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

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

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

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

#rockbox log for 2008-11-15

00:00:42 Join Thundercloud [0] (
00:02:27*bluebrother has an idea
00:03:04Zagorcould the problem be in the converters?
00:03:28Zagor(or whatever the tools that create pdf and html from tex are called)
00:03:35 Quit ap0 ("Baļ")
00:04:09bluebrotherthe pdf is created directly, so there is no actual conversion.
00:04:48 Quit {phoenix} (Read error: 104 (Connection reset by peer))
00:05:16 Join dany_21a_ [0] (
00:17:16 Quit dany_21a (Read error: 113 (No route to host))
00:18:10kugelbluebrother: I'd be able to test your finding on my fuze with a microsd, but unfortunately, due to the lack of buttons, I cannot get past the main menu anyway
00:20:50kugeldany_21a_: hey, how's your new fuze going?
00:21:06kugelLambdaCalculus37: have you tried the final backlight fading patch?
00:21:25 Quit Horschti (Read error: 110 (Connection timed out))
00:21:37LambdaCalculus37kugel: Not yet. Beer called.
00:21:45LambdaCalculus37I have a little time, so let me try now.
00:22:01 Join Horschti [0] (
00:22:22 Nick Horschti is now known as Horscht (
00:24:00bluebrotherkugel: my findings on fuze?
00:26:08kugelbluebrother: test your findings on my fuze
00:26:30*bluebrother wonders what his findings were
00:27:18*pixelma guesses kugel confuses the Dominiks
00:27:28bluebrotherI know some basic english at least ...
00:27:41kugelbluebrother: sorry, I read your sentence wrong
00:27:52*bluebrother felt like getting mixed up with someone else :)
00:27:52kugeland pixelma is right :S
00:28:16kugelsorry :/
00:28:36*kugel read "wonders what findings were"
00:28:57*bluebrother hands kugel an s/were/are/ for that misunderstanding :)
00:29:37kugelthanks :) I'm gonna keep it for future reference
00:30:15kugelthe "his" was causing the missunderstand though
00:30:33 Join einhirn [0] (
00:30:43kugeldomonoky: I'd be able to test your finding on my fuze with a microsd, but unfortunately, due to the lack of buttons, I cannot get past the main menu anyway
00:30:47kugelanother try :p
00:34:02dany_21a_hi kugel (and others)
00:34:25 Quit nplus (Remote closed the connection)
00:35:15dany_21a_have got my (new) fuze yesterday... prior i have killed (a merly new) fuze... dont know if it helps someone, have put some pictures of the inside online (is already linked in the wiki)
00:36:58 Quit ender` (" The last time someone listened to a Bush, a bunch of people wandered in the desert for 40 years.")
00:37:07dany_21a_i have installed the current svn version on the defunct (the buttons dont work anymore - except the homebutton) - it only reaches "Loading firmware"... will try the CCU_IO later
00:38:41LambdaCalculus37kugel: Building now... gotta go soon though.
00:42:24domonokykugel: still testing would be nice, to see if the sd-driver works without the CCU_IO... (if it doesnt, you wont see the main-menu .-) )
00:44:13 Join hollanbm [0] (n=hollanbm@
00:44:34hollanbmhey all, i was wondering if anyone had experience configuring loader2
00:44:47Lloreanhollanbm: Loader2 is iPodLinux's software, you should ask them.
00:44:49BigBambihollanbm: You need to aks ipodlinux about that
00:45:14hollanbmi just figured that since i needed to add a line for rockbox i should ask here
00:45:23hollanbmdo you happen to know there channel
00:45:32LambdaCalculus37Try #ipl or #ipodlinux
00:45:59scorche#ipodlinux not #ipl
00:46:19LambdaCalculus37scorche: Thanks, I knew it was one of the two but wasn't 100% sure.
00:47:15hollanbmi got it
00:47:19hollanbmthx guys!
00:49:29hollanbm(sighs) there channel is dead basically
00:49:45hollanbmanyone here think they might be able to help?
00:49:52LambdaCalculus37Then ask on their mailing list, but very honestly, we can't help you out with Loader2.
00:50:46hollanbmI would think people would use there bootloader versus the rockbox one, that way one can choose which OS he would like
00:50:50BigBambihollanbm: This is a Rockbox channel, we support the Rockbox bootloader
00:51:06hollanbmokay, well then I'm having problems with it
00:51:08BigBambiYou can chose with the Rockbox bootloader too
00:51:15BigBambiIt just doesn give you a menu
00:52:04hollanbmwell everytime i plug my iPod running rockbox into my computer through USB it reboots directly into Apple OS
00:52:07hollanbmwhich i dont want
00:52:16LloreanYou don't want USB?
00:52:26hollanbmno, i dont want to use/see apple OS at all
00:52:31BigBambiYou have to
00:52:40BigBambiRockbox doesn have USB yet for your ipod
00:52:43LloreanRockbox doesn't have USB yet, so the only way to connect to your computer is through Apple right now
00:53:00BigBambiBesides, that isn't the bootloader
00:53:03hollanbmthrough the emergency disk mode right?
00:53:08BigBambiOr the OF
00:53:20hollanbmits when i disconncect that it reboots into apple OS
00:53:28LloreanDo you have Hold on?
00:53:34LambdaCalculus37Because you're probably leaving the hold switch on.
00:53:43BigBambiThere you go then
00:53:49hollanbmwhat does that mean?
00:53:52BigBambiThat is how you dual bot with the Rockbox bootloader
00:54:01hollanbmthe hold switch on reboot?
00:54:08LambdaCalculus37Clicking the hold switch on when you turn on your iPod boots the OF.
00:54:10BigBambion boot, yes
00:54:14BigBambiThe manual tells you this
00:54:47hollanbmI did not know the bootloader could dual boot which is why i thought i had a problem
00:54:56hollanbmtime to get rid of that loader2 shit then
00:54:58BigBambiThat is why reading the manual is recommended
00:55:06 Quit Zagor ("Client exiting")
00:55:25LambdaCalculus37Yes, reading is fundamental. ;)
00:56:16hollanbmthx guys!
00:56:58hollanbmgrrr, the rockbox utility will not let me overwrite loader2
00:57:19hollanbmit says the iPod could not open the iPod
00:57:27BigBambiAdmin rights?
00:57:38hollanbmI will try running the program as sudo
00:57:39BigBambiAs in, do you have admin rights
00:57:40LambdaCalculus37And how does an iPod recursively open itself?
00:57:49BigBambihollanbm: That is also in the manual I believe
00:57:55BigBambihollanbm: Please read it
00:58:07hollanbmthe manual for the rockbox utility?
00:58:17LambdaCalculus37And Rockbox itself.
00:58:40LambdaCalculus37Rockbox Utility has a wiki page.
00:58:59hollanbmIve read the wikis
00:59:00 Join Slack_ [0] (
00:59:11BigBambihollanbm: The Rockbox manual
00:59:25LambdaCalculus37kugel: Successful build, but I've gotta go now. My LUG guys are showing up for some beer time.
00:59:37*LambdaCalculus37 is off to enjoy some beer
01:00:20 Quit LambdaCalculus37 ("Ka-chunka")
01:00:50BigBambiActually, I'm not sure the manual does tell you to use admin/root rights with rbutil
01:02:12hollanbmgrrr damn linux wont let me open the rbutility from the command line
01:02:39 Quit hollanbm ("Leaving")
01:03:50linuxstbHmm, that's the second person in about two days who used loader2 because they didn't think the Rockbox bootloader could dual-boot, rather than because it has a pretty menu...
01:04:29BigBambiPeople should read the manual/search... :)
01:06:54 Join domonoky1 [0] (
01:08:14BigBambiIs it all ipods and sansas that need root/admin to install the bootloader?
01:08:22 Join Slack__ [0] (
01:08:22 Quit Slack_ (Operation timed out)
01:08:28BigBambiAnd do any others?
01:08:41LloreanIt should be possible to do it without root/admin for Sansas, right?
01:08:53BigBambiI'm not sure :)
01:08:59LloreanWell, they require it now.
01:09:15linuxstbIt depends on your OS as well - you don't need admin/root rights, just write access to removable raw disk devices.
01:09:33BigBambiI'm adding a note to the manual to run rbutil as root/admin where required
01:09:41LloreanBut I *think* that we, as an alternative, can just put a PP5022.MI4 on the device containing a dump of its current firmware with a Rockbox bootloader and next reboot it'll upgrade itself?
01:09:54BigBambilinuxstb: So are there any OSs I should exclude?
01:10:02LloreanBigBambi: I *think* current versions of RBUtil are supposed to tell you when you need more privileges?
01:10:28BigBambiLlorean: I'll try it out
01:10:32linuxstbI'm almost certain you don't need to be root on OS X.
01:10:40BigBambiLlorean: Enough people seem to miss it though
01:11:02linuxstbBut I think you do on Windows, and on most Linux distros (in their default configuration).
01:15:38linuxstbBut I thought rbutil used the USB PIDs to detect devices?
01:16:40BigBambiI think it does where possible
01:17:32linuxstbIf it detects a device via USB PIDs, but then can't find the raw disk device, then the reason is almost certainly going to be permissions.
01:17:41linuxstbSo it could give a more helpful error.
01:17:46BigBambiVery true
01:18:16BigBambiI haven't tried it, Llorean seemed to think that it might
01:18:36BigBambiStill, I think a note in the manual would be useful - we do tell people to read it :)
01:18:39pixelmathere's a hint for it in the "manual bootloader installation" part and a note that if you are stuck in the process using the Rockbox Utility you should have a look at that section, but: I know the installation chapter leaves a bit to be desired (and I started collecting info for a restructuring but having to write a lot of stuff in English...)
01:19:05pixelmathe note is in the "automated installation" section, of course
01:19:24BigBambiI don't mind the writing bit, I get a bit lost in all the options and latex stuff
01:21:02pixelmaI should really get my notes into a list form and ask for help in the forums or so :\
01:21:32pixelmahelp in making a text out of it
01:21:34BigBambiI will gladly write stuff :)
01:23:28 Quit domonoky (Read error: 110 (Connection timed out))
01:23:32 Part toffe82
01:24:08BigBambiAt the moment I have just a little warning in the automatic section saying "\opt{ipod,sansa}{\warn{On Linux and Windows, it is necessary to run \textsc{Rockbox Utility} with root or administrator rights respectively in order to install the bootloader. If you are simply updating Rockbox, this is not required.}}"
01:24:13 Quit sin613 (Read error: 110 (Connection timed out))
01:24:30BigBambiDo you think it is worth adding something like that, or would you rather wait for a bigger rework?
01:26:47pixelmaI don't know when this rework will happen, so better add it now to have something at least
01:27:14BigBambiOK, well I can't add it so I'll check I haven't broken anything then post it someplace :)
01:27:33 Part dany_21a_
01:27:44BigBambiAlthough it is just that line in installation.tex
01:27:51linuxstbamiconn: Have you looked at the latest bmp resize patch on flyspray?
01:28:19pixelmaBigBambi: somehow I too think that the Rockbox Utility tells you about missing rights, but it could be that it only works in XP or so
01:28:58 Quit bluebrother ("leaving")
01:28:58BigBambipixelma: It may well do, but I think we have enough people that ask to justify it - at least this way when we say RTFM it is justifyable :)
01:29:33pixelmayeah, agreed
01:29:46domonoky1yes, there is rights checking in rbutil, but i think it works only on windows...
01:30:08domonoky1so having it in the manual would be good :-)
01:30:25*linuxstb still wonders why the Windows version of rbutil is distributed as a zipped exe file...
01:30:51domonoky1stupid web "filters" ...
01:31:16linuxstbReally? No-one else seems to have problems - many Windows programs are distributed as .exe files.
01:31:43linuxstbAnd if it really is a problem, just make a zip file available as a second download option.
01:31:52pixelmaBigBambi: I just wonder about the \textsc , is it used somewhere else (directly)?
01:31:55domonoky1maybe we should just try, and see if we get complains :-)
01:32:29BigBambipixelma: Yes, in the line below where I added it - that? why I used it :)
01:33:12BigBambiThe section that starts "\warn{When first starting \textsc{Rockbox Utility} run ``Autodetect'',
01:33:12BigBambifound in the configuration dialog"
01:34:31pixelmaok, I usually just copy from existing parts too :)
01:34:51BigBambiOften the best way, not always :)
01:35:04 Quit domonoky1 (Read error: 104 (Connection reset by peer))
01:36:45Unhelpfulah-HA... the missing files weren't *from* svn, they were generated
01:45:00 Quit Nico_P (Remote closed the connection)
01:45:37***Saving seen data "./dancer.seen"
01:50:58 Quit culture (Read error: 110 (Connection timed out))
01:57:22kugeldomonoky: (for the logs) I still cannot get past the bootlogo (the bootlogo of the main binary, not bootload). The bootlogo is always shifted left by some 20pixel. Sometimes it even says "No partition found". This is my svn diff, maybe you made some more changes?
01:59:36kugeldomonoky: (for logs again). If I don't remove (i.e.if I don't do "#if defined(HAVE_MULTIVOLUME) && 0)") the init of the external sd card I always get ata error -1 (like before). So removing that CCU_IO didn't change anything for me. Maybe I haven't disabled it enough?
01:59:52 Quit kugel ("ChatZilla 0.9.83 [Firefox 3.0.3/2008101315]")
02:03:31 Quit tyfoo ("Carpe diem")
02:04:35 Quit gevaerts (Nick collision from services.)
02:04:45 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
02:09:48 Nick fxb__ is now known as fxb (
02:09:53 Join HellDragon [0] (
02:11:46 Join obo [0] (n=obo@rockbox/developer/obo)
02:21:44 Join ap0 [0] (
02:27:23 Quit advcomp2019 (Read error: 104 (Connection reset by peer))
02:30:07 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
02:31:26 Quit reacocard (".")
02:32:54 Quit Schmogel (Read error: 104 (Connection reset by peer))
02:40:48 Quit MethoS (Remote closed the connection)
02:41:47 Part pixelma
02:42:14 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
02:43:03 Quit PaulJam (".")
02:47:18 Quit n1s ()
02:47:36pixelma2hmm... just noticed that the last H100 build also threw such a weird error as seen in OndioSP and Player builds before lately. But then the H100s have something like rombox too (although the error message doesn't mention it, just trying to find relations between the targets that fail)
02:48:14 Join obo_ [0] (n=obo@rockbox/developer/obo)
02:49:49 Quit obo (Read error: 110 (Connection timed out))
02:58:12 Quit Thundercloud (Remote closed the connection)
03:00:42 Quit ap0 (Read error: 104 (Connection reset by peer))
03:03:15 Quit einhirn (Read error: 131 (Connection reset by peer))
03:17:26 Join reynaldo_abc [0] (n=bea6c81c@gateway/web/cgi-irc/
03:19:32reynaldo_abcI will be installing rockbox in my mp3 in a few days, i just came to the question that if i use rockbox utility does it install the rockbox bootloader too? or i have to manually install?
03:19:37 Join hillshum [0] (
03:23:25 Quit reynaldo_abc ("CGI:IRC (EOF)")
03:25:55soapit installs the bootloader as well.
03:30:23advcomp2019soap, he left, but i see the person on abi tho
03:34:15 Part Llorean
03:34:41 Quit fyre^OS (Read error: 104 (Connection reset by peer))
03:34:44 Join Llorean [0] (
03:40:29 Join fyrestorm [0] (
03:42:16 Quit avis (Remote closed the connection)
03:43:24 Nick plus_M_ is now known as notplus_M (
03:45:39***Saving seen data "./dancer.seen"
03:45:42 Quit hillshum (Read error: 110 (Connection timed out))
03:45:47 Nick fxb is now known as fxb__ (
03:46:03 Join Darksair [0] (n=user@
03:47:32 Join homielowe [0] (
03:47:55 Join gabe565 [0] (
03:48:47 Quit gabe565 (Read error: 104 (Connection reset by peer))
04:09:47 Join gabe565 [0] (
04:14:49 Quit gabe565 ("ChatZilla 0.9.83 [Firefox 3.0.4/2008102920]")
04:28:19 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
04:29:38 Join miepchen^schlaf_ [0] (
04:38:15 Quit agaffney (Read error: 104 (Connection reset by peer))
04:38:28 Join agaffney [0] (n=agaffney@gentoo/developer/agaffney)
04:39:27 Quit _lifeless (Remote closed the connection)
04:40:10 Join _lifeless [0] (n=lifeless@
04:40:22 Join reacocard [0] (n=reacocar@
04:45:06 Join blkhawk- [0] (
04:46:01 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:59:32 Join Ginger [0] (
05:01:11 Quit Ginger (Client Quit)
05:01:24 Quit blkhawk (Read error: 110 (Connection timed out))
05:22:21 Quit Darksair (Connection timed out)
05:25:13 Quit obo_ ("bye")
05:28:22 Join kkurbjun [0] (
05:37:47 Join Hillshum [0] (
05:38:46HillshumWill the VMWare RB environment image run under VirtualBox?
05:45:43***Saving seen data "./dancer.seen"
05:46:31HillshumOkay, so I logged in at a command line, how do I get X going?
05:48:31HillshumI've pushed Ctrl+Alt+F7, but that doesn't work
05:55:22blkhawk-dont ask me
05:55:29blkhawk-i just woke up
05:55:46ameyerlog into a console and run "startx"
05:55:49ameyerI think
05:57:16 Join Darksair [0] (n=user@
05:57:36Hillshumlots of error data
05:58:27ameyerif it's anything like running Linux on actual hardware, that's probably not actually errors
05:58:57ameyerand X should eventually start
06:00:54Hillshumit didn't
06:01:13HillshumI think the Disk image isn't fully working
06:03:23HillshumScreen flashes, comes back with text
06:03:40HillshumShouldn't tty7 have X already?
06:04:47ameyerdepends on wheter gdm/kdm/xdm is running
06:04:51ameyeror not
06:06:39HillshumHow do I check?
06:07:12ameyerps aux | grep dm
06:07:16ameyerI think
06:07:51Hillshumone line?
06:13:47Hillshumi get nothing from that
06:16:24ameyerlooking at the wiki page, I'm completely lost
06:17:43*ameyer doesn't think he's ever successfully dealt with XFree86
06:21:21Hillshumi think it's supposet to work fine :)
06:22:47HillshumI accedently ctrl alt deled it and it spat out some stuff indicating some x stuff was running
06:25:15 Join blkhawk [0] (
06:25:27Hillshumlast thing on tty1 is "Starting X display manager: xdm"
06:25:58ameyerI know this isn't the Linux way, but reboot?
06:26:06ameyeryour vmware image obviously
06:26:19Hillshumit then goes blank and looks like it is going to x
06:26:24Hillshumdid once
06:28:10Hillshumas always
06:31:51Hillshum"Right click on Xshells and Eterms" ??
06:34:10ameyerthat'd require X and Fluxbox to be working
06:34:22 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
06:36:09 Join blkjawk[work] [0] (
06:38:22Hillshumhere's the last bit of output
06:41:43 Quit blkhawk- (Read error: 110 (Connection timed out))
06:45:00 Quit Hillshum ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
06:52:32 Quit blkhawk (Read error: 110 (Connection timed out))
06:57:45synergistwindows live's in ur persnal shit, hostin ur pics
06:58:01*synergist felt oddly compelled to say that, apologies. think it's too much coca cola
06:59:23ameyerwindows live?
06:59:32ameyerisn't that an oxymoron?
07:15:43 Part kkurbjun
07:17:18 Join BjoernErik [0] (
07:17:31 Quit Bjoern-Erik (Read error: 60 (Operation timed out))
07:19:49 Join bertrik [0] (
07:21:47 Join J-23 [0] (
07:24:17 Quit XavierGr ()
07:26:03 Quit miepchen^schlaf_ ()
07:45:44***Saving seen data "./dancer.seen"
07:56:50 Quit J-23 (Remote closed the connection)
07:58:02 Join pabs [0] (
08:03:51 Quit bertrik ("Leaving")
08:06:15 Join J-23 [0] (
08:10:33 Quit J-23 (Remote closed the connection)
08:15:36 Quit reacocard (Read error: 110 (Connection timed out))
08:22:19 Quit Darksair ("Emacs = ESC-Meta-Alt-Ctrl-Shift")
08:24:32 Join J-23 [0] (
08:42:46 Quit BlakeJohnson86 (Read error: 110 (Connection timed out))
08:44:51 Join BlakeJohnson86 [0] (
08:50:10 Quit scorche (Nick collision from services.)
08:51:02 Join scorche [0] (i=Blah@rockbox/administrator/scorche)
08:54:56 Join Rob2223 [0] (
09:00:42 Nick BjoernErik is now known as Bjoern-Erik (
09:09:43 Quit pixelma2 (Nick collision from services.)
09:09:52 Join pixelma2_ [0] (n=marianne@rockbox/staff/pixelma)
09:09:54 Nick pixelma2_ is now known as pixelma2 (n=marianne@rockbox/staff/pixelma)
09:10:00 Quit amiconn (Nick collision from services.)
09:10:06 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
09:12:41 Quit Rob2222 (Read error: 110 (Connection timed out))
09:15:19 Quit BHSPitLappy (Remote closed the connection)
09:17:57 Quit pixelma2 ("-")
09:18:09 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
09:24:15 Join n1s [0] (n=nils@rockbox/developer/n1s)
09:33:58 Join nplus [0] (n=nplus@141.25.Globcom.Net)
09:38:22 Quit jhulst_ (Remote closed the connection)
09:45:47***Saving seen data "./dancer.seen"
09:49:14 Join MethoS [0] (
10:05:49 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
10:27:29 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
10:47:25 Join aaronwi [0] (n=noone@unaffiliated/aaronwi)
10:57:28aaronwiany more progress on the creative zen:M ? I know its been going on for a year now, lost track
10:58:16 Nick JdGordon|afk is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
10:58:19n1saaronwi: all info on in progress ports are (or at least should be) in their respective forum threads
11:09:39 Join culture [0] (
11:10:36 Join ender` [0] (
11:21:34 Quit cg (Read error: 104 (Connection reset by peer))
11:22:16 Quit tessarakt ("Client exiting")
11:24:43 Join cg [0] (
11:27:07 Quit JdGordon (Remote closed the connection)
11:29:24 Join JdGordon [0] (
11:34:21 Join {phoenix} [0] (
11:39:39 Join homielowe_ [0] (
11:39:39 Quit homielowe (Read error: 104 (Connection reset by peer))
11:42:57 Quit JdGordon (Read error: 104 (Connection reset by peer))
11:45:51***Saving seen data "./dancer.seen"
11:46:24 Join JdGordon [0] (
11:49:21n1shmm, the "generic_i2c" driver in firmware/drivers is not used anywhere and not even in SOURCES, and from looking at the history it seems it was never used, should it go away?
11:51:32 Quit Seed ("cu, Andre")
11:58:22 Join tyfoo [0] (
12:14:37 Quit aaronwi ("rm -rf / ...o crap")
12:20:03 Join freqmod_qu [0] (i=quassel@2001:700:300:1430:213:d3ff:fee9:5ed0)
12:35:49 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
12:44:45 Quit J-23 ("Flying cow pressed ^D on my keyboard.")
12:44:54 Join J-23 [0] (
12:55:28 Join Lear [0] (i=chatzill@rockbox/developer/lear)
13:00:01 Nick fxb__ is now known as fxb (
13:28:56 Join kugel [0] (n=chatzill@unaffiliated/kugel)
13:31:30 Quit J-23 ("Flying cow pressed ^D on my keyboard.")
13:31:52 Join reacocard [0] (n=reacocar@WL-112.CINE.HMC.Edu)
13:32:10 Join J-23 [0] (
13:40:27 Join Darksair [0] (n=user@
13:43:14 Join cg_ [0] (
13:44:54 Join MethoS- [0] (
13:45:28 Join moos [0] (
13:45:52***Saving seen data "./dancer.seen"
13:46:03 Join merbanan [0] (n=banan@
13:49:27 Quit MethoS (Read error: 60 (Operation timed out))
13:55:04 Quit reacocard (".")
13:57:38 Quit cg (Read error: 110 (Connection timed out))
14:02:43 Quit J-23 ("Flying cow pressed ^D on my keyboard.")
14:02:49 Join J-23 [0] (
14:03:05 Join Thundercloud [0] (
14:06:41 Join ap0 [0] (
14:07:41 Quit advcomp2019 (Read error: 60 (Operation timed out))
14:11:04 Join Schmogel [0] (
14:13:31 Quit cg_ (Read error: 60 (Operation timed out))
14:14:18 Join cg [0] (
14:28:11 Join gabe565 [0] (
14:31:05 Part pixelma
14:31:29 Join pixelma2 [0] (n=marianne@rockbox/staff/pixelma)
14:36:22 Join mrkiko [0] (
14:37:15 Quit cg (Read error: 110 (Connection timed out))
14:46:24 Quit gabe565 ("ChatZilla 0.9.83 [Firefox 3.0.4/2008102920]")
14:47:18 Quit JdGordon (Read error: 104 (Connection reset by peer))
14:47:39 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
14:49:33 Join reacocard [0] (n=reacocar@WL-112.CINE.HMC.Edu)
14:49:34*bluebrother found something regarding \settings{} in the manual
14:50:16 Quit _lifeless (Remote closed the connection)
14:52:25 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
14:52:54 Join lasser [0] (
14:56:27 Join J-23_ [0] (
14:56:39 Quit J-23_ (Remote closed the connection)
14:57:18 Join _lifeless [0] (n=lifeless@
14:59:31 Quit J-23 ("Flying cow pressed ^D on my keyboard.")
14:59:55 Join J-23 [0] (
15:04:01 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
15:04:07 Quit Thundercloud (Remote closed the connection)
15:27:58 Quit n1s ()
15:31:56 Quit reacocard (".")
15:41:29 Join webguest14 [0] (n=430aeeaf@gateway/web/cgi-irc/
15:42:51bluebrotherwhat do people think about FS #9554? I think it's as pointless as the "bug" it refers to
15:45:08JdGordonin the bug task he says the rockbox way makes more sense...
15:45:53***Saving seen data "./dancer.seen"
15:46:16bluebrotherso why does it deserve a patch at all?
15:46:26linuxstbbluebrother: Do you have an iaudio remote?
15:46:37bluebrotherlinuxstb: no.
15:46:54 Join MethoS-- [0] (
15:46:57 Quit webguest14 ("CGI:IRC (Ping timeout)")
15:47:09bluebrotherbut if he even tells the Rockbox way makes more sense than the OF one, why is there a point in modifying Rockbox to follow the OF mapping?
15:47:56JdGordonits the same guy... maybe he has split personality and one of him likes the of more and it drives him (more) crazy when his other half gets into rb?
15:48:02linuxstbHe doesn't say "makes more sense", he says "probably makes more logical sense"... I think an iaudio user/dev should judge.
15:48:03*JdGordon should probably goto bed
15:49:02 Nick JdGordon is now known as JdGordon|zzz (n=jonno@rockbox/developer/JdGordon)
15:49:38bluebrotherwell, "consistency is more important for a control that relies on muscle memory" sounds pretty much to me as it's simply a "I'm used to the other way".
15:49:59linuxstbIn which case, is there a reason for Rockbox to do it the opposite way?
15:50:25bluebrotherI guess so ;-)
15:50:43linuxstbWithout knowing what the remote looks like, I would imagine it's a pretty arbitrary choice.
15:51:34 Join Xerion [0] (
15:55:25bluebrotherfound an image:
15:57:49soapdoes any other target / remote have a wheel of such orientation?
16:00:06bluebrotherthe h100 lcd remote has a similar "wheel", though it's labelled << / Menu / >>
16:00:54 Quit MethoS- (Read error: 110 (Connection timed out))
16:00:59bluebrotherjust checked, that has >> go down in lists, << go up
16:01:04soapand does moving your thumb toward the top of the remote have the same keymapping as this remote?
16:01:23soapahh - so it's a side to side wheel, not a top to bottom wheel?
16:01:31soap(h100 LCD remote)
16:01:32bluebrotherthat is at the top of the remote.
16:02:34soapLooking at the pictures of the remote and at the patcher's description, the Rockbox mapping seems logical.
16:03:03soapI was just curious if it was consistent across targets with similar wheels (if any existed).
16:03:32bluebrotherthe leftmost one. It's the "wheel" thats on the top left (upper right on the picture)
16:04:14bluebrotheralso, from the m3 pictures I think the button in question is on the top edge too
16:04:29soapahh, that remote looks designed to be used rotated 90 degrees from the iAudio one.
16:05:15linuxstbComparing with the wheel targets, the iaudio behaviour seems more consistent - i.e. "+" moves forward, "-" moves back.
16:05:21 Quit bmbl ("Woah!")
16:05:28soapyes, looks like the M3 is on the side, thus the wheel goes up and down. I assume the iRiver one, though is a landscape oriented screen, thus the wheel is going side to side.
16:06:05bluebrotherchecking config-iaudiom3.h also tells that the lcd is in landscape mode, so the picture is rotated 90 degrees
16:06:19soapso westerners would assume the iRiver one goes "forward" when the wheel is moved to the right.
16:06:43bluebrotherif forwards maps to down on lists ;-)
16:07:25bluebrothercomparing with the ipods scroll wheel, the "east" part is up -> down in lists, the "west" part up -> up in list
16:07:55soapis the M3 remote is held with the menu wheel on the bottom of the unit, moving the wheel right = forward = menus should go down, IMHO.
16:08:11soapwhich is counter to how it is mapped now.
16:10:03soap*"IF the M3 remote is held..."
16:10:16 Join cg [0] (
16:10:38 Quit {phoenix} (Remote closed the connection)
16:10:50bluebrotherfrom reading the task + moves up in RB right now. Which is, compared with the wheel on scrollwheel targets, consistent −− moving to the right on the bottom part of the wheel moves up in lists
16:12:06bluebrotherwell, the picture looks to me as the remote is held that you can read the "HDD ..." text, and unless I get the picture wrong this means the Menu / vol button is at the bottom with vol+ being right
16:13:29bluebrotherin which case right -> up seems more logical to me than the other way round
16:13:32soap"moving to the right on the bottom part of the wheel moves up in lists" I totally agree that normally counter-clockwise rotation = up, but is that a valid consideration when your only interface with a wheel is upside down?
16:14:10soapis it really a clockwise / counterclockwise interface at that point or a left / right interface at that point?
16:14:39bluebrotherwell, at least I would assume this.
16:14:42soapI would argue that when you only have 60 degrees of exposed wheel it is now a left / right interface and not a wheel interface.
16:15:27bluebrotheron the h100 remote it is a "wheel", but you can't turn it completely. Still I would consider it as a wheel (or wheel-like) interface.
16:17:05 Join LinuxMafia [0] (
16:17:19LinuxMafiahi all
16:18:12LinuxMafiai formatted my sansa loader /rockbox loader by mistake , what should i do?
16:18:53bluebrotherwhat means "formatted loader"?
16:19:26LinuxMafiabluebrother, there is 20MB for sansa loaders right? i formatted it
16:20:03LinuxMafiabluebrother, now when i turn on my sansa nothing happens , just the circle light on
16:20:28bluebrotherthat's the recovery partition.
16:20:54bluebrotherthat's not the loader itself. check
16:21:07 Join obo [0] (n=obo@rockbox/developer/obo)
16:21:31LinuxMafiabluebrother, but laoder wont load now
16:23:45bluebrotherso what? You need e200tool to fix this.
16:23:51bluebrotherthe wiki page describes this.
16:24:23LinuxMafiabluebrother, thanks alot
16:24:32LinuxMafiacan i do that easily?
16:25:22LinuxMafiathanks alot
16:25:36bluebrotherwell, depends on your computer knowledge. Why don't you read the page first, then ask questions?
16:27:21linuxstbIsn't that just the firmware partition? IIUC, the recovery partition is just a ramdisk visible in recovery mode.
16:27:27LinuxMafiabluebrother, i got it
16:27:34LinuxMafiai have to do manufactore recovery
16:35:17 Join nelek [0] (
16:36:55nelekwhere can i see the status of proposed new rockbox functionality?
16:37:33neleki proposed a file system driver a while ago and it was added somewhere and i'd like to follow up on if something has happened with it or not from time to time
16:38:10linuxstbThere's not really anywhere like that.
16:42:34bluebrotherfile system driver? Doesn't sound to me like something that has an actual chance of getting worked on anyway
16:44:01LinuxMafiai can not mount the device
16:44:42bluebrotheryou don't need to mount it for e200tool
16:44:49LinuxMafiai hold down the center buttom , but this is all i get
16:44:55LinuxMafiausb 2-5.2: new full speed USB device using ohci_hcd and address 12
16:44:55LinuxMafiausb 2-5.2: configuration #1 chosen from 1 choice
16:45:11LinuxMafiai dont know which device it is
16:45:34linuxstbYou don't need to know - e200tool will find it.
16:45:47nelekbluebrother : why?
16:46:05nelekat the time they said it was a valid proposal
16:46:26linuxstbnelek: What kind of filesystem driver? And who were "they"?
16:46:56LinuxMafiawow thanks
16:47:02neleka hfs+ driver
16:47:24mrkikoNo guys - "they" are very hostile in developing another fs driver
16:48:14neleki know, i got a load bs first but it makes sense to have a hfs+ driver
16:48:56neleksince a lot of ipods are using it by default
16:50:24nelekand performance of FAT on a Macintosh system is terrible
16:50:49linuxstbnelek: The fact is that a HFS+ driver will be a _lot_ of work for someone, and no-one cares enough to do that work.
16:51:07linuxstbWhy don't you do the work?
16:51:17neleki can't program
16:51:22linuxstbYou can learn.
16:51:43neleki already have two jobs :(
16:52:13mrkikolinuxstb: I think I know a little about programming ... but I can easyli implement things with pc because I can read manuls. Reading rockbox source file actually is problematic for me
16:54:47linuxstbnelek: But seriously, I can't answer for everyone. But personally, FAT32 is "good enough" for me, and there are other things I would prefer to spend my limited Rockbox time on. We all do this for a hobby, so have our own priorities.
16:55:31neleki know, i can only ask for something that would improve the product and make a lot of mac users happy
16:56:35linuxstbThat's another issue - there are very mac users who contribute to Rockbox.
16:56:43linuxstb^I mean very few mac users
16:56:58LinuxMafiai dont know which rom file i have to copy
16:57:08neleksince the ipod is an apple product its strange that your product actually works better with pcs than with macs
16:57:32neleki could contribute, i just can't program
16:57:55linuxstbRockbox isn't a product - it's a hobby.
16:58:05domonokynelek: rockbox isnt a "product", and it runs on much more systems then ipods... so fat32 is the best choice (works everywhere)...
16:58:18linuxstbAnd it works on 28 different devices (only a few of which are ipods)...
16:59:27neleki know
16:59:34markunwhat would be the advantage of HFS+?
16:59:53nelekit works better with mac systems
17:00:13markuncan you give an example?
17:00:31nelekthe performance degradation on macs is considerable when you move your ipod to hfs+
17:00:37nelekeerm to fat32
17:00:49 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
17:01:14markuncopying is a lot slower you mean?
17:01:25linuxstbnelek: So do you not use Rockbox currently, or do you suffer with FAT32?
17:01:30bluebrotherwell, to give you a number: I wrote a rather basic FAT16 implementation about two years back. It doesn't deal with FAT32 or FAT12, it doesn't deal with long filenames etc.
17:01:47neleklinux i dont use rockbox atm because of it
17:01:50bluebrotherit took me a complete work to get the basics done. HFS is much more complicated.
17:02:20nelekmarkun yes
17:02:20bluebrotherhaving HFS support is of gain for a small number of users.
17:02:38bluebrotherbut the work needed is quite noticably.
17:03:15bluebrothercomplete week, not work :)
17:03:17nelekwell, it improves your product considerably with the interaction of two products which are of the same company
17:03:26bluebrothernelek: Rockbox is NOT a product
17:03:32bluebrotherface it, if you like it or not.
17:03:38domonokyfat32 is just the logical choice for a mobile device, everything understands fat32, even some stereo/ car audio system.
17:03:45nelekhow do you mean not a product
17:03:57domonokyif mac is bad with fat32 complain to apple :-)
17:04:06bluebrotherit is not a product. It's a project. This is a vital difference
17:04:14nelekno, fat32 is a crippled FS
17:04:16markunwe're not trying to sell anything
17:04:26nelekbut microsoft isn't fixing it
17:04:40nelekits a free product...
17:04:40bluebrotherFAT32 is a rather easy file system. But it works quite reliably.
17:04:48bluebrotherit is NOT a product.
17:04:48markunI think it's a good thing it's not being changed
17:05:02domonokynelek: wrong, fat32 is just a very simple FS, which is why it works everywhere.. its not crippled (apples fat32 driver maybe cripplet).. :-)
17:05:16nelekits too limited
17:05:35neleki cant even store an image backup on a file system like that
17:05:39markunnelek: but don't get us wrong, if someone writes a HFS+ driver we wouldn't be unhappy about it
17:05:41LinuxMafia i get this message
17:05:45bluebrotherbesides, FAT is an old filesystem. There is nothing one could call crippled because it doesn't support newer features someone might want to have.
17:05:50neleki don't have metadata, ...
17:05:53LinuxMafiaload main image failed
17:05:53markunI just don't think it would be in the default rockbox builds
17:05:55*domonoky mentions that rockbox doesnt even need any users... we would be fine to use rockbox just ourselfs :-)
17:06:05LinuxMafiaswitch to recovery mode
17:06:06bluebrotherand why should MS "fix" it? Its working, and for the ones who want more features they developed NTFS.
17:06:59nelekbecause they can
17:07:01LinuxMafiabluebrother, any idea?
17:07:05bluebrotherno they can't.
17:07:07 Quit faemir (Read error: 104 (Connection reset by peer))
17:07:12markunthere isn't another FS which works everywhere and is easy to implement, is there?
17:07:13bluebrother"Fixing" FAT would mean breaking it.
17:07:50bluebrotherso, apart from breaking on old, working file system they could implement a more sophisticated one. Which they did. It's called NTFS.
17:08:08nelekapple fixed hfs+
17:08:22nelekthey had a 2 gb limit, they overcame it
17:08:26*domonoky also doesnt understand what in fat32 would need fixing.. big file support ?
17:08:28nelekthey added journaling
17:08:46bluebrotherso what? I don't care about apple. They have HFS on their machines only, so they don't need to worry about compatiblilty
17:08:51domonokyand journaling on a flash disc would be bad....
17:08:56bluebrotherMS can't do that because FAT is ... almost *everywhere*
17:09:31bluebrotherwhich means they would break almost everything that uses / used FAT.
17:09:45bluebrotherso they're much better off developing a new filesystem.
17:09:47nelekthe point is they extended their existing code with backward compatiblity, as filesystem extensions instead of a complete rewrite
17:09:53bluebrotherwhich, as I said several times, they did.
17:10:10domonokyand the 2gb/4gb limit in fat32 is by design, you would need a fat64 for it :-)
17:10:11bluebrotherNTFS is a complete different filesystem compared to FAT
17:10:35 Quit Bensawsome ("The awsome is gone :(")
17:10:38markun... which we also wouldn't support
17:10:44LinuxMafiawhat .rom file should i use
17:10:53nelekNTFS support doesnt make sense
17:11:04nelekits not used on any mp3 player
17:11:29 Join Bensawsome [0] (n=Bensawso@unaffiliated/bensawsome)
17:11:37bluebrotherand FAT64?
17:11:51nelekdoes that even exist?
17:11:51LinuxMafiaLinuxMafia k # ls /mnt/cd
17:11:59domonokynelek: lets stop this discussion here, if someone wont to do the neccessay work,fine. but "we" wont do it.. .
17:12:07LinuxMafiait is all i have in recovery partion , it formated
17:12:10nelekthe ipod is only one of the 26 players but its by far the most popular player around
17:12:33bluebrotherwell, this discussion is pointless. MS definitely had reasons for not "fixing" FAT but rather developing NTFS.
17:12:42markunnelek: it's not the most popular rockbox target among us devs..
17:12:48bluebrotherthe "popular" Ipods aren't supported in the newer versions anyway.
17:13:26linuxstbnelek: Please.... There's nothing you can say to convince the people here to spend months of their spare time implementing HFS+
17:13:39bluebrotherbesides, Rockbox is about fun, not about "crippled filesystems".
17:13:56nelekyou guys are so typically dev
17:14:04markunthanks :)
17:14:18nelekalways finding excuses why your product doesnt work.
17:14:23markunwe're not listening to our 'customers'? :)
17:14:23bluebrothernelek: feel free to go and do a basic FAT implementation yourself. Then you'll understand why FAT is a reasonable filesystem and why it's not "crippled".
17:14:33bluebrotherrockbox is NOT A PRODUCT!
17:14:50bluebrotherwell, nelek isn't a customer anyway, even if we had a product ...
17:15:05domonokynelek: again, rockbox is not a product, its just our hobby... we are just so nice to share it with you....
17:15:13nelekits a product of your labor, weither you like it or not. its available for end users so its a product
17:15:33bluebrothermarkun: we're not listening to our non-customers and non-users ... I kinda like it this way :P
17:15:45bluebrothernelek: no. Product is something else.
17:15:56nelekdoesnt matter if its free or of charge. someone produces it and someone consumes it. the definition of a product
17:15:59domonokyas we dont need users, we also dont need to listen to them... :-)
17:16:05bluebrothersomething being available doesn't make it a product. Period.
17:16:18linuxstbnelek: Where does that definition come from?
17:16:37markunnelek: we have a feature request section in the forums. Maybe you have more luck getting people convinced there.
17:16:50bluebrothernelek: so you are a product? Someone (your parents) "produced" you and now someone (like your employer) is "consuming" you "product"?
17:17:04bluebrotherthis definition is plain stupid
17:17:26neleki'm a product of my parents love, true
17:17:28nelekyou too
17:17:42bluebrotherso you get consumed?
17:18:01 Join sin613 [0] (
17:18:12nelekthis is entirely besides the discussion
17:18:26bluebrotherwell, it's your definition ...
17:19:08nelekbut its ok, you made it clear that hfs+ is not going to happen
17:19:23nelekyou could have told me the first time, then i wouldn't have bothered to come back
17:19:26*domonoky defines that rockbox is a hobby,not a product. we just do what we want :-)
17:20:17nelekhave fun with your product
17:20:22 Part nelek ("Infinite Loop")
17:20:50bluebrotherinfinite loop? In being a product?
17:21:02linuxstbSo, anyone want to test this hfs+ patch I've been working on?
17:23:24mrkikofat is simply tupid and inconsistent filesystem, aniway - true: coding a filesystem is very hard
17:27:26LinuxMafiaBlakeJohnson86, i get error : it cannot load the .rom file
17:28:27LinuxMafiabluebrother, *
17:29:20 Quit obo ("bye")
17:29:47LinuxMafianow it sais wwritinng boot rom
17:30:49LinuxMafiathis helped "don't know if it will, but I think it might. If it won't, renaming it to sdbootrom.bin might help."
17:31:05 Quit MethoS-- (Remote closed the connection)
17:33:32 Join saratoga [0] (n=41becb3b@gateway/web/cgi-irc/
17:37:47 Join obo [0] (n=obo@rockbox/developer/obo)
17:41:11LinuxMafiahow long it takes?
17:45:06 Join LambdaCalculus37 [0] (
17:45:56***Saving seen data "./dancer.seen"
17:51:18 Nick fxb is now known as fxb__ (
17:57:45mrkikoI re-read the nelek discussion...
17:58:09mrkikocrazy... :D
17:59:48 Quit Lear ("ChatZilla 0.9.83 [Firefox 3.1b2pre/20081114034305]")
17:59:48 Quit obo ("bye")
18:02:16 Quit saratoga ("CGI:IRC")
18:02:29 Join Seed [0] (
18:06:55 Quit Horscht ("I am root. If you see me laughing, you better have a backup")
18:08:02Nico_Pwow crazy indeed. I think I would've told him to leave us alone and go bug apple about FAT32 performance
18:10:05 Join obo [0] (n=obo@rockbox/developer/obo)
18:11:01ender`people have a very hard time understanding that (at least with free software) something won't happen just because they wish for it
18:12:19ender`(or worse, demand it)
18:15:42 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
18:24:12 Quit Llorean (Read error: 54 (Connection reset by peer))
18:27:31 Join aarcane [0] (
18:28:18 Join dany_21a_ [0] (
18:32:38 Join miepchen^schlaf [0] (
18:39:48 Join Llorean [0] (
18:45:09 Join waran [0] (
18:47:08 Part waran
18:47:09 Join fdinel [0] (
18:47:16 Quit bmbl ("Woah!")
18:48:55LambdaCalculus37On the H100 and H300 pads, what is the BUTTON_ definition for the A-B button?
18:50:16domonokysee firmware\target\coldfire\iriver\button-target.h
18:50:29domonokyit seems to be BUTTON_MODE
18:52:01LambdaCalculus37domonoky: Thanks. :)
18:54:01 Join DerDome [0] (
18:58:34 Quit esthar ("KVIrc 3.4.0 Virgo")
19:00:47LambdaCalculus37domonoky: Thanks for pointing me to button-target.h; I'll take it from there.
19:03:53*LambdaCalculus37 is a little puzzled as to a logical button layout for Rock the Wumpus on the H10
19:04:03 Quit Darksair ("People who are zhuangbility want to show their niubility but only reflect their shability.")
19:04:22LambdaCalculus37Any H10 owners have a good suggestion for a button layout?
19:05:15 Quit mrkiko (Read error: 60 (Operation timed out))
19:07:46 Join mrkiko [0] (
19:11:43 Part pireau
19:18:19 Join pabs_ [0] (
19:18:41 Quit pabs (Read error: 110 (Connection timed out))
19:18:47 Nick pabs_ is now known as pabs (
19:20:29 Join Thundercloud [0] (
19:23:44 Join verbal [0] (
19:25:17 Quit homielowe_ ()
19:27:44*LambdaCalculus37 works on his Wumpus plugin
19:27:47 Nick blkjawk[work] is now known as blkhawk (
19:28:24 Nick blkhawk is now known as blkawk (
19:28:30 Nick blkawk is now known as blkhawk (
19:38:52 Quit sin613 (Read error: 110 (Connection timed out))
19:44:24 Join toffe82 [0] (
19:45:52 Join sin6131 [0] (
19:46:01***Saving seen data "./dancer.seen"
19:47:24 Join FOAD_ [0] (
19:51:43 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
19:55:41 Quit kugel (Read error: 110 (Connection timed out))
19:56:51 Quit advcomp2019 (Read error: 104 (Connection reset by peer))
19:57:25 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
19:59:05 Join kugel_ [0] (
19:59:05 Nick kugel_ is now known as kugel (
19:59:11 Quit FOAD (Read error: 110 (Connection timed out))
19:59:11 Nick FOAD_ is now known as FOAD (
19:59:32kugeldomonoky: did you read the logs?
19:59:52domonokyno.. maybe i should :-)
20:01:10kugeldomonoky: yep, i left you a message
20:02:44domonokykugel: two remarks: did you remove the CCU_IO in both the bootloader an main binary ? ( doesnt work for me). And indeed i have more changes.. (funmans attempts to make sd-driver more stable, but i dont know if that matters)...
20:03:06 Quit verbal (Read error: 110 (Connection timed out))
20:03:16kugeldomonoky: yep I did it for both
20:03:36kugelwell, I guess that highly matters
20:03:57domonokyi could perhaps give you patch of my changes...
20:04:30kugelor, commit it. are the changes strongly believed to be wrong or why do you keep it local?
20:05:22domonokyone part are funmans changes.. which i dont really know... and the CCU_IO i am not sure about other targets..
20:07:47kugeldomonoky: well then, thanks
20:08:01domonokyi often get checksum invalid with this, but when it works i can browse files... and bootup is much faster...
20:08:49kugelcould someone commit the of offset for the fuze at least?
20:09:29kugelit's the same as for e200v2, so the line would be #if defined(SANSA_E200V2) || defined(SANSA_FUZE)
20:09:53LambdaCalculus37Very, very early work on my Wumpus plugin (and proof that I've still got a lot to learn ;) ):
20:10:10LambdaCalculus37I only have keymaps laid out right now.
20:11:14 Join krazykit [0] (
20:11:24kugelwhat's wumpus if I may ask?
20:11:51LambdaCalculus37kugel: Old computer game from the "stone age" days.
20:14:54kugeldomonoky: can you commit that?
20:15:13kugeldomonoky: that also makes sure the offset is only added for the internal memory
20:18:52domonokykugel: why put the if outside of the ifdef ?
20:20:06kugeldomonoky: why not? there may be future targets with another offset and microsd slot
20:20:30kugeldrive is declared no matter of HAVE_MULTIVOLUME
20:20:51domonokyah, oki... will commit it.
20:21:20kugeldomonoky: it's "const int drive = 0" if HAVE_MULTIVOLUME isn't defined
20:25:51kugeldomonoky: thanks
20:26:02kugeldomonoky: still no main menu
20:26:51domonokymake at least 10 tries to be sure :-)
20:27:11kugelalso, I think we should make sd_init NOT error out if there's no microsd inserted ;;
20:27:44domonokyand wait a long time after bootup.. my m200v4 took a long time to boot before my fix :-)
20:27:48kugelthat's the infamous ata error -2
20:28:57domonokyyou could make a patch that solves this.. :-)
20:29:14kugelnot sure what the correct fix would be
20:29:34kugelI'd just remove the "return ret" after the init of the external card
20:29:48kugelso it would still be initialized, but not error out if it fails
20:30:21 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
20:32:01domonokyhm, the proper fix would probably be to implement "card_detect_traget()" and/or "card_enable_monitoring_target()" ...
20:32:44kugeldomonoky: that'd be my quick fix
20:32:47domonokyso it detects if there is a external card. then it should try to init the non-existing card..
20:33:50 Quit ap0 ("Baļ")
20:33:55domonokyah, yes, that wll of course need fixing.. :-)
20:34:25kugeldomonoky: well, sd_init_card returns very early if there's no card found. Using this function directly would probably create less "overhead" then adding a monitoring function
20:34:30 Quit sin6131 (Read error: 110 (Connection timed out))
20:35:52domonokyremoving the return, is a good quick fix. But will need proper handling later (with respect to hot-swap)...
20:36:25kugelyep, sure
20:37:39kugeldomonoky: I get data aborts and a garbaged screen when trying to boot rockbox
20:38:34kugeland I still don't know why my screen is shifted left by some pixels
20:38:38kugelthat's probably related
20:38:46domonokyand thats with which changes ?
20:39:20kugeldomonoky: the patch you gave me, the one patch you committed and the quick fix about external sd init
20:39:48kugelit happens with and without microsd, so it's unlikely related to the latter one
20:39:52domonokytry it with the CCU_IO enabled (and change both bootloader and rockbox)...
20:40:20domonokyand make a few tries, the driver is still very unstable...
20:40:21kugeldata abort at 0x3005c520
20:40:29kugelI did aout 10 tries
20:40:56kugeldomonoky: lol success
20:41:14domonokythe 11th try was successfull ?
20:41:17kugelon the 11th try
20:41:44kugelbut it crashed immediately after seeing the main menu
20:42:10domonokywhat crash ?
20:42:36domonokybut crashes are to be expected with a unstable mass-storage driver..
20:42:58kugelyea, I don't complain :)
20:42:59domonokysome crashes are also related to not havig write acess, i think....
20:43:41kugeldomonoky: btw: it said "no .rockbox directory, installation incomplete"
20:44:33domonokyalso related to the driver.. i think :-)
20:44:50kugelmost likely
20:46:15domonokycan you load rockbox from the sd-card ?
20:46:31 Quit jhulst (Read error: 113 (No route to host))
20:46:42 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
20:46:42domonokywould be interessting to know if the external sd still works without the CCU_IO...
20:47:29bluebrothergrrr. Package soul helps with the small caps problem in the manual, but it conflicts with hyperref :(
20:47:29 Quit Nico_P (Read error: 104 (Connection reset by peer))
20:48:27 Quit linuxstb (Read error: 60 (Operation timed out))
20:49:13kugeldomonoky: yes, it's possible
20:49:36 Join AndyIL [0] (i=AndyI@
20:49:55kugelwhich reminds of another think that should be fixed in svn
20:50:02kugelthing rather
20:52:54domonokykugel: provide patches,and i will take a look :-)
21:00:08kugeldomonoky: ok let me test it fist
21:00:22 Quit AndyI (Read error: 110 (Connection timed out))
21:03:00 Quit GodEater_ (Remote closed the connection)
21:03:30 Quit Thundercloud (Read error: 54 (Connection reset by peer))
21:05:08bluebrotherpixelma2: around?
21:07:52kugeldomonoky: card_detect_target() isn't implemented, is it?
21:08:16domonokykugel: take a look at the code...
21:08:54kugeldomonoky: I see a stub, but I get undefined reference compiling error
21:09:10 Part dany_21a_
21:10:58 Quit linuxstb_ (Read error: 104 (Connection reset by peer))
21:12:05kugeladding a prototype doesn't help?
21:13:31 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
21:14:10kugelah got it
21:14:18domonokytake care about the #ifdefs...
21:18:28 Quit tyfoo ("Carpe diem")
21:28:17 Quit Bawitdaba (Read error: 104 (Connection reset by peer))
21:32:42 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
21:34:53 Quit krazykit ("Connection reset by beer")
21:35:41 Join ajonat [0] (n=ajonat@
21:37:21kugeldomonoky: no, I can't boot from microsd anymore
21:37:36kugelit did work though without all these changes
21:37:57domonokyoki, so we need the CCU_IO thing on v2 with sd-cards.
21:38:05 Quit miepchen^schlaf (Connection timed out)
21:38:15kugeldomonoky: it didn't work with CCU_IO either
21:38:46domonokydoes it work with svn ?
21:38:54kugeldomonoky: btw, that's the issue I talked about:
21:39:26kugelbootloader doesn't use BOOTDIR at all (which is defined in the target config). it's even more hardcoded to /.rockbox
21:39:43kugelI think linuxstb and Bagder agreed with me here
21:40:51bluebrotherthy not BOOTDIR "%s", firmware?
21:41:10bluebrotherinstead of "%s/%s"
21:41:13kugelbluebrother: does that make any difference?
21:41:44bluebrotherwell, it's only one replacement −− BOOTDIR is a const string anyway
21:41:59bluebrotheror is it a pointer to some const data?
21:41:59kugelso no. but I don't really care
21:42:06kugelit's a #define
21:43:01bluebrotherwell, it does make a difference: in the first case you have a string "%s/%s" plus the value of BOOTDIR. In the second you have the string BOOTDIR"/%s". Which is 2 characters less, plus you have only one replacement.
21:43:09bluebrotherso while the difference is small there is one.
21:44:04kugelit's not a difference which actually matters though
21:44:17kugelbut as I said, I don't about which of the two
21:44:39bluebrotherdepending on your constraints it can matter ...
21:44:49bluebrotherit might not matter in this case.
21:46:02***Saving seen data "./dancer.seen"
21:47:53 Join homielowe [0] (
21:49:53kugeldomonoky: would you commit that quick fix about the external sd init?
21:52:15bluebrothergrrr. Why isn't the settings searching in the pdf working anymore?
21:53:39 Quit toffe82 (Read error: 110 (Connection timed out))
21:54:58 Quit nima (Read error: 110 (Connection timed out))
21:58:02 Join bertrik [0] (
22:13:25domonokykugel: could you update the DeviceDetection page with the USB-ID of the fuze ?
22:24:48 Quit J-23 (Nick collision from services.)
22:26:35 Join J-23_ [0] (
22:27:25kugeldomonoky: done
22:27:43 Quit J-23_ (Nick collision from services.)
22:29:11 Join rhinobanga [0] (n=57c2a3b2@gateway/web/cgi-irc/
22:29:15 Join J-23 [0] (
22:30:10rhinobangaevening folks ... any developers about? I'd like to discuss a change I've made to the rockbox code.
22:30:56kugelrhinobanga: just begin to talk, eventually someone will reply
22:31:30 Join tessarakt [0] (
22:31:44rhinobangaI have changed the code to dynamically create album art
22:32:09bluebrotherdynamically _create_?
22:32:23rhinobangaif the cover*.bmp file is not there and folder.jpg exists it will create cover*bmp automatically
22:32:39bluebrotherhow that? Using a default bitmap?
22:32:50kugelrhinobanga: as in renaming or actually converting to bmp?
22:32:53bluebrotherah, by converting that?
22:33:08rhinobangayes, read the jpg and create the bitmap
22:33:20bluebrotherdoes it scale or only convert?
22:33:30bluebrotherand is this done as plugin or in the core?
22:33:56bluebrotheryou know there is this resizing patch that's being worked on? It's in the tracker
22:34:00*kugel suggest rhinobanga to put it on the tracker so everyone can have a look at
22:34:25rhinobangaI remember you folks were talking about refactoring the whole metadata subsystem
22:34:41bluebrotheralso, if you are decoding jpg in the core, why write a bmp file?
22:34:45rhinobangaso I wanted to know if anyone would be interested in this
22:34:53bluebrotherwhy not put it into the metadata buffer directly?
22:35:38rhinobangabecause it was the easiest/quickest way to get it up and running
22:35:58rhinobangaby creating the cover_X_Y.bmp file dynamically I didn't really have to do much
22:36:34rhinobangaI know there are loads of memory issues however I am only interested in getting my iPod working
22:36:45bluebrotherthat's of course a way but I don't think such an attempt will make it into the core at all
22:36:55bluebrotheryou ipod isn't working without album art?
22:37:16kugelrhinobanga: the best way to share the code and get opinions is to put it on the tracker or mail at the mailing list, irc is not really proper for that (in general)
22:37:30rhinobangaI only have folder.jpg and I don't want to pollute my main music repository with all these *.bmp's
22:38:00kugelwith bmp resize you only need 1 bitmap
22:38:14bluebrotherkugel: well, I fear this will be a rather big blob nobody is willing to completely look at. It does multiple things at once ...
22:38:30LambdaCalculus37None of which sound very cleanly implemented.
22:39:02rhinobangagiven that every wps can request different sizes I thought it would be easier/quicker doing it this way without having to change to much
22:39:17bluebrotherrhinobanga: if you're interested to get the feature in svn I'd rather suggest trying to help out with the resize patch, or have a look at refactoring the jpg decoding so it can get integrated into the core
22:39:21rhinobangaand the creation of the .bmp is only done when it needs to be
22:39:45LambdaCalculus37rhinobanga: That implementation sounds like using a sledgehammer to crack a nut.
22:39:45*kugel agrees with bluebrother
22:40:26bluebrothercreating an intermediate bmp sounds like the wrong solution to me, at least from a Rockbox point of view (of course it's perfectly legal if you "just" want to get it "done" for yourself)
22:40:29rhinobangamaybe ... but it works and it means no preprocessing of existing artwork has to be done
22:40:53bluebrotherwell, but this kinda rules it out for getting accepted.
22:40:53rhinobangathere is nothing worse than getting a wps with a different artwork resolution
22:41:12rhinobangathen having to go throu and regenerate it
22:41:20bluebrotherthere's a really simple solution for that: don't change themes ;-)
22:41:34bluebrotheryou could do that with a plugin.
22:41:35rhinobangaespecially when you have a filesystem containg over 1000+ albums
22:41:54rhinobangalol ... true ... but I like to experiment ... I am a programmer after all :D
22:42:17rhinobangahow far off is the "official" solution to this?
22:42:22LambdaCalculus37I still think it would be much easier to just add JPEG decoding to the core, along with album art resize.
22:42:41LambdaCalculus37But there are patches in the tracker with some work done.
22:42:51kugelrhinobanga: you're completely ignoring the bmp resize patch
22:43:00bluebrotherwell, IIRC there is general consensus that we want scaling in the core and jpg decoding
22:43:16bluebrotherkugel: how do you know he isn't incorporating it? ;-)
22:43:30rhinobangawhy have a bmp and folder.jpg in the same directory?
22:43:45kugelwell, you can delete the folder.jpg of course
22:43:46bluebrotherwhy not? It doesn't hurt
22:43:52rhinobangapersonally I like clean folders in my main repositories
22:44:02rhinobangasingle file ... single solution
22:44:16LambdaCalculus37Then here's a solution: keep the JPEG files for your album art in another folder, away from your music.
22:44:42rhinobangaexcept all the windows stuff I do won't find it
22:44:50kugelrhinobanga: you can also put the album art into /.rockbox/albumart so you don't have any bmp/jpg in your music folder
22:45:11LambdaCalculus37Although that will require you to rename everything to match album titles.
22:45:13rhinobangamissing the regeneration point
22:45:29bluebrotheryou could pregenerate for all possible sizes ;-)
22:45:42bluebrotherbut I agree, this is somewhat missing the initial point
22:46:12bluebrotherstill, there are different problems that should get addressed separately instead of simply pushing stuff into the core
22:46:23kugelrhinobanga: I see better chances if you help out with bmp resize or get jpeg viewer into the core
22:46:34kugel(and get the wps show jpgs)
22:46:34*bluebrother wonders what's the binsize delta of that patch
22:46:54 Join n1s [0] (n=nils@rockbox/developer/n1s)
22:47:02bluebrotherI see _no_ chance of such an attempt getting into svn at all.
22:47:09rhinobangaas I said the solution is great for me, all I have to do is upload my album folder containg my .M3U, .JPG and .MP3s and let rockbox handle any generation it needs
22:47:19rhinobangaI wasn't intending it to
22:47:33bluebrotherwell, that's a valid point.
22:47:42kugelrhinobanga: expect decreased battery life
22:48:03rhinobangawhy? it only generates the bmp onoce as needed
22:48:03kugelrhinobanga: again, put it on the tracker and you'll get some more opinions
22:48:20bluebrotherrhinobanga: (a) disc spinup, (b) decreased audio buffer size
22:48:20rhinobangathen when the artwork is requested again the .bmp file is there
22:48:42rhinobangaboth of which are a once off hit
22:49:10kugelstill, the ram used for audio buffer is decreased
22:49:25LambdaCalculus37This is especially bad for lower mem targets.
22:49:32kugelwhether your algorithm is doing something or not
22:49:32rhinobangaonly for generating the .bmp
22:49:33bluebrotherwell, you're free to do it this way if you like. But it's not the way it will get done once Rockbox has this functionality
22:49:43bluebrotherso I seem to miss the point why you want to discuss this ...
22:50:00bluebrotheryou need decoding buffers for the jpg handling
22:50:24rhinobangaagain a once off hit, then from then on the .BMP file is sitting there
22:50:44kugelrhinobanga: rockbox doesn't use dynamic memory management. The RAM you need for your stuff is "wasted", no matter if you already have bmps
22:50:45bluebrotherno, as the album art is read to the audio buffer
22:51:13*domonoky mentions that his m200v4 shows 214k free for the audio buffer :-)
22:51:28 Join robin0800 [0] (
22:51:29bluebrotherwhich you can't use for decoding the jpg while music is playing ...
22:51:53rhinobangamy iPod has 64MB of memory
22:53:29rhinobangaas I said, I just wanted to know how far off any official solution was to this problem before I posted anything on the forum
22:53:48bluebrotherwell, from my view you're completely off
22:53:53rhinobangait sounds like it's going to be a while yet
22:54:14LambdaCalculus37It's when Mr Someone decides to do it.
22:54:40 Quit lasser ("ChatZilla 0.9.83 [Iceweasel 3.0.3/2008092816]")
22:55:08bluebrotherfeel free to jump in helping out with the resize patch or integrating jpeg properly into the core
22:56:32 Join {phoenix} [0] (
22:56:42rhinobangaI'll maybe have a look, I was just sitting at my desk yesterday and got fed up with RB not automatically utilising my folder.jpg so I thought I'd do something about it.
22:58:10 Join Zagor [242] (
23:00:05Zagorrhinobanga: for what its' worth: I applaud your initiative, even if it is not the way we want it done in Rockbox. any way you slice it, your solution sure beats whining about it!
23:02:09 Join webguest91 [0] (n=430aeeaf@gateway/web/cgi-irc/
23:02:53 Quit webguest91 (Client Quit)
23:03:18rhinobangasure does ... defo one of the perks of open source :D
23:04:12rhinobangaI'll post a msg on the iPod forum sometime this week and see if anyone else is interested in this
23:04:38 Join miepchen^schlaf [0] (
23:05:30 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
23:06:17 Join webguest20 [0] (n=3ea72868@gateway/web/cgi-irc/
23:07:00saratogaspeaking of resizing, what is needed for the bmp resize patch to go in?
23:08:16saratogait looks like most of the comments have been addressed, and the patch has been refactored into smaller units for each function
23:08:43 Join obo_ [0] (n=obo@rockbox/developer/obo)
23:10:18 Quit obo (Read error: 110 (Connection timed out))
23:10:23ZagorUnhelpful: here? which target did you build, which failed?
23:11:16 Quit webguest20 ("CGI:IRC (Ping timeout)")
23:11:29UnhelpfulZagor: gigabeat s. it worked fine until i tried your patch on top of a clean source repo, and then it missed those generated files
23:11:52Unhelpfulmight it be a good idea to put everything generated in build dir?
23:12:06Zagorit is
23:12:28Zagoror do you mean the builddir root?
23:12:57Unhelpfuli mean in the builddir, and not in the source dir. currently, a number of things under tools are built in tools.
23:13:25Zagorah yes, tools are special. that
23:13:34Zagorthat's because they don't change for different targets
23:15:03Unhelpfuli can see why they're there... and git clean can make them go away ;)
23:15:37Unhelpfulwhich is where i got into trouble, i initially didn't realize that iaudio_bl_flash.{c,h} were generated, and thought that my repo was corrupted in some way :/
23:16:12linuxstbsaratoga: Like any patch, it needs a committer to take responsiblity for it, and start committing the patches one by one...
23:18:19*LambdaCalculus37 goes to look at the patch
23:18:23saratogalinuxstb: but all of your initial concerns have been addressed?
23:18:30*domonoky wonders how rockbox uses i2s, i only see i2_reset() functions...
23:18:47bertrikwhat is this "rockbox" you speak of?
23:19:34ZagorUnhelpful: yes, it is not a build system for novices :-) but a simple rule is that all files in your builddir are generated. so when it compiles a file that is in your builddir, that file was generated from something else. such as all bitmaps.
23:19:59 Quit rhinobanga ("CGI:IRC")
23:22:27Unhelpfulthe "old" makefile generates iaudo_bl_flash.* in tools. the unified one does not work as-is for me unless those are already present. probably i should change the patch i posted to put those files in builddir instead of just copying what the non-unified build does?
23:23:18Zagoroh sorry I misread. yes that is a bug I have to fix.
23:23:44Unhelpfulat least the generated sysfont.h was already going in the right place ;)
23:23:47 Quit bertrik ("Leaving")
23:23:49kugellinuxstb: and you don't want to be that committer?
23:24:43linuxstbsaratoga: I don't know. I've had a look at the 01 patch, and am not sure about the change from a 32-bit to 24-bit type for a pixel. I haven't looked at anything else.
23:24:45linuxstbkugel: No
23:25:46ZagorUnhelpful: you don't have to post a new patch if you don't want to. I'll look at it.
23:25:48Unhelpfulmaybe it would make sense to have a commonbuild dir for putting objects built from tools? it's the only way i can think of to share them between builds without polluting the src, but i think it would end up being just as ugly
23:26:15saratogalinuxstb: ok then, I've slowly been reading through how this patch works
23:26:50saratogaabout the ram use comment you made, are you just readnig the "RAM usage:" line out of rockboxinfo.txt?
23:27:04saratogai assume theres more because the increase there is quite small
23:29:27UnhelpfulZagor: should iaudio_bl_flash.* stay in tools, actually? it would seem putting them there is justified by the same argument that we put .o files and binaries there
23:29:40Unhelpfulin which case i don't need to do anything to the patch i posted
23:31:24Zagoryes, it makes little sense putting generated files in different places just because they have different names. I say we either fix a "pretty" way of doing it (keeping tools entirely clean) or we put all generated files side-by-side with the .o files.
23:31:43Zagorfor now, I'll settle with the latter
23:32:41kugellinuxstb: btw: the original bmp resize added less binsize (don't know about ram though)
23:32:44*LambdaCalculus37 wonders whether to draw the Wumpus graphics using external bitmaps, or to draw using C code
23:34:26LambdaCalculus37Meh... I'm kinda hungry now and can't really think anymore about coding.
23:34:31ZagorLambdaCalculus37: what does it look like? any screenshots of other versions?
23:34:37Unhelpfulthen the only thing i might change in what i posted would be to make the sysfont.h dep a little cleaner... how do the generated dependencies work? could it possibly be included in those?
23:34:53LambdaCalculus37Zagor: Here's what I'm aiming for:
23:35:25ZagorUnhelpful: yes, it can. it is supposed to be handled already, I'll look into why it isn't.
23:35:31kugelLambdaCalculus37: showing bitmaps is boring
23:36:04LambdaCalculus37kugel: Rockblox appears to be generating its graphics using C code.
23:36:31LambdaCalculus37Might make it easier for generating the map and the Wumpus mouth closing on you when you die.
23:36:37ZagorI can imagine some people wanting to prettify it on the more bling-bling targets. but hey, it's your plugin!
23:37:17LloreanI'd say generally speaking bitmaps are the way to go.
23:37:30LambdaCalculus37Zagor: I'll study invadarox and rockblox more closely, then... those two plugins appear to generate all of their graphics internally.
23:37:52*Unhelpful could swear that when he played hunt the wumpus "back in the day" that much of the point was that you didn't know which way things connected.
23:38:16LambdaCalculus37Llorean: Bitmaps may be easy for stuff like the bloodstains, the Wumpus (on the map), and the player, but generating a random map using bitmaps may be harder.
23:38:24amiconnUsing bitmaps should be faster than doing lots of partial drawing
23:38:25LloreanLambdaCalculus37: I'd say it won't be.
23:38:51LloreanYou only have four "arcs" for connecting rooms.
23:38:53ZagorLambdaCalculus37: random maps are just a random collection of bitmaps, with some rules to the randomness
23:38:53linuxstbsaratoga: Yes, my comments were based on looking at the RAM usage in rockbox-info.txt.
23:39:11amiconnRockblox graphics are simple and btw rockblox only draws the blocks. The background is a bitmap
23:40:08LloreanI think using bitmaps leaves you more room to differentiate graphics quality based on what the target can do, without sacrificing code readability.
23:40:30amiconnThat too
23:40:33LambdaCalculus37Llorean, amiconn, Zagor: Looking on it now, I'm thinking that bitmaps for the curves in four directions, and each "room" will work more.
23:40:37saratogalinuxstb: i must be doing somethign wrong since the patch is only giving me a ~9KB delta
23:40:46LambdaCalculus37Very good advice. :)
23:40:58 Quit amiconn (" /me would prefer updates not requiring reboots....")
23:41:31LambdaCalculus37In the meantime, I'm heading off. See everyone later!
23:41:48 Quit LambdaCalculus37 ("Ka-chunka")
23:42:43linuxstbsaratoga: Is that with all 4 patches applied?
23:42:54saratogalinuxstb: yes, on the sansa if that matters
23:43:33 Join dirwiz [0] (
23:43:41linuxstbThen maybe it's been improved since the first version - e.g. I think he changed the large bitmap buffers to 24-bit, rather than 32-bit.
23:44:15saratogathat seems to good to be true, but maybe
23:44:29 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
23:44:49soapCould the buffers be shrunk even more on 16bit colour targets?
23:46:04***Saving seen data "./dancer.seen"
23:48:06saratogaah ok redid it and got 20KB which seems more reasonable
23:48:24 Join toffe82 [0] (
23:49:58 Quit n1s ()
23:50:36saratogaif we ever get JPEG support, we could shrink the memory costs quite a lot by making resize jpeg only, and using the JPEG DCT trick for most of the resize
23:51:01kugelis it still max 500x500 or uses it another algorithm for higher bitmaps? I'd agree with linuxstb that it should be at least 512x512 since people tend to have bitmaps with 501x503 and the like
23:51:47saratogathe whole area sampling thing is not needed for jpeg
23:52:21amiconnWhy is there a limit at all?
23:52:50amiconnI'd like to see an algorithm that never loads more than 1..2 pixel lines at once, and scales on the fly
23:52:52soapwhat will the negative delta be of removing the current album art code which looks for cover.100x100.bmp and such? Very small I'd imagine.
23:53:11saratogaamiconn: with jpeg + bilinear thats possible with good results
23:53:43saratogawith pure bmp you can't use the DCT in the jpeg decoder to do antialaising for you, so you need to load more lines to get a good image
23:53:45amiconnBesides, 500x500 would not even cover the whole m:robe 500's screen...
23:54:10saratogafrom the comments it sounds like the current patch does not have a size limit
23:54:43Unhelpfulrequiring conversion to jpeg would probably be reasonable (for things that are bmp/png/whatnot already)
23:54:46saratogathough i do'nt understand the resize well at all
23:55:28saratogain terms of binsize, either requiring small bmp or jpeg of arbitrary size would be ideal i think
23:55:48saratogai could be wrong but it seems like its that resize by a factor > 2 thats painful
23:55:58kugelsaratoga, amiconn: if I got it right, then only the area sampling is limited and it will use "Nearest Neighbour or bilinear algorithm"
23:56:13kugelon higher bitmaps
23:56:21saratogaif its a factor of less then 2, I think area sampling is bilinear
23:56:38saratogasince then you only need 2 cache lines and much less RAM
23:56:58Unhelpfulsaratoga: a big enough jpeg might still force you to use larger sampling factors, if the size is more than... i'd think two macroblocks to target pixel?
23:57:14saratogai think area sampling is just basically a lowpass filter, with the number of cachelines being the order of the filter
23:58:48amiconnImho bilinear interpolation is sufficient

Previous day | Next day