archos. Iriver H300 is the same as iPod Photo, and H100 is as dpassen1 said I think1
23:58:43bluebrother^I also think a \note{} command would be good.
23:58:56tucozwhat does that do?
23:59:01scfwhats is screen resolution of H300?
23:59:01tucozah, I get it

Previous day | Next day