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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2006-03-27

00:00:52 Nick Paprica is now known as Paprica[sleep] (i=Paprica@rockbox/developer/paprica)
00:00:52DBUGEnqueued KICK Paprica[sleep]
00:00:53Paul_The_NerdNow I need to make tagcache only scan my /Music folder.
00:01:11 Join bagawk [0] (n=lee@unaffiliated/bagawk)
00:03:14kclafwhere is the tagcache thing in rb ?
00:03:19kclafis it a plugin or smthing ?
00:03:27kclafcan't find it
00:04:02Paul_The_Nerdkclaf: It's in the playback menu.
00:08:02 Quit tucoz ("Leaving")
00:08:10 Quit damaki_ (Read error: 104 (Connection reset by peer))
00:09:39 Join damaki [0] (
00:10:52linuxstbI've made a start on a TagCache wiki page - please help flush it out. I think it would be useful for Slasheri if we could document the obvious bugs in one place for him.
00:11:03 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
00:11:31kclafnice, gonna check the page
00:11:40Paul_The_Nerdlinuxstb: That first "Ipod-specific" one happens on H120 too, and I think it was reported on H3x0 from back when old tagdb was being used.
00:11:45 Quit Nico_P ("Konversation terminated!")
00:11:54b00st4paul_the_nerd: u intalled doom?
00:12:03linuxstbPaul_The_Nerd: OK, I'll fix. I'm sure the data abort will happen on Archos as well.
00:12:10 Join Nico_P [0] (
00:12:24 Quit Nico_P (Remote closed the connection)
00:12:38Paul_The_Nerdb00st4: I suppose you could say that, yes. :)
00:12:44BagderLinusN: did you notice that we have PLAY and REC swapped in button.h for the x5?
00:12:49b00st4what files do i add to sources?
00:12:55amiconnhej LinusN
00:12:57Paul_The_Nerdb00st4: The .patch should do it for you
00:13:02LinusNBagder: yes, so i've been told
00:13:14BagderLinusN: tree.c assumes rec for the menu... :-)
00:13:16LinusNhi amiconn
00:13:43amiconnLinusN: Do you think it would be a good idea to commit wavplay?
00:14:02LinusNamiconn: yes i do, just so we don't forget it
00:14:18b00st4should work for 0.87 too
00:14:31amiconnImho it's also a nice preview for true pcm playback integration
00:14:39preglowanyone opposed to me adding another option for the piezo speaker click on clickwheel event thing?
00:14:40Paul_The_Nerdb00st4: Are you on iPod or on iRiver?
00:14:56Paul_The_Nerdb00st4: Use the last .diff from Llorean, and the .zip before it.
00:15:02Mikachupreglow: you committed it?
00:15:14preglowMikachu: no, but i'm thinking of getting it in a commitable state
00:15:27preglowand i for one don't want the thing on, so i think an option is required
00:15:34b00st4there is no Llorean :D
00:15:46linuxstbpreglow: No objections. What are you going to make the default? I think it should be enabled to help blind users.
00:15:46b00st4Karl Kurbjun (kkurbjun)?
00:15:47Paul_The_Nerdb00st4: Paul Louden
00:15:49JdGordonwtf? i just updated to the very latest cvs source and still dont have the tagcache options.. ?
00:15:57Paul_The_Nerdb00st4: Is it DarkkOne there then?
00:16:04preglowlinuxstb: yeah, blind people is the only reason i ever implemented it
00:16:09preglowi'll certainly never use it
00:16:15b00st4so not the latest src
00:16:18linuxstbpreglow: Nor will I.
00:16:29 Join Moos [0] (
00:16:34 Join tvelocity [0] (
00:16:35Mikachumy beeper plugin makes my nano shut down without fail within 5 minutes every time
00:16:51preglowMikachu: perhaps all plugins shut down after 5 min?
00:16:54Mikachui think maybe it doesn't like using the cpu too much
00:17:03Mikachuit has happened in pacman in more extended playing too
00:17:06t0masMikachu: what are you trying to implement?
00:17:06Paul_The_Nerdb00st4: The latest ones are H3x0 specific and I don't *think* they've been fixed for iPod yet
00:17:07Mikachubut not anywhere else
00:17:13preglowlinuxstb: but yeah, since we have no current lang files, i was thinking we might actually want to make it default to on for now
00:17:20preglowlinuxstb: current voice files, i mean
00:17:21amiconnLinusN: The good (or bad?) thing is that wavplay increases the probability of the sporadic MMC driver hang on Ondio, probably because it puts heavy load on both serial interfaces
00:17:48t0masMikachu: timer plugin?
00:17:51amiconnMaybe that will help me figure out what's going on...
00:18:08Mikachut0mas: no, what the plugin does is not so important... just that the nano seems to shut down when you use 100% cpu for too long
00:18:09LinusNamiconn: interesting
00:18:24Mikachuapparently i'm the only one who made it past level 3 in pacman on a nano
00:18:24t0masMikachu: yeah, but I am interested in what it does..
00:18:26amiconnOn recorder it works like a charm, apart from the disk spinning up every 10 seconds with 44.1 stereo wavs
00:18:48Paul_The_NerdMikachu: That's odd, since before the CPU speed adjustment, it ran at 100% cpu constantly
00:19:05amiconnLinusN: I already have a suspicion, but no idea how to verify it...
00:19:09Mikachui don't mean having the frequency at 75 mhz, i mean actually using the cpu instead of idling
00:19:11t0masMikachu: ok
00:19:31CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
00:19:31*t0mas goes to bed...
00:19:34t0mas0:19 already
00:19:36t0masstupid summertime
00:19:37Mikachume too
00:19:56amiconnEven then, Ondio runs more stable than anything swcodec atm :/
00:19:58Paul_The_NerdMikachu: Aaah. Well, how do you know it's using 100% of the cpu, instead of just being constantly boosted?
00:20:11b00st4hm, it fails all, but im too tired to fix that now...
00:20:13JdGordonhey LinusN, u gonna commit that patch, or u want me to fix that 1 mistake first?
00:20:17amiconnWtf... tagcache adds a whopping 8KB to the binary? Certainly fishy, since it is meant to replace tagdb...
00:20:28preglowamiconn: has it yet, though?
00:20:34b00st4im out pz, hx for ur help guys, would neva have made this :D
00:20:44b00st4sleep well everybody
00:21:14 Quit b00st4 ()
00:21:21LinusNJdGordon: i think i'll fix it
00:21:56LinusNhave you compiled it on all targets?
00:22:45Bagder"pz, hx for ur"
00:22:51Bagdernow what language is that?
00:23:13Bagderthose kids...
00:23:24LinusNyou're too old
00:23:32Bagderand grumpy!
00:23:36*Bagder shakes fist
00:23:48kclafi wish they had put a hdd led on the ipod
00:24:01kclafcant know when tagcache update is done
00:24:14MikachuPaul_The_Nerd: i don't have the cpu scaling enabled.. and i know pacman uses nearly all cpu power, and i know i'm only yielding without sleeping in my plugin
00:24:32preglowMikachu: have you found some pitch value that sounded the most pleasing? :-)
00:24:41*amiconn likes DST
00:24:44linuxstbkclaf: You should see a "virtual LED" in the status bar (unless you have disabled the status bar...)
00:24:45Mikachuheh, no, but you can use my beeper plugin to find one :P
00:24:54*Bagder votes for DST time zone all year around
00:25:02preglowgmt+1 !
00:25:05Mikachui think we should switch time zones every day
00:25:06linuxstbkclaf: Or go to Info->Debug->Tag Cache
00:25:15Mikachui want australian time tomorrow
00:25:17Paul_The_NerdMikachu: I'm just curious if it's 100% use, or something else. I thought pacman didn't use *all* of it on Nano, since it's throttled to 30fps, and hits more than that without throttling.
00:25:45kclaflinuxstb : just next to the clock, it that it ?
00:25:51MikachuPaul_The_Nerd: well, the only other thing that is common to them is i touch the scroll wheel constantly for 10 minutes
00:26:08Paul_The_NerdMikachu: Well then, it could be a scroll wheel driver problem. So there. :-P Hehehe
00:26:08linuxstbRockbox ran happilly for months at a constant 75MHz. So I can't see how pacbox constantly boosting the CPU can cause any problems.
00:26:22 Quit ender` (" Q: What's another name for the "Intel Inside" sticker they put on Pentiums 4? A: Warning label.")
00:26:33Mikachulinuxstb: surely running real instructions rather than sleeping makes a difference, regardless of the frequency?
00:26:37linuxstbHow does the idle timer work? Does touching the scrollwheel affect it?
00:26:50Mikachuand like i said i disabled the cpu scaling
00:27:12JdGordonu guys have it easy with dst.. we were supposed to come out of it sat night (2 days ago), but no.,. because of the blody commonwealth games we dont get out till next w/e.. so my comp is showing the time an hour behind
00:27:15linuxstbkclaf: Yes, in the top-right corner.
00:27:16JdGordonbloody games!
00:27:24Mikachui have to sleep now.. sorry to bring something up and leave
00:27:27JdGordonanyway, gonn be late to uni.. cyaz
00:27:46kclafok now i'm sure, rb really rocks
00:29:43webguest16JdGordon: That's precious.. so a random cultural event changes when dst is?
00:29:47 Join herz42 [0] (
00:31:31kclafok i think i made it crash
00:31:32 Quit Paul_The_Nerd ("Leaving.")
00:31:42kclafby browsing all songs
00:32:02 Join ashridah [0] (
00:32:43kclafand now data abort at 00016C54 by browsing artists
00:33:28preglowtagcache still has some way to, yes
00:33:45 Join nnod [0] (
00:34:13 Part nnod ("Leaving")
00:34:21kclafyea i know, im just saying the issues i get, if that could help in any way
00:35:02linuxstbkclaf: Do you have tagcache on disk or in RAM?
00:36:32kclafi read your wiki page about ipod issue
00:38:31amiconnBagder: Is there a reason for swapping the internal codes of REC and PLAY for X5? I can't see one...
00:38:40Bagderbecause they were wrong
00:38:56Bagderthe play button sent a REC code
00:39:01Bagderand vice versa
00:39:32preglowanyone thought of a nice way to detect if a file has updated tag data?
00:39:40amiconnI mean this one not the following
00:39:51amiconnThis merely swaps the bits used for REC and PLAY
00:40:23amiconnShouldn't change behaviour at all, unless some place doesn't use these constants
00:40:30Bagderhm, no it was probably just a stupid ifix
00:40:30LinusNBagder: you should have changed the button scanning code instead
00:41:24amiconnLinusN: YOu forgot to bump the api...
00:41:48 Quit charkins ("Leaving")
00:42:12amiconnI'll have to bump it as well for wavplay.
00:42:50LinusNtoo late
00:43:16 Quit bam_ (Remote closed the connection)
00:47:13amiconnBagder: Builds seem to hang again...
00:47:24Bagderthere's a single server doing this to us
00:48:01 Join linuxstb_ [0] (
00:48:03BagderI'm disabling it
00:48:13 Quit linuxstb (Nick collision from services.)
00:48:19 Nick linuxstb_ is now known as linuxstb (
00:48:45preglowsweet god
00:48:47preglowi am such a fool
00:49:18linuxstbIs anyone else having problems with tagcache displaying (and playing) albums in a random order (i.e. ignoring the tracknumber tag)?
00:50:20 Join damaki_ [0] (
00:51:00Bagderthe 8th server was done after 265 seconds
00:51:06Bagderthe ninth after 745
00:51:25kclafi get freezes only by browsing artists
00:52:08midkayit'd be cool to like.. have plugins have their own minimum api requirement.. and so if the api version is, say. 1.6, and a plugin only requires 1.4, then it can safely run rather than 'incompatible version'.. or something.
00:52:27amiconnBagder: Christi's server seems to have path problems. "m68k-elf-gcc: No such file or directory"
00:52:38Bagderyes, I disabled m68k for her server just now
00:56:04preglowoooh, david's fixed the long issues
00:56:05*preglow tests
00:58:16preglowamiconn: now wavpack works on amd64 too
00:58:40preglowamiconn: which makes libmad the only problem left
00:59:31LinusNmidkay: that's exactly how it works today
00:59:33midkayBagder, if you changed the mapping in button.c, don't you need to switch it back for tree.h?
00:59:41midkayLinusN, hmm?
00:59:54Bagdermidkay: no, it was supposed to be switched in both places
01:00:23midkaydoesn't that negate the change, though? swapping them in one place, and then swapping them again elsewhere?
01:00:34midkayah, never mind..
01:00:59amiconnpreglow: nice :)
01:01:11amiconnwavplay committed.
01:01:50preglowbut ok, minipoll: i've added a tick for each clickwheel emitted to make it easier for blind people to recognize when something happened. should this click be emitted as part of the button driver or as part of the list widget?
01:01:55preglowyea or nay
01:01:59amiconnBagder: What caused/causes the holes in the build table?
01:02:10Bagderssh failures
01:02:22Bagderthat isn't properly taken care of by the master script
01:02:32midkaypreglow, it should be disablable, right?.. button driver, i think.. thus global..
01:02:50preglowoption, yes
01:03:10midkayi say global, then, for sure..
01:03:11preglowonly problem with the button driver issue is that the tick sound not always will correlate to movemt
01:03:15preglowlike in games and such
01:03:27*amiconn thinks the build table looks like a swiss cheese today
01:03:48LinusNamiconn: if your cheese is that green, throw it away
01:03:49midkayamiconn, i was looking at the design at the right, with all the reds. it looks like a face with a W mouth..
01:03:52 Join tucoz [0] (
01:04:04midkaypreglow, that's true..
01:04:20tucozlinuxstb: I am not sure, but it looks like it sorts the files correctly, just inversed order
01:04:26midkaybut that'd be .. 'rare'?
01:04:36midkaythere might be a few games with such a problem, but..
01:04:45LinusNpreglow: a global button click/beep would be nice on all platforms
01:05:19LinusNtime to sleep
01:05:23LinusNnite all
01:05:23preglowLinusN: i'm not quite certain i'm up for that
01:05:29preglowi want this commited so i can do other work :>
01:05:31amiconnpreglow: Nice type in wav2vw.c... "chuck header"
01:05:33LinusNpreglow: booooo
01:05:33preglowLinusN: gnight
01:05:40 Part LinusN
01:05:53 Quit damaki (Read error: 110 (Connection timed out))
01:06:54preglowthe problem i have with the ticking on the ipod is: 1) i want each tick to be uniform length, say 10ms 2) the asynchronous manner of the clickwheel handler makes it very hard to achieve without using timer.c
01:06:58amiconnWas there before, I think
01:07:47amiconnpreglow: Yeah, but using timer.c means pretty much giving up backlight fading, right?
01:07:51preglowamiconn: exactly
01:07:57amiconn(at least fade in)
01:08:00preglowamiconn: which shouldn't be a major point for blind people, granted...
01:08:11preglowamiconn: but other people might want to use it too
01:08:17 Join bam_ [0] (
01:08:29 Part tucoz
01:09:16preglowamiconn: i can start a tick sound from button_tick, but the scroll events never go through that
01:09:24amiconnpreglow: Perhaps do both piezo-on and piezo-off in a tick task, and only set a flag in the wheel handler?
01:09:26preglowhmm, perhaps i can set a flag
01:09:39preglowheh, yes, we seem to agree
01:09:49amiconnWhy do scroll events not go through button_tick?
01:10:16*amiconn didn't look very closely at iPod button handling
01:10:19preglowmy initial button handler just didn't do it for some reason
01:10:25amiconnThat'll probably change soonish
01:10:25preglowand it's been kept that way ever since
01:10:38preglowbut yeah
01:10:48preglowbutton dispatches happens ever, what, 0.1 secs?
01:11:09***Saving seen data "./dancer.seen"
01:11:12preglowis that often enough for the click wheel?
01:11:16preglowdid you win the ipod?
01:11:51amiconnButton dispatches happen every tick, i.e. 10 ms
01:12:53preglowoh, right
01:12:58amiconnpreglow: This one:
01:13:00preglowi thought i saw some code to handle it less often
01:13:39*linuxstb remembers amiconn saying he will never buy an ipod :)
01:13:52preglowamiconn: goodie, i expect you'll have rockbox running on it in a jiffy :)
01:14:00preglowwill be very fun to see how the codecs perform on those things
01:14:30amiconnlinuxstb: Yeah, true, I said that. It's still valid for the ones with plastic cases
01:14:54amiconnpreglow: The mini has PP5020 (2G: PP5022) so I don't expect surprises
01:15:03preglowso i see
01:15:13preglowwell, that's good
01:15:21preglowit should perform very well, then, what with the b&w small screen
01:15:31amiconnI want to hack the grayscale lib...
01:15:43linuxstbAny idea if it's a 1st or 2nd gen mini yet?
01:15:49amiconnNot yet
01:17:49linuxstbIt's possible we may need separate builds for the two minis - I think there are differences in the button drivers, and the PP5022 has 128KB of IRAM, compared to 96KB in the PP5020.
01:18:26preglowamiconn: so the POLL_FREQUENCY code is just in case the tick freq changes?
01:19:13linuxstbIt appears the PP5022 can be clocked at up to 100MHz according to the product brief...
01:19:26preglowlinuxstb: it can, but then we need to reinit some hardware,i think
01:19:27amiconnYes. It's also a leftover from back when we didn't poll that often
01:19:30preglowlinuxstb: i tried it
01:19:49linuxstbYou have a PP5021 though...
01:20:16preglowthat can be clocked higher than 75mhz, too, no?
01:21:26 Join Guest1221 [0] (
01:21:36preglowok, the flag idea worked just peach, so that'll stay
01:21:37Guest1221Quick question?
01:21:42preglowGuest1221: shoot
01:21:47Guest1221What is Rockbox
01:21:59Guest1221it says open source something
01:22:02Guest1221i don't get what that means
01:22:08Guest1221im on that site atm
01:22:23preglowopen source means the source code is open for everyone to see
01:22:31pregloweveryone can help in developing it
01:22:45linuxstbIt's a replacement for the software (firmware) that runs inside various mp3 players.
01:24:12Guest1221can i change it back
01:24:30Guest1221like if i d/l to my Iriver if i don't want it can i put in the old firmware?
01:24:48preglowyou can use both
01:24:50linuxstbBut you can have both installed at the same time - you choose which one to use when you turn on your player.
01:25:27linuxstbWhich player do you have?
01:25:33Guest1221Iriver H340
01:25:53midkayGuest1221, you might try clicking the "Why choose rockbox" link on the front page.. for a list of features specific to your model. that's why i made it; that's why it's there..
01:26:46preglowwhere in the menu system can my click option go, btw?
01:27:01Guest1221like for firmware?
01:27:01midkaypreglow, i'd say "system" fits best..
01:27:15Guest1221It like settings| firmware upgrade|yes
01:27:16 Quit ashridah ("uni")
01:28:24Guest1221either settings or general
01:29:01preglowsometimes the cursor lags a bit when i enter a menu and use the clickwheel
01:29:04preglowanyone else noticed it+
01:29:32midkaypreglow, you mean navigation with the wheel being slow?
01:29:41preglownot slow
01:29:53preglowwhen i first scroll a bit, the cursor just lags for a second
01:29:53linuxstbI think I've noticed that - the clickwheel not immediately responding?
01:29:57preglowlinuxstb: yes
01:30:05preglowit's really annoying with the piezo patch
01:30:12preglowsince you hear when it's supposed to move
01:30:14midkayi haven't noticed that, but.. the slowness of the wheel itself is so annoying..
01:30:18Guest1221thnx guys
01:30:22 Quit Guest1221 ("Later")
01:30:24midkayi always have to boost to 75mhz to make it usable
01:31:02linuxstbthere will be even more complaints about the wheel when people start using tagcache
01:31:30preglowbrowsing long lists?
01:31:40preglowyeah, indeed
01:31:44amiconnWow, wavplay.c takes the lead in plugin source file size.
01:31:48preglowdidn't lostlogic have some acceleration magic going, btw?
01:32:00amiconnThis is cheating though, most of it is just 10 giant arrays
01:32:48bam_where do I get an install for m68k-elf-gcc?
01:33:01bam_im in kubuntu
01:33:39preglowno install, you get to build it yourself!
01:33:43preglowand what a grand adventure it will be
01:34:10 Quit bam_ (Remote closed the connection)
01:36:26preglowpretty, pretty please
01:36:31preglowsomeone explain me how this iram can be so slow
01:40:35midkaybecause,, it is.
01:40:43sharpei've an idea!
01:40:59midkaysharpe, get back to work. i didn't say you could stop. :(
01:41:00preglowif we somehow aren't exploiting the iram properly in some fancy way
01:41:09preglowby god, how things speed up then
01:41:59sharpemidkay, it basically works, it just doesn't have the normal zip handler in it...
01:42:10midkaysharpe, what?
01:42:22sharpethe wpsloader
01:43:31sharpebut i had another idea for something, that may be useful, or useless.
01:44:04sharpea cpu emulator!
01:45:54 Join oftenwinter [0] (i=oftenwin@
01:46:03 Join didj [0] (
01:46:35sharpewell, it at least sounds interesting
01:47:35 Join bam_ [0] (
01:48:04bam_anyone got the crosscompiler setup on debian based distro?
01:49:36midkaysharpe, wtf? :)
01:50:39 Quit Moos ("Glory to Rockbox !!!")
01:50:39sharpei was just thinking about stuff, and i thought about a cpu emulator, then thought, 'hey, why not make one for rockbox?'
01:51:03 Join Aditya|Shower [0] (
01:52:12linuxstbbam_: Yes - those CrossCompiler instructions work fine.
01:55:56bam_which version of binutil's
01:57:26midkaysharpe, sounds great, do it. :)
01:58:50sharpebut what cpu?
01:59:02midkayhmm. an athlon 64 3700+!!
01:59:15linuxstbbam_: 2.16 is fine (as the wiki page says).
01:59:23sharpelet me rephrase that, what cpu architecture? :)
02:01:32sharpethis? ?
02:01:47sharpeer well, the processor it uses anyway
02:03:00 Join damaki__ [0] (
02:03:47midkayjust kidding.
02:03:50midkaythat was a random number.
02:03:53midkayand you went looking for it.
02:04:03midkay*giggles to self* that's funny to me.
02:04:12sharpeand it's actually a processor
02:04:56midkayhyaaaaa. hahahahayhyahyahaha.
02:05:07oftenwinterdoes anyone know if it's possible to map the analog audio output to the line-in or digi-out jacks?
02:06:01sharpewhat about the 8086? :D
02:07:05sharpeor the 4004...
02:09:28oftenwinterany ideas on the plausibilty of it at least?
02:12:57 Quit obo (Read error: 110 (Connection timed out))
02:13:10 Quit muesli__ (Read error: 110 (Connection timed out))
02:14:26oftenwinteri guess not
02:14:30 Join jd|uni [0] (
02:15:21bam_configure: error: no acceptable cc found in $PATH
02:15:21bam_<−−any ideas?
02:15:39 Quit Rick (Read error: 104 (Connection reset by peer))
02:15:47jd|unibam_: u got the correct version of gcc installed and setup?
02:16:00jd|unifor the target?
02:16:07bam_but I shall re-install...
02:16:12sharpehey midkay, what about the 6502? :)
02:16:26jd|uniu prob just forgot to put it in your path
02:16:34midkaysharpe, what about screw it and work on the wpsloader! :((((
02:16:42 Quit didj (Read error: 110 (Connection timed out))
02:17:06sharpeit's just an idea, not going to start on it for a while...
02:18:08 Join Rick [0] (
02:18:41 Quit damaki_ (Read error: 110 (Connection timed out))
02:18:53midkaysharpe, don't make me whip you.
02:19:06sharpewow... that is quite possibly the most stupid toy i've seen.
02:19:14sharpe'Thumb Warriors!'
02:22:29midkayhaha. skannerz..
02:22:31 Join RoC_MM [0] (
02:22:37midkay"innovative technologies".
02:22:43midkaywhat a load of beeeeee essssssssssssssssss.
02:25:13 Quit dpro (Read error: 110 (Connection timed out))
02:25:30jd|unimm.. nice.. im in the comp lab at uni and this big spider is rcawling around the desk... and i tihnk its got a white mark on its back.. means its pretty dangerous!
02:25:33midkaysharpe, so what's the status?
02:25:42sharpeof what?
02:25:44midkayjd|uni, pet it and touch it and hold it.
02:25:57midkaysharpe, you know what.
02:25:58sharpeand be sure to play with it
02:26:26midkaymolest it.
02:26:31sharpeand if you want to be friendly with it, cup it in your hands and shake it around.
02:26:41sharpehow do you molest a spider?
02:26:50midkayrape it with a piece of dust?
02:27:47bam_whats this mean?
02:27:48jd|unihaha wierdso
02:27:50bam_checking whether the C compiler (gcc -no-cpp-precomp ) works... no
02:27:50bam_configure: error: installation or configuration problem: C compiler cannot create executables.
02:28:11bam_how do I correct it
02:28:15jd|unidid u not do −−enable-languages=c ?
02:28:28sharpeyou know midkay, if we had the 6502 emulator for rockbox, we could do an atari emulator :)
02:28:32bam_errr, no and how?
02:28:54midkaysharpe, if we had the athlong 64 fx-60 emulator we could do doom3.
02:28:59midkayathlong. haha. oops.
02:29:06midkayha. that's funny. because it's almost like a sexual reference.
02:29:12jd|unibam_: u following the crossconmpiler wiki page corectly? u need that when u run configure for gcc
02:29:14midkayschlong. athlong.
02:29:28bam_lemme look again...hummm
02:29:29sharpeprocessor oriented sexual innuendo
02:30:25midkaysharpe, you want my "intel inside" you?
02:32:22sharpei hate you.
02:33:47midkayi have an athlong that makes cool fireworks-like FX when you do stuff with it!
02:36:07jd|uniso anyway.. has any1 else not got the options for tagcache even after updating to the latest cvs?
02:39:47 Quit bam_ (Remote closed the connection)
02:41:50 Join merbanan [0] (
02:52:48 Quit jd|uni ("CGI:IRC (EOF)")
02:54:46 Quit Kohlrabi (Nick collision from services.)
02:54:54 Join Kohlrabi [0] (
02:55:02 Quit oftenwinter ()
02:58:48 Quit [TCK] (Read error: 110 (Connection timed out))
03:01:57 Quit midkay ("Leaving")
03:02:52 Join merbanan_ [0] (
03:05:51 Join crixu [0] (n=crixu@
03:06:37crixuhello...somebocy here, that can tell me if program listed on work for an acer laptop ??
03:06:38 Quit merbanan (Read error: 104 (Connection reset by peer))
03:07:06 Quit webmind (Remote closed the connection)
03:07:11 Join webmind [0] (
03:11:10***Saving seen data "./dancer.seen"
03:14:42sharpewhy do i want to bake a cake that is set on fire?
03:17:34 Quit mog_work (Read error: 113 (No route to host))
03:18:37 Quit crixu ()
03:20:44 Join actionshrimp [0] (
03:23:07 Quit herz42 ("Und wech")
03:34:03 Quit bagawk (Read error: 110 (Connection timed out))
03:35:42 Join bagawk [0] (
03:51:57 Join mog_work [0] (
03:57:37 Join midkay [0] (n=midkay@rockbox/developer/midkay)
03:58:58 Quit tvelocity ("Ex-Chat")
04:21:32 Quit actionshrimp ("( :: NoNameScript 4.01 :: )")
04:25:45 Quit midkay (Read error: 104 (Connection reset by peer))
04:25:58 Join midkay [0] (n=midkay@rockbox/developer/midkay)
04:40:53 Join TCK [0] (
04:52:14 Quit Kohlrabi (Read error: 104 (Connection reset by peer))
04:53:11 Join biffhero [0] (n=rob@
04:55:41mikearthurAMP: + LAPTOP = :(
04:58:11biffheroouch, the amp fell on the laptop?
05:01:19 Join station [0] (
05:01:29mikearthur30kg amp
05:03:19 Quit Daishi ("Client exiting...")
05:06:56biffherohow many "pound" ($$) laptop?
05:09:53mikearthursecond hand like 2 years ago
05:10:04mikearthurif it was a nice one, I'd be raging
05:10:36 Join Paul_The_Nerd [0] (
05:11:11***Saving seen data "./dancer.seen"
05:12:59mikearthurneed a new laptop though now :S
05:13:12biffherohello paul. this is the second night in a row that you have logged in about 15 minutes after me. are you following me? :-)
05:13:21biffheroyes, can you get a replacement screen?
05:15:31Paul_The_NerdUsually I'm on much earlier, but I've been busy this weekend. Heh
05:21:16 Join Sinbios [0] (
05:21:45 Nick Aditya|Shower is now known as Aditya (
05:24:47 Quit webguest16 ("CGI:IRC (EOF)")
05:25:11 Join bam_ [0] (
05:27:17 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
05:28:25biffheroare themes supposed to jack with the button layouts?
05:29:32Paul_The_Nerdbiffhero: They can't really change the button layou.
05:30:52Paul_The_Nerdlostlogic: Oh, you're here. Yeah, the shuffle freeze at end of playlist thing has been around for a decent while. Occurs when the "Repeat" mode is "Shuffle", doesn't seem to matter whether normal shuffle is on or off. And I haven't verified it's still here after your recent changes, but seeing that thing in the mailinglist reminded me of it
05:31:28biffheroPaul_The_Nerd: thanks, I just downloaded a big .zip file of .wps-es, and I find that when I load certain themes, my menu press (ipod 5g) doesn't take me back out to the menu anymore
05:31:35lostlogicPaul_The_Nerd: that's what I thought... not that I don't want to fix that bug too, but it's important to know what changes cause what bugs, and gl wasn't being very helpful in separating them out.
05:32:02biffheroalso, a lot of the WPS have the text 'background.bmp|' on the screen. or 'back.bmp|'
05:32:15lostlogicbiffhero: sounds like you have mismatched wpss and build
05:32:25Paul_The_Nerdlostlogic: Yeah, I know. The one I described is *very* reproduceable, and has been in for a few months at least. And it happens on iPods and iRivers.
05:32:47biffherolostlogic: oh yeah, good point. maybe the wps parser has changed, you think?
05:32:52lostlogicPaul_The_Nerd: that's what I thought, I hope gl's is the same (I'm pretty sure based on his meandering description)
05:32:56lostlogicbiffhero: no
05:33:09lostlogicbiffhero: some WPSs are designed for the experimental builds which are not official rockbox builds
05:33:14Paul_The_Nerdbiffhero: If the backgrounds are a different reso from your screen, then you'll get that
05:33:16lostlogicbiffhero: those will not work correctly in an official build
05:33:33lostlogicor maybe I'm talking out my ass and paul is right.
05:33:37*lostlogic shuts up.
05:33:56Paul_The_Nerdlostlogic: There's not a WPS pack for 5G I think, so I'm betting he got the H300 ones
05:34:09lostlogichmph, nod
05:34:12Paul_The_Nerdbiffhero: Were they listed as 5G WPSes or just general 24-bit ones?
05:34:58biffherogeneral 24-bit, from "Ultimate 24Bit WPS Collection" thread. some say specifically i-this, or i-that, but most are probably for the h300
05:35:54Paul_The_Nerdbiffhero: The "Ultimate...collection" is for the H300 yes. Any of them with backdrops shouldn't work. As well, some may have unsupported tags in them.
05:36:17biffheroalso, the thread was in the "for H3xx Series" forum (there isn't a forum for iPods), so that is probably my problem.
05:36:26Paul_The_NerdI would guess so. :)
05:36:51biffheroas for what build I am running, it is a cvs build060324-2233 + scrolling patch
05:37:41Paul_The_NerdIf you strip the background images out, and any of the WPS %tokens that aren't supported by your build, they should work. :-P
05:39:52biffheroI ofund a scrolling patch ...
05:40:17biffherothat was needed for some of the 5G wpss on
05:40:34dpassen1I use that patch, as well
05:41:29Paul_The_NerdYeah, the scroll margin patch is pretty nice, but not the way Rockbox wants to do that in the end.
05:41:32biffheroit seems as if rockbox was not architected with multiple resolutions / bit depths in mind. Not a slam, just an observation. I am sure that the devs will make it right. The movement on this project seems way faster than the one for ipodlinux
05:41:48Paul_The_Nerdbiffhero: Why do you say that?
05:41:55 Join Arrogant [0] (
05:42:08lostlogicbiffhero: it's the wpss that aren't ... and can't be setup for many resolutions in one file because they are pixel specific screen layouts
05:42:08Paul_The_Nerdbiffhero: The *only* restriction for resolutions is on the backdrop image being the same as the screen size.
05:42:40lostlogicbiffhero: in a way you are right though, rockbox was originally created for the archos player with a 2 line character cell display ;)\
05:42:42biffheroyes, exactly my point. the pixel specific screen layouts will be eliminated in the future, no?
05:42:53lostlogicbiffhero: no, how would you propose to do that?!
05:43:01biffherohrmm... good question
05:43:17biffherolet's take a different problem, one that I probably understand better. :-)
05:43:25biffheroon the retailOS for the iPod
05:43:26lostlogicscreen layout is very screen size specific, there's no real way to genericize that any more than e can make generic the drivers for hardware that is only in one model ;)
05:43:28Paul_The_Nerdbiffhero: They could only be eliminated if the images were vector based and could be scaled easily to varying resolution sizes
05:43:59Paul_The_NerdThat's IT! XML WPSes using SVG for the images!
05:44:04biffherothe .jpg file that I put in the image gets resized based upon what I am doing with the ipod. the .bmp loader that I saw in one patch was only for 100x100 .bmp files
05:44:11*lostlogic beats paul with .svg
05:44:26biffherothere is nothing wrong with resizing images on the fly, is there?
05:44:42lostlogicwhat .jpg file −− that you view wiht the image viewer?
05:44:54biffheroof course, that's for album art, not for the actual wps decorations....
05:45:01Paul_The_Nerdbiffhero: Well, that requires more CPU cycles than just loading a properly scaled one. Plus, do you make the images for the smallest screen and size up so no data is lost, or the largest and size down?
05:45:22 Quit amiconn (Read error: 110 (Connection timed out))
05:45:22 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
05:45:23lostlogicand waste space and ram and clock cycles onthe smaller screen and likely less pweorful players
05:45:24biffherolostlogic: no, that's in the retailOS, where the jokers don't have to support that many platforms.
05:45:28Paul_The_Nerdbiffhero: Also, what about screens that are taller than they are wide, instead of wider than they are tall? Different aspect ratios or pixel aspects mean different needs
05:45:38biffherothat's true.
05:46:13Paul_The_Nerdbiffhero: In the end, it's much easier to just have WPS_220x176x16 and WPS_320x240x16 versions and so on
05:46:17lostlogicalbumart is at this point in its infancy, who knows where it'll end up ;)
05:46:26biffherothis could also be done on the way in, with a front-end "loader" app, which would know the target, their screen sizes, and then do the appropriate resizing on the load, to save space
05:46:31Paul_The_NerdYou know exactly what type of screen the WPS will work on, but nobody uses proper naming
05:46:50biffheromaybe the naming is all that is borked?
05:46:55lostlogicat devcon, we discussed creating a dir structure so taht only WPSs that would fit the screen of the target would be built and included
05:47:01biffherogood one
05:47:01Paul_The_Nerdbiffhero: No, it's the fact that the creators don't bother to name it.
05:47:34Paul_The_Nerdbiffhero: In the CVS, there are many sizes of several of the WPSes, and the program that makes the .zip only uses the largest size that's smaller or equal to the screen
05:47:48biffherothat's wise
05:47:58biffherobut the theme loader could know what resolution the screen is, and then look at the files in the .wps at load time, and refuse to load the wrong one
05:48:30biffheroI guess it is 'ok' to load too small ones, and not load too large ones
05:48:39Paul_The_Nerdbiffhero: But a smaller .wps could work just fine. As well, a text-only WPS can work on ANY screen size, but look best at 320*240 in one font, and 160*120 in another font
05:49:00biffheroyes, the fonts on this 5G are often too small
05:49:23Paul_The_Nerdbiffhero: Yeah. But really, it's the WPS creator's responsibility to tell the user what screens it's intended for.
05:50:07biffheroor, could it include .bmp files of multiple sizes? (header bar at the top of the screen, for instance)
05:50:33Paul_The_NerdWell the header bar is part of rockbox, and that's something entirely
05:51:04biffheroI meant the header image bar that often is too small width-wise, not the super-teeny text one
05:51:05lostlogiccan disable the status bar and have that information entirely in the WPS now, afaik
05:51:21biffherothat status bar doesn't show up in most themes for me.
05:51:23Paul_The_NerdBut, each image would need different pixel coordinates depending on the screen size, as well as different sized images. It's really best to have MyTheme_160x120x2.wps and MyTheme_176x132x16.wps MyTheme_220x176x16.wps
05:51:34biffheroyeah, ok
05:52:01Paul_The_NerdYou either have to make *everything* proportional positioning, and use some form of vector graphics, or stick raster and just deal with the limitations
05:52:24biffherobackground images are pretty cool, can a WPS change the image on song change?
05:52:35lostlogicsome day (never) some one (who?) could make a vector WPS format that was build time parsed (no)
05:53:35 Join Guest2273503 [0] (
05:53:36 Quit Sinbios (Nick collision from services.)
05:53:41Paul_The_Nerdbiffhero: A wps cannot at this time do that. One could potentially offer the choice for album art to be the backdrop instead
05:54:03 Nick Guest2273503 is now known as Sinbios (
05:54:09Paul_The_Nerdlostlogic: Yeah, I don't think vector based WPSes would be good in *any* way other than eyecandy. :-P
05:55:50biffherothis really is pretty amazing software. has anyone heard anything about the actual hardware makers abandoning their internal firmware development for the use of rockbox?
05:56:21lostlogichah, not likely
05:56:36Paul_The_NerdEither way, the iRivers at least were basically abandoned before Rockbox adopted them
05:56:53biffherohardware, or software, or both?
05:57:12Paul_The_NerdWell, there was ONE software update semi-recently
05:57:22Paul_The_NerdBut the H1x0 and 3x0 lines were both discontinued I believe. I know the 1x0 was
06:01:38 Join bam__ [0] (
06:01:54 Quit bam_ (Success)
06:03:30 Join Rob2222_ [0] (
06:10:18 Join gtkspert [0] (
06:13:41biffherough, where is the setting for the select bar to be an inverse bar or a triangle pointer?
06:14:05biffheroidjit, fount it
06:14:12bam__general settings display...I think..
06:15:30biffherogeneral settings, display, lcd settings, line display
06:19:20RoC_MMis there a way to put the scroll bar on the right instead of the left?
06:20:22 Quit Rob2222 (Read error: 110 (Connection timed out))
06:22:15Paul_The_NerdRoC_MM: No.
06:22:24Paul_The_NerdRoC_MM: Well, technically yes. But no "easy" way
06:26:48Paul_The_NerdThe problem is that this is open source, so in nearly all cases, "is there a way to do X" is "Yes" it's just several of those times "Yes" is "If you code it to."
06:27:23RoC_MMsounds familiar
06:27:52Paul_The_NerdI've perhaps said it many times here and in the forums
06:27:55Paul_The_NerdI'm very literal
06:30:09JdGordonany1 play bubbles on the ipod?
06:31:10Paul_The_NerdI do
06:31:44JdGordonis there a spare button on the start screen?
06:31:48JdGordonor button combo
06:32:13JdGordon.. and room for 1 more line of instructions?
06:32:14Paul_The_NerdSELECT|LEFT or SELECT|RIGHT could work. *maybe* SELECT|PLAY but I'm not sure
06:32:49Paul_The_NerdUnfortunately, my Nano isn't at hand so I can't check the screen right now.
06:34:13JdGordonmenu|select looks free.. is it a pain to do that combo?
06:34:29Paul_The_NerdAnything | Select is fine
06:34:43Paul_The_NerdBut menu|select is *most* often used for Exit, so that would be unexpected for anything else
06:35:16JdGordonye, but exit is menu|_rel so...?
06:35:43JdGordoni tihnk i might leave that for the moment..
06:36:20Paul_The_NerdI'm just saying menu|select is not intuitive for doing anything on the ipod, because in many other plugins it's exit. I have no real objection to it, but other people might be confused
06:43:32midkaymenu|select is pretty much "the" quit button when menu isn't.
06:43:47midkayi can't think of it serving any other purpose anywhere in rockbox, for that matter..
06:43:49midkayi'd go with something else.
06:44:40 Quit Vertigo_t (Read error: 110 (Connection timed out))
06:44:41JdGordonhow do u do the key stuff when u want a diff action for short and longpress? have _pre for the button, then |_repeat for long and |_rel for short press... but how do u get it to not trigger the _rel action after longpress?
06:45:19JdGordon.. or is _repeat|_release both used on the release after longpress?
06:45:20midkayi never needed _pre for anything (not even sure what it's for)..
06:45:33JdGordonPaul_The_Nerd: ?
06:45:35midkayi believe _repeat "cancels" the _rel.
06:45:47Paul_The_NerdI don't know.
06:45:58midkaycan't be sure, though - but i've never had any repeat/rel problems just using them like that..
06:46:09JdGordonok, ill play
06:46:36 Join Vertigo_t [0] (
06:46:53lostlogic_pre is in order to distinguish between a 2 button combo being released (first one button then the other is released) and just one of the buttons being pressed and released
06:47:10lostlogicI believe that only _rel is set on a long release, but I can't swear to it
06:47:26midkaylostlogic, you mean it only triggers after you hold and release?
06:47:26JdGordonis _pre needed to check for long press?
06:47:38lostlogicJdGordon: shouldn't be
06:47:44lostlogicmidkay: right
06:47:45midkayi think it's the opposite. at the least, it definitely triggers even on a short press/release...
06:48:03lostlogicmidkay: _rel means teh button was released
06:48:09lostlogicno matter what the preceeding situation was
06:48:09midkaylostlogic, right.
06:48:22midkayand you say you think it only triggers after a long press.
06:48:48midkayi'm saying i think it only triggers after a short press. i know it definitely triggers after a short press, i'm just not sure whether or not it does after a long one.
06:49:24lostlogicI wasn't claiming only after a long press
06:49:42lostlogicI believe it triggers on _any_ release of the button long, short, or combo
06:49:48midkayah, alright
06:53:47JdGordonhas any1 actually played chesbox?? the comp takes way too long for moves :p
06:53:56 Join Iguana [0] (
06:54:41JdGordonhaha.. chessbox doesnt redraw after every button press.. which means loading the playback menu screws the whoel thing up :D
06:56:55Iguanarockbox crashed my iriver
06:57:56RoC_MMto latest?
06:58:11Iguanai did?
06:58:22RoC_MMtry a build from two or three days ago
06:58:25IguanaI never used rockbox in my whole life and i did the whole thing
06:58:46RoC_MMrockbox boots and shows the menu right?
06:58:47Iguanathen it frose on 2.19 P.M
06:58:54Iguanait did all tha
06:58:59Iguanathen something happened
06:59:10Iguanaand there was a screwed up battery and all these other little icons
06:59:11RoC_MM ...something horrible
06:59:21IguanaI was in TEARS :(
06:59:22RoC_MMI'd try a build from a few days ago
06:59:23 Join Paul_The_Ner1 [0] (
06:59:29Iguanaim not trying it at all
06:59:32biffherodon't get it wet!
06:59:41Iguanaim scared the whole thing will die
06:59:53Iguanait kept all my files though im happy for that ^_^
07:00:25 Part biffhero
07:00:42RoC_MMare you trying the 2.5 release or a CVS build?
07:01:01RoC_MMI don't think there's any reason to think you'll lose your files
07:01:03 Join biffhero [0] (n=rob@
07:01:11IguanaWell i had to reset the whole thing
07:01:14Iguanacause it was crashed
07:01:18Iguanaand frozen
07:01:23Iguanaand none of the other buttons worked
07:01:52Iguanaand i honestly have no idea what release or CVS build i was using o.o
07:03:22 Quit mikearthur (Remote closed the connection)
07:03:30RoC_MMit freezing is not all that bad
07:03:36RoC_MMjust a bug
07:03:42 Join B4gder [0] (
07:03:49RoC_MMdo you still have the zip file you downloaded?
07:03:51Iguanayeh but
07:03:57IguanaI deleted the whole thing
07:04:04RoC_MMthat's fine
07:04:20ArrogantRockbox keeps loading even when I hold down menu
07:04:21ArrogantWhat's going on
07:04:38IguanaNa i was a bit scared cause my iriver had never crashed before in its life
07:04:40RoC_MMon this page, did you choose from the top of the page where it says 2.5 or the bottom?
07:05:03RoC_MMCVS=Daily build...not an official release, it's a "state of the code, no guaruntees"
07:05:13IguanaDaily Builds
07:05:19IguanaLatese H300
07:05:24Paul_The_Ner1Arrogant: You're not pressing menu early enough, probably. Tap menu to turn on the ipod, then immediately press and hold it right after.
07:05:52ArrogantI think maybe the battery is dead.
07:06:01ArrogantThat might not help.
07:06:21Paul_The_Ner1"dead" as in "needs charging" or "broken"?
07:06:37RoC_MMthere may have been a regression Iguana ...a regression is where something that worked before was broken by a new change to the code...I suggest trying a build from 2 days ago:
07:06:45 Quit Paul_The_Nerd (Nick collision from services.)
07:06:49 Nick Paul_The_Ner1 is now known as Paul_The_Nerd (
07:07:04 Quit station ("bye")
07:07:07Iguanaerr im not going to try anything for a while >.>
07:07:17RoC_MMthere is no risk
07:07:32RoC_MMand you just need to download that and extract that to your iriver.
07:07:33Iguanayou can say that
07:07:40RoC_MMyou've already installed the boot loader, no more commands to issue
07:07:50Arrogantneeds charging
07:07:50IguanaI deleted boot loader as welll
07:07:59IguanaI deleted the WHOLE thing lol
07:08:14Iguanaafter i reset it it workded well
07:08:19Paul_The_NerdArrogant: Ah. Yeah, well it's likely that Rockbox will boot, or at least start booting, with less power than the Apple OS will
07:08:25Iguanabut i didn't know whether it would do something funny so i deleted it
07:08:35RoC_MMLike it says:
07:08:41RoC_MMThese are automated daily builds of the code in CVS. They contain all the latest features. They may also contain bugs and/or undocumented changes... The top line is the latest. This build is for iriver H300.
07:08:46Paul_The_NerdIguana: deleted the file, or reflashed to an un-patched .hex?
07:08:54ArrogantIt popped off after a few seconds. I'm gonna get a bit more power before I mess with it.
07:08:56RoC_MMhe reverted
07:09:09Paul_The_NerdArrogant: Good plan
07:09:33Paul_The_NerdRoC_MM: Well, I noticed he kept saying he "deleted" everything, and I know some people think that deleting the patched .hex, if they kept it around, would remove the bootloader.
07:09:33Iguanaerr second one or first one
07:09:54Iguanathe original one
07:10:03Iguanawith out the firmware software thingo
07:10:16Paul_The_NerdThere's no real reason to ever remove the bootloader unless you're about to send it in for warranty repairs.
07:10:53Iguanalets just say the small writing when i first loaded the bootloader freaked me out a bit
07:11:07Iguanaand i couldn't play my videos either
07:11:12RoC_MMbootloaders output text
07:11:14Iguanathey weren't showing on it
07:11:17***Saving seen data "./dancer.seen"
07:11:21RoC_MMfont size is usually small so it can fit a lot on the screen
07:11:28Paul_The_NerdWell Rockbox doesn't support video, so you have to boot into iRiver to use video
07:11:45Paul_The_NerdIt seems odd to be trying an experimental firmware if something as minor as extra text freaks you out
07:12:00Paul_The_NerdRockbox is an entirely alternative firmware, it's not just a patch on the original firmware or anything.
07:12:11IguanaI was experimenting all last night with themes and mods and all
07:12:27RoC_MMI'm rather intrepid but the bootloader flashing text that I didn't understand faster than I could read did make me uneasy the first time I saw it too
07:13:33Paul_The_NerdRoC_MM: Well it is made pretty clear that it doesn't have a "release" version yet.
07:14:19RoC_MMdidn't stop me from installing it on my ipod
07:20:21 Quit imphasing (Read error: 110 (Connection timed out))
07:22:12 Part Paul_The_Nerd
07:23:07 Nick Arrogant is now known as arrogant[buffy] (
07:28:02 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
07:28:04 Part LinusN
07:28:27 Join LinusN [0] (
07:31:02 Join webguest67 [0] (
07:31:37webguest67anyone running rockbox on a ipod video?
07:32:47lostlogicit's off at the moment...
07:33:02biffheroRoC_MM: I _like_ that attitude. That's why I have rockbox on _my_ 5G 60G Video
07:33:34webguest67well, i posted already in the forum, but does anyone else hear a low constant clicking after turning the unit on in rockbox?
07:33:41biffheroI don't
07:33:52webguest67what headphones are you using?
07:33:55lostlogicwebguest67: does it go away once you start playback?
07:33:59biffheroAmerican Airlines
07:34:10webguest67lostlogic: no it does not
07:34:20webguest67but the playback masks it
07:34:21lostlogicwebguest67: I don't believe I do −− there's background noise _until_ the audio codec is initialized for me.
07:34:25lostlogicShure E3
07:34:56Bg3rmorning :)
07:35:05biffherowait one, I don't do a lot of real work in the rockbox, I keep going back to retailOS
07:35:06webguest67I use Sennheiser HD280pro closed cans in a very quiet environment
07:35:08lostlogicdamnit, if you're awake that means I'm up too late.
07:35:20webguest67using the stock ipod buds its pretty much impossible for me to hear
07:35:31lostlogicwebguest67: and the ipod with retailos does not make this noise?
07:35:32B4gderlostlogic: hey, its euro hour coming up here, go to bed! ;-)
07:35:33biffherodon't hear it in my AA ones
07:35:47biffhero2135PST, lostlogic, where are you?
07:36:04 Quit darkless (Client Quit)
07:36:18webguest67but i can still hear the scroll wheel movement with the buds when i turn the db level up past -5db
07:36:18webguest67not that i ever listen to music that loud
07:36:36biffheronope, don't hear it. turning volume up
07:36:41webguest67but the stock firmware does not make any sound
07:36:44*lostlogic fetches his IPV to investigate
07:36:53biffherois -5db the same as 100% volume?
07:37:04webguest67i belive it goes up to 5 or 6 db
07:37:22B4gderthe top volume differs between targets
07:37:22RoC_MM6db is 100%
07:37:36RoC_MMbut usually on the ipods you don't go higher than -20
07:37:40RoC_MMor so
07:37:43RoC_MM-15 maybe
07:37:43biffheromoving to another skin, to tell me the actual db setting
07:37:58RoC_MMdb depends on the level of the music in the song
07:38:01lostlogichah, I never listen above -45 unless I'm in a particularly pissy mood
07:38:23JdGordon-45?? thats way to quiet.. who can u hear that low?
07:38:34RoC_MMhe may be using replaygain
07:38:35webguest67i need to turn up the volume to around -20 / -18db to drive my HD280s
07:38:55lostlogicI do use replaygain, but it reduces the volume on all tracks
07:39:02B4gderyou guys listen to MUSIC on your players? ;-]
07:39:03lostlogicShure E3 are pretty low impedence.
07:39:06biffherowould it follow that my sucky headphones need to be turned up louder?
07:39:16lostlogicI can hear the click wheel clicks anywhere over -11dB.
07:39:36lostlogicbut I don't hear any low constant clickign
07:39:36webguest67and clicks also mean just rotating as well, yes?
07:40:05lostlogicwebguest67: hrm?
07:40:22biffheroare these them?
07:40:28lostlogicbiffhero: not necessarily, some pro cans are very high impedence
07:40:46webguest67would just rotation the wheel qualify as a click?
07:40:48lostlogicyeah, I have the E3
07:40:53Iguanamy seinhessiers broke :(
07:40:59IguanaIm buying sony LOL
07:41:20lostlogicwebguest67: no, only when it actually triggers a rockbox event, the clicks are actually rockbox sending the volume control changes to the hardware based on this.
07:41:36Slasherimorning :)
07:41:40webguest67ah, so rotating my finger around is essentially clicking
07:41:43lostlogicwebguest67: if you crank the volume and return to the file browser, I bet you get no clicks, I don't
07:41:54webguest67lemme check
07:42:03lostlogicwebguest67: as defined bya pple firmware, the piezo clicks only when the finger rotates a 'notch'\
07:42:38 Quit arf-arf (Read error: 110 (Connection timed out))
07:42:48webguest67lostlogic: is the scroll wheel noise only coming from the right speaker?
07:42:55webguest67that is how it is happening on mine
07:43:08lostlogicwebguest67: definitely stronger in the right, I didn't pay close attention though
07:43:33lostlogicIguana: why sony?
07:43:35lostlogicIguana: get shure :-P
07:44:02webguest67lostlogic: you are right, the scroll wheel does stop making noise when it is in the menu
07:44:14Slasherii have also E3 and they are really high impedance
07:44:20webguest67but i still get the constant clicking
07:45:03webguest67for example, what happens when you pause the track, turn the volume up to 6db (dont play the track back, im sure that would hurt). do you hear a constant clicking in the right ear?
07:45:21lostlogicSlasheri: ermh? That doesn't sound right... 26ohm, isn't very high
07:45:23 Join iNT0XiC8D [0] (
07:45:38lostlogice5 are 110 ohm for comparison
07:45:40RoC_MMit's not 6db
07:45:46RoC_MMit's volume level 6
07:45:52Slasherilostlogic: well, anyway much higher than any other of my headphones :)
07:45:59lostlogicwebguest67: no, no constant clicking
07:46:09Mikachui looked at headphones yesterday, two of them were 16 and 32
07:46:12RoC_MMif the track is silent, putting it at 6db will produce silence, and thus not 6db
07:46:21Slasherihuh, using 110 ohm with dap might be quite impossible without some noise attenuator
07:46:34lostlogicSlasheri: yeah, I was just thinking that
07:46:51lostlogicI'm glad e3 are as high as they are −− in a quiet office environment -57 on the ipod is louder than I'd prefer :-P
07:47:52Iguanacan you get shure in australia? o.o
07:48:13lostlogicIguana: sells them, they ship worldwide, don't they?
07:48:25lostlogicalso the apple store sells at least e2 and e3 if you have apple stores
07:48:26Iguanaim not into internet buying O.o
07:48:27Sinbiosi think ship to australia
07:49:10biffherodick williams (or whatever his name is) can order them in, no?
07:49:44 Join Paul_The_Nerd [0] (
07:49:54Paul_The_Nerd would y'know, be a single song.
07:50:36Iguanadick smith
07:51:18Paul_The_NerdMan, I hate it when my computer gets all sluggish, and then my typing shows up in another window
07:51:56IguanaI have a Mac
07:52:44Paul_The_NerdIt's more a side effect of my horribly slow HD.
07:52:55webguest67lostlogic: when the volume is set so the ipod says -15db (the volume that is adequate for my headphones), I measured the clicking noises to be roughly at -73db
07:53:09webguest67so they are quite quiet
07:53:49webguest67but my sennheisers picks them up in a quiet enviroment
07:54:08biffheroPaul_The_Nerd: my KDE went so bonkers one time that my konsole window slammed an entire ssh session into IRC, _including_ the passwords. The other terminal was good, no echo, but it showed up in irc. ugh
07:54:45lostlogicwebguest67: are the click clearly periodic or semi-random?
07:54:54webguest67clearly periodic
07:55:02lostlogiconce per second?
07:55:09webguest67methinks it might be the ipods way of checking to see if the scroll wheel is pushed
07:55:16webguest67about 1.7 times a second
07:55:24webguest67its kinda hard to say
07:55:38lostlogicare they also right ear only?
07:55:56lostlogicdo they follow the volume control in loudness?
07:56:12 Nick Sinbios is now known as Sin|showah (
07:57:08webguest67lostlogic: i used adobe audition and it seems that they are roughly 118 BPM
07:57:26webguest67yes, louder volume increases the loudness
07:57:32 Join webguest76 [0] (
07:57:45lostlogicwebguest67: do you build your rockbox from source or do you use a binary build?
07:57:53webguest67cvs daily builds
07:58:17lostlogicwebguest67: can you try connecting to a charger and see if the clicks are still present?
07:58:41webguest67plug into usb? to keep it from going to disk mode, i hold menu, yes?
07:59:06lostlogicThe first thing that comes to mind to me is that battery level is read from the ADC 2x per second and powering up the ADC may load the PMU just enough to tick the amp on your high impedence phones
07:59:10lostlogicwebguest67: yeah, hold menu
08:01:25webguest67lostlogic: no it is still the same. I noticed that the clicking exhibits different behavior at different screens. At the now playing screen, it is a constant 118BPM click. When file browsing, it does a different variation
08:01:49Paul_The_Nerdwebguest67: If you use a different WPS, is it different?
08:02:01webguest67yes it is different
08:02:07Paul_The_NerdSo, it's the LCD refresh?
08:02:21webguest67dunno, i will switch to the default rockbox and check. 1 second
08:02:25lostlogiclcd should update 30-40x per second
08:02:43lostlogicI'm inclined to think it's ticking during I2C based on what we've found out so far
08:03:00lostlogicwhich also makes sense since our I2C driver is guessed and definitely not the correct use of the hardware
08:03:28lostlogicmy shures are clearly not high enough impedence for me to hear it
08:03:52Paul_The_NerdI have PX100s, which aren't exactly either
08:05:08webguest67I am getting a pair of Creative EP-630s tomorrow, so chances are I will not notice the ticking
08:05:16webguest67but it is still a bit concerning
08:05:20LinusNlostlogic: you have a nano?
08:05:33webguest675G video
08:05:34lostlogicwebguest67: nothing to be concerned about
08:05:40lostlogicLinusN: 4g like webguest67 said
08:05:54LinusNwas looking at this:
08:06:28LinusNthe patch looks kinda naive, but i don't know much about ipod
08:06:49B4gderpreglow and cassandra are nano persons
08:07:10 Quit sharpe (Read error: 110 (Connection timed out))
08:07:13lostlogicthat should be all that is necessary for basic battery monitoring functionality −− preglow should really do it though since he has the device
08:07:13 Quit webguest76 ("CGI:IRC (EOF)")
08:07:29B4gderI have a fairly nice uplang for langv2 now as well
08:07:34lostlogicit's the exact same in terms of actually reading the level between the two, we'll just need different scaling and percent values for each ipod
08:07:45lostlogic(maybe not even different scaling)
08:07:54webguest67RoC__MM: Changed back to the default wps, still the same ticking
08:08:30Paul_The_NerdLinusN: I've *just* done the first two +s of that patch, and they seem to work *decently* at least. Dunno what the latter two defines actually *do*
08:09:09B4gderthere's only three things left for langv2 to get committed: 1) more testing by me 2) a convert-all to v2 script by me and 3) amiconn's english lang fixes
08:09:21LinusNPaul_The_Nerd: they activate the charging status control, so the charging animation stops when it is full
08:09:24lostlogicPaul_The_Nerd: you'll want them... shows the neat charging icon, and actuall reports that the hardware isn't charging any more when the battery is full respectively
08:09:46LinusNB4gder: you da man
08:10:07Paul_The_Nerdlostlogic, LinusN: Gotcha.
08:10:45Paul_The_NerdRegardless, the iPod charges while plugged in, reports somewhere between 90 and 100% full when fully charged (which is about as reliable as the Retail OS gets as well).
08:10:47lostlogicPaul_The_Nerd: those two will work on 5g and nano, but likely not on 4g until someone looks up the right port bits for them
08:11:00Paul_The_Nerdlostlogic: I was only commenting on Nano anyway. :)
08:11:36lostlogicI think the nano battery is flakier in terms of full charge discharge voltage than the bigger batteries, from what I've heard abou tit so far
08:11:46Paul_The_NerdOh, definitely
08:12:04lostlogicI guess that is the life of a small LiPoly cell
08:12:12lostlogic(or is it LiIon instead?)
08:12:13Paul_The_NerdOh so very small.
08:12:19Paul_The_NerdI think it's LiIon, but I'm not 100% sure
08:12:42lostlogicI seem to recall that poly only made sense on the larger cells, but I could also be hallucinating that.
08:13:28 Quit RotAtoR ("zzzzzzzz")
08:13:37Paul_The_NerdI'm almost certain the 400mah replacements I saw for it were LiIon
08:14:09lostlogicand on that lithium related note, I sleep.
08:17:12 Quit webguest67 ("CGI:IRC")
08:22:52 Join Sinbios [0] (
08:25:30 Join damaki_ [0] (
08:25:39 Quit RoC_MM ("Leaving")
08:26:07 Join Cassandra- [0] (
08:26:51XavierGrLinus: Do you happen to know any details about the failuer in USB Bootloader Mode in H300?
08:27:15XavierGrIt seems that many people (including me) have this problem.
08:29:27 Join ender` [0] (i=ychat@
08:32:18XavierGranyway I have to go
08:32:21XavierGrLater all!
08:32:29 Quit XavierGr ()
08:34:16bam__so...did I miss anything?
08:36:27JdGordononly the horrible massacre of all the damn swedes here
08:36:47JdGordona giant leprecawn crossed the waters from ireland and ate em all
08:39:33 Quit Sin|showah (Read error: 110 (Connection timed out))
08:39:49LinusNamiconn: moo
08:39:57 Quit Cassandra (Read error: 110 (Connection timed out))
08:39:58 Nick Cassandra- is now known as Cassandra (
08:40:01bam__well damn he must have been very hungry
08:40:08*amiconn is editing ArchosFAQ :)
08:40:12*LinusN was spared
08:40:16bam__bet he got indegestion from all that hair
08:40:24JdGordon.. i mean.. yay :D
08:40:45LinusNamiconn: :-)
08:40:58bam__so anyone know if the play-stop-play bug has been fixed?
08:41:59 Quit Iguana ("Later")
08:42:13midkaybam__, it was right after it was introduced.. days ago..
08:42:20midkayif we're thinking of the same bug.
08:42:29 Quit damaki__ (Read error: 110 (Connection timed out))
08:43:09bam__i think so
08:43:27bam__where you press stop the play again and its paused
08:44:17midkaymaybe not the same one then..
08:44:25midkaynever had such a problem, myself..
08:44:31 Join kkurbjun [0] (
08:44:37LinusNmidkay: we're getting complaints about the button combo that pops up the pitch screen
08:44:39bam__hmmm I have tthe 23rd release
08:44:55midkayLinusN, hm?
08:45:37bam__lemme check the latest release(compiled that is)
08:45:44B4gderkkurbjun: how you like my idea of getting your doom version into CVS and giving you CVS commit access to work on it there?
08:46:32kkurbjunB4gder, That would be great if that's what works best for everyone.
08:47:21B4gderkkurbjun: I think it would allow more people to help out with the work.
08:47:25Paul_The_Nerdkkurbjun: Does it compile for iPod again?
08:47:25midkayoh, forgot about id3 display.. and never even knew there was a directory skip. :)
08:48:05B4gderyeah, we'd need it to build fine for all targets before commit
08:48:10kkurbjunB4gder, Yeah, that would be great if more people worked on it.
08:48:27midkay"First we lost the "quick" menu (which now has to be turned off expicitly, rather than disappear automatically when you release the mode button)" - i hope that wasn't my doing? i don't remember touching any button code..
08:49:05midkayit's even working for me now, i don't have to turn it off explicity.. he must be mistaken..
08:49:09kkurbjunPaul_The_Nerd, No, it builds, but it doesn't run yet
08:49:09kkurbjunB4gder, right now the only targets I can test or build for is the H300
08:49:26Paul_The_Nerdkkurbjun: What's wrong on iPod?
08:49:34kkurbjunthe ipod port compiles, but doesn't actually run
08:49:45kkurbjunit's a problem with alignment again
08:50:00Paul_The_NerdWhat was wrong with the working version that you scrapped it?
08:50:01kkurbjunit works fine in the sim and on the H300
08:50:17Paul_The_NerdYeah, but that's *one* target.
08:50:26kkurbjunPaul_The_Nerd, the new version adds many many more features
08:50:32LinusNmidkay: do you mean the archos version?
08:50:49midkayLinusN, yes..?
08:50:55LinusNjust checking
08:51:08Paul_The_Nerdkkurbjun: Have you worked on the alignment issue then?
08:51:46LinusNmidkay: hmm, it seems he has a h100
08:51:57kkurbjunPaul_The_Nerd, I would love for it to be working on the Ipods now, I don't really have a way to test the problem, or verify "fixes"
08:52:28 Join webguest35 [0] (
08:52:37midkayLinusN, i'm quite sure didn't touch button handling, so it must have been something else...
08:52:43midkayif it ever even worked like the archos.
08:53:01kkurbjunlinuxstb was able to fix the old version, but it doesn't seem that the problems are in the malloc code anymore
08:53:28LinusNmidkay: i don't think you did that
08:53:49midkayeither way, i still need to finish that update i've been working on. :)
08:53:51Paul_The_Nerdkkurbjun: Well, I'll take a look at it tomorrow, and at least document where I get data aborts or what not. I may not be able to fix things, but at the very least I can help figure out where to look again.
08:54:08kkurbjunPaul_The_Nerd, That would be great
08:54:11midkayhmm, to save me searching through files - there is a lcd_getmargins, right?
08:54:33Paul_The_Nerdkkurbjun: I'd do it now, but my iPod's been loaned out for the night, in an attempt to win a friend over to Rockboxing her H120.
08:55:03LinusNmidkay: grep is your friend
08:55:16kkurbjun: ), that's fine, I have other things I need to work on for the next couple of days anyway
08:55:18midkayLinusN, i sure could.. :)
08:55:20LinusNand all lcd functions are declared in lcd.h
08:55:20amiconnGetting doom to run on H1x0 using the grayscale lib shouldn't be a big problem... :)
08:55:24LinusNextern int lcd_getxmargin(void);
08:55:24LinusNextern int lcd_getymargin(void);
08:55:40midkaythere we are. thanks. :)
08:55:55kkurbjunamiconn, I plan on looking at that soon, the grayscale lib runs on the sims now right?
08:56:00midkayi'll probably then, tomorrow, update the quick screens, add support for them on the ipods, and fix the pitch screen..
08:56:15amiconnSome (probably tagcache) message now overwrites the logo at boot on archos :(
08:56:22amiconnkkurbjun: Yes
08:57:13Slasheriamiconn: ah, so the message "Committing tagcache" overlaps with the logo then..
08:57:18B4gderkkurbjun: and for x5 support, it shouldn't require much on your behalf other than a different lcd size and button mapping
08:57:31bam__guess that bug is
08:57:33amiconnSlasheri: The mesage was " " this time
08:57:45LinusNinformative :-)
08:57:46B4gderbut of course it can be left for an x5 owner to fix later on
08:57:54amiconn(and tagcache is not enabled right now)
08:58:11kkurbjunB4gder, I can give you a version that compiles cleanly on the H300 and the sims, would it be possible to start with those as the two targets for now, like I said it compiles on the ipod, but it doesn't run yet because of alignment issues. I guess I could start working on completing some other ports before the features are finished..
08:58:34B4gderkkurbjun: I think that's a fair start, yes
08:58:52Slasheriamiconn: tagcache always checks at boot if there is anything to commit, and it will display the message very shortly (user can't see it) if there is nothing to do
08:59:52amiconnOkay. So we need to adjust the test position on archos
09:00:00kkurbjunB4gder, I should have time to work on it after wednesday. I think I can have a clean release by that night
09:00:13kkurbjunafter wednesday afternoon that is
09:00:31*amiconn just statred putting some stress on the wav playback plugin, by playing a whole CD as wav, on recorder, running from battery
09:00:32Paul_The_Nerdkkurbjun: have you made any changes not in the patch tracker?
09:00:53Paul_The_Nerdkkurbjun: Specifically, if I'm going to be crashing my ipod with it tomorrow, I'd like to at least do it with sources that will crash in the right place. :)
09:01:19amiconn670MB wav file...
09:02:05kkurbjunPaul_The_Nerd, yes, but the changes that I have made shouldn't affect the crashes... I can give you a copy of the source I have now, but it really compiles ugly right now if you want to work with that
09:02:40Paul_The_Nerdkkurbjun: I'll just use what's on the tracker if you're fairly confident it won't change the crashes. I trust you.
09:02:52Paul_The_Nerdamiconn: How's it going with that?
09:03:31 Join darkless [0] (
09:03:37amiconnIt's running fine...
09:03:40Bg3rany thoughts about this from the ipod owners ?
09:03:46 Join markun_ [0] (
09:04:02 Join Zagor [0] (n=bjst@rockbox/developer/Zagor)
09:04:28 Join HuntermicAtWork [0] (
09:04:31amiconnLinusN: On archos tt's possible to get rid of the pop caused by the hard MAS reset when switching between playback and recording
09:05:00LinusNamiconn: oh, was there a pop? :-)
09:05:19amiconnWe can reset the DSP separately. This has the additional advantage that all settings of the audiocodec part stay intact
09:05:27amiconnLinusN: Yes
09:05:40 Quit markun (Read error: 54 (Connection reset by peer))
09:05:53LinusNthat's very good
09:05:56Paul_The_NerdBg3r: Some (several (many)) people have requested that feature in various locations. I'm somewhat indifferent about it. I think some thought needs to go into what should really happen when an earphone is unplugged, as there may be some alternatives that are better than simply pausing (though I haven't thought of any yet. So maybe not.)
09:05:58kkurbjunPaul_The_Nerd, Here's the test code: I don't know if this version will fit on the ipod's plugin buffer, if it doesn't let me know and I'll just scrap the whole prboom set of menu's as it adds quite a bit to the codesize.
09:06:33Bg3rPaul_The_Nerd i'm asking because this guy seems active :)
09:07:00Bg3rit'll be good someone to comment his work = (bool *)1 should be (void *)1
09:07:26amiconnLinusN: That's what I'm using to get the MAS back into mp3 mode before leaving wavplay, without the need to set all sound settings again.
09:07:38LinusNBg3r: otherwise it looks really nice
09:07:50Paul_The_NerdBg3r: Well, I see no reason not to really. People like that feature.
09:08:04amiconnThe only sound setting handled in the DSP is the stereo matrix, so only that needs to be set again
09:08:10Bg3ryep, but because i don't own iPod i can't test it, so i won't be the one to commit it
09:08:13amiconn(channel config + stereo width)
09:08:29 Quit iNT0XiC8D ("click...")
09:08:38*amiconn got a better understanding of the MAS by working with the PCM codec
09:09:29Paul_The_NerdBg3r: Unfortunately, as I said earlier, my iPod's out for the night. Sounds like it should be tested on a 5G anyway, since the creator has a Nano
09:10:32Bg3rso, linuxstb | lostlogic :P
09:10:43amiconnThe PCM codec also has the stereo matrix processing for recording. While it doesn't sound extremely useful, there's at least one thing it can do: switch channels for those units where L and R are swapped for recording...
09:11:08LinusNah, good
09:11:08Bg3ramiconn does this pcm codec allows recording too ?
09:11:20***Saving seen data "./dancer.seen"
09:11:29Bg3rah, ignore me
09:11:29 Join einhirn [0] (
09:12:29midkaykkurbjun, is that test source for ipod?
09:13:18Paul_The_Nerdmidkay: It compiles but will/should have data aborts.
09:13:18kkurbjunmidkay, I don't think it will work for the Ipod right now, it is just the latest source that I am working from
09:13:48midkaykkurbjun, ah, alright. apparently Paul_The_Nerd's ipod isn't available atm, was wondering if you needed a tester or something.
09:14:43kkurbjunmidkay, if you can find where the data abots are occuring and tell me if the code fits in the plugin buffer that would be helpful.
09:14:57midkaykkurbjun, oh, sure..
09:15:12bam__for those interested the fav's plugin is very easily edited in a nice hex editor...for predefined playlist building
09:15:23 Part bagawk
09:15:33bam__makes for easier otf playlist saving
09:15:34Paul_The_Nerdbam__: There's also the source...
09:16:03bam__only if your not an idiot and can get the cross compiler running :)
09:16:46bam__if I could get the cross compiler running, I could have some real fun...
09:16:54LinusNbam__: aaah, you're that idiot everybody's talking about :-)
09:17:04bam__eh, probably
09:17:24bam__eh the hex editor has limitations
09:17:31Paul_The_Nerdbam__: Windows, Linux, MacOS?
09:17:44Paul_The_Nerdbam__: What problems do you have?
09:18:22bam__I have bintools and gcc installed from the feed, I suppose its just configuration thats left
09:18:42Paul_The_NerdThe feed?
09:18:52bam__running ubuntu
09:18:56bam__err kubuntu
09:19:20Paul_The_NerdYou may just want to download the source and follow the full CrossCompiler instructions in the wiki
09:19:31Paul_The_NerdTake a wee bit longer, but *probably* end up with more consistent results
09:19:31midkaykkurbjun, shall i use the 0.81 diff?
09:19:44bam__yea...I should try again..
09:19:57bam__well off to try....
09:20:06biffheroBg3r: regarding 4899, have to go to bed now, hope to try it tomorrow
09:20:12kkurbjunmidkay, oh, thanks for that, I forgot the test needs a new diff
09:20:15Paul_The_NerdI used the CrossCompiler page for my original Cygwin install
09:20:18kkurbjunone second
09:20:21midkaykkurbjun, oh, alright :)
09:20:44biffheroon my iPod, it pauses on unplug, and then only lights up the LCD on plug-in
09:21:39Bg3rbiffhero your ipod is ?
09:21:40biffheroBg3r: retailOS is as follows: if "hold", unplug will pause, plugin does nothing
09:22:10biffheroif !hold, unplug pauses, plugin does nothing
09:22:24biffheroI thought that plugin would turn the LCD on, but it doesn't
09:22:41Paul_The_Nerdbiffhero: Odd, on the Nano, replug resumes playback I think(it takes a second to respond though)
09:23:43biffherowhat do you want me to apply that patch to? a new CVS download, or a daily, or a cvs 2006-03-24 ?
09:24:06kkurbjunmidkay, kkurbjun/rockdoom-sourcetest.diff">
09:24:20biffheroI really like the 'stop on unplug' because I often twist around and unplug the thing, or drop it, or whatever.
09:24:43Paul_The_Nerdbiffhero: You'd apply it to a clean CVS source ideally...
09:24:49LinusNbiffhero: apply it to the latest cvs
09:26:26biffheroI just thought of something. It might act differently on a plugin event depending on whether it is only in LCD light-off, or whether it is in LCD all-off. I will let it shut off the LCD here, and then plug in to it. waiting...
09:27:09biffherothere it is. pluging in the headphones when the LCD is off turns on the LCD and turns on the backlight
09:27:18biffheroit does not start playback.
09:27:20 Quit darkless (Read error: 113 (No route to host))
09:29:01 Quit Paul_The_Nerd ("Leaving.")
09:29:25 Join darkless [0] (
09:30:53midkaycrap, in VMware how can you scroll up or anything?
09:31:07midkaybeen bugging me for a while..
09:31:13midkaykkurbjun, also - doesn't fit in RAM, apparently.
09:32:52kkurbjunmidkay, thanks, I'll be abandoning the new menu's then
09:32:57webguest35Isn't really a vmware issue
09:32:57midkayaha. thanks.
09:33:05webguest35just linux-console
09:33:15midkaywebguest35, it's not an issue at all :) just my lack of knowledge..
09:33:28midkaykkurbjun, so.. shall i recompile with more RAM, or.. wait for a new version?
09:34:23webguest35midkay: I just meant that it isn't specific to vmware at all
09:34:37midkaywebguest35, ah.
09:34:42kkurbjunmidkay, you can recompile it with more ram for now, I'll release a new version when I have time to work on it
09:34:50midkaykkurbjun, alright
09:36:34bam__oh anyone have the at&t true voice crystal file?
09:36:35 Join aliask [0] (
09:38:32 Join petur [0] (
09:38:53aliaskHey is anyone having REALLY weird stuff happening when installing latest CVS?
09:38:56midkaykkurbjun, sorry, i have no base wads.. i need prboom?
09:39:06 Quit nkh^away (Read error: 110 (Connection timed out))
09:39:25kkurbjunmidkay, yes, kkurbjun/prboom.wad">
09:39:39midkaythx for direct link :)
09:40:03kkurbjunno problem
09:40:54midkaydata abort at 01F56578
09:40:57 Nick merbanan_ is now known as merbanan (
09:42:07kkurbjunmidkay, if you can add a delay in printf and find out where exactly that is in the code that would be extremely helpful
09:42:35midkaykkurbjun, i have no idea how you debug these or even use the numbers.. if it's simple, i can probably manage - what exactly? :)
09:43:31B4gdermidkay: you can use the .map file to figure out within what function that address is located
09:44:16midkaywhere do .map files end up?
09:45:02B4gderthe plugin maps are in apps/plugins and I guess the doom one is either there or in its own dir
09:45:31B4gder(I'm not familiar with the doom build integration)
09:46:25midkayi have doom.rock in /.rockbox - i don't see any .map files anywhere
09:46:27 Nick arrogant[buffy] is now known as Arrogant (
09:46:35B4gderthey're in the build tree
09:46:44midkayoh, that.
09:47:28midkay .text.Z_ChangeTag
09:47:28midkay 0x01f56578
09:47:30midkaylooks right..
09:48:20midkayfwiw, kkurbjun, i got a massive number of errors during compile, if you weren't aware
09:48:59kkurbjunmidkay, yes, I know this version has alot of warnings
09:49:12midkaykkurbjun, alright
09:49:55 Join austriancoder [0] (n=austrian@
09:49:58kkurbjunmidkay, it looks like there is a problem with the malloc code
09:50:15midkaykkurbjun, major?
09:51:52kkurbjunmidkay, I'm not sure where that problem is, but an abort at that address would indicate that. major enough to make doom not run..
09:52:02midkaykkurbjun, hmm..
09:53:52kkurbjunmidkay, thanks for that test, that at least gives me something to work with
09:54:33linuxstbSlasheri: Is it a known tagcache problem that tracks within an album don't appear in the correct order?
09:54:34midkaykkurbjun, no problem - anything to get doom committed quicker ;)
09:54:41kkurbjun: )
09:55:03kkurbjuntime to sleep
09:55:08midkaykkurbjun, night :)
09:57:43 Quit Arrogant (Read error: 104 (Connection reset by peer))
09:58:03 Quit Rick (Read error: 104 (Connection reset by peer))
09:58:04 Join Arrogant [0] (
09:58:11 Join Rick [0] (
10:00:17Slasherilinuxstb: yes, i will add sorting code to the browser soon
10:01:08linuxstbSlasheri: OK, thanks. I just wanted to make sure it wasn't something strange about my tags.
10:01:17Slasherigood :)
10:01:33linuxstbDid you see the TagCache wiki page I started?
10:01:49Slasheriin fact, i think that optional sorting mode by track order would be nice too
10:01:55Slasheriyes, i just read that
10:02:08Slasherii already tried to fix the ipod problem, somebody should test it now again
10:02:28linuxstbWhy optional? At least in album view, I don't think it makes sense for the tracks to not be in tracknumber order.
10:02:29LinusNSlasheri: how is chunked browsing?
10:02:33ArrogantMy battery won't charge.
10:02:46LinusNArrogant: which model?
10:02:48linuxstbSlasheri: I'm recompiling for the ipod now.
10:02:54ArrogantIt's only 3 months old
10:03:13linuxstbArrogant: Have you tried forcing your ipod into disk mode, or starting the Apple firmware?
10:03:22ArrogantI tried starting the Apple firmware
10:03:26Slasherilinuxstb: Hmm, true. I will still think about that. Doing the tracknumber ordering requires new tag to the tagcache engine
10:03:28ArrogantIt shuts off before it starts
10:03:33ArrogantIt doesn't have enough power to do it
10:03:48SlasheriLinusN: tagcache engine now supports it (sorting), but UI doesn't yet work right
10:03:51ArrogantAnd instead of charging, it boots rockbox
10:04:00Arrogantand sits there.
10:04:38ArrogantLike just now, I plugged it in and it loaded Rockbox.
10:05:00ArrogantWhich would be fine if it actually gained some power in the process.
10:05:09linuxstbHow do you know it's not charging?
10:05:30LinusNSlasheri: good, i was just wondering, since chunked browsing affects the sorting
10:05:51ArrogantIt keeps shutting off while trying to start up
10:05:55SlasheriLinusN: yep, the sorting should work fine now on all targets :)
10:08:07linuxstbArrogant: Hold down SELECT+PLAY when you turn on your ipod - this will force it into disk mode. Then try charging from there.
10:13:22 Quit darkless (Read error: 113 (No route to host))
10:13:28 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
10:13:56 Quit Arrogant (Read error: 104 (Connection reset by peer))
10:14:18linuxstbSlasheri: I still get a data abort if I try to load tagcache in ram. It was at 00030bcc - about half way through tagcache_get_next()
10:14:26linuxstbBut this seems later than before.
10:14:26 Join darkless [0] (
10:15:40Slasherilinuxstb: ah, interesting
10:16:40linuxstbIs a reboot required for the "load to ram"/"keep on disk" option to take effect?
10:18:39linuxstbMaybe it needs a message similar to the dircache?
10:20:12linuxstbBut anyway, it seems to be working very well, even at this stage. It solves the itunes problem nicely - it indexes both the itunes loaded files and files I've just copied to my ipod.
10:20:21Slasherigood idea, i will add that message
10:20:30Slasherinice :)
10:21:44crwlany idea why I got album contents in reverse order, ie. the last song first etc.?
10:22:15 Join Arrogant [0] (
10:23:43Slashericrwl: track number sorting is not yet working
10:24:01crwlok :)
10:26:28Slasheribut now going to eat something (to wilhelmiina) ;)
10:27:20ZagorBg3r: is there any point in having two "filter first" options? wouldn't it be better to merge them into one, and have the main and remote work the same?
10:28:01austriancoderwhat kind of tools do i need to "build" the manual on linux?
10:28:13linuxstbSlasheri: I've just added a note about track order within an album to the wiki page.
10:28:38linuxstbaustriancoder: LaTeX. Which Linux distributution do you have?
10:28:47austriancoderlinuxstb: gentoo
10:29:53linuxstbI've no idea about gentoo. But I would expect every Linux distro to have latex packages.
10:30:42austriancoderlinuxstb: i will look for latex packages for gentoo
10:34:23linuxstbSlasheri: I don't think you should add files to tagcache if get_metadata() returns false. A return value of false should mean that the file is unplayable by Rockbox - not that there were no tags.
10:37:31 Quit TCK (Read error: 104 (Connection reset by peer))
10:37:32 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
10:43:52Bg3rZagor some people like the remote's buttons to work immediately, while the ones on the main unit just to "show" the screen on first press
10:43:59Bg3rfor example h300
10:44:37 Quit amiconn (Read error: 110 (Connection timed out))
10:44:40 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
10:50:22austriancoderas tagcache gets into rockbox, i have a nice idea. Tagcache should update its cache with every rockbox start. To speed it up, we could add a hashshum to dirs so we can see where some modifications has been done. And tagcache only looks at the changed dirs.... what do you think?
10:52:04midkayaustriancoder, doesn't sound like anything revolutionary - still probably takes a while :) an option for a rescan on startup or manual updates would be best, definitely - since i only add music once in a while, it's rare - no need to waste time rescanning each boot..
10:52:11midkayanyways - bed for me. night all.
10:52:29amiconnAfaik, tagcache updates dynamically when dircache is enabled.
10:52:42amiconnSlasheri: Please correct me if I'm wrong...
10:52:55austriancoderamiconn: whats dircache and how to enable it?
10:52:59amiconnWithout dircache, dynamic updates would be too slow
10:55:41austriancodergot dircache running :) Ok.. so no need for checksums
10:59:06 Quit Arrogant ("Leaving")
10:59:39 Quit damaki_ (Remote closed the connection)
11:00:28Slasheriamiconn: correct if loading to ram is also enabled
11:00:38Slasherilinuxstb: ah, i will fix that
11:01:13Slasheriaustriancoder: yep, tagcache already detects new files automatically when cache is loaded to ram
11:05:15austriancoderSlasheri: cool
11:05:19 Part austriancoder ("Kopete 0.11 :")
11:05:20 Join damaki [0] (
11:07:29 Join tianjing [0] (
11:07:52amiconnSlasheri: What's still missing is removing deleted files, right?
11:08:26Slasheriamiconn: yep, true. But i will automate that also in future
11:08:48Slasheriit shouldn't be very hard to add
11:11:24***Saving seen data "./dancer.seen"
11:20:42 Join bobTHC [0] (n=bobTHC@
11:20:54bobTHChi peeps !
11:22:36bobTHCwelldone Bagder for the plugin wavplay !!!!
11:22:50LinusNthat was amiconn
11:23:13bobTHC:) so well done amiconn !
11:24:05B4gderI don't do good stuff
11:24:15 Join Jungti1234 [0] (n=jungti12@
11:24:16bobTHCu do
11:24:28 Join austriancoder [0] (n=austrian@
11:25:08bobTHChave we any chance to be able to implement some other codec in the mas ?
11:26:03amiconnThe PCM codec can be used for some more formats than the plugin now supports, like WAV with 8bit PCM, ADPCM
11:26:14amiconnWe can also support AIFF, and PCM recording
11:26:20amiconnThat's about it
11:26:42bobTHCalready not bad for the hardware
11:26:44amiconnWell, if someone codes an efficient MOD player...
11:27:03Jungti1234hmm sorry
11:27:25*amiconn continuously played wav today for ~1.5 hours
11:27:37amiconnBattery is still above half full
11:27:46Jungti1234I didn't look for copyright of Rockbox logo.
11:28:01Jungti1234Now, I found it.
11:28:18Jungti1234It was GPL.
11:28:24webguest35That version is
11:28:27bobTHCor some DSP tricks like the famous m3po "Variable playback speed is possible without affecting pitch"
11:28:29webguest35the large one may not be
11:29:20 Quit quobl_ (Remote closed the connection)
11:29:39 Join quobl_ [0] (n=quobl@tor/session/x-a314e6a2c6715462)
11:32:55Slasherilinuxstb: i have done a new fix, would you like to try a patch?
11:33:54 Quit austriancoder ("Kopete 0.11 :")
11:33:59LinusNbobTHC: the m3po playback speed is not a dsp trick
11:34:31bobTHChow they handle that ?
11:34:42LinusNadding/removing mp3 frames
11:35:31bobTHCall is done by the MAS ?
11:35:37LinusNno, by the sh1
11:35:46bobTHChuu, strange
11:36:09LinusNwe can do that as well if we want to, when we implement the frame walker for playback
11:36:20linuxstbSlasheri: Sure.
11:37:05bobTHCoki LinusN, another good news ;)
11:37:06LinusNi'm tempted to put together a proof of concept, but i will probably never find the time
11:37:32LinusNi have it all worked out in my head
11:37:49linuxstbWon't files using the bit reservoir cause problems?
11:37:58bobTHCas usual it's the 1 target ;)
11:38:39LinusNlinuxstb: we reformat the stream and convert it to cbr in realtime
11:39:07linuxstbSounds clever.
11:39:20LinusNcbr 320kbit/s
11:40:00*linuxstb has no idea what's inside an mp3 frame.
11:40:01LinusNthen we can strip any frame we like
11:41:03LinusNthere might be some obscure 320kbit/s stream out there that still makes use of the bit reservoir, but they are hopefully quite rare :-)
11:42:25bobTHCbottleneck or calculation problems ?
11:46:05Slasherilinuxstb: done,
11:46:07 Join tianjing_ [0] (
11:46:17Slasheriyou should first remove all .tcd files to regenerate the tagcache
11:46:32Slasherilinuxstb: ups, the url is
11:46:54linuxstbOK, I'll test it now.
11:48:15preglowLinusN: doesn't almost all cbr files use the bit reservoir?
11:48:24LinusNyes they do
11:48:48LinusNthat's why we convert to 320kbit/s
11:48:54LinusNto make room
11:49:12amiconnVBR files can use the bit reservoir as well (and often do)
11:49:27webguest35Hrm, anyway to figure out sf->flyspray item ids?
11:49:53webguest35That is, if an item mentions a SF id, which flyspray task that means
11:50:13amiconnLinusN: What's the X5V?
11:50:20 Join Matze [0] (
11:50:21LinusNx5 without fm radio
11:52:29 Quit tianjing (Read error: 110 (Connection timed out))
11:54:07 Join didj [0] (
11:54:52 Join willis_ [0] (n=willis@
11:55:29willis_anybody know how to get back to the WPS while browsing in the file browser?
11:55:41LinusNwillis_: what model?
11:55:48amiconnAh ok. Something like the archos *_SP devices then...
11:55:53LinusNamiconn: yes
11:56:02LinusNwillis_: press Play
11:56:44willis_LinusN, ah you're right... i swear that used to open directories instead : )
11:57:41linuxstbSlasheri: It still crashes in the same place.
11:57:44aliaskOn the coding side of things, how easy would it be to implement getting back to the WPS from the file browser when trying to go "up a level" while at root?
11:58:03LinusNaliask: dead easy, and very unwanted
11:58:14aliaskWhat are the reasons against this?
11:58:22willis_I have to say I would prefer that method, it seems logical
11:58:30LinusNlogical in what way?
11:58:54aliaskWell to me, the base screen is the WPS, so you are sort of exiting the browser. (I know that this isn't the case in rockbox though)
11:58:59willis_well there is no logical file structure beyond root (to the left that is) so it must go somewhere and why not to the playings creen
11:59:07LinusNaliask: to rockbox, the main screen is the browser
11:59:08aliaskIt seemed rather intuitive when it was like that in the iRiver firmware.
11:59:10Slasherilinuxstb: ah :/ hmm, could you find the exact location where it crashes?
11:59:39aliaskLinusN: I know, but intuitively the WPS is, I mean the player's main function is playing music, not browsing for it.
11:59:49peturLinusN: the settings menu behaves the same way: if you reach the top, it goes back to where you were
12:00:12LinusNaliask: intuition is very platform dependent
12:00:31aliaskLinusN: Yes, that's hard to argue with.
12:00:38LinusNpetur: that's because Left leaves a menu
12:00:58LinusNyou don't leave the browser, since it's the main screen
12:01:23willis_LinusN, well then isn't that the idea that left leaves a selection action.. therefore left should leave the song selection screen i.e. the file browser I think that's how it works in my mind
12:01:38willis_LinusN, but pressing play is easy as well : )
12:01:43LinusNwillis_: the browser is the main screen
12:01:55willis_LinusN, but from the wps it is a song selection screen
12:02:14LinusNno, not in rockbox
12:02:27willis_LinusN, also you could think of it as a back button in a web browser
12:02:45LinusNnow *that* would be confusing :-)
12:03:12LinusNbecause then you never know where you end up
12:03:20willis_the user wants to go back to what they were previously doing... i.e. listening to music
12:03:21*amiconn HATES that discussion about going "back" to the wps
12:03:27LinusNme too
12:03:36webguest35aliask: But how intuitive is the WPS screen as base screen if there's no music playing? Wouldn't you rather be able to find some music then?
12:03:36aliaskWe'll leave you alone then :)
12:03:50amiconnThe wps has never been the main screen in rockbox, and hopefully will never be
12:04:15amiconnWPS makes no sense without music playing. File browser always makes sense
12:04:34aliaskLet me make it clear that I am perfectly happy with the way rockbox is.
12:05:15willis_yeah of course... but as soon as you start the wps... it takes over, people don't generally start music playing and then stop it later to leave their mp3 player on to do other thigns
12:05:35aliaskI think that both sides have good arguments though. To me it's sort of balanced, so why bother changing the way it is.
12:05:45willis_so in a sense it becomes the default screen and all actions return to it.
12:05:50 Join b00st4 [0] (
12:05:55LinusNone major problem with left-to-wps is that it makes blind navigation harder
12:06:09willis_LinusN, that's a good point
12:06:27LinusNone left too many and you end up in wps, and the voice stops talking to you
12:06:57aliaskYeah, I do find myself pressing left about 20 times so that I know I'm at root.
12:07:00 Quit willis_ ("Leaving")
12:07:51linuxstbSlasheri: I'm not sure how I can find the exact C instruction it's crashing at. I've found the asm instruction... It's possibly around line 535 of tagcache.c, but it's just a guess.
12:08:10 Quit petur ("CGI:IRC 0.5.7 (2005/06/19)")
12:09:04Slasherilinuxstb: ok, thanks
12:09:42Slasherilinuxstb: hmm, in fact that seems to be the correct place
12:12:19 Join webguest10 [0] (
12:12:27webguest10hi @ all
12:12:42webguest10have you seen this? −−>
12:13:38aliaskI think there's a wiki page that has a large list of other "rockbox"'s
12:13:51aliaskB4gder with the details faster than me!
12:14:07 Join Moos [0] (
12:14:46amiconnlinuxstb: Mapping from asm to C shouldn't be hard if only moderate -O levels were used
12:14:53 Quit Sinbios (Nick collision from services.)
12:15:16 Join Sinbios [0] (
12:16:50preglowhave anyone ever trademarked rockbox? :>
12:16:50 Quit webguest10 (Client Quit)
12:17:33aliaskpreglow: Let's hope not... (if it would even affect us that much)
12:19:52 Quit nudel ("At Argon, we're working to keep your money.")
12:20:11amiconnB4gder: Regarding langv2.... How does that target mapping work?
12:20:42amiconnI need to understand that b/c I have to port parts of genlang2 to vbscript for building voices...
12:20:47B4gderits not very advanced yet, but basically I use the ARCHOS define from the makefile
12:21:05B4gderto allow the language system to pick different phrases
12:21:10B4gderh120: this text
12:21:15B4gder*: the rest uses this
12:21:27amiconnWhat about texts identical for multiple targets?
12:21:51B4gderit does a wildcard match
12:21:55B4gderand I intend to allow a list
12:22:04B4gderlike "h120, nano: heeey"
12:22:37amiconnE.g. the 'Yes' button: PLAY for player, all recorder. NAVI for H300, SELECT for h100, h120, iPods. RIGHT for Ondio
12:23:29 Join DrMoos [0] (
12:23:30 Quit Moos (Read error: 104 (Connection reset by peer))
12:23:42B4gderwith what I explain above, that should be no problem
12:23:51b00st4how can i add the clock to the statusbar of my wps?
12:24:14B4gderamiconn: for doing voice clips, I figure I should do a mode that extracts them all somehow for your scripts
12:24:14amiconnAnother suggestion: If a certain string is empty for a target (i.e. supposedly unused), could you modify the LANG_* identifier in the .h to allow catching the situation were it would be used unknowingly?
12:24:33aliaskWow, I didn't even realise that there had been much work into langv2 yet. Good to see though :)
12:25:02amiconnLike: LANG_FM_SCAN when it's used, but LANG_FM_SCAN__EMPTY when it's supposed to be unused?
12:25:29amiconnOtherwise we might accidentally have empty strings displayed
12:25:53*webguest35 stares at the mess of margin-patches, peakmeter and scrollbar extension patches
12:25:54Slasherilinuxstb: hmm, if i have for example struct test { short data }; should sizeof(struct test) give 2 or 4 bytes?
12:25:56webguest35viewports now!
12:25:59amiconnThis way gcc would barf about LANG_FM_SCAN not being defined...
12:26:22amiconnwebguest35: Post-3.0 stuff
12:26:27linuxstbSlasheri: I would guess 2 bytes.
12:26:51webguest35amiconn: Yeah, it's just sad to see the amount of hackish patches created
12:27:12Slasherilinuxstb: ah, then it might be a problem. please try following: tagcache.c line 72. Right after "short tag_length" add short padding;
12:27:19Slasheriand try again :)
12:27:44amiconnSlasheri: Short data only needs to be short aligned
12:28:48linuxstbamiconn: Look at line 72 of tagcache.c. I'm guessing Slasheri then casts tag_data from a char to something else.
12:29:52LinusNyou will have a hard time aligning that data
12:30:11Slasherihmm, or in fact that tag_data should remaing as char string
12:30:14 Join muesli__ [0] (n=muesli_t@
12:30:40B4gderamiconn: I'll focus on the needs first and do the fancy good-to-haves later on
12:30:51 Join nudel [0] (
12:32:53amiconnB4gder: I think this is an essential feature to protect us from empty-string accidents. I also think this shouldn't be hard
12:33:00B4gderI agree
12:33:07B4gderbut I still focus in the required stuff first
12:33:49 Join Lecksey [0] (
12:34:57 Join XavierGr [0] (
12:35:17amiconnlinuxstb, Slasheri: tagfile_entry.tag_data *has* to be padded to an even length at least on arm and SH1, otherwise it will crash
12:35:18Slasherilinuxstb: and btw, did you have the latest cvs snapshot before you applied that patch?
12:35:56Slasherior never mind in fact :)
12:36:14XavierGrDoes anyone know why the IriverBoot topic is cloased and I can't add some information?
12:36:23Slasheriamiconn: hmm, interesting
12:36:46Slasheriamiconn: ah yes, it should be padded now (with the patch i provided for linuxstb )
12:37:01Slasheribut it's still crashing somewhere :/
12:37:42webguest35XavierGr: locked by another one editing?
12:38:07XavierGrno, I have tried 3 times now and I get the same error.
12:38:17webguest35what is the error?
12:38:42XavierGr"/home/bjst/rockbox_html/twiki/data/Main/IriverBoot.txt is not writable"
12:39:05*webguest35 pings Zagor
12:40:01linuxstbSlasheri: That last change (line 72) seems to fix it :)
12:40:12Slasherilinuxstb: oh, great! :)
12:40:24Slasherithen i will have to figure out why it fixed that and commit the fix
12:41:48kclafsweet sweet
12:41:56*kclaf kisses slasheri
12:42:22kclafill test it when itll get commited
12:43:12Slasherigood, i will do that soon
12:44:07 Nick DrMoos is now known as Moos (
12:45:14 Join dpro [0] (
12:46:17Slasherikclaf: done
12:49:20 Join petur [0] (
12:49:59kclafok thanks, do i have to force a tagcache update ?
12:51:36Slasherikclaf: you should remove all of the .tcd -files in .rockbox directory and then reboot twice (with dircache and tagcache loading to ram enabled)
12:54:33linuxstbWhich of the Limits affect the size of the lists in the tagcache browser? Are both "Max files in dir browser" and "Max playlist size" relevant?
12:55:28B4gderimho, the tag cache should use aligning on all targets
12:56:04Slasherilinuxstb: at least the "max files in dir browser" affects the browser
12:57:28SlasheriB4gder: Hmm, i will consider changing that in future. It would increase memory consumption something like between 10-20 KiB
12:58:01Slasherino, 20-40 KiB in fact at least
12:59:02kclafer.. am i drunk or theres no more IPOD 5G bleeding edge build on CVS page ?
12:59:03XavierGrB4gder: Can you realease the lock of IriverBoot temporary to add a comment?
12:59:30LinusNXavierGr: fixed
12:59:51LinusNit looks like zagor had manually edited the file, perhaps removing wiki spam
13:00:32B4gderSlasheri: yes, but it would improve the code
13:00:41B4gderand it would make the caches binary compatible
13:00:45b00st4i often get this error, can any1 help me?
13:00:52B4gderthe current ifdef is a bit naive
13:01:10SlasheriB4gder: true
13:01:11B4gderisn't ppc also a bit stricted on aligning?
13:01:31*B4gder thinks of sim on Mac OS
13:01:53LinusNb00st4: you need to be in the correct directory to apply the patch, and the patch has to be properly created
13:01:57linuxstbWouldn't endian differences would prevent binary compatibility?
13:02:04XavierGrOk Linus, thanks. It is a comment for the H300 USB bootloader mode on some users.
13:02:06B4gderoh right
13:02:16B4gdervery true, scrap that comment
13:02:25b00st4and how i know the directory?
13:02:48b00st4or can i just set -p4 or something to search deeper for the file
13:02:56 Nick tianjing_ is now known as tianjing (
13:03:10 Join [IDC]Dragon [0] (
13:03:41LinusNkclaf: no, you're not drunk
13:04:22b00st4aight, made it up, thx
13:04:38*[IDC]Dragon cheers for wav playback in cvs
13:04:51[IDC]Dragonmade me compile it this morning
13:05:08[IDC]Dragonand try out Great Giana Sisters.wav
13:05:32[IDC]Dragonhaving 8 MB helps ;-)
13:06:05XavierGrgotta run. Later all!
13:06:10 Quit XavierGr ()
13:09:15[IDC]Dragonamiconn: do you read?
13:11:28***Saving seen data "./dancer.seen"
13:12:17LinusNb00st4: looks like one of the build servers is acting up
13:16:14amiconn[IDC]Dragon: Yes
13:17:49 Join Sin|showah [0] (
13:18:12[IDC]DragonI'm thinking about giving recording a try, if I find some sleepless hours or such
13:18:35 Quit Sinbios (Nick collision from services.)
13:18:36 Nick Sin|showah is now known as Sinbios (
13:18:37[IDC]Dragonyou have that optimised recording transfer somewhere?
13:19:02[IDC]Dragonor any other thoughts about the parallel i/f
13:20:10 Join spiralout [0] (
13:20:38 Join KN|stiff [0] (
13:21:06Lynx_wow, tagcache is done?
13:21:26spiralouthi i don´t want to be rude but has someone please a precompiled version of the h140 manual pdf from cvs?
13:22:06spiralouti tried serveral times to do it myself but i have problems with cygwin
13:22:25preglowa good point
13:22:39preglowwe might want to distribute those
13:22:54 Join c0utta [0] (
13:22:55spiraloutyes this would be nice
13:23:20linuxstbLynx_: It's now in CVS, but there's still work to do.
13:23:40Lynx_is it functional yet?
13:23:44preglowin a way
13:23:47preglowthere's a long, long way to go
13:23:56preglowspiralout: i can compile you one in a sec
13:24:01*LinusN looks at the latest bug report and cries
13:24:06preglowspiralout: but be aware it's still very unfinished, and probably incorrect at places
13:24:39spiraloutfine thank you
13:25:23 Join damaki_ [0] (
13:25:37spiraloutincompleteness isn´t a problem i know how rockbox works but i want to sell my player to a friend and maybe its easier to get in with a manual ..
13:25:59 Join webguest76 [0] (
13:28:19spiraloutthank you so much... great rockbox team !
13:28:19 Quit webguest76 (Client Quit)
13:28:56 Join webguest44 [0] (
13:29:55 Nick webguest44 is now known as obo (
13:30:03dprohi, I was wondering if anyone was working on updating the docs/ I looked over the plugins api file last night and it seems kind of outdated ...
13:30:17preglowspiralout: like i said, it's a draft manual, it's unfinished, and probably inaccurate in places
13:30:24dpro(like mpeg_bla_bla being audio_bla_bla now etc)
13:30:25amiconn[IDC]Dragon: Maybe I still have the asm loop somewhere. If not, it shouldn't be hard to redo that
13:30:30preglowspiralout: don't sell it as anything else
13:30:54preglowi guess this draft watermark thing tucoz put in is good after all :)
13:31:03oboBg3r: ping?
13:32:39amiconn[IDC]Dragon: The main problem was that I didn't trust my loop because of the transfer error, but as you know it turned out that it wasn't my fault...
13:33:39 Quit Lecksey ()
13:35:38 Quit Jungti1234 (Read error: 110 (Connection timed out))
13:40:48 Quit damaki (Read error: 110 (Connection timed out))
13:43:53 Join ashridah [0] (
13:44:58 Quit Sinbios ("If the definition of a klutz is someone who doesn't have eyes on their ass, then yes, I suppose I am a klutz.")
13:45:06spiraloutyes i know as mentioned it´s a friend of mine i also said him that rockbox is unfinished and not in an official state yet...the manual is only that he don´t have to ask me all the time you know
13:46:28 Join DrMoos [0] (
13:46:38 Quit Moos (Read error: 104 (Connection reset by peer))
13:47:14 Join sharpe [0] (
13:47:30 Nick DrMoos is now known as Moos (
13:48:32b00st4 <<what can i do to fix this error?
13:49:05LinusNb00st4: talk to the patch author
13:49:38webguest35Also, you should trim your screenshots
13:50:16b00st4dunno, what patch affected screen_access
13:51:36LinusNb00st4: the thing is, when you apply patches, you're more or less on your own
13:52:15LinusNonly the patch author can help you
13:52:45b00st4damn, i dont know, what patch it was, applied severals, but they all seem to succeed
13:53:15B4gderthen start over and rebuild between each patch
13:53:32b00st4jap, i gotta do it like that...:/
13:53:33dprob00st4: grep screen_access /where/your/patches/are * # ??
13:53:49dpromake that /where/your/patches/are/*
13:54:07b00st4screen_access is not the name of the patch
13:54:30dprob00st4: but it's probably /in/ one of the diffs
13:54:46b00st4has to be, ur right
13:55:45amiconnBagder, kkurbjun: What was the opinion about putting a proper malloc in the plugin lib?
13:55:58amiconnShould help solving a number of problems...
13:56:42B4gderand also introduce a few
13:56:57B4gderbut I think I'm in favour of adding it anyway
13:58:14 Join Jungti1234 [0] (n=jungti12@
13:58:32b00st4it was the peakmeter.patch
13:58:35b00st4editor in here?
13:59:51b00st4and what is the command to reject a patch
14:00:04B4gderpatch -R
14:00:22b00st4patch -r < name?
14:00:35preglowB4gder: why would it introduce anymore than we already have?
14:01:09B4gderbecause writing code with pointers and allocated memory is somewhat more tricky
14:01:11preglowthe only problem i can think of is leaking memory, and our current malloc implementation leaks memory by default...
14:01:22preglownet gain, no loss
14:01:39B4gderso it'll be writes to random memory addresses etc
14:01:39linuxstbMy only argument against putting malloc in the plugin lib is that it will encourage people to use it. But the problems caused by bad malloc implementations are probably worse.
14:02:12amiconnB4gder: It's not intended for writing new code, mainly for making it easier to make imported code work
14:02:18amiconnLike doom...
14:02:26preglowif we already know how to tell people not to use malloc
14:02:34B4gderI bet people will start using it the minut it arrives
14:02:34preglowbut i'm not about to tell people to rewrite doom not to use malloc
14:02:51preglowand good luck to them
14:02:58linuxstbWe could still reject patches that use malloc without a good reason.
14:02:59preglowas long as they don't use the audio buffer
14:03:28B4gderas I said, I am in favour but I don't think we should pretend that it won't add a few new interesting possibilities to do wrong
14:03:55preglow'course not
14:04:15preglowthis also calls for a new entry in the rockbox code rules about malloc, though :>
14:04:26B4gderhehe, yeah
14:04:38 Join CaT[tm] [0] (
14:07:20 Quit DJ_Dooms_Day ("Trillian (")
14:07:36 Join marevalo [0] (
14:08:12 Quit ashridah ("Leaving")
14:08:13 Quit safetydan ("Leaving")
14:11:25 Join ashridah [0] (
14:13:36linuxstbHowever, I don't think the current problems with Doom on the ipod are related to malloc though - the malloc implementation aligns the allocated memory to (IIRC) 16-byte boundaries.
14:13:51 Quit aliask ("Gnight")
14:13:55 Join DJ_Dooms_Day [0] (
14:16:25preglowstill doesn't say anything about the access pattern within the allocated memory
14:16:50preglowwhat are the current problems with ipod doom, btw?
14:19:38ashridahwell. the tagcache still seems to lock up on H140
14:20:24 Quit DJ_Dooms_Day (Read error: 104 (Connection reset by peer))
14:20:57JdGordonhey all
14:21:29 Join netcrusher88 [0] (
14:22:02Slasheriashridah: hmm, how?
14:22:10 Quit spiralout ("ChatZilla 0.9.70 [Firefox]")
14:22:40ashridahit generates a tag cache up to 100%
14:22:40preglowlocks up in the root menu sometimes here
14:22:45preglowjust plain hangs everything
14:22:51ashridahthen i enable "id3 database" in show files, and hangs when i exit the menus
14:23:00ashridahbacklight stuck on.
14:23:06Slasheriah, that's weird
14:23:06preglowditto here, yeah
14:23:09netcrusher88does ipod version require a reformat if it's already fat32 format?
14:23:14preglowhappens rather sporadically
14:23:18preglownetcrusher88: no
14:23:28Slasherisome problems with the ui.. remaining from the old tag database
14:23:30ashridahit does it consistently here
14:23:49netcrusher88does it have a way to build a database of id3 tags and browse like that?
14:23:53preglownetcrusher88: no
14:23:59preglownetcrusher88: you can browse the file system just as it is
14:24:00Slasherisomeone should redesign the ui for tagcache.. :)
14:24:00linuxstbpreglow: Doom just crashes with a data abort during initialisation IIRC. I think the problem is due to the use of the "packed" attribute in some struct definitions - which is needed in order to read the binary structures from disk, but I think is also used for RAM accesses.
14:24:04preglownetcrusher88: as a matter of fact, that's the default
14:24:26preglowlinuxstb: they should just redo the read routines if the structs to be read aren't too many
14:24:45preglowlinuxstb: how does idoom handle that?
14:25:02netcrusher88just read the faq, thanks :)
14:25:29linuxstbpreglow: I think rockdoom is now based on prboom, and idoom is based on the original doom (which didn't have these problems).
14:25:46 Join PhR3aK [0] (
14:25:47JdGordonhas any1 had a look @ ?
14:26:09 Part netcrusher88 ("Leaving")
14:27:07ashridahgreat. now it's hung at the directory cache enable/disable menu screen
14:27:21Bg3rJdGordon :))
14:27:46JdGordoni rekon that would be good for 3.0...
14:27:53 Join Febs [0] (
14:28:20linuxstbashridah: There's a TagCache wiki page - it explains a workaround for the "freeze after enabling the id3 database browser" problem. But if you find new problems, please add them there.
14:28:36Bg3rJdGordon if noone looks at it, i think to do it till 3rd of april...
14:29:02JdGordonu mean keep nagging about it? thats what i plan to do :D
14:29:18oboBg3r: you were looking at 4899??
14:29:46ashridahlinuxstb: sneaky
14:29:48Bg3robo yes, but i don't own a iPod
14:30:02Bg3rJdGordon i mean to think about commiting it
14:30:06ashridahwoot. 'tagcache' progress is a few hundred % now :)
14:30:15Bg3rashridah cool
14:30:17JdGordonBg3r: i dont have commit acess..
14:30:23JdGordonthats why im nagging..
14:30:27Bg3rJdGordon but i have...
14:30:38JdGordonso nu? do it! :)
14:30:53Bg3rlooking at the multiple-fm-presets patch atm
14:30:56oboBg3r: ah, okay, np
14:31:01Jungti1234hi JdGordon
14:31:07JdGordonhey Jungti1234
14:31:16amiconnlinuxstb: gcc *should* take care of accessing unaligned data in packed structs. Maybe that's another of those infamous gcc quirks...
14:31:47 Join DrMoos [0] (
14:31:49LinusNamiconn: yes, but it's easy to fool gcc with casting pointers
14:32:06 Quit Moos (Read error: 104 (Connection reset by peer))
14:32:26amiconnIsn't there a special attribute for pointers, telling gcc that the pointer might be unaligned?
14:32:38LinusNi have no idea
14:34:03LinusNsounds like something that would inflate the code size a lot
14:34:14 Join XavierGr [0] (
14:35:18ashridahah there we go. damn. that's pretty nifty.
14:35:19ashridahfast too
14:35:31JdGordonyay, tagcache finally works for me :)
14:37:30 Join Moos [0] (
14:37:34 Quit DrMoos (Read error: 104 (Connection reset by peer))
14:38:56 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
14:40:55safetydanSilly question. Can you just add decibel values together? So if I have +3db gain on the Replaygain pre-amp and -5db gain on the eq pre-amp, the end result is -2db gain right?
14:41:30safetydanmy life is easy then
14:41:35preglowwhat's up?
14:42:07safetydanadding the eq pregain setting
14:42:20safetydanI'll just add the values to the replay gain one
14:42:27preglowjust be careful about overflow
14:42:49Jungti1234What is 'Central European Time'?
14:43:11 Quit Moos (Read error: 104 (Connection reset by peer))
14:43:18 Join Moos [0] (
14:43:31JdGordonsilly q.. is there any way to disable tagcache? i think my mp3s arnt tagged well enough to bother with
14:43:48dproJungti1234: that's GMT+1
14:44:13dproJungti1234: paris, berlin, vienna etc ...
14:44:52dprobobTHC: okok ;)
14:44:52 Nick Paprica[sleep] is now known as Paprica (
14:44:52DBUGEnqueued KICK Paprica
14:45:12Jungti1234hi Paprica
14:45:28Jungti1234then, Why don't use UTC?
14:45:58 Join ender1 [0] (i=ychat@
14:46:09JdGordonLinusN: ..?
14:46:23 Quit ender` (Nick collision from services.)
14:46:26 Nick ender1 is now known as ender` (i=ychat@
14:46:45LinusNJdGordon: i've seen it, but i haven't had time to work on it
14:47:01 Quit Moos (Read error: 104 (Connection reset by peer))
14:47:06 Join DrMoos [0] (
14:48:03bobTHCGMT == UTC + temporal correction
14:48:41JdGordondoes any1 know how to make a remote control for the irivers? apart from the actual connector, is it fairly simple?
14:49:23Jungti1234bobTHC, dpro: Thanks :)
14:49:23LinusNno it isn't that simple, especially not the lcd
14:49:36JdGordona non-lcd one of course..
14:49:51LinusNthat's not *that* hard
14:50:14LinusNit will probably be hard to find the connector
14:50:46JdGordonim thinking maybe hacking up an old pci card to get the connector.. but it looks like the gap might be too narrow
14:51:01preglowhell yes
14:55:30Papricacould you commit it?
14:55:38LinusNyou can
14:55:46Papricai know, dont have a time =\
14:55:55Papricai'm learning to test =\
14:56:30LinusNi don't have time either
14:56:33LinusNany takers?
14:56:40XavierGrHmm CET?
14:56:47safetydanLinusN, Paprica, I'll commit it
14:56:49XavierGrIs there EET?
14:56:54XavierGror WET :D
14:56:56Papricasafetydan, thanks
14:57:01Bg3rXavierGr EET is our ...
14:57:08Bg3rWET is GMT probably :)
14:57:20B4gdertime zones names are fluent
14:57:31Bg3rGMT = GMT + 0; CET = GMT + 1; EET = GMT + 2
14:57:44linuxstbSo what timezone is central Europe in now? CEST?
14:57:56B4gderaccording to linux, yes
14:58:01Bg3rshould be, + daylight ...
14:58:30XavierGrGod with this Daylight thingy. I lost 1 hour of sleep!!!
14:58:45Jungti1234Korea is GMT+9
14:59:06Bg3rXavierGr haha yeah, but u got 1 hour more some months ago
14:59:34XavierGrindeed. I just nag about it when I am lazy
14:59:59CassandraOK. My build host should now be coping with SDL properly.
15:00:12Jungti1234Korea <= 8 => Europe ?
15:00:16B4gderCassandra: you checked your m68k gcc path too?
15:00:37amiconnpreglow: Any news from the gcc guys?
15:00:51CassandraDidn't know there was a problem, what's the arch?
15:01:08B4gderCassandra: the path for m68k-elf-gcc seemed wrong
15:01:46amiconnCassandra: See here:
15:01:48safetydanPaprica, done
15:03:05Jungti1234that is only hebrew support?
15:04:22safetydanAnother silly question. Why does apply_gain() in dsp.c do this?
15:04:23safetydan struct dsp_config *my_dsp = dsp;
15:04:23safetydan if (my_dsp->replaygain)
15:04:30safetydanWhy create a temp pointer?
15:04:35PapricaJungti1234, yep
15:05:08CassandraWeird. The compiler was somewhere totally different to where I thought it was.
15:06:18CassandraOK - looks good now.
15:06:23 Join netcrusher88 [0] (
15:06:38B4gderok, putting back m68k on your host
15:06:54netcrusher88is there a way to change the zoom level in the picture viewing mode in the ipod 5g version?
15:07:08 Join Kohlrabi [0] (
15:07:25CassandraB4gder, you got that I can do SDL builds, right?
15:07:30XavierGrnetcrusher88: currently no
15:07:34B4gderyes, I've put that back too
15:07:48B4gderyou've got a full line now
15:07:52PhR3aKwhy do the daily builds not work on my x5? the cvs builds work...
15:08:21CassandraIt'd be handly if the build system had some way to report brokenness directly.
15:08:21LinusNPhR3aK: define "work"
15:08:27netcrusher88why does it keep telling me that some jpg's are unsupported, is it certain color depths?
15:08:35B4gderCassandra: yeah
15:08:44PhR3aKerm... the daily builds dont boot
15:08:44webguest35netcrusher88: progressive jpegs probably
15:09:03LinusNPhR3aK: what does the bootloader text say?
15:09:07netcrusher88instead o interlace
15:09:35PhR3aKwait... i take a look
15:10:18PhR3aKsomething like model name: h120
15:10:19PhR3aKsum: [some_hex]
15:10:19PhR3aKresult: -5
15:10:56netcrusher88are there plans to add support of progressive jpgs in the future, just out of curiosity?
15:11:04B4gdernetcrusher88: yes please
15:11:08PhR3aKmaybee a silly question... but what is the diffrence between cvs builds and daily builds
15:11:10Bg3rnetcrusher88 do it :)
15:11:20linuxstbCassandra: It would be handy, but it's difficult to determine if a problem is due to errors in Rockbox, or errors in the build environment.
15:11:31***Saving seen data "./dancer.seen"
15:12:03netcrusher88lol wish i knew how
15:12:18B4gderyou can always learn!
15:12:25netcrusher88oh, i will
15:12:34CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:12:34*Cassandra thinks the distributed build system rocks greatly though. When I suggested it (no idea if I was the first to suggest it or not) I didn't seriously expect someone to implement it.
15:12:37netcrusher88and i'll get my CS major friend to help out, maybe
15:12:39linuxstbPhR3aK: A daily build can be thought of as the last CVS build of the day. So it's slightly more likely to be stable than one built during a busy day's development.
15:12:44LinusNPhR3aK: when did you download that daily build?
15:13:11netcrusher88can rockbox coexist with ipodlinux?
15:13:28PhR3aKok... sry i think ive tried out an older one... i'll try the newest one...
15:13:43Cassandranetcrusher88, yes. Just use the ipodlinux bootloader.
15:14:05netcrusher88thanks everyone
15:14:09PhR3aKbut sound doenst work anyway right? so nothing new about sound driver?
15:14:25linuxstbnetcrusher88: Yes. If you have a working ipodlinux installation using ipodloader2, then it's easy to add Rockbox. But doing the ipl installation is the hard part at the moment.
15:14:29LinusNPhR3aK: no, that still doesn't work
15:14:52 Join quobl [0] (n=quobl@tor/session/x-9ab524fec4a839a2)
15:15:36PhR3aKahh ok... it works... thx linus!
15:15:44preglowamiconn: the guy who said he'd have a look was moving last weekend, i'll start badgering him again in a couple of days
15:16:06 Quit petur ("CGI:IRC 0.5.7 (2005/06/19)")
15:16:14PhR3aKnice to see rockbox on my x5 :)
15:16:17netcrusher88yeah, noticed that the ipl is the hard part, but i think i got it
15:16:50 Part netcrusher88 ("Leaving")
15:17:26LinusNPhR3aK: would be even nicer to hear it :-)
15:17:57PhR3aKyep right... any idea how long it will take untill sound works?
15:18:00 Nick DrMoos is now known as Moos (
15:18:02B4gdersound is just sooo overrated ;-)
15:18:11LinusNno, it all depends on how much time i get to work on it
15:18:34LinusNit shouldn't be that hard though
15:19:01*Moos sending to LinusN some free time :-)
15:19:37LinusNehum, that is gratis, not free :-)
15:19:57*LinusN makes fun of free software nitpicks
15:19:58 Quit JdGordon (Read error: 104 (Connection reset by peer))
15:20:07 Join SereR0KR [0] (
15:20:19Mooshehe :-)
15:21:06*Moos is sending 100 000 euros instead ;-)
15:21:35LinusNmuch better
15:21:45Mooshaha :D
15:22:36*ashridah sends LinusN large amounts of alcohol in an effort to sabotage rockbox
15:22:48ashridah<LinusN> *hic*
15:22:58Mooshaha XD
15:23:25ashridah<LinusN> 10 print "Rockbox Rules" 20 goto 10
15:23:36LinusNhic...gentllemenn, we have schound!
15:24:10Bg3rLinusN u can even more :P
15:24:32ashridah* LinusN commits a whoppee cushion
15:25:28 Join jrok10 [0] (
15:26:13Bg3rseems preglow is on the cherry too
15:26:18LinusN /me does an "ooopsch" commit
15:26:24*Jungti1234 don't know meaning of the word... :(
15:27:13*Jungti1234 can't smile so.
15:27:54 Join diablogmp3 [0] (
15:28:04 Part jrok10
15:28:05PhR3aKis the battery status working on the x5? looks like it works...
15:29:44amiconnLinusN:^What are you sinking about?
15:30:57diablogmp3i have a question : will the cpu frequency changing on iPod be implemented soon ?
15:31:40linuxstbdiablogmp3: It's already implemented for the Nano and 5g/video. It's causing problems on the 4g, and it's hard to say when those will be solved.
15:31:51 Quit ashridah ("sleep time.")
15:31:54diablogmp3ok thanks !
15:33:01diablogmp3we, french ipod users, love Rockbox so we are awaiting for every major changes on Rockbox
15:33:49PhR3aKi mean does the battery bar work... because the status of power handling is 10%
15:33:56diablogmp3just few persons hesitate to install Rockbox because they consider that, for the moment, it's unstable...
15:34:23safetydanIs it better for dsp.c functions to be passed data or for them to grab data from global_settings?
15:34:37 Quit crashd ("Changing server")
15:36:13LinusNamiconn: :-)
15:38:21LinusNBagder: i'm wondering about your rec->play button switch in tree.h
15:38:40LinusNi thought we wanted rec to be the menu button
15:40:07B4gderI was pondering on using REC for stop...
15:40:31B4gderbut my swap in there was just because as I wrote in the commit message
15:40:34B4gderI liked that way
15:41:01LinusNok, i don't mind, but i think the Ondio way is better
15:41:10B4gderhow is that?
15:41:12LinusNhold Play to stop
15:41:25LinusNshort PLay = pause
15:41:28LinusNlong play = stop
15:41:31linuxstbThe ipods do exactly the same thing.
15:41:32HuntermicAtWorkdoes we have sound on the X5?
15:41:37LinusNhuntermic: no
15:41:45MoosYes I use Ondio too, and this way is better imho
15:41:48HuntermicAtWorkow, ok
15:42:45B4gderI think if other targets do one way it makes sense to do the same
15:43:21amiconnLinusN: I wonder why you prefer the Ondio way. I can't say I like it.
15:43:39 Quit marevalo ("Abandonando")
15:43:46LinusNbecause it is at least almost intuitive
15:43:49amiconnI would prefer modal operation. First press of Off would pause, second press Stop
15:44:04LinusNamiconn: what off button?
15:44:05amiconnWhen paused, Right (=play on Ondio) would unpause
15:44:24LinusNx5 has no off
15:44:31LinusNthat's the whole point
15:44:37amiconnWe would lose the ability to skip forward while paused, but imho it would be more intuitive
15:45:09linuxstblostlogic: Your last commit (clear the lcd before shutdown) should apply to at least the Nano and 4g Color as well. I am also curious to know why white was chosen, and not black when clearing the screen. Doesn't that mean there is a flash of white before shutdown?
15:45:13amiconnLinusN: I know. Similarly, Ondio has no (designated) Play, and Right is taken in wps for skip / seek fwd
15:45:49LinusNamiconn: then i was wrong about the ondio way, and should have said the ipod way
15:45:51amiconnlinuxstb, lostlogic: I'd rather switch off the lcd & backlight, like we do on other platforms
15:46:11LinusNfor the targets without Stop, i think long-play is the best choice
15:46:50amiconn(to be precise, we don't switch off on H1x0 and archos, but rather set contrast to minimum which has the same effect)
15:49:42linuxstbamiconn: If you can point us in the direction of hardware documentation... The problem is that the lcd image fades out over the period of a couple of minutes when the ipods are put into standby mode.
15:51:31 Part LinusN
15:51:58PhR3aKLinusN: Is the "off" button on the x5 currently set as something like a "back" button... because you can turn games off and so on with this botton
15:52:37Jungti1234He left.
15:52:41PhR3aKthats pretty bad because if you hold this button too long the x5 turns off
15:52:56PhR3aKok ^^
15:53:14B4gderPhR3aK: so don't hold it too long
15:53:26 Join imphasing [0] (
15:53:31B4gdermost players have the same thing
15:53:36amiconnlinuxstb: Is the lcd controller part of the PP, or is it separate?
15:54:10linuxstbI believe it's separate.
15:54:30linuxstbBut on the 5g it's accessed via the Broadcom chip.
15:54:39linuxstbs/it's/the lcd/
15:55:19linuxstbBut it should be possible to track down the lcd controller datasheets for the earlier ipods.
15:57:49Spida <- hm, I have seen a different (but interesting) idea for designing methods to enter characters with a limited amount of keys. Each key chose between a subset of charcters that gets smaller on each keypress. like a menu/submenu-system where ich menu is limited to the number of physical keys.
15:58:45 Join virtue|blot0 [0] (
16:00:12 Quit diablogmp3 ()
16:00:48 Join godzirra [0] (
16:00:51godzirrahiya folks.
16:01:03webguest35Someone port dasher to rockbox!
16:01:06godzirraWas just curious if anyone else using the ipod version of rockbox has trouble with a wall adapter to charge a 5th gen?
16:01:19webguest35Dasher being
16:01:20godzirraEverytime I plug it in, it reboots in a continous loop until I unplug it.
16:01:27webguest35(text-input method)
16:02:04lostlogiclinuxstb: preglow: either of you have the datasheet that is 'similar' to the 8758 handy?
16:02:38godzirraI'm guessing not a whole lot of ipod rockbox users alive yet?
16:02:56 Join spongeboy [0] (n=Administ@
16:03:38lostlogiclinuxstb: yes, there is a white flash on shutdown, I dunno why white instead of black, either one will cause a flash, I guess black would be less intrusive
16:05:03spongeboyhey, i'm new to this and have a nice easy question - what does the function led(0) and led(1) do?
16:06:24spongeboyit. Looking at metronome.c and its defined at the top of the file....
16:07:18spongeboyguessing it turns on a led (duh!) for SH7034 players.
16:08:21spongeboyJust want to say thanks for the good work everyone. i'm really enjoying rockbox, keep up the good hacking.
16:09:36lostlogiclinuxstb: is there a way to set the LCD to its solid color background instead of its backdrop background (like lcd_clear_display, but)?
16:10:05godzirraAnyone know if Ipod rockbox is any closer to displaying covers from the id3 tags?
16:10:19Mikachui don't think it's a wanted feature
16:10:30 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
16:10:33Mikachuthat's the impression i got from watching discussions at least
16:10:52godzirraHrm. I wonder how come? People would rather have it loaded as a seperate bmp?
16:11:08tucozIs it ok if I commit this? (looks nice in the sim at least)
16:11:10godzirraIts a feature I'd want... you could use it or not use it based on the wps file. Would be the ideal thing for me anyways. :)
16:11:23Mikachuwell, it seems silly to have the jpg in every file in an album
16:11:43Mikachuand bmp are faster to decode
16:12:03godzirraYeah.. I suppose thats true.
16:12:03amiconnlostlogic: Remove the backdrop, then lcd_clear_display();
16:12:06Zagortucoz: can't see why not
16:12:16tucozZagor, ok. I'll do it then
16:12:25lostlogicamiconn: yep, that's what I just did
16:12:38godzirraMikachu: My main problem is that I have covers in every album because thats the way my media tagger has 'em, and I don't have a script to go through and pull covers out into bmp files. ;)
16:12:46Mikachuis lcd_clear_display() faster than just memsetting the framebuffer to some value?
16:13:12Mikachugodzirra: shouldn't be too hard though
16:13:23webguest35You can put bmp in id3 as well, if you're so inclined
16:13:33godzirrawebguest35: Right, but rockbox doesnt read the imge from id3.
16:13:40linuxstblostlogic: The datasheet we used for the 5g's audio codec is the WM8983 - you can easily find it on the Wolfson website.
16:13:41godzirraMikachu: how? perl?
16:13:59webguest35godzirra: No, but mikachu was using bmp as an argument against using the id3tag
16:14:08godzirrawebguest35: Ah.
16:14:17Mikachui don't think bmp in id3 is a standard, but don't know
16:14:26Mikachugodzirra: shouldn't be too hard in linux :P
16:14:43godzirraYeah.. I just don't have linux setup to connect to my ipod. ;p
16:14:43Mikachuall you need is a program that extracts the jpgs in a semi-automated fashion
16:14:51Mikachuthat's trivial
16:14:54godzirraYeah, I know.
16:15:16godzirraI still think reading frmo an id3 tag should be an option though. ;)
16:15:18webguest35Mikachu: You can put any file you want, the mime-type is part of the frame
16:15:25Mikachuah, okay
16:15:36lostlogiclinuxstb: thanks −− any idea why all the bitshifting and weird is done in the wmcodec_send()?
16:15:40Mikachui don't know if jpeg decoding in the core is wanted
16:15:45webguest35So you could put image/gif if you wanted
16:16:01Mikachuor if the wps should load some code into plugin mem or something (eventually)
16:16:28linuxstblostlogic: You send a 16-bit value down the I2C bus to the codec. The top 7 bits are the register address, and the bottom 9 bits are the data.
16:16:48linuxstbIt's explained in the datasheet IIRC.
16:17:07webguest35The "type" field of the APIC frame is interesting.. you can set type to things like "Cover (front)", "Composer" etc.. and then "A Bright coloured fish"
16:17:23Mikachucan you have several apic frames thne?
16:17:33webguest35But only one of each type
16:17:51lostlogiclinuxstb: gotcha, thanks. I'm going to abuse it in the hope of eliminating popping on volume change (only audible at higher volume levels)
16:17:53Mikachucan type be arbitrary or a predefined set?
16:18:01webguest35It's a predefined set.
16:18:13webguest35 section 4.14
16:18:14Papricasomeone could help me with VMware?
16:18:45 Join mikearthur [0] (
16:19:06webguest35Man, if you used all frames of id3v2 you could practically convert it to a full website about that song
16:19:29Mikachui don't think i understand a bright coloured fish, why would you put a joke in a standard?
16:20:14webguest35I guess they just did.
16:20:25webguest35It's not like it matters, there are 255 values to chose from
16:26:16dprohmmm ... bmp2rb doesn't seem to work like it says in the usage statement ...
16:26:54CassandraMikachu, because they're terribly dry reading.
16:27:04dproI just try "bmp2rm -f 5 -i mypic -h . " now what's wrong with that ?
16:27:30dproyes there is a mypic.bmp ;)
16:27:50CassandraYou'd also really hate RFC1149.
16:28:52CassandraRFC1149 is the internet standard for IP over avian carrier.
16:29:02CassandraGo look it up if you're interested.
16:29:02bobTHCavian carrier ?
16:29:15dproCassandra: hehe good one has even been implemented twice
16:29:41preglowdoes most colour displays even have a contrast setting?
16:29:43dproCassandra: I think once somewhere in israel and then again in bergen
16:29:48bobTHCpigeon ,
16:29:53Cassandradpro: Indeed.
16:30:34CassandraAll you youngsters, missing out on some of the best jokes on the Net.
16:30:36dprothere are lots of excellent april 1st rfcs aroung ;)
16:30:41 Join Kyomi [0] (
16:30:54KyomiHey.. anyone know where to get that H320 optimized build of RB?
16:30:55Cassandra(although to be fair the RFC predates my presence on the net by about 2 years.
16:31:06preglowKyomi: it's capped the experimental build now
16:31:19preglowKyomi: and i think the place to find it is in the misticriver forums
16:31:25*Kyomi searches on misticriver somemore
16:31:35dproso back to my question what's wrong with my bmp2rb syntax ?
16:31:37KyomiHehe.. I was a member of that even before I saw rockbox :)
16:32:18 Quit Febs (Read error: 110 (Connection timed out))
16:32:29Cassandradpro: Not the foggiest. If I knew, I'd have answered you.
16:35:22KyomiI've heard that RB 3.0 might include the Doom plugin
16:35:56linuxstbdpro: I think you also need "mypic.bmp" at the end of the command-line.
16:36:02amiconndpro: You didn't give it a filename...
16:36:37dproah cheers
16:36:49amiconnThe -i option is only used (and necessary) for changing the logical name of the bitmap in the generated source
16:36:52dproamiconn: looks much better now
16:37:00dproamiconn: ic
16:37:13amiconnIf you have mypic.bmp and want it to be called mypic, there's no need for -i
16:39:07 Quit Jungti1234 ()
16:41:41 Quit XavierGr ()
16:42:13 Join Spida_ [0] (
16:42:28KyomiWould there be any reason to upgrade to the experimental build?
16:42:59preglowKyomi: well, review the feature list and see
16:43:13preglowKyomi: the experimental builds grows smaller as we commit patches, and it's kinda hard to keep track what new ones come
16:43:33 Part spongeboy
16:43:38KyomiI've just found where to get it.. nothing on where to see what is in it
16:45:41linuxstbKyomi: I think there are two major (for some users) features only in the experimental build - scroll margins and album art. Plus Doom, but I'm hoping to commit that to CVS tonight.
16:46:44KyomiDoom ^^
16:46:46KyomiDoes it work?
16:46:54KyomiAt least somewhat?
16:47:33KyomiAlthough.. I tried installing it earlier
16:47:46KyomiAnd when it rebooted itself the audio didn't paly
16:47:53KyomiThen I reverted back the the daily build
16:48:04KyomiAnd then had to shutdown/restart to get the audio back
16:48:29KyomiI'm guessing the installation is the same? Copy/paste/reboot?
16:48:44KyomiOr does the games folder go in a different place
16:49:19safetydanpreglow, would you rather data be passed to functions in dsp.c or have the dsp functions pull data from global_settings?
16:49:23linuxstbIIRC, you need a /games/doom/ folder, with prboom.wad plus whichever wads you want to play.
16:49:48KyomiYou mean I could take the wad from the original?
16:49:55KyomiOr hexen/heretic wad?
16:50:05linuxstbYes - I believe it can handle a lot of different wads.
16:50:08amiconnB4gder: Does the build system already use post-build transfer?
16:50:18B4gderamiconn: nope
16:50:31B4gderamiconn: but we now have pretty detailed logs for each build
16:50:34Kyomilinuxstb: Despite the obvious innuendo of that statment... I need to get that going :)
16:50:44KyomiToo bad I have to wait another 2 hours >_<
16:50:53KyomiI have class coming up in 10 mins
16:50:55B4gderand it shows the transfers to not use a very significant time
16:50:57amiconnWhen thinking about it, post-build transfer requires separate handling of the bleeding edge downloads
16:51:32amiconn...because bleeding adge builds wouldn't be available immediately after the build round ends
16:51:32B4gderthey are already popping in when the builds are done
16:51:45B4gderthey would just pop in slightly later
16:52:04amiconnYes, but I can imagine impatient people downloading too early, and getting the previous bleeding edge
16:52:15 Quit Spida (Success)
16:52:34B4gderright now I believe we remove the previous ones when a build round starts
16:53:00Mikachulostlogic: shouldn't something like middle gray be the best to clear to? though my knowledge of lcds is pretty limited
16:54:04KyomiSooo... theoretically... Doom should work?
16:54:27CassandraB4gder, is my server working OK now?
16:54:50Cassandra(Would be handy if you could link the build times on the by machine table to the builds themselves.)
16:55:10lostlogicMikachu: yes, most likely
16:55:18B4gderseems to be working fine, and yeah I'll work on getting those times link to the actual builds
16:55:21lostlogicMikachu: that's about what my background is set to and it looks dandy :)
16:55:42KyomiI need to get my doom wad quickly
16:55:42CassandraThanks. That way we can at least easily find broken builds.
16:55:57Cassandra(Maybe colour the link according to build status too.)
16:56:02B4gderI'll also link to the master build log
16:56:09B4gderfor each build
16:56:15CassandraGood plan
16:57:10*Cassandra wonders why the titles on the cvs build table insist on reloading each time I access the page.
16:57:18*amiconn wonders what server uses that strange, ever-changing hostname
16:57:20Mikachulostlogic: so maybe it should be set to that before shutdown?
16:57:32CassandraIt takes about three seconds to display the build table.
16:57:43amiconnSometimes it's called 'arm', now it seems to be called 'sdl', ...
16:57:59KyomiHow does this work?
16:58:13KyomiYou just slap the wad files in there and it'll let you select which one?
16:58:33KyomiWolfenstein 3d doesn't have a wad... I dont think
16:58:50B4gderamiconn: yeah, its a flaw in the "retry build on a different host" logic
16:58:58webguest35Cassandra: That is indeed strange - the files send a perfectly valid Last-modified (which is in the past)
16:59:15CassandraHmmm. My box didn't show on the last build run. What gives?
16:59:45Cassandrawebguest35, could just be delay in the renderer, I suppose.
17:01:28CassandraI am such a geek. I just bought my sister a DIMM for her birthday.
17:01:41amiconnB4gder: I should probably put my voice generation scripts in cvs, in a sub-dir below tools/
17:01:45 Join thegeek_ [0] (
17:01:59B4gderok, I've now added a link to "buildmaster log" on each server stats page
17:02:04B4gderamiconn: a good idea
17:02:08Cassandraamicon: Talkbox? Sounds like a good idea.
17:02:29CassandraI think you should change its name to Rocktalk though.
17:02:30amiconnI should also probably try to handle SAPI with activestate perl, this way I'd have less problems with the v2 conversion
17:02:31B4gderthe log is somewhat crude
17:02:37B4gderbut still a log
17:02:41amiconn(but more problems understanding the logic myself :/ )
17:03:37amiconnCassandra: I mean MakeVoices.vbs, voicefont.c and wavtrim.c
17:04:08KyomiOh my god
17:04:11Mikachuplease don't put .vbs in cvs, i would have to buy a new harddrive if i accidentally downloaded it
17:04:12KyomiThis is awesome
17:04:34KyomiHow can I switch games though?
17:04:34amiconnMikachu ???
17:04:41KyomiI have doom2 heretic and hexen wad in there
17:04:48Mikachunevermind me, i just don't like vbscript
17:04:55lostlogicMikachu: I tend to prefer setting it to the user's background color as that should be less of a flashy change before shutdown, where setting to a particular color could be a large change from their chosen background.
17:05:15Mikachulostlogic: well it should flash to the color that stays after the lcd is shut down
17:05:54CassandraHmmm. Looks like the build system failed to retreive the buildlog from byzantium.
17:06:00Mikachulostlogic: that should be less of a statement and more of a question
17:06:05CassandraCan't see why. It's here on the server OK.
17:06:32lostlogicMikachu: the original patch did that −− flashed white, but it was fugly... maybe I should flash it white after the backlight goes off.
17:06:34KyomiGod thats annyoing
17:06:35Cassandraamiconn, oh right. Yes. Them too.
17:06:42KyomiThe forward button should be forward
17:07:14KyomiThe music doesn't work in this version?
17:08:12 Quit thegeek (Connection timed out)
17:08:30preglowmno, don't think so
17:08:43linuxstbKyomi: No, the music has never worked in Doom on Rockbox. The sound effects should work though.
17:09:07KyomiHow do you open doors?
17:09:15 Quit KN|stiff (Read error: 104 (Connection reset by peer))
17:09:17CassandraI think if the cat wants to go out in the rain it really shouldn't dry off near me.
17:09:20 Quit Zagor ("Client exiting")
17:09:30Kyomiopen doors/press buttons
17:10:22linuxstbI don't know.
17:10:44KyomiDo you know how to switch wad files?
17:10:50KyomiI have doom heretic and hexen
17:11:25KyomiAnd there are two music sliders... one for the music I guess and the other for sfx
17:11:25Mikachuthey aren't doom
17:11:28linuxstbNo - I've only ever tried it with a single wad.
17:11:29KyomiI'm guesing they already know that
17:11:33***Saving seen data "./dancer.seen"
17:11:41KyomiI you can only do doom wad files?
17:11:45Mikachuheretic and hexen used other engines
17:12:04Kyomioh well.. doom is fun; )
17:12:13KyomiNow I just need to find a way for cheat codes :)
17:12:22Mikachudo you compile from source?
17:13:48Kyomiargh.. no strafe... looks like no rocket launcher on first level :(
17:13:53dpro'struct plugin_api' has no member named 'lcd_invertrect'
17:13:54dprohuh ?
17:14:08MikachuKyomi: do you need the strafe to jump to the thing?
17:14:19amiconndpro: You're trying to apply a patch from the stone-age...
17:14:21Kyomino wait
17:14:26Kyomirun XD
17:14:32KyomiI need run to jump on the vent
17:14:36KyomiOr walk fast or something
17:14:42Mikachudoesn't he run by default?
17:14:45KyomiStrafe would be VERY useful
17:14:46dprono I'm trying to write sth. and I do what the docs say ;9;2~)
17:15:04amiconndpro: Which docs?
17:15:26KyomiI keep falling short of the vent
17:15:36amiconnRegarding gfx, should contain all you need
17:16:01amiconnAh that one. It's way out of date...
17:16:06dproamiconn: ah cheers ...
17:16:31KyomiIt'd be wonderful if navi was fire
17:16:37Kyomirec was change weapons
17:16:42Kyomia-b was open stuff
17:17:07KyomiAnd + was actually move forward
17:18:18KyomiNow only if I could warp to lvl 31 to test how this thing REALLY wors
17:20:24KyomiHoly mother of god is this hard
17:21:30dprook so does this automatically get added to the plugins menu or do I have to do something to actually see it ?
17:21:30*dpro thought he'd rather ask before ;)
17:22:39Mikachuthe plugins menu is just the filebrowser in /.rockbox/rocks
17:22:51dproah I guess I should add it to apps/plugin.c
17:22:59dproMikachu: ah ok then
17:23:05B4gderand if you put the plugin in the regular plugin dir and built it so, it'll be put on the proper place with make zip
17:23:27KyomiYou know... I'd love to have a strafe left/right
17:23:28 Quit PhR3aK (Read error: 104 (Connection reset by peer))
17:23:32 Join PhR3aK [0] (
17:23:40MikachuKyomi: i think you have made that point clear :)
17:23:43KyomiIs there a way someone could code that like... quickly? ^^;
17:23:46 Quit PhR3aK (Client Quit)
17:23:57dprothe moment of truth ...
17:24:11KyomiI mean dont get me wrong
17:24:19KyomiIt's AWESOME they did all the work for it :)
17:24:30KyomiDo you think they would ever get the music working?
17:24:40KyomiOr maybe allow a music.. menu in it
17:24:44preglowit's ordinary midi
17:24:45dprosync && umount /ipod # let's see
17:24:55KyomiSay... get all of the music from it.. convert to ogg/mp3
17:25:04KyomiThen make a playlist
17:25:07Mikachui think doom uses the iram so it stops the playback
17:25:14preglowthat'd probably take more cpu than playing midi
17:25:22KyomiAnd just have the music repeat for the current level
17:26:55dprowell almost ... just the invert part doesn't seem to work the way I expected ...
17:27:06Mikachudpro: what are you making?
17:27:27dpromikachu: nothing just trying to get a picture to flash invert/normal mode
17:27:31KyomiYeah ummm...
17:27:42KyomiYou wont be able to get past level 2 without running
17:27:52KyomiYou need to run to get into that "fortress"
17:27:57Kyomito get the red keycard
17:28:16Mikachushouldn't that just be int_16 | 0xffffffff for every pixel?
17:28:46Mikachudisclaimer, that may not be the right number of fs
17:28:48KyomiThats... odd
17:28:58KyomiMouse Sensitivity is gamma level
17:30:32KyomiSo theres no way I'm going to get hexen or heretic to work... right?
17:30:39Mikachunot with doom
17:30:48Kyominot even with that zdoom thing
17:30:58 Quit B4gder ("time to say moo")
17:30:59Mikachuif you feel like porting zdoom to rockbox
17:31:19KyomiI dont even know how to port
17:33:30KyomiBut do you think heretic/hexen is beyond the scope of the h320?
17:36:38CassandraSeems unlikely.
17:36:57KyomiThat it would run? or not
17:37:34 Join Febs [0] (
17:38:24 Quit imphasing (Read error: 110 (Connection timed out))
17:39:17 Join crashd [0] (
17:44:50 Quit quobl (Remote closed the connection)
17:46:38preglowsame engine, no?
17:46:47Mikachuhexen is definitely not the same engine as doom
17:46:53*preglow retracts
17:46:54Mikachuit has scriptable events and whatnot
17:47:04preglownever was a big hexen fan
17:47:11 Quit Febs ("Reboot.")
17:49:11KyomiHow about Heretic?
17:49:21KyomiI just liked hexen for the weapons :)
17:49:46KyomiI'd bet Hexen would run noticably slower though.. if it ever did get converted
17:50:04KyomiNow... Only if I knew how to port... I'd get Wolfenstein 3D on here :)
17:53:28preglowthat should run fast as hell
17:53:47 Nick markun_ is now known as markun (
17:57:30lostlogicAm I correct that the minimum settable volume level for a device should be one lower than the minimum non-mute level available?
17:57:56Mikachuso that the minimum volume level is muted? that seems logical to me at least
17:58:14lostlogicmaking some math corrections in sound.c for ipoops if this is correct.
18:00:48 Quit darkless (Read error: 113 (No route to host))
18:04:17 Join Febs [0] (
18:04:22preglowlostlogic: afaik, that's usually correct
18:05:59kkurbjunlinuxstb, The idoom port used the packed_attribute in some areas of the code also. I was reading a faq that seemed to indicate that that attribute is ok to use on the arm compiler
18:06:01 Join AwayKrimi [0] (
18:06:12 Nick AwayKrimi is now known as nkh^away (
18:07:52kkurbjunamiconn, I think a proper malloc would be nice, doom has a full malloc which made life alot easier. The code could be reused also as it should be pretty fast. I don't think it's the most memory efficient, but its focus is on speed.
18:08:04 Join Nico_P [0] (
18:08:09preglowkkurbjun: we've already got a malloc ready for use
18:09:30preglowkkurbjun: why would you need fast malloc anyway? using a malloc where you need speed isn't very clever
18:10:43 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:11:27linuxstbkkurbjun: What's the current status of Doom now? If we ignore the ipod/alignment problems, is it working OK?
18:11:28kkurbjunpreglow, doom uses alot of it's own malloc calls in game. Having the malloc will be nice for other ports though.
18:11:56kkurbjunlinuxstb, yes, it's working great, I need to clean up my current code and make a release of that
18:12:35linuxstbOK. I'll wait for that, and then look at trying to commit it. After it's committed, I'll try and solve the ipod problems.
18:12:58kkurbjunlinuxstb, sounds good
18:13:29 Quit Febs ("CGI:IRC (EOF)")
18:13:48dprohmm if I set the backlight timeout to never in my plugin will I have to manually reset it to the value in user_settings or will the menu code take care of that ?
18:13:57Mikachuyou have to reset it
18:14:03dprok thx
18:14:52 Quit [IDC]Dragon ("CGI:IRC (EOF)")
18:14:53kkurbjunpreglow, I know malloc is traditionally very slow, the best fit algorithm would be much more efficient with memory. The zone allocator could end up badly fragmented depending on how it's used.
18:15:20 Join Febs [0] (
18:15:37kkurbjunthe best fit seems like it would be a much better canidate for inclusion
18:16:49dpronow this whole thing really starts to become fun - I guess the hassle starts once I consider writing /portable/ code with all them dreaded #ifdefs ;)
18:17:09 Quit YouCeyE (Read error: 110 (Connection timed out))
18:17:58dproanyone wrote some pcm player / mixer for ipod for a little drummachine or looper thing ?
18:20:04preglowsomeone did it for h1x0
18:20:23 Quit obo ("CGI:IRC")
18:22:33webguest35Whatever happened to that? Was a nifty little toy
18:22:44preglowlost interest, i'd wager
18:23:10webguest35There should be rules against that
18:24:16 Join quobl [0] (n=quobl@tor/session/x-d1f3f83dbe38a111)
18:24:48 Join nls [0] (
18:33:08nlsgood evening folks
18:33:43rob-sorry, ignore that m
18:33:49nlsok :)
18:34:47nlsI'm about to commit BatteryBenchmark plugin page for the manual and wondered if we should ask people to submit results to the wiki
18:35:14nlsbut make it multi target or something
18:35:51nlsBatteryRuntime maybe?
18:36:06webguest35You could just make an index page
18:36:14webguest35there's also IpodRuntime afaik
18:36:36webguest35So BatteryRuntime with links to IriverRuntime and IpodRuntime sounds good?
18:36:57webguest35And move generic description from those two pages to BatteryRuntime?
18:37:13nlsyep sounds nice, wanna do it?
18:38:30 Quit mog_work ("Leaving")
18:38:38webguest35not registred too lazy etc. :-s
18:39:04nlsI'll look into it then..
18:43:06bobTHCamiconn u forgot to change the version to 2.5 in the archos FAQ ;)
18:43:09*amiconn wonders why it's necessary to use special tricks for reducing clicks resulting from volume changes on newer targets
18:43:55amiconnThe audiocodec part of the MAS does that automatically. Just set a new volume, the MAS will softly ramp up/down internally
18:43:55 Quit DT291 (Read error: 104 (Connection reset by peer))
18:44:13 Join DreamTactix291 [0] (
18:45:40preglowamiconn: that's software based
18:46:00preglowyou'll seldom see something like that managed in hardware
18:46:03amiconnbobTHC: Feel free to correct that. I just wanted to change the answer regarding PCM, didn't check whether other answers are outdated
18:46:27bobTHCi know mate, just kiddin' ;
18:46:30amiconnpreglow: Hmm? How could that be software based
18:46:37preglowi can code something exactly like that for swcodec platforms too, but do we really want to waste time/cpu on that?
18:46:42preglowamiconn: i figured the decoder did it?
18:46:49preglowamiconn: since that'd be pretty logical
18:46:53amiconnThe volume ramping is done in the audiocodec part, not the DSP
18:47:07amiconnIt even works with the DSP part disabled
18:50:41preglowbut no
18:50:45 Join dpassen2 [0] (
18:50:52preglowaudio quality isn't the first priority for these people, it seems
18:51:35amiconnGot the same impression
18:52:08 Join xmixahlx [0] (n=xmixahlx@
18:52:10preglowi'm eager to see how the quality of the eq in the 5g audio chip is
18:52:20amiconnANother example is that the UDA1380 seems to have a lower S/N ratio than the MAS' DAC
18:52:31preglowit's got a five band one, pretty much like eq.c, but with coarse, fixed parameter step
18:52:44preglowlinuxstb: you tried using it yet, btw?
18:52:58preglowthe eq part in the nano sucks ass
18:53:05preglowit's so incredibly crappy i want to cry
18:53:43amiconnQuite noticeable hiss with my earphones, even at rather low volume settings
18:54:46*amiconn is curious how the iPod DAC compares to both UDA1380 and MAS
18:55:57preglowprobably unfavourably
18:56:08preglowgod knows
18:56:20preglowthe earphones i'm using now are so crappy it's a miracle i can hear anything
18:56:41amiconnThe UDA also hass significantly worse crosstalk attenuation than the MAS
18:56:47bobTHCidiotphiles knows
18:56:49amiconn(at least the headphone amp)
18:59:29linuxstbamiconn: There are 3 or 4 different DACs used in the iPods. The 5g seems to be regarded as the best sounding, and the 4g is supposed to have a noisy headphone out, a problem fixed if you use the line-out. Not sure about the minis.
18:59:44linuxstbpreglow: No, I haven't tried the 5g's EQ yet.
18:59:54amiconnI'll soon see myself...
19:00:03dpassen2The Shuffle supposedly has the best sound of them all, though it is an entirely different chipset.
19:00:03amiconn(or rather, hear)
19:01:08linuxstbYou'll have to finish the Rockbox port first though... The mini uses the same DAC as the 3g - and audio isn't working there yet. But it should just be a case of bug-hunting - it's implemented, but not working.
19:01:56 Quit virtue|blot0 (Read error: 110 (Connection timed out))
19:02:20 Join bluey [0] (
19:02:59*linuxstb wonders if slimx will return
19:03:29 Join obo [0] (
19:05:18tucozlinuxstb, could you perhaps call for help with the 3g over at ipodlinux?
19:05:28tucozor maybe you have
19:05:39preglowthere's no hurry...
19:05:50preglowcodecs will probably be a nightmare on it, so just lets drop the entire port
19:05:58preglowok, that's settled then
19:06:01tucozslower processor?
19:06:06*bobTHC wonders how they made to compare DACs, double blind ?, datasheets specs ? direction of the wind ? size of the cptain beard ?
19:06:38amiconnpreglow: Clock up to 90MHz instead of just 75...
19:06:38linuxstbbobTHC: The last two are the most likely. You don't want science getting in the way.
19:06:49preglowamiconn: i don't think pp5002 goes much higher than 75
19:07:07linuxstbWe could start using the second core...
19:07:08amiconnThe PP5002 factsheet says 90MHz
19:07:29amiconn...and the PP5020 factsheet says 80MHz
19:07:30preglowmyes it does
19:07:36linuxstbI think the IPL people have said they had heat problems over 75MHz. I could be wrong though...
19:07:47preglowi say we just stick to 75 mhz
19:07:48amiconnThat might be
19:08:00preglowsolve our problems in other ways
19:08:07preglowlinuxstb: yeah, we could start using the second core
19:08:10bobTHClinuxstb > ;) the irivers have problems above ~120MHz while the CPU itself is specced 140MHz
19:08:20preglowbut before we start working on that, i want to do some tests to see if iram is in fact useless
19:09:20amiconnPerhaps IRAM has some extra settings?
19:09:30preglowperhaps, but i can't imagine why that should be
19:09:37preglowsram is sram
19:10:11preglowbtw, i need to move the clickwheel piezo ticking out of firmware/ altogether if i want it to be an option
19:10:41preglowamiconn: becase firmware/ code can't check with apps/ code to see if the setting is on?
19:10:48amiconnThere are numerous things in firmware/ controlled by options
19:10:51preglowamiconn: or should i just add another button.c function to switch it on and off?
19:10:53 Join actionshrimp [0] (
19:11:05amiconnNo, the app code would need to apply that setting, via a function
19:11:10preglowyes, ok
19:11:13preglowi'll just do that, then
19:11:24amiconnSame way as sound.c, backlight.c, ... work
19:11:35***Saving seen data "./dancer.seen"
19:15:21biffherosomeone wanted a test on iPod 5G for (pause playback on headphone remove)
19:15:21 Quit Kohlrabi (Read error: 104 (Connection reset by peer))
19:15:21 Join Kohlrabi [0] (
19:16:14 Quit webguest35 ("CGI:IRC (Ping timeout)")
19:16:28nlswill databox and searchengine plugins be useable with tagcache
19:16:49obobiffhero: it wasn't me, but I'm interested to hear what you think
19:18:27preglowlinuxstb: you know if we currently power down any of the parts in the codecs we don't use?
19:18:53 Join imhasing[school] [0] (
19:19:11 Quit actionshrimp ("( :: NoNameScript 4.01 :: )")
19:19:13biffheroobo: there was some question as to the design. Someone thought that retailOS paused on unplug, and re-started on plug-in. I verified that it did not do that on my 5G, but it paused on unplug, and _if the LCD had shut down, not just turned off the light_ it would turn on the light and turn on the LCD on plugin, only if the hold button was not held
19:19:51preglowit pauses on unplug on nano, does not resume on plugin
19:20:05biffheropreglow. thanks, someone last night thought it might resume on plugin
19:20:12 Join darkless [0] (n=darkless@
19:20:24oboyeah, I wrote it to resume on plugin, but I don't use retail...
19:20:57oboWhen you say LCD shutdown - do you mean the hibernation mode that retail goes into?
19:21:07biffheropreglow: can you verify what it does on plugin in the two states 1. backlight off, LCD on, and 2. backlight off, LCD off
19:21:11linuxstbpreglow: I think it's the reverse - we should power-up only the parts of the codecs which are used. And I think we do that. But the codec handling needs a thorough review - someone needs to study the datasheets and work out the best configurations. I've just done the minimum to get them working in most cases.
19:21:11tucoznls, I have no idea. What is databox?
19:21:33tucozIs that the query making program?
19:21:52biffheroobo: I think so. it takes a little while to get into that state. I don't think it is the same sleep mode as when you hold down play for n seconds
19:22:14preglowlinuxstb: that's what currently happens in 8975, i see
19:22:14biffheroobo: will verify that they are different (or not) in a second.
19:22:36preglowbiffhero: lcd is always on
19:23:06preglowplus, i don't have any music that retailos recognizes anymore, so i can't test
19:23:12biffheropreglow: if I wait another minute or so on my 5G, the lcd goes blank, don't know if it goes off or not
19:23:13tucozmaybe we should just assume that tagcache is the only database option for the 3.0 release?
19:23:15 Join mirak [0] (
19:23:28preglowbiffhero: ahh, right, sleep
19:23:29 Join Rondom [0] (
19:24:04preglowbut why do we care about this anyway?
19:24:20preglowduplicating behaviour isn't generally our preferred way of development
19:24:21biffherosomeone wanted to know what the retailOS did
19:24:32nlstucoz, i think so
19:24:38obohow about a three state option? no effect on phones, pause only, or pause and resume?
19:24:47preglowwhat about not adding it at all? :P
19:24:56biffherothat sounds good
19:24:57obothere is that too :)
19:24:59linuxstbpreglow: Get's my vote :)
19:24:59preglowi was no big fan of the behaviour, heh
19:25:08 Join webguest35 [0] (
19:25:15biffherothe default could be 'no effect' on all platforms but the ipods.
19:25:31preglowbiffhero: not all platforms are able to support it at all, so yes
19:25:36obocan any other platforms detect it?
19:25:45linuxstbobo: No.
19:25:55biffheroI don't know how important the principle of least surprise is here. do you try to make the default behaviour on a platform match their retailOS at all?
19:26:35amiconnlinuxstb, obo: The archos player and recorder might be able to detect earphine plugin
19:27:13biffheroobo: I can't say that the timeout-sleep is any different than the play/pause-sleep, when I force the sleep with the keypress, a headphone insert turns on LCD and backlight
19:27:17amiconn(exploiting the fact that an ordinary 2-ring 3.5mm jack will ground the remote pin)
19:27:25obookay - so it I rewrite it to be 3 state that would be better? I don't have an archos to test with - so someone else would have to add that bit...
19:27:35 Quit bluey ("Leaving")
19:27:51webguest35Principle of least surprise shouldn't take into account the retailos, but Rockbox isolated, imho
19:27:53biffheroobo: that sounds most featureful
19:28:11preglowbiffhero: plugin turns it on again when it's sleeping
19:28:18webguest35I don't see how you could expect it to not resume on plugin
19:28:27webguest35But that's me
19:28:28preglowi certainly would not
19:28:39preglowi'd even be pissed if it did
19:28:42biffherowebguest35: when I plug it in to my otterbox, I don't want it to start playing. same with my aux-in on my car
19:28:43preglowpissed off, even
19:28:49webguest35If it stops automatically, I'd expect it to start automatically again
19:28:58biffheropreglow: better than being pissed on, I guess. ;-)
19:29:07webguest35Ah, I meant if it had stopped by unplugging
19:29:14obookay, I'll see what I can do
19:29:19*lostlogic spams CVS with a 3rd commit for the same feature.
19:29:19webguest35I see what you mean now
19:29:21preglowbiffhero: perhaps
19:29:52biffherowebguest35: I disagree, since I stop it when I get in the car by unplugging the headphones, but plug it in to the car jack before starting the car
19:30:28 Join Bger [0] (n=Bager@
19:30:38webguest35*shrug* I'd expect it to be a new playback state "on hold"or something, that is resumed on plugin.
19:30:50webguest35But I don't have an ipod, so have your fun :)
19:31:11linuxstbI don't think the playback code needs more complications at this point...
19:31:28biffheroam i correct that there is no actual cvs checkout per se, but only a download of a tarball of the cvs checkout?
19:32:46 Join lodesi [0] (
19:33:17biffheroI am testing the patch at for obo, I am using the tarball from .. Is that recent enough?
19:33:23Nico_Pcan someone point me to where the usb screen is shown ?
19:33:41lostlogicNico_P: you're goig nto fix USB from WPS?
19:33:46Nico_Pcan't find it
19:34:05preglowlinuxstb: do you know where i can find the pinout of the nano connector?
19:34:09Nico_Pat least if you're talking about the backdrop bug
19:34:22lostlogichmm, wasn't talking about backdrop bug
19:34:23lostlogicoh well
19:34:27Nico_Pwhat then ?
19:34:29lostlogicscreens.c usb_screen?
19:34:40lostlogicI think on iriver and ipod that goin gto USB mode from WPS doesn't work most of the time
19:34:56lostlogicit shows the screen, but doesn't connect to the computer on iriver and shows the screen but doesn't reboot on ipod
19:35:16Nico_Pthat i have absolutly no idea how to fix :p
19:35:35Nico_Pi found screens.c usb_screen, but i was looking for what called it
19:35:56webguest35grep's your friend
19:36:26 Join bluey [0] (
19:36:27Nico_Pdidn't quite figure how to use it yet... i tried but it did nothing
19:36:31Nico_Pi'm new to linux
19:36:42webguest35grep -r usb_screen apps/*
19:36:49webguest35should help
19:37:18tucozNico_P, find -name "*.[ch]" | xargs grep something
19:37:21Nico_Paahh... works much better now
19:37:43Nico_Pwell "grep -r usb_screen apps/*" worked fine
19:37:52tucozyes, that works fine
19:37:58linuxstbpreglow: (I'm assuming they are all the same)
19:38:20preglowlinuxstb: the nano isn't even smaller or anything?
19:38:27preglowconnector, that is
19:38:27 Join austriancoder [0] (n=austrian@
19:38:29linuxstbI wouldn't know.
19:38:54webguest35Paging Cassandra
19:39:17linuxstbI'm pretty sure they are the same - I haven't seen cables/accessories specifically for the Nano.
19:40:48linuxstbpreglow: Yes, they seem to the same. e.g.: (works on all ipods)
19:42:08lostlogicgrep -r −−include="*.[ch]" usb_screen apps <−− like tucoz' but no pipe ;)
19:42:17austriancoderwhere in rockbox can i see how much discspace is left?
19:43:56linuxstbMain menu->Info->Rockbox Info. But Windows doesn't keep that information current.
19:44:05preglowlinuxstb: when looking into the connector, it looks like only four of the pins are actually connected...
19:44:17austriancoderlinuxstb: thanks
19:44:28preglowman, it's hard to count these pins
19:44:29Mikachupreglow: they call it universal dock so...
19:45:01Mikachui'd also assume the usb cable only has the 4 usb pins connected
19:45:11preglowyeah, it's the usb pins that are
19:45:13preglowat least three of them
19:45:23preglowi just can't count my way to the last one without some kind of magnifier
19:45:51Mikachui only have 4 goldy bits on the connector on the cable
19:46:10preglowme too
19:46:14preglowand the three first are usb
19:46:22Mikachui thought usb was 4 pins
19:46:41Mikachudata+, data-, 5v and ground
19:46:55kclafhmmm is that normal that i still experience data abort with TagCache on Ipod5G (load to RAM enabled)
19:47:03preglowthe ground is common
19:47:07linuxstbkclaf: I don't experience any.
19:47:21blueyanyone uses a Rsync Script for syncing his player?
19:47:29Mikachupreglow: wikipedia says 4 pins...
19:47:45kclaflinustb : really ? i cleaned my previous files, re-updated, rebooted
19:47:47 Quit ender` (Read error: 110 (Connection timed out))
19:47:48preglowit's probably the usb ground pin
19:47:55preglowMikachu: it _is_ four pins, i know that
19:48:12preglowwell, ok, at least we don't have to worry about enabling recording for nano
19:48:18Mikachuoh now i understand what you said i think
19:49:03preglowthe shield could very well have been usb gnd tyoo
19:49:04preglowtoo <-
19:49:12preglowbut yeah, only usb is connected for nano
19:49:37preglowso ok, we can power down line out 2 as well, on nano
19:49:41preglowehh, line out
19:50:25preglowCassandra: reading this? the nano seems to have no serial port
19:50:40biffherohow do I get back to the WPS screen (ipod 5G)?
19:50:46linuxstbDoes the nano have a remote connector next to the headphone socket?
19:50:47preglowbiffhero: press play
19:50:51preglowlinuxstb: yes
19:51:00linuxstbThat's the serial port as well.
19:51:14preglowthe headphone socket?
19:51:23biffheropreglow: thx. is that only supposed to work from filebrowser, or from menu as well?
19:51:24Mikachuis the remote connector inside the uni dock contact?
19:51:27linuxstbNo, the remote connector next to the headphone socket.
19:51:30preglowbiffhero: file browser
19:51:36preglowlinuxstb: then how come the pins for serial aren't connected?
19:51:45biffheroobo: the pause on unplug restarts on plugin, and it is cool
19:51:46preglowlinuxstb: ahh, _remote_ connector
19:51:48preglowlinuxstb: no, no remote
19:51:52preglowlinuxstb: just the docking port thing
19:52:52linuxstbAre you 100% sure that the Nano's dock socket only contains USB? I thought it could charge via firewire?
19:53:12obobiffhero: cool sounds good :)
19:53:25amiconnpreglow: Are you currently looking at the inside of your nano?
19:53:25CassandraThere are pins for it. I can't think of a good reason why they wouldn't be connected.
19:53:31biffheroobo: I will test the tri-state patch when it shows up
19:53:36 Quit bobTHC (Read error: 104 (Connection reset by peer))
19:53:52obookay, if you're still in here I'll shout when I'm done
19:53:53preglowi see only four pins, and those line up with pins 27, 25, 23 and i _think_ 16
19:53:59preglowamiconn: no, just the cable
19:54:07linuxstbThis seems to suggest something is there:
19:54:08preglowamiconn: stuff might be connected on the inside, of course
19:54:56amiconnyes, that's what I mean
19:55:19preglowyes, that does indeed suggest that
19:55:44amiconnThe cable connector most likely just has the pins present which are needed for that cable's purpose
19:56:06preglowyeah, entirely probable
19:56:21amiconnI wonder why no iPod rockbox dev seems to have opened his iPod so far
19:56:28preglowbecause it's flaming impossible
19:56:33preglowwithout breaking it
19:56:49preglowespecially the nanio
19:56:53amiconnI am very curious what's inside the shell, and will try to find a way
19:56:54Mikachui saw some site that claimed to have ipod opening tools, but i don't know if they work obviously
19:57:17amiconnThere must be a way, since obviously the thing was put together somehow
19:57:32 Quit imhasing[school] ("psy..")
19:57:34Mikachumaybe they've glued it together
19:57:43preglowno, they haven't
19:57:59preglowall you need is a stiff guitar pick and some use of force
19:58:14preglowi've got the first, but haven't dared force it hard enough yet
19:58:15amiconnPerhaps the nano is different, but iirc someone (apple itself?) offers changing batteries
19:58:20Mikachuheh there must be a way
19:58:22preglowamiconn: nono, it's entirely possible
19:58:27preglowamiconn: there's even a video of people doing it
19:58:36Mikachuamiconn: i think their battery exchange plan actually entails giving you a new ipod
19:58:41kclafremoved * .rockbox, removed rockbox.ipod, reinstalled latest bleeding edge making sure i dont have cache issue with previous version, forced update on tagcache, rebooted twice −−> "Data abort at 00016C5C"
19:59:07 Join ender` [0] (i=ychat@
19:59:36Bgerkclaf what the .map file says about 00016C5C ?
19:59:51amiconnpreglow: Yeah, perhaps apple itself, but others?
20:00:05amiconnI remember having seen replacement batteries for mini on eBay...
20:00:53linuxstbReplacement batteries normally come with plastic tools for forcing ipods open. e.g.
20:01:31preglowi seem to have lost the only pick i had which could be used
20:02:20*preglow tries with a knife
20:03:06Bgeri think we can lose one of the main ipod devs ....
20:03:17preglowno, i think i'll ambandon that course of action
20:03:39Mikachuaren't there like broken ipods for sale here and there?
20:04:00linuxstbThere will be soon...
20:04:07amiconnMikachu: Probably.
20:05:01amiconnBroken units would surely be helpful to identify components. Opening a working unit would have the additional advantage of being able to do measurements to figure out how things work together
20:05:03preglowit's really bloody easy to scratch this thing
20:05:22preglowi don't recommend putting a stanley knife to it
20:05:44amiconnWhen writing the MMC driver, I had my Ondio running open and powered from a lab psu most of the time
20:05:47Bgerthere are no map files in the bleeding edge builds, i suppose ?
20:05:57amiconnThis way I could use my scope to watch signal
20:06:33Mikachupreglow: i get scratches on mine when i wipe grease off with soft cloth
20:06:47preglowyeah, it's a really sturdy piece of equipment
20:06:53preglowmade for mountain climbing
20:07:06Mikachui think their real profits lie in the socks
20:07:29preglowthan and the 20-packs of nano tubes
20:07:43Mikachuthey actually change $29 for some wool
20:07:45preglowbought primarily by people who are not right in the head
20:08:08Mikachua guy in front of me at a lecture had one, i had to try very hard to not laugh in his face (or neck)
20:08:45MikachuIt doesn't matter which iPod model you have because iPod Socks fit all of them. And it doesn't matter what your mood is because each iPod Socks package includes six bright colors so you can pick the one that feels best.
20:09:50preglowdamn, this metal casing's so sharp it bloody damaged a jazz3 pick
20:10:02*preglow gets mitten
20:10:27preglowMikachu: yes, with colours hand-picked by a grandma kite-high on acid
20:10:41Mikachuaren't there some glass scratches repear kits in regular hardware stores?
20:10:55preglowMikachu: i really don't care too much about my ipod looks
20:11:05preglowMikachu: as a matter of fact, i want the little bastard to look neglected
20:11:14preglowMikachu: and abused
20:11:21Mikachuhaha, i agree with that
20:12:54preglowi've got it now!
20:12:59preglowbusted my pick right up, it did
20:13:06Mikachuyou opened it?
20:13:41preglowi'm in the process, but it's really tight
20:13:49 Join linuxstb_ [0] (
20:13:50safetydanwoo... working eq pregain
20:14:01 Quit linuxstb (Nick collision from services.)
20:14:04 Nick linuxstb_ is now known as linuxstb (
20:14:12preglowi really wonder if this'll every sit tightly again
20:14:55safetydanI wonder if we need an "auto-gain" mode
20:15:00 Join carini [0] (
20:15:28Mikachupreglow: you can always put it in a sock and nobody will know
20:16:13amiconnsafetydan: In fact the MAS has that
20:16:23 Quit nkh^away (Connection timed out)
20:16:28amiconnIt's an option in rockbox for archos
20:17:08amiconnAVC (Automatic Volume Correction)
20:17:26safetydanIs that related to equalizer type functions?
20:17:48amiconnNo, but an el cheapo replacement for replaygain
20:18:11safetydanyeah just read the manual entry for it
20:18:47safetydanI was thinking that the eq pregain might need an option to automatically pick the lowest gain to avoid clipping
20:19:59preglowsafetydan: no, screw that
20:20:04preglowsafetydan: really easy to do manually anyway
20:20:23preglowok, it's almost close to open, but i feel like i have to bloody kick it to get it completely open
20:20:29safetydanyeah more work for me, was just thinking about it while tinkering
20:20:58 Quit darkless (Read error: 110 (Connection timed out))
20:21:11 Join darkless [0] (
20:21:13obolinuxstb: your comment on 4899 - would button.c be the right place for the GPIO function?
20:22:08amiconnpreglow: Btw, I wonder why EQs usually allow positive channel gains, therefore requiring pregain.
20:22:20amiconnWould be easier to allow only negative channel gains...
20:22:29Febsping tucoz
20:22:29linuxstbobo: I'm not sure... Headphone detection (at least the way it's done on the ipods) doesn't seem to fit anywhere.
20:22:35tucozFebs, yep
20:22:57preglowamiconn: because it's possible?
20:23:09austriancoderhow must i change the Makefile of a codec to link codec X against my codec y?
20:23:19preglowamiconn: of course you should always try to cut instead of boost, but if you want to boost, then hey, why not
20:23:37FebsWhen we're breaking chapters into different files, should we: (a) input all sections and subsections from the main.tex file, or (b) input sections from the main.tex file, input subsections from their respective sections, etc.
20:23:50 Quit Rondom ("I'm leaving on a jetplane, don't know when I'll be back again...")
20:24:43linuxstbaustriancoder: What are you trying to do? All the codecs (i.e. apps/codecs/*.c) are linked against all the codec libraries.
20:24:59tucozFebs, I think we should go for option a). That is, the main file could contain some text but for larger sections we could split that to a separate file.
20:25:13Mikachuhah, maybe you can use pulling the headphone coord as a button in rockboy
20:25:32austriancoderlinuxstb: i am trying to compile libtheora.. and i needs to be linked agains libtremor
20:25:37safetydanpreglow, speaking of possible things, should we have the ability to have 24 dB band gain when the pregain maxes out at 12 db?
20:25:41Mikachuwould make for interesting gameplay
20:25:57linuxstbaustriancoder: You mean the library itself requires linking against another library?
20:26:00preglowsafetydan: pregain "maxes" out at -12 db?
20:26:05 Join webguest13 [0] (
20:26:23austriancoderlinuxstb: yes.... i want to decode a video with libtheora and ogg is used for audio
20:26:38 Quit nls (" HydraIRC -> <-")
20:26:38 Quit webguest13 (Client Quit)
20:26:43Mikachuaustriancoder: you mean vorbis
20:26:47safetydanpreglow, the format of apply_gain is s8.23 so 12.7 or so dB is the max
20:26:52linuxstbCan't you just link your "theora.c" codec against both libraries?
20:26:59linuxstb(which will happen automatically)
20:27:03 Part carini
20:27:18preglowsafetydan: yeah, but that's _gain_, we don't want the pregain to be able to gain, only cut
20:27:54austriancoderlinuxstb: libtheora need internal some ogg vorbis stuff from libtremor
20:27:54preglowat least i can see no big reason to allow it to do gains, and if 12 dB is the max, so be it
20:28:11tucozFebs, I think it is easier that way, as we know that there is a main.tex file in each chapter directory. That file 'controls' that chapter. It will be messy if included files also includes other files.
20:28:28preglowmy nano case doesn't really fit that well anymore, no
20:28:36preglowwhat a bunch of fucking retards appe are
20:28:42Febstucoz, I think you're right.
20:28:43austriancoderlinuxstb: i need the functions in bitwise.c from tremor to compile libtheora
20:29:13tucozpreglow, "as a matter of fact, i want the little bastard to look neglected .. and abused"
20:29:23linuxstbYou shouldn't need to link bitwise.c into libtheora though. Just include the .h file, and then your "theora.c" links against both libraries.
20:29:26preglowtucoz: but not falling apart, there's a difference :)
20:29:31amiconnpreglow, safetydan: s8.23 should allow for 24dB gain
20:29:48austriancoderlinuxstb: okay.. will try it
20:29:57amiconn(if the value is linear as I presume)
20:30:16tucozbbl, reboot
20:30:17 Part tucoz ("Leaving")
20:30:59linuxstbpreglow: Did you see anything interesting inside your Nano?
20:31:16preglowlinuxstb: i couldn't get it completely open and didn't feel like deforming it further in the attempt
20:31:39safetydanamiconn, yes of course it should... braino on my part
20:32:00Mikachupreglow: my frying pan was a bit deformed, i fixed it with a hammer
20:33:38preglowMikachu: cast iron then, i assume
20:33:52Mikachuno, teflon and stuff
20:34:07Mikachui figured either it breaks or i fix it, it was pretty much annoying to use since it only got warm in one corner
20:34:19preglowyeah, a lot of them has a tendency to bend
20:34:21preglowfor some bloody reason
20:34:28preglowi had one that only got hot in the middle
20:34:34Mikachui may have rinced it with water too soon once or twice
20:35:04safetydanso... should I call it "Precut" instead of "Pregain"? :)
20:35:05preglowthough, one'd think banging one of those with a hammer would make it a bit less non-sticky, heh
20:35:07Mikachui used a rag between the hammer and the pan too
20:35:31Mikachui'm not sure how it will respond next time i heat it up though
20:36:03preglowexplode in your face!"
20:36:09Mikachucan't you just automatically pregain the largest gain of the bands but *-1?
20:36:17Mikachupreglow: oh no !11
20:37:41preglowMikachu: i don't like the idea of my software second guessing me
20:37:53preglowif i don't want the cut, then i don't want it bloody auto-compensating for me
20:38:07Mikachuhm, maybe not a good idea no, if you know the audio is not near clipping
20:40:01safetydanwell if you really want to get in to it, there's no stopping you setting +24 db pregain in a settings file. The UI won't let you do that though.
20:40:52preglowso, what's the gui range?
20:41:02safetydan-24.0 to 0.0
20:41:50safetydanwhich come to think of it, means I should be using 8 bits in the config sector not 9
20:41:57 Quit SereR0KR ("XChat Aqua")
20:42:18 Join Paul_The_Nerd [0] (
20:44:13preglowwhy? should 6 bits be enough for that?
20:44:32preglowincluding sign
20:44:43safetydan.1 increment
20:45:34 Quit Paprica (Read error: 110 (Connection timed out))
20:49:36 Join Chamois [0] (
20:51:35 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
20:53:21 Join PhR3aK [0] (
20:57:27safetydanhrm... .1 step size might be a tad small
20:58:05 Quit dpassen2 ("CGI:IRC (EOF)")
20:58:56preglowperhaps 0.25 would be better
20:58:58preglowor even 0.5
20:59:01preglow0.5 is nice, yeah
21:00:18 Quit bluey ("Leaving")
21:01:03amiconnWell, as 1dB is afaik the smallest human-noticeable change, 1dB (or perhaps 0.5dB) step size should be sufficient
21:01:13tucozpreglow, how do you play a file from the file-tree on the ipod?
21:01:24tucozor enter a directory
21:01:40 Part austriancoder ("Kopete 0.11 :")
21:03:46safetydanShould the equalizer band gain step size be 0.5 dB as well?
21:06:01 Join ep0ch [0] (
21:09:40preglowamiconn: i'm pretty sure i can hear smaller increments than 1 db when you're talking the entire frequency band
21:10:31preglownot much smaller, though
21:10:40preglowsafetydan: i dunno
21:10:44preglowsafetydan: i don't use the bloody thing
21:10:56preglowFebs: any opinions? :)
21:11:34safetydanpreglow, we really need someone on the team who uses it since I don't use it either :)
21:11:38***Saving seen data "./dancer.seen"
21:12:44kclafwhen i get data abort browsing the id3 database, which .map file should i check ?
21:13:44 Quit PhR3aK ("get satisfied! :: (Gamers.IRC) ::")
21:15:20Mikachuwhy are you coding on things you won't use? :)
21:16:15ep0chi think the eq gains and pre-gain step size should be consistent imho
21:19:55 Quit lodesi ("Leaving")
21:20:03 Join lodesi [0] (
21:20:42safetydanMikachu, so we can tick that feature off the list :)
21:22:49safetydanspeaking of things I will use, I need to get RVA2 tag support in
21:22:58safetydanhalf my files seem to use that instead of the TXXX tag
21:23:13safetydanno wonder I thought I hadn't replaygained half of them
21:24:10Mikachui wish easytag would support taglib properly soon
21:26:40preglowMikachu: for fixed point coding experience on my part
21:26:48preglowMikachu: and because of the chicks
21:27:12Mikachulots of chicks in cs
21:28:02preglowjust ask'em if they want to come home and see your fixed point eq code
21:28:03preglownever fails
21:29:08Mikachui'll try it out
21:29:12*Paul_The_Nerd needs to remember that
21:29:30Mikachuis it okay if i pretend i wrote rockbox' fp code?
21:29:39preglowMikachu: sure, as long as you send compensation beer
21:29:52Mikachuyou can get compensation speech instead
21:29:54preglowMikachu: i've got a list ready you can just give to the systembolaget people
21:30:24Mikachui haven't been to systembolaget since i was little
21:30:52preglowthen here's the excuse to reintroduce you to that most splendid of establishments
21:31:13preglowyou can show them my imdct code as well
21:31:25Mikachuin jpeg.c?
21:31:27preglowhell, give them imdct36() and have the vomit in your lap
21:31:30preglowno, libmad
21:31:47Mikachui'm not into vomiting
21:31:56preglowthe huge bloody blob that makes your eyes water
21:32:06preglowi desperately want to get rid of it, but for some reason, it's quite fast
21:32:27safetydancordic can do dct...
21:32:57preglowcan it paint my walls?
21:33:03safetydanpossibly not
21:33:14Mikachutoast bread?
21:33:21preglowwhat about washing dishes?
21:33:21safetydannot likely
21:33:26preglow'cause i'm quite bored by that now
21:33:41safetydanthey have machines for that you know
21:34:19preglowmine is busted and waiting for the repair guy
21:34:24preglowliving in the meantime is agony
21:34:35Mikachui don't have one!
21:34:54preglowstudents don't deserve one!!
21:34:56Mikachudoing dishes and biking to school is when i actually use my ipod :)
21:35:10preglowi just recently got one myself
21:35:31safetydandishwasher wouldn't fit in my flat
21:35:34preglowi stopped listening to music while biking when i noticed it doesn't go too well with the speeds i tend to travel at
21:35:36crashdstupid keyboard
21:35:36Mikachui have to set my speakers ridiculously (that's hard to spell) high to hear it over the tap
21:36:11preglowthe entire way to the uni was one big downhill slope
21:36:18FebsI just committed a number of new files for the manual to CVS. I figured out how to do cvs add and cvs commit to add new files, but could someone step me through how to commit an update to an existing file?
21:36:35Mikachujust cvs commit it
21:36:37preglowFebs: what do you think about the eq? do we need smaller steps than 0.5db?
21:36:44 Join tianjing_ [0] (
21:36:53Mikachuit's like a new file only you don't need to add it first
21:37:20 Nick tianjing_ is now known as spacemail (
21:37:39Febspreglow, my gut reaction is no.
21:37:44preglowmine too
21:37:51safetydanbloody hell, there isn't actually much that cordic can't do, FFT as well
21:37:54bam__so hows the tag cache working?
21:37:56BgerFebs and if you want to be sure what did you change, u can use a cvs diff -u before commiting ...
21:38:02preglowbam__: working somewhat
21:38:07kclaflinuxstb : i have found the issue, it was caused by the voice menus ;x
21:38:14preglowi always do 'cvs commit filename(s)'
21:38:27 Join Henry [0] (
21:38:31preglowkclaf: what issue?
21:38:35bam__somewhat? so what are the issues, anything really bad?
21:38:48 Quit Henry (Client Quit)
21:38:49safetydanbam__, some lockups
21:39:01bam__and second anyone have a copy of the at&t voices file?
21:39:04preglowsafetydan: where are you seeing this?
21:39:07bam__ewwww, lockupos
21:39:08kclafpreglow : data abort on browsing the tagcache db
21:39:18bam__guess I shall wait :)
21:39:19safetydanpreglow, google for cordic fft
21:39:21preglowkclaf: weird...
21:39:22Paul_The_Nerdbam__: We can't distribute that at the moment.
21:39:39kclafstrange thing that it doesnt data abort when i normally browe files or menus
21:39:42bam__I know license problems
21:40:08bam__but if someone not associated with the project were to say email me the files...:)
21:40:09preglowsafetydan: nice to see there's a patent on it
21:40:23 Join Henry43 [0] (
21:40:50safetydanpreglow, yeah, not sure how old that is though
21:41:15safetydanlots of good ieee papers too
21:41:22Mikachubam__: optimist :)
21:41:59safetydanbam__, the microsoft voice ones aren't too bad
21:42:06Paul_The_Nerdbam__: Why not just use the voice files that *are* available for the moment?
21:42:13preglowthey're not up to date
21:42:13Mikachudoes festival suck?
21:42:20preglowMikachu: not very much
21:42:25preglowbut it's not the bees knees
21:42:35Paul_The_Nerdpreglow: Neither are the AT&T ones, they're just less "not"
21:42:36Mikachuapart from only being english (and spanish?)
21:42:57KyomiDoom2 is...odd
21:43:05KyomiFirst time I run it, it works fine
21:43:27KyomiI go back and play it again and it crashes.... so... delete the config file and it works... but theres no sfx
21:43:46 Quit tianjing (Read error: 110 (Connection timed out))
21:46:11Paul_The_NerdKyomi: Which version of Doom2 are you using?
21:47:29 Nick spacemail is now known as tianjing (
21:50:06 Quit Moos ("Glory to Rockbox !!!")
21:50:31 Join Moos [0] (
21:51:09 Quit Matze ("Miranda IM! Smaller, Faster, Easier.")
21:52:10tucozFebs, nice :)
21:53:30FebsThat will definitely make it easier to udpate.
21:54:08tucozFebs, yes. Did you make sure that no begin - end itemize spans over several files?
21:54:49tucozIt looks like it, just wante do make sure. (if that is the case, then it would be hard to maintain ;-))
21:55:42 Join SereR0KR [0] (
21:56:06tucozLooks great
21:56:38 Quit SereR0KR (Client Quit)
22:01:35 Join SereR0KR [0] (
22:01:57 Quit SereR0KR (Remote closed the connection)
22:02:25muesli__does winter and summertime change all over the world at the same date/time?
22:03:12muesli__damn..what a pitty
22:03:18 Quit t0mas (Read error: 110 (Connection timed out))
22:03:28 Join herz42 [0] (
22:03:34Mikachuthat's why there is a special timezone package for glibc
22:03:48Mikachuand a standard for specifying daylight savings switch date
22:04:17 Quit Bger ("[BX] Who ate my nuggets?!")
22:04:17 Join SereR0kR [0] (
22:04:28muesli__i guess its a pain in the ass work to include all timezones all over the world
22:04:50Febstucoz, I checked all of the begin/end{itemize} statements, and I am fairly certain that they are all correct.
22:05:10Mikachuit wouldn't be my first choice of things to do on a sunny sunday afternoon
22:05:23tucozFebs, good. I just wanted to make sure
22:05:53tucoza bit scared of hard to find errors
22:07:13 Quit SereR0kR (Client Quit)
22:07:24 Join SereR0KR [0] (
22:09:37 Quit lodesi (Read error: 113 (No route to host))
22:12:27 Join lodesi [0] (
22:13:28 Join wuzzles [0] (
22:13:44wuzzleshello all
22:13:55muesli__how much ram eats up tagcache btw?
22:14:28tucozmuesli__, check the audio info in the debug menu
22:14:42tucozthat is the audiobuffer
22:15:22 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
22:15:24tucozFebs, do you think we should provide a menu-overview in the manual? Like a tree for fast reference
22:16:51*amiconn is impressed by archos PCM playback :)
22:17:00FebsThat's probably a good idea.
22:17:07amiconn5.5 hours continuous playback with 44.1kHz 16 bit stereo
22:17:25tucozwow, that is really good.
22:18:00amiconnThat's with 2500mAh cells in a recorder v1
22:18:10dwihnoThere is PCM playback on archos?
22:18:22 Join maeck [0] (
22:18:36linuxstbSo if FLAC could happen, it should get even longer playback?
22:18:39tucozamiconn, I have no clue on batteries. Is that better than the battery on an iriver h100?
22:18:40amiconnCD quality means spinups every ~10 seconds, with the HD spiining approximately half of the time
22:18:55Mikachudo you really save battery if you spin up every 10 secs?
22:19:56amiconntucoz: Yes it is. Otoh, the 2.5" HD in the archos needs more power than the 1.8" drive in the iriver
22:20:13tucozI see
22:20:30tucozI am impressed nevertheless
22:20:31amiconnMikachu: I don't know, probably not much
22:22:52 Quit SereR0KR ("XChat Aqua")
22:24:07preglowamiconn: might it not be better to keep the disk spinning at all times?
22:24:57wuzzles−−> does directory cache not work on the video ipods yet?
22:25:28 Join drspoon [0] (
22:25:54 Join yeahx [0] (
22:26:33yeahxwow they're are some cool new things lately
22:27:07preglowlike what?
22:27:12preglowwuzzles: it should
22:27:28Mikachupreglow: the proper response would be 'yes we r teh roxx0rz'
22:27:38linuxstbwuzzles: Yes, it's worked for a while. Are you having problems?
22:27:46yeahxwell not so new maybe but the backlight fade thing
22:27:51 Join b00st [0] (
22:27:59yeahxcant turn backlight off still but it fades :)
22:28:06preglowahh, there's still that bug, yes
22:28:09preglowi wonder what causes that
22:28:15preglowluckily i have no inclination to look into it
22:28:32yeahxand Im about to figure out how to make the pacman thing work
22:28:51yeahxwhere to get and put the "rom" it says is misssing
22:30:07Mikachudid you look at
22:30:17wuzzlesnaw i dont remember if it was working or not - i am at my workplace.
22:30:28yeahxI was probably on my way there, was looking anyway
22:31:13drspoonrecently i saw something was added to stop the bitmap wps's from being built for players
22:31:15drspoonsurely it would be good to only build wps's designed for target screen size to be included?
22:31:50yeahxoh ok, thats where to put the files anyway
22:31:52wuzzleslast night i noticed that the channels were reversed when listening to my wavepack files, tho.
22:32:34wuzzlesor i had a suspicion that they were
22:33:14yeahxjust have to get the MAME rom I guess
22:34:30Mikachuyeahx: it's hard not to find them with google once you know the filenames
22:35:18wuzzlesor if you know the pacman zip name, may pull something up
22:36:10 Part Paul_The_Nerd
22:36:13 Quit b00st4 (Read error: 110 (Connection timed out))
22:36:33godzirraAny ipod devs on here? :)
22:36:40godzirraI mean now. As opposed to earlier :)
22:37:36godzirraGuess not. =)
22:37:41godzirraOh well... where do you report bugs?
22:38:03Mikachutry the first news entry on the home page
22:38:08Mikachusorry, second
22:38:32godzirraSorry. ask a stupid question.
22:38:42Mikachui don't have any
22:39:16godzirraHmm.. I'm not sure if my bug is related to the first bug on the list.
22:39:49godzirraI dont think so since its not usb..
22:40:11preglowwhat bug?
22:40:48yeahxoh yeah I know it wont be too hard to find
22:41:07 Join imphasing [0] (
22:41:20godzirraI plugin my rockbox enabled ipod to a wall charger (video ipod) and it goes into an endless "connected/disconnected/reboot" loop until I unplug it.
22:41:39yeahxboot into apple firmware
22:41:40Mikachuwasn't that fixed recently?
22:41:43godzirraIt acts like its trying to connect via usb, failing, rebooting into rockbox, seeing a usb c onnection, then trying again.
22:42:01godzirraMikachu: how recently? I don't have the daily build on there. Mines 4 days old or so?
22:42:13Mikachuno idea
22:42:33godzirrayeahx: Still a bug if I have to boot into apples firmware to charge, I would say. :)
22:43:09Mikachugodzirra: try holding menu when you plug it in
22:43:16yeahxthats what I used to do, always before connecting to the computer
22:43:27godzirraok.. lemme try that.
22:43:27drspoonis it just me or would mapping REC to player 2 up be better than PLAY on the h300s?
22:43:58Mikachudrspoon: player two in what?
22:44:04obobiffhero: new version if you want to test
22:44:10drspoondoh. i meant to say - pong
22:44:22godzirraholding menu when you plug it in seems to work.
22:44:25drspoonlook at a pic of the player and you'll see why
22:45:06safetydanokay, there's clearly something I'm missing here... why would this (precut << 23) / 10 result in apply_gain doing the opposite of what I want?
22:46:40herz42maybe it doesn't know what you want from it?
22:46:57Mikachudoes rockbox have a dwim() function in the core?
22:47:17herz42never saw this. what should it be?
22:47:21Mikachudo what i mean
22:47:22safetydanI think so... or at least there's a patch for it :)
22:47:33herz42ah, that would be a great feature...
22:47:54Mikachui think it would increase the code size a bit though :/
22:49:00safetydanand result in some very strange bugs
22:50:36safetydanah well, I guess rockbox can live without an eq precut settings for another night
22:51:11 Join PhR3aK [0] (
22:51:12herz42what should that calculation do?
22:51:38safetydanconvert precut to s8.23 format and divide that by 10
22:51:45safetydan(since precut is actually 10 times the needed db)
22:52:11herz42and how many bits does it have initially?
22:52:22safetydan8 max
22:52:26linuxstbMikachu: That ipod USB wall charger bug hasn't been fixed, but you probably saw Cassandra optimistically close the bug report on the bug tracker.
22:52:27 Quit PhR3aK (Client Quit)
22:52:59safetydanapply_gain acts weird, like if I precut by 0.5 dB it's fine, but cutting by 2 dB actually gets louder
22:53:13herz42usually integer div doesn't work for fixpoint operations
22:53:19preglowsure it does
22:53:23preglowif you're dividing by an integer
22:53:30herz42but you need to compensate, or?
22:53:52preglowonly if you're dividing by a fractional
22:53:57preglowand he's not doing that
22:54:14preglowsafetydan: all you need to be careful about in the case is overflow
22:54:46safetydanoverflow isn't an issue at the moment I hope since I'm just subtracting away from 0
22:55:15preglowsafetydan: what variable are you setting?
22:55:37safetydangain in apply_gain()
22:55:46 Join miner49er [0] (
22:56:04preglowthat variable isn't in db format, but you knew that, right?
22:56:42safetydanerr... no
22:56:50preglowthen hooray! there's your bug
22:56:51safetydanthat'll probably explain that then
22:57:09safetydanWhat format is it?
22:57:11preglowdb -> amplitude factor formula is: a = 10^(db/20)
22:57:27preglowso you need to squeeze your number through that first
22:57:34preglowfirmware/replaygain.c, function convert_gain
22:57:36preglowthat's how
22:57:41safetydanso that's what that's for
22:57:58safetydanexcept that runs at s7.24 which won't handle -24 db
22:58:32 Quit wuzzles ("CGI:IRC (EOF)")
22:58:49preglowno, you need to use long long
22:59:02preglowso the calc doesn't overflow before the divide by tenm
22:59:15Mikachulinuxstb: ah okay
22:59:24Mikachulinuxstb: i thought someone made some progress in figuring out the bits too
22:59:42 Join t0mas [0] (
23:00:05*t0mas mumbles something about bad ISP's
23:00:10t0maslike Planet Internet form the Netherlands...
23:00:10preglowsafetydan: however, you can do it like ((precut << 16)/10) << 7 with no loss in precision, i think
23:01:02Bagderhey t0mas
23:01:08Mikachuif you make the ui use binary, you could save some divisions :)
23:02:19preglowsafetydan: forget that, of course there'd be some loss of precision
23:02:40preglowmikachu has a point!
23:03:06Mikachui do? i was mostly joking
23:04:13preglowso am i
23:04:20Mikachui know :)
23:04:29preglowi usually do whenever i throw around exclamation marks
23:05:24safetydanhrmm... according to the comments in convert_gain it bottoms out at -23
23:06:19herz42safetydan: is it too inaccurate to multiply with an approximated 1/10?
23:06:37preglowyes, don't go that way
23:06:43preglowjust use a 64 bit int if that's the whole point
23:06:47preglowthis isn't speed critical code
23:07:03 Join [IDC]Dragon [0] (
23:07:46t0masBagder? hands off the script?
23:08:02safetydaneh, got what I want from get_replaygain_int
23:08:14safetydanokay seems to work, so time for a commit
23:08:17Bagdert0mas: I added a buildmaster log link in all server stats pages, as you might've seen
23:08:38t0mashave you checked m68k and sdl for Cassandra?
23:08:46 Join quobl__ [0] (n=quobl@tor/session/x-16640c035e74a8a8)
23:08:48Bagderyes, they're enabled again
23:08:54preglowcan anyone tell me what the fp_mul code in replaygain.c is up to? i don't understand why it does stuff so complicated
23:08:57t0masok, I'll remove the comment
23:08:59Bagdert0mas: check this =>
23:09:15Bagdert0mas: where the server 'arm' is the interesting one
23:09:34*t0mas searches for arm...
23:09:36t0masnothing found
23:09:56safetydanpreglow, possibly because they're a port of Java code :)
23:09:59 Quit miner49er ("CGI:IRC (Ping timeout)")
23:10:06 Quit quobl (Remote closed the connection)
23:10:47preglowsafetydan: sweet moses
23:11:02Mikachuwho would port java code? java code is supposed to be burned
23:11:06safetydanoMathFP is a J2ME fixed point library
23:11:40***Saving seen data "./dancer.seen"
23:11:40 Join solexx_ [0] (
23:11:41safetydancomment at the top of the file says the routines are based on that so...
23:12:29Mikachuhooray for html tags in comments
23:13:58safetydanyeah good ol' javadoc
23:14:15preglowthe fp_mul code doesn't seem to handle the case of LONG_MIN*LONG_MIN very well
23:14:35t0masBagder? where is that arm example you wanted to give?
23:14:39preglowlucky for us we'll never see those values
23:14:42t0masI don't see any server calling itself arm
23:15:16Bagderread it again
23:15:27*t0mas tried to find it in the buildlog
23:15:37Bagderno, the buildlog doesn't show it
23:15:50Bagderbut I'm quite sure it happens because of the 'retry' functionality
23:16:27Bagdert0mas: you can select other recent builds with 11 servers to see the same effect
23:16:56t0masthat's a never done build
23:17:04t0masit reports that there is 1 build not done
23:17:06t0masin the end
23:17:23Bagderit still shouldn't claim such a server name
23:17:33Bagderbut yeah, I've noticed that too
23:17:37t0masI'll change it
23:18:48t0massleep(10) = 10 seconds of waiting in perl?
23:19:07Bagder"perldoc -f sleep"
23:19:11 Quit Chamois (" HydraIRC -> <- Go on, try it!")
23:19:56 Quit Nico_P (Remote closed the connection)
23:20:02t0mashm... can I read access a var in perl without synchronizing with other threads?
23:20:15BagderI don't know
23:21:43godzirraSo out of curiousity, how do I set the time on rockbox?
23:21:48godzirraIts off by like 30 minutes =)
23:21:59preglowt0mas: what perl version?
23:22:08t0masnew enough for the new thread version
23:22:14preglowahh, ok
23:22:26preglowyeah, i believe you can
23:22:39t0masI tought I have read about it too
23:22:43t0masbut I'm not sure...
23:23:13Mikachugodzirra: settings etc etc set time
23:23:18t0masok, nasty one...
23:23:24t0mascan I use some goto like thing in perl?
23:23:40 Quit solexx (Read error: 110 (Connection timed out))
23:24:19godzirraI really wish it would stop resetting my wps.
23:24:47godzirraback to rebooting endlessly with a wall charger.
23:24:48Bagdert0mas: there's a goto
23:24:58t0masyeah, how do I set a lable?
23:24:59t0maslike basic?
23:25:52BagderI don't remember basic nor how goto in perl works, but I would expect so
23:27:10godzirraIck.. goto?
23:28:03t0masfound it
23:28:09*t0mas is in a hacker mood
23:28:14preglowgoto works fine in perl
23:28:21preglowlabels are used in lots of contexts
23:28:36preglowthey're nice for nested while loops
23:28:57t0masperl even has a special goto statement
23:29:01t0mascalled "redo"
23:29:10t0masto redo a loop :)
23:29:45t0masBagder: servers now wait in a loop untill the builds are done
23:30:03t0masreporting their waiting status instead of the done state
23:30:15 Part yeahx
23:30:19t0maswe might want to change it... so only servers with not done targets wait... but that's far more work
23:30:23t0maswith no practical effect
23:30:53 Quit drspoon ("Chatzilla 0.9.72 [Firefox]")
23:31:01Bagderwell a server can take another server's failed build, can't it?
23:31:31t0masno, sorry...
23:31:38b00stis david dent here?
23:31:42t0masI meant that only servers capable of building some still left target should wait
23:31:49Bagderah, ok
23:31:51t0masno use for an ARM only server to wait when all arm targets are done
23:31:59 Join tianjing_ [0] (
23:32:02t0masbut hey... no practical benefit from that... just looks better :)
23:32:03Bagderstill no big deal
23:32:17 Quit safetydan ("Leaving")
23:32:28t0maswhere can I see the running log? still just dbglog?
23:32:45t0masI've setup some server to do sh1 builds
23:32:45Bagderit moves it when a build is completed
23:32:47t0maswhile it can't do that
23:32:54preglow7.3 seconds per build is completely respectable
23:33:09Mikachudo you rebuild even if the commit was in manual/?
23:33:11BagderI think we can officially say that we don't want P3 cpus
23:33:14Bagderin the farm
23:33:22BagderMikachu: no
23:33:23preglowtoo slow?
23:33:26*t0mas cries out loud
23:33:28Bagderpreglow: yes
23:33:39t0masp3 SMP 2x800 mhz iirc...
23:33:44t0masmy testing server
23:33:52Bagderwell a dual p3 could be fine
23:33:56t0masmight even be 900 ;)
23:34:06preglowwhy not just time a build and see
23:34:17t0masidle P3's can be fast...
23:34:22 Quit tianjing (Read error: 110 (Connection timed out))
23:34:26Bagdera 500Mhz p3 takes some 4-500 seconds
23:34:30t0masbusy dual xeons can be quite slow :)
23:34:53 Quit Aditya ("Trillian (")
23:34:54t0maswell... not just state "we don't want them"
23:35:04t0masbut do ask some questions before setting up new ones
23:35:30Bagdernah, I rather set a limit. a single P3? no thanks
23:35:48Bagderfor my own simplicity
23:35:52*t0mas spots something bad in the logs :)
23:36:04 Join Aditya [0] (
23:36:06t0masyour failure detection part... it reports a missing zip file
23:36:10t0masbut doesn't try to rebuild
23:36:28Bagderwasn't sure how far I'd go force rebuilds
23:37:19BagderI can see a need for some kind of protection for the case where all servers fail
23:37:37 Quit Rob2222_ ()
23:38:01Bagderneed to rebuild on failed zip then I guess
23:38:11 Join yeahx [0] (
23:38:19t0masyes, I think so
23:38:40 Join Rob2222 [0] (
23:38:49 Quit Rob2222 (Read error: 104 (Connection reset by peer))
23:39:03t0maswill you edit?
23:39:21Bagderyou go ahead
23:39:35 Join Rob2222 [0] (
23:40:00b00sthmm ipod_unplug.patch error on this one
23:40:21b00sthunk #1 failed at 3830
23:40:26t0masBagder: those ld errors... didn't get a rebuild
23:40:29t0mason what condition?
23:40:32b00stany coder of this here?
23:40:57obob00st: yup
23:41:36oboanymore details about the failure?
23:41:38Bagdert0mas: what you said, it doesn't rebuild when it failed getting a zip
23:41:59t0masyou check for a build to finish
23:42:01b00stcould make a screenshot
23:42:08t0masthat should report a failed build
23:42:42Bagdert0mas: the build finishes fine
23:42:50t0masoooh wait
23:42:52t0masld errors
23:43:00t0masthose are real errors...
23:43:12t0masmight be caused by some cvs commit
23:43:17Bagderah, yes
23:43:37*t0mas goes to find something else to test
23:43:49godzirraAnyone have the link handy that shows how to rebuild your itunes database? :)
23:43:53Bagderand in fact the lack of a zip can also happen in legit case
23:43:58Bagderin a
23:44:11herz42hmm, think I know the reason for the 'backlight is always on if switched off' problem on ipods
23:44:14godzirranevermind.. found it :)
23:44:32b00sttakes too long for me to type..
23:45:19herz42gcc for arm doesn't have -fsigned-char, so the -1 in the table for off gets 255 -> timeout of 4+min
23:45:48obohehe - okay, I guess it's failing after the commit at 21:20 - give me a minute or two and I'll re-do it
23:46:21preglowherz42: the arm char is signed by default, afaik
23:46:31herz42but if it is changed to signed, another recently introduced bug is revealed: keypresses don't work anymore if the backlight is statically switched off (that should be the case for other targets also)
23:46:32t0masBagder: I've just made fail a h120sim
23:46:48t0masit should be rebuild by the first servers that looks for sdl builds
23:46:49b00stpowermanagement_20060316.patch is also failing now...
23:46:53herz42preglow: I have changed to signed and backlight goes dark...
23:47:06godzirraIs there a way to boot into the ipod firmwarE? (normal firmware, not emergency disk mode?)
23:47:10preglowperhaps that's where our current backlight bug stems from
23:47:53herz42but when fixed, only scrolling works on ipod (and long keypresses) , no normal pushing anymore
23:47:54b00stsees, that there is something changed with english.lang
23:48:16godzirraanyone know?
23:48:18b00stsame with the unplug patch..
23:48:19Mikachub00st: open english.lang and english.lang.rej in an editor
23:48:31herz42what is caused by "if (!remote_filter_first_keypress || is_remote_backlight_on()" in button.c
23:48:42obogodzirra: hold menu as the apple logo appears
23:49:28b00stand than mikachu?
23:49:36Mikachui was hoping it would be obvious
23:50:03b00stu know, im a noob still :D
23:51:14b00stand ther eis no lang.rej
23:51:57b00stonly the original
23:52:20Mikachuright, look in the .patch instead
23:52:25b00stmaybe cuz i did −−dry-run?
23:52:29Mikachuand just add the lines starting with a +
23:52:35Mikachubut remove the +
23:52:43t0masBagder: while having killed 2 builds... it still made it
23:52:53t0masonly in a bit longer time
23:53:13t0maswith some bugs left
23:53:30t0mas176 Server finished, but 2 builds left. Sleep and recheck.
23:53:30t0mas181 Server done (DBG left=0).
23:53:50t0mas303 Server finished, but 1 builds left. Sleep and recheck.
23:54:01godzirraAck! Anyone have a copy of foobar 2000 0.8?
23:54:03 Part yeahx
23:54:07t0masso at T + 181 there were no builds left...
23:54:08godzirraI can only find 0.9 and foo_pod wont work with 0.9
23:54:13t0masand at T + 303 we had 1 left
23:54:19b00stmikachu: there are no lines with +a in front
23:54:26b00sti searched for it
23:54:40Mikachuhm, is it a patch with lots of < and >?
23:54:51Mikachuthey're impossible to read imo
23:54:55 Join OPP [0] (
23:54:58 Quit imphasing ("Lost terminal")
23:55:01b00stno, not really
23:55:03OPPanyone here working with X5
23:55:12Mikachuwell, going to bed now
23:55:14Bagdert0mas: that's because the last failure occured at 300
23:55:27 Quit tucoz ("Leaving")
23:55:31OPPi just want to know if the X5 will have no file limit and ID3 tag databasing
23:55:31t0masah ok
23:55:34b00stok, thx anyways, gn8
23:55:46*t0mas edits
23:55:51b00stmaybe there will be some fixes tomororrow
23:56:34b00stim away too...tomorrow guys and good night
23:56:50 Quit b00st ()
23:57:14t0masBagder: fixed
23:57:23t0masit only counts them as done when they really are done
23:57:42linuxstbOPP: Rockbox no file limit and ID3 tag databasing - so the X5 port will as well.
23:57:47Bagderwell, it makes sense to show how many there are in the queue to build as well
23:58:02Bagderwhich it then showed before
23:58:04Bagderbut nevermind
23:58:10OPPk just making sure, thanks linuxstb
23:58:20OPPkeep up the good work
23:58:31*t0mas goes to bed
23:58:38t0masc u tomorrow
23:58:58OPPi cant wait for id3 databasing my x5 looks messy

Previous day | Next day