Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2007-02-28

00:00:08pixelmadan_a: do you want another hwcodec tester too?
00:00:25dan_apixelma: The more the merrier!
00:00:26linuxstb_Does the patch affect code-size on single-core targets?
00:00:45dan_alinuxstb_: Yes - it should shrink the core by about 100bytes
00:01:08linuxstb_How do you manage that? Did you add extra code in the past, just so you can remove it and gain points?
00:01:32dan_aYes - I waited until there was a table I could get green points on!
00:03:33LinusNdan_a: what an odd patch, you don't use svn?
00:03:53LinusNdan_a: 2 failed hunks
00:04:16dan_aLinusN: I use git for local development, because branches are really easy to do so I can work on 10 different things at once
00:04:43LinusNi use 10 repos for that
00:04:48dan_aI bet the failed hunks are the change I've just posted
00:04:59LinusNmain.c and config.h
00:05:06preglowdan_a: sounds clever
00:05:07dan_aI wanted to save bandwidth on the SVN server
00:05:15preglowi also want to have loads of repos, but i don't have space for ten checkouts
00:05:29dan_aI'll update the patch
00:06:31pixelmadan_a: the 3g build has some red...
00:10:50 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
00:10:54 Join miepchen^schlaf [0] (
00:12:09preglowwhen you bloody hire an arm programmer that has not yet learnt the use of the post increment addressing mode
00:12:14preglowyou hired the wrong guy
00:12:30*preglow gazes at portalplayer
00:12:45 Quit Thundercloud_ (Remote closed the connection)
00:14:39 Join Thundercloud [0] (n=thunderc@
00:14:56saratogathe bootloader isn't written in c?
00:15:31preglowif it is, i sure hope it wasn't ads they used
00:15:54preglowbeing the Offical arm compiler
00:15:58preglowthey say it's very good
00:16:10Lloreanamiconn: I remember you had a decent reason why Play shouldn't go to the WPS from the menus, but I can't remember what it was.
00:16:16 Part Domonoky_
00:16:29 Quit secleinteer (Remote closed the connection)
00:16:44 Join secleinteer [0] (n=scl@
00:17:32 Quit lee-qid_ (Read error: 104 (Connection reset by peer))
00:17:40pixelmaLlorean: maybe because "play" or even more accurate is "select" on Ondios from the menus
00:18:17pixelma"play" = "right" , I mean
00:18:57 Quit Everybody ("( :: NoNameScript 4.02 :: )")
00:19:00pixelmabut I'm not sure if it was his reason
00:19:12Lloreanpixelma: Well, by "play" I refer to the "resume playback and/or re-enter the WPS button"
00:19:44pixelmaok.. then I don't know
00:19:50LloreanRather than the button specifically marked "play", it just so happens to align with the play button on most of the targets I own.
00:19:54LloreanI suppose I should refer to it as "resume"
00:20:08LloreanBasically, I seem to recall there being a good reason not to resume from the menus, but can't recall it right now.
00:21:32LinusNdan_a: compiling...
00:22:10LinusNdan_a: /home/linus/rockbox2/firmware/export/config.h:360:1: warning: "CURRENT_CORE" redefined
00:22:53 Quit printfXh4 (No route to host)
00:23:26dan_aLinusN: I'll get onto it - that's what happens when you rush through updating a patch because you've got a red build table
00:23:31pixelmasame here - building for OndioFM
00:24:13LinusN/home/linus/rockbox2/firmware/export/config.h:284:1: warning: this is the location of the previous definition
00:27:13preglowLinusN: i2c device addresses are only 7 bits, yes?
00:27:40preglowi wonder what this ipod bootloader precaution of clearing bit 16 is, then
00:28:15preglowi've basically found and deciphered both i2c sender and receiver functions, now
00:28:30preglownow to find out where they're used
00:28:46preglowthere are a couple of differences to what we do in our drivers
00:29:01preglowwell, in the one we have that resemble the one in the bootloader, at least
00:29:58 Join donvito [0] (
00:31:07 Quit bluebrother ("leaving")
00:31:15saratogapreglow: this is using the emulator or digging through the dissassembled binary?
00:31:17 Join Thundercloud_ [0] (n=thunderc@
00:31:19preglowsaratoga: the latter
00:31:31preglowi just did an objdump and annotate the source code as i go
00:32:02 Join Febs_ [0] (
00:32:04 Nick Febs_ is now known as Febs (
00:34:06preglowseems i've found the equivalent of our pcf50605_read too
00:34:17preglowso at least i'm getting closer to what i want
00:34:34dan_aLinusN: The latest version compiles cleanly on several PP and non-PP targets
00:34:47 Quit mirak (Remote closed the connection)
00:37:32 Quit CriamosAndy ("( :: NoNameScript 4.03 :: )")
00:37:38 Quit Mouser_X (Nick collision from services.)
00:37:47 Join Mouser_X [0] (
00:39:58 Quit midkay (Read error: 104 (Connection reset by peer))
00:40:34preglownew patch from tomal
00:40:46 Join midkay [0] (n=midkay@rockbox/developer/midkay)
00:44:01dan_apreglow: FWIW, I've just got checked, and my git repo is 45MB smaller than my SVN repo, and that's with 14 different branches
00:44:40preglowdan_a: svn keeps a full copy of the repo for diffing
00:44:44preglowdan_a: so it doesn't surprise me
00:44:50dan_aSo does Git
00:44:53Mikachudoesn't git keep a full copy of all revisions even?
00:45:16Mikachupossibly better compressed
00:46:09nlsisn't svn totally uncompressed on disk?
00:47:09 Quit Thundercloud (Read error: 110 (Connection timed out))
00:47:23nlswhich makes grepping the source som much fun! :-P
00:47:54 Join Everybody [0] (
00:48:07Mikachunls: | grep -v .svn
00:48:26Mikachuor use ctags, then you almost never have to resort to grepping
00:50:56 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
00:51:48pixelmadan_a: the kocop patch seems to not break anything here - anything you want me to test in particular?
00:52:19LinusNdan_a: works fine on fmrecorder - rolo too
00:52:28dan_apixelma: I'm pretty sure that if audio works everything should
00:52:47dan_aLinusN: Do you have any problems with the general approach?
00:53:31dan_a(And thank you to both of you for testing)
00:53:40JdGordonmorning all
00:54:08JdGordon7 new emails in tehe rockbox filter.. 6 from the one fs page :p
00:54:22LinusNdan_a: i haven't reviewed the code
00:55:58 Quit ender` (" When I was a child... We had a quick-sand box in the backyard... I was an only child... eventually.")
00:56:11roolkuHi JdGordon, I hope you don't mind me fixing the radio menu - it bugged me when I wanted to copy the mechanism and it didn't work as expected...
00:56:19 Quit linuxstb_ ("CGI:IRC")
00:56:21nlsJdGordon: I reported two more menu things in the thread
00:56:49 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
00:56:53JdGordonroolku: I havnt looked at last nights commits yet, but no problem :)
00:57:42 Join Ribs [0] (n=ribs@
00:58:20JdGordonand thanks nls
00:58:31linuxstbdan_a: I've just tried your latest kocop11.2 patch, and it breaks rolo on my Color.
00:58:47LinusNlinuxstb: party pooper
00:58:55linuxstbEverything I try to rolo (Rockbox, diagmode and diskmode) just freeze on Executing....
00:59:01*dan_a looks around for forehead
00:59:09linuxstbLinusN: He only fixed it 20 minutes ago...
00:59:14*Mikachu provides dan_a with a surrogate forehead
00:59:21*linuxstb will test again...
00:59:25*dan_a slaps surrogate forehead
01:00:23dan_alinuxstb: I've not tested rolo with the latest kocop - I've been developing the rolo fix independently of that
01:01:09Mikachui got a Checksum Error on apple_os.ipod
01:02:14Mikachudiagmode and diskmode works, and diskmode actually connects too
01:02:20 Nick Everybody is now known as Everybody|away (
01:02:25linuxstbWith the COP patch or without?
01:02:38 Join Tacoman359 [0] (i=40fc2ae8@gateway/web/cgi-irc/
01:02:50linuxstbYes, they all work for me as well without the COP patch.
01:02:58Mikachuapple_os too?
01:03:08linuxstbNo, that's never worked...
01:03:43 Join Febs_ [0] (
01:03:50 Quit obo ("bye")
01:04:11Tacoman359can anyone point me towards a working half-life doom addon? i have copied the one off of my half-life cd but it said the wad doesn't follow the standard doom graphics
01:04:31scorcheTacoman359: no, we cant
01:04:47Tacoman359ahh k
01:06:14 Quit Tacoman359 (Client Quit)
01:06:50Mikachuit would be swell if rolo turned off the backlight too, diskmode doesn't
01:08:46linuxstbMikachu: Yes, I was thinking the same.
01:09:15*Mikachu considers reverting the 'boot OF instead of diskmode on usb cable insert' hack
01:09:41Mikachudo you know if diskmode enables the higher current charging mode?
01:10:40linuxstbNo idea, but it's how I normally charge.
01:10:57Mikachui suppose it isn't painfully slow then
01:11:14 Quit entheh ("^~")
01:11:45dan_aIf audio isn't playing, what does the COP thread do?
01:11:47linuxstbI normally just leave it charging overnight, so haven't noticed.
01:13:27 Join fasmaie [0] (
01:14:47*LinusN goes to bed
01:14:59 Part LinusN
01:16:58 Join Thundercloud [0] (n=thunderc@
01:18:04 Join Thundercloud__ [0] (n=thunderc@
01:20:54***Saving seen data "./dancer.seen"
01:21:57 Part toffe
01:22:17 Quit Febs (Read error: 110 (Connection timed out))
01:22:53 Quit m0s- ()
01:23:00 Quit Ribs ("eh eh ehhhh!")
01:23:58 Join Febs__ [0] (
01:23:59 Nick Febs__ is now known as Febs (
01:26:56 Quit Mouser_X (Read error: 110 (Connection timed out))
01:27:00 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
01:27:20 Join miepchen^schlaf [0] (
01:30:38 Quit jhMikeS (Nick collision from services.)
01:30:44 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
01:32:32 Quit inversions (Read error: 110 (Connection timed out))
01:34:11pearldivercould anyone try to play FLACs with crossfade please?
01:34:21 Quit Thundercloud_ (Read error: 110 (Connection timed out))
01:35:22 Quit saratoga ("CGI:IRC (EOF)")
01:35:27 Join Thundercloud_ [0] (n=thunderc@
01:40:08 Quit Thundercloud (Read error: 110 (Connection timed out))
01:42:05 Join rotator [0] (n=e@rockbox/developer/rotator)
01:42:30 Quit Febs_ (Read error: 110 (Connection timed out))
01:43:31linuxstbpearldiver: I'm trying now - what am I looking for?
01:43:46pearldiverartifacting at the start of the next track
01:44:03pearldiverhalf a second sound, might be a bit of the previous track
01:44:20pearldiverwas introduced around 0223, 24
01:44:30pearldiverdont have any problems with 0222 build
01:44:42 Part nls
01:46:01 Quit Nico_P (Remote closed the connection)
01:46:07 Join Mouser_X [0] (
01:47:38linuxstbpearldiver: I can't notice anything (on an ipod).
01:47:51 Join Thundercloud [0] (n=thunderc@
01:48:26pearldiverleave it for a few tracks
01:48:59linuxstbSo it doesn't happen on every crossfade?
01:49:11pearldiverit happens for me every 2nd one or so
01:52:06pearldiver"scratchy" sound
01:52:20pearldiverwhat are your settings for crossfade?
01:54:23 Quit Thundercloud__ (Read error: 110 (Connection timed out))
02:01:50 Quit fasmaie (Read error: 54 (Connection reset by peer))
02:02:12 Quit darkless ("Leaving")
02:02:18 Join darkless [0] (
02:05:21linuxstbpearldiver: I never use crossfade, but for these tests I just set everything to 3 seconds, and fade-out mode to mix. I still haven't noticed any problems...
02:05:33XavierGris there a new H300 bootloader?
02:05:42pearldivermy fade-out mode is crossfade
02:06:02 Quit Thundercloud_ (Read error: 110 (Connection timed out))
02:06:16 Join lost|X40 [0] (
02:06:36linuxstbpearldiver: OK, I've changed it.
02:07:03 Quit lost|X40 (Client Quit)
02:07:09 Join lost|X40 [0] (
02:08:03 Quit barrywardell ("Konversation terminated!")
02:09:03*linuxstb wonders what the first characters in the latest commit message should be
02:09:27 Join ncognito [0] (
02:09:32fejfighterXavierGr: from the wiki, it appears not
02:09:46fejfighterjust an update to the svn
02:09:51 Join billodo_ [0] (i=billodo@nat/sgi/x-5ca657df9c4597b0)
02:10:05roolkulinuxstb: sorry, I c&p the hyphen from flyspray
02:10:26XavierGrhmm I wonder if I should compile my own bootloader?
02:10:33 Quit Mouser_X (Nick collision from services.)
02:10:36 Quit billodo (Read error: 113 (No route to host))
02:10:39linuxstbroolku: No problem.
02:10:42 Join Mouser_X [0] (
02:11:05roolkuI tried to edit it, but got the reply "Repository has not been enabled to accept revision propchanges; ask the administrator to create pre-revprop-change hook"
02:11:29XavierGrbah I won't do it, I am a chicken into these sort of stuff
02:11:34linuxstbpearldiver: Yes, I just got a little bit of corrupt audio...
02:11:50linuxstbSo it seems the crossfade setting is needed, rather than mix.
02:12:19 Quit ncognito (Remote closed the connection)
02:12:23linuxstbpearldiver: Did you try other codecs?
02:12:39pearldiverlinuxstb tried mp3s, no problems
02:12:56 Quit lostnihilist (Read error: 60 (Operation timed out))
02:12:58 Join ncognito [0] (
02:14:06pearldiversometimes the corruption sound is quite loud
02:14:09linuxstbDo you have replaygain enabled?
02:14:19pearldiverand i noticed that it might be a bit of a previoys track
02:14:27 Quit roolku ()
02:14:29pearldiverno replaygain
02:14:46linuxstbI do, so we can rule that out.
02:14:52 Join billodo [0] (i=billodo@nat/sgi/x-8450591abd8b90d0)
02:16:19linuxstbYou said the 22nd Feb daily build is OK? What about the 23rd?
02:17:01pearldiveri didnt build anything that day
02:17:25pearldiveronly 0222 and 0224
02:17:30pearldiver0224 has the problem
02:17:35pearldiver0222 doesn't
02:17:44linuxstbYou could try the 23rd daily build.
02:17:58pearldiveryeah let me try it
02:18:08linuxstbBut jhMikeS's big commit on the 24th seems a likely candidate.
02:19:43pearldiveri see his changed for coldfire
02:20:13preglowlinuxstb: not surprising, is all
02:20:23linuxstbI think they affected all targets, but he also added coldfire optimisations.
02:21:57linuxstbBut I've only heard a glitch once, all the other transitions since that one have been fine.
02:21:58pearldivertrying 0223
02:23:20preglowthe bootloader does a shitload of writing to the pcf
02:24:00linuxstbOff-topic question, but does anyone know why I can't email a sourceforge mailing list I'm the admin for? My emails just disappear, they don't even appear in the admin interface...
02:24:19preglowsourceforge for you
02:24:26XavierGrbah iriver H300 bootloader won't let me scramble it onto the final H300.hex file
02:24:33XavierGrit says incorrect file size
02:25:30linuxstbI seem to recall someone saying it's now too big...
02:25:59XavierGrdamn it if only was here to explain to me his recent change in the bootloader
02:26:25preglowi find the code that reads the pcf int regs
02:26:43pearldiver0223 - no problems, smooth
02:26:44preglowbootloader digging is more fun when stuff starts to trickel out
02:26:44XavierGrI am very itchy about the H300 bootloader and the USB problem is a real hassle
02:26:56preglowtricle too
02:27:09 Quit billodo_ (Read error: 110 (Connection timed out))
02:27:17XavierGrbut I will have to wait... even if I want to I can't flash it right now, the scramble tool just won't let me to do it
02:27:33pearldiveralso, it takes like 4 seconds for EQ to kick in while playing FLAC?
02:28:00preglowpearldiver: it always does that unless you've spent some time in the eq screen
02:28:01pearldiverlinuxstb 0224 is where it all starts
02:28:13pearldiverthe corruption
02:28:45pearldiverpreglow but if i just do EQ on / off, same?
02:28:45pixelmahmm... seems like no one has corrected Bjrn Stenberg in onplay.h (for Kate ...err... JdGordon ;) ) so far but I'm about to leave...
02:29:04LloreanXavierGr: For the H300 and H100 it's not wise to compile your own bootloader.
02:29:24linuxstbpixelma: OK, I'll do it.
02:30:25Mikachuan apparently not perfect text editor
02:30:29linuxstbHmm, it looks fine to me - iso8859-1 encoded like all the other files...
02:31:52linuxstbAh yes, plus main.c, screens.c and sound_menu.c....
02:32:17XavierGrLlorean: yup, but if things are like the wiki says it is pretty straight forward
02:32:19linuxstb... and gui/gwps-common.c
02:32:31XavierGrLlorean: but something is wrong with the bootloader code atm
02:33:01XavierGrthe new bottloader.bin file (compiled from current svn) is wrong, cause the old one still builds fine
02:33:09XavierGrso i will have to wait for linus
02:33:13LloreanXavierGr: Even if it's straightforward, if anything at all goes wrong you have a bricked unit. Better to wait for a binary bootloader that you know will work.
02:33:45XavierGrLlorean: of course you will have a bricked player, that's the sweet risk you take when you flash it alone :P
02:33:50pixelmalinuxstb: thanks for looking a bit closer too :) I'm really off now
02:34:19 Part pixelma
02:34:43XavierGris there any md5 built in command on linux?
02:34:57Mikachuwould you believe it is called 'md5sum'?
02:35:01LloreanXavierGr: The thing is, the flash can succeed, but if what you've compiled won't execute properly for some reason, it'll also be bricked.
02:35:18XavierGrLlorean: yes I realise that
02:36:02XavierGrflashing with impoper file or code will end up bricking the player.
02:36:42XavierGrthat's why most of us are waiting Linus to try it first because he can unbrick it with his magic :P
02:37:38LloreanI was just warning because it seemed like you were trying to.
02:38:06 Join blithe_ [0] (
02:38:14LloreanI can't imagine bootloader USB mode being so important as to even take a very faint risk of bricking a player.
02:38:24XavierGrI was trying to built the bootloader, I know very well what will happen if there is something wrong with it
02:38:49XavierGrbut even if I would take the risk (which I won't) I can't make the final bootloader file
02:38:50preglowlinuxstb: the very first bloody thing the bootloader does is enable wake on alarm...
02:39:22XavierGrbecause the tools won't let me do it (there is obviously an error)
02:39:30JdGordoncan someone paste the o with the dots please?
02:39:40linuxstbpreglow: Now you need to write yourself a flash driver and patch the bootloader...
02:40:01preglowlinuxstb: exactly!
02:40:33linuxstbDidn't I say rtc alarm would be easy?
02:41:06XavierGrnow that I think of it I remember Linus saying something about the bootloader size
02:41:19XavierGrbut I can't remember exactly what he said about it
02:41:34preglowlinuxstb: i would have been done days in the past if had taken the sane route and just implemented date increment logic
02:41:41preglowwhat is that, five lines of c code?
02:41:57JdGordonyou got the alarm working?
02:42:16Mikachuyeah, mine and linuxstb's ipods wake up every day at 10pm now
02:42:19preglowJdGordon: yeah, long ago, i just need a reliable way to figure out if we woke on alarm
02:43:21 Quit blithe (Connection timed out)
02:43:22XavierGrwell anyway linus' commit seems to be just a minor reorganization in the bootloader sequence, I hope it does the trick in the end
02:43:53preglowwhat, the alarm disable code i gave you didn't work?
02:44:02Mikachui didn't try it
02:44:10preglowthen no wonder it always wakes
02:44:16 Join TwilightInZero [0] (
02:44:18Lloreanpreglow: Why exactly do we need to know the alarm woke it?
02:44:25MikachuLlorean: so we can turn it off
02:44:28TwilightInZeroHi there, guys.
02:44:30Mikachuthe alarm, that is
02:44:53LloreanWe could just turn it off, if it's on, no matter what.
02:44:55*JdGordon looks forward to alaram on the h300 :p
02:45:13Shaidjust check when rockbox is loading whether time now => alarm time
02:45:17MikachuLlorean: you mean if you set the alarm and reboot, the alarm is unset?
02:45:17Shaidif so, turn it off
02:45:33linuxstbpreglow: So you don't set the date fields in the alarm at the moment?
02:45:35LloreanMikachu: Yes.
02:45:41Shaidthough that'd stop it turning on for the next day.
02:45:48preglowlinuxstb: if i were to do that, i would have to calculate them first
02:46:20Shaidmaybe if it's within >5mins of alarm time turn it off.
02:46:28LloreanMikachu: Or, never unset it automatically, and put a menu entry for it.
02:46:29Mikachuactually i don't know if it turns on every day, i always have it plugged in
02:46:30preglowfor some absurd reason, i don't want any date increment logic, but i see i'll probably need it
02:46:57linuxstbDoesn't the date/time setting screen already do that?
02:47:11preglowi set alarm seconds to 00 and minute and hour as specified, and mark the rest as irrelevant
02:47:20JdGordonhang on.. whats so hard about date increment logic?
02:47:26preglowit's not hard
02:47:57preglowlinuxstb: i haven't seen any code like that, at least
02:48:05preglowbut now that you mentiond it, i think amiconn talked about it once
02:48:46preglowmktime has some such logic, i see
02:48:51preglowbut i can't use that
02:49:05TwilightInZeroYou guys seem to be involved into some important code-related conversation, so I'll go ahead and ask my question so you guys can get back on that...
02:49:09XavierGrbah! I just read in the logs that Linus send a new bootloader for the USB problem but that person didn't replied to it
02:49:50TwilightInZeroI'm having major issues setting up a compiling environment. I first attempted to use cygwin, but if I try to do anything at all in it, it says this: "bash: fork: resource temporarily available."...
02:50:05XavierGrah mikachu thanks about the md5sum command
02:50:19linuxstbTwilightInZero: Do you have two cygwin.dll files on your computer? That's a common cause of cygwin trouble.
02:50:22TwilightInZeroThen I tried the VMWare one, and it worked alright, except that Samba didn't work at all. In other words, \\debian\user gave an error in Windows Explorer.
02:50:27TwilightInZeroTwo of 'em?
02:50:37TwilightInZeroI'll give a quick search real quick.
02:50:56TwilightInZeroTo finish off my question, I last tried coLinux, but it couldn't connect to the SVN server.
02:50:57SoapTwilightInZero: the most common cause of Samba problems is ZoneMFingAlarm.
02:51:15scorcheSoap: heh
02:51:27TwilightInZeroSoap: I don't have ZoneMFingAlarm. ;)
02:51:36Soapfor the coLinux install did you enable ICS on your primary network connection?
02:52:00*linuxstb thinks maybe TwilightInZero should choose one system to debug...
02:52:09TwilightInZeroSoap: I'm not sure about that. I'll check real quick while I'm running this search
02:52:15*TwilightInZero probably should, but he's not that smart. :P
02:52:38linuxstbI guess there's more chance of a solution if you have three problems.
02:52:49TwilightInZeroI just want to get one working.
02:53:02TwilightInZeroIt doesn't matter to me which one.
02:53:14Soapget them all working and report back some speed benchmarks!
02:53:20linuxstbpreglow: I'm not sure how it works, but the data/time settings screen seems to have a lot of knowledge about dates.
02:53:30preglowi'll have a look once i'm through with this
02:53:57TwilightInZeroGood idea. Ah, here we go... ICS is disabled. Shall I go ahead and check that off?
02:54:19SoapICS needs to be enabled on your primary network connection.
02:54:33 Join pabs_ [0] (
02:54:50TwilightInZeroAlrighty, then... I'll enable it and try coLinux again, post-haste.
02:55:32linuxstbpreglow: timedate_set() in menus/settings_menu.c
02:57:04LloreanOh, hey, MP3 optimizations from Tomal
02:57:45 Join Thundercloud_ [0] (n=thunderc@
02:59:29TwilightInZero"Post-haste" is gonna last longer than I expected... I have to to move the VMWare image onto my iPod so that I have room for the coLinux one. Sorry, guys.
02:59:54*linuxstb sleeps
03:00:16SoapLlorean: where?
03:00:19LloreanSoap: Tracker
03:00:31LloreanThe very topmost item
03:01:11 Join aliask [0] (
03:01:43 Quit billodo (Read error: 104 (Connection reset by peer))
03:02:40preglowthe bootloader, that turd, just reads the int flags and does nothing with them
03:02:46preglowdate increment logic it is
03:03:51 Join Thundercloud__ [0] (n=thunderc@
03:04:06LloreanAt least more is known about the flash bootloader now.
03:04:23preglownot that it'll help much
03:04:54linuxstbIf we want to flash rockbox, we need to document all the hardware init'ing it does...
03:05:04preglowtrue enough, i guess
03:05:19preglowat least i've mapped out whatever it does with the pcf fairly well
03:05:26linuxstbNow we can rolo the diskmode app, that's the biggest issue.
03:05:51preglowi guess i can do a proper i2c multiple read routine too now
03:06:04LloreanIs the diskmode app the one you get from Select+Play, or the one you get from connecting a USB cable in Rockbox?
03:06:10preglowalso, our i2c routine differ somewhat from the ones in the bootloader
03:06:14LloreanDid we ever determine if there was a difference between those other than the string displayed?
03:06:25 Quit pabs (Read error: 110 (Connection timed out))
03:06:28preglowthey're the same in different modes, afaik
03:06:37 Join billodo [0] (i=billodo@nat/sgi/x-34745024f78cf36f)
03:06:57linuxstbYes, I doubt there would be two separate disk modes in flash.
03:07:53linuxstbI've never noticed a difference, apart from the forced (SELECT+PLAY) disk mode requires a reset to leave.
03:08:37preglowbut ok, with date increment logic i can set all the alarm fields to the exact time
03:08:38LloreanThat and I believe one of them actually says "Disk Mode" and the other doesn't. Or some string. I can't remember what exactly it said right now, but it doesn't display in both.
03:08:43preglowso that the alarm will never trigger again
03:09:08preglowbut that still does not solve the problem of the ipod triggering the alarm, the user shutting it down immediately, then turning it on again
03:09:13preglowthe alarm will then trigger again
03:09:26linuxstbLlorean: Yes, I've just tested, and the "rockbox" disk mode doesn't display "Disk Mode"
03:09:52Lloreanlinuxstb: It doesn't seem likely they're actually different, I was just curious why there are any differences like that.
03:10:39linuxstbRolo'ing the diskmode app gives you the "Disk Mode" text...
03:11:32 Part ncognito ("WeeChat 0.2.3")
03:12:40 Quit bospaadje ("great minds run in great circles")
03:13:56 Join bospaadje [0] (
03:15:30aliaskI spy something to fix in svn! Barely deserves a commit though.
03:15:34 Quit Thundercloud (Read error: 110 (Connection timed out))
03:15:37SoapI hate to ask (not test) but with the recent commit can you Rolo apple firmware?
03:15:43aliaskIn roolku's last commit theres a comment which is wrong
03:16:04aliask /* Copy Album (assumed max 4 letters + 1 zero byte) */ should be Copy Year
03:16:47ShaidSoap: Don't think so.
03:16:49LloreanSoap: I don't believe so.
03:16:53 Quit fejfighter ()
03:18:23*JdGordon thinks we should have a comment-fixing party... quite a few wrong ones in the code :'(
03:18:44aliaskAnd that'd actually make it worth a commit rather than just a single fix :)
03:19:52JdGordoncant think of any other specific ones atm tho... :p
03:20:13 Part Llorean
03:20:57***Saving seen data "./dancer.seen"
03:21:28 Join Thundercloud [0] (n=thunderc@
03:25:12 Join fejfighter [0] (
03:26:56 Quit Thundercloud_ (Read error: 110 (Connection timed out))
03:28:45 Join toffe [0] (
03:32:21TwilightInZeroWhew... I got coLinux to download the current SVN.
03:32:40TwilightInZeroThank you very much, Soap.
03:33:01SoapI still suspect VMware's problems are firewall related.
03:33:11 Join Daxim [0] (
03:33:22Soapbut glad co works for you.
03:33:41TwilightInZeroMaybe. I'll experiment with that a little later. I still need to see if Samba works for coLinux.
03:34:12Daximhey there. I've read in the rockbox manual for ipod video that the original firmware can be restored
03:34:29Daximis that 100% safe? I don't want to brick a ~300 device :/
03:34:37Soaprestored as in removing the rockbox bootloader, or restored as in dualbooting?
03:34:51SoapOh, if you brick your iPod you would be the first.
03:34:54Daximrestored as in using apple's stuff again
03:35:05TwilightInZeroYou can switch back and forth whenever you want, Daxim.
03:35:07Soap300 euros? Is that the 80GB or the 30GB?
03:35:32TwilightInZeroDo you mean $300 US, perhaps?
03:36:35TwilightInZeroWell, regardless of that, you can boot into the Apple OS by flicking on the hold switch when you first boot and the apple logo appears.
03:38:54 Quit Thundercloud__ (Read error: 110 (Connection timed out))
03:39:27TwilightInZeroHmm... what do I do with that IP address, Soap?
03:39:42Soaphave windows explorer open it.
03:39:56SoapIt is the local address of the coLinux samba server.
03:41:28 Quit fejfighter (Remote closed the connection)
03:42:03TwilightInZeroI entered it directly into the address bar and Firefox tried to open it. :/
03:43:41TwilightInZeroI recieved "Windows can't find \\".
03:43:53 Quit Mouser_X (Read error: 110 (Connection timed out))
03:44:11SoapI gotta go, hmm
03:45:36TwilightInZeroWell, I don't want to hold you back.
03:45:54TwilightInZeroI'll look around elsewhere for the answer.
04:04:06TwilightInZeroArgh... I can't find anything in samba's config that'll help me. It doesn't help that I'm linux-dumb. :/
04:07:40 Quit Thundercloud (Remote closed the connection)
04:08:14 Join SirFunk [0] (
04:12:38 Quit perplexity (Read error: 113 (No route to host))
04:14:07Soapsamba should be configured right.
04:14:25SoapThe issue I believe is windows side.
04:14:39Soapfrom the command line can you ping
04:18:19TwilightInZeroYeah, I can ping it
04:18:32Soapwindows firewall up?
04:18:41TwilightInZeroIt is indeed up.
04:18:52Soapare you behind a NAT router?
04:19:06Soapdrop windows firewall then.
04:19:43TwilightInZeroIt's now deactivated...
04:19:51SoapI am sure there is a way to get this working while having windows firewall up, but if you are behind a NAT box (and not in the DMZ, and don't have silly ports forwarded) there is little risk to dropping it.
04:20:24TwilightInZero\\ didn't work.
04:21:58Soapfrom windows file explorer?
04:22:14Soapdoes Colinux show up in "My Network Places"?
04:23:09TwilightInZeroI tried it from windows file explorer, and it doesn't show up in My Network Places.
04:24:16SoapI'm currently fresh out of ideas (outside the obligitory REBOOT!) ;)
04:25:10TwilightInZeroI guess I could use that as a last resort. :/
04:25:21Soapit's a sucky answer.
04:26:14TwilightInZeroWell, I have to go.
04:26:20TwilightInZeroThank you for your help.
04:26:36TwilightInZeroGood night.
04:26:41Soapdo you have "Client for Microsoft Networks", "File and Printer Sharing for Microsoft Networks" enabled on the TAP adapter?
04:26:44 Quit TwilightInZero ("Chatzilla 0.9.77 [Firefox]")
04:27:58 Quit Daxim ("pro quo, Agent Starling")
04:30:36 Join perplexity [0] (n=joust@
04:47:32scorchei havent been paying attention to the conversation, but have you tried simply restarting the image?....that has worked for me a few times after samba doesnt work for some odd reason
04:55:51 Join fejfighter [0] (
05:02:24 Join daurnimator [0] (n=fake@unaffiliated/daurnimator)
05:11:37 Join printfXh4 [0] (
05:14:08 Quit perplexity (Read error: 110 (Connection timed out))
05:18:45 Quit pill (Remote closed the connection)
05:20:58***Saving seen data "./dancer.seen"
05:22:02 Join Mouser_X [0] (
05:26:57 Quit fejfighter (Read error: 110 (Connection timed out))
05:27:08 Quit Mouser_X (Nick collision from services.)
05:27:19 Join Mouser_X [0] (
05:32:32 Join alienbiker99 [0] (
05:36:32 Join fejfighter [0] (
05:37:36 Quit atsea-145193 (Excess Flood)
05:38:59 Join atsea-145193 [0] (i=ariel@gateway/tor/x-8c699f896f8f242e)
05:39:59 Quit rotator ()
05:50:25 Quit alienbiker99 ("( :: NoNameScript 4.02 :: )")
05:57:40 Join perplexity [0] (
05:58:00 Quit fejfighter (Read error: 110 (Connection timed out))
06:07:19 Join JerryLange [0] (
06:19:04 Quit Mouser_X (Nick collision from services.)
06:19:13 Join Mouser_X [0] (
06:23:46 Quit Febs (Read error: 110 (Connection timed out))
06:25:23 Quit Mouser_X (Nick collision from services.)
06:25:32 Join Mouser_X [0] (
06:26:33 Nick pabs_ is now known as pabs (
06:46:00donvitoif anyone here knows, any progress with the sansa E2xxR?
06:56:17printfXh4Why, that is a very good question, donvito.
06:57:12donvitoyes, yes it is
06:57:22donvitoare you a fellow R owner
06:57:30printfXh4I own a Sansa E260.
06:57:31printfXh4So no.
06:58:12donvitoi see, i have a e260r
06:58:27printfXh4What's the difference?
06:58:31donvitoi know that someones working on it, and probably making some progress, i just want in on it haha
06:58:45donvitor is for retarded.... just kidding
06:59:13donvitor is for the best buy only rhapsody edition, leave it to my parents to get the one that isnt rockbox compatible
06:59:20donvitobut hey, it was a very nice gift
07:02:13donvitoi meant the unit is retarded, not you
07:02:28donvitolol i thought about how that came out, no offense meant!
07:06:26printfXh4None taken. :P
07:06:42printfXh4I don't use Rockbox on my Sansa actually.
07:06:47printfXh4It doesn't look usable enough to me.
07:06:53printfXh4Once they have sound running, I might consider.
07:07:21donvitorgr that.
07:07:57donvitoi can't wait, i used to own an archos xxxxm can't remember, but i used to use rockbox way back in the beginning of it, i loved it!
07:08:00donvitoso yea, i can;t wait
07:08:10donvitorockbox > stock firmware, period
07:14:37daurnimatoryou can't use usbhost in rockbox
07:17:52scorchedaurnimator: then code us a usb stack
07:18:15daurnimatorport rockbox to my player!
07:18:28*daurnimator waits for JdGordon
07:21:00***Saving seen data "./dancer.seen"
07:24:02 Quit pearldiver (Read error: 110 (Connection timed out))
07:29:49*donvito is away "Hunting white sharks at the baltic sea" Log: on Pager: off
07:30:36scorchedonvito: please turn that script off for at least this channel
07:30:56 Join Shaid` [0] (
07:31:54donvitook sorry, i honestly thought it was the channel, but its my stupid client
07:32:06donvitowhats the accepted away protocol here?
07:32:13donvitolike _ before name or what
07:32:35scorchea nice silent /away should turn your nick grey in the userlist
07:32:49donvitorgr, ty.
07:34:38*amiconn wonders which client(s) do that
07:34:55donvitodid it do it again?
07:34:56scorchewell, it should do some sort of marking in the userlist on most clients
07:35:02scorcheoh...the script
07:35:09donvitoits its anaconda
07:35:42amiconnHydra doesn't show 'away' status; you only get to notice it when opening a query
07:35:53perplexityX-chat does
07:35:57donvitook for real i need to go away this time, haha, stupid class at 9, with a test, bye guys
07:37:25amiconnx-chat :\
07:37:42scorcheyou do not like xchat?
07:38:15 Join Shaid2 [0] (
07:38:21 Quit Shaid (Nick collision from services.)
07:38:31 Nick Shaid2 is now known as Shaid (
07:38:40scorchewhy not?
07:39:30amiconnI don't like its nick colouring scheme and its window layout. And the win32 version I once tried was ustable, crashing every few hours
07:39:47amiconnOn linux it's stable at least, hence usable
07:41:21scorchewell, did you compile it yourself?
07:41:41scorchei have had good luck with the silverex version if you are too lazy
07:41:58scorcheand the coloring can be changed
07:42:20scorchethe only beef i have with it, is that i cant move the tabs around
07:44:05 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
07:46:08 Quit amiconn_ (Client Quit)
07:48:13daurnimatorxchat is terrible :S
07:48:54amiconnWell it's usable but I'd miss a lot of the flexibility I have in Hydra
07:49:06scorchei should try hydra one of these days
07:50:19JdGordonkonversation ftw :p
07:50:44amiconnNo kde for me...
07:51:10 Quit Shaid` (Read error: 110 (Connection timed out))
07:52:16bagawkirssi for me :)
07:56:11qwmirssi. <3
07:56:17qwmkde is nasty
07:56:24qwmscorche !
07:56:27qwmi'm behaving!
07:58:20 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
08:01:14qwmhow's that neat mpegplayer progressing?
08:01:29JdGordonmornign LinusN, im sick of this patch already :p
08:01:39LinusNJdGordon: already??? :-)
08:01:48JdGordoni mean more than before
08:02:09XavierGrHello Linus
08:02:18XavierGrWhat's up with the update on the bootloader?
08:02:45XavierGrI tried to make the bootloader but the scrambler will refuse to create the hex file
08:02:45LinusNJdGordon: hmmm, to be honest, i don't care much if it returns to the root menu, since the browser location is remembered
08:03:00LinusNXavierGr: reconfigure
08:03:18XavierGrI think I did...
08:03:28XavierGrso is it safe to make the bootloader with your changes?
08:03:34XavierGrare you going to realease one?
08:03:41LinusNnot yet
08:03:54LinusNi'll fix rtc and charging first
08:04:01JdGordonand alarm?
08:04:03XavierGrah nice :) at last!
08:04:13LinusNJdGordon: rtc == alarm
08:04:19JdGordonyay :)
08:04:43 Join redbreva [0] (
08:04:45LinusNthe sad part is that i still don't fully understand the usb on the h300
08:05:25LinusNmy findings with the multimeter contradict the OF disassembly
08:05:41LinusNso i'll have to do some more research
08:05:46XavierGrdid the guy yesterday contacted you about the test bootloader you gave to him?
08:06:27XavierGrit was pondlife?
08:06:46LinusNor was there another guy?
08:06:47XavierGruhm I think it was someone else
08:06:54LinusNin the logs?
08:07:11XavierGrit was at the time that you asked someone for an H300 that has the USB bootloader problem
08:08:22JdGordonLinusN: oh, did you fix the nonlcd problem in the bootloader? (I tihnk it was that one which always thought hold was on?)
08:08:34LinusNJdGordon: not yet
08:08:49XavierGrLinus: so does your commit fixes the USB bootloader mode in the end?
08:09:40LinusNi only see pondlife there
08:10:06LinusNand yes, it worked for him
08:10:18XavierGrbah idiot I am! Sorry...
08:10:37XavierGrperfect! I am gonna test it too, when I manage to make the bootloader
08:11:10amiconnLinusN: Does it fix the nasty click?
08:12:10JdGordonLinusN: last thing.. are you going to add a killswitch for the bootloader this release?
08:12:16LinusNamiconn: yes
08:12:27LinusNJdGordon: possibly
08:12:48JdGordoncool.. just as long as its not really possible to activate it accidently :p
08:13:01LinusNwe'll see
08:14:10*amiconn wonders hy we would want that
08:14:24JdGordonfor warrenty returns
08:14:31JdGordonif the disk isnt accessable
08:14:43LinusNi.e fraud
08:15:58amiconnJdGordon: Well if the disk isn't accessible you can't remove rockbox from it either. So why worry about the bootloader?
08:16:17 Quit Mouser_X (Read error: 110 (Connection timed out))
08:16:20 Join decayedcell [0] (
08:16:27XavierGrbecause if the disk isn't readable then iriver wouldn't know :P
08:16:56JdGordonif the bootloader works but the disk doesnt, iriver will see the rb bootloader and deny the return
08:17:12JdGordonnot that it really matters anymore seen all hxxx's would e out of warrenty by now
08:17:22amiconnIf the disk isn't accessible that doesn't mean it can't be made working again
08:17:43XavierGrnot to mention that iriver won't be able to replace his unit with another H100
08:17:45amiconn...and then iriver would still see rockbox
08:18:17XavierGryes but still it would mean that they will have to repair the disk/board and then search his files
08:18:44DataGhostafaik they're not even allowed to do that
08:18:47LinusNamiconn: i still have to fix the click for the application
08:18:53DataGhostby company policy
08:19:12DataGhostbecause there might be illegal stuff on it and they don't want it to be their responsibility
08:19:28XavierGrDataGhost: well my (un)-repaired H300 came with some mp3s from the technician LOL
08:19:49LinusNXavierGr: any luck creating the H300.hex file?
08:19:52DataGhostwell that's forgivable
08:20:02XavierGrsucky though I immediately removed them nothing of my pleasure
08:20:17XavierGrLinus VMware refuses to svn update right now
08:20:43XavierGrI will let you know, though I might not manage to make it before I go to work
08:21:16*JdGordon is game - i mean.. crazy... Ill try a new bootloader
08:22:03XavierGrwell I am afraid to try it even if I manage to create the firmware file :P
08:22:06 Part toffe
08:22:51JdGordonare the instructions the same as whats on the wiki?
08:23:35XavierGrI think yes
08:23:47XavierGrit worked for me until the scrambling
08:24:21 Part decayedcell
08:24:33XavierGrLinus: It should work normally with all (official iriver) H300.hexes right?
08:26:01JdGordonscramble worked..
08:26:14XavierGrI guess I did something wrong
08:26:22JdGordonLinusN: have you got a md5 of the hacked .hex I can compare against just to make sure?
08:26:36XavierGrthe command was right it is just that the tool would give me a file size error
08:26:51LinusNXavierGr: that is wrong
08:27:02LinusNit should not give a file size error
08:27:20LinusNis your svn up-to-date?
08:27:21XavierGrlet me try again then
08:27:38XavierGrIt should be but I am just svn updating as we speak
08:27:59LinusNafter updating, you must run ../tools/configure again
08:28:10LinusNotherwise the binary will be too big
08:28:30JdGordonok... here we go...
08:30:39JdGordontick tock...
08:30:50JdGordonstupid thing takes a while doesnt it ? :p
08:31:09XavierGrhmm i think it will work now, because the bin is 3.99 and not 4mb
08:31:34JdGordonwell it booted back into rb so it worked :)
08:31:54XavierGryup I have the hex now
08:31:57XavierGrit seems that I didn't update my repository
08:32:03XavierGrdamn it and I was almost sure about it
08:32:14JdGordonlogging on the remote :) awesome
08:33:05JdGordonLinusN: ill keep an eye on usb and let you know if its any worse than before (mine either works for days on end or doesnt :p )
08:33:09XavierGrJdGordon: md5sum?
08:33:34JdGordon14c404d9366c6ca7a8d9353ecc043d19 h300.hex after patching the 1.30E
08:33:51XavierGrah I used 1.29 korean
08:34:04XavierGrok here it goes
08:34:09*XavierGr crosses fingers
08:35:12DataGhostaaaaaaaaand then the power went offline. too bad
08:35:38XavierGrit works :P
08:35:54DataGhostthat's what I thought this morning
08:36:00*JdGordon has a stupid version string in the bootloader now
08:36:16JdGordonversion: rootpatch-12500M or something :p
08:36:17XavierGrhmm strange why it doesn't respond to button presses when in USB bootloader mode?
08:36:25DataGhostand after that I noticed that I'd been up all night, as planned... but didn't do the planned work. instead, I kept on tinkering with my iPod
08:36:29XavierGr(the USB mode works though for the very first time)
08:36:42DataGhostand now I feel really tired and I still have to build 8 reports in MS access -_-
08:37:06 Join gtkspert_ [0] (
08:37:15LinusNXavierGr: you mean to turn on the backlight?
08:37:30LinusNi can fix that
08:37:37XavierGrah great the USB bootloader mode works great, thanks Linus
08:37:49XavierGrwhere it is a rather small detail
08:38:01 Quit redbreva (Read error: 104 (Connection reset by peer))
08:38:06 Part gtkspert_
08:38:07XavierGrbut yeah I suppose it would be good to be responsive on keypresses even on USB bootloader mode
08:38:52XavierGrand to think that all these months it was a minor reorganaziation of the USB initialization sequence (if I am right on this) :P
08:39:15LinusNXavierGr: there is more to it that i haven't yet understood
08:39:37JdGordonLinusN: usb screen on the remote is stuffed... it displays the battery log on all lines
08:39:47XavierGrhmm strange, when I charge it at off it won't load the original firmware
08:40:12LinusNXavierGr: that is not good
08:40:24LinusN(i.e not intended)
08:40:32 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
08:40:36 Join RadioactiveMan [0] (i=c0231115@gateway/web/cgi-irc/
08:40:36XavierGrwell I just checked into mains
08:40:46XavierGrit wont do anything until I start rockbox
08:40:53XavierGrlet me check if OF works :O
08:41:45XavierGrok OF works
08:42:04XavierGrbut it won't auto power-on on adapter power
08:42:13*JdGordon seconds that
08:42:22 Join Timm3h [0] (
08:42:24LinusNodd indeed
08:42:28XavierGrwhen I plug the mains it just sits idly until I power on rockbox or OF
08:42:44LinusNeven after running the OF?
08:43:50XavierGrafter running the OF it will charge normally
08:44:06XavierGr(even if I power it down, then OF charging screen will appear)
08:44:07 Join kaaloo [0] (
08:44:19JdGordonnot here...
08:44:37JdGordonI just went into Of, then turned off.. tryed charging and nothing is happennig
08:45:20XavierGrbut I have to leave now, I will test more thoroughly when I get back
08:45:22LinusNi see the problem
08:45:26XavierGrthanks again Linus bye
08:46:07LinusNand it is potentially harmful
08:46:19JdGordonthe fix or the problem?
08:46:23LinusNif you leave the charger connected
08:46:25XavierGreven when runing rockbox?
08:46:41XavierGrso shouldn't I leave it to charge into rockbox?
08:46:50LinusNno, only in the case when it doesn't react when you insert the charger
08:46:54JdGordonwhy is the backlight being turned off in the usb loop?
08:47:02LinusNto save battery
08:47:19XavierGryeah better that way
08:47:35LinusNi think we could leave it on if the charging is enabled
08:47:36XavierGrjust it would be nice to have backlight on keypress as you said
08:47:38JdGordonbut the buttons dont work.. so you can forget its connected
08:47:51LinusNwill fix that
08:48:08LinusNi think i'll set USB bootloader mode to "fixed" in IriverBoot
08:48:59JdGordonLinusN: remote_line needs to be set before check_battery()
08:49:08LinusNi see
08:49:14JdGordonwell.. actually, anywhere line=0;
08:49:41 Join Rob222241 [0] (
08:51:27JdGordonand last thing, at line 513
08:53:48JdGordonhmm... automatic reset if rockbox.iriver isnt there?
08:54:23JdGordonI thought I saw text error messages, or are they not ready yet?
08:54:32LinusNreset instead of poweroff?
08:54:46LinusNtext error msg should work
08:55:06JdGordonpoweroff is what I meant
08:55:18JdGordonbut it gives no error, just shutsdown
08:56:26LinusNJdGordon: it's because there is no delay, so you don't see it
08:56:46JdGordonoh.. shouldnt it freeze on the screen?
08:57:06LinusNit actually tries to restart the OF, but the OF shuts down immediately because ON is not held
09:01:14LinusNoops, i see a bug
09:01:50 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
09:07:12JdGordona bad bug?
09:07:22 Quit Rob2222 (Read error: 110 (Connection timed out))
09:09:07 Join illriginal [0] (
09:09:49*petur just saw two bad bugs
09:10:08JdGordonkill then quick!
09:10:17GodEateris saratoga around ?
09:10:35GodEaterI think he might have been a bit hasty locking a forum thread...
09:12:01JdGordonLinusN: doing a remote_type() check should be enough to get aroudn the h300 non-lcd remote right?
09:12:33GodEateranyone else care to unlock it (assuming you agree with me) ? It's in the Gigabeat install forum...
09:12:34JdGordone.g (assuming I didnt stuff up the logic :p
09:14:45JdGordonyeah, that logic is wrong
09:16:58 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
09:17:03LinusNJdGordon: it might not be that easy, since the remote_tick handler must be running before we can detect the remote type
09:17:20LinusNi.e the button check must be done later in the code
09:17:32JdGordonoh ? nuts :p
09:18:11amiconnLinusN: Do we have adc?
09:18:11JdGordonspeaking of tik handlers.. what happens if it takes longer than 1 tick to complete?
09:18:23LinusNamiconn: yes
09:18:27amiconnIf so, the remote check could be done "manually"
09:18:39LinusNJdGordon: then you're in for a surprise :-)
09:18:56LinusNamiconn: yes
09:19:02JdGordonhaha ok
09:19:15amiconnA side effect of checking the accessory voltage is that it also tells you the hold state for the lcd remotes
09:19:16amiconnNo extra check necessary
09:19:52JdGordonLinusN: please let me know if you find any bugs bad enough that I should go back to the old bootloader
09:20:08GodEaterLlorean: thanks for that - I got the impression he was trying to install Rockbox too - it does sound like he's managed to bugger his unit up though
09:20:23LinusNJdGordon: there is a bug that can give you an I04: error when loading rockbox
09:20:47LinusNJdGordon: committing a fix for that in a minute or so
09:21:05***Saving seen data "./dancer.seen"
09:21:08LinusNjust want to test it on my h300 first
09:23:29JdGordonaudio thump and zzap is still there btw..
09:23:34JdGordon.. on boot
09:24:08LinusNnot good
09:24:50LinusNi think i want to split the h300 and h100 bootloader source file
09:24:59*petur just pressed stop while a song was playing and backlight off. It never came on again, hard reset needed :(
09:25:00LinusNit's a major #ifdef hell right now
09:25:21 Join Jaxx [0] (n=atul@
09:25:30LinusNpetur: i saw that too a few days ago
09:25:47LinusNforgot to follow up though
09:26:31peturand maybe related: on three occasions I had the first keypress eaten: pressing stop only switched on the backlight. (and yes, I have that setting off)
09:28:14LinusNpetur: target?
09:29:13 Quit linuxstb (Read error: 113 (No route to host))
09:29:18Jaxxhave there been any battery life improvements for the ipod mini or later firmwares? We still dont use the hardware chip to decode the mp3s?
09:29:44peturJaxx: hand over those datasheets
09:30:20Jaxxaah. so cpu is the way to go? until someone finds the datasheets i.e :)
09:30:52GodEaterthere isn't a hardware mp3 decoder in the mini is there ?
09:31:06Jaxxrockbox is so great in features .. if only it had 14 hr battery life on the ipod port
09:31:20 Quit GodEater ("CGI:IRC")
09:31:33 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:32:07JaxxGodEater1: judging by the battery juice that rockbox sucks compared to the apple firmware .. i am guessing that it is chip based decoding
09:32:38GodEaterJaxx: well I'm pretty certain that's a false assumption - ipods don't *just* decode mp3s - what about AAC ?
09:33:27GodEaterthe working theory on battery life issues with ipods and other PortalPlayer based targets here is that we're not powering down enough of the iPod when it's not required. Because we don't know how.
09:33:27 Quit idnar (Read error: 60 (Operation timed out))
09:34:01 Join idnar [0] (n=mithrand@unaffiliated/idnar)
09:34:01Jaxxu mean caching the next ~5 songs into the 32 MB ram and shutting off the hdd?
09:34:11ShaidI'm going to take a multimeter to mine while it's playing/idle/etc and see if I can work out what's drawing lots of power when it shouldn't
09:34:25GodEaternot precisely that, but that general idea yes Jaxx
09:34:57ShaidI'd also love to get my hands on the broadcom datasheet.
09:35:02Shaidwonder who I need to kill for it?
09:35:36Shaidis it possible the broadcom chip is powered on, draining battery, on the 5gs?
09:35:51GodEatervery possible
09:35:56LinusNthat's possible
09:36:06GodEateralthough that won't apply in the mini case that Jaxx is asking about
09:36:15JaxxShaid: the broadcom chip is responsible for?
09:36:23GodEaterJaxx: Video decoding
09:36:27ShaidJaxx: It's the video/3d effect chip on the 5th gens
09:36:33Shaidalso has 64meg of ram in it
09:36:41Shaidwhich would be nice to get access to...
09:36:58Jaxxis the 80 gig supported?
09:37:09ShaidI'm not even sure who's working on it.
09:37:14*GodEater look ssidelong at LinusN
09:37:21*Shaid does too
09:37:28 Join decayedcell [0] (
09:37:39*LinusN hides in shame
09:37:48Jaxxwhts the largest hdd rockbox supports? and on which platform can i get that ?
09:37:59GodEater160GB on an Archos JBR I believe
09:38:20GodEaterwell - assuming you want the disk to fit in the original case
09:38:22ShaidI'm tempted to do the 3.5" hdd 'mod' to my 4th gen
09:38:28Jaxxdoes it involve hardware decoding on the Archos? is battery life good?
09:38:28Shaidjust for fun
09:38:36GodEaterif you don't care about that then we can have a REALLY big disk :)
09:39:00*GodEater looks forward to his petabyte iPod sometime soon
09:39:20Shaidthere's an idea
09:39:25Shaidexternal raid controller
09:39:36GodEaterJaxx: yes, the archos uses a hardware decoder, battery life is better than the OF#
09:39:48GodEaterit is, however, ugly as sin (IMO)
09:39:48Shaidis the ipod a SWCODEC platform?
09:40:01GodEateryes it is
09:41:24 Quit GodEater ("CGI:IRC")
09:41:38 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:42:59Jaxx599 bucks for the 504 160 gig .. and it even wont fit in my pocket
09:43:11 Quit illriginal ("Leaving")
09:43:22GodEaterJaxx: that's not a JBR
09:43:25LinusNJdGordon: i can't hear a thump when booting
09:43:25Shaidif you can get one, the Toshiba Gigabeat X60 is a good player for rockbox.
09:43:32Shaidjust make sure it's not an S60.
09:43:45GodEateror the F60 - that's nice too
09:43:48GodEaterand easier to find
09:47:15 Part Jaxx
09:47:26LinusNi wonder why i can't hear a thump when booting my h300
09:48:02LinusNi hear a slight click when rockbox has started, but nothing else
09:50:17GodEaterLinusN: clearly you don't have it hooked up through a 1000W Sub Woofer ?
09:51:57 Join Mouser_X [0] (
09:52:18LinusNGodEater: hehe, no, only headphones :-)
09:52:27 Quit decayedcell (Remote closed the connection)
09:52:44JdGordonLinusN: I think it might depend on whats plugged into the headphone outs... with my stereo its very annoying, also before it was doing an annoying zzap sort of sound
09:52:53JdGordonfor abiout half a second
09:53:04LinusNwhen does the zzap come?
09:53:23JdGordonboth stop just before the lcd turns on
09:53:52LinusNso it's basically right when you press ON?
09:54:24JdGordonpretty much..
09:54:35LinusNJdGordon: is the problem the same when you boot the OF?
09:54:49 Join pill [0] (
09:54:50LinusNor even when you use the original rom without the bootloader?
09:56:04JdGordonits no where near as bad holding rec
09:56:10JdGordondunno about using the original bootloader
09:56:14LinusNthe UDA1380 reset is performed right before lcd_init()
09:56:28LinusNi could do it earlier
09:56:58LinusNJdGordon: do you build your own bootloader?
09:57:10JdGordonI dont think rockbox inits the sound hardware properly.. I get whitenoise in rockbox (with nothing playing) and none in the OF
09:57:43LinusNcan you move the call to audiohw_reset(); to right after power_init()?
09:57:59JdGordonbah, sorry, the OF has the whitenoise also
09:58:07JdGordonthats safe?
09:58:12LinusNor perhaps right after the ON button check
09:58:16LinusNyes, it's safe
09:58:51 Join B4gder [0] (
09:59:34 Join vjt [0] (
09:59:52 Join kubiix [0] (
10:00:48vjthello, rockbox rocks! :) i'm about to compile it with the COP patch for 5g ipod, which arm-elf-gcc version should i use? 3.3? 4.0? this post tells me to use 4.0.2, but it's also from march 2006
10:00:50*JdGordon wishes the flash process was a bit faster :p
10:01:07GodEatervjt: 4.0.3
10:01:17vjtGodEater, thank you
10:01:52JdGordonLinusN: nothing :) ill try a few more times to make sure tho
10:01:56GodEatervjt: if you used the tool/ script - that's what will get installed for you
10:02:06LinusNJdGordon: putting it there is however *not* safe for the h100 bootloader
10:02:34LinusN(because of the sleep call)
10:02:37JdGordonbah, nope.. didnt fix it
10:02:56LinusNJdGordon: but is it better?
10:03:28JdGordonno change... its not every boot unfortunatly
10:03:31GodEaterJdGordon: Are you going to be commiting the root menu patch today ? Or are there still issues to thresh out ?
10:03:51JdGordonIf im allowd to I probably will :)
10:03:52LinusNJdGordon: would be interesting to know if it is better with only OF installed
10:04:01JdGordonbut I have to go now for a few hours
10:04:12JdGordonLinusN: ok, ill check for you when i get back unless someone else does
10:04:18LinusNJdGordon: thx
10:04:48 Quit JerryLange (Read error: 104 (Connection reset by peer))
10:07:03GodEaterah - I notice root_menu.patch doesn't apply cleanly to svn today
10:07:09GodEaterno that's a lie
10:07:13GodEaterit just doesn't build
10:08:09 Join Llorean [0] (
10:15:54 Join roolku [0] (
10:18:20DataGhostI really feel like a truck drove over me or something :/
10:18:29DataGhostand still at least 6 hours of hard work remain
10:18:31 Join ender` [0] (
10:18:51LloreanDataGhost: Have you come within 5 feet of me in the last 6 days or so?
10:19:00*LinusN enjoys some "nice" 8051 hacking
10:19:02LloreanWhatever I have seems to be very contagious, and with symptoms somewhat similar to that.
10:19:21DataGhostwell I stayed up this night to finish the work for today
10:19:22*GodEater thought Llorean meant he'd run over DataGhost in his truck
10:19:27peturLinusN: my H300 just crashed again. Was playing with backlight off and I hit play to pause music. Shall I FS this?
10:19:34DataGhostbut instead I got iPL to run off fat32 without symlink issues etc
10:19:35 Join pondlife [0] (
10:19:41LinusNpetur: definitely
10:19:57DataGhostand I still need to do the work I planned for this night, only in less time
10:20:03pondlifeLinusN: Morning, any more bootloader mods coming up, or shall I go for it?
10:20:46LinusNpondlife: would be very nice if you tried the latest svn
10:20:55pondlifeYep, am about to
10:21:14pondlifeJust thought I'd hold off if you were about to update anything.
10:23:06pondlifeDid you fix: "08.46.07 # <LinusN> and it is potentially harmful"
10:23:15LinusNno, not yet
10:23:30pondlifeOK, so just be careful with the charger
10:23:37LinusNjust don't leave the charger connected if it doesn't boot when you insert it
10:25:35 Join decayedcell [0] (
10:26:05 Join inversions [0] (
10:26:05pondlifeHmm, wiki search isn't finding anything for the word "bootloader"
10:26:42LinusNi guess it's "boot loader"
10:27:07LloreanI believe that's what was settled on for the manual.
10:27:10LloreanAt least
10:27:14pondlifeNope, not on
10:27:28pondlifeIt worked last week!
10:27:31 Join BigBambi [0] (
10:27:45pondlifeAnyway, I found my content, now to get scrambling
10:28:58vjtGodEater, no, i didn't run it. i'm waiting for the vmx download to complete, and i'll build inside vmware :)
10:29:16vjti am currently really enthusiast about rockbox, it is the firmware i always wanted for my player
10:29:23 Quit RadioactiveMan ("CGI:IRC (EOF)")
10:29:30 Join juhramcsn [0] (i=c0231115@gateway/web/cgi-irc/
10:30:50LinusNvjt: same here :-)
10:31:19 Quit juhramcsn (Client Quit)
10:31:21 Join juhramcsn2 [0] (i=c0231115@gateway/web/cgi-irc/
10:32:25linuxstb_Shaid: My understanding is the the LCD is connected to the Broadcom chip in the 5gs, and all LCD updates go via the broadcom chip - so it's running during normal usage.
10:33:44 Quit GodEater ("CGI:IRC (Ping timeout)")
10:33:45linuxstb_But I don't think that's the cause of the reduced battery life - the 5g doesn't have worse battery life than the other ipods.
10:35:20LinusNhow much power does the ipod draw when both cores are idle?
10:35:43 Join lee-qid [0] (
10:38:40 Quit lee-qid (Read error: 54 (Connection reset by peer))
10:39:13 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
10:40:07 Join lee-qid [0] (
10:40:30pondlifeLinusN: USB bootloader mode now seems fine, 3 out of 3 tests
10:41:30 Join Ribs [0] (n=ribs@
10:42:44LinusNhow much do we know about the Samsung SA58700 chip?
10:43:23LinusNcalmrisc crap, no data sheets i suppose?
10:44:03B4gderwhat is using that?
10:44:34 Join bluebrother [0] (i=QOSfLCKH@rockbox/staff/bluebrother)
10:44:51LinusNiriver E10, just had one in my hand
10:44:52pondlifeLinusN: Normal boot now gives two clicks which weren't there before
10:45:04pondlifeBut not a biggie.
10:45:14B4gderLinusN: aha... I doubt that is calmrisc though
10:45:57B4gderyep, ARM940T
10:46:21B4gderbut calmrisc core within it
10:47:37LloreanThe E10 is the television remote / MP3 player, right?
10:48:27B4gderit seems so, yes
10:49:32B4gderbut very few buttons, I bet it would be awkward to use as a remote for real
10:49:43LloreanVery probably.
10:49:55LloreanI used to have a PDA whose IR function I used for such things.
10:50:30B4gderI got myself a "universal remote" thing for my living room just a while ago, and with a bazillion buttons it still is hard to get all devices controller in a nice and intuitive way
10:50:41LloreanBecause every IR-equipped device I've owned has come with a 'programmable' or 'multi-purpose' remote that seems unable to be programmed to interact with any of the other ones despite the vast list of 'supported' devices in the product manual.
10:51:21DataGhostLlorean that probably just means that they don't interfere :P
10:51:34LloreanWhen I visit my parents, they have one with an LCD screen, and these crazy multipurpose buttons. You press one thing and it turns on a TV, the DVD player, a receiver unit, and ejects the DVD tray, for example.
10:51:48B4gdermy remote can "learn" from the original remote, and put that on any key
10:52:03LloreanI wish I had one that could learn. I keep meaning to buy one.
10:52:06B4gderand such "macro" functions
10:52:34LloreanI just want to have the volume keys correspond to one device, the channel keys to another, a power key for each, and maybe input selection for the TV as an added bonus.
10:52:47B4gdercontrolling TV, DVD, VCR and STB
10:55:30 Quit lee-qid (
10:55:30 Quit BigBambi (
10:55:30 Quit pill (
10:55:30 Quit linuxstb_ (
10:55:30 Quit perplexity (
10:55:30 Quit SirFunk (
10:55:30 Quit lost|X40 (
10:55:30 Quit secleinteer (
10:55:30 Quit amiconn (
10:55:30 Quit tchan (
10:55:57 Quit Mouser_X (Read error: 110 (Connection timed out))
10:56:05NJoinlee-qid [0] (
10:56:05NJoinpill [0] (
10:56:05NJoinlinuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
10:56:05NJoinperplexity [0] (
10:56:05NJoinSirFunk [0] (
10:56:05NJoinlost|X40 [0] (
10:56:05NJoinsecleinteer [0] (n=scl@
10:56:05NJoinamiconn [0] (n=jens@rockbox/developer/amiconn)
10:56:05NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
10:56:07NJoinBigBambi [0] (
10:56:12 Quit tchan (Connection reset by peer)
10:56:24BigBambiEveryone is back
10:56:39DataGhostyou're back
10:57:01BigBambilol, there were about 6 of us left then - the biggest netsplit I've ever aseen
10:57:03 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
10:57:15LloreanFrom our side, about six of you left, a rather small one. :-P
10:57:15DataGhostI knew someone was going to say that :)
10:57:21DataGhostit was quite.. what Llorean said
10:57:25linuxstb_LinusN: I think someone posted some ipod power measurements in the forum recently, but I can't find the thread again now.
10:57:32 Quit secleinteer (Connection reset by peer)
10:57:46 Join secleinteer [0] (n=scl@
10:57:58Lloreanlinuxstb_: Someone did, though there was a _little_ uncertainty about the circumstances of a couple of the measurements if I recall
11:01:57LloreanThere is it
11:02:07LloreanLinusN: see link two lines up.
11:06:10 Quit secleinteer (
11:06:10 Quit tchan (
11:06:10 Quit BigBambi (
11:06:10 Quit linuxstb_ (
11:06:10 Quit pill (
11:06:10 Quit lee-qid (
11:06:10 Quit lost|X40 (
11:06:10 Quit amiconn (
11:06:10 Quit perplexity (
11:06:10 Quit SirFunk (
11:06:41NJoinsecleinteer [0] (n=scl@
11:06:41NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
11:06:41NJoinBigBambi [0] (
11:06:41NJoinlee-qid [0] (
11:06:41NJoinpill [0] (
11:06:41NJoinlinuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
11:06:41NJoinperplexity [0] (
11:06:41NJoinSirFunk [0] (
11:06:41NJoinlost|X40 [0] (
11:06:41NJoinamiconn [0] (n=jens@rockbox/developer/amiconn)
11:06:51 Quit secleinteer (Connection reset by peer)
11:07:13 Join secleinteer [0] (n=scl@
11:07:24Lloreanlinuxstb_: Did you see the link I gave, for the forum post, or did it get lost in the split?
11:08:25DataGhostthe split was 5 minutes later :)
11:09:43LloreanDataGhost: Sometimes I seem them very late.
11:14:41DataGhostanyway, I'm off to uni
11:16:34 Part Llorean
11:18:35 Part decayedcell
11:19:46 Quit atsea-145193 (Excess Flood)
11:21:08***Saving seen data "./dancer.seen"
11:21:44 Join atsea-145193 [0] (i=ariel@gateway/tor/x-3a0339e3308a5eb3)
11:24:34 Nick kclaf is now known as kclafk (
11:25:44 Join BoeMatic [0] (
11:28:36 Quit juhramcsn2 ("CGI:IRC (EOF)")
11:28:49 Quit inversions (Read error: 110 (Connection timed out))
11:30:25 Join decayedcell [0] (
11:33:47 Quit BoeMatic (Remote closed the connection)
11:34:13*GodEater has just played around with dan_a's idea of using git to host his local repository - it works quite well
11:37:20LinusNthe latest root menu patch works fine
11:38:56markunLinusN: almost ready for a commit?
11:39:19LinusNlooks like it - haven't tested that much yet
11:39:30 Quit GodEater ("CGI:IRC (Ping timeout)")
11:40:22 Quit decayedcell (Remote closed the connection)
11:43:26 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
11:44:23GodEaterI think it's ready for commit - so does JdGordon - I think he's just waiting for the nod from the rest of the committers to make sure they don't have any last minute qualms
11:46:10 Join Rondom [0] (
11:47:39LinusNi personally think it's fine, but i haven't tried all targets
11:48:39LinusNwe will of course get our fair share of "this sucks - i can't do X anymore with Y clicks" etc, but that's to be expected
11:49:03LinusNi think this is a major step forward
11:49:09GodEateryou can't please everyone
11:49:10 Join Nico_P [0] (
11:49:13peturwhat about manual updates?
11:49:27LinusNpetur: that will be sooooooo lovely to do :-)
11:49:28GodEatergood point
11:53:22LinusNthe manual will need a major overhaul, and unfortunately the menus will change a lot in the time to come, because of the root menu
11:53:39LinusNso the manual will be in flux
11:54:06*bluebrother wonders if JdGordon will add the manual changes to his patch
11:54:39*bluebrother stops dreaming
11:54:41LinusN"we will not accept the patch until you have added the manual changes too" :-)
11:54:58peturpoor Jd
11:55:01*bluebrother raises his hand for voting :)
11:55:47Kasperlei have a wps file here that doesn'T seem to work correctly. and those parts that seem off are not covered by the docs, either (at least not like that). one example: %s%m|49|318|%?ia<%ia |%?d2<%d2|(root)>>
11:55:58LinusNi think the root menu is destined for svn inclusion, and there is no reason to wait
11:56:23bluebrotherKasperle: isn't %m part of the scrolling margins patch?
11:56:25LinusNbut it is vital that the keymaps are up-to-date for all targets
11:56:46BigBambi%m isn't official, as bluebrother says
11:57:03 Join fejfighter [0] (
11:57:07Kasperlethat at least explains why i can't find it in the docs ;)
11:57:08BigBambiWhere is the WPS from?
11:57:18BigBambiIt should say what patches you need when you get it
11:57:55*bluebrother thinks about a packaged wps format that can handle situations like that ...
11:58:00Kasperleit's this:
11:58:08bluebrotherdamn, I'm dreaming again ;-)
11:59:13Kasperlei guess i'll try that album art patch, then ;)
11:59:30BigBambiAs bluebrother said, I think %m is scrolling margins
11:59:30LinusNi like the skin proposal in the wiki
12:00:11BigBambiGiven it is a Senab theme, you might want to look at his unofficial build and see what is in it, seeing as it almost certainly will have the patches you need
12:00:20BigBambiLinusN: Have you a link?
12:02:23bluebrotherI'm not sure if I get the exact point. So it's moving the skin functionality out to separate code?
12:02:28 Join CaduMimi_ [0] (
12:02:28bluebrotherin terms of loadable code?
12:02:34 Nick CaduMimi_ is now known as cadu (
12:02:58LinusNbluebrother: no, it's a suggestion for a system that can apply a skin to all screens
12:03:07caduhello guys, i have a small question about the toshiba gigabeat series
12:03:15LinusNmuch like the wps is today, but for all screens
12:03:16bluebrotherah, so mostly a wps-for-other-screens thing?
12:03:31GodEatercadu: go ahead
12:03:31caduyesterday i talked with a couple of pals here about it and both say this player is okay for use, good :P
12:03:38bluebrotherok. Sounds much more reasonable to me than what I thought first ;-)
12:04:26bluebrotherI'd be interested in something like that. But no time for the next 3 month :(
12:04:36 Join decayedcell [0] (
12:04:41cadutoday i did a quick read at rockbox's toshiba gigabeat homepage... it shows some devices at 50% support like power management / lcd driver...
12:05:03caduany caveats at all ? i've been thinking about getting me one of those all day long :)
12:05:21GodEatercadu: That may just mean those items have not been fully optimised yet
12:05:24markuncadu: which one are you planning to get?
12:05:29LinusNcadu: it's still a work in progress, but i hear that it works just fine
12:05:47GodEatercadu: it's still a fantastic player IMO
12:05:51LinusNcadu: make sure you get an F or X series, and not S
12:06:07 Join tick [0] (i=c27f0812@gateway/web/cgi-irc/
12:06:25cadumarkun, haven't taken a look at the stores yet, but as i'm in japan, i would not expect anything but the bleeding edge models...things here vanish from the stores quickly, and are quickly supplanted by new models and whatever :P
12:06:36tickLinusN: I'd rather interpret the skin proposal in the bluebrother's way
12:06:45markuncadu: I could split out the some of the items on the todo so you can see which part of the power management is done and which not (wakeup from sleep for example)
12:07:15tickNice idea indeed
12:07:25BigBambiLinusN: nice idea, but I think it's have to be WPS like rather than moving code around - the vast majority of people who would want to write skins (me included) don't have the knowledge either in C or rockbox to do it that wy, whereas I can deal with WPS style tags
12:07:30cadumarkun, nice, but i see these parts aren't really *vital* for the device operation....
12:07:47caduwakeup from sleep is a nice feature who pause it for a while, i usually turn the damn thing off already :P
12:08:16tickBigBambi: there could be an "interpreting" skin that would parse a wps file. Other skins could hard code the displaying
12:08:34LinusNBigBambi: from the wiki page in question: "So it is possible to easily change some Parts of the user interface without programming in C"
12:08:42BigBambiOK, missed that bit
12:08:50BigBambiGood plan though
12:09:13tickLinusN: I don't get this point. How's that supposed to be?
12:09:16markuncadu: you can turn it off with a small battery switch, but when you shutdown from within rockbox it just enter the sleep mode. But right now we just reboot rockbox when we wakup from sleep.
12:09:38BigBambiWould go nicely with viewports...
12:09:54cadumarkun, so these smart portable audio players don't get really 'off' at all ?
12:09:54markunBigBambi: yes, that's what I thought as well
12:09:57bluebrotherso it's basically a "wps" like parsing for list views, wps, radio and recording screen?
12:10:10LinusNhmmm, maybe i am interpreting it wrong...
12:10:21cadumarkun, like the ATX vs AT mobo paradigm
12:10:39tickSo a skin could be something like the plugin API now: just a bunch of funcs for rendering of different screens.
12:10:40LinusNit certainly looks like he is suggesting a plugin-like interface after all
12:10:42markuncadu: well, they preserve what's in RAM and a few special registeres, but the rest is turned off.
12:10:53LinusNand that plainly sucks
12:11:26BigBambiI don't know from an architecture point of view, but from a me being able to write a skin, it would seem to make it hard
12:11:49tickLinusN: why that? The current state of the things could be easily implemented with the skin concept
12:12:02markunLinusN: I read his proposal as separating the interface form the rest or rockbox. Might be nice for blind users as well with text-to-speech.
12:12:07cadumarkun, usb and cradle charging are okay? rockbox detects it or it's just charging in hardware-only ?
12:12:11pondlifeIs there a limit to the number of times flashing can take place before any hardware fails?
12:12:29LinusNi think skins with C code is doomed to break
12:12:32pondlifei.e. 1000 flashes and that's yer lot
12:13:00markuncadu: it's hardware controlled and rockbox detects when the battery is being charged.
12:13:23tickLinusN: it will be dangerous, yes. But very flexible.
12:13:41LinusNit will be a burden to keep synchronized
12:14:13tickSo the skins should be tested thoroughsly. But the API would stabilise quite fast I think
12:14:19*pondlife just started his H300 with a non-LCD remote
12:14:19LinusNand a nightmare to support when there are 3rd party skins
12:14:42perplexitypondlife ?? nice! current svn code ?
12:14:48cadumarkun, nice :)
12:15:05pondlifeA small mod or 2, still working out what
12:15:13cadui think i'll buy (whatever is available at the stores, Fxx or Xxx) this weekend :P
12:15:28markuncadu: are you a programmer?
12:15:37cadumarkun, mind you, i'll arrive home...turn it on to see if it's okay, and flash rockbox onto it :P
12:15:37*fejfighter offers thumbs up to pondlife
12:15:48 Quit fejfighter ()
12:15:54markuncadu: right now rockbox doesn't require any flashing
12:16:12cadumarkun, i can do some basic C/pascal , some knowledge in linux/networking, can do some Python, but that's it
12:16:34markuncadu: well, if you want to help out, there is always something to do.
12:16:42cadumarkun, oh i've read it..the firmware stays on a partition on the hard drive....
12:16:43tickLinusN: but still: it's a nice idea IMO
12:16:43 Quit tick ("CGI:IRC")
12:17:05*JdGordon returns
12:17:26JdGordonbluebrother: you want me to edit the manual? it will be full of typos and sucky english... and macros :D
12:17:26markunbut right now I don't know what else can be done to improve the battery life for example.
12:17:50cadumarkun, well i don't think i could help with coding, i don't have any useful C or arm/whatever asm knowledge at all , i can do some hacks when i want or when i'm pissed off (like i did to xmms...patch applied, and some simple mods on Nethack for gp2x)
12:17:52bluebrotherhehe ... LaTeX macros are quite different to C preprocessor ones :)
12:18:10*JdGordon hates LaTeX
12:18:24*bluebrother doesn't understand this
12:18:28markuncadu: perhaps we can port some nice things over from the gp2x software?
12:19:12cadumarkun, i would like to see the opposite happening too....a rockbox port to gp2x :P
12:19:30markunanyway, I'm several hours late for work already (strange dreams). Talk to you later.
12:19:41JdGordonall it needs it cross compiler for sdl arm and you should be ok
12:19:43caduit has 2 arm processors (one lacks a MMU and has some specialized instructs) and all hardware documented :P
12:19:50perplexityI reckon rockbox would make an awesome userspace application for my palm :)
12:20:03caduJdGordon, i would love userspace rockbox on gp2x (arm linux really)
12:20:13markuncadu: good thing all our hardware was documented as well (besides the LCD module)
12:21:00JdGordonLinusN: I dont think you need to worry about the thump sound... I just put on the OF bootloader and its only a bit quieter.. and happens at a differnt place, but is still there
12:21:10JdGordonmaybe its faulty hardware and not software?
12:23:09JdGordon... and if your there.. just making sure the svn bootloader should be safe?
12:23:55JdGordondid die?
12:25:25JdGordonghosts somewhere between it and me then?
12:26:40JdGordoncan someone dcc me bootloader/main.c from svn please?
12:30:07BigBambiJdGordon: you stil need it?
12:30:48BigBambiIs that working?
12:31:16BigBambiI'm at work, maybe firewall
12:31:19BigBambione mo
12:31:39JdGordoncan you email to jdgordy AT please?
12:31:42JdGordonif thats easier
12:32:37BigBambiwill do
12:33:22BigBambishould be gone
12:33:23Timm3hg i knew that
12:34:31JdGordonhmm.... still waiting...
12:34:43BigBambiOK, different address
12:34:50BigBambiYou are going to get this if it kills me now
12:35:19JdGordonthat can be spammed.. i never check it :)
12:35:57BigBambiI meant I'd send it from a different address - which I just have
12:36:00BigBambiBut I'll send it to that too
12:36:10 Join Mouser_X [0] (
12:36:45JdGordonHMM.. that optusnet email expired :p
12:36:50JdGordonor i forgot the pass
12:36:50BigBambiRight, I've sent it to gmail twice and optus once
12:37:04BigBambiYou are destined never to have the bootloader
12:37:08JdGordonwtf is with my internet?
12:37:19*JdGordon has no rockbox atm :'(
12:37:40luckzqtscrobbler seems to get stuck at 'checking last submission date' works for everyone else?
12:38:14BigBambifor me at least
12:39:15JdGordon1 to 40 of 1308 Messages in the optus email :p
12:39:25JdGordonall forum replies from 1 which wont allow my gmail
12:39:58JdGordonand one email with main.c :D
12:40:12BigBambiI dared to dream and it can to pass
12:42:37*JdGordon wonders if B4gder is blocking my ip as a subtle way of saying he doesnt want the root menu patch commited :D
12:43:30 Join webguest84 [0] (i=863b2ffd@gateway/web/cgi-irc/
12:43:32B4gderwhat did you try to do that could be blocked?
12:43:52JdGordonI cant access svn. at all.. https, svn up.. ping, nothing
12:43:58JdGordonhttp even
12:44:08webguest84do you know a webclient for irc with which I can join _my_ channels ?
12:44:08B4gderJdGordon: from what IP?
12:44:08webguest84I m behind a firewall
12:44:21webguest84those ****** blocked IRC, ... even POP
12:44:22pondlifeJdGordon: Works here FWIW
12:44:29webguest84I mean they blocked POP in a university network
12:44:41webguest84hanks b4gder
12:44:52 Quit webguest84 (Client Quit)
12:45:28JdGordonB4gder: .. although i was being funny about blocking my ip..
12:45:37 Join DataGhostUni [0] (
12:45:51B4gderI can have a look anyway
12:45:51JdGordonbut its wierd, because works, so its not a cable gone down somewhere??
12:46:17B4gderthen it sounds like you have something in between blocking that port or something
12:46:37GodEaterif he can't even ping it's not a port issue
12:46:43B4gder k
12:46:47B4gderno, but some kind of filter
12:46:52JdGordonis svn and www the same compuyter?
12:46:53B4gderif works
12:46:53GodEaterI should really try not to teach you guys to suck eggs shouldn't
12:46:57GodEater+I ?
12:46:58B4gderan no
12:47:01B4gderthey're not the same host
12:47:08 Join Thundercloud [0] (n=thunderc@
12:47:13JdGordonshould they have similar ips?
12:47:18B4gderJdGordon: can you browse ?
12:47:40JdGordonno, timeing out
12:47:54GodEaterJdGordon: have you run a traceroute ?
12:48:09B4gderI can't traceroute to your machine from the server
12:48:43JdGordonare box boxes in sweden?
12:48:52JdGordonmaybe there is a link down somewhere?
12:48:53B4gderboth are in Sweden, yes
12:49:10GodEaterin a sealed nuclear proof bunker...
12:49:10JdGordonaliask: daurnimator: can you guys get to
12:49:31daurnimatordont' think so
12:49:37daurnimator.... how was i here :P
12:49:41JdGordon:) so im not crazy
12:50:10JdGordonpondlife: you?
12:50:10BigBambiHmmm, I can't ping you either JdGordon
12:50:15GodEaterJdGordon: sledgehammer to crack a nut solution here, but you could maybe try TOR ?
12:50:18BigBambiexplains why DCC didn't work
12:50:23*JdGordon doesnt respond to ping...
12:50:27BigBambiah, ok
12:50:34pondlifeJdGordon: I'm in the UK, and all seems fine from here
12:50:44BigBambiI'm also in the UK, and all works Ok here too
12:50:46JdGordonI thought you were in fromt he gold coast?
12:50:54pondlifeI wish
12:51:01BigBambiI am in my head
12:51:25*GodEater wishes any of the UK's beaches could be described as gold
12:51:35BigBambiOr beaches
12:51:44pondlifeSome of the Welsh coast is lovely, pity about the weather
12:51:45JdGordontraceroute seems to die after wherever that is?
12:52:20 Join Febs__ [0] (
12:52:22 Nick Febs__ is now known as Febs (
12:52:35GodEateraccording to geo ip
12:52:51 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
12:52:53GodEaterThe Singtel internet exchange
12:53:07pondlifeI can get to both Singtel and svn.Rockbox
12:53:12GodEaterphone them and shout at them
12:53:19 Join miepchen^schlaf [0] (
12:53:22JdGordonits 11pm here.. no point riningg
12:53:50GodEaterthey should answer the phone 24 hours
12:54:02ShaidJdGordon: let me see if I can get there
12:55:53aliaskSingtel owns optus.
12:56:12JdGordonB4gder: are the servers in the same room or same isp?
12:56:26B4gderquite far away, network wise and physical
12:56:55JdGordonok, that must be it then...
12:57:27GodEaterI find it bizarre that the routing dies so far away from Sweden though
12:57:47GodEateryou'd have thought that at that distance you wouldn't even get the website either
12:57:57aliaskSingtel's international lines are particularly unreliable in my experience with them.
12:58:08aliaskEither crappy speeds or dropouts
12:58:37B4gderyeah, but clearly these must be using two different routes
12:59:03JdGordonisnt the point of tcp/ip that if one route dies another can take over?
12:59:29GodEaterthat's the theory - but only if the router has alternate routes configured
12:59:38GodEaterwhich in this case it appears it doesn't
12:59:53JdGordonok, well it seems that if i ssh into my uni I can get to it
13:00:03*GodEater finds a lot of supposedly network savvy people don't really get tcp/ip
13:00:59 Join DataGhostUni2 [0] (
13:01:00JdGordonstupid singtel :'(
13:01:24aliaskSadly we don't have much choice in the matter
13:01:25 Quit DataGhostUni (Nick collision from services.)
13:01:27 Nick DataGhostUni2 is now known as DataGhostUni (
13:01:29*GodEater is in the privileged position of counting one of the fathers of packet switched networking as one of his friends
13:02:16JdGordondaurnimator: your not with optus tho.. so why cant you get through? telstra wouldnt go through singtel would they?
13:02:54aliaskNope, but I dont think he's with telstra either iirc
13:03:04daurnimatorits telstra
13:03:08daurnimator(the backend)
13:03:16Shaiddamn, World of Warcraft + VMware at the same time = bad.
13:03:19daurnimatoras it happens, i work for my isp ;)
13:03:35JdGordonits either telstra or optus lines.. noone else has their own network
13:03:41 Nick linuxstb_ is now known as linuxstb (n=linuxstb@rockbox/developer/linuxstb)
13:04:12ShaidJdGordon: I can't get their either.
13:04:18daurnimatorthen theres layer 2 traffic
13:04:20ShaidSo iinet is in same boat as Optarse.
13:04:23daurnimatorWCG, SPT etc
13:05:16 Nick kclafk is now known as kclaf (
13:07:14 Part decayedcell
13:07:44pondlifeLinusN: In the H300 bootloader, the remote type hasn't been detected when you work out hold_status.
13:08:47pondlifeThe detection can't start until lcd_remote_init() has been called.
13:08:55LinusNpondlife: i know
13:09:01pondlifeAh, ok.
13:09:02daurnimatorJdGordon: what you think about that new port? ;)
13:09:18LinusNpondlife: i discussed it with JdGordon earlier today
13:09:27*pondlife missed that
13:09:37pondlifeWill look at the logs
13:10:01LinusNpondlife: i thought you had it working?
13:11:42pondlifeOnly by redetecting later
13:12:04JdGordonLinusN: 2 easy things... the remote backlight isnt turned off in usb, and buttons still dont turn on the display (which im hoping you are going to fix)
13:13:14 Nick blithe_ is now known as blithe (
13:13:46LinusNpondlife: i see
13:14:03GodEateramongst the million and one other things on LinusN's to-do list :)
13:14:11JdGordonOK, so assuming I can access svn tomorow, Im OK to commit the root menu? all keypads are done (well the menu_wps action is there for all)
13:14:18LinusNGodEater: amen to that
13:14:35LinusNJdGordon: fine with me at least
13:14:44*JdGordon feels more comfortable about playing in the bootloader so can do those backlight fixes..
13:14:48GodEaterget's my vote too - for what it's worth
13:15:18JdGordon... with OFF in the wps going to root?
13:15:34LinusNJdGordon: sounds ok to me
13:15:41 Quit ender` (" Programming is like sex becuse: 11. But some people will never realize how bad they are, and youre wasting your time trying)
13:15:45pondlifeJdGordon: Do you mean STOP?
13:15:51amiconnpondlife: Those flash roms are usually specced for 100.000 cycles
13:15:55JdGordonOFF == STOP
13:16:07LinusNJdGordon: you will of course be the blind community's enemy #1 for a while :-)
13:16:10pondlifeWhy not follow playlist on that one?
13:16:15 Join barrywardell [0] (n=barry@
13:16:25pondlifeIt does at the moment
13:16:27JdGordonLinusN: I know.. and I dont really like that... :(
13:16:49JdGordonI tihnk maybe changing off to GO_TO_PREVIOUS instead of ROOT would make more ppl happy tho
13:16:50pondlifeTry using it with your eyes closed and just voice on
13:16:54LinusNpondlife: why go to the browser at all when you stop the music?
13:17:09JdGordonpondlife: only 2 of the root items are currently voiced
13:17:31pondlifeLinusN: Because I want to select another album
13:17:33JdGordonwho makes the voice files?
13:17:44LinusNJdGordon: not voiced, as in "no voice support", or as in "needs updated voice file"?
13:18:30JdGordonthe latter
13:18:58 Join ender` [0] (
13:18:59pondlifeIMHO it's important to maintain the current basic stop/play
13:19:06pondlifePause is not good as voice stops.
13:19:07LinusNpondlife: sure, but is that what the user would expect, for instance after resuming playback after boot?
13:19:43pondlifeAfter boot, I'd be happy to go to the root, but would prefer to go to the last played in the browser
13:19:54 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
13:20:03pondlifeBut stop is quite common, booting is not
13:20:11vjti've compiled a rockbox with the cop patch for the ipodvideo, it is available at vjt/"> if someone needs it :)
13:20:38pixelmayes, I'd expect to be in the screen I came from too
13:20:40LinusNpondlife: all i want is a consistent and intuitive behaviour
13:20:44pondlifeMe too.
13:21:11***Saving seen data "./dancer.seen"
13:21:12pondlifeBut when going back to the menus or browser, I want to always go to wherever I last was.
13:21:21LinusNi guess previous screen might be good enough, but what if the last screen was the plugin browser?
13:21:31pondlifeThis should cover the settings and everything ideally
13:21:55pondlifei.e. regard it as just 2 parts - the menu heirarchy and the WPS.
13:21:55JdGordonLinusN: you cant go plugins > wps
13:22:09pondlifeWhat about pressing play?
13:22:11JdGordonit will always be either root or files
13:22:16JdGordonits disabled
13:22:28 Quit aliask ("Chatzilla 0.9.77 [Firefox]")
13:22:37LinusNindeed it is
13:22:44pondlifeWhy is that?
13:22:47 Quit DataGhostUni ("Peace and Protection 4.22.2")
13:22:50JdGordonbecause.. :)
13:22:57pondlifeWhy not have play always work as a resume? Consistency ;-)
13:23:08JdGordonthat was being consisteant with svn
13:23:22pondlifeWell, we can improve it, can't we?
13:23:31LinusNthe code in svn is not very consistent, is it? :-)
13:23:42pondlifeStop means go to heirarchy, play means go to WPS
13:23:52LinusNi would expect Play to take me to WPS at all times
13:23:59JdGordonamiconn: have you still got everything setup to make voice files?
13:24:19JdGordonits a simple change.. but it does feel wierd going wps > plugins :p
13:24:29 Join Thundercloud_ [0] (n=thunderc@
13:24:34JdGordoni mean, that will happen if stop == go to previous
13:24:37pondlifeCurrently, in settings, play selects. That's been wrong for a long time..
13:24:54pondlifeThat's the one that's annoyed me most
13:24:57amiconnI didn't make any new voice file since the langv2 transition... and I'm not allowed to publish the bearable ones :(
13:25:01LinusNJdGordon: that's why i think it should go to the root when you stop
13:25:19LinusNthen the file tree is one click away
13:25:22pondlifeI don't understand?
13:25:30pondlifeWhy not go back to the last place?
13:25:35LinusNand since it remembers the location, it shouldn't be a problem
13:25:43JdGordonamiconn: ok, i was wondering why the ones from you on the wiki are so old :p
13:25:58LinusNpondlife: because it's potentially confusing
13:25:59pondlifeSeems inconsistent to me. Why should the browser be special...
13:26:00pixelmaLinusN: still one click too many
13:26:20LinusNpixelma: yes, if you want to go to the browser
13:26:22pondlifeIf it goes back to the heirarchy, it's only one long left back to the root.
13:26:48LinusNpondlife: the last screen could be the plugin browser or the theme browser
13:26:50pondlifeI know you don't see it as a heirarchy, but I have to for consistency...
13:26:53pondlifeWhy not
13:26:57JdGordonpixelma: did you try the latest patch? hopefully left will act properly for you in the tree
13:27:06pondlifeUseful if selecting a theme, certainly
13:27:06amiconnI should adapt the vbscript one day.... probably before the per-target voice split :/
13:27:32pondlifeI can see that plug-ins might be a bit pointless, but consistency is much improved, no?
13:27:44pixelmawant to repeat my suggestion: make it go to the screen one has set as start screen - could make everyone happy
13:27:50LinusNpondlife: consistency is always going back to the root, imho
13:28:08pixelmaLinusN: that's what I do the most, I don't change settings too often...
13:28:11pondlifeTrue, but that's a useful feature lost.
13:28:11LinusNpixelma: sounds reasonable
13:28:20pondlifeCan I set start screen to be browser?
13:28:30pondlifeSo "root"="browser"
13:28:42JdGordonstart != root tho
13:28:50pixelmaJdGordon: yes, repeat left works now
13:28:59GodEaterstart is first screen displayed, root is always the root menu
13:29:02LinusNmy only concern is that it may not be obvious to the user what screen it will return to when you stop
13:29:14pondlifeThe last one they selected
13:29:35GodEaterthat wouldn't make any sense to me if I'd been listening to music for 6 hours
13:29:39pondlifeAlthough I could live with the start screen I guess, as long as follow playlist works properly
13:29:44GodEaterI'd have forgotten where I was by then
13:29:49JdGordonHow about we leave it stop > root, and if we get heaps of requests for something else we change it?
13:29:54pondlifeStop -> start
13:30:12pondlifeWhich will be root for most
13:30:13markunJdGordon: sounds fine
13:30:24LinusNpondlife: that might be kind of irritating it the start screen is the wps
13:30:37*amiconn sure as hell won't set the root menu as his start screen, and would hate to be thrown there by a simple 'Stop"
13:30:41LinusNor the fm radio
13:30:51 Quit Thundercloud (Read error: 110 (Connection timed out))
13:30:56*pondlife wants to use the browser much more than the root
13:31:05LinusNok, so last screen it is
13:31:10*pixelma too
13:31:16 Join fasmaie [0] (
13:31:27pondlifeLong left takes you back to root from most places too, right?
13:31:33LinusNit's probably not as confusing as i fear
13:31:34pondlifeSo it's easy to get back
13:31:36linuxstbWhy don't we use the "startup screen" setting as the place that STOP takes you to?
13:31:46LinusNpondlife: Menu takes you to the root
13:31:46pondlifeBecause it might be WPS or radio
13:32:02pondlifeLinusN: Of course, even better
13:32:04LinusNlinuxstb: read my comment a few lines up
13:32:29JdGordonI think start screen with a sanity check would be best
13:32:43pondlifeI prefer last
13:32:50linuxstbI still think STOP should go to the root menu, otherwise it's no longer a root menu.
13:32:50pondlifeStart screen is a compromise IMHO
13:33:00pondlifelinuxstb: MENU does that
13:33:23linuxstbAnd SELECT takes you to the file browser...
13:33:34pondlifeSelect selects
13:33:38JdGordonwhatever happens... can someone please build a updated voice file? make voice does nothing :p
13:34:12MikachuJdGordon: did you configure for voice?
13:34:14B4gderthe last time I tried, make voice built a voice file
13:34:16pondlifeHow about select in the WPS goes to the last screen but doesn't stop playback?
13:34:28JdGordonMikachu: configure seems to have done nothing
13:34:40JdGordonwhich target needs to be chosen?
13:34:54GodEaterJdGordon: any of them I'd guess
13:35:01GodEaterand the V for voice
13:35:01JdGordonpondlife: select goes to the last browser, and doesnt stop
13:35:22B4gderJdGordon: you should select the target you want the voice for
13:35:29B4gderthey will be target-specific soonish
13:35:42JdGordonconfigure for voice does nothing :(
13:35:58GodEaterJdGordon: does here
13:36:01linuxstbpondlife: What you're saying is that you don't want a root menu, but want to keep the existing Rockbox behaviour 100%?
13:36:17JdGordonGodEater: is it telling you You need Festival, eSpeak or Flite in your path to build voice files ?
13:36:36pondlifeI want consistency
13:36:38JdGordonespeak btw is installed and in the patch
13:36:40linuxstbIn my mind, the root menu is the place you go to when you exit a screen.
13:36:48JdGordon^ amen
13:36:49LinusNpondlife: and going to root is not consistency?
13:36:53linuxstbBut in addition, we have short-cuts between screen.
13:36:58GodEaterGodEater: yes it is
13:37:05GodEatertalking to myself
13:37:08GodEaterJdGordon: yes it is ;)
13:37:14GodEaterand I have all three
13:37:20B4gderJdGordon: it works for me with festival installed
13:37:22JdGordonwhisky tango foxtrot!
13:37:24GodEateralthough I think the Festival install is broken
13:37:43pondlifeSorry, on the phone...
13:37:48GodEaterI don't appear to have a festival_client :(
13:37:58GodEaterso I just went with espeak instead
13:38:03GodEaterwhich appeared to work
13:38:25JdGordonB4gder: is there a script i can run instead of make voice?
13:38:31GodEaterafter you've done make voice - what then ? I'm left with a bazillion .mp3 files
13:38:47pondlifeLinusN: Going to root would be consistent, yes. But going to root on MENU and last screen on STOP is also consistent and better (IMHO).
13:38:56B4gderGodEater: there should be a single file with them all merged
13:39:02LinusNi can go with last screen
13:39:10B4gderJdGordon: sure, just do what make voice does manually
13:39:15GodEaterB4gder: called what ?
13:39:24GodEaterah - english.voice
13:39:30JdGordonwhich is? I get no Makefile
13:39:59GodEaterJdGordon: do you *have* those things in your path ?
13:40:11 Part fasmaie
13:40:22pondlifelinuxstb: FWIW, I certainly do want a root menu, I like the idea of going from file to database browser and settings in a single heirarchy.
13:40:26JdGordonI have espeak
13:42:20*JdGordon gives up... GO_TO_PREVIOUS it is
13:42:38linuxstbI just think "last screen" isn't logical. Does it really matter if it takes a couple of clicks to enter the file browser after you press STOP? Or just press SELECT, then STOP.
13:43:03*JdGordon agrees soo much there... exit shoud goto root
13:43:05amiconnYes it does. It's cumbersome if you're used to classic rockbox navigation
13:43:26JdGordonit takes 5 min to get used to the new menu
13:43:35amiconnI once tried Ondio original firmware... and was immediately annoyed by the root menu
13:43:50amiconnIt requres extra button presses
13:44:16GodEaterhaha - espeak sounds like robbie the robot
13:44:25amiconnThe root menu is probably a good idea... but only if you're not forced to use it everytime
13:44:26 Quit barrywardell ("Konversation terminated!")
13:44:34pondlifeamiconn: My thoughts exactly
13:44:38 Quit Spida (Read error: 113 (No route to host))
13:44:39markunGodEater: but it's a nice 1 man project
13:44:46linuxstbamiconn: It requires one extra press. Just press SELECT to go the file browser, then stop playback.
13:45:20pixelmawhich could be avoided...
13:45:44pondlifeIt just seems more logical to regard the WPS as an overlay on the menu system (including the browser etc.)
13:45:46linuxstbIf you're intending to choose a new album, why stop playback at all? Just use SELECT to go the browser, and choose your next song.
13:45:48GodEatermarkun: Only one guy works on it ? Wow - impressive!
13:46:10markunGodEater: it was designed to run on very simple machines (12MHz ARM initially) and to be understandable, but did not have to be human sounding.
13:46:16JdGordonpondlife: if thats the case then for sure you would expect off to take you to the root menu
13:46:17pondlifelinuxstb: If you want to go to the root menu, why not just press MENU?
13:46:28pixelmalinuxstb: currently you're taken back to the root menu when playback has stopped
13:46:29pondlifeNo, to the last screen I was in
13:46:37GodEatermarkun: well it's pretty good - I wasn't complaining, I actually like the mechanized voice sound :)
13:46:51pixelmalinuxstb: I mean at the end of the playlist
13:46:52markunGodEater: well, he got some help with support for other languages
13:46:53pondlifeI want to go to the root menu only on boot up or when I navigate left into it.
13:47:03JdGordonGodEater: can you put up voice files for the patch please?
13:47:13Shaidthe amiga could do robotic speach synthesis on a 7mhz 68k.
13:47:29ShaidI was impressed back in 1987.
13:47:30markunGodEater: if you have some spare time, a port to rockbox is on my wishlist
13:47:45linuxstbpixelma: Ah, I can see how the "stop at end of playlist" case is different. I agree it makes sense there to take the user to the browser.
13:47:47ShaidBut now in 2007 it's not really alot better...
13:47:57GodEaterfor comparison, flite is terrible
13:47:57JdGordonespeak doesnt use as many languages as rockbox does tho
13:48:03markunShaid: yes, I was impressed too, but wasn't there some special hardware for it?
13:48:25GodEaterJdGordon: would that mean building a voice file per target ?
13:48:26markunJdGordon: I added Dutch, didn't take me long (although it's far from perfect)
13:48:34pixelmalinuxstb: and in this case I don't see why there should be a difference depending on the reason why playback has stopped
13:48:37markunJdGordon: so other people can just add their own languages
13:48:38*linuxstb is starting to get persuaded...
13:48:43JdGordonGodEater: I dont thin so.. dont know tho
13:48:43pondlifeWhat's the best quality voice generator that allows the end product to be distributed?
13:49:13GodEaterI wish I could try out festival - the gentoo ebuild doesn't seem to install the binaries that the make procedure expects though
13:49:37*JdGordon has already caved in.. we are goign to previous.. im _not_ changing that to previousbrowser
13:49:48linuxstbWhy previous, and not previousbrowser?
13:49:59pondlifeNo, previous is the one. Previous browser would be wrong.
13:50:01JdGordonbecuase tihnk its wrong to
13:50:04ShaidI think the Paula chip helped, but it was mostly a kernel device
13:50:10*linuxstb is confused now
13:50:10Shaidof course, my amiga days are quite some time ago.
13:50:19JdGordonmarkun: how do you add voices? just lots of reaing or is there set words to say?
13:50:20*pondlife can't see the point in PrevousBrowser
13:50:41linuxstbWhat other previous screens could be useful when you press STOP?
13:50:46JdGordonthe point is select in the wps :p
13:50:49pondlifeDatabase? Settings?
13:50:59linuxstbDatabase is one of the previous browsers.
13:51:00JdGordonfiles, database, root
13:51:09JdGordonrec? fm?
13:51:11 Join l33ch [0] (i=c32e462c@gateway/web/cgi-irc/
13:51:19GodEaterB4gder: which version of festival does rockbox expect ?
13:51:27pondlifeHmm, play -> radio at last??
13:51:42JdGordonwe can go from any screen to any screen with this
13:51:47LinusNi think rockbox will never be able to evolve if we always have to retain the old behaviour
13:51:55l33chhi all
13:52:02*GodEater agrees with LinusN
13:52:13Shaidwell, yeah, evolution means change, it's part of its nature, really.
13:52:13linuxstbCan someone summarise the current/proposed behaviour? I'm confused now.
13:52:14*pondlife does too
13:52:20markunJdGordon: well, I just started to add the most likely phoneme for each letters in the dutch language, then just made it read some words and corrected the mistakes by coming up with new rules
13:52:21JdGordon... and retain the charcell compatability :p
13:52:24Shaidwe're not reactionists, are we?
13:52:27pondlifelinuxstb: It's much simpler
13:52:29LinusNi can go with previous screen
13:52:46JdGordonmarkun: ok, sounds like fun :)
13:52:46pondlifeJust navigate the menus, PLAY -> WPS, STOP -> menus
13:52:50markunJdGordon: let me check if I have my initial file, perhaps you will them see what I mean
13:53:02JdGordonna dont woryy about it
13:53:05pondlifebrowser is regarded as part of the menus
13:53:08JdGordoni was just curious..
13:53:16*JdGordon only speaks english
13:53:27linuxstbStarting with the WPS, is the proposal that: SELECT goes to last browser, MENU goes to root menu, and STOP goes to the previously entered screen?
13:53:30JdGordonpondlife: no its not...
13:53:43JdGordonlinuxstb: yes
13:53:45LinusNlinuxstb: yes
13:54:02pondlifelinuxstb: Yes!
13:54:03JdGordonpondlife: they arnt part of the menu, the menu just ties everything to gether
13:54:17pondlifeNo, I mean conceptually in the UI
13:54:22linuxstbThen I don't understand the STOP behaviour - either the root menu or last browser would seem more useful.
13:54:23pondlifeI know they're not really
13:54:29JdGordonthere also..
13:54:43*pondlife would like SELECT to go to previous too, not previous browser
13:54:52LinusNconceptually, the root menu is, well, a menu from where you can go to different screens
13:55:04pondlifei.e. SELECT in WPS is same as STOP but playback continues.
13:55:50linuxstbThat just seems very confusing - I would prefer buttons to be predictable in terms of going to absolute places.
13:55:50LinusNpondlife: i'd rather see SELECT in WPS as a shortcut to the browser
13:55:52*JdGordon finaly sees a massive advantage of svn over cvs... I can sitll make a patch :)
13:56:06pondlifeFair enough
13:56:08markunJdGordon: example of spanish:
13:56:59markunJdGordon: the english rules are very very complex compared to this as you can imagine :)
13:57:26JdGordondoes it handle non-english alphabets?
13:57:34Mikachuyes, swedish works fine
13:57:49markunyes, utf-8 is the default encoding
13:58:07JdGordonthose rules... does it try to transliterate the text?
13:58:34JdGordonlooks interesting :p
13:58:43markunJdGordon: well, if by that you mean change text to phonemes then yes
13:59:00Mikachumarkun: is there any way to handle words that are pronounced differently based on context?
13:59:10*JdGordon is a comp geek... not an english one.. wtf is a phoneme?
13:59:16markunMikachu: not really (like 'read' and 'read' right?)
13:59:54JdGordonor, letters being pronounced differently if they are in the middle or at the end of the word?
14:00:01Mikachuthat's no problem
14:00:06 Quit JETC- (Read error: 54 (Connection reset by peer))
14:00:13l33chI noticed problems while shutting down rockbox ! It does not power off but it's like standby :( Do anybody know about it ? port is iRiver H10 :) I have r12504M - 070227 with kocop11.diff :)
14:00:21 Join JETC- [0] (
14:00:31markunJdGordon: if you are a comp geek you could try wikipedia :)
14:00:38JdGordonnot that geekyy :p
14:00:43 Join barrywardell [0] (n=barry@
14:01:48markunthe stress in words can also be tricky to get right sometimes
14:01:59markunwithout adding too much rules
14:02:15JdGordonWritten in C++. :( how hard is porting c++ to c?
14:02:19roolkupetur: are you here?
14:02:21markunJdGordon: not hard
14:02:30markunit's almost all just normal C
14:02:45peturroolku: 10% ;)
14:03:12markunJdGordon, Mikachu: compare the english rule file to the spanish one ;)
14:03:15roolkupetur: I believe you wrote the i2c eeprom stuff for h300?
14:03:50JdGordonhaha wow!
14:03:52peturroolku: not really, just adapted it a bit to work on h300
14:03:55roolkupetur: any chance of a quick test if my re-shuffeling has broken it?
14:04:08markunl33ch: and without the patch it works fine?
14:04:12roolkupetur: yes, that is what I mean.
14:04:26roolkupetur: if you have the time here is a build:">
14:04:55l33chmarkun: I go to compile the latest source and i'll try it ... wait a few minutes
14:04:59peturroolku: no time now, but tonight ok ?
14:05:00pondlifeJdGordon: Sorry, but I get a patch error with the latest root menu in bootloader/main.c.
14:05:06markunl33ch: you can also download the latest source.
14:05:14markuneh, latest build :)
14:05:31JdGordonpondlife: ignore it... yshouldnt be there
14:05:45roolkupetur: yes, no problem. I might commit it anyway and fix later - it is not used on h300 as far as I know
14:06:01peturit isn't no....
14:06:12peturjust in the debug menu
14:06:29 Quit barrywardell ("Konversation terminated!")
14:08:52*LinusN fears an intense debate once the root menu is committed
14:09:48JdGordonthats wa sinevitable tho
14:09:55LinusNof course
14:10:03JdGordonas long as we are mostly in agreement its all that matters
14:11:12preglowand then it's time to start working on being able to skin it!
14:11:15LinusNi'm just glad that i'm apparently still young enough to be able to learn new things
14:11:41LinusNlike using the root menu
14:11:42preglowthe day i don't, i think i'll just throw myself in a lake
14:12:04preglowi think the root menu rocks
14:12:05peturand swim a bit...
14:12:14LinusNi love the root menu
14:12:34LinusNbut i think "root menu" is a somewhat misleading name
14:12:34preglowmakes rockbox a nicer place overall
14:12:38GodEaterI also fear the change of fearing change if you see what I mean ;)
14:12:45preglowand probably easier for noobs
14:12:53preglowLinusN: agreed
14:13:15LinusNi think perhaps "Start menu" or "Main menu"
14:13:25LinusN"root" implies a hierarchy that doesn't exist"
14:13:28preglowstart menu gets my vote
14:14:06preglowit's not like we really need to call it anything anyway
14:14:11preglowapple calls their main menu "ipod"
14:14:17LinusNwell, we might have to, in the manual
14:14:33pixelmaso JdGordon would have to rename root_menu.c ? ;)
14:14:36LinusN"Menu" takes you to the xxxx
14:14:45LinusNpixelma: lol * 2
14:14:52linuxstbI always thought the idea of the root menu was to create a hierarchy...
14:14:58JdGordonI was actually going to ask if root_menu.* should be moved and renamed
14:15:14LinusNlinuxstb: where is the hierarchy?
14:15:22JdGordonmoved into gui or menus and named ?
14:15:24preglowi think of the point of the root menu as just organising things in a more logical place
14:15:26linuxstbThe root menu is at the root, and everything else underneath.
14:15:47LinusNyou can go from any screen to another
14:15:54LinusNvia the root menu
14:15:56Febs"Main Menu" might be confusing as there is a main menu now which serves a different function.
14:16:07linuxstbLinusN: Exactly - _via_ the root menu.
14:16:12JdGordonMain Menu is renamed to Settings in te patch
14:16:22JdGordonnot via... you can go any to any skipping the root!
14:16:23markunbut start menu is also not right if you have the settings menu defined as the start screen :)
14:16:24linuxstbApart from the small number of short-cuts which by-pass the root menu.
14:16:34pondlifeLinusN: What's the harm in regarding it as a heirarchy?
14:17:00preglowyou might get surprised when it doesn't act as such
14:17:17roolkuroot menu is not so intuitive for technically challenged users - I like main menu
14:17:17preglowanyway, i believe it's quite redundant
14:17:23preglowpeople will use it, find out how it works, and then remember that
14:17:28markunpondlife: same problem as "why doesn't left bring me back to the WPS?"
14:17:34preglowyou don't need a perfect concept of anything to be able to use it
14:17:51FebsBut you do need to be able to have a consistent name for it to be able to document it.
14:17:59LinusNyes we do
14:18:11LinusNi think "Main menu" is the best
14:18:14linuxstbHow about "Rockbox menu" - with "Rockbox" as the title?
14:18:23pondlifeI vote "Main menu"
14:18:25LinusNlinuxstb: not bad
14:18:41markunbut then again, all menus in rockbox are a rockbox menu
14:18:48linuxstbBut not The Rockbox Menu.
14:18:56markunah no :)
14:19:00preglowmain menu gets my vote
14:19:09LinusNok, so we have "Main menu", "Start menu" and "Rockbox menu"
14:19:12GodEaterhe definite artcile
14:19:14FebsNotwithstanding my earlier comment, I think "Main Menu" makes sense as well.
14:19:19GodEaterI vote Rockbox menu
14:19:20preglowhmm, rockbox menu isn't bad either :P
14:19:26linuxstbMain Menu works, but is boring...
14:19:28LinusN(here we go again) :-)
14:19:32pondlifeNah, there are lots of Rockbox menus ;)
14:19:35*GodEater remembers this isn't a democracy :)
14:19:40preglowi definitely like rockbox menu now
14:19:58JdGordonthe title now is Rockbox Menu (it should be ROCKbox but meh..), settings is Settings
14:19:58pondlifeno, wait, preglow changed his mind ;)
14:20:18preglowit's set now that i've cycled through all three alternatives
14:20:23JdGordonpreglow: I cant commit :'( link is still down in singapore
14:20:34preglowis that so
14:21:34 Join Obsys [0] (
14:22:30bospaadjebtw, i just read the wiki page on the root menu - regarding the ipod keymap, any change in that? I saw a suggestion to use short select to stop
14:23:20 Quit Ribs ("eh eh ehhhh!")
14:24:14pondlifeYou know, if it goes in as Root Menu (TM), it'll become another Tagcache (TM) eventually ;-p
14:24:19FebsI predict "Rockbox menu" will cause confusion.
14:24:38preglowyou're predictions are noted
14:24:57bospaadjeI like "Rockbox menu"
14:25:11*linuxstb is moving towards "Main Menu"....
14:25:13LinusNi think "Rockbox menu" should work
14:25:16bospaadjesince it's essentially 'the' menu rockbox uses
14:25:34preglowwelcome to the rockbox decision making circus
14:25:41pondlifeNo, there are lots of menus in Rockbox!
14:25:44pixelmaJdGordon: I don't see a change in the stop behaviour with the latest patch
14:25:51LinusNall other menus have names too, like "Settings" etc, so I don't think there would be much confusion
14:25:54bospaadjethe rest is submenus (sort of), right?
14:26:09preglowroolku: do we really want runtime detection for something almost no one will have?
14:26:09Febs"Go to the Rockbox menu." "I'm in the Rockbox menu. The one with the sound settings." "No, go to the ROCKBOX menu."
14:26:20LinusNhowever, "Main menu" has more meaning to it
14:26:24markunAny left-to-wps guys in here?
14:26:39JdGordonpixelma: oh? nuts :p
14:26:39GodEatermarkun: stirrer :)
14:26:46pondlifeJdGordon: It's almost working
14:26:49roolkupreglow: I implemented it as a suggestion from LinusN
14:26:56pixelmaJdGordon: maybe because the patch didn't apply cleanly? I'll try again
14:26:58pondlifeBut Follow Playlist is only going to the folder still, not the file.
14:27:04GodEaterJdGordon: not worth trying to setup TOR to commit now ?
14:27:38 Join vertic23 [0] (
14:27:41roolkupreglow: the problem is some users of the MOD are "hardware-only" and cannot compile
14:27:42JdGordonGodEater: no, 12.30am here.. cbf... ill do it tomroow when everyone is asleep in case it breaks builds no one will notice :)
14:27:44linuxstbHow about referring to it as the main menu, but keeping the "Rockbox" title onscreen?
14:27:53pixelmapondlife: that worked for me with the previous patch (at least when going to filebrowser from WPS)
14:27:54GodEaterJdGordon: hahaha - good plan :)
14:28:10preglowlinuxstb: sounds extra good
14:28:14bospaadjelinuxstb, refer to it as "Rockbox main menu" then..
14:28:17pondlifepixelma: Try this. Play track 2, then press STOP
14:28:22LinusNpreglow, roolku: much like the 8Mbyte recorder build
14:28:32markunThe Rockbox main root menu ;)
14:28:34JdGordonpondlife: flow playlist only works if audio is actually playing
14:28:37GodEatermarkun: what is the "STOP" button on the Gigabeat ?
14:28:38pixelmapondlife: then I'm back in the root main menu atm
14:28:41LinusN"Rootbox" :-)
14:28:44preglowLinusN: yeah, but that doesn't add binary size, does it?
14:28:44GodEateris it "A" ?
14:28:57preglowbut i guess binary size isn't that much of an issue on h1x0
14:28:57pondlifepixelma: Go into Files and find a track 02...
14:29:07pondlifeThen play it and press STOP. What happens?
14:29:13markunGodEater: it's A, but I wish it was POWER
14:29:13LinusNpreglow: a few hundred bytes i guess, but only on the h100, i hope
14:29:38*GodEater is bemused by how it appears impossible to ever have the WPS display the "stop" graphic
14:29:39preglowand i definitely guess it's worth it to avoid having another build
14:29:43preglowor have someone maintain custom build
14:29:56pixelmapondlife: in the filebrowser?
14:30:02pondlifeNo, in the WPS
14:30:05JdGordonpixelma: stop works correclty here....
14:30:35pixelmapondlife: I said root menu atm, I'll try again
14:30:36markunmaybe we should even put some code in a #if 0 to jump to the WPS when you press left tree to keep some people happy :)
14:30:36pondlifeJdGordon: Can you try this too? Navigate to a track 2 in the file browser and play it. Then press STOP in the WPSS
14:30:44JdGordonpondlife: just tried.. works fine
14:30:48JdGordonoh, stop...
14:30:49 Join Arathis [0] (
14:31:05pondlifeSELECT works, but STOP almost works
14:31:15JdGordonbut like I said.. follow isnt supposed to work if audio isnt playing
14:31:20pondlifeAha, I see
14:31:23preglowroolku: i guess you don't have the ability to wake up on rtc alarm?
14:31:23linuxstbJdGordon: How is "previous screen" decided? If you enter the WPS from the root menu, is the root menu the previous screen?
14:31:32LinusNpreglow: sure we do
14:31:50linuxstbJdGordon: So which screens have a shortcut button to the WPS?
14:31:54pondlifeI think of follow playlist as meaning "update the previous screen"
14:32:07JdGordonbrowsers, menus atm...
14:32:12pondlifePLAY should work from everywhere soon, right?
14:32:18JdGordonshortcuts can be added on a target by target basis after
14:32:19LinusNpondlife: i think of follow playlist as "update the browser"
14:32:32preglowLinusN: is it wired correctly for it too?
14:32:46roolkupreglow: alarm work with the MOD
14:32:50pondlifeLinusN: OK
14:32:53preglowmore files i need to include in my patch :/
14:33:09preglowi renamed HAVE_ALARM_MOD to HAVE_RTC_ALARM
14:33:51JdGordonLinusN: do you know what needs to be fixed for the h300 bootloader for the alarm to turn it on?
14:34:09preglowalard mod is applicable in this case, but not in the majority case
14:34:19*JdGordon thinks he remembers preglow saying his patch was for the rtc in the h300
14:34:20*preglow wants new fingers
14:34:21Mikachuhaha, a lard mod eh?
14:34:39linuxstbJdGordon: The ipods and h300/x5 have the same rtc chip
14:34:45preglowJdGordon: it is, kinda, but the drivers need to be merged
14:34:54 Join tucoz [0] (i=528612c1@gateway/web/cgi-irc/
14:36:03tucozLinusN: I see you committed some changes to the hxxx bootloaders. Is this the same bootloader as the one Slasheri has been working on wrt rambox etc?
14:36:20 Quit Mouser_X (Read error: 110 (Connection timed out))
14:36:43tucozMy concern is mainly the noise on boot
14:37:40LinusNtucoz: i haven't worked on that issue, no
14:37:45tucozi see
14:37:49 Join norbusan [0] (
14:38:04preglowLinusN: seems like it gained about 6kb
14:38:14JdGordonLinusN: as long as the AC is plugged in with rockbox on its safe?
14:38:17roolkuhm, I wonder why recorder 8MB went up 40 bytes and Rec down 16 - compiler version differences?
14:38:31JdGordoncompressed bins
14:38:35preglowroolku: for rec, perhaps, but rec 8 meg shouldn't vary that much
14:38:53tucozin any case, it's not that bad nowadays as the noise is only ~0.5s compared to the 1-2s it was earlier :)
14:39:10LinusNpreglow: ah yes, all the alarm setting ui stuff went in as well
14:39:11tucozs/is/lasts for/
14:39:31roolku...and the clock plugin
14:39:59LinusNroolku: the plugins don't count in the size table
14:40:53roolkuI am also suprised that the h300 build didn't change as I juggled the code around a bit
14:41:54 Join pearldiver [0] (
14:42:13tucozshould the rtc code really be included in the standard core for the h120 as this requires a hw mod?
14:43:00 Part norbusan
14:44:31*tucoz reads log
14:45:14tucoz14.27.41 # <roolku> preglow: the problem is some users of the MOD are "hardware-only" and cannot compile
14:46:02tucozi am impressed by people able to dissect their player and solder a chip in that tiny piece of machinery and _not_ being able to compile rockbox
14:46:29roolkutucoz: I was suprised as well
14:46:30preglowi'd be able to teach my mother to compile rockbox, heh
14:46:48 Quit JdGordon ("zzzzzzzzzzzz")
14:47:02LinusNit's not about being able to compile rockbox, it's about keeping a patch in sync
14:47:14tucozthat is a valid point
14:47:22LinusNand that patch is not trivial
14:47:38linuxstbWhy don't we just add it as an extra build then?
14:47:44LinusNwe could
14:47:45preglowah, but you can commit it without adding it as default for all h1x0 builds
14:47:46tucozlike the 8MB?
14:47:52preglowbut anyway, i don't mind much
14:47:58linuxstbSeems like run-time detection of RTC would be more trouble.
14:47:59preglowi don't think there will be many optional features
14:48:02LinusNi don't mind either
14:48:15tucozme neither. was mainly thinking of policies
14:48:23preglowit's easier for everyone and we've got kilobytes to burn
14:48:23LinusNespecially since i intend to mod my h140 soon :-)
14:48:32tucozwhat is counted for as an 'official' mod
14:48:33preglowLinusN: it is tempting indeed...
14:48:49preglowbut i don't know if my h120 casing can take another opening up
14:48:53roolkuI think there are now 14 modded h1x0 out there
14:49:16LinusNabout 12 more than the 8mbyte mod :-)
14:49:22tucozis it possible to f
14:49:33tucozfit that chip inside a h120?
14:50:25tucoznot that i will go anywhere near my h120 witha soldering iron combined with my lack of soldering skills
14:50:36tucozthanks LinusN
14:50:41 Quit cadu (Read error: 104 (Connection reset by peer))
14:50:49Slasheritucoz: ah, that issue. can you compile rockbox yourself?
14:51:01LinusNbtw, the pictures in the forum post should be moved to the wiki
14:51:53tucozSlasheri: sure. but not at the moment. I just moved to a new apartment, and i wait for my internet connection to be set up
14:51:55LinusNand perhaps degrade the ds1339c variant to "optional"
14:52:08Slasheriyou should try adding audiohw_reset() to the apps/main.c
14:52:15amiconnpreglow: The bin for recorder8mb is as compressed as the normal recorder bin
14:52:18Slasherido it just after the kernel init stuff and see if the problem is gone
14:52:33Slasheriif it works, that can be committed
14:52:39amiconnBut the binaries aren't identical; there are a few special measures added for 8MB
14:52:42roolkuLinusN: the startup problem with the ds1339c doesn't exist anymore
14:53:22tucozok. I can take my laptop to a friend some time later this week. I'll let you know if it helps
14:53:25roolkuLinusN: but I will reorganise the wiki page a bit
14:53:32LinusNroolku: nice
14:53:35Slasheriroolku: hmm, rtc runtime detection. interesting :)
14:53:39tucozSlasheri: thanks btw
14:53:41preglowamiconn: could you give me quick summary as to what needs to be done to get rid of long calls? postfix all declarations by IDATA_* and prefix all definitions by the same?
14:53:47Slasheritucoz: lets hope it helps
14:54:10amiconnpreglow: ICODE_ATTR, not IDATA
14:54:19amiconnAnd only functions, of course
14:54:43preglowah, yesyes
14:54:45preglowwrong thing came out
14:54:58amiconnThere are some places with special ICODE_*
14:55:24 Quit petur ("worrrrk")
14:56:25roolkuSlasheri: well strictly speaking it only detects if there is an I2C device at that address, but the likelyhood of someone soldering in a temerature sensor with the same hardware address are minute
14:56:40amiconnpreglow: There's an easy way to find out where the declarations and definitions are clashing: add __attribute__((long_call)) to ICODE_ATTR for arm and the try to compile...
14:57:06tucozsee you
14:57:09 Part tucoz
14:58:06pondlifeI really like the root menu, but wish very much that Files always took me to the root directory.
14:58:20preglowshut up!
14:58:23Slasheriroolku: but i think that is a good feature
14:58:42pondlifeI mentioned it once but I think I got away with it.
14:58:50pixelmapondlife: you missed the discussion yesterday...
14:58:53 Join webguest53 [0] (i=c0231115@gateway/web/cgi-irc/
14:58:53 Join OgMaciel [0] (n=omaciel@ubuntu/member/gnukemist)
14:59:00*pondlife looks at logs
14:59:10roolkuSlasheri: might encourage more people to do the mod
14:59:29*preglow needs another h1x0
14:59:32pondlifeIt just seems very wrong to go right from root menu to files, then right takes you somewhere else.
14:59:35Slasheriroolku: and it didn't even increased the binary size, except for h1xx of course.. that is great
14:59:39pondlifeSorry, I meant left.
14:59:43LinusNpondlife: then don't use Right
14:59:47 Join Criamos [0] (
14:59:49 Join TheSphinX^ [0] (
14:59:58LinusNpondlife: use Select
14:59:59Mikachuyou could confuse people immensely by making right go to / and select go to last folder :)
15:00:14*pixelma has to use 'right' but oh well...
15:00:19LinusNpixelma: hehe
15:00:29TheSphinX^hi there
15:00:32pondlifeIs there no select on Ondio at all?
15:00:52LinusNpondlife: this is why i want to stress that it is *no hierarchy*
15:01:01pixelmapondlife: there is but it is used as "resume"
15:01:07pondlifeWell, why not make it a hierarchy?#
15:01:16LinusNthe main menu is not the parent of the root directory
15:01:25pondlifeIt's almost there now, surely that would be an improvment?
15:01:31LinusNabsolutely not
15:01:34pondlifeI misunderstood the whole patch then.
15:01:39 Join x1jmp [0] (
15:01:44pixelmathat didn't come out right
15:01:56LinusNi don't want to start over browsing from the root just because i have entered the main menu
15:02:28LinusNthe browser is a separate component, which happens to be selectable from the main menu
15:02:35LinusNand from the wps
15:02:50LinusNand it remembers the position for convenience
15:02:52pondlifeDarn it. I thought this was aimed at clearing up this mess.
15:02:57LinusNit is
15:03:02LinusNand it is no mess
15:03:12LinusNwhere is the mess?
15:03:23pondlifeIn my head I guess!
15:03:39LinusNwe have finally gotten rid of the existing hierarchy
15:03:47LinusNand now we have a flat structure
15:03:52pondlifeHierarchies are good though
15:04:13LinusNnot when you mix the file system hierarchy and the menu hierarchy
15:04:28pondlifeIt certainly looks like a hierarchy to the uninformed (i.e. me).
15:04:42LinusNi.e the main menu is not the parent of the root folder
15:04:54pondlifeI understand that now, but didn't get that from using the UI
15:05:38pondlifeI saw MENU -> root, PLAY -> WPS, STOP -> last place in a hierarchy.
15:05:49LinusN*but* i think Left-to-mainmenu is a bad thing in this
15:06:05pondlifeYes, very probably now.
15:06:26LinusNleft-to-anything from the file root is a bad thing imnsho
15:06:37pondlifeMy main objection to left-to-WPS was for blind users, so you can be sure that an extra left won't go somewhere unvoiced.
15:07:12pondlifeIf the root IS regarded as a level above the browser, it works fine. Left at root does nothing I assume...
15:07:19*pondlife tries
15:08:06pixelmabtw... don't know what I did wrong earlier but "stop" works as expected now
15:08:15pondlifeI just can't get my head around the paradigm of this UI.
15:08:25pondlifeSeems to be lots of bits glued together
15:08:28markunpondlife: then just wait till it's finished
15:08:51preglowpondlife: you're trying to think too much about it
15:08:52preglowjust use it
15:08:58pondlifeI am trying to use it.
15:09:14pondlifeI'm finding it illogical
15:09:24pondlifeThen I'm thinking "why?"
15:09:25preglowi think it worked pretty great
15:09:31markunpondlife: what would be more logical then having 1 place to access everything?
15:09:36pondlifeIt's better
15:09:38GodEaterme too - seems very intuitive to me
15:10:07 Quit pill (Nick collision from services.)
15:10:11 Quit Febs (Read error: 110 (Connection timed out))
15:10:19 Quit webguest53 ("CGI:IRC (EOF)")
15:10:19pixelmahmm... maybe the left-to-main-menu is really the thing that's causing confusion
15:10:25 Join pill [0] (
15:10:32LinusNpixelma: i agree
15:10:42markunpixelma: I think so too
15:10:56LinusNleft-to-main-menu implies a hierarchy that doesn't exist
15:10:57markunLinusN: damn lag :)
15:11:04preglowLinusN: after a fair deal of disassembling, i've noticed i can expect no kind of signal from the pcf that an alarm has occured. any ideas on how to deal with that? :/
15:11:29 Quit Kittt0s (Read error: 104 (Connection reset by peer))
15:11:39LinusNpreglow: i guess you will have to stick to comparing the alarm and current time
15:11:43 Quit pill (Nick collision from services.)
15:11:56preglowLinusN: then we'll have multiple triggerings if the user switches it off after an alarm occurs, then quickly on again
15:12:00 Join Kitt0s [0] (
15:12:00 Join tick [0] (i=c27f0812@gateway/web/cgi-irc/
15:12:01 Join pill [0] (
15:12:14LinusNpreglow: how come?
15:12:25preglowLinusN: well, the alarm time and current time will still match
15:12:40LinusNpreglow: really? how?
15:12:51tickLinusN: with the root menu, I'd say LEFT should NOT go to the root menu from the root dir. Before root menu I was for LET-to-WPS
15:12:51Mikachucan't you disable all the other flags when they match the first time?
15:12:52 Quit pill (Nick collision from services.)
15:13:02 Join _pill [0] (
15:13:14preglowLinusN: i disregard seconds completely, since i don't know how long a startup will take. so i just compare minutes and hours, and you've got ample time to switch rockbox off then on again before the current minute changes
15:13:20LinusNtick: full ack
15:13:31Mikachupreglow: maybe you can encode state in the second register?
15:13:31preglowbut i guess i could do some kind of threshold based on the seconds value, but that might break nicely during a long dircache update
15:13:41LinusNpreglow: surely it can't take a minute to boot?
15:14:04preglowLinusN: i don't know, i just know it can take long
15:14:07LinusNand couldn't you clear the rtc wakeup interrupt flag?
15:14:09preglowwhat with dircache, database, etc
15:14:12tickLinusN: b/c if we'd do this we'd make the user to feel the root menu as the parent of the root dir. Which is not the case.
15:14:14 Join Pill [0] (n=Pill@wikimedia/Pill)
15:14:26preglowLinusN: all interrupt flags and interrupt masks are clobbered by the bootloader
15:14:29LinusNpreglow: you check the alarm interrupt *after* the database init?????
15:14:32preglowLinusN: including the wake on rtc flag
15:15:00LinusNtick: exactly
15:15:17preglowLinusN: ah, right, i haven't actually checked the order yet, i just did a stupid assumption :)
15:15:19PillCould you answer me one question, this user _pill, has he just changed his nick some times?
15:15:36tickLinusN: does JdGordon have the same opinion? How it's implemented now (I mean the latest patch)?
15:16:11MikachuPill: look in the log
15:16:15preglowLinusN: rtc_init is called pretty quicly, so i guess i could do a threshold based on the seconds value
15:16:19pondlifeLinusN: Sorry, phone rang. I agree that the left to menu makes it feel like a hierarchy.
15:16:27LinusNit currently does left-to-mainmenu
15:16:32pondlifeHence most of my confusion
15:16:33PillMikachu | where?
15:16:42LinusNand that should be removed, methinks
15:16:47l33chbtw..when I build bootloader from source codes is it only for iRiver H10 UMS .. not MTP ? I have MTP model ... coz when I put the new one on player it wrote some error and goes to emergency mode.
15:16:54 Join Domonoky [0] (
15:16:57preglowMikachu: what second register, btw?
15:17:29Mikachupreglow: i may be completely confused about how it works, i meant the place you tell the rtc chip what second to wake up on
15:17:48preglowMikachu: i can't put anything in that, then it becomes a valid wakeup time
15:18:14Arathisl33ch: 5/6 or 20 gb?
15:18:45preglowbut no, i'm leaning towards just putting in a seconds threshold now, and encoding the full date in the alarm register
15:18:53l33chArathis: 6GB
15:18:53preglownot just minute and hour with the rest marked as irrelevant as now
15:18:55 Part pondlife ("lunch...")
15:18:56PillMikachu | Nick changes aren't logged, thanks anyway.
15:18:56 Quit tick ("CGI:IRC (EOF)")
15:19:04 Part Pill ("-")
15:19:08preglowthis will be vastly overkill for h3x0 and x5, but
15:19:58Arathisl33ch: anyway, I have a 20GB MTP and the bootloader works fine. You could either post the error msgs or just try the precompiled at (5GB-MTP for you)
15:20:05LinusNpreglow: full date is a good thing
15:20:18Mikachupreglow: you can't mark the seconds as irrelevant and still store something in them then, i guess?
15:20:30TheSphinX^ is this right?
15:20:47TheSphinX^arg looks only good in my notepad
15:21:12***Saving seen data "./dancer.seen"
15:21:23preglowMikachu: not without using undefined behaviour, no
15:21:46preglowLinusN: sure, i just need to stuff in or adapt some of the leap-year logic and stuff
15:21:56preglowrockbox has some already, just not usuable by me for this purpose, i think
15:22:29Mikachupreglow: could you save something in nvram.bin then?
15:23:13l33chArathis: I use this precompiled from that link ... but I wanted to try if new is working too coz I had problems with shutting down player ... now it looks working after upgrading to the latest firmware build. I go test another Theme maybe it was caused of it ;-)
15:23:43TheSphinX^LinusN: : <- is there any false in it?
15:24:02LinusNTheSphinX^: it's a total mess
15:24:04Mikachuit looks even stranger now
15:24:18Arathisl33ch: When did you last update your bootloader? They were changed at the end january
15:24:34TheSphinX^ok i try to let it look better in the browser, <3 notepad ^^
15:24:34preglowMikachu: i'm not too familiar with that
15:24:37markunTheSphinX^: what kind of font are you using?
15:24:51markunTheSphinX^: try one with fixed width in notepad
15:24:53LinusNTheSphinX^: it looks bad in my notepad too
15:25:10markunTheSphinX^: (like courier new or something)
15:25:44amiconnpreglow: Maybe I don't understand the real problem, but shouldn't the fact that a valid wake-up time is set tell you that it woke up from alarm?
15:26:16Mikachuhm yeah, can't you just set the wake-up time to invalid after it matches once?
15:26:20LinusNTheSphinX^: btw, the fm and recording will not be accessible from the menu in the future
15:26:27amiconnI thought that the alarm wake-up disables alarm as about the first thing it does...
15:26:46amiconnSo normally there shouldn't be a valid alarm time, only in the wake-up case
15:26:59LinusNTheSphinX^: there will be no duplicates
15:27:19l33chArathis: I bought this player at the begging of February so I put this bootloader from 30 of January or was it 31 ;)
15:27:27 Quit miepchen^schlaf (Read error: 113 (No route to host))
15:27:32 Join miepchen^schlaf [0] (
15:27:48preglowamiconn: normal targets disable alarm in other means, i think
15:27:51l33chbegin :)
15:27:51Arathisl33ch: 31st, but okay then ^^
15:28:00preglowamiconn: so that the alarm time registers are used to store the alarm time for the next time you want to set it
15:28:12TheSphinX^LinusN: better ? :)
15:28:15Arathisl33ch: If the bootloader will still produce errors please post them
15:28:33LinusNpreglow: you could cheat and set it to some time in the very distant future
15:28:41l33chArathis: oki now i'll test this iCatcher Theme
15:28:43Mikachupreglow: won't it work just as fine if you use a config setting for it instead?
15:28:53preglowMikachu: it will, i'd rather just avoid it
15:29:35LinusNTheSphinX^: you can remove all links from the menu to the other screens
15:29:39Mikachuso if you leave the registers intact anyway, can't you just change the seconds field to mark that it's been triggered?
15:29:51markunTheSphinX^: what's 'root'?
15:29:56Mikachulike if it's :01 you want to trigger and if it's :02 you don't
15:30:05markunTheSphinX^: and you are missing the settings menu
15:30:09LinusNmarkun: i think "root" is the Root menu
15:30:22markunand menu is the settings menu?
15:30:23l33chthx for help ;-) later guys and maybe some woman :)
15:30:25LinusNaka Rockbox menu, aka Main menu
15:30:49markunor quick menu?
15:30:51 Part l33ch
15:30:57LinusNTheSphinX^: and WPS is on the same level as the other screens
15:31:14LinusNas is the "menu"
15:31:35preglowLinusN: hmmm, that solution would be clean in the code sense, and dirty in the conceptual sense :)
15:31:38preglowi think i like it, heh
15:31:52LinusNpreglow: :-)
15:31:53*amiconn is still annoyed by himself missing this fine fmr on ebay :(
15:32:11*LinusN feels bad for amiconn
15:32:25amiconnThe thing went for just 63EUR ...
15:32:42amiconn...including 1 year warranty
15:32:49LinusNi missed a fine gigabeat because the auction ended in the middle of the night
15:34:58markunLinusN: I think I bought mine at 2:00 or something :)
15:35:26markunand my girlfriend's iriver T30 at maybe 4:30
15:36:28 Quit vertic23 ()
15:36:57LinusNTheSphinX^: your latest bmp is still wrong
15:37:55 Join tvelocity [0] (
15:38:39preglowi've actually gotten up in the middle of the night to shotgun a couple of desirable auctions
15:38:53GodEaterI thought the term was "snipe"
15:39:05preglowit is, but i haven't used it for years, so forget
15:39:13preglowboth are about shooting :)
15:40:06GodEatershotgun sounds a bit more random though
15:42:11preglow_everybody_ is selling chargers for h1x0
15:43:19*TheSphinX^ LinusN work in process
15:43:42 Quit Obsys ("Chatzilla 0.9.77 [Firefox]")
15:43:49 Join Llorean [0] (n=Llorean@
15:43:58markunLinusN, TheSphinX^: what about this?
15:44:19LinusNmarkun: much better
15:44:55markunLinusN: what other shortcuts are there? They might differ from target to target of course
15:45:20LinusNcan't think of any right now
15:46:09TheSphinX^markun i think from wps to db/file ther must be placed an "or"
15:46:23LinusNyes, somehow
15:46:57LinusNbut that's a minor detail, i'd say
15:47:36preglowdoesn't seem like h1x0 is easy to come by now
15:47:40 Join Febs__ [0] (n=chatzill@
15:47:43 Nick Febs__ is now known as Febs (n=chatzill@
15:48:06 Quit GodEater ("CGI:IRC")
15:48:21 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
15:48:21markunLinusN, TheSphinX^: should there be a setting for that?
15:48:45markunlike 'default browser'
15:49:12TheSphinX^i thought the lastused of them
15:49:56LinusNhmmm, where does it go if you resume a playlist created by the database?
15:50:22*LinusN tries
15:50:25 Quit Galois (Remote closed the connection)
15:50:46 Join Galois [0] (
15:51:01markunLinusN, TheSphinX^:
15:51:36LinusNmarkun: wooooooo
15:52:56amiconnHmm, does Stop do anything atm in the root menu?
15:53:03 Quit Nico_P (Remote closed the connection)
15:53:04 Join pondlife [0] (
15:53:33*amiconn still didn't try it.... for some reason
15:53:51LinusNamiconn: it stops the playback
15:54:32LinusNok, if you start playing in the database, turn it off and then resume, Select in WPS takes you to the file browser
15:55:05LinusNfair enough, i guess
15:55:40LinusNperhaps it should remember that it was in the database, but i wouldn't call it a showstopper
15:57:02amiconnI think rockbox should remember the last browser used for playlist creation. This way it could also store a pseudo path when the last browser was the db, and implement 'follow playlist' this way
15:58:06pondlifeWhat does the SVN code do?
15:58:28markunpondlife: did you look at my pdf?
15:58:30LinusNamiconn: i agree
15:58:31Lloreanamiconn: Follow Playlist in DB still wouldn't work.
15:58:41pondlifemarkun: No, where is it?
15:58:43LinusNpondlife: the svn code has only one browser
15:59:01amiconnLlorean: Why not?
15:59:10Lloreanamiconn: What if you entered a "Genre" heading from the main DB, to insert a whole album, then left and entered an "Artist" one to see the whole list of artists, and inserted two more albums from there?
15:59:37LloreanHow is it to know, from a filename, which series of filters you followed to get to that song in the database?
15:59:47amiconnHmm, I think playlists created from the db should store db pseudo paths instead of filesystem paths
16:00:02LloreanDynamic ones probably could/should.
16:00:03LinusNthen the playlist code would be a nightmare
16:00:47amiconnThe db->file mapping has to happen anyway at some point
16:00:59amiconnIt's just that this point would be deferred
16:01:13Lloreanamiconn: What about the Playlist Catalog then. Once you save them, they should become standard M3U, and then once those are loaded, they become un-followable again, right?
16:01:26roolkuamiconn: since the path can change dynamically I don't think this is a good solution
16:01:34LinusNLlorean: yes, but playlists are not database related
16:01:42preglowLinusN: what would you say is a reasonable maximum time from alarm to rockbox start? 15 secs?
16:01:47preglow10 would perhaps be more sane
16:01:58preglowLinusN: from alarm to rtc_init(), even
16:02:00LinusNpreglow: 10 sounds fair
16:02:08roolkuamiconn: for example you can have playcount in the path, which will move the song out of the playlist as soon as it is played
16:02:15 Join pixelma_ [0] (i=pixelma@rockbox/staff/pixelma)
16:02:33preglowLinusN: i'll just use a date in 2099 as an "alarm disabled" date
16:02:47*amiconn didn't think about playcount - never used such obscure features
16:02:55LinusNpreglow: sure
16:02:57pondlifepreglow: Short-termism, there :)
16:03:07amiconnpreglow: Don't forget about the "stupid iriver offset"
16:03:22preglowamiconn: i did forget about that
16:03:27roolkualso songs will appear in many places in the database tree, I would suggest follow playlist only to work for the file browser
16:03:32LinusNlame iriver developers....grr...
16:04:05markunpoor korean guys.. working their asses of to bring use these mp3 players and this is what they get.
16:04:27LinusNroolku: i agree fully, however, it could remember the last browser when resuming, starting in the database root
16:04:40LinusNmarkun: :-)
16:04:46B4gderI watched the "I blow the bus up with my iriver h10" 24 episode yday
16:05:02B4gderI think I'll avoid those h10s them seem nasty
16:05:10roolkuLinusN: for my usage pattern that would not be so good, but maybe less confusing for others?
16:05:15 Quit pixelma (Nick collision from services.)
16:05:15 Nick pixelma_ is now known as pixelma (i=pixelma@rockbox/staff/pixelma)
16:05:16LinusNB4gder: which rockbox build did he use? :-)
16:05:25B4gderit wasn't that easy to detect
16:05:30B4gdermust've been a custom build!
16:05:45TheSphinX^LinusN: , markun, amiconn: ?
16:05:58roolkuLinusN: I would like to find associated files to the song I am playing (i.e. lyrics, cue, albumart) without having to switch file view back and forth
16:06:30LinusNroolku: how would you do that?
16:06:45preglowand here's people complaining about just the eq blowing up
16:06:48LloreanSelect in WPS always going to Filetree?
16:06:53preglowthey have no sense of perspective!
16:07:10LloreanI still like an option in the WPS Context menu for "Go to currently playing file"
16:07:15LinusNTheSphinX^: almost correct
16:07:26LinusNLlorean: that would be a nice feature
16:07:29roolkuI see the database as a query engine that helps me create playlists, rather as an alternative tree
16:07:44LinusNTheSphinX^: you forgot the "follow playlist" option
16:07:57*TheSphinX^ ups
16:08:04pondlifeHmm, whatr should PLAY do in the WPS context menu?
16:08:07LinusNroolku: i believe that's a good way to see it
16:08:07roolkuso browsing is really in the file tree, where I have all files
16:08:20LinusNpondlife: go to wps?
16:08:27pondlifeYes, it should!
16:08:39roolkuand there is only one instance of every file and it doesn't move around
16:08:46amiconnmarkun: The H300 rtc implemetation in the original firmware has broken leap year handling
16:08:54pondlifeIt doesn't here with the root menu patch, don't know if that's a new bug though.
16:08:54markunamiconn: I know
16:09:18TheSphinX^i've an question (before is focus totaly on root_menu)
16:09:18TheSphinX^how aboaut adding these lines to the viewers.config:
16:09:18TheSphinX^dib,rocks/rockpaint, 01 10 01 10 01 10
16:09:18DBUGEnqueued KICK TheSphinX^
16:09:18TheSphinX^mpeg,viewers/mpegplayer,5D 7F 5D 7F 5D 7F
16:09:18TheSphinX^? =)
16:09:31amiconn...because they use a 35 year offset to how the rtc chip is supposed to be used
16:09:45markunpondlife: I think it's more important what the patch should do, then what it actually does right now
16:09:55preglowamiconn: ok, so i should use 2064 as "alarm disabled" year, then?
16:10:06amiconnRockbox uses 36 years and at least gets leaps years right this way
16:10:08pondlifeBut it's about to be committed isn't it? I'm just testing.
16:10:44pondlifeI know what I want the UI to be, and it's not happening...
16:11:09amiconnpreglow: Not sure whether iriver of tests the alarm year at all. If it doesn't, the year shouldn't matter much
16:11:20preglowamiconn: indeed
16:11:20amiconnIirc the rtc driver handles the special offset for h300
16:11:30preglowit does
16:11:36preglowmno, timefunc.c does
16:13:36 Quit Domonoky (Read error: 54 (Connection reset by peer))
16:13:55 Join perl|work [0] (
16:16:45 Join illriginal [0] (
16:18:27 Join Mouser_X [0] (
16:18:41amiconn+6K just for RTC... which 99.9% of all H1x0s don't have... :\
16:19:40 Join desowin [0] (
16:20:00LinusNamiconn: does it matter?
16:20:24GodEaterspeaking as an H140 owner, it doesn't matter to me
16:20:30LinusNme neither
16:20:33preglowit's worth it just to not have to bother with another build
16:20:34preglow6k i snothing
16:20:56LinusNi won't be able to measure the battery or performance impact in any way
16:21:16 Quit lost|X40 (Remote closed the connection)
16:21:33LinusN6kbytes out of 32mbytes
16:21:37desowinodd, today I wasn't able to reproduce last song not scrobbled with my patch, can anyone test it ?
16:24:14GodEatermarkun: Can you take a look at legion's post in the gigabeat install forums? Is he flogging a dead horse, or is it possible to get his Gigabeat working without the original firmware being on it ?
16:24:56markunGodEater: of course, with the GBSYSTEM.ZIP from the wiki you don't need the original firmware
16:25:02LloreanGodEater: I believe, the GBSYSTEM.ZIP from the install page actually requires 'dummies' of all the files the flash loader needs to detect other than the one file our bootloader replaces.
16:25:23Lloreanerr, s/requires/contains
16:25:24 Join lostnihilist [0] (
16:25:27GodEaterright - in that case I'll bear with him and try and help some more :)
16:26:12markunI guess just replacing 1 files is not so easy for everybody..
16:26:41Lloreanmarkun: People seem to want to reformat their gigabeat first.
16:27:30preglowLinusN: would you say relying on static function variables to be inited to 0 by the bss clear is good?
16:28:14LinusNi believe so, but i often init them anyway for clarity
16:28:31preglowyeah, that's what i think is best anyway
16:28:41preglowreal binary size isn't affected anyway, just size on disk
16:30:17 Quit Mouser_X (Nick collision from services.)
16:30:26 Join Mouser_X [0] (
16:31:38 Quit Rondom ("Ex-Chat")
16:35:29 Part LinusN
16:36:35 Quit Thundercloud_ (Read error: 110 (Connection timed out))
16:40:16 Quit jaebird (Read error: 113 (No route to host))
16:40:21amiconnpreglow: Initing static variables to 0 does exactly that - relying on the bss to be zeroed out
16:40:54preglowamiconn: they won't be included in the bss if i init them to zero, will they?
16:40:59amiconnSo writing = 0; doesn't change anything in the binary, hence it should be done to make clear this variable is initialised
16:41:27amiconnThey will, unless you use -fno-zero-initialized-in-bss
16:41:32amiconn...but we shouldn't use that
16:41:49preglowah, ok
16:41:59preglowi've seen some places to rely on static vars being inited to zero
16:42:04preglownamely the archos rtc alarm code
16:43:07amiconnI learned about this stuff when trying to find out why some plugins acted funny after running others
16:43:34amiconn...because a long time ago the plugin loader didn't clear the plugin bss....
16:48:21CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
16:48:21*GodEater ponders ruining his 83 days uptime for a kernel upgrade
16:48:29preglowoh noes! ruining uptime!
16:49:26GodEaterit's a serious consideration :) 83 days is a record for me
16:49:49preglowi've had server boxes that lasted four-five times that
16:50:01Mikachuyou must have several local root exploits open with that much uptime
16:50:07B4gderso set back the time before you shutdown, so that you can advance it 84 days immediately after boot-up! ;-)
16:50:58preglowB4gder: doesn't look like the zune-linux dudes are too keen on answering...
16:51:10preglowthey're probably too busy talking together some place with no logging
16:51:16*amiconn upgrades too often to even get near this uptime
16:51:35*markun doesn't care much for uptime
16:51:37GodEaterI might as well do it - the yearly building power down is soon
16:51:47GodEaterso my uptime will go anyway
16:51:51Mikachuget a ups!
16:52:08GodEaterhah - business justication ?
16:52:22GodEater"I want a l33t uptime" isn't going to cut it I think
16:52:24markunMikachu: but how to connect the server to the UPS without taking the power off..
16:52:37Mikachujust plug it in to the local power net
16:52:38GodEaterit's only a desktop anyway
16:52:45preglow83 days uptime on a desktop????
16:52:54preglowi had eight on mine yesterday and it wasn't a pleasant sight
16:53:05Mikachumine is 12 days currently
16:53:12Overandwindows box - \\ZAPHOD has been up for: 9 day(s), 19 hour(s), 13 minute(s), 0 second(s)
16:53:13*Llorean wishes he could ask "How could the Rockbox guys get involved if Zune-linux won't share any information" but it still hasn't sent me an activation email
16:53:14Mikachuthe average is around 40 i think
16:53:19 Join toffe [0] (
16:53:29preglowi shut my computer off at nights, ordinarily
16:53:43 Quit amigan (Read error: 110 (Connection timed out))
16:53:52markunLlorean: on which forum?
16:53:57 Join Du4LiTy [0] (
16:53:58Lloreanmarkun: The Zune-Linux one.
16:54:18preglowLlorean: bagder already pretty much did that, didn't he?
16:54:21LloreanI even sent an email to their admin yesterday, but haven't received a response.
16:54:21x1jmpDoes anyone believe in the Zune-linux project?
16:54:25Lloreanpreglow: They responded to him.
16:54:38LloreanOr not "they"
16:54:40LloreanSome user there.
16:54:42Llorean"Maybe if the Rockbox guys get invovled with the Zune Linux Project they could help get it done faster. And then work on Rockbox."
16:55:17LloreanAs part of a very slightly larger response that tries to excuse them for not sharing details with the excuse "They want to be sure they get the credit" basically
16:55:48Mikachumost running programs i have, have been running for over 9 days
16:56:02Du4LiTyhey guys, i just got rockbox on my ipod video
16:56:13Du4LiTyi was wondering what you guys use to organize your music libraries?
16:56:35x1jmpThere was a post on a forum apparently from the zune-linux asking how to program...
16:56:45Llorean \Artist\Album\TrackNum-Title.Extention
16:56:55x1jmps/zune-linux/zune-linux author
16:57:01markunx1jmp: I don't
16:57:04 Join Thundercloud_ [0] (n=thunderc@
16:57:04Mikachu( ps axf -o etime,comm )
16:58:11Mikachuirssi is only 7 minutes younger than the kernel, and that's the end of my uptime report
16:59:31x1jmpI was even surprised that zune-linux is still a topic
16:59:35 Quit TheSphinX^ (Read error: 104 (Connection reset by peer))
16:59:46 Quit GodEater ("CGI:IRC")
17:01:52x1jmpWhy are some voice strings just empty? Is it intended not to have every message spoken?
17:02:16 Join raphi [0] (
17:02:53Du4LiTyso you guys just drag and drop you music into the ipod hdd?
17:03:01Du4LiTyor any player w.e.
17:04:38LloreanThat's one of the big advantages of Rockbox over many players' native firmware, and a reason for a lot of people changing to it.
17:04:53Du4LiTyi see
17:05:21Du4LiTyhow would you add album art, use an ID3 tag changer or osemthing?
17:05:34LloreanRockbox doesn't support album art.
17:05:43LloreanAnd the patched versions that do, don't use it in ID3 tags.
17:05:49B4gderDu4LiTy: use whatever tool you like for id3 tags
17:06:16Du4LiTyI saw some themes that showed album art, haha
17:06:28Du4LiTyi don't care much baout it, my friend just uses album art, i'm not so big about it
17:06:31LloreanThey require modified versions of Rockbox
17:06:38 Quit perplexity (Read error: 113 (No route to host))
17:07:19Du4LiTyi see
17:11:31 Quit B4gder ("Time to say moo")
17:11:55raphihey dudes, when will there be an fast/stable player for the iaudio X5? is there much priority assigned at this atm??
17:12:41Kasperleare there reasons other than "nobody wrote it" for rockbox not supporting album art embedded in ID3 tags?
17:12:42 Join perplexity [0] (
17:12:53MikachuKasperle: yes, they can be bmp, jpg or png
17:12:58preglowKasperle: i believe there's a patch for it, but also having them in tags is a waste
17:13:19Mikachuthe spec says the tag can refer to tags in other mp3 files, but i don't think anyone ever has implemented that :)
17:13:21*preglow now has better working ipod rtc alarm
17:13:34Kasperlepreglow: <3
17:13:44preglowbah, if they're going to depend on each other anyway, why not just keep it as an external file?
17:14:07LloreanKasperle: Having them in-tag means you have multiple copies of the album art, one per song, for each album.
17:14:21Mikachui don't know, maybe they were high when they made the spec. that would explain the bright coloured fish too
17:14:24Kasperlejust wondering because i tagged a lot of mp3s to include album art (just used mp3tag in windows, which does that quite nicely)
17:14:55Lloreanraphi: What do you mean by "fast/stable"?
17:14:57Mikachui found an album once where 33% of the total size was the cover picture, each mp3 file had a 2MB png in it
17:14:58 Quit Mouser_X (Nick collision from services.)
17:15:04Lloreanraphi: General consensus seems to be that Rockbox works fairly well on the X5
17:15:07 Join Mouser_X [0] (
17:15:15KasperleMikachu: weee
17:16:02raphiLlorean: I mean fast and some more videocodecs....
17:16:12preglowit is fast
17:16:23preglowif you want more video codecs, find someone to implement them
17:16:28Lloreanraphi: Ah, you didn't say anything in your message about "video" at all.
17:16:46raphilLorean: when i convert a movie to a mpg it is so slow...
17:17:06LloreanjhMikeS: I honestly can't interpret your most recent forum post at all. Any hope of you using proper English like we ask everyone else to? ;)
17:17:17linuxstbraphi: It's simply that the CPU in the X5 can't handle it.
17:17:21raphioppps, forgat that :-D the problem is also the a/V sync...
17:17:24markunraphi: videos will always have to be converted, even if we have support for other video codecs
17:17:25Lloreanraphi: The X5 itself is quite slow.
17:17:56jhMikeSLlorean: ;) :\ :0
17:18:05raphibut as i've seen on the box that it can even handle xvid...
17:18:12raphiI'm a little bit confused..
17:18:12linuxstbraphi: When you encode your .mpg file, you need to reduce the fps.
17:18:26raphiwhat about 25fps? too much??
17:18:26linuxstbxvid at a very low framerate.
17:19:16markunraphi: anywy, it's very hard for us to do, don't expect any miracles
17:19:25 Join Rondom [0] (
17:19:31raphiwhat format do you recommend? i think mpg, but what bitrate, soundformat etc...?
17:19:34Lloreanraphi: I believe the Xvid that the native firmware supports is something like 10fps?
17:19:44linuxstbAccording to the PluginMpegplayer page, the X5 can decode full-screen video at an average of 18fps, and wide-screen (160x96) video at 25fps. But those are just averages - during periods of high motion, the framerate will drop.
17:20:07linuxstbAlthough with more optimisation, those speeds should be increasable.
17:20:08raphito be honest... I haven't really tried the original firmware :-D I preferred yours!!
17:21:03linuxstbraphi: But back to your original question, mpegplayer isn't a priority for anyone at the moment.
17:21:13***Saving seen data "./dancer.seen"
17:21:42dmdfanOut of curiosity, what would it take from Rockbox or any other mp3-player to play let's say 160x120 video on a 320x240 lcd or vice versa?
17:21:51jhMikeSx5 joystick battery usb *grumble*
17:22:30markundmdfan: it would take a scaler?
17:22:42 Join jaebird [0] (
17:22:44XavierGralthough I am bit too late on that: I prefer the name Rockbox Menu for the Root menu
17:22:55dmdfanmarkun: I thought something like that
17:23:07preglowlinuxstb: did you narrow down the crossfade problem to jhMikeS' commit?
17:23:16raphiOk, I'll try a bit with the different settings.... Maybe I'll then tell what is working fine!!
17:23:28XavierGrnow add a (TM) to it and patent it too, to be trendy :D
17:24:31raphito the devs: really great project!! rules more than the original firmware!! :-D
17:24:43 Join |Rincewind| [0] (
17:25:02linuxstbpreglow: No, I didn't try any other builds. pearldiver seemed to be able to recreate it easier than me, but I don't know how far he got.
17:25:03jhMikeSpreglow: now what is it? It's sounding like I need a standard checklist to cover all the combinations of settings.
17:25:38linuxstbjhMikeS: crossfade with FLAC files seems to glitch after the 23rd February...
17:26:04jhMikeSodd, I tested with flac. Any particular conditions/settings?
17:26:08linuxstbNot consistently, and I could only recreate it one.
17:26:15linuxstbpearldiver reported it.
17:26:34linuxstbMy settings were 3s for everything, and crossfade mode set to "crossfade".
17:26:40markunaka perl|work
17:26:49 Quit perplexity ("* the time for going has come *")
17:27:05 Join CriamosAndy [0] (
17:27:14linuxstbjhMikeS: I guess it could be an ARM-only issue - I tested on an ipod and I think he has a gigabeat.
17:28:11linuxstbdmdfan: Rockbox is struggling already to play videos, so wasting CPU cycles on scaling doesn't seem a good idea.
17:28:33jhMikeScrossfade? None of the dsp stuff should be crossfade related. (Kept reading it as "crossfeed"). Will do some probing though.
17:29:16linuxstbYes, crossfade.
17:29:19markunmaybe we should give crossfeed a different name because it's causing too much confusion :)
17:29:28preglowwe probably should
17:29:39preglowi don't think crossfeed is a very helpful name
17:29:51LloreanAre there other suitable names, or would we just need to make something up that's more descriptive?
17:30:04dmdfanlinuxstb: I understand, maybe in 5 years :)
17:30:05preglowwell, what it does is simulate listening over stereo speakers
17:30:09jhMikeSit is probably only confusing to your's truely
17:30:15preglowwhich doesn't make for a short name
17:30:28linuxstbSomething like 3D Rocksound Superwow
17:30:35*jhMikeS though crossfeed was the audio term for it anyway
17:30:35LloreanjhMikeS: Well it's very easy to read one as the other.
17:31:21linuxstbI do always look for crossfade in the sound settings menu though...
17:31:32LloreanI think some of the settings seem to be in strange places.
17:31:49 Quit bonbonthejon (Remote closed the connection)
17:31:54LloreanI always look for Replaygain in Sound Settings, because it and the EQ live in the same place in my mind.
17:32:05jhMikeScrossfade has another ai term though that I can't seem to recall right now
17:32:10LloreanEven though I understand that Replaygain is definitely 'playback' oriented.
17:32:17markunpreglow: I downloaded the HRTF database and scilab today, but didn't have much luck with converting the matlab files
17:32:19linuxstbAfter the root menu is committed, a rearrangement of the settings hierarchy could be nice.
17:33:23preglowmarkun: tried octave?
17:34:25perl|workjhMikeS hey, im with the crossfade and FLAC issue
17:34:39*Llorean wonders what logic puts "Beep Volume" in the playback menu.
17:35:03Mikachui don't see how replaygain is playback related
17:35:05perl|workgigabeat target
17:35:10Mikachuit only affects the volume of the music
17:35:18jhMikeSperl|work: give me a rundown on anything that makes it consistent. It showed up on the 23rd of Feb
17:35:36 Quit illriginal ("Leaving")
17:36:07perl|workyes, 0222 doesnt have it
17:36:13FebsIf volume is a "sound setting," then it seems to me that Replaygain should also be a "sound setting."
17:36:19perl|workdaily 0223 doesnt have it as well
17:36:24LloreanMikachu: I can see it as being 'playback' oriented since it depends on music being played and tags in files.
17:36:29markunpreglow: no, I never heard of it until now
17:36:30preglowi think any setting that directly affects the sound is a sound setting
17:36:30linuxstbperl|work: But daily 0224 does?
17:36:31perl|workhome built 0224 is where it starts
17:36:31LloreanMikachu: But I don't think it's a Playback option
17:36:33preglowlike replaygain
17:37:05perl|workno EQ, no replaygain, just crossfade and FLACs
17:37:07linuxstbperl|work: Which svn revision is your 0224 build?
17:37:21Mikachui moved it to the sound menu in my tree i think
17:37:34perl|worklinuxstb cant check now, at work
17:37:43linuxstbYou don't have your player?
17:37:57jhMikeSperl|work: only between two FLAC files
17:38:09FebsNow that there is a separate "Settings" menu, does it make sense to have a "General settings" menu?
17:38:11perl|workjhMikeS all FLAC files, randomly
17:38:26perl|workcurruption "scratchy" sound, half a second
17:38:36perl|workat the start of the next track usually
17:38:56jhMikeSonly thing I did on the 23rd was fix an overflow issue on the spc codec echo
17:39:02perl|worklinuxstbi have 0222 build on it
17:39:54LloreanFebs: I think most of us can agree, a significant reorganization could benefit everything under "Settings" now.
17:40:26LloreanProbably more categories than we had before, now that it's a nested menu.
17:40:35|Rincewind|Febs, I made a comment on the root menu with a suggestion for a way to get rid of general settings. It just has to wait until root menu is in
17:40:52|Rincewind|*on the root menu flyspray entry, i mean
17:40:52perl|workjhMikeS 0223 was OK
17:41:00linuxstbLlorean: Maybe introduce an "Advanced settings" menu...
17:41:30Lloreanlinuxstb: I think I'd still rather have all "Sound Settings" in one place, rather than some be classified as "Advanced" and hidden somewhere else.
17:42:12linuxstbProbably. I agree splitting things will just be annoying.
17:42:21|Rincewind|I thought it was agreed on that anything with "advanced" menues is not the way it should be done
17:42:24jhMikeSok, so some time on the 24th ... probably 17:06 huh?
17:42:36 Quit Criamos (Read error: 110 (Connection timed out))
17:42:46Lloreanlinuxstb: An 'advanced' menu is an option, but I imagine it's not a necessary one. I guess time and discussion will tell.
17:43:07Mikachuone way to do it is have an advanced option and suddenly some more options show up in the regular sound menu
17:43:22perl|workjhMikeS what happened around 17:06? heh
17:43:37*amiconn thinks 'playback' is the proper place for both replaygain and crossfade
17:43:45LloreanMikachu: Then it'd need to be disabled by default, so users aren't constantly asking where options are in the forums, in which case it's mostly pointless.
17:43:47perl|worki was checking log, i ignored your updates when i saw coldfire...
17:43:52preglowi think crossfade belongs in playback
17:43:57preglowbut not replaygain
17:43:59*amiconn wonders how many features in rockbox he'll prbably never use
17:44:01Lloreanamiconn: I think it is for Crossfade, but not Replaygain.
17:44:52jhMikeSthink I know what it might be... I can revert one tiny two line change with no harm to anything and see
17:44:54LloreanMentally, Replaygain is to me "track-specific EQ Precut"
17:44:55preglowamiconn: i wonder how many rockbox features i have coded that i'll never use
17:44:55|Rincewind|It would be nice to remove some entries from playback. There are about 15 in there, I think 10 items should be a max for one menu
17:45:35 Quit Mouser_X (Nick collision from services.)
17:45:41Febs|Rincewind|: just read through your suggestion on FS. I like it.
17:45:46 Join Mouser_X [0] (
17:46:14jhMikeSperl|work: or can you test a two line patch for me?
17:46:33perl|workjhMikeS just in 6 or so hours
17:46:44perl|workbut put it online please, ill test as soon as i can
17:47:05markunwhat, putting patches online?
17:47:30preglowLlorean: or album specific
17:47:36preglowLlorean: but i agree
17:47:38Lloreanpreglow: Or album specific, yes.
17:47:58preglowanyway, do anyone wanna test ipod alarm now
17:48:14perl|workmarkun to resolve FLAC playback with crossfade
17:48:24LloreanI cannot test anything until at least this evening.
17:48:51preglowfor those who do:
17:50:19 Quit |Rincewind| (Remote closed the connection)
17:50:31jhMikeSperl|work: does it so happen that it's the fade jumping in large increments?
17:51:56 Join |Rincewind| [0] (
17:53:45Mikachupreglow: it started and began playing music, when i rebooted it didn't resume, so it seems to be working
17:54:26perl|workjhMikeS its weird actually, when the fade out starts, the upcoming tracks has this corruption sound and the old one just stops right away
17:54:32perl|workhard to explain
17:54:44preglowMikachu: it should work now, yes, despite the gigantic hack
17:54:50preglowMikachu: it'll turn on again one time in 2099 :)
17:55:21preglowi've put an comment to that effect in the source, so it's all good
17:55:23Mikachuwill it start playing music at a set time even if rockbox is on?
17:55:40Mikachuwould it be possible to detect the alarm triggered from inside rockbox?
17:55:43preglowthe alarm is really supposed to shut rockbox down when you set it
17:55:47preglowbut that doesn't happen for some reason
17:55:50perl|workjhMikeS my crossfade settings: 3, 4, 3, 6, crossfade
17:55:52preglowsure, possible it is
17:56:04preglowbut it's not specified behaviour right now
17:56:20preglowif anyone with a 3g ipod would test the patch, i'd be happy
17:56:28preglowdo we have any of those around? :>
17:57:30*Mikachu remembers to revert patch to avoid conflicts later
17:57:46preglowalways a good idea
17:57:49preglowone i too seldom remember
17:57:54jhMikeSI changed something to have the dsp process the entire decoder chunk if it could but I crossfade applies the same volume level to any given chunk passed in
17:59:03pondlifeDid anyone mention ?
17:59:47Lloreanpondlife: It's been discussed in the forums some, I suppose.
18:00:04pondlifeAh, ok
18:00:05Kasperlepreglow: "wake-up alarm"?
18:00:08Kasperleis that it?
18:00:48jhMikeSbah...will double check on that anyway
18:01:03MikachuKasperle: yes
18:01:20Lloreanpondlife: Actually, wait, different patent suit.
18:01:29Kasperlei guess i'll see what it does in 5 minutes ;)
18:01:59|Rincewind|Slasheri: I saw a bootloader commit in the svn log. Are the changes useful for iriver H1x0 with flashed [rock|rom]box?
18:02:19pondlifeI thought so. This ones a bit more serious-looking.
18:02:37jhMikeSyep, I sure do get a little glitch doing 1s
18:02:51Lloreanpondlife: Actually, this one's less serious looking, since it's on a hardware device only.
18:02:56Lloreanpondlife: Less serious, for Rockbox, at least.
18:03:27perl|workjhMikeS so you figure out where it comes from?
18:03:30LloreanSince Rockbox does not include a screen, a battery, or a storage device, I think we don't fall under it. :)
18:04:00 Part Llorean
18:04:36jhMikeSperl|work: well, can't be positive but I'll switch something back and see it is stops
18:04:56perl|workjhMikeS thanks for looking into it
18:05:39Slasheri|Rincewind|: those changes are useful, but not very urgent. So it's best to wait for the next official release of the bootloader v7 and then upgrade to it
18:05:39 Quit pondlife ("disconnected has pondlife")
18:05:45Kasperleso rockbox did wake up on my 5.5g
18:06:55|Rincewind|right now I don't even have zpre3 installed (no rombox, and a build from early january), I am a little scared to update
18:07:11Slasherihmm, so you have the v6?
18:07:12 Join perplexity [0] (
18:07:30 Part raphi
18:07:31|Rincewind|i have the first bootloader that provides rockbox in flash
18:07:50|Rincewind|I want to update, but I am not certain about the steps
18:07:50Slasherioh.. really? and you are still running rockbox from flash?
18:08:16|Rincewind|I haven't updated rockbox since rombox was commited.
18:08:41|Rincewind|but now I have time to do some programming again so I want to run a recent build
18:08:42Slasheriok, but then i really suggest updating to the 7-pre3.. it contains plenty of fixes
18:08:51Slasheriand no patching is needed
18:09:13 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
18:09:21Slasherijust install the latest version of rockbox and copy that bootloader to /bootloader.iriver
18:09:54Slasherithen boot from disk, flash bootloader, .rockbox/rombox.iriver and /rockbox.iriver and reboot
18:10:00jhMikeSperl|work: that sure was the reason. the real problem is that it matters at all to the pcm buffer.
18:10:08 Join inversions [0] (
18:10:11|Rincewind|I read the wiki page and don't understand the bit about changing the boot sequence when installing a new build.
18:10:14GodEatermental note to self: when building new kernel, make sure to include drivers for the SATA bus.
18:10:51|Rincewind|it says I have to change the boot sequence everytime I install an new build. I imagine this is rather painful when programming...
18:10:59Slasheri|Rincewind|: it's easy, that menu will automatically show after first boot. And can be switched off with hold switch during boot
18:11:04Kasperlepreglow: that's pretty nice, two things come to mind, though: 1: if the ipod is on hold, it will boot the original firmware with the current bootloader, 2) is there a way to have it use the piezo to make some noise when it wakes up from rtc?
18:11:06Slasherino, you don't
18:11:47Slasherinormally there is no need to change the boot sequence. But if you do a lot of debugging, you might want to change it to boot from disk temporarely
18:11:52perl|workjhMikeS aha
18:12:05|Rincewind|that makes sense
18:12:36markunSlasheri: there are some plans to flash a rockbox bootloader to the Gigabeat. Do you think I should extend iriver_flash copy it to a new plugin?
18:12:39 Quit Rondom ("Ex-Chat")
18:12:44|Rincewind|so when I have the boot loader and rombox installed, in the future I just copy the new build to the player, Rolo and flash again?
18:12:47Slasheriand when running rockbox for longer time without upgrading/programming, it might be the best choice to run rombox (the hardest one to upgrade)
18:13:17Slasherimarkun: hmm, it depends how much that code needs changing
18:13:22|Rincewind|why is rombox hard to upgrade?
18:13:29Slasheriif it needs major changes, then new plugin would be the best
18:13:30jhMikeSperl|work: I'll undo the change and test again too see that it comes back since if that's the "fix" it should
18:14:05Slasheri|Rincewind|: it's not "hard" but you need to load rockbox explicitly from disk or upgrading it isn't possible
18:14:31|Rincewind|ok, makes sense, if I think about it. Cant upgrade running code
18:14:38amiconnSlasheri: Just rolo should do, shouldn't it?
18:14:46Slasheriamiconn: that should do it too
18:15:12amiconnThat's the recommended sequence on archos when upgrading flashed rockbox
18:15:39amiconnWith flashed rockbox.ucl rolo is recommended, with rombox.ucl it's required
18:16:27|Rincewind|Ok, I try it. Thanks Slasheri!
18:16:56Slasherigood luck :)
18:16:59 Quit datachild (
18:16:59 Quit Kasperle (
18:16:59 Quit qwm (
18:16:59 Quit preglow (
18:17:09NJoinpreglow [0] (
18:17:30NJoinKasperle [0] (
18:17:34 Quit perplexity (Remote closed the connection)
18:17:37NJoinqwm [0] (
18:18:41|Rincewind|for the first time in months my iriver sees a vanilla rockbox again...
18:18:55|Rincewind|I hope it doesn't get upset ^^
18:20:13BigBambiSlasheri: bootloader 7pre3 contains many fixes iincluding me not being able to turn on with remote :)
18:20:27jhMikeSdoes rolo detection just use filesize to detect a changed fw on disk? It only seems to pop up the screen when the size changes.
18:21:28SlasheriBigBambi: ah, you were still having that issue
18:22:05BigBambiyeah :)
18:22:17BigBambiI went back to v6 and it lets me turn on with my remote fine
18:22:41Slasheriweird because it works fine for me..
18:22:58 Quit billodo (Read error: 104 (Connection reset by peer))
18:23:13NJoindatachild [0] (
18:23:22Slasherihow far you could get with 7pre3 and remote?
18:23:34 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:23:58*jhMikeS usually takes silent responses to certain questions to mean "hey, stupid, you know you're able to figure that out yourself" :)
18:24:44BigBambiSlasheri: It said hold was on when it wasn't and turned off again
18:24:58 Join mattzz [0] (
18:25:12BigBambiSo it was spurious hold detection rather than not turning on per se
18:25:29SlasherijhMikeS: hmm, it was something like that. Or something to do with the startcluster
18:25:41SlasheriBigBambi: ah
18:26:08 Join endor [0] (
18:26:20Slasheriprobably a small sleep or yield would fix that
18:26:44BigBambiI think you said something before about the ADC not settling in time being a possibility
18:26:47jhMikeSI thinking it should compare the header checksums really to be reliable
18:27:01SlasherijhMikeS: indeed
18:27:01BigBambiDid you change the wait time in 7pre3 compared to 6
18:27:18SlasheriBigBambi: yes, i removed it
18:27:23endorwould anyone be able to direct me to something to read about disassembling firmware compiled on sigmatel chip? which program would I use for that? is there any information about this?
18:27:24|Rincewind|hm, firefox doesn't want to save bootloader.iriver either it shows garbage on screen or wants to save with the name IrriverFlashing.htm
18:27:31jhMikeSand check whenever usb is unplugged so I don't have go back to the root
18:27:41BigBambiSlasheri: aha! I suspect we have our cuplrit!
18:27:44 Quit GodEater ("CGI:IRC (Ping timeout)")
18:27:57Slasheri|Rincewind|: did you try "save as"?
18:28:31|Rincewind|yes, that gave me "IriverFlashing.htm"
18:29:13SlasheriBigBambi: hmm.. but i am sure that will be fixed for the v7 at least. Or probably i could do 7pre4 too and it upgrading to it would be optional
18:29:18|Rincewind|I'm on ubuntu linux
18:29:57Slasheri|Rincewind|: hmm.. try konqueror if you have KDE installed
18:30:05Slasherior just wget the file
18:30:07BigBambiSlasheri: I don't mind about a pre4 - I wanted to use 7 mainly for the voltage readout in bootlader USB. I would of course prefer it fixed in 7 release :)
18:30:26|Rincewind|I try wget. Just have to remember the syntax...
18:30:36SlasheriBigBambi: hehe :)
18:30:49Slasheri|Rincewind|: just give the full url to it
18:31:12|Rincewind|then it saves it in some strange folder hierachy
18:31:23endorhm, was that a stupid question? :)
18:31:25Slasherino.. just a moment, i will try
18:32:16|Rincewind|lol, the file it downloaded has this name: IriverFlashing?template=oopsaccessdenied;def=no_such_attachment;param1=viewfile;p
18:32:27Slasheriwget ';filename=bootloader.iriver'
18:32:34 Quit mattzz ("Leaving")
18:32:37Slasherithat should do it. Then:
18:32:43|Rincewind|ok, i left out the quotes
18:32:44Slasherimv 'IriverFlashing?rev=2;filename=bootloader.iriver' bootloader.iriver
18:32:59Mikachu-O bootloader.iriver
18:33:01Mikachu(not -o)
18:34:57|Rincewind|ok, got it. how can I check md5 checksum?
18:35:28 Quit tvelocity ("Αποχώρησε")
18:35:31|Rincewind|ahh, thanks, forgot the sum
18:39:17LearShould the H1x0 RTC mod really be enabled by default?
18:40:05SlasheriLear: is there any reason why not?
18:40:43Lear6 kB of code that only benefits a few users? :)
18:41:10Slasheriand we have 32 MB of ram :)
18:41:27LearYes, I know...
18:41:58Slasherii would be more worried if that feature would eat something over 30-50 kB
18:42:35Slasheriit's still very useful for those who have did the mod
18:42:46Slasheriand probably encourages more people to do it too
18:43:03bluebrotherhow are plugins handled that use the rtc? Like the clock?
18:43:29roolkubluebrother: they get a constant time
18:43:44|Rincewind|the iriver_flash plugin says "Erasing 101%" not very comforting...
18:43:59Slasheri|Rincewind|: hehe, that's just a GUI glitch :)
18:44:23Slashericode simpliness vs UI correctness
18:44:27x1jmpI just get up to 98%
18:44:52*|Rincewind| should make a screen shot and send it to TheDailyWTF
18:45:01roolkubluebrother: what would be the best way to include it in the manual?
18:46:23bluebrotherroolku: I think adding it to some appendix.
18:46:43bluebrotherif a user does the mod he is somewhat experienced ... at least we should safely assume that
18:47:19bluebrotherso the clock plugin is now build for h100 series?
18:47:24roolkubluebrother: yes
18:47:34 Join perplexity [0] (
18:48:07|Rincewind|ok, the flashing worked. Nice to have 300kb more audio buffer.
18:48:25roolkubluebrother: maybe the rtc related stuff could be included as with the other targets, but with a big red warning that a hardware mod is required?
18:48:37bluebrothersounds reasonable
18:48:57bluebrotherI thought about improving the html output ... someone gimme some time!
18:49:13roolkubluebrother: how is it done with the alarm clock mod for Archos?
18:49:17pixelmareminds me of the radio situation in the Recorder V2 manual...
18:49:22bluebrotherwith for example making the warnings more "impressive"
18:49:38bluebrotherdon't know −− never checked that.
18:49:43 Quit wLLm (Read error: 54 (Connection reset by peer))
18:49:59 Join w1ll14m [0] (
18:50:19 Join Thundercloud__ [0] (n=thunderc@
18:51:27endorhm, too bad :(
18:52:27pixelmabluebrother: there's a note at the beginning of the radio chapter for those
18:54:22 Join petur [0] (n=petur@rockbox/developer/petur)
18:54:41 Quit Thundercloud_ (Connection timed out)
18:57:36 Part endor
18:58:05 Quit OgMaciel (Read error: 110 (Connection timed out))
18:58:51 Join OgMaciel [0] (n=omaciel@ubuntu/member/gnukemist)
18:59:02 Quit Lear ("Chatzilla 0.9.77 [Firefox]")
19:00:16 Join web-taz [0] (
19:00:43 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
19:01:39saratogafor the ipods, whats the purpose of the bootloader builds?
19:01:49saratogai thought the ipodpatcher tool installed the bootloader
19:02:03 Join webtaz [0] (
19:02:23bluebrotherbut the bootloader itself need to be build
19:02:25saratoga(i'm thinking of the -Boot builds on
19:02:53saratogaso it just displays the status of the bootloader and core builds seperately?
19:02:55bluebrotherah. To make sure they build clean?
19:03:21bluebrotherif not build with the others it would be hard to find when they get broken accidentally
19:03:33bluebrotherthe same applies e.g. for the iriver bootloaders
19:03:45saratogaah ok
19:03:58bluebrotheryou usually install only a released bootloader as the iriver series can get messed up even worse than the ipods
19:04:00saratogawas thinking they were seperate builds for some reason
19:04:11saratogadidn't realize they go into the same .zip
19:04:27bluebrotherI don't think they do
19:04:44*amiconn would think the rtc mod should be disabled by default
19:04:45bluebrotheryou usually install a bootloader that got released, i.e. put on the download page
19:04:46Bagderno, the bootloaders aren't provided as zips from
19:04:59amiconnNo other mod is enabled by default in the builds
19:05:03preglowthis one is :)
19:05:16bluebrotherit's similar to the sims: they are also build regularly but you can't download them
19:05:17preglowwe've got space to burn, why bother about another set of builds just for 6kb?
19:05:36amiconnpreglow: E.g. because it has other implications that just the 6K
19:05:45preglowamiconn: which are that?
19:06:10|Rincewind|how much space does the iriver flash have btw?
19:06:28*Slasheri isn't suprised to what amiconn thinks ;)
19:06:33bluebrother4MiB iirc
19:06:40amiconnA few minutes earlier.... e.g. the clock plugin is now built for h1x0... and documenting a feature that isn't available for >99% of all H1x0s is ugly to say the least
19:07:19Slasheri|Rincewind|: 2 MiB
19:07:21bluebrotherI agree
19:07:25SlasheriH300 has 4
19:07:27amiconnIt's one more thing to confuse unexperienced users... and users up to doing the mod should also be up to building with the option enabled
19:07:32*bluebrother stands corrected
19:07:52|Rincewind|2 mb and has to hold rockbox, rombox and dircache, right?
19:07:55amiconnThere are several other mods which are all disabled by default
19:08:29Slasheri|Rincewind|: just bootloader, rockbox and rombox
19:08:32amiconnAlarm mod for recorder v1, 8MB (somewhat special case), backlight for Ondio, autorock...
19:08:45preglowamiconn: true
19:09:02|Rincewind|I thought with flashed rockbox the dircache hibernates?
19:09:11Slasheri|Rincewind|: bootloader is 64 KiB fixed, rambox and rombox has both about 1 MiB
19:09:42Slasheri|Rincewind|: yes, but it hibernates on disk
19:09:52|Rincewind|ah ok
19:09:54Slasheriflash is not big enoug to hold that
19:10:09bluebrotherbtw, as we're about restructuring some menus: what do you guys think about my last patch for organizing the plugins folder in FS #5464?
19:10:19pixelmajust make it a simple define to uncomment like it is for backlight modded Ondio? I think that's pretty handy
19:10:42 Quit webtaz (Read error: 60 (Operation timed out))
19:10:47*|Rincewind| thinks how cool it would be to have his favorite plugins load from flash
19:11:13Slasheri|Rincewind|: or just preload them to ram? :)
19:11:28|Rincewind|that would be nice, too
19:11:35amiconnSlasheri: The fact that there are plenty of resources still doesn't justify wasting them
19:11:52Slasheriamiconn: indeed, but i don't think that is a waste though
19:11:54|Rincewind|but not all of them, just the 3 or 4 that I use regularly
19:11:57amiconnBloatware is the domain of some other software company...
19:12:08preglowall others?
19:12:19amiconnSurely not
19:12:23preglowalmost all others
19:12:24Slasherithat is still a major feature to some users (and growing number of them)
19:13:03amiconnSlasheri: Sure, and I'm not saying that the feature should be removed
19:13:42amiconnIt should just be disabled for official builds, easily enabled by users experienced enough to do the mod
19:13:53pixelmabut honestly - I think people who do the mod should be able to compile... it's not that hard
19:14:19amiconnI still wonder why the 8MB builds are provided explicitly, imho they shouldn't
19:14:45amiconn...for the same reasons I just stated regarding the other mods
19:14:56 Join redbreva [0] (
19:15:50saratogadoes the current ipodpatcher work correctly with the 3G?
19:16:39Du4LiTyMeh, i think my ipod hdd is dying, keeps freezing in reguler firmware and rockbox
19:18:59preglowamiconn: i don't think the h120 manual should mention should mention the rtc
19:18:59 Quit web-taz (Read error: 110 (Connection timed out))
19:19:06preglowamiconn: but the inclusion of the clock plugin is worse to hide
19:19:15 Join webguest90 [0] (i=509c2f25@gateway/web/cgi-irc/
19:19:39 Quit webguest90 (Client Quit)
19:20:06bluebrotherso disable the mod for regular builds? I think that's the best solution
19:21:15***Saving seen data "./dancer.seen"
19:21:57saratogaalright i'm stuck, trying out the new ipodpatcher on my 3G seems to have broken my old bootloader
19:22:08 Quit kaaloo ("Leaving.")
19:22:12|Rincewind|maybe leave it in for one more day, so that all the modders can get the latest rockbox and then disable it.
19:22:19saratogausing -d to remove it and then reinstalling does not help
19:22:39pixelma|Rincewind|: and confuse all the others with that?
19:24:30|Rincewind|well, who is confused? People who don't care had a feature they didn't use and the next update it is gone again. Modders who don't want to compile can use the build from this date or use the REP, where it will be activated, I am sure
19:24:42bluebrother|Rincewind|: I don't see any use of that −− modders can enable it themselve and how do you ensure all modders get the build this day?
19:25:05|Rincewind|well, they had there chance and there are the archives
19:25:16pixelma|Rincewind|: the clock plugin won't go away when updating (unless they delete it)
19:25:27bluebrotherif the REP will enable it I don't see any reason why those guys can't wait a couple of more days until that gets updated
19:25:39|Rincewind|I know about the clock, but there is nothing that can be done about that now.
19:25:59bluebrotherthe svn builds aren't archived. And having a mod in the daily archive would be even worse IMO
19:26:08pixelmashould I start demanding an official buils for backlight modded Ondio now?
19:27:50*amiconn thinks that the progress bar of a .cue-enabled mp3 looks quite strange
19:28:16*amiconn tried .cue on his h140 first to see whether it works, and how it's supposed to work
19:29:27perl|workwhy is it strange? it shows the track marks (nicely i think)
19:29:40 Join web-taz [0] (
19:29:42bluebrotherhow does it work? Never tried it
19:29:56amiconnThe track marks are so close to each other that it doesn't tell much
19:30:22perl|workhow long is your file and how many tracks are in it?
19:30:34amiconnI ripped about the only album I have where .cue would make sense if rockbox wouldn't support gapless anyway
19:30:37pixelmaperl|work: think of smaller displays?
19:30:38amiconnIt's 36 tracks
19:30:40perl|workah h140
19:30:46perl|worki see i see
19:31:08bluebrotherh100 isn't small ... think of the archos devices :)
19:31:12pixelmahow will that look on an Archos screen
19:31:24pixelmatoo slow...
19:31:26 Join barrywardell [0] (n=barry@
19:31:40perl|workbluebrother you enable .cue support from the options, start playing an mp3 file, it looks for the .cue for it
19:31:44 Join obo [0] (
19:31:45bluebrotherthe ipod mini has a smaller display than the h100
19:32:06bluebrotherperl|work: I know that, but how is it represented? Does it get a single entry in the playlist or multiple?
19:32:26bluebrotherand how are the tags handled if it's a single entry? They "simply" change?
19:32:32amiconnperl|work: It's "enable the option, *reboot*, then use it"
19:32:56*bluebrother should get something with cue file
19:32:59perl|worksingle entry, you see the track marks on the progress bar and can skip through the tracks, each track tags are being showed
19:33:04jhMikeSpreglow: when you think low latency for playback output, what magnitude comes to mind for time? I have it fairly well worked out by now.
19:34:10preglowjhMikeS: hmm, would have to try, 100ms?
19:34:18preglowjhMikeS: how would you go about it?
19:34:36jhMikeSmultichannel with automatic envelopes. probably a thread to do the work
19:35:04preglowautomatical envelopes? eh?
19:35:16preglowi mean how would you handle the actual low-latency part?
19:35:33preglowand where does envelopes fit in?
19:36:13jhMikeSthe dsp processes the channels, sums them, and then sends them to the part between the DSP and DMA is the latency
19:36:42jhMikeS100ms should be an eternity.
19:36:46preglowso you're moving dsp to the very latest part of the chain
19:37:01amiconn.cue support doesn't work at all on archos
19:37:02jhMikeSAuto envelopes come in with fading
19:37:02preglowi expect you'll have trouble making some people accept the pcm buffer should be 32 bit :>
19:37:21jhMikeSThe upstream part won't have to can be codec depth
19:37:23amiconnNeither do the trackmarks appear, nor is it possible to skip between them
19:37:33preglowjhMikeS: well, yes, 32 bits
19:37:45jhMikeSfor the most part but not a requirement
19:37:53preglowi guess
19:37:57preglowwell, sounds like pure gold to me
19:38:31jhMikeSFour channels: codec1, codec2, voice, beep. beep doesn't need to take buffer space though
19:38:34preglowgonna see if i have time to do some arm resampler work today
19:38:53jhMikeSshould give spc a kick
19:39:16 Join Rondom [0] (
19:39:47 Quit saratoga ("CGI:IRC (EOF)")
19:40:12jhMikeSChannels should probably have some reserve buffer but most memory would be linked together on the fly to form a stream for a channel from the common pool
19:40:36 Join austriancoder [0] (n=austrian@
19:40:44amiconnbluebrother: You can't have mutiple playlist entries with .cue and keep .m3u handling compatible at the same time
19:40:52amiconn.m3u only knows file paths
19:41:19jhMikeSand the dsp/pcm_buffer will cease to be separate entities in essence
19:43:52 Join Thundercloud_ [0] (n=thunderc@
19:44:03jhMikeSI guess the pcm buffer would reduce to a simple queue with a few chunks for DMA to send out
19:44:35preglowbut yeah, 100ms should be an enternity in that regard
19:44:59preglowbut you really don't need lower latencies than that unless you're performing or something
19:45:01preglowand need quick feedback
19:45:04preglowbut lower is better :)
19:45:17preglowwould be nice to be rid of the current low latency hack
19:45:23preglowand hack it is, an awful one
19:46:08 Quit redbreva ("Chatzilla 0.9.77 [Firefox]")
19:47:01 Quit Mouser_X (Read error: 110 (Connection timed out))
19:47:46 Quit web-taz (Read error: 110 (Connection timed out))
19:49:10 Nick Everybody|away is now known as Everybody (
19:49:15jhMikeSpreglow: one thing is that to have channels pause, start and stop without interrupting others is the once it's out to DMA, DMA plays to empty. That was a pause can be done in the middle of a voice clip with no interference.
19:49:47jhMikeS.1s should be fine for that really
19:50:20preglowlinuxstb: around?
19:50:57*w1ll14m reinstalls his linux machine, it suffered a lot from some powerloss damage :(
19:51:22amiconn100s is already a noticeable delay. Probably not annoying though
19:51:35jhMikeSit comes to mind that a lot of hacks will disappear. just shove data in the channel and the channel starts
19:51:42amiconnShould be <30ms to be unnoticeable afaik
19:51:52preglowamiconn: sure, i think even 100ms will do very nice
19:52:01preglowamiconn: and in any case be a massive improvement on what we have now
19:52:19jhMikeSI'll use the lowest I presents no problem
19:52:21*amiconn wishes Nico_P were around
19:52:24w1ll14mwhats the latency we have now ?
19:52:33amiconnNot sure what to do with .cue support on archos
19:52:33jhMikeS3 seconds
19:52:51w1ll14mhmmm sure thing 100ms would be faster
19:53:20 Quit BigBambi ("Leaving")
19:53:29jhMikeSsuppose even if it's noticeable, what we have now is utterly annoying ;)
19:53:52preglowto say the least
19:54:54w1ll14mdamn ... compiling the a new kernel always takes a lot of time :|
19:55:20w1ll14mand it's annoying because i don't have a dns any more.... lolk
19:55:24amiconnThe cuesheet viewer works, but nothing else
19:55:37jhMikeSall this crossfade stuff will be basically as simple as switching the input channel for the codec and applying envelopes
19:56:33jhMikeSexact same mechanism for fade in/out
19:57:03jhMikeSand if I feed like being fancy, soft muting for channels starts and stops
19:57:43*amiconn prefers to stay fade-free
19:57:57amiconnBut yeah, there are other advantages
19:57:57preglowthen you don't need to enable fades
19:58:09preglowi don't like long fades, but short ones would be good
19:58:16preglowno clicks on pause, that way
19:58:24preglowalso, this way we can do fades when voice clips occur
19:59:04jhMikeSwon't matter...every channel should have a volume control and mute ... a mixing board really
19:59:46amiconnEvery channel meaning every single channel, or a stereo pair?
20:00:12peturjhMikeS: I experienced some interesting timings when recording: It took 30+ seconds to flush. But the cause is I filled up my disk pretty much (less then 2GB left) and it's fragmented again.
20:00:13jhMikeSJust stereo pair really
20:00:20|Rincewind|could this make some fancy mixer stuff possible? Like playing a karaoke song and using the mike to sing to it.
20:00:26 Quit Thundercloud__ (Connection reset by peer)
20:00:41jhMikeSpetur: so how did it handle the strain?
20:01:09peturjhMikeS: should try again as it was a short test and the flush was at stop
20:01:17jhMikeSbut no skips?
20:01:29peturit was very short ;)
20:01:37peturso no flushing inbetween
20:01:46peturbut I'll try later tonight
20:02:22jhMikeSJust answer! :) It still matters even in that case.
20:03:09amiconnjhMikeS: Do single channels and support arbitrary panning... ;)
20:03:11jhMikeSIt should have stopped the encoding at the correct point.
20:03:24jhMikeSamiconn: If it
20:03:29jhMikeS's wanted, sure
20:04:15jhMikeScould add a channel dsp stage to pan too
20:04:24bluebrotheramiconn: why can't m3u and cue work simultaneously? The playlist could still include the mp3 file but the playlist viewer could present the indivifual tracks
20:04:44bluebrotherit could work this way for other multi-track formats as well, like sid
20:08:33 Join Shaid2 [0] (
20:09:50jhMikeSI'd like to have every dsp stage take the same parameters so the stages can be linked in any order really.
20:10:06 Nick Everybody is now known as Everybody|away (
20:10:20jhMikeSMaybe just one struct pointer
20:11:11jhMikeSI'm thinking all this might add up to less code than what's in pcm_buffer.c now. Won't get too ahead of myself though.
20:15:39*bluebrother commented on FS #6630 once more ...
20:15:57 Quit printfXh4 (Read error: 110 (Connection timed out))
20:17:00austriancoderplease review
20:17:27markunaustriancoder: check the logs :)
20:17:46bluebrotheraustriancoder: we talked about that a bit earlier today. Sounds a bit like you want to have some binary skins ...
20:18:14austriancoderbluebrother: ah.. will check the logs.. yes binary skins
20:18:28jhMikeSpreglow: what do suggest, if soft muting is implemented, to do when the mute runs passed available data? it'll probably have to synthesize something there. finish with the last sample value?
20:18:32bluebrotherI don't see any advantages in binary skins
20:19:02bluebrotherI'd rather use a more powerful markup like the current wps does
20:19:07austriancoderbluebrother: you can abstract the way everything is drawn in rockbox.. menus link on the original sansa firmware are possible
20:19:32preglowjhMikeS: why would you want to be able to do them in any order?
20:19:46bluebrotherbut do we want that much abstraction? I like it pretty much RB looks similar on all devices
20:20:22austriancoderbluebrother: even with compiled skins.. rockbox will look the same on every target, which uses the skin
20:20:37markunaustriancoder: I think it's a good idea to separate the UI from the rest of rockbox
20:20:48bluebrotherand that would work with different screen sizes?
20:21:17pixelmabluebrother: about the root menu comment - that was also discussed here earlier - the outcome was the it couldn't work (at least without sanity check) because the user can set radio, recording etc. as start screen
20:21:21austriancoderbluebrother: i hope i get this..
20:21:46markunpixelma: and especially the WPS..
20:22:11bluebrotherpixelma: I thought about that too. But if a user sets recording as startup screen I don't see a problem
20:22:14austriancoderbluebrother: it should be possible to give relative sizes and positions
20:22:25bluebrotheronly for resuming. But I overlooked that ;-)
20:22:48jhMikeSpreglow: at the very least upsteam or downstream resampling but the design might allow it anyway and be simpler for it
20:22:57bluebrotheraustriancoder: how should that work with the current fixed fonts?
20:23:08markunaustriancoder: separating the UI could also be good for a better text-to-speech intgerface, what do you think?
20:24:06*amiconn thinks we shouldn't abstract too much
20:24:14 Quit dmdfan (Read error: 145 (Connection timed out))
20:24:22austriancoderbluebrother: fonts.. a nice subtopic... as every skin can define the used font on its own. It can use rockboxs installed fonts or may ship one with the skin
20:25:13austriancodermarkun: I also think that seperation UI from general rockbox code give a cleaner code
20:25:30austriancoderamiconn: why?
20:25:50markunamiconn: are you afraid it will become more complex or just add too much code?
20:26:18markunor performance?
20:26:54 Quit perplexity (Read error: 110 (Connection timed out))
20:27:06amiconnComplexity, code size, performance, ...
20:27:24 Quit Shaid (Read error: 110 (Connection timed out))
20:27:29bluebrotherusers will create broken themes because they oversee the limitations of the device?
20:27:43 Join Shaid [0] (
20:28:07bluebrotherwhile I agree that a certain amount of abstraction would be good I also think we shouln't abstract too much.
20:28:18bluebrotherat least, it's a music player
20:28:32|Rincewind|how will a user create a skin if the skin is a binary file?
20:28:48bluebrotherwith some magiv from the tools/ folder?
20:28:49austriancoderbluebrother: but look 2 years in the future... 600 mhz cpu, ram, wifi..
20:28:57bluebrothernooo ...
20:29:04amiconnBinary themes would be a nightmare to support in regard to the variety of architectures, lcd sizes and depths, available ram, number of displays...
20:29:05 Quit Shaid2 (Read error: 110 (Connection timed out))
20:29:06austriancoderbluebrother: no.. the skins are coded in C and compiled
20:29:26markunaustriancoder: yes, I'm also a bit worried that the nice and simple DAPs we now have will not be around forever..
20:29:31bluebrotheraustriancoder: remember how old the arcos devices are ... and still supported and working with Rockbox
20:29:36austriancoderamiconn: if it is designed in a good way.. than i see no problem
20:30:14bluebrotherbut that is basically moving out the GUI into a plugin
20:30:32austriancoderbluebrother: yep..
20:30:41 Join Nico_P [0] (
20:30:42bluebrotherwhich requires compiling, keeping track of the ABI, and adjusting for each device you want to use it for
20:30:58austriancoderbluebrother: we need to yet also
20:31:22bluebrotherjust imagine a monochrome theme on a color unit ...
20:31:24pixelmawasn't it agreed on that core functionality should not become a plugin?
20:31:38bluebrotherwe need to do for the plugins, but that is not core functionality
20:31:39austriancoderbluebrother: if the user wants it
20:32:27 Join dmdfan [0] (
20:32:32austriancoderbluebrother: for this.. the current used "skin" will be the default skin for rockbox
20:32:48pixelmaalso: wouldn't that mean that you couldn't use other plugins with this
20:33:00austriancoderbluebrother: so there is no change in the UI... but if the user installs a new Skin on its own.. he could get a better looking UI
20:33:07*jhMikeS thinks separating base functionality from display of the information/interactivity with user would simplify things...example: recording screen...probably others.
20:33:35bluebrotherI still think this is too complicated
20:33:47 Quit donvito ("Anacna "It's better to be rich and healthy than poor and sick"")
20:33:47jhMikeSas to full snap in binary skins...hrm
20:34:06bluebrotherat least, it's not KISS anymore
20:34:06|Rincewind|should the skin handle buttons as well?
20:34:14markunbluebrother: I think for the grayscale targets you would compile rockbox with the old UI, not the skinnable one
20:34:14 Quit linuxstb (Remote closed the connection)
20:34:14 Join kaaloo [0] (
20:34:22jhMikeSSimple code is usually the most flexible
20:34:42bluebrothercustomizable buttons? That would be even more hell to support
20:34:45 Join linuxstb [0] (i=5343d4aa@rockbox/developer/linuxstb)
20:34:51austriancoderbluebrother: yep.. adding an #define , for instance
20:34:54linuxstbpreglow: pong.
20:35:08austriancoderno buttons.. only the calls to lcd_xxx are replaced
20:35:10preglowlinuxstb: yo, keen on trying out the new alarm patch?
20:35:19preglowlinuxstb: if i get successful reports on a couple more ipods, i'll just commit
20:35:23preglowso long i've just got nano
20:35:40|Rincewind|i'm just asking how far the skin goes, if it provides more then just visuals for the standard screens, then the skin doesn't need buttons. but if the skin can define some fancy screens as well...
20:35:46linuxstbpreglow: Sure.
20:36:09preglowlinuxstb: the only ill effects this time is that your ipod might wake up some day in 2099
20:36:17|Rincewind|argh, my logic is skrewed again
20:37:07austriancoderok guys.. so i think you need a proof-of-concept to get convinced - or?
20:37:33bluebrotherdon't guess that will convince me ...
20:37:39bluebrotherbut feel free to try ;-)
20:39:50markunaustriancoder: perhaps it would be a good start to split the UI and keep the old interface
20:39:50markunif it keep rockbox nice and simple you can start on your own interface
20:41:05jhMikeSthe idea of switching the hw samplerate is a bit bothersome. On coldfire output FIFOs are supposed to be in a reset state before touching the IISCONFIG clock bits. This hasn't been the case though since I arrived on the scene.
20:42:13austriancodermarkun: yeah.. i will do a a simple proof-of-concept on one of the screens
20:42:50markunjhMikeS: and what about changing the samplerate because a song is changing but some of the other channels are also playing..
20:43:18preglowjhMikeS: why's that a problem?
20:43:40 Join entheh [0] (
20:44:29jhMikeSmarkun: exactly
20:44:35amiconnNico_P: I tested .cue support on archos... and apart from the cuesheet viewer it doesn't work at all
20:44:38jhMikeSpreglow: clicks
20:45:06preglowjhMikeS: clicks when you change sample rate? so what? you shouldn't be playing anything while you do it anyway
20:45:21preglowchanging sample rate is never supposed to be a glitch free happening
20:46:05Nico_Pamiconn: how so ?
20:46:06markunjhMikeS: and can the channels have different samplerates?
20:46:07jhMikeSpreglow: that's why I hate it. it puts so many restrictions on things. late dsp would help though
20:46:15Nico_Pit crashes ?
20:46:20amiconnNico_P: No trackmarks and no skipping possible
20:46:32Nico_Pamiconn: have you enabled it in the settings ?
20:46:48amiconnI tried it on H140 first, to check how it should look
20:46:49jhMikeSmarkun: they'd be filled with data at whatever the codec gives it
20:46:52preglowjhMikeS: how would late dsp help?
20:47:11amiconnThe track marks made the progress bar look weird in my test case
20:47:12jhMikeSpreglow: because we apply the needed resampling just before sending to DMA
20:47:18Nico_Pamiconn: i'll try on a sim... what archos model do you have ?
20:47:37Nico_Pamiconn: wierd how ?
20:47:38amiconnBut I doubt that you can test on the sim - hwcodec playback isn't simulated
20:48:12amiconnNico_P: 36 non-equally spaced marks on a 160 pixel wide bar do look weird...
20:48:12markunjhMikeS: it's maybe not ideal, but what about setting the target samplerate manually?
20:48:55Nico_Pamiconn: i can imagine that :p
20:49:01jhMikeSmarkun: you mean as in a user option? what I'd hope is that DSP is applied on the lowest load side of things.
20:49:05amiconnNico_P: The hwcodec playback engine is separate from the swcodec one. It's not even in apps/
20:49:25linuxstbIt just sounds like the cuefile is never read...
20:49:46Nico_Pyes that's probably the problem
20:49:49amiconnBut I don't really understand hwo and where that's supposed to happen
20:50:36linuxstbIt will be in the get_metadata() function - this is the (swcodec only!) function which reads the metadata from a file before it's loaded into the audio buffer.
20:50:41Nico_Pamiconn: in principle, it happens at the same time as the loading of the track
20:50:47jhMikeSa good thing is that this should work with current playback and any revision of it and is something I'd like in place before changing all that stuff.
20:51:00linuxstbIt's in apps/metadata.c
20:51:39amiconnmetadata.o is now present in the archos builds, but afaik it isn't supposed to be
20:51:50amiconnNot sure atm
20:51:51 Join mashon [0] (i=04e0ea5a@gateway/web/cgi-irc/
20:52:23jhMikeSI think I might need a new type kernel of kernel object to control the thread...we don't have events with a wake counter.
20:52:51 Join barrywardell_ [0] (
20:53:22linuxstbMaybe it's used by tagcache...
20:53:53 Part kaaloo
20:54:46Nico_Pso where is the HWCODEC playback code ?
20:55:26linuxstbfirmware/mpeg.c etc
20:55:44Nico_Pok, i'll have a quick look
20:55:53mashonperhaps you can all help.. Im new to rockbox but own a nano 2. Why wont the nano 2 run rockbox? Is it just undeveloped or are there some sorts of nasty binary signing going on? In other words: can we expect sometime to have 2nd gen or not?
20:56:31 Join Thundercloud__ [0] (n=thunderc@
20:56:43linuxstbThe 2nd gen Nano is completely different internally to the other ipods.
20:57:09linuxstbNico_P: The problem is that firmware code can't call apps/ code...
20:57:40preglowmy favourite issue
20:57:41Nico_Plinuxstb: hmm that's a problem indeed
20:57:56preglowi had to install a plugin from firmware/sound.c to apps/dsp.c
20:58:13jhMikeSpreglow: one of my top 10 at least...not that I have a list written down
20:58:29linuxstbNico_P: I think you have two options - move the cuesheet code to firmware/ or unify the playback engines...
20:58:52amiconnlinuxstb: Some other features already use callbacks afaik
20:58:55Nico_Plinuxstb: i'll leave the second to amiconn if he doesn't mind
20:59:17linuxstbamiconn: Thanks for a sane third option...
20:59:50Nico_Pi'll have a look
20:59:59amiconnlinuxstb: I think playback engine unification will probably be your most desired move once you start working on the AV300...
21:00:55preglowamiconn: could you test the alarm patch on your mini?
21:01:08 Join Lear [0] (i=chatzill@rockbox/developer/lear)
21:01:09preglowif linuxstb gets it working on his 4g and you on your mini, i'll just commit it
21:04:43 Quit mashon ("CGI:IRC (EOF)")
21:05:05 Quit barrywardell (Read error: 113 (No route to host))
21:05:59linuxstbamiconn: I was hoping someone else would have done it by then...
21:06:48 Join Vyrus001 [0] (
21:07:08amiconnBah, this checkout is literally taking ages
21:08:40*amiconn is now building rockbox for recorder from just before .cue support
21:08:50amiconnI want to see the .map ...
21:10:21amiconnlinuxstb: I don't think that's a one-man job...
21:10:35amiconnI already suggested it for devcon
21:10:45amiconnBtw, any news regarding devcon?
21:11:21linuxstbYou can count me in to work on it, but I doubt I'll make devcon.
21:12:09linuxstbpreglow: Hmm, my 5g woke up, but my Color didn't. Let me test the Color again...
21:12:55preglowlinuxstb: oh, hmm, i'll include a couple more things to the patch, then
21:15:25preglowlinuxstb: there's a new one there now if it still didn't work out
21:15:56preglowthis time i explicitely set wake on rtc and unmask the interrupt, not that it should matter
21:17:38linuxstbOK, compiling now. Second test didn't work either.
21:18:35Kasperlewas the bootloader changed to reboot into apple os when the ipod is on hold, btw? or did i just not notice that earlier?
21:19:04 Quit Timm3h (Read error: 110 (Connection timed out))
21:20:05Nico_Pamiconn: anything i should know about your pre-cue investigation ?
21:20:07preglowlinuxstb: btw, do you have a 4g flash image lying around?
21:20:21 Quit Thundercloud_ (Connection timed out)
21:20:27amiconnNico_P: I just wanted to know whether metadata.o is included there. It is
21:20:31 Join Shaid2 [0] (
21:20:34Nico_Pah ok
21:21:04amiconnBut I'm fairly sure metadata reading is done in a different place on hwcodec, but I don't remember where atm....
21:21:11linuxstbpreglow: Yes, I've a nice collection of flash images now... Do you want them all?
21:21:17***Saving seen data "./dancer.seen"
21:21:18preglowlinuxstb: sure, that would be extra col
21:21:21preglowcool as well
21:21:52linuxstbThe alarm just woke my Color...
21:22:03linuxstb(your second patch)
21:22:19preglowdid it now
21:22:37preglowi'll have a peek at your flash image in a sec to find out why
21:22:52preglowit was almost certainly the wake on rtc bit toggle
21:23:00*dmdfan thinks if anybody wants to test one particular .cue + .mp3 on their H1x0/H3x0 series
21:23:08 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
21:26:22 Join decayedcell [0] (
21:27:06 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
21:27:23 Quit amiconn (Nick collision from services.)
21:27:23 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
21:27:34jhMikeSpreglow: Is it some big deal switching params around in functions to use a single pointer? I find for cf it would be easy to adapt them to that but maybe it's time I got an arm based player so I can to keep my changes from being aggrivating.
21:28:08preglowi don't really mind
21:28:43 Join Mouser_X [0] (
21:29:11jhMikeSGigabeat is ARM based, right? I want one of the 300MHz beasties anyway :)
21:29:17preglowit's arm based, yes
21:29:26preglowand i don't know if i want you to have a target that powerful :P
21:30:22*amiconn thinks jhMikeS should get a 3g
21:31:04amiconnPerhaps we would see some things go realtime then...
21:31:20 Quit decayedcell (Remote closed the connection)
21:31:36preglowisn't musepack already realtime?
21:31:51peturjhMikeS: recording now didn't give _that_ big flush times and the code handled it without a glitch...
21:33:05 Join decayedcell [0] (
21:33:07jhMikeSpetur: I'm glad it got some stress testing and passed :)
21:33:38jhMikeSamiconn: what's the advantage of a 3g over a Gigabeat?
21:33:46 Quit Thundercloud__ (Connection timed out)
21:33:54jhMikeSor is that irony?
21:33:58amiconnIt's slower arm... even slower than all other ipods
21:34:19amiconnNot by clock frequency but due to a broken cache implementation
21:34:55jhMikeSgreat...and here I want to listen to SPCs with all the fancy dsp
21:35:04 Nick Everybody|away is now known as Everybody (
21:35:52*dan_a would pay into the "get jhMikeS a 3g" fund
21:36:36jhMikeSif it's free I certainly won't refuse it ;)
21:38:40dan_aI'll keep an eye out. People's old, dying, MP3 players tend to migrate towards me, and then magically start working properly
21:39:02 Quit Shaid (Read error: 110 (Connection timed out))
21:39:09 Join Shaid [0] (
21:39:33*amiconn somehow prefers older mp3 players over newer ones
21:39:43LinusNamiconn: it's an age thing .-)
21:40:50*amiconn looks at LinusN and has strong doubts that this could be the reason
21:41:15 Quit Shaid2 (Read error: 110 (Connection timed out))
21:41:25amiconnage(LinusN) > age(amiconn) afaik
21:42:59 Part decayedcell
21:46:09pixelmamaybe someone has an idea... I'm still trying to figure out why I had a scrobbler-log all of a sudden the other day on my Ondio, where it's been said that it wouldn't work on Archos
21:46:10jhMikeSso what's a fair going price on a Gigabeat anyway?
21:46:22preglowlinuxstb: so, you conclude it works nice on 4g now, tjen?
21:46:23preglowthen too
21:46:33linuxstbYes, two tests, both worked.
21:47:23linuxstbYour second patch changed the alarm to 1min intervals - is that what you're planning to commit?
21:48:12preglownot at all
21:48:15preglowit's just for easy testing
21:48:40markunjhMikeS: about $100 for a F40
21:48:42preglowthe 4g bootloader code is quite similar to the nano code
21:48:45preglownot surprising, i guess
21:49:12linuxstb4g or color?
21:49:14 Join amigan [0] (i=dcp1990@unaffiliated/amigan)
21:49:23preglowi have a nasty habit of calling both 4g
21:49:27|Rincewind|can someone tell me how to include my own .c and .h files in a svn diff?
21:49:52linuxstbI think of the color as a 4g, but try to distinguish...
21:49:54pixelmaI wanted to find out what caused the scrobbler-log to be created. It contained the content of two folders that I listened to but IIRC I listened to some more that day. I let the feature set to on but nothing's been added since then - does anyone have an idea what I could look for?
21:50:05bluebrotherdiff -u /dev/null file.c >> diff
21:50:15bluebrotheror svn add them first
21:50:21|Rincewind|i want to use svn diff
21:50:21linuxstbYou can just do "svn add file.c" before the diff.
21:50:22bluebrotherbut I prefer the first way
21:50:38linuxstbsvn add doesn't require write access, unlike cvs.
21:50:47|Rincewind|is there a bad side effect if I svn add?
21:51:12bluebrotherdon't know ... maybe you'll need to svn remove them at some point again ;-)
21:51:46obopixelma: I don't know why it's broken on Archos units - one user reported that it had worked intermittently, but it seems to be a problem with a callback within the HWCODEC (mpeg.c) code
21:51:53linuxstbI can't imagine what kind of side-effects you're worried about.
21:51:57|Rincewind|well, I am starting with a fresh source tree quite often anyway. I give it a try
21:52:25 Quit Lear ("Chatzilla 0.9.77 [Firefox]")
21:52:31x1jmpand how to create a patch from an already patched source?
21:53:19pixelmaobo: that's why I wanted to find out what happened. I first thought it could have to do mith the fact that I watched the playlist almost at creation (or right after) but couldn't reproduce it so far
21:53:25|Rincewind|x1jmp: svn diff > patchfile.patch
21:54:12x1jmp|Rincewind|: that would include the already applied patch as well
21:54:22 Join Llorean [0] (
21:54:32markunx1jmp: you can edit the patch
21:54:42linuxstbx1jmp: Or revert the first patch.
21:54:44bluebrotheror try patch -R first
21:55:07obopixelma: I wish I had more info - without a unit to hand it's a bit hard to tell what is going on. I had a couple of people try debug builds for me - they didn't get any log entries whilst testing
21:55:41bluebrotheri.e. reverting the patch first
21:55:41pixelmaobo: the funny thing is that the revision that the log states is the same as the last commit that had to do with the scrobbler log (the callback thing) - I even installed this one again but to no avail
21:55:41markunI think it's a good idea anyway to check the patch with and editor, in case it includes some unwanted changes
21:56:58x1jmpreverting wouldn't help as I want to create a patch based on another patch/source
21:57:21*amiconn is totally confused about this cue sheet patch
21:57:42amiconnHow is the wps supposed to trigger a sub-track change, as opposed to a normal track change?
21:57:58 Part Llorean
21:58:07|Rincewind|x1jmp: you have patched your source with a few patches and want to create a patch with all of the changes?
21:58:53XavierGr*sigh* I can't really believe that I missed an F40 sold for 50 bucks last week in ebay...
21:59:02pixelmaobo: is this debug build the one in the tracker? Maybe I could play around with it for a while
21:59:04x1jmp|Rincewind|: just like a 'diff patched_source new_patched_source'
21:59:25pixelmaobo: or the patch to build it on my own
21:59:52obopixelma: I think I put it in the tracker, but I've been away for a couple of weeks, and have a memory like a sieve - one sec...
21:59:54|Rincewind|I use emacs for this, the command is ediff-directories
22:00:35*x1jmp is not familiar with emacs, just vim
22:00:58markunXavierGr: damn!
22:01:29|Rincewind|ediff-directories compares two source trees and gives you the differences between them
22:01:50linuxstbx1jmp: Personally, I would produce a patch including all changes - I don't like patches which depend on other patches. But if you want to, just diff two copies of the rockbox source - diff -ru version1/ version2/
22:02:01obopixelma: yup, FS #6639 - it only really debugs the scrobbler code - it needs something in mpeg.c instead to find out what is going wrong
22:02:52 Join printfXh4 [0] (
22:03:24x1jmplinuxstb: that's what I used once, but it's nasty to copy the whole rockbox tree all the time
22:04:01linuxstbYou only have to do it once. You just apply the patch to it, then reverse it.
22:05:49*amiconn *guesses* the cuesheet callback would need to be hooked somewhere in add_track_to_tag_list()
22:06:30x1jmpThough everytime I do an svn up the other tree must be updated, too
22:06:44pixelmaobo: so you are saying that this patch won't help you with debugging?
22:06:58 Join alienbiker99 [0] (
22:07:49|Rincewind|x1jmp: then you have to do the svn up in the other tree, too
22:08:31|Rincewind|you can write a short script to do that if you plan to have this situtation longer
22:08:40obopixelma: it make me think the error is in the Archos playback engine - I'm not that familar with that code, and I don't really know where to go from here
22:10:12pixelmawhich again leads to "playback engine unification" ;)
22:10:16*pixelma hides
22:11:49 Quit Mouser_X (Nick collision from services.)
22:11:58 Join Mouser_X [0] (
22:12:23x1jmpdid anyone have a look yet on the new dict plugin I posted on the patch tracker (6697)?
22:12:41x1jmpI'd like to know if it could replace the current dict plugin
22:14:42linuxstbI've never used the current dict plugin, so have no idea...
22:14:45 Join Llorean [0] (
22:15:43Nico_Pamiconn: to know if there has been a change of "subtrack", i compare the elapsed time and the indexes of the current track and the next
22:15:58Nico_Pit's done in update() in gwps-common.c
22:17:01Nico_Pon line 2613
22:17:13linuxstbx1jmp: I don't think the root menu part of your patch is svn material though - none of the other viewers get that special treatment.
22:17:23amiconnNo, I mean how does the wps tell the engine to change sub tracks?
22:18:52Nico_Pamiconn: i don't get what you mean
22:19:05Nico_Pwhen the users presses NEXT or PREV ?
22:19:32Nico_Pit's handled in gwps.c IIRC
22:20:29 Quit FOAD ("Lost terminal")
22:20:50x1jmplinuxstb: ATM the root menu part is just a personal preference
22:21:21amiconnThe only code in gwps.c that was added with cue sheet support seems to suppress directory changes when playing a track with cue sheet
22:21:32Soap /me is curious to see how many "Add X to Root Menu" unsupported builds erupt.
22:22:04 Join FOAD [0] (
22:22:40Nico_Pamiconn: look at line 406
22:24:35jhMikeSthat's gigabeat S would be nice to get working ... 532MHz ... hrm
22:25:28preglowmy, how overkill
22:25:45amiconnOk, so it seems most things are handled by gwps... but the cue sheet isn't loaded on archos, because archos doesn't use get_metadata() afaiu
22:25:50jhMikeSpreglow: sounds good to me. can't wait for something 1GHz
22:26:31jhMikeShas an FPU too
22:26:39preglowwould simplify my life a good deal
22:26:54amiconnMetadata loading on archos seems to happen in add_track_to_tag_list() which in turn calls mp3info() directly
22:27:21amiconnNo need to handle other things than just id3 tags here (yet)
22:28:22jhMikeSThis time a Freescale ARM based CPU ... can't get away from Freescale ... :\
22:29:50amiconnfpu? bah...
22:30:29amiconnThose high-perf thingies take away all the fun
22:30:33linuxstbA "vector FPU" iirc
22:30:59 Join Juice^ [0] (n=Juice@
22:31:00preglowyeah, one of those arm simd things
22:31:17Soappreglow: How about I send you a iPod line-in dock connector, and in linuxstb's words "answer the line-in recording question once and for all."
22:31:26preglowit would take the fun away from achieving things on limited hardware, but it also allows you to do stuff you can't on limited hardware
22:31:47preglowSoap: haha, i'm really not interested in ipod recording at all
22:32:19preglowdoes nano even do line in?
22:32:19linuxstbpreglow: It's just that no-one has tested line-in recording on the nano, so it would be nice to know if it works.
22:32:26preglowoh, right
22:32:28Soap_that's_ the question preglow.
22:32:34Soapoops, too late.
22:32:37preglowsure, i'll take a line-in docs connector...
22:32:44jhMikeSi'd just optimize until a space-time anomoly forms during playback
22:33:10hcsrendering audio FROM THE FUTURE!
22:33:16preglowalso FROM SPACE
22:33:38*preglow synthesis a sine wave with plenty of vibrato
22:33:54preglowi really, really need to learn how to write
22:33:56preglowthis is getting silly
22:33:58markunjhMikeS: toffe