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 2005-11-15

00:00:12XavierGrso to use e.g. 7 pointer chars, ok
00:00:18Bagderuse one
00:00:27amiconnWaaargh! who messed with the builds?
00:00:36]RowaN[is crossfeed not the same thing as stereo width?
00:00:40XavierGrthen the same problem applies
00:00:46BagderXavierGr: same as what?
00:01:02preglowmarkun: so i see
00:01:08XavierGrBagder: If there is only one string columns will still exceed 80.
00:01:30Bagdernot if you do what I told you
00:01:45XavierGrwait a sec bevuase I got really confused.
00:01:49BagderI don't know what more I can say
00:01:50preglowmarkun: the generic version should be enabled for everything but coldfire, apparently
00:01:53_FireFly_markun : i have maybe found some better values for High-Pass filter
00:02:09preglow_FireFly_: if you did i'm willing to bet there's now too much bass
00:02:34 Join tvelocity [0] (n=tony@
00:02:57XavierGrBagder as you can see this is the original
00:03:26XavierGrI know that I can substitute it with a *string but again IN the source the line will be more than 80 chars
00:03:28Bagderyes, now insert " " on 10 places
00:03:38_FireFly_preglow: if you want to try here are the new values:
00:03:42Bagderor rather " [newline] "
00:03:46_FireFly_#define HIGH_NEG 0xBBBBBBBCL
00:03:46_FireFly_#define HIGH_COMP 0x77777777L
00:03:56XavierGrah then I should break the " to change line?
00:04:18BagderXavierGr: for example, after each \n you stop the string with a "
00:04:25Bagderand add a new one to start the next line
00:04:31XavierGrok got it
00:05:13XavierGrI just didn't know that I could break the string just by terminating the " and start again with "
00:05:54Bagderthen you learned something
00:06:05XavierGrindeed thanks
00:06:30_FireFly_preglow: i didn't spoke about the bass :)
00:07:27 Quit Febs ("CGI:IRC (EOF)")
00:07:44]RowaN[is crossfeed different than stereo width?
00:08:48 Quit LedFloyd ()
00:09:21_FireFly_for me with the new settings the decrease of the volume is a bit less and for the dull sound also
00:09:53]RowaN[i didnt think stereo width worked on the iriver build, unless its been recently introduced
00:10:12ashridahheh. having the status bar on wiggs out that 'boxes' wps
00:11:02]RowaN[thank god for custom wps.. i just tried that boxes wps.. and then immediately reverted back =p
00:16:06*preglow sings the "too few registers" blues
00:17:57*ashridah watches preglow do the stack shuffle
00:18:03solexx_I am sorry if I am annoying you, but I am also interested in the new "crossfeed dsp effect"
00:18:10solexx_what is it?
00:18:25 Join ehntoo [0] (
00:18:29ashridahsolexx_: basically a reduced volume version of the left channel is bled into the right channel and vice versa
00:18:58ashridahit basically simulates listening via speakers
00:19:07preglowi don't suppose anyone knows how i can access static variables from asm? :>
00:19:08ashridah since that's what happens normally, and most cds are mixed with that in mind
00:19:22solexx_I will try it tomorrow
00:19:28ashridahinstrumental jazz in particular is often recorded with completely separate instruments on different channels
00:20:26solexx_Sounds interesting. But I am not sure I'll like it.
00:20:45solexx_It's fun to hear instruments only in one ear :)
00:21:16solexx_but well, I never had really good speakers...
00:23:00ashridahsolexx_: actually, it's hard to concentrate on the music. the effect won't be very noticable, just easier to listen to.
00:23:32preglowthis making all variables static really doesn't ease assembler versions of routines
00:23:42ashridahmono recordings suffer from similar issues, really. you find it harder to listen to them. if you timeshift it a tiny bit and play that in the other ear, it actually get's easier
00:24:45markunashridah: I could do it for mono as well..
00:25:13ashridahmarkun: don't bother on my account, i don't have any mono recordings
00:25:53amiconnBagder: wrong fix...
00:26:15amiconnIt's no longer called BL_IRIVER, but BL_IRIVER_H100, and some more files need fixing...
00:26:36Bagderwell, it did fix a few errors ;-)
00:26:42Bagderbut i'm working on a better one
00:27:26amiconnBah, I didn't really manage to do much today. At least the comments on my gcc bug reports are coming in quickly
00:27:52preglowamiconn: url?
00:31:10 Join muesli_- [0] (
00:33:53 Quit ender` (Read error: 113 (No route to host))
00:36:33 Part bitmastro
00:38:26 Quit _FireFly_ ("Leaving")
00:42:53Bagderthe h1x0 sims are broken
00:42:57Bagdereven after my fix
00:45:53Moos23:58<markun> preglow: ACC etc are not declared for non coldfire platforms (ipod)
00:46:26Bagderand not for simulated coldfire platforms either
00:46:39preglowhe didn't make any non-coldfire code
00:46:41preglowhe's working on it now
00:47:08markun..and on the phone :)
00:47:33Bagderyou do it over the phone? ;-P
00:47:39 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
00:47:57*Bagder hits the bed
00:47:59 Nick paugh is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
00:48:06XavierGrThe ondio doesn't have a disk right?
00:48:12markungood night
00:48:17 Part AliasCoffee ("Leaving")
00:49:59MoosXavierGr: yes flash device and MMC card
00:50:37XavierGrso my plugin that detects hd spins will not work on this target.
00:51:00Moossurely not :)
00:51:04amiconnIt will if it checks often enough
00:51:27amiconnThe mmc driver mimics the ata driver behaviour where it makes sense
00:51:33XavierGrmy code uses ata_disk_is_active function
00:51:44Moosoh yeah
00:51:54XavierGrevery time the disk begins a new spin it will log the battery results
00:52:17amiconnYes, ata_disk_is_active is implemented for the mmc driver, check ata_mmc.c
00:52:40amiconnThe mmc is considered active when the mutex is locked
00:52:54XavierGrhmm then it shouldn't be a problem.
00:53:03amiconnHowever, detecting probably doesn't make much sense on the Ondio, just write in regular intervals
00:53:24XavierGrDo all the targets have BUTTON_OFF BUTTON_ON?
00:53:41amiconnOndio only has BUTTON_OFF
00:54:04linuxstb_The ipod probably won't have either.
00:54:46XavierGramiconn, linuxstb_: Please tell me a button to replace then.
00:54:51XavierGrit is as it sounds
00:54:58XavierGron to confirm off to cancel
00:55:22 Quit actionshrimp ("a bird in the bush is worth two in your house")
00:56:19amiconnOn Ondio: cancel = BUTTON_OFF, confirm = BUTTON_RIGHT
00:56:37amiconnRight doubles as Play on the Ondio
00:56:48XavierGrok then
00:57:06amiconnSo you can say Play to confirm, Off to cancel
00:57:37amiconnSame goes for archos recorders (they have a real BUTTON_PLAY)
00:57:47XavierGrThat is convinient no need to ifdef the text then.
00:57:50linuxstb_I haven't thought about the ipod button mappings at all yet. I need to go through every plugin at some point, so don't worry about the ipod for now
00:57:53amiconnFor archos player it would be BUTTON_PLAY and BUTTON_STOP
00:58:35XavierGrabout the LCD_CHARCELLS: How many lines of text can it render?
00:58:53amiconn2 lines, 11 chars each
00:58:54***Saving seen data "./dancer.seen"
00:59:13XavierGrwith the normal text rendring functinos?
01:00:00XavierGrI mean players don't use another function for text dislay, right?
01:00:21amiconnlcd_puts() and lcd_puts_scroll() are there
01:00:33amiconnlcd_putsxy() obviously is not
01:00:56XavierGrI will drop those for this target then.
01:01:27XavierGrso my plugin can run on all targets, if I am not mistaken...
01:01:30 Join muesli- [0] (
01:09:02amiconnThe build table does look ill
01:09:57ashridahhm. is it possible to add a slight margin between the box and the text scrolling inside it for the 'boxes' wps?
01:11:27XavierGrsorry to ask this at such time; But what happened to wps, and what's all this talk about wps boxes e.t.c?
01:11:45XavierGrwhat's the new story here?
01:12:12ashridahXavierGr: recent commits added graphical wps's with bitmap graphics and whatnot.
01:12:38ashridahcheck out the 'wps' directory inside the rockbox directory, and rockbox's build process will include them.
01:12:57ashridahthere's an ipod-alike and a rockbox-original one
01:13:00XavierGrso a daily build will now include default images for WPSs
01:13:28ashridahi don't know, probably. i don't really use daily builds
01:13:52ashridahyou still need to select one first, afiak.
01:13:56XavierGrinteresting I will have a look
01:13:58ashridahafaik even
01:14:26XavierGrmajor activity these days. Good to see so many things happening.
01:15:14ashridahmm. looks like there's been some work on the h300 front
01:15:50amiconnA bit too many quick-shot commits for my taste :/
01:16:42ashridahamiconn: is it relatively easy to setup a script that builds every platform?
01:16:43XavierGralways very thoughtfull before any commit eh amiconn?
01:17:27ashridah(such as, say, the one that already does this for the daily builds)
01:17:27amiconnashridah: It should
01:17:31 Quit Sandking ()
01:18:12amiconnXavierGr: Usually I test-build for every platform and sim my changes *might* affect.
01:18:22 Quit muesli_- (Read error: 110 (Connection timed out))
01:18:30ashridahit's a pity cvs doesn't have a way to 'reject' a patch if it fails some process.
01:18:35 Join Sandking [0] (
01:18:46ashridahalthough that'd be a pain at some points
01:18:50XavierGramiconn: that is good and considerate.
01:19:17XavierGrAll must build some of the major branhes of the targets before commiting
01:19:19amiconnIn case I'm not sure I rather do one more build to make sure
01:19:32amiconnI just don't want to break things.
01:21:19preglowas long as people fixes the breakage, i don't mind
01:21:31XavierGrremind me again: Red colour on the daily builds table shows errors, while yellow show warnings?
01:21:37preglowXavierGr: yes
01:23:02amiconnSome changes (e.g. asm memcpy()) require even more caution. A subtle bug therein could cause all sorts of weird effects although the build runs ok...
01:24:18amiconnI have a test plugin for it that tests all possible combinations of source alignment, destination alignment and length that might be handled differently in the function
01:25:29 Join lokki [0] (
01:25:48 Join pinkutank [0] (n=ddd@
01:26:23lokkijust updated to the latest daily build and checked the to wps's out
01:26:43lokkii think you should definitly add a third one
01:27:23lokkiand i think it should be the default one (the one that used to be default)
01:28:05lokkibecause once you selected a wps from the menu, you can't go back to the default wps or can you?
01:28:31amiconnHmm, as you mention it... a while ago I noticed that it's not possible to reset some file-related settings (font, wps, language) to their builtin default
01:29:09lokkiyeah, but at least you can select the rockbox default font in the font menu
01:29:22amiconnOf course it's possible to have a file with the identical font, wps or language as the builtin default, but it's still not the same
01:29:36lokkii agree
01:30:14lokkipersonally i like the default wps a lot better than the bitmapped ones
01:32:01lokkiallready have a driver for you rtc xshock?
01:32:05amiconnI also don't care about bitmaps, but my wps is a bit different from the default
01:32:14 Join DJDD_ [0] (
01:32:43amiconnI'd like to see some larger fonts included...
01:33:01markunamiconn: I have some lager fonts, unicode too..
01:33:09lokkihmm, i usually just turn the icons off and use nimbus 12...
01:33:28lokkithat gives me a nice and clean look
01:33:58amiconnFor headphone and home use I'm using the default font on iriver, but for in-car I would like a larger one, maybe 20 pixels
01:34:38lokkisorry i meant xtal-14 not nimbus 12
01:34:42amiconnThe default could be a bit larger on iriver, but that's not a priority
01:34:53lokkiyes i agree
01:35:06lokkibut 20 pixels is very large
01:35:22markunlokki: XShocK said he had the his iriver displaying time and date, so I guess he made the driver.
01:35:27amiconn10x7 for H1x0 would roughly resemble the physical size of the current default 8x6 on the archos screen
01:35:29XavierGrgargantuan I could say
01:35:35preglowi think the default is fine on iriver
01:36:00XavierGrwell I could be better with a slight bigger, but anyway I use chicago 12 for my taste.
01:36:29lokkias long as we all can choose our fonts it doesn't really matter...
01:36:45XavierGr:) that's why I love Rockbox
01:36:47amiconnThe default is always used for a number of things
01:37:07lokkiyes, for debug menu...
01:37:31amiconnand vkeyboard, recording screen, numbers in the status bar, ...
01:38:28lokkii think for the vkeyboard it's just fine so you can navigate through all the symbols
01:38:45lokkistatusbar could be bigger and recording too
01:39:32 Join ironrok [0] (
01:39:42pinkutankdoes anyone have a compiled jepg plugin that will allow file switching / browsing without interrupting music?
01:40:01markunThe unicode patch give more possibilities and more problems. People might want to select a vkeyboard with hebrew for example.
01:40:04pinkutankI couldnt compile my own, I dont have the time to learn it atm
01:40:32XavierGrpinkutank: I made the patch to auto scroll but I haven'e compiled with the while playing
01:40:59lokki<markun>: when did he say that? i couldn't find it in the logs
01:41:25pinkutankso you have the plugin ready to use patched?
01:41:58pinkutankI d really like play and browse, I'm surprised no one uses it, I would but I'm not good ay programming ; never compiled anything before
01:42:00XavierGrI have the plugin that scrolls files but it will stop music.
01:42:15pinkutankstill can I have it, its better than the original
01:42:36XavierGrokay, you have a late daily build right?
01:43:21amiconnLinusN: Will your next bootloader (I guess there will be H100, H120 and H300 versions) incorporate the SDRAM init fixes?
01:43:21markunlokki: Can't find it either. Maybe I dreamed or logbot didn't record it.
01:43:46pinkutankthe eq was on development, what happened to it?
01:44:15XavierGramiconn, LinusN: If a new bootloader is made I think that we need to rethink if the file needs to be in the root directory
01:44:27XavierGrand if it is then at least make it delete proof.
01:44:33XavierGr(read only)
01:45:01markunlokki: in rockbox-20051113.txt: 21.03.54 # <XShocK> now my iriver shows time and date.. :))
01:45:05XavierGrpinkutank: where to send it?
01:45:18XavierGr(type an emai with "at")
01:46:05pinkutankgot it?
01:46:26lokki<markun>: thanks... the possibility of a clock are endless...
01:46:27XavierGryes I have the file but you must tell me where to send it.
01:46:30pinkutankI am assumin no one has compiled a non interrupting jpeg viewer
01:46:36pinkutankI pmed you the email
01:46:45preglowmarkun: seems my optimised crossfeed is faster, yes
01:46:56XavierGrpinkutank: only registered members can private chat
01:47:00pinkutankoops forgot
01:47:04markunpreglow: nice work
01:47:16amiconnXavierGr: YOu can allow privmsgs from unregistered users
01:47:25pinkutanki am registered
01:47:25markunI find it a bit difficult to tell the exact boost ratio
01:47:30amiconnIf you're registered yourself, that is
01:47:42 Join |joshn| [0] (
01:47:45preglowmarkun: i just noticed that flac boosts with crossfeed enabled
01:47:51 Part ironrok
01:47:52preglowmarkun: it doesn't with the optimised version
01:48:07pinkutankno one uses the jpeg viewer?
01:48:13pinkutanki read manga
01:48:46lokki<markun>: btw do you know what plugins there are for other rtc capable architectures (eg calendar alarm etc)
01:48:55markunno idea
01:49:17amiconnlokki: calendar and clock
01:49:32|joshn|LinusN: saw your raft of H3x00 patches today.
01:49:39 Join Mordov [0] (
01:49:43lokkithat would be awesome...
01:49:54markunalso bounce uses the clock
01:49:57|joshn|Is rockbox currently booting on the 3x0, outside of BDM?
01:50:04amiconnWake-up alarm is a feature that not only requires an rtc, the rtc must also be able to power up the unit somehow
01:50:31lokkiyeah that seems logic
01:50:35amiconnThe archos fm and v2 recorders can do that, and the v1 recorder can do it as well with a small hardware mod
01:50:45LinusN|joshn|: no
01:50:48amiconnFor these, the alarm function is built into the core
01:51:03XavierGrLinusN: what can you currently do with the H300?
01:51:13lokkido you think the hx0 could be modded in such a way to?
01:51:21XavierGrI mean what parts of rockbox are implemented?
01:51:26|joshn|I'd like to help with it, but I'm not quite ready to got that far to start now
01:51:28lokkitoo i mean
01:51:41preglowmarkun: i'll commit it once i'm certain it produces the same results
01:52:06amiconnlokki: Way harder, since there is no rtc to begin with. One would have to add the rtc, including the powerup functionality
01:52:31preglowmarkun: btw, i'll remove the crossfeed enabled check outside of the function, so we at least don't waste a function call just to find out it isn't needed after all
01:52:36amiconnThe archos recorder v1 already has an rtc, just the powerup functionality needs to be added with a small mod
01:52:55markunpreglow: ok
01:53:02pinkutankbefore I sleep , once more. does anyone have a compiled jpeg viewer that will view jpegs without stoppin g playback
01:53:09markunlooking forward to your changes in the morning :)
01:53:16markungood night all
01:53:34 Quit ]RowaN[ (Read error: 110 (Connection timed out))
01:53:38lokki<amiconn> but just a rtc wouldn't be very hard, and that would allready be a lot
01:53:50LinusNXavierGr: i can write to the lcd, i can read the battery level and most of the keys
01:54:01LinusNbacklight works
01:54:56XavierGrata driver?
01:54:57amiconnLinusN: [23:34:47] <amiconn> LinusN: The H300 baclight is hooked to some kind of pwm output, so it's dimmable in hardware?
01:57:48LinusNit is, but the frequency is so low so it looks awful
01:57:57LinusNXavierGr: not yet
01:58:09preglowthe pwm frequency actually can't be adjusted?
01:58:29LinusNit is set to the highest
01:58:50preglowcan the same pin be used as gpio?
01:58:59amiconnHehe, on the archos we can make the backlight flicker or blink...
01:59:11lokkigood night...
01:59:19amiconnrecorder, taht is
01:59:20 Quit lokki ("CGI:IRC")
01:59:45LinusNpreglow: the backlight pwm can be gated with a port pin
02:02:03 Quit muesli- (Read error: 110 (Connection timed out))
02:02:18LinusNthere is a chance that we can dim the backlight on the h300 as well, but at the moment i con't care at all
02:03:10preglowno reason you should
02:05:29 Join JotaCE [0] (i=JotaCE@
02:07:18 Quit Moos ("Glory to Rockbox")
02:08:19JotaCEanyebody can give me the ArchosUnlock.exe program ?? please send to :D
02:08:48pinkutankstill no one carrying a noninterrupitng lpeg plug in i presume?
02:09:43preglowsomeone would tell you if they did
02:10:59pinkutanksorry, i thought maybe someone came in recently
02:11:19 Quit pinkutank ()
02:12:19preglowanyone here intimately familiar with how crossfeed sounds that wants to test an optimised version?
02:18:35XavierGrok that's it, for today tommorow I will submit my plugin.
02:18:42XavierGrgood night everyone
02:21:30JotaCEgood night
02:22:10preglowwhere'd backlight fade times go?
02:25:43preglowright, more in cvs
02:26:16 Quit JotaCE ()
02:52:31Mordov!HELP can't get the DB to work on my ihp120... I have dowloaded the DB file but it think I do the java thing wrong...
02:53:21MordovI dont uderstand the how to page on how to run the java thingy..
02:54:17Mordov4. Run "java -jar SongDB.jar −−strip /mnt/archos /mnt/archos" from inside the .rockbox directory
02:54:23Mordovhow do I do that :)
02:58:55***Saving seen data "./dancer.seen"
03:05:59 Part LinusN
03:08:36 Quit |joshn| ("KVIrc 3.2.0 'Realia'")
03:09:48 Join Vlad0man [0] (
03:15:54 Join solexx [0] (
03:23:30Mordovhum seems to work now... don't know whay but whay know ;) this software is great, trows out stuff you don't need and ad stuf you need (like snake ;)
03:24:09Mordov-software +firmware
03:25:43 Quit Vladoman (Read error: 110 (Connection timed out))
03:28:14 Quit solexx_ (Read error: 110 (Connection timed out))
03:29:17TiMiDamiconn: you will be happy with my next commit : 2kb firmware size reduction :P
03:29:34TiMiDon the recorder
03:33:49XavierGrHiyo TiMiD!
03:35:50XavierGrhow can I add a file to my patch?
03:36:15XavierGrI type cvs diff -u > file but that will not add the new file to my patch.
03:41:53 Quit DJDD_ (Read error: 110 (Connection timed out))
03:49:22 Join RotAtoR [0] (
03:54:41 Join rubberglove [0] (
03:55:23rubberglovehi all. i have a question about a compile error that has me confused
03:56:27rubbergloveipod_fw.c:28:19: errno.h: No such file or directory
03:56:45rubbergloveipod_fw.c:29:20: unistd.h: No such file or directory
03:56:58rubberglovebut both files are there...
03:57:16TiMiDXavierGr: cvs add file, then cvs diff
03:57:34TiMiDrubberglove: you tried to compile for ipod ?
03:57:54rubbergloveno. i was just running 'make' in the tools directory
03:57:55TiMiD(because all the other targets seems to work )
03:58:20rubberglovesame with the daily build and cvs
03:58:28TiMiDare youunder linux ?
03:59:05rubberglovenope. cygwin. the 'rockbox development kit' actually.
03:59:18TiMiDhmm I can't help you then
03:59:19rubbergloveit works fine normally...
04:00:00TiMiDI just compiled it 30min go and everything was just fine
04:00:14TiMiDweell I'm gonna sleep
04:00:50XavierGrrubberglove I get the same errors on making the tools
04:01:11XavierGrjust ignore those errors
04:01:48rubbergloveah.. cool. i sure don't need the ipod firmware patcher anyway...
04:05:21XavierGrTiMiD: It seems that the command you gave me needs access to the repository so...
04:08:16 Quit rubberglove ("CGI:IRC (EOF)")
04:32:50 Join Paul_The_Nerd [0] (
04:33:46Paul_The_NerdSo, I'm having a new and interesting problem. Sometimes when I hit next, a lossless .wv will report having a total length of 0:00 and not play. (Usually if I skip 2 or 3 songs)
04:34:13Paul_The_NerdThese songs play normally, and it's not consistent to when it will trip up
04:46:49 Part Midgey34
04:58:59***Saving seen data "./dancer.seen"
05:21:05 Join Jungti1234 [0] (n=jungti12@
05:22:07DBUGEnqueued KICK Jungti1234
05:22:07*Jungti1234 님은 부재중이십니다.(go to school)
05:22:25 Quit HCl (Remote closed the connection)
05:22:27 Join HCl [0] (
05:23:09 Quit markun (Remote closed the connection)
05:23:16 Join markun [0] (
05:33:10 Quit t0mas (Read error: 110 (Connection timed out))
05:41:55 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
06:58:21 Quit RotAtoR ()
06:59:03***Saving seen data "./dancer.seen"
07:00:22 Quit tvelocity ("Leaving")
07:10:20 Quit goa (Read error: 104 (Connection reset by peer))
07:11:07 Join goa [0] (
07:31:44 Join mashalla [0] (
07:53:52 Quit mashalla ()
07:55:35XShocKmarkun, sorry. i wasn't here. I made a code that reads and writes date and time ( it is a mild change in fmradio_i2c.c code for driver, to read out a register).
07:57:30 Join ender` [0] (i=ychat@
07:58:12XShocKso it displays time, date, but I haven't had yet intergrated clock into rockbox, it is just a menu item in debug_menu that show me the time and date
08:01:52 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
08:03:30XShocKi think it will not be hard at all to add ake-up functionality to iriver, it just requires an rtc chip with alarm, and also an amplifier(i supose rtc alarm aoutput is very weak.) .I guess a resistor + transistor will do the job of amplifying, and feed that signal to play button.
08:05:18XShocKi will try to find a replacement for my alarmless rtc with one with alarm.
08:05:35XShocKi don't really need that functionality, but will do it just for the sake of it.
08:08:28Slasherii just might put a bluetooth chip inside the player so i could use cell phone for example to control it while it's inside pocket :)
08:08:55XShocKthats interesting. :))
08:09:42Slasherihehe :) and i have all necessary code ready.. jst need to find a good place for that chip
08:15:21XShocKyeah. putting chip inside might be a problem if you do it on pcboard.
08:15:53XShocKi made it on a pcboard, 1.5mm. it baery fit.
08:16:16Slasherithe chip is quite small but too thick tu put under battery
08:16:16XShocKi ll go sleep.
08:16:24Slasherigood night :)
08:17:33CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
08:17:33*Jungti1234 님이 부재중에서 돌아오셨습니다.(go to school)
08:18:44Jungti1234hi everybody
08:29:46 Join Sandking_ [0] (
08:34:44 Nick Vlad0man is now known as Vladoman (
08:35:55 Join Zagor [0] (n=bjst@pdpc/supporter/sustaining/Zagor)
08:41:20 Quit Sandking (Connection timed out)
08:45:37 Join einhirn [0] (
08:54:28 Join _FireFly_ [0] (
08:59:05***Saving seen data "./dancer.seen"
09:02:32 Quit Jungti1234 ("Bye Bye~")
09:18:19 Join ansivirus__ [0] (
09:20:04 Quit ansivirus_ (Read error: 110 (Connection timed out))
09:21:00 Quit _FireFly_ (Read error: 110 (Connection timed out))
09:33:41 Join B4gder [0] (
09:47:42preglowmarkun: man, the crossfeed volume drop is really noticable
09:48:03markunpreglow: yes, maybe time to look for some other filters..
09:48:29markunI don't know how complex the shelving filters are
09:48:43preglowdepends on how steep you want them to be
09:48:56preglowthis'll break my asm opt :P
09:49:15preglowwanna try it out to see if it sounds the same, btw? i haven't bothered to wav write or anything
09:53:40 Join Kaggen [0] (
09:53:48 Join webguest19 [0] (
09:54:22markunpreglow: I think yours sounds louder, but I might be mistaken.
09:54:27webguest19yeah ! i just saw the iriver h3xx daily bulids and only one yellow error !!!
09:54:44webguest19i hope rockbox will launch on h3xx soon !
09:54:45 Quit Kohlrabi ("Leaving")
09:57:33preglowlinus most certainly seems to do work on it these days
09:57:53preglowlets hope some other devs jump on it as well, then
09:57:59preglowmarkun: if so, then that's a bug
09:58:08preglowmarkun: but the crossfeeding in itself works ok?
09:58:11B4gderas soon as the bootloader hits the fan, I'm sure the h3x0 port will boom
09:59:14 Join webguest45 [0] (
09:59:18markunpreglow: Yes, I think there is, really sounds louder, but it still sounds ok on the 1 song I tried so far.
10:00:51 Quit webguest19 ("CGI:IRC")
10:02:46 Quit webguest45 (Client Quit)
10:05:39 Quit Rick (Read error: 104 (Connection reset by peer))
10:06:01 Join Rick [0] (
10:20:38 Join webguest01 [0] (
10:27:35 Quit webguest01 ("CGI:IRC")
10:40:32 Join _FireFly_ [0] (
10:46:55 Join sublah [0] (
10:47:59 Quit sublah (Client Quit)
10:48:49 Join linuxstb__ [0] (
10:49:10 Nick linuxstb__ is now known as linuxstb (
10:49:18linuxstbpreglow: An interesting libfaad thread:
10:49:33linuxstbSeems like the ffmpeg people are very interested in writing their own AAC decoder, using libfaad only as a way to get documentation on the format.
10:52:53preglownow there's some remarkable news
10:54:16preglowlet's just hope all their codecs are as nice as the flac one, heh
10:54:30 Join linuxstb__ [0] (
10:55:06solexxwhat are the "iriver h1x0 boot" build targets?
10:55:32linuxstb__The bootloader - the small program that is patched into the original iriver firmware and then flashed into your player.
10:56:17solexxthx. Has it been updated recently for H120?
10:56:34solexxiow: is there any need to reflash?
10:56:39linuxstb__preglow: They seem to have good intentions - i.e. writing small and efficient codecs.
10:56:56KaggenHi, I having some problems building m68k-bdm-1.30, I can't seem to get the includes correct, the best i've got is a ld-error otherwise it complains on missing header files.
10:56:59linuxstb__So I'm sure it will be significantly better than libfaad if they do it.
10:57:01KaggenAny ideas?
10:57:43preglowlinuxstb__: is that a stated intention or just what it looks llike?
10:57:46linuxstb__solexx: No, even developers only run official releases of the bootloader - i.e. the ones that Linus makes available on the Wiki page or are released as part of the fw patcher.
10:58:31solexxlinuxstb__: thanks a lot
10:58:58preglowreads the word 'seem'
10:59:00preglowbut yes
10:59:05preglowi haven't looked at any of their other codecs
10:59:08linuxstb__preglow: From reading that thread, it seems one person has already started work on documenting the format, and the other developers seem keen on the idea. So I don't know if/when anyone will work on it.
10:59:08***Saving seen data "./dancer.seen"
11:00:21linuxstb__No, I haven't really studied their other codecs either. But a RealAudio (cook) decoder was recently added to CVS. That could be interesting for people wanting to play webcasts on their Rockbox, but it's floating point.
11:02:09amiconnI do especially like the last sentence of the last comment: "I'm sure it could be implemented in less than 10% of the code that FAAD
11:02:18preglowand i believe them
11:02:30linuxstb__That's exactly what they did with libFLAC.
11:02:34preglowand looking at what a cute little decoder they made for flac... i believe them
11:03:05preglowwent from one of the slowest to our fastest codec
11:03:13preglowwith only a simple optimisation
11:03:16preglowand some iram
11:04:02 Quit linuxstb_ (Read error: 110 (Connection timed out))
11:05:03wubblathe FLAC decoder is rockbox's fastest one?
11:05:09wubblaeven faster than mpeg?
11:05:14preglowby far
11:06:27wubblaso there should be space for optimizations in the mpeg-codec, right? :D
11:07:15Kaggenis it only Linus who have been working with the BDM?
11:07:16preglowyes, but most of the obvious stuff has been done
11:07:20preglowprobably can be done better, though
11:09:39 Quit linuxstb (Read error: 110 (Connection timed out))
11:11:24preglowthere, optimised crossfeed, if anyone would care to compare its sound with the old one, please go ahead
11:12:31preglowamiconn: seems the gcc devs have taken the usual "let's pretend it's not there" route on your bug
11:13:56 Join amiconn_ [0] (
11:14:26preglowi've seen that way too often
11:16:11markunpreglow: the check in replaygain could also be moved out of the function
11:16:45preglowyes, i know
11:16:49preglowi'll deal with it later
11:19:08wubblahm... i just cannot find any mpeg-decoder in apps/codecs ...?!
11:19:15 Part Kaggen
11:19:17preglowmpa.c ?
11:19:28 Join Kaggen [0] (
11:20:08wubblapreglow: ah... sorry 'bout that :-)
11:28:07wubblahas anyone ever compared iRiver's original firmware with Rockbox's in terms of "sound-quality" and battery-life?
11:31:03 Quit amiconn (Read error: 110 (Connection timed out))
11:31:03 Nick amiconn_ is now known as amiconn (
11:31:39preglowyes and yes
11:32:02preglowin terms of sound quality there have been done a couple of simple waveform based tests done via spdif recording
11:32:12preglowfor most codecs apart from mp3, rockbox and iriver are essentially equal
11:32:20preglowmp3 differs somewhat more, but not much
11:32:26preglowas for battery life, i can't remember
11:32:30preglowroughly equal
11:33:06markunany idea what causes the difference for mp3? According to the libmad site it should be very accurate
11:33:10wubblaso, mp3 was better on rockbox?
11:36:10preglowmarkun: probably bugs and simplifications thanks to optimising
11:36:18preglowwubbla: slightly worse
11:36:48preglowi can probably have it more precise, but that would mean sacrificing performance
11:38:09preglowand i'm actually even not willing to do it before someone shows me some blind-test that proves it's audible
11:38:14preglowperformance is king
11:38:36Zagorpreglow: now that you've revealed it, the hordes will come stomping :-)
11:38:47preglowhopefully with abx tests in hand!
11:41:53 Join Lost-ash [0] (
11:42:25 Quit ashridah (Read error: 110 (Connection timed out))
11:42:56preglownow back to some ipod work
11:45:58linuxstb__preglow: What are you working on? Interrupts?
11:46:22 Nick linuxstb__ is now known as linuxstb (
11:47:07preglowi'll just see what i can do in crt0.s
11:47:23preglowi still haven't got much overview over what needs doing on the interrupt front
11:48:18preglowi've never actually set up interrupts from scratch before
11:48:24preglowjust hooked existing ones, etc
11:48:39preglowand this is arm, besides
11:48:46preglowbut, i guess ipl has what's needed
11:52:37 Join ep0ch| [0] (n=ep0ch|
11:54:13ep0ch|i don't think the mp3 output issue is audible at all
11:54:58ep0ch|but we're only accurate to about 11.5 bits
11:55:39ep0ch|i might try some different mp3s like low bitrate cbr etc and see if i can spot a differene
11:57:02ep0ch|if i were to play a 22khz file via optical out, does the rockbox resampler kick in, or does the optical out do native 22khz?
12:00:24preglow11.5 bits?
12:00:29preglowthat's got to be wrong
12:00:36preglowep0ch|: resampler kicks in
12:00:42ep0ch|well the difference was about +/-20
12:00:55ep0ch|which is about 4.5 bits
12:01:05preglowep0ch|: spdif doesn't even support 22khz
12:01:10ep0ch|preglow: ok
12:01:12preglowthen i've clearly done something wrong in optimising
12:01:24linuxstbpreglow: I'm thinking that we should perhaps not byte-swap the framebuffer, but just leave it in native format - I think this will make it easier to share code (and perhaps more importantly, bitmaps) with the H300. I'm not sure if that optimisation is worth the inconvenience it causes.
12:02:23preglowdo we byteswap the framebuffer?
12:02:36XavierGrhmm made my first battery test with my plugin. Results were dissapointed. Though Istarted the test with battery at 50 %
12:02:44linuxstbBecause the lcd_update function needs to send a big-endian word to the LCD driver.
12:03:07XavierGralso runtime results agree with the plugin.
12:03:10 Join webguest93 [0] (
12:03:12preglowlinuxstb: any reason why we just don't plain switch endianness? arm does allow us to do so
12:03:43linuxstbI think the ARM is less efficient in big-endian mode - but I've got no solid evidence to back that up.
12:03:47preglowshould be no difference if you ask me. the endianness change only affects the memory interface
12:03:56 Quit webguest93 (Client Quit)
12:04:04 Join webguest40 [0] (
12:04:04XavierGrhere is the benchmark results if anyone is interested
12:04:19preglowinternally it should be very simple to implement
12:04:59ep0ch|XavierGr: i;m interested
12:05:09ep0ch|thats just over 6hrs with 50%
12:05:27ep0ch|i think it would be nice if the log file also wrote the battery size in there :)
12:05:35XavierGryes but with a 1900mAh battery
12:05:38preglowXavierGr: what codec?
12:05:50XavierGrmp3 192kbps
12:06:02XavierGrthis test is not so good
12:06:11XavierGrI will try again
12:06:20preglowbut in my experience it takes longer for a battery to go to 50% than from 50% to 0
12:06:28XavierGrcause I let the player play and it skipped foldres and such
12:06:39XavierGrso it may have played other codecs too.
12:06:49preglowyou should always repeat a large album
12:06:49ep0ch|XavierGr: is this plugin available to others to test?
12:06:58XavierGryes of course
12:07:25XavierGrI say that on the comments too, and I was the first to break that rule, lol
12:07:39amiconnXavierGr: Do you use sleep() in your battery log thread?
12:07:56XavierGrhmm no
12:08:13preglowthen it probably eats a lot of battery
12:08:13XavierGrI just scan for disk activity
12:08:31XavierGrbut I yield every time
12:08:37XavierGror it doesn't matter?
12:08:47amiconnIt does matter
12:08:52preglowwhy do you scan for disk activity?
12:08:57amiconnIf nothing to do you should use sleep()
12:09:06XavierGrthen someone will have to say to me when to call a sleep.
12:09:07preglowright, to avoid spinning up the disk
12:09:19XavierGrI will pastebin the code.
12:09:25amiconnpreglow: There is (currently) no interface in the ata driver that actively reports activity
12:09:50 Join DangerousDan [0] (
12:09:58amiconnXavierGr: I place of rb->yield() just use rb->sleep(HZ/10)
12:10:16preglowwell, you can probably just sleep for long amounts of time
12:10:18amiconnI'm quite sure that checking every 1/10 s is enough to catch a spinup
12:10:21preglowat least a number of seconds
12:10:58preglowas is now your plugin probably uses a little amount of power itself
12:10:59XavierGr1/10 seconds that's too fast.
12:11:18XavierGrthe disk spin minimum is 3 sconds
12:11:41preglowthen hooray, sleep for a couple of seconds
12:11:53preglowyour plugin should have very little overhead in that case
12:12:00amiconnOne second should be sufficient
12:12:18amiconn...just to make sure not to miss a spinup
12:12:26XavierGrso you say to change the yields in the thread with sleeps? Ok got it
12:12:52amiconnThe point is that as long as not al threads are sleeping, the scheduler cannot use the sleep instruction of the cpu
12:13:17amiconnXavierGr: Probably not all yields, just the one in the loop that checks for disk activity
12:14:05XavierGrwell it is one yield inside this loop
12:15:37preglowsleeps wont happen anyway when you use codecs
12:16:51linuxstbWhy not?
12:17:04XavierGrI could always set the amount of sleep from the configuration file (spin disk time - 2)
12:17:31preglowlinuxstb: because the codec plugins themselves never sleep
12:18:48linuxstbDoesn't the pcmbuf_insert_split() block - meaning the codec thread is sleeping?
12:18:52 Nick Lost-ash is now known as ashridah (
12:19:11preglowno, it means it's yielding
12:19:23preglowybkess that function sleeps, of course
12:19:26preglowhaven't checked that
12:20:41preglowunless, even
12:20:58Slasheriit sleeps
12:21:09Slasheriat least it should sleep :)
12:21:26preglowSlasheri: btw, there's a bug report on the forum
12:21:35 Quit webguest40 ("CGI:IRC (EOF)")
12:21:42linuxstbI think it does sleep - see line 313 of playback.c
12:21:43Slasheripreglow: hmm, i will check later
12:21:55Slasherino 150km away from workstation :)
12:22:00linuxstbThat's the loop waiting for there to be space in the output buffer.
12:24:02 Join mashalla [0] (
12:31:08XavierGrok can someone say to me how to add a file to my patch?
12:31:43XavierGrI mean I type cvs diff -u > file but this will output all the changes to files that already exist in cvs
12:32:26_FireFly_with cvs add but for this you must have write access to cvs
12:32:50XavierGryes that was told by timid yesterday
12:33:00_FireFly_the normal diff has an option −−new-file but this isn't supported from cvs dff
12:33:02XavierGrso I can't right?
12:33:26linuxstbJust do a normal diff between two directories - instead of using "cvs diff".
12:33:28_FireFly_you can if you make a diff localy between two local copies
12:33:42XavierGrok then shoot
12:33:52XavierGrjust add a −−new-file handler?
12:34:39_FireFly_yepp it's a option of diff
12:35:06_FireFly_simply diff -u −−new-file org-dir patched-dir
12:35:20_FireFly_all new files in patched-dir will be added to the diff
12:35:25XavierGrok thanks
12:39:27_FireFly_in svn it is possible to use a external diff command to generate diffs and no network access is needed :)
12:39:48preglowgood, good! now go annoy bagder about it!
12:40:14_FireFly_preglow do you mean me ??
12:40:19preglowanyone :P
12:40:24preglowhe already knows i want svn
12:41:13ep0ch|how can i compile without remote support?
12:41:29preglowyou need to hack a header
12:41:36preglowwhy do you want to do that, btw?
12:41:39ep0ch|(i dont use the remote)
12:41:50ep0ch|and want rockbox lean and mean :)
12:42:06_FireFly_maybe we should add an function which returns if remote is connected or not
12:42:38_FireFly_and use this function to determine if the screen-access functions should call any remote-lcd functions
12:43:47preglowep0ch|: apps/screen_access.h
12:43:54preglowep0ch|: line 34, change the 2 to 1
12:43:59preglowep0ch|: untested, that should work
12:44:12preglowyou could also undefine HAVE_REMOTE_LCD, yes
12:44:32_FireFly_it should work if only HAVE_REMOTE_LCD is unset
12:44:51preglowand that's probably defined in config-h100.h
12:44:51_FireFly_because the value of NB_SCREENS depends of HAVE_REMOTE_LCD
12:45:04ep0ch|where is HAVE_REMOTE_LCD defined?
12:45:12preglowep0ch|: read a couple of lines up ;)
12:46:04 Join Febs [0] (
12:46:21amiconn_FireFly_: The remote lcd drawing (to the remote framebuffer) also needs to be done when the remote is not connected
12:46:35ep0ch|i was thinking there would be one big file in rockbox with lots of defines
12:46:57amiconnRockbox doesn't know beforehand when it will be connected
12:47:27preglowi've been wondering, currently the remote and unit is always in the same state, yes? would it be easy/feasible to make it possible for the two to be in different states?
12:47:27_FireFly_amiconn this could be always be done but no lcd-update or other lcd-calls which will change direktly somthing on the remote-lcd
12:47:34preglowfor example one in wps, the other in filetree
12:47:54_FireFly_with my wps-widget it is possible only some modification are needed
12:48:02_FireFly_in my widget
12:48:12preglowcurrently it seems the two share button handlers, that would obviously need to change
12:48:15_FireFly_but then we need two seperate gui-threads for this
12:48:28_FireFly_and button handler
12:48:55amiconn_FireFly_: If you had a look at the remote lcd driver, you would know that's what it does
12:49:10amiconnlcd_update() does immediately return when the remote is not connected
12:49:49amiconnlcd_remote_update() even
12:51:03_FireFly_but what is when the remote was plugged in on boot but after then removed ?? then is remote_initialized=true
12:52:14amiconncheck remote_tick(), line 467
12:54:05_FireFly_maybe i'm stupid but this won't be get true :
12:54:05_FireFly_if (countdown >= 0)
12:54:05_FireFly_ countdown−−;
12:54:05DBUGEnqueued KICK _FireFly_
12:54:05_FireFly_ if (countdown == 0)
12:54:07_FireFly_if (countdown >= 0)
12:54:30_FireFly_forget it
12:54:37_FireFly_i'm stupid :)
12:54:53 Join alThor [0] (
12:55:54_FireFly_but not all lcd-function check if the remote is plugged in
12:56:39 Part Kaggen
12:56:51preglowthey shouldn't
12:56:59preglowjust lcd_remote_update, afaik
12:57:21amiconnAll functions which access the hardware should (and do) that
12:57:45amiconnThe drawing functions must not
12:57:58_FireFly_what about lcd_remote_backlight_on
12:58:19markunany idea which chip the Toshiba Gigabeat X30 uses?
12:58:31amiconnThere it's not worth to check, it just sets a port bit
12:58:49amiconnIt's not exactly an lcd driver function either
12:58:49 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
12:59:10***Saving seen data "./dancer.seen"
12:59:29 Join kaggen_ [0] (
12:59:45 Quit kaggen_ (Client Quit)
12:59:46 Join kaggen_ [0] (
12:59:58 Nick kaggen_ is now known as kaggen (
13:01:01B4gdermarkun: it sure looks as if it could be a coldfire one
13:01:16_FireFly_the only bug which i had descovered is when you plug-in the remote-plug only half and then complete the remote-lcd stays blank
13:01:21B4gderso many similarities with h3x0 and x5
13:03:38 Quit mashalla (Read error: 104 (Connection reset by peer))
13:04:20amiconn_FireFly_: Yes, that's because the driver tries to init too early. I haven't found a solution for this problem yet
13:05:03amiconnHmm, maybe there is one, just can't test it now...
13:05:52CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
13:05:52*preglow has a 1500 line patch for libmusepack
13:05:53_FireFly_it would be nice if we could determine if a call to the remote is really executed on the remote
13:05:59preglowanyone remember me saying i wouldn't work on libmusepack? i sure don't
13:06:15amiconn_FireFly_: We can't. The data transfer is write only
13:06:43_FireFly_yeah but it would be nice if this would work
13:07:27_FireFly_i had tested a little bit about this and found out, that the command is simply echoed from the output to the input when no remote is connected
13:07:41_FireFly_it seams to me
13:07:45markunB4gder: couldn't find any information with google so far
13:07:53ep0ch|who knows about gui\list.c?
13:07:53B4gderme neither
13:08:03ep0ch|regarding the remote
13:08:09B4gdertheir firmware upgrade is *huge*
13:08:12 Quit kaggen ("Leaving")
13:08:16B4gder13.5MB bin file
13:08:23amiconn_FireFly_: This echo will most likely also happen when the remote is plugged
13:08:38_FireFly_ep0ch| TiMiD has coded it but just ask maybe i can help you
13:09:00ZagorB4gder: !!!
13:09:14Zagornew version of the demo tunes?
13:09:14B4gdergb2010ASfwup.bin: romfs filesystem, version 1 13911104 bytes, named ROM Disk.
13:09:38ep0ch|_FireFly_: I'm trying to compile without remote support on the ihp, so i've change the firmware/export config file to reflect this
13:09:47 Join justsomeperson [0] (
13:09:47preglowopen it!
13:09:50preglowmount it!
13:10:04B4gderand look at this:
13:10:10B4gderfile ELF 32-bit LSB executable, ARM, version 1 (ARM), for GNU/Linux 2.0.0, dynamically linked (uses shared libs), stripped
13:10:12_FireFly_the list uses the screen struct of screen-access
13:10:19ep0ch|_FireFly_: however in list.c i get compile errors because SCREEN_REMOTE is undeclared
13:10:26B4gder(also part of their update firmware package)
13:10:38_FireFly_so if you unset HAVE_REMOTE_LCD so NB_SCREENS is 1 so no remote
13:10:48preglowwho needs specs when they hand it to you on a plate anyway
13:11:07_FireFly_ep0ch| then this is a bug
13:11:24B4gderwell, we don't know which arm
13:11:54ep0ch|ok, i'll see if i can get it to work nicely
13:11:55*B4gder does the strings game
13:12:20B4gdernow that is fun
13:12:35B4gderglibc 2.2.5, gcc 2.95.3 ;-)
13:13:09B4gderfunny, but I can't find any pointers to the source code
13:13:28_FireFly_ep0ch| simple cover the defines for the remote-buttons which #ifdef HAVE_REMOTE_LCD
13:13:30Zagorhow surprising...
13:13:30_FireFly_in list.h
13:13:37preglowB4gder: a simple instruction search should verify if it's arm 5e or later, at least
13:13:53B4gderyes I guess
13:14:04B4gderI could mount the file and scan
13:14:11 Nick alThor is now known as rand_althor (
13:14:25 Join kaggen [0] (
13:14:33ep0ch|_FireFly_: is it better to do this in list.h?
13:14:48ep0ch|for the LIST_RC_* defines
13:15:02_FireFly_i mean that :)
13:15:24B4gderputting a different romfs on an x30 shouldn't be that hard then I guess ;-)
13:16:21_FireFly_[13:13] <_FireFly_> ep0ch| simple cover the defines for the remote-buttons which #ifdef HAVE_REMOTE_LCD
13:16:21_FireFly_[13:13] <_FireFly_> in list.h
13:16:35_FireFly_^^ :)
13:16:39preglowb4gder: is this a new player?
13:17:04B4gderI don't know, I've never heard of it before markun mentioned it
13:17:14Zagoryes it's new
13:17:38Zagorreleased in japan in september.
13:17:42B4gder for an insane flash-orgy site about it
13:17:47markunBut there are other Toshiba players like F20, maybe they are more or less the same
13:17:59preglowmacromedia still hasn't made a 64 bit flash plugin
13:18:01B4gderthey are available with 60GB 1.8"
13:18:42B4gder"28 user selectable equalizers"
13:18:45ZagorB4gder: yeah, it's their own disks so you'd somewhat expect that :-)
13:18:57B4gderyes, no surprise there
13:19:05preglowwhich reminds me
13:19:13preglowi was considering commiting an eq soon
13:19:19preglowwithout shelving filters
13:19:21markunMaybe they will be the first with a single platter 40 gb then too..
13:19:26Zagor"GB means one billion bytes when referring to hard drive storage capacity."
13:19:42Zagori love that
13:20:35Zagor"Toshiba announced today another Gigabeat player, with a 30GB hard disk. This player manages to be 20% smaller than it's previous generation, F Series, and features a screen with a resolution of 240 x 320."
13:21:07Zagor"will be available in Piano Black, Pure White and Crimson Red in Japan, on October 22"
13:21:11Zagorvery new, then
13:21:23B4gdertheir 40GB is as thin as their 10GB
13:21:27preglowtime to warm up some gpl zealots on the missing source
13:21:30preglowthat should make our job pretty easy
13:21:32B4gderhinting it is a single platter
13:21:44B4gderor thick 10GB version
13:22:13B4gderor possibly wrong spec sheet of course
13:22:56markunI can't find for which gigabeat this firmware is:
13:22:57B4gderat least the 60GB is claimed to be thicker, 3/4"
13:23:42B4gderI download my version here:
13:24:22B4gderI am a "HKSAR End-user" ;-)
13:26:24B4gderstrings gb2010ASfwup.bin | grep -i "free software"
13:26:24B4gder# Copyright (C) 1996-2001, 2002 Free Software Foundation, Inc.
13:28:01linuxstbB4gder: Where did you get the .bin file from ? I only got an .exe file.
13:28:14B4gderI ran that
13:28:22B4gderit extracts 4 files
13:28:38B4gderI can make them available if you wanna have a look
13:28:54linuxstbIt's OK. Wine did the job.
13:29:22B4gderI'll upload them anyway, in case someone else wants to play
13:30:03markunyes, wine worked fine
13:30:34linuxstb"strings" give a _lot_ of information. Looks like all the symbols are still there.
13:30:52linuxstblibm seems to be there.
13:32:07linuxstbYou can mount the .bin file using "mount -o loop -t romfs" to see the filesystem.
13:34:27 Join DJDD_ [0] (
13:36:49linuxstbThe firmware update itself seems to be in some .DAT files - and they look like they are encrypted.
13:38:18ep0ch|i think the settings menu needs a DSP menu
13:38:26ep0ch|its crowded in there
13:38:31linuxstbBut it looks like you could execute code by making it pretend to be a firmware updater.
13:38:59markunlinuxstb: yes, that would be a nice trick
13:39:05linuxstbThere seems to be a binary LCD driver in an unstripped .o file as well - legna-lcd-gb4.o
13:39:24linuxstbSo that should be very easy to reverse.
13:41:16preglowep0ch|: i agree
13:42:21 Quit justsomeperson ("CGI:IRC")
13:46:04linuxstbmarkun: Are you planning a Rockbox port to that new Toshiba then? :)
13:46:17markunHmm.. :)
13:46:37B4gdergo go go
13:46:42markunIt would be a nice candidate I think :)
13:47:23markunBut it's not for sale here I guess.
13:48:03markunLet's hope they don't go out of production by the time the port is finished..
13:48:13linuxstbWhy break with tradition?
13:48:57ep0ch|so basically your doing the ipod port to kill it off, good plan :)
13:49:33markunB4gder: The single platter 40gb is the F40? only shows one 40gb one
13:50:49linuxstbI'm assuming that if you write a Linux application to run on the Toshiba, then it could gain control of the whole device and start another kernel - i.e. Rockbox?
13:51:11B4gderan application can't do that
13:51:29linuxstbWhat about uclinux?
13:51:47B4gderright, in uclinux it should be possible I guess
13:52:47B4gder"Warning: Do not operate the product while walking,"
13:53:39markunLook at the traffic, not your device is maybe what they want to say
13:53:45linuxstbNo GPL/LGPL text anywhere then?
13:55:04B4gder"You may fall or cause a traffic accident,"
13:55:06B4gdernot that I've seen
13:55:08B4gderI'm reading the user's manual now
13:55:36ep0ch|how can i do a cvs up which will overwrite my local changes?
13:57:00preglowyou may fail
13:57:24B4gderep0ch|: Use the `-C' option to overwrite locally modified files with clean copies from the repository
14:02:12 Quit ashridah ("Leaving")
14:03:54linuxstbThe Toshiba HDD Audio Player uses built-in software components distributed as free software according to end-user license agreements set forth by third parties or copyright notices. For information about specific agreements on these software components, please read the attached License Information regarding Software Used on Toshiba HDD Audio Players .
14:04:11markunB4gder: Does it have a clock?
14:04:28linuxstbThat's a quote from the license document - but there is no "attached License Information"
14:04:42B4gderI don't spot any mention of a clock
14:05:01B4gdermy acroread crashed so I stopped reading that ;-)
14:05:13Zagorlinuxstb: maybe you get it in the accompanying papers when you buy it
14:06:57linuxstbI found some info at the end of the user manual PDF. It refers to the GPL and LPGL licensed components (kernel, glibc) and copies of the licenses are on the CDROM.
14:07:09ep0ch|_FireFly_: i'll put my patch in the tracker, that fixes not having remote defined, just changes two files
14:09:09linuxstbmarkun: Go shopping :)
14:13:58 Quit _FireFly_ (Read error: 110 (Connection timed out))
14:17:56ep0ch|please can someone kill patch 1357357 in sourceforge (i didnt press the tickbox to upload a file)
14:18:51preglowcan't you upload one after having submitted it?
14:19:38ep0ch|ahh didnt think of that, still i have two entries in SF
14:22:23ep0ch|heh or i could just delete it myself :)
14:25:21 Join MrShlee [0] (
14:27:28 Quit linuxstb (Read error: 104 (Connection reset by peer))
14:27:32 Join linuxstb [0] (
14:30:29 Quit DJDD_ ("Trillian (")
14:37:44 Quit MrShlee ("Changed by popular Request - "I am the heart of your darkness, you stinking oompa-loompa morons!"")
14:40:58 Quit kaggen ("Leaving")
14:41:29preglowwhat the hell
14:41:41preglowwhen i try mpc.codec from the bleeding build, i just get noise
14:42:24ep0ch|fine here
14:42:52ep0ch|(but i did my own compile)
14:43:08preglowcould you just try the one in the last bleeding edge one?
14:43:18preglowi can't get it to behave coherently here
14:43:29 Join linuxstb_ [0] (
14:44:18ep0ch|not having remote support knocks nearly a whole 10k off rockbox.iriver
14:44:39preglowi refuse to believe haven't pointed this out by now
14:46:15ep0ch|latest bleeding edge is good on mpc
14:46:24preglowso i must be doing something wrong, then
14:46:41preglowi'm just testing my latest musepack changes
14:47:21ep0ch|any chance you could put some random stuff into iram for mp3 and vorbis
14:47:38 Join webguest16 [0] (
14:48:46webguest16as i can se on the h3xx wiki page you have managed to make a firmware for h3xx :D just the bootloader to finish and then it will be realesed ? or?
14:49:01B4gder"released" ?
14:49:16B4gdernot even the h1x0 version has been released
14:49:30B4gderusable: yes, released: no
14:49:35preglowep0ch|: i'd rather put some important stuff there
14:49:37webguest16yeah but a version the we can run on the h3xx:D
14:49:45preglowep0ch|: i'll look into it some day
14:50:47webguest16just the bootloader and 1 error left until we can run it on h3xx ?
14:51:06B4gder"just" that, yes
14:51:06 Quit linuxstb (Read error: 104 (Connection reset by peer))
14:51:12 Join linuxstb__ [0] (
14:51:40webguest16haha ;) will the bootloader take long then?
14:52:19B4gderthe bootloader has always been what we're waiting for for h3x0
14:52:25B4gderits been so for many months
14:52:46webguest16yeah but you have also eliminated all the error in this months:D
14:52:55B4gderthey were never the problem
14:53:16B4gderI mean, not the big problem
14:53:59webguest16but as i can se linus can run the startup screen on the h3xx so the bootloader can be that far away now?
14:54:19B4gderI don't know
14:54:22XavierGrwell there it goes
14:54:34XavierGrBattery Benchmark plugin is now on Patchtracker
14:54:56XavierGrPlease test it if you are interested, it can become valuable at times
14:55:12preglowep0ch|: there's some good iram usage for you!
14:55:18preglowunfortunately, no one uses layer 1 files
14:57:22XavierGroh ep0ch, I remember you said that you were interested. So please if you have some time please try it, and tell me about it.
14:58:05XavierGrAlso, linuxstb_: If you are going to make some benchmarks again for any of the codecs, please use it to see credibility.
14:58:06 Quit webguest16 ("CGI:IRC")
14:58:31XavierGrCurrently I charge mine and will test first thing first.
14:59:12***Saving seen data "./dancer.seen"
14:59:17linuxstb__XavierGr: It would be useful to do some tests with and without your plugin - we don't want a battery monitor that causes Rockbox to use more power.
14:59:34XavierGrof course!
14:59:58XavierGrI will do such a test and let you know.
15:06:03 Quit linuxstb_ (Read error: 110 (Connection timed out))
15:06:22ep0ch|XavierGr: i'll look at it
15:06:23linuxstb__preglow: Is "swab32" (i.e. swap bytes 0 & 1 and bytes 2 & 3 in a 32-bit word) an easy operation to do on the ARM?
15:08:47ep0ch|XavierGr: does it write the battery size into the log file, it would be useful when comparing results
15:09:10preglowlinuxstb__: yes, with the rev instruction
15:09:13preglowi think we've got that
15:10:02preglowthe sheet i've got says it's a rev 6 instructio
15:10:40TiMiDep0ch|: if you don't want remote, compile with logf support, this disables the remote
15:11:02TiMiDor just define NB_SCREENS to 1 in screen_access.h
15:11:57TiMiDchanging it in list.h would be odd
15:11:57ep0ch|TiMiD: i've changed the export/config-h120.h and submitted a patch as it didnt compile
15:12:06amiconnlinuxstb__: I'd rather leave the optimisation in (different byte order in framebuffer)
15:12:15preglowmarkun: it looks like my asm optimised crossfeed is very efficient
15:12:31preglowmarkun: at least judging by the boost percentage
15:13:35 Quit nevs27 (Connection reset by peer)
15:13:57linuxstb__preglow: Looks like REV16 will do exactly what is needed.
15:14:40preglowlinuxstb__: it will, if we've got it
15:15:00TiMiDamiconn: nearly 2kb binary size decrease with the incoming multi-screen playlist viewer
15:15:57linuxstb__amiconn: I'm thinking that if we write an assembler optimised lcd_update, then the overhead will be minimal. But I'm still undecided.
15:16:15preglowanyone here with some mpc files that want to test a faster decoder?
15:16:22ep0ch|go on
15:16:43ep0ch|i have a benchmark of boost ratio for what it once was
15:16:43 Join DangerousDan [0] (
15:17:19preglowep0ch|: in which case i look forward to hearing the new results
15:17:22preglowit should be quite a bit faster
15:17:45 Join Sandking [0] (
15:18:06ep0ch|XavierGr: please can you clean that patch up? it references rockbox-devel or something odd like that
15:18:30amiconnpreglow: Your layer1 opt uses the wrong attr. Should be ICONST_ATTR for constant data
15:18:47preglowamiconn: ahh, sorry about that, it's an old opt i just commited now
15:19:09amiconnTiMiD: Nice :-) Is this already removing all old dupes, or are there still some of them?
15:20:10amiconnlinuxstb__: Maybe the overhead is minimal, but sure > zero. Up to now the philosopy always was to use native format internally.
15:20:17 Join Kohlrabi [0] (
15:20:33amiconnIt's not really a problem to have 2 16bit formats, imho
15:20:44preglowthink i agree there
15:21:10B4gderI'm on that train too
15:21:11preglowif there is something obvious we can do to lessen cpu load, i think we should jump at it
15:21:32linuxstb__OK, I'm convinced :)
15:22:07markunNo ogg on the gigabeat. Rockbox to the rescue!
15:22:11ep0ch|preglow: before my mpc file was at 29.5% boost
15:22:28ep0ch|preglow: now.... about 17.5%
15:23:04markunSomeone should optimize Tremor. It's probably the slowest codec right now..
15:23:08linuxstb__Can I assume that lcd_update_rect() will be called with an even value for x and width, or does the function need to check that?
15:23:12ep0ch|no aac is :p
15:23:38ep0ch|that's a lie .mod and sid is really the slowest :D
15:24:02TiMiDamiconn: there are a lot remaining
15:24:17TiMiDit didn't removed dupes :p
15:24:27preglowmarkun: it's not straightforward
15:24:39preglowep0ch|: and those are pretty high bitrate, yes?
15:24:43TiMiDI just used gui_list instead of the original home made list
15:24:49ep0ch|preglow: yeah 235
15:24:57markunI know
15:25:40preglowep0ch|: and nothing wrong?
15:25:45preglowep0ch|: if not, i'll commit it now
15:25:47ep0ch|sounds fine
15:26:26ep0ch|but i haven;t done an output analysis (yet)
15:26:36preglownothing should have changed
15:26:50ep0ch|ok i'll take your word for it
15:30:00 Quit Sandking_ (Connection timed out)
15:30:18 Join NicoFR [0] (
15:30:36ep0ch|oh hangon my ihp has frozen at the end of an mpc :s
15:31:01preglowat the very end?
15:31:12ep0ch|it says 2 seconds from the end
15:31:31ep0ch|letme find a paperclip and do more tests
15:31:48preglowand you're certain it didn't do so before?
15:31:52XavierGrep0ch: yes the battery size is mentioned in the log.
15:31:58preglowcan't see exactly how i can have triggered that
15:32:23XavierGrep0ch|: what about the patch? what it says?
15:32:36preglowep0ch|: oh, but it does indeed happen :V
15:32:59ep0ch|grrrr my reset switch has vanished
15:33:48preglownot good
15:34:06ep0ch|nothing like a good bug to fix
15:34:40preglowand at the very end
15:34:48preglowlike i said, you're sure it didn't happen before?
15:35:15ep0ch|let me try with the old codec
15:35:48preglowi'm pretty sure it didn't
15:35:59ep0ch|me too
15:36:07preglowdamn, i'm bouncing completely against the iram wall here
15:36:20XavierGrep0ch|: Try a different -px handler (it worked for -p1 for me) and let me know when you test it.
15:36:21ep0ch|shame fastforward doesnt work...
15:36:36ep0ch|XavierGr: ok
15:37:54ep0ch|XavierGr: oh yeah that seems to have applied cleanly, thanks
15:38:09amiconnTiMiD: So... when all other conversions are done and the duplcate splash(), statusbar, buttonbar can be removed we'll save even more?
15:40:16TiMiDyes that's theidea ;)
15:41:51 Quit Zagor ("Client exiting")
15:42:50ep0ch|preglow: first track played fine with the old mpc
15:44:20preglowyes, i pretty much know what's wrong
15:44:49preglowi must be overwriting some memory somewhere
15:44:58preglowthe playback layer is what's crashing
15:45:22amiconnTiMiD: Sounds good :)
15:48:10 Join yngwi [0] (
15:50:20yngwihi anyone in here?
15:50:57ep0ch|nah we've all gone down the pub
15:51:57yngwigotcha! do you know anything about building the crosscompiler with cygwin on windows?
15:52:27ep0ch|yngwi: for cygwin i followed
15:52:33markunB4gder: I think Gigabeat is not UMS..
15:54:37yngwido you know are there problems when cygwin is on my system when i install the devkit u pointed at with your link?
15:55:38linuxstb__markun: The license docs mention linux-wlan, samba and dhcp - does it have any networking?
15:55:50yngwibecause i tried the simple giude already and it didn't work
15:56:03linuxstb__yngwi: Go to the CrossCompiler wiki page.
15:56:20linuxstb__That gives instructions for building the cross-compiler from source.
15:56:27yngwithats exactly my problem
15:56:30markunlinuxstb__: wlan would be fun
15:56:54yngwii did that and id used to work on another pc but when trying at home i cannot get it to compile (the binutils)
15:57:24linuxstb__Which target are you compiling for?
15:57:30markunlinuxstb__: there is a wifi cradle for sale
15:57:57linuxstb__markun: I'm quite tempted myself...
15:58:52markunThe F41 has a nice single platter 40GB drive, but the X30 is quite a bit smaller. Difficult decision
15:59:41linuxstb__yngwi: Do you have all the tools required to build binutils? You need Perl, flex, bison, m4, and maybe others.
16:02:06yngwiyes i think
16:02:23yngwiat least i tried to follow the instructions the best i could
16:02:42yngwiand as i said i managed to build it without probs the last time i tied
16:02:56yngwiit was at the pc of my now ex-girlfriend :-/
16:03:28linuxstb__What is the error message you get?
16:05:19yngwimake[1]: *** [fibheap.o] Error 1
16:05:21yngwimake[1]: Leaving directory `/home/nix/build/binutils/libiberty'
16:05:28yngwimake: *** [all-libiberty] Error 2
16:05:34yngwithis are the last 3 lines
16:05:48linuxstb__There must be another error above that.
16:06:35 Join _FireFly_ [0] (
16:06:51Mordovcould some one help me? Mt ihp120 stops when scanning trugh the list of artist, it completly lock's up and I have to setset the player... any way to make tihe firmvare more stale?
16:07:26yngwi../../../binutils-2.16/libiberty/fibheap.c:395: warning: implicit declaration of
16:07:28yngwi function `memset'
16:07:36ep0ch|Mordov: that's with the iriver firmware or rockbox?
16:07:42yngwithis is the line above the 3 lines
16:07:54yngwisorry for bothering you with noob problems
16:08:05Mordovrockbox ofeurce... wahy else would I be here;=
16:08:31Mordovand excause my english :)
16:08:53ep0ch|have you tried a scandisk/chdsk/fsck?
16:09:08Mordovno I will do
16:09:31linuxstb__yngwi: That's just a warning - you can safely ignore it.
16:09:39linuxstb__That wouldn't cause the error.
16:09:46Mordovfirst 5 times it stopd at pixis but now it goes towo lines longer down and stop at placebo
16:10:05ep0ch|what does iriver firmware do?
16:10:15Mordovdoing just fine :)
16:10:32yngwi../../../binutils-2.16/libiberty/fibheap.c: In function `fibheap_delete_node':
16:10:34yngwi../../../binutils-2.16/libiberty/fibheap.c:285: error: `LONG_MIN' undeclared (fi
16:10:35yngwirst use in this function)
16:10:37yngwi../../../binutils-2.16/libiberty/fibheap.c:285: error: (Each undeclared identifi
16:10:38 Quit NicoFR ()
16:10:38yngwier is reported only once
16:10:40yngwi../../../binutils-2.16/libiberty/fibheap.c:285: error: for each function it appe
16:10:41yngwiars in.)
16:10:50Mordovhow do I load original firmvare?
16:10:55yngwithis are the first lines with error in it :-)
16:10:57Mordovhave to test agein to be surten
16:10:57yngwiof course from the bottom
16:11:13_FireFly_Mordov: hold rec while press on
16:11:16ep0ch|Mordov: hold record and play while booting
16:12:05Mordovbest thing abaout rockbox (and the says alot cause it is god) myst be the startup time :)
16:12:41ep0ch|it might get a little quicker in future too
16:12:55wubblaep0ch|: why's that?
16:13:03ep0ch|rockbox in firmware
16:13:05 Quit B4gder ("time to say moo")
16:13:26yngwii checked it and i think i have all that tools from the cross compiler page...
16:13:38Mordovok iriver firmvare seems to be fine,,, but sometme rockbox lat's me scan thrugh the whole list to,,,,
16:13:42ep0ch|as long as Slasheri doesnt break his unit anytime soon
16:13:47MordovI'll scan and chkdsk :)
16:13:52yngwii also have downgraded to gcc 3.3.3
16:14:01ep0ch|Mordov: would you say you have a lot of direcories/files?
16:14:19Mordov100mb left
16:14:35Mordov1800MB somthing used
16:14:43ep0ch|well if you have a lot of files/directories try changing the limits in settings
16:15:03ep0ch|General -> System -> Limits
16:15:13Mordovlimit the files in my folders?
16:15:20ep0ch|increase them
16:15:42Mordovah, I lookd at that yesterday.. I'll check it agein
16:15:56ep0ch|default is 400, maybe you've changed it or exceeded it
16:15:59Mordovah, the olug unplug also rox:)
16:17:00 Part XavierGr
16:17:42Mordovep0ch| looks like that did the trick :) thanks
16:18:01Mordovand keep up the brilliant work:)
16:18:20yngwiyeah, brilliant work..
16:19:33Mordovum, no it didn't:( sometimes it let's me scan the whole list.... I'll do scannig and come back :)
16:21:23 Join webguest65 [0] (
16:21:30webguest65Hello people
16:21:33linuxstb__yngwi: Which version of gcc are you using? I think that may be the problem. Try upgrading to the latest standard cygwin gcc.
16:22:24yngwihmm on the wiki it says i need gcc-3.3.3 in order to compile 3.4.4 for iriver correct, so i have gcc-3.3.3
16:22:30webguest65preglow:; are you here?
16:23:27preglownope, i'm heading out now
16:23:43preglowif it's possible, i can read it when i get back
16:24:01linuxstb__yngwi: Sorry, I didn't see that (I'm not a cygwin user). You haven't done anything obviously wrong, so I'm not sure what the problem is.
16:24:17yngwiits pretty strange
16:24:27 Quit Mordov (Read error: 104 (Connection reset by peer))
16:25:00linuxstb__One option would be to download the development kit, and then copy the contents of the /opt/m68k from there and copy it into your cygwin installation somewhere.
16:25:19linuxstb__The DevKit is basically a very minimal cygwin with pre-compiled binaries of the cross-compilers.
16:25:33yngwidoes it interfere with cygwin?
16:25:37linuxstb__I've no idea.
16:25:45webguest65preglow: it was just for ask you what's about your 1500 lines libmusepack code, commit will come soon?
16:25:48yngwihmm i guess ill try that, thanks
16:26:24linuxstb__If you do use that devkit, then you will get an error message when compiling the "tools" directory in Rockbox (related to ipod_fw.c). You can safely ignore it.
16:26:43yngwiok thanks..
16:26:45 Join tcmjr [0] (
16:26:59linuxstb__That's a very recent problem - a new devkit will hopefully be released soon to fix it.
16:28:32tcmjrHello all.
16:30:20 Join b0br [0] (
16:32:17 Join muesli_- [0] (
16:32:59 Join Zagor [0] (
16:33:53 Quit webguest65 ("CGI:IRC")
16:40:00linuxstb__Sorry about the log message on last CVS commit - I accidentally pressed ENTER before I had finished typing the cvs commit command.
16:40:22linuxstb__The commit itself turned out to be fine, but that was just luck.
16:40:38linuxstb__It could have been nasty though....
16:42:37linuxstb__For the record, the commit was an optimisation to the ipod lcd_update_rect() function.
16:43:51yngwisay... linuxstb_, the error you told me about, is it supposed to happen when i build the tools?
16:44:08yngwii guess so
16:44:22_FireFly_yngwi: yepp
16:44:44yngwioh sorry i forgot what you wrote 10 lines or so above
16:44:59yngwiFireFly, thanks
16:45:05 Join Mordov [0] (
16:46:00Mordovhmm, I did the scanning bit... but I think it's somthing to do with the Tag DB cause it only hangs when I use the tag wiev function
16:46:16Mordovand the song it hangs oncahnges
16:46:31amiconnlinuxstb__: You can't assume that x is even in an lcd_update_rect() call
16:47:05linuxstb__amiconn: I guessed that.
16:47:21linuxstb__Is there a clever way to correct the x and width values?
16:48:36amiconnYou'd need to do a similar thing as the 1 and 2-bit drivers do for the line blocks
16:48:43 Quit Mordov (Client Quit)
16:48:58amiconnRound down x_start to an even value, and round x_end up
16:49:01yngwihmm it seems it didn't work too well
16:49:43yngwiwhen compiling i got errors
16:50:04yngwimake[3]: *** No targets specified and no makefile found. Stop.
16:50:05yngwimake[2]: *** [libmad] Error 2
16:50:07yngwimake[1]: *** [build-codecs] Error 2
16:50:08yngwimake: *** [all] Error 2
16:51:02_FireFly_has the dir apps/codecs/libmad files in it ??
16:51:06 Join mashalla [0] (
16:51:15 Part b0br
16:51:29 Join Mordov [0] (
16:51:32linuxstb___FireFly_: Yes, but not in the source tarballs
16:51:40_FireFly_btw do you use the bleding edge archive ??
16:51:43 Join Jungti1234 [0] (n=jungti12@
16:52:28linuxstb__Bagder: You around?
16:52:29yngwithe dir "C:\rockbox\home\guest\source\apps"? yes
16:52:47linuxstb__yngwi: You've come at a bad time - the source archives are currently broken.
16:52:58linuxstb__Can you use CVS?
16:53:11_FireFly_the bleeding edge source archive is broken for a while
16:53:18_FireFly_at least some months
16:53:23yngwioh i guess thats the problem then
16:53:35linuxstb__yngwi: Go here:
16:53:55linuxstb__and type the two commands under Anonymous read-only checkout
16:54:04linuxstb__That will get you the very latest source.
16:54:12yngwiok.. thanks
16:54:29Mordovwhat does the voice UI do? can I speak to the player or is it for the player speaking to me?
16:54:42linuxstb__The player speaking to you.
16:54:54_FireFly_it's mainly for blind users
16:54:54Mordovah ok that whay I could'n make it work ;)
16:55:03Mordovwas my guss :)
16:55:07_FireFly_yepp you need voice files
16:55:39MordovI'm tring to uppdate DB file now, hope that helps stopoing to lockup...
16:56:24Mordovit's only lockup in tag mode.. and I deleted som files with Tag mode yesterday and that gave me errors.. are one suposed to be able to delete in Tag mode?
16:59:16***Saving seen data "./dancer.seen"
17:00:59yngwiok thanks guys... i think that was the problem
17:04:08 Join muesli- [0] (
17:05:53_FireFly_TiMiD: are you here ??
17:10:18rand_althoranyone more than Linus that have used the BDM-interface?
17:10:44Mordovhmm nothig helps...... this is frustrating. rockbox still locksup when I brows the artist
17:11:43Mordovwhen I brows all the tag options..
17:15:19Mordovhmm sometimes it works and let me run through it some time's and sudenly it stopps...
17:15:43 Quit muesli_- (Read error: 110 (Connection timed out))
17:19:14 Quit muesli- (Read error: 104 (Connection reset by peer))
17:23:47Mordovhmm I can eaven play the filses it lock's un on if I search for them instad of finding them on the list
17:23:58Mordovall sugestions are wellkome :)
17:26:09 Quit Jungti1234 ("Bye Bye~")
17:29:25markunrand_althor: someone has a bdm for his iaudio
17:29:46markunbut we didn't hear from him in a long time
17:30:39_FireFly_TiMiD: ??
17:32:53 Join muesli_- [0] (
17:33:36 Quit muesli_- (Client Quit)
17:46:30preglowlinuxstb__: it is possible to enter a comment even now
17:46:34preglowlinuxstb__: for the commit, that is
17:47:44preglowcvs admin -m <revision>:<message> <file>
17:48:05 Quit arkascha (Remote closed the connection)
17:48:27linuxstb__OK, I'll give it a go.
17:49:19preglowwebguest65: the musepack commit will just be general speedups, nothing more, but it introduces a bug, so might not be commited for a while
17:50:55linuxstb__Has anyone used cvs admin before? Is the following syntax correct: cvs admin -m "1.10:This is the text of my message" file.c
17:52:15preglowi think if you move the quote to after the colon
17:53:41linuxstb__I don't think that matters - the shell should treat them both the same.
17:53:53_FireFly_cvs admin -m 1.7:"Truncate four-digit years to two in input." date.c
17:54:00_FireFly_found on the net
17:54:01preglowlinuxstb__: that's true
17:54:17linuxstb___FireFly_: Thanks. I'll give it a shot.
17:55:08preglowit wont be visible until the next commit, and i've got a little one queued
17:55:42linuxstb__"cvs log file.c" will tell me if it worked.
17:56:56linuxstb__Lovely. Worked fine on the first file.
17:57:35_FireFly_you must do it for each file seperatly
17:58:43linuxstb__Yes, I did do. In fact, it gave me the chance to make more specific log messages for the changes.
17:59:09linuxstb__It should be OK now. History has been rewritten.
18:00:53linuxstb__preglow: If you're looking for a quick and easy assembler function, there are the swap16 and swap32 functions in firmware/export/system.h
18:02:42 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
18:07:26markunI started a thread on a gigbeat user forum:
18:08:28yngwiyeeh i have compiled the simulator...
18:08:58linuxstb__yngwi: congratulations. What are you planning to do with it?
18:09:11yngwibuild my dream wps :-)
18:09:19yngwibut i encountered another problem
18:10:14yngwiwhen trying to play a file it writes (in a windows box) the application or dll \archos\_temp_codec.dll is no valid windows disk
18:10:17 Join nevs27 [0] (i=lino@
18:10:17yngwior so
18:10:29yngwiand the sim display says codec error
18:10:34yngwiwhat could that be?
18:12:36 Join Moos [0] (
18:15:07yngwiok now its working
18:15:23yngwidid something wrong, but whatever
18:15:34yngwithanks for the great help and work :-)
18:16:39 Join Philip_0729 [0] (
18:17:44yngwii've got a question, though.
18:17:55yngwihow does the simulator simulate battery drain?
18:27:17 Quit mashalla (Read error: 110 (Connection timed out))
18:36:20Mordovbah.. now I can't make DB files anymore,,, teh process seems fine but I get No id3tag DB error when starting, and the file wount uppdate
18:37:58Mordovbuhu I have eaven rainstaled rockbox firmvare
18:40:15 Quit Philip_0729 ("Miranda IM! Smaller, Faster, Easier.")
18:41:07 Quit ep0ch| ("gone")
18:44:16 Quit Mordov ("")
18:59:17***Saving seen data "./dancer.seen"
19:01:34 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:03:24 Join Mordov [0] (
19:03:55 Quit Mordov (Client Quit)
19:04:26 Join Mordov [0] (
19:05:31 Join ender1 [0] (i=ychat@
19:06:31Mordovschuld the .tagdb get bigger as the CMD run's and index is written? cause mine dont...
19:11:28 Quit ender` (Nick collision from services.)
19:11:39 Nick ender1 is now known as ender` (i=ychat@
19:14:56MordovI get this error nomater what I do. please helpe :) not an ID3 rockbox DBarchive....
19:18:41preglowi don't use the db, sorry
19:25:19Mordovit's not writing anything to the DB file anymore... I have tryed it on to computers
19:25:48Mordovrockbox.tagbd file is empty
19:26:06Mordovcan someone send me a file so I know how it's suposed to look?
19:26:34preglowsure you're not using the java prog wrongly?
19:26:41preglowi don't have java, so i can't test
19:27:33 Quit Febs ("CGI:IRC (EOF)")
19:27:34Mordovit might be that, but I have done excatly what written on the page... and tha scould be easy...
19:28:39preglowi don't think the java db generator is actively maintained anymore
19:28:44preglowit might have gone stale
19:28:46preglowi have no idea
19:28:59Mordovit worked earlyer today... that what's the big thing :)
19:29:25Mordovamd the it started to lockup when browsing tag's and I have been struggelign since :)
19:29:35Mordovis pearl better?
19:29:37Mordovfro PC
19:32:36preglowthe perl one doesn't support as much as the java one does, i think
19:32:39preglowlike i said
19:32:44preglowi don't use the db, so i don't know much about it
19:33:14*HCl rubs his eyeballs
19:34:20Mordovok, where can I get hold of someone who knows abaout the DB
19:34:51*HCl looks around a bit
19:35:03markunMorning HCl :)
19:35:18HCli'm tired & bored
19:35:25HCli probably need coffee
19:35:40preglowwell the bored part can be fixed now!
19:36:35Mordovis teh DB file suposed to get biger as the writing goes on do you know HCI?
19:37:39 Quit Moos (Read error: 110 (Connection timed out))
19:37:54HClthe database gets written at the end
19:38:04Mordovgood to know :)
19:38:17HClit can't write it earlier, because it needs to sort it
19:38:23HCland it needs all entries to be able to sort
19:38:33Mordovmine don't write at all;)
19:38:48Mordovand I all worked so fine this morning...
19:39:57yngwiis there a wps tag for repeat-shuffle?
19:41:08yngwioh, i get it
19:41:14yngwiforget about the question
19:41:42 Join arkascha [0] (
19:47:38TiMiD_FireFly_: yes ?
19:48:50_FireFly_a simple question abbout scrollbar.[ch]
19:49:08_FireFly_why do you include there lcd.h twice ??
19:51:31Mordovdo I have to have a iRiver firmvare DB file to make it work???
19:56:49 Quit rand_althor ()
19:58:36Mordovwhere can I get hold of someone who knows abaout the DB file.. I can't get it to work and I don' want to go back to cracpy iRiver frimvare
19:58:42MordovI neeeeeeeed the Tag DB
19:58:54_FireFly_TiMiD: ??
19:59:23markunLooks like this player has a removable HD :)
20:03:21 Join actionshrimp [0] (
20:03:50 Join Lear [0] (
20:05:19 Join _arkascha [0] (
20:08:08Learpreglow: does your crossfeed speedup change the volume in any way (compared to the original code)?
20:17:44 Join bbad [0] (n=bbad@
20:19:36 Quit arkascha (Read error: 110 (Connection timed out))
20:29:38_FireFly_TiMiD: hello ??
20:36:15 Join joehmisticriv [0] (
20:38:03Mordovcan some one please send me a ID3 tag DB file... that might help
20:39:35Mordovor can some one say SHUT UP STUPID FUCK WE DONT WANT TO HELP YOU... ignorace is not a gift ;)
20:39:46 Quit _arkascha (Read error: 110 (Connection timed out))
20:39:47 Join tucoz [0] (
20:40:20tucozMordov, what is the problem with the id3db?
20:41:11MordovI't deosnt write... I got not a id3 tag database eror when startingh
20:41:20tucozIt should be as simple as download SongDB.jar from
20:41:35tucozand run java -jar SongDB.jar music
20:41:52Mordovit all worked this morning
20:42:12Mordovbut I tryed to reinstall all because of some other bug and now I don't get any info in teh file
20:42:26Mordovit makes the file but don't write anything
20:42:34tucozOk, weird
20:42:40tucozI'll try something
20:42:54tucozhmm, I just remembered. I don't have java
20:43:23Mordovand run java -jar SongDB.jar music is the music bit important??
20:43:27 Join gromit` [0] (
20:43:31Mordovmin file don't have the word
20:43:49tucozMordov, no just the name of the directory where your music is :)
20:43:53Mordovah :)
20:44:17Mordovwas irretaed earlyer thatr it wrote the whole disk.. but now I wish I had sutch smalll problems :)
20:45:08Mordovcoud you send me a DB file???
20:45:09tucozhmm, have you tried running the command on your host-computer on a directory with mp3's?
20:45:47tucozOn your pc
20:46:02Mordovok, I'll try
20:46:41 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
20:47:07Mordovtaht worked
20:47:12Mordovlike a dream
20:47:14tucozok, good
20:47:23Mordovso it's rockbox end....
20:47:32 Join DangerousDan [0] (
20:48:20tucozMordov: linux?
20:48:33MordovI'm a stupid microsoft user :)
20:48:49tucozOk, try to run the command from somewhere else. Like in your C:\ dir
20:49:01tucozthat is, go to C:\
20:49:49tucozrun java -jar SongDB.jar (name_of_iriver_dir)\music or whatever
20:50:32tucozI remember earlier that it took ages on the iriver to generate a db, but really fast on the pc
20:50:34_FireFly_Mordov: did you also used the parameter −−strip ??
20:50:43Mordovwow I copied teh file I made in the C: test into the iRiver and now it all works agen
20:50:52tucozMordov: good
20:51:14Mordov_FireFly_ I have tryed everything I could think of;)
20:51:18tucozthen it is not working as it should on the iriver. Do not know why though
20:51:26Mordovguess it all started when I deletedthe old....
20:51:41Mordovmaby java was writning to the ghost of it
20:51:47tucozHCl: if you read this, then you might have some info on this.
20:51:53_FireFly_you need the strip parameter afaik
20:52:11tucozah, that is probably true
20:52:21tucozor, is it?
20:52:24_FireFly_to fix the paths
20:52:35_FireFly_as far as i know = afaik ;)
20:52:56tucozDoes it work now, in rockbox I mean?
20:53:03Mordovso what code schoud I have in my makedbj file?
20:53:24Mordovall works fine. after copying teh DB file I made on c:
20:53:36tucozMordov, what is makedbj? the batch file?
20:54:12_FireFly_i think rb can't the file if the path looks like this c:\mp3s\
20:54:16Mordovwhere do I put the __strip
20:54:42tucozI think it is −−strip
20:55:12Mordovand that comando does??
20:55:33tucozand that would be java -jar SongDB.jar −−strip path_to_directory
20:55:50Mordovok, what does it do? I didn't get that ;)
20:56:38Mordov[20:49] <_FireFly_> you need the strip parameter afaik:)
20:56:40Mordovgot it :)
20:57:35tucoz"4. Run "java -jar SongDB.jar −−strip /mnt/archos /mnt/archos" from inside the .rockbox directory"
20:58:14TiMiD_FireFly_: sorry I got disconnected
20:58:22tucozit strips away what you write after −−strip
20:58:25joehmisticrivdevelopment of the iriver port is definately underway then?
20:58:29TiMiDI fixed it
20:58:46joehmisticrivapologies if thats been asked a thousand times.
20:58:57tucozjoehmisticriv: H3x0 you mean?
20:59:01Mordovjava -jar SongDB.jar /music
20:59:09Mordovthis is the only comando that writes the file :)
20:59:12joehmisticrivyeah :)
20:59:21***Saving seen data "./dancer.seen"
21:00:00tucozSeems like it, doesn't it? I am quite certain that all those commits regarding that platform has a obvious intention.
21:00:24_FireFly_it seams it doesn't matter if you use −−strip or not
21:00:58tucozMordov, if it works, just use what works
21:01:03Mordovthanks for the help guys :)
21:01:16tucozI mean on your iriver
21:01:32Mordovif it anit broken whay fixit :)
21:01:44joehmisticrivooh another question..
21:01:49joehmisticrivim looking at this iaudio x5l
21:02:08joehmisticrivand was just wondering if it would be worth the upgrade from my h340
21:02:15Mordov(except I'm a repair dude at work so I earn money from fixing ting that are working ;)
21:02:23joehmisticrivive read the review of it but am still undecided
21:03:03tucozjoehmisticriv: It looks nice, but I think a possible rockbox port is further in the future than that of a h3x0 port
21:03:16Mordovhave someone moded a better joystick? the original on the 120 sux
21:03:34joehmisticrivyeah that crossed my mind
21:03:48tucozA port was started a while ago, but the person starting it hasn't been seen for some months
21:04:13Learmordov: it isn't so bad if you do something that gives you a better grip of it, imho...
21:04:49Mordovhmm good idea, I wqas thinking of throwing it away and exchange it with some rubber cirkle
21:05:00Mordovpuls instal of leaning ;)
21:05:07Mordovbah PUSH
21:06:22tucozjoehmisticriv, but the hardware on the iaudio is quite similar to that of the h340
21:06:36 Quit bbad ("KVIrc 3.2.0 'Realia'")
21:06:52joehmisticrivi like the button layout on the h340 because ive mastered changing tracks/volume etc while its still in my pocket...
21:07:19joehmisticrivyeah...the 5fps increase on the video playback
21:07:31joehmisticrivand for such a small display 5fps is going to make a difference
21:07:36joehmisticrivthe h340 is slightly choppy
21:07:42joehmisticrivbeen using it to watch simpsons episodes
21:07:52tucozok, 10 fps ?
21:08:48tucozI do not think rockbox could enhance that very much. Maybe by using a uncompressed tailor-made format. But not when showing avi, mpg's or whatever else.
21:08:48joehmisticrivyeah on the iriver, according to this review the x5 is 15
21:10:05tucozjoehmisticriv, maybe the ipods could do better? There has recently been started a ipod porting effort. Seems like it is progressing nicely.
21:11:16Learmordov: e.g., hot glue works quite well.
21:11:29joehmisticrivonly problem with that is most of my music is ogg...ripped from cd
21:11:36MordovLear :D nice
21:12:49tucozjoehmisticriv: ah, the ipod-firmware do not play those. Right. Well, I believe rockbox will play oggs on ipod when that time comes.
21:12:50joehmisticrivalso, i like using the hot swappable drive. Simply drag and drop
21:13:18joehmisticrivand im a bit of a perfectionist
21:13:20Mordovthis is strange... it write DB files when I run java -jar SongDB.jar /MUSIC/techno
21:13:30joehmisticrivthe ipod tend to scratch easily. my iriver isn;t perfect but its damn good..
21:13:30Mordovnot with java -jar SongDB.jar /music
21:13:37joehmisticrivbrb gf demands we watch a film together :)
21:13:39MordovI have to sue a sub folder
21:13:45tucozok, bye
21:15:12Mordovwould I help with an god old /S ?
21:17:32tucozwhat do you mean?
21:17:36tucozlike in dir /s
21:19:01tucozyou probably know this, but by running java -jar SongDB.jar, without any options or directories then you will get all available options printed out.
21:22:19 Join Cassandra [0] (
21:23:20_FireFly_Hi Cassandra
21:23:31Mordovby runnig java -jar SongDB.jar it does the bug dance.. I only get written index if I write java -jar SongDB.jar /music/some random folder
21:23:54Mordovjusat music dont cyt it
21:24:12CassandraBadger: regarding the guy who's having problems reverting to default WPS. Any reason why we shouldn't just check in a copy of the default WPS to the WPS collection?
21:24:37_FireFly_Cassandra: do you have also the problem with your wps (boxes) that the images for the hold didn't get cleared when hold is off ??
21:25:04 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
21:25:39CassandraYes. I believe that's a general bug with graphical WPS code.
21:26:06CassandraSeeking in the track, or switching to the menu and back will clear it.
21:26:18_FireFly_yepp, the problem is if the image-tag is not in the same line as it is shown on screen
21:26:34_FireFly_but this is only right if in the same line is shown some text
21:26:57_FireFly_if only images is shown in this line it doesn't work
21:26:59 Join |joshn| [0] (
21:27:30_FireFly_if no text is shown in this line then the images doesn't get cleared if they should
21:28:46_FireFly_i have a solution for this but i don't know if this works correctly for all wpses which has this problem
21:28:47tucozor maybe that's where you read it :)
21:29:11_FireFly_but for this the wps had to be modified
21:30:01CassandraFirefly - I think it's better to fix the bug in the WPS code.
21:30:31Slasheriyes, the bug should be fixed
21:30:42_FireFly_yepp i had tried some to fix it but it didn*t worked correctly
21:31:17Cassandratucoz: That's what I'm talking about. I think the simplest solution is just to check a copy of the default config into CVS then people can revert the same way as they do with fonts.
21:32:05_FireFly_in one possible solution(i thought it were a solution) some images get also cleared which shouldn't or/and some text wasn't shown
21:32:05CassandraThe other bug that needs fixing is of course remote lock staying on when the remote isn't plugged in.
21:32:06amiconnHmm, still no Linus :/
21:32:08Slasheribtw, i will soon have a stable iriver_flash plugins, that allows to flash the rockbox firmware (but not yet to revert back to original fw)
21:32:13amiconnAny H300 owner in here?
21:32:28 Join Midgey34 [0] (
21:32:36_FireFly_Cassandra: for the remote-hold bug in wps have i a patch
21:32:37CassandraSlasheri: Like to live dangerously, do you?
21:32:56Slasherihehe, how? :)
21:33:19Slasheriall tests so far have been quite safe
21:33:21_FireFly_it adds a function which indicates if remote is plugged in and this function is used in wps-code
21:34:47tucozSlasheri: do you know why rockbox attempts at playing .WMA files?
21:35:00Slasheriamiconn: no h300 and in fact probably i wouldn't even by it anymore because it's discontinued. But iaudio would be interesting to buy and start some porting to it :)
21:35:10Slasheritucoz: nope, checking that
21:35:14tucozShouldn't the codec switch-case catch that?
21:35:21tucozok, :)
21:35:39Slasheriapps/tree.c: { "wma", TREE_ATTR_MPA, Icon_Audio, VOICE_EXT_MPA },
21:35:47Slasherithere is an entry for wma in tree.c
21:35:55Slasherithat should be probably removed
21:35:55_FireFly_this is only for creating playlists
21:36:14Cassandra*nods* I agree.
21:36:14amiconnSlasheri: Isn't the iaudio the one that has the display on the remote only?
21:36:35CassandraOh, I bet I know why it's there.
21:36:41Slasheriamiconn: eh, really? At least iaudio x5 should have a nice color screen on the main unit
21:36:44tucozamiconn: that is the m3 afic
21:36:48Slasheriunfortunately it lacks the optical out :/
21:36:52CassandraI bet wma does something similar to wav with encapsulation.
21:37:00markunamiconn: only the M3 and M5 have not LCD on the main unit
21:37:19 Join davidc__ [0] (
21:37:53CassandraBloody Microsoft and their 'bright' ideas.
21:37:53amiconnThen M3 or M5 are out of question for me
21:40:10davidc__Any of the rockbox dev's working on the iPod port around? [/me is from the iPodLinux project.. wanted to discuss some things]
21:41:34tucozdavidc_: call for preglow or linuxstb
21:42:19CassandraDamn - I've just realised the good reason.
21:42:33CassandraThe defaults are different for the player and others.
21:43:18tucozBut the build script could bundle the different defualts, couldn't it?
21:43:44davidc__anyhow, preglow / linuxstb__ - message me when you're around
21:43:57Cassandra*nod* Yes, it could.
21:45:11tucozAnyway, I think the theme loading sounds nice. i.e. the loading of a font, and the wps.
21:45:59tucozBut, I do believe to have a show status bar tag in the wps is a good thing. It is nice to be able to see that stuff when not in wps as well.
21:46:59Mordovnow it put theDB file on root..... the strange part is that it worked this mornig so there must be somme setting that is alterd and does not reset when I reinstall rockbox...
21:47:13_FireFly_tucoz: i have a patch for this
21:48:27tucoz_FireFly_: ok, I hope it is cvs-material
21:48:57amiconnMordov: The tagdb file goes into /.rockbox, not in into the root
21:49:32MordovI know tahts whay it so strange it automaticly put it in root....
21:49:36 Quit joehmisticriv ("CGI:IRC (Ping timeout)")
21:49:50_FireFly_tucoz: for the old wps-code it is already on tracker
21:50:01_FireFly_i have a new one for my wps-widget
21:50:02Mordovthink I soon have experienced every bug with the DB file today :)
21:50:19Cassandratucoz: I think a show statusbar tag is a good idea. I want to play about with the WPS code some time when I'm not sleeping 16 hours a day. (The joys of chronic fatigue.)
21:52:40 Join Coldtoast [0] (
21:52:51Coldtoasthi all
21:53:12Coldtoastanybody recorded using the h1x0's internal mic?
21:54:32tucozCassandra: seems like _FireFly_ has something up his sleeve :)
21:55:25Coldtoastif so, do you get much HDD wirring noise?
21:55:31CassandraThis is good.
21:55:57crwlColdtoast, yes, quite a lot
21:56:00Coldtoast is mine
21:56:09Coldtoastdoes it sound less than that?
21:56:14tucozthat chronic fatigue stuff do not sound like a nice thing. Hope it gets bettis
21:56:17ColdtoastI just noticed how noisy the drive is
21:56:34MordovI was finaly able to read the last line: exception in thread "main" java.lang.outofmemoryError: java heap space
21:56:39Coldtoastbut I dunno if I just noticed it or if it's just become louder
21:56:43Mordovwhat does this meen anh how to fix
21:57:13Midgey34amiconn: what do you need an H300 owner for?
21:58:02amiconnI need a front-image (preferably a scan) of the main unit, and of the lcd remote
21:58:27amiconnI'm working on the H300 win32 simulator
21:58:37Midgey34hmm, I might be able to get a scan of the main unit
21:58:43Midgey34what res do you prefer?
21:59:42 Quit Sandking (Read error: 113 (No route to host))
22:00:14amiconnColdtoast: Thanks, it looks like it's uable
22:00:23Midgey34Coldtoast: That differs very slightly from the actual unit, there's no "design by inno" and the screen is different
22:00:25Coldtoastlooking for a pic of the remote
22:00:36 Quit Kohlrabi (Nick collision from services.)
22:00:42 Join Kohlriba [0] (
22:00:53Mordovamiconn: figured out whay it when to root... I started it from there ;)
22:00:58*Mordov is a noob
22:01:21amiconnMidgey34: I need to cover the screen area with black anyway. The "deisgn by inno" is a bit ugly though
22:01:36Midgey34well I mean the silver outline
22:01:49Midgey34multi-codec jukebox is on the side
22:02:06Midgey34its not that big of a deal though
22:02:15Coldtoast is prolly a bit small
22:02:40Midgey34well that's an accurate picture
22:03:17Coldtoastthere you go
22:03:43 Quit ehntoo ("Leaving")
22:05:18yngwiyey my first wps with images is complete though it is only a ripoff :-)
22:05:49amiconnColdtoast: That's also a bit small...
22:05:52dpassen1let's see a screendump
22:06:04Coldtoasthmm. ok. can't find anythin gbigger on google images
22:06:12amiconnI need an image that doesn't need to be scaled up
22:07:39amiconnThat means, the lcd area should be 220x176 pixels or bigger
22:07:59yngwidpassen1 do you mean me?
22:08:00amiconnThe outer areas aren't that important; I'll crop them anyway to save space
22:08:20Mordovcan some one please explain this erro msg? exception in thread "main" java.lang.outofmemoryError: java heap space
22:08:35Coldtoasthere you go amiconn
22:08:51yngwiif yes, i plan to post it in the forum bzw. in the wps showroom in the wiki
22:08:55Coldtoastthat should DEFINITELY be big enough
22:09:23amiconnWow... yes this one looks good :)
22:09:42Coldtoastnot the most awesome pic of the outer bit I guess
22:09:47Coldtoastbut f'n HOOOOGE
22:12:16ColdtoastI can find loads of pix of the LCD remote
22:12:24 Quit gromit` ("Leaving")
22:13:13Coldtoast is the biggest h3x0 one tho
22:13:39amiconnHmm, are there two remotes?
22:13:49Coldtoastthere's teh LCD one
22:13:54amiconnI thought the H300 also has an lcd remote?
22:13:58Coldtoastand the non-LCD one
22:14:14Coldtoastthere is now, of course
22:14:23Coldtoastthe original h300 remote had no LCD tho
22:14:31Midgey34amiconn: here's a scan of my h320, not sure if its helpful
22:14:34amiconnThe lcd version is what I would need
22:14:45Coldtoastdon't you have the h100 remote?
22:14:55Midgey34there is an lcd version, but I don't think it's sold outside of korea
22:15:22amiconnColdtoast: Is this identical to the h100 lcd remote?
22:15:30Coldtoastpretty sure, yep
22:15:39Midgey34I thought it differed
22:15:46Midgey34and made use of a joystick
22:15:58Coldtoastoh yeah! there is that one
22:16:08CassandraBloody Apple. Can't install the new Quicktime without iTunes as well.
22:16:09Coldtoastit's ugly
22:16:17Coldtoastyes you can Cassandra
22:16:37CassandraCan you? Have Apple hidden away a *real* installer somewhere?
22:17:42CassandraNice of them to mention that, don't you think?
22:18:23_FireFly_aah i have a solution for the image get not cleared problem in wps :)
22:18:33Midgey34that's the h300 remote
22:18:39Midgey34but the picture won't be helpful
22:18:40CassandraFIrefly: Cool.
22:19:06Coldtoaston that page, they have the link there. not really hidden
22:19:15Coldtoastunder "Upgrade Now"
22:19:31Coldtoastppl just skim tho
22:19:33_FireFly_but i don't know if it works with all wpses correctly(clears areas which shouldn't)
22:19:40 Join rand_althor [0] (
22:19:54_FireFly_Cassandra: but with yours it works
22:20:21dpassen1is crossfeed supposed to lower the volume significantly?
22:20:25CassandraAh yes, I see it.
22:20:56CassandraFirefly: Well, needs to be a general fix, really.
22:21:14Cassandradpassen1, don't think so. ReplayGain does though.
22:21:27dpassen1i have replaygain enabled already
22:21:33Coldtoastwhat does crossfeed do anyway?
22:21:38_FireFly_it's maybe not the best how to do it but the problem seams to be only for image-tags which are used in conditionals
22:21:42dpassen1but toggling crossfeed (after a few seconds) will lower the volune
22:21:59dpassen1crossfeed feeds a low volume section of the left channel to the right channel and vice versa
22:22:07dpassen1for more natural listening (less stereo width)
22:22:43CassandraI'm not sure why that's been implemented as it has.
22:22:44_FireFly_and with a delay to simulate the way through the head
22:22:47Coldtoasthmm. ok. Cos for me, all I seemed to notice was the significant volume drop
22:23:07CassandraI think a more general stereo width function akin to the Archos one would've been a better idea.
22:23:14Coldtoastsounds almost like you're deliberately putting each channel out of phase a bit
22:23:18CassandraAnd not significantly more difficult.
22:24:11_FireFly_am i right that the image problem in wps only referst to images which are used in conditionals ??
22:24:45dpassen1regardless, i'm not sure why it lowers overall volume
22:25:05 Nick paugh is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
22:27:02CassandraNo idea, firefly.
22:27:16 Quit Coldtoast ("Peace and Protection 4.22")
22:29:36_FireFly_it works also correctly for my wps
22:29:46_FireFly_no corruption of the display
22:31:00 Join Paul_The_Nerd [0] (
22:31:28_FireFly_it works correctly with all wpses on my device (3 for main and one for remote)
22:32:02_FireFly_the fix is relative simple
22:32:47Mordovwhay have I sudenly run pout of heap space??
22:33:19Mordovdoes it use the iRiver as cache?
22:33:23_FireFly_i assume that the problem is only when images are used in conditionals so i simple clear all images in a conditional which i found in the skip_conditional function
22:33:24 Join Zoric [0] (
22:34:38_FireFly_*clear the area if the image
22:34:50_FireFly_on the display
22:37:35tucozMordov: That means that java has run out of heapspace. Nothing to do with iriver
22:38:24Mordovbut I don't see whay it has run out suddenly and I can't see what it does teh same on both my computers
22:38:33Mordovand I can't fix it..
22:41:23_FireFly_if someone want to test my fix can use my build(h120/140) it can be found here:
22:43:22tucozMordov, try java -jar -Xmx256m SongDB.jar
22:44:15linuxstb__davidc__: Are you still around?
22:45:34Paul_The_NerdWhat does that fix _FireFly_ ?
22:45:40amiconnlinuxstb__: Do the colour lcds have contrast adjustment or not?
22:46:08amiconnI ran into a build problem with the simulator involving lcd_default_contrast()
22:46:22linuxstb__amiconn: I think so, but I haven't explored it.
22:46:51amiconnIf they don't I would remove the function, otherwise it needs to be moved into lcd-16bit.c
22:47:25amiconn...because the hardware specific driver files don't get built for the simulators
22:49:22 Quit Lear ("Chatzilla [Firefox 1.5/undefined]")
22:49:50_FireFly_Paul_The_Nerd: to fix this problem for some wpses: Cassandra: do you have also the problem with your wps (boxes) that the images for the hold didn't get cleared when hold is off ??
22:59:25***Saving seen data "./dancer.seen"
23:02:05 Part tucoz
23:06:37 Join matsl [0] (
23:07:21ZoricHi. Is rockbox on iRiver still known to have problems playing mp3s with a sample rate other than 44,1khz ?
23:07:43Bagdernot that I know at least
23:08:43Bagdersourceforge updated its look
23:09:01Paul_The_NerdThe world is ending, apparently
23:09:06Paul_The_NerdPenny Arcade changed too.
23:09:22Zoricbecause when listening to an audiobook of the latest Harry Potter, which is MP3 64kbps 22050hz, fastforwarding doesnt work correctly, nor does the progressbar or timedisplay. And rockbox reports the file as being MP1 128kbps O_o
23:09:51Zoricbut it plays correctly (sound)
23:09:53Bagderis it CBR or VBR?
23:10:06Zoricchecking in foobar
23:10:46Zoricdont think its VBR... suppose its CBR
23:12:08Zoricyup its CBR
23:12:59 Quit rand_althor ()
23:16:02linuxstb__amiconn: I don't think you can change the contrast on the ipod's colour lcd. The apple firmware had a contrast setting for the b/w lcds, but it doesn't exist on my colour model. Also, the IPL source only has contrast code for the b/w lcds.
23:16:39amiconnI'll just move the function for now, it can be removed later anyway
23:17:06amiconnI just noticed that the lcd_get_*ground() functions for 16 bit lcd are imprecise
23:17:22linuxstb__A bug in the green?
23:17:36amiconnNo, in all colour channels
23:17:58linuxstb__I noticed a mistake when I made the bmp writer, but forgot to change those functions.
23:18:11davidc__linuxstb__: check your privmsgs
23:18:20amiconnThe problem is with chopping and restoring low bits
23:18:37amiconnA simple example (rgb444)
23:19:12amiconnIf you set white,{0xff,0xff,0xff) is chopped to {0xf,0xf,0xf}
23:19:16 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
23:19:44 Nick linuxstb__ is now known as linuxstb (
23:19:47amiconnThe simple restore method that just involves left-shifting then yields {0xf0,0xf0,0xf0} which is certainly not white anymore
23:20:31 Join b0br [0] (
23:21:13amiconnThere is a simple solution for this problem, which I have to implement in the simulator lcd_update*() functions anyway
23:21:26 Part b0br
23:21:44_FireFly_isn't it easier/better to use the lower 4 bits(in rgb444) and save the upper 4 bits for restoring ??
23:22:12amiconnYOu have to shift down when setting the colour, otherwise the colour would be wrong
23:22:32amiconnAs mentioned, it is simple to return the correct (rounded) colour later
23:23:02amiconnAnyway, it's not necessary to do this for lcd_get_*ground()
23:23:14_FireFly_i don't understand this why to use the upper bits instead of the lower bits
23:23:51amiconnThe return value does not need to be shifted up, as it's defined to return values in the native range (0..MAX_RED) etc
23:23:53davidc__anyhow linuxstb, that bootloader looks good
23:24:01davidc__does it work on all generations?
23:24:28linuxstbNo - only the devices we have - which is a 60GB color (lcd type 1) and a Nano.
23:24:41linuxstbIt should work with lcd type 0 color, but it's untested.
23:25:28linuxstbWe're just targetting the PP5020 devices initially.
23:25:29davidc__Hmm.. we can port it to all gens as we have the testing resources
23:26:24Paul_The_NerdIsn't the nano PP5021?
23:26:25davidc__we would also want to add hfs+ support
23:26:32davidc__PP5020 is very similar to 21
23:26:38linuxstbThe bootloader is basically a cut-down version of rockbox - using the rockbox drivers for ATA, FAT32 etc. So porting the bootloader would mean that Rockbox versions for those devices are close.
23:26:40davidc__only some iram size changeas
23:26:52davidc__ah, hrm
23:26:56linuxstbDoes the Nano have a different amount of IRAM then?
23:27:06davidc__yeah.. the 21 has an extra 24k IIRC
23:27:14linuxstbThat will please preglow a lot.
23:27:21linuxstbHe's the dev with a Nano,
23:27:38Paul_The_NerdHow different is the 5G ipod then?
23:27:49davidc__Very different
23:28:06davidc__some of it is the same
23:28:12*Paul_The_Nerd shrugs
23:28:12davidc__but the video path is definitely different
23:28:18davidc__and the audio path is most likely different
23:28:29davidc__they use the broadcom chip for decoding the video / lcd drawing
23:28:29Paul_The_NerdNot that it matters to me, since I own none of them
23:28:31davidc__we know that much
23:28:39davidc__and might use it for mp3 decoding
23:28:44linuxstbIs it possible to access the LCD without the broadcom chip?
23:28:48Paul_The_NerdThough there's a high probability of me purchasing a Nano once Rockbox will let me play at least oggs on it.
23:28:59amiconnThe sim is showing colour :-)
23:29:07davidc__linuxstb: it is possible, but extremely slow
23:29:20linuxstbToo slow to be usable?
23:29:24davidc__aka, for a basic UI, its possible if you're only redrawing miniscule portions of the display
23:29:37davidc__but for things like iPL works on like video decoding, its completely unusable
23:29:44davidc__IIRC, on the order of seconds to redraw
23:30:05Paul_The_NerdWhen you're measuring in SPF it's not good.
23:30:26davidc__well, the way to do it proper is to do your drawing on the broadcom chip
23:30:36davidc__but the thing is
23:30:42davidc__the broadcom uses a custom instruction set
23:31:05BagderDSPs tend to it seems
23:31:57linuxstbI have a question about the ipod grayscale LCDs - how are the pixels packed in the framebuffer? Our existing LCD code for 2bpp displays packs the pixels vertically - 4 vertical pixels per byte. Is the ipod the same?
23:31:58davidc__yeah, it seems to be a variant of ARC.. but we're not sure yet
23:32:16davidc__they're horizontal, from right to left IIRC
23:32:22davidc__it was a long time ago I poked at that stuff
23:32:23 Join Sandking [0] (
23:32:28 Quit _FireFly_ ("Leaving")
23:33:07linuxstbThat makes adding support for the greyscale LCDs more work. But only a small amount of the driver needs to be functional for the bootloader to display text.
23:33:48davidc__well, how modular is your driver code?
23:33:51 Join RotAtoR [0] (
23:33:58davidc__oh, and detecting ipod generations is a real bitch
23:34:02davidc__there are a lot of little tiny things
23:34:19davidc__you might just want to out-and-out take our detection code
23:34:21linuxstbWe avoid that - we do it at compile-time and produce different builds for different targets.
23:34:44linuxstbSee our build table here for the current targets:
23:34:55linuxstbEvery CVS commit generates a new set of builds.
23:35:54davidc__cool.. yea.. we went for the "universal" option.. its less code to sling around
23:36:01davidc__but it does make for some very interesting kernel init codew
23:36:09linuxstbI can imagine.
23:36:23Bagderless code?
23:36:33davidc__er, less binaries
23:36:34Bagderhow can it be less code to do run-time checks?
23:36:40davidc__*been a long day ;)
23:37:10Bagderwe actually do run-time checks on the Archos player target
23:38:00linuxstbdavidc__: Has IPL switched to subversion now?
23:38:15davidc__the new source tree that we just announced is in svn
23:38:23davidc__but the kernel and old pz is still in cvs @ sf
23:38:37davidc__sourceforge has been really flaky lateley, so I dunno how long we're gonna be there
23:38:56Bagdersourceforge been flaky for years
23:39:08Bagderespecially CVS
23:39:17davidc__heh, we havent had any anonymous cvs access for a few weeks now
23:39:23davidc__apparently it is scheduled downtime
23:39:29davidc__according to sourceforge
23:39:36davidc__2 weeks scheduled downtime.. my ass.
23:40:35davidc__we should look into sharing code.. we're doing a lot of hand optimizing of assembly for the ipods to do all sorts of fast stuff
23:40:43davidc__those things are probably portable
23:40:57davidc__[aka graphics blits, blends 888->565]
23:41:19linuxstbWe'll mainly be looking at optimising our audio codecs - they are already very well tuned for use in portable players.
23:41:28markunBagder: thanks for the Gigabeat Surgery site
23:41:48davidc__Do you use custom codecs?
23:41:49BagderI didn't add that, someone else did
23:41:52davidc__or a port of other ones?
23:42:16Bagderwe use other's, ported and optimized by us
23:42:17linuxstbWe use the standard libraries - libmad for MPEG, Tremor for Ogg.
23:42:43davidc__cool - just so you know... we got fairly poor performance out of those
23:42:50davidc__gcc is really bad at optimizing for arm
23:42:52Bagderwe don't ;-)
23:43:00davidc__well, for the iPod I mean ;)
23:43:20davidc__we got much better performance out of the codecs that real released
23:43:30BagderI'm surprised
23:43:33linuxstbWe found one of the biggest gains were using IRAM efficiently. There is a big difference in speed between the DRAM and IRAM on the iRiver devices.
23:44:13linuxstbpreglow: The man with extra IRAM :)
23:44:13davidc__yea.. we've been looking into using that for accelleration, but the ipod has nasty cache coherency issues
23:45:30davidc__so the iram ends up being used for all sorts of things that require cache coherency
23:46:05linuxstbWe use 48KB of IRAM for the codecs and the other 48KB for general use - thread stacks and some core functions.
23:46:24davidc__yea, theres nowhere near that much iram
23:47:00linuxstbWhat are the caching issues we need to be aware of?
23:47:02davidc__[on the ipod that is]
23:47:06davidc__there are two cpu cores right?
23:47:16davidc__they both have separate caches
23:47:19davidc__they are non-snooped
23:47:30davidc__so basically, you've gotta handle cache coherency on your own
23:48:02preglowlinuxstb: i'd gladly take more!
23:48:26davidc__preglow: your nano has some extra IRAM iirc
23:48:39preglowargh, i was hoping all rockbox archs had 96kb
23:48:47davidc__no, just your nano
23:48:59preglowhow much?
23:49:27davidc__jas.. checking my notes
23:49:34preglowdavidc__: what, separate caches? like 48kb each?
23:49:48amiconnpreglow: This isn't true anyway; SH1 just has 4KB
23:49:58preglowamiconn: well, sh isn't swcodec
23:50:31amiconn[23:48:29] <preglow> ... > all rockbox archs < ... ;-)
23:50:56*amiconn is picky today
23:51:21preglowno, i knew sh has less iram
23:51:28davidc__each processor has an 8kb unified cache
23:51:34davidc__non snooped between the two
23:51:36davidc__and its not writeback
23:52:20preglowdata only?
23:52:35amiconnafaik unifiied means code+data
23:52:48preglowi'm a bit shaky on the terms
23:52:50amiconnbetter than coldfire
23:53:27preglowwe basically propose to use one of the cores for codecs only, so i don't think we'll have too many coherency issues
23:54:19davidc__wait, 5020 _does_ have 96k iram
23:54:30davidc__the newer ones were 24 k up from that
23:54:55linuxstbdavidc__: Do you (i.e. IPL) know how to adjust the CPU frequency?
23:54:59 Join pinkutank [0] (n=ddd@
23:55:09davidc__look @ the kernel init code
23:55:12preglowthe iram is shared between cores, i take it?
23:55:19davidc__preglow - yeah, its non-cached
23:55:23davidc__and no wait state
23:55:28preglowof course, heh
23:55:29davidc__so basically, thats what you get to use for IPC
23:55:49davidc__besides the mailboxes, which we don't ahve all figured out yet
23:56:09preglowsounds exciting
23:56:13linuxstbIs this standard ARM stuff, or specific to PortalPlayer's implementation?
23:56:16preglowdavidc__: so you're one of the ipl reverse engineering people?
23:56:25davidc__well, leachbj is the crazy reverse engineer
23:56:37preglowdavidc__: nice anyway, i appreciate your work
23:56:44linuxstbAs do I.
23:56:51davidc__eh, I'll pass on the compliments
23:57:01davidc__I don't do much RE because he always beats me to it ;)
23:57:15pinkutankhellp, sorry for repeating after yesterday, but anyone with a non interrupting compiled jpeg viewer on?
23:57:59linuxstbdavidc__: Which ipods do you own? We need a tester for the "old" LCD code for the 4G Color/Photo.
23:58:11preglowpinkutank: why don't do it yourself?
23:58:38davidc__I've got a old lcd photo
23:58:41davidc__and a 1G
23:58:52davidc__er, old lcd color

Previous day | Next day