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 2005-10-02

00:06:55 Join matsl [0] (
00:28:51 Quit XavierGr ()
00:29:05 Quit matsl (Remote closed the connection)
00:38:16 Quit Moos ("Glory to Rockbox")
00:41:38 Join DangerousDan [0] (
00:52:05 Join muesli- [0] (
01:09:58 Quit ender` (Read error: 110 (Connection timed out))
01:10:28 Quit Colddy (Read error: 104 (Connection reset by peer))
01:14:45***Saving seen data "./dancer.seen"
01:16:28 Quit t0mas (Read error: 104 (Connection reset by peer))
01:16:55 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
01:21:31 Quit muesli- (Read error: 110 (Connection timed out))
01:32:46 Join BoD[] [0] (
01:32:51BoD[]Hello world !
01:33:33 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
01:34:15BoD[]today I bought an ipod
01:34:48BoD[]hoping that one day rockbox will run on that ;)
01:35:18linuxstbYou are obviously a gambler. Which iPod did you buy?
01:36:05BoD[]:) the "photo" 20gig
01:36:38BoD[]i *almost* got the 60gigs but they didn't have it and I didn't want to wait, (and it was very more expensive)
01:38:23 Nick BirdFish[alt] is now known as BirdFish (n=bradbox8@
01:40:46BoD[]I first bought an iriver h340
01:41:22BirdFishWhat happened to it?
01:41:49BoD[]but didn't like it because no touchpad or wheel, and usb host didn't work (with my sony), and can't read the screen when it's not lit
01:41:56BoD[]oh and a bit "thick" too
01:41:59BoD[]so I returned it
01:42:11BoD[]then I got a mpio hd300 40gig
01:42:23BoD[]its firmware is ridiculous
01:42:26BirdFishIt's a good thing that you can't read the screen when it's not lit
01:42:32BirdFishThat feature saves power ;)
01:42:40BirdFishAnd size is always better compared to ipod :D
01:42:50BirdFishYou would have been much better off
01:43:17BoD[]it's thicker than the ipod i got (but it's 20gigs so I guess it's normal)
01:43:39BirdFishDid you check into the IAudio X5?
01:43:43BoD[]the mpio 40gig was thick too
01:43:54BirdFishnote: you might still have time to return the ipod ;) ;)
01:43:56BoD[]iaudio x5? hmm no is it good ?
01:44:30fuzzie20:13 <linuxstb__> Not really. I've been busy on the iPod.
01:44:32BirdFisheCost has them for dirt cheap!
01:44:37fuzzie^- much hope for rockbox on the iPod
01:44:39BoD[]yes but does rockbox run on it ? ;)
01:44:46BirdFishNot currently
01:44:56BirdFishBut the manufactured firmware is good just the same
01:45:08BirdFishAnd a port for the X5 has been opened ;)
01:45:20BoD[]can you randomize everything ? ;) (with the mpio : you can't)
01:45:45BirdFishYou can set the screen's brightness
01:45:49BirdFishChange the background images
01:45:59BoD[]uh oh what's that joypad thinggy ? ;)
01:46:05BirdFishIt plays FLAC with manufactured firmware! :D :D :D
01:46:16BirdFishIt's professional, that's what it is ;)
01:46:53BoD[]is it really practical to browse 200 artists fast ? like a wheel or touchpad is
01:47:01BirdFishIt's the only player in the world right now (other than the deprecated Karma) that comes preloaded ready to play FLAC
01:47:12BirdFishNot sure about that
01:47:21BirdFishMine's on it's way
01:47:22 Quit t0mas (Read error: 104 (Connection reset by peer))
01:47:23BirdFishI can't wait
01:47:40BoD[]ok you'll tell me :)
01:47:44linuxstbThe iPod had ALAC, but that's about the only good thing I can say about Apple's firmware.
01:47:50linuxstb(has ALAC)
01:48:15BoD[]but anyway the story is
01:48:16BirdFishThe one thing that it doesn't have currently but will have in about a week or two is a browse function by ID3 tag
01:49:38BoD[]the story is I tried 2 models, they sucked, I had to return them and that's very anoying, so I just opted for a model I know works ok.
01:49:41fuzziethe advantage of the iPod seems to be the wheel rather than the firmware
01:50:20BoD[]well :)
01:50:24BirdFishBoD[]: I had the same thought a couple times ;)
01:50:26BoD[]the mpio has a touchpad thing
01:50:27BirdFishBut I've held out
01:50:27fuzziewhich is why it'd be nice to see rockbox on it.
01:50:41 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
01:50:42fuzziei wish i knew enough to help linuxstb out :)
01:50:43BoD[]but the firmware is so ridiculous that the thing is unusable
01:50:44BirdFishipodlinux is leet too
01:50:52BirdFishno offense rockbox
01:50:58linuxstbfuzzie: Just buy an ipod and learn...
01:50:58BirdFishI just happen to get off to linux :P
01:51:21fuzzielinuxstb: what model(s) would i need, in order to use whatever you're working on?
01:51:41linuxstbA new color model - I have a 60GB color.
01:51:52BoD[]hey that's what I have ;)
01:51:58BoD[]maybe I could help ;) yeah right
01:52:03BirdFishfuzzie: you've got enough money to randomly buy a few models just to work on firmware?
01:52:07BirdFishSweet ;)
01:52:13fuzziei have enough money to buy *one*
01:52:24fuzziewell, actually, in reality i have enough money to buy a whole heap
01:52:29fuzziebut that money is kinda earmarked 'savings'
01:53:02BirdFishfuzzie: if you don't mind my asking, what is your profession? I mean, how did you come about learning to program firmware?
01:53:10fuzziei don't :)
01:53:16BirdFishI'm getting ready to decide my major in college and was just wondering
01:53:16fuzziehence why i was saying above, i wish i knew enough
01:53:27BirdFishI thought you meant about the model ;)
01:53:33linuxstbThere's not a great deal of difference between the models - so it doesn't really matter which you buy. As long as it's one of the newer PP5020 based models, and as long as it isn't the Nano (at the moment).
01:53:39fuzziei have done so embedded hardware work
01:53:44fuzzieerr, some
01:53:46fuzzielinuxstb: would a Mini work?
01:54:16linuxstbYes - anything supported by iPodLinux -
01:54:16BirdFishThe nano is cool in regards to the size factor
01:54:37fuzzielinuxstb: thanks
01:55:24BirdFishBut I'm still going to rebel against anything Ipod for a bit
01:55:32linuxstbfuzzie: I assume you know about the LCD problems with the new colour ipods?
01:55:57BirdFishWhat about them linuxstb
01:56:14BirdFishI haven't heard about them
01:56:40linuxstbBirdFish: Color ipods sold in the last couple of months have a very slightly different colour LCD - so the ipodlinux LCD driver doesn't work with them.
01:57:07linuxstbBut I'm expecting (hoping) the ipodlinux guys to find the solution soon.
01:57:40BirdFishthe color version isn't completely supported yet anyways, is it?
01:58:08fuzzieBirdFish: but, well, if you want to do firmware work, either CS or CS/EE would be more than good enough skills-wise
01:58:12linuxstbSee the link I just posted. All the important hardware has fully working drivers
01:58:36linuxstbIt's more than enough for a usable Rockbox port.
01:59:01BirdFishfuzzie: I'm thinking about CS&E
02:01:41dpassen1i'm Computer Engineering
02:01:52BirdFishLike it?
02:01:56dpassen1so far
02:02:10dpassen1not yet, but u
02:02:19dpassen1*I'm still towards the beginning
02:02:35dpassen1the pre-requisites were the tough part
02:02:50dpassen1Calc 1,2,3 Linear Algebra, Diff Eq, Physics 1 and 2, Chem 1 and 2, etc.
02:02:57fuzziea lot more difficult than CS, imo
02:04:03dpassen1its funny watching people drop as i advance
02:04:15dpassen1many switched to Information Systems
02:05:52fuzzieick. :)
02:06:11dpassen1they used to abbreviate it as IFSM - I Failed Science and Math
02:06:16BirdFishI didn't really like what CE involved. I'm more into the software stuff.
02:07:00fuzziewell, any decent CS course ought to give you a chance at some embedded work
02:07:54 Join darkless [0] (
02:08:00dpassen1i hope to be able to contribute to rockbox in time
02:08:30darklessdoes anyone have a link to the latest dircache patch by Slasheri?
02:10:59dpassen1i dont know well enough to say which is recommended
02:11:18 Quit t0mas (Connection timed out)
02:15:11darklessthanks a lot, dpassen1. The second revision is much faster, because it uses fat_opendir instead of opendir :-)
02:15:30dpassen1glad to help
02:15:48fuzzieheh, libmad on ipodlinux isn't realtime?
02:17:24 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
02:17:58linuxstbfuzzie: Most things are not realtime on ipodlinux.
02:18:52linuxstbThat's why I think there's a need for Rockbox.
02:19:16fuzzieLinux overhead?
02:19:26fuzzieI must admit my first reaction was 'that's going to suck overhead-wise'
02:19:34fuzzieBut then I realised, I don't know if that's really true..
02:19:49 Join ashridah [0] (
02:20:55linuxstbAt the moment, they are only using one of the two ARM processors. They probably haven't done as much low-level optimisation as we've done with the codecs either.
02:21:25linuxstbI don't know how well they are using the IRAM for example.
02:26:31 Quit dpassen1 ()
02:30:09MipsIrvanyone here have experience with armboot?
02:31:16 Part MipsIrv
03:09:44preglowlinuxstb: remember how much iram the portalplayer cores have got?
03:14:48***Saving seen data "./dancer.seen"
03:23:08 Quit preglow ("leaving")
03:26:13 Quit BoD[] ("TCL 4EVA")
03:26:26 Quit paugh ("Leaving")
03:31:25 Quit sangre (Read error: 104 (Connection reset by peer))
03:57:56 Quit RotAtoR ()
04:05:55 Join QT [0] (i=as@madwifi/users/area51)
04:16:16 Quit QT_ (Read error: 110 (Connection timed out))
04:24:21 Quit darkless ("Chatzilla [Firefox 1.0.7/20050915]")
04:46:47 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
05:00:00 Join pheon [0] (
05:04:15 Join JoeBorn [0] (
05:04:28 Join Paul_The_Nerd [0] (
05:14:51***Saving seen data "./dancer.seen"
06:29:27 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
07:00:08 Join amiconn_ [0] (
07:00:08 Quit Mxm`Pas`Bien (Read error: 104 (Connection reset by peer))
07:00:13 Join Maxime [0] (
07:01:23 Join Lynx0 [0] (
07:09:36 Quit ze (Read error: 104 (Connection reset by peer))
07:10:26 Join ze [0] (
07:14:53***Saving seen data "./dancer.seen"
07:17:18 Quit gromit` (Remote closed the connection)
07:18:21 Quit amiconn (Read error: 110 (Connection timed out))
07:18:21 Nick amiconn_ is now known as amiconn (
07:19:57 Quit Lynx_ (Read error: 110 (Connection timed out))
07:19:58 Nick Lynx0 is now known as Lynx_ (
07:57:09 Join Gibbed [0] (
07:57:09 Quit Rick (Nick collision from services.)
07:57:19 Nick Gibbed is now known as Rick (
07:59:30 Join Gibbed [0] (
07:59:32 Quit Rick (Nick collision from services.)
07:59:42 Nick Gibbed is now known as Rick (
08:02:24 Join Gibbed [0] (
08:02:26 Quit Rick (Nick collision from services.)
08:02:36 Nick Gibbed is now known as Rick (
08:33:58 Join webguest90 [0] (
08:34:12 Quit webguest90 (Client Quit)
08:42:34 Quit ashridah (Read error: 110 (Connection timed out))
08:43:07 Join ashridah [0] (
08:44:40 Join Vlad0man [0] (
09:02:59 Quit Vladoman (Read error: 110 (Connection timed out))
09:14:54***Saving seen data "./dancer.seen"
09:21:22 Quit webguest95 ("CGI:IRC (Ping timeout)")
09:40:02amiconnBagder, Zagor or LinusN?
09:40:37amiconnSomething is wrong with the IRC logs. Current log works, but older logs don't show up on the irc page
09:40:58amiconn(currently missing are 20050930 and 20051001)
09:52:48Zagorhmm, strange
10:01:23 Join ender` [0] (i=ychat@
10:19:52markunender`, how did the unicode build work for you?
10:20:12ender`fine, however the japanese font doesn't work - i only get garbage
10:21:06markunStrange. It used to work.
10:21:32markunAh, did I give you the 18x18? That doesn't work no.. I have a smaller japanese font that does work.
10:22:03markuncurrently fonts can't be bigger than 16 pixels.
10:22:18ender`the filename says 12x13ja
10:27:56markunI get garbage too..
10:54:42linuxstbMorning all. Can someone explain the stride parameter to lcd_mono_bitmap_part() ?
10:57:06amiconnThat's simple. 'stride' says how many bytes to advance for each row. This is necessary when drawing partial bitmaps
10:57:36linuxstbSo it refers to the input bitmap?
10:58:18amiconnSee how lcd_mono_bitmap() calls lcd_mono_bitmap_part()
10:58:59amiconnAll *_*_bitmap_part() functions have a stride parameter
11:00:09amiconnBeware that stride means the number of bytes, not the number of pixels
11:00:56amiconnThe numbers are identical for the core mono bitmap format of all current targets, because of the 'orientation' of a byte in a monochrome bitmap
11:02:13amiconnI'd suggest to keep this orientation for >=8 bit displays, as there is no need for a specific byte orientation in mono bitmaps, and it's the most common mono bitmap format in rockbox
11:02:46amiconnThe only graphics engine with a different mono bitmap format so far is playergfx
11:03:02linuxstbYes, I've got no intention of changing the mono bitmap format.
11:07:50amiconnThe graphics routines for >=8 bit displays will be rather trivial compared to the current ones
11:08:20amiconnlcd_draw_bitmap_part() will essential be a sequence of memcpy()
11:09:01linuxstbYes, an exact number of bytes per pixel obviously simplifies things.
11:10:23 Join _FireFly_ [0] (
11:12:36 Join linuxstb_ [0] (
11:14:58***Saving seen data "./dancer.seen"
11:25:11Slasherihehe, with dircache creating the root playlist takes only a few seconds :)
11:25:16 Join linuxstb__ [0] (
11:25:58Slasherinow updating the code not to spin up disk when modifying the playlist
11:27:08 Quit linuxstb (Read error: 113 (No route to host))
11:29:44 Part amiconn
11:31:37 Join Moos [0] (
11:31:52 Join Paul_The_Nerd [0] (
11:33:34 Join linuxstb [0] (
11:40:13 Quit Maxime (Read error: 104 (Connection reset by peer))
11:41:21markunSlasheri: I noticed something strange with the pcmbuf_beep.
11:41:55 Quit linuxstb_ (Read error: 110 (Connection timed out))
11:42:00markunIt works when I skip during the loading of the first part of the first track, but when that is loaded it doesn't beep anymore.
11:42:15markunWhen the rest of the buffer is loaded the beep works again.
11:44:09Slasheriah, that is probably caused because of a too short pcm buffer. We need to write a fix for this
11:44:27 Join Maxime [0] (
11:46:29 Quit linuxstb__ (Read error: 110 (Connection timed out))
12:03:33 Join guillaume [0] (
12:08:12solexxis anybody here familiar with the wps parsing code?
12:10:23 Join arkascha [0] (
12:10:44 Join webguest52 [0] (
12:11:26arkaschaSince upgrade to 2.5 my iriver H120 refuses to play any oggs with a 'codec failure'. This worked in 2.4. Any ideas anyone?
12:12:02markunarkascha: Are you building from cvs?
12:12:21arkaschano, release cause I thought that the new 2.5 holds all the stuff for iriver...
12:13:10solexxarkascha: rockbox was not released for iriver yet
12:13:16markunarkascha: there has not been a release for iriver. Use the daily build.
12:13:42arkaschaI see. I'll use a daily snapshot again. Thanx :-)
12:15:35_FireFly_the latest daily-build is from 20050930
12:16:04_FireFly_why doesn't exist a daily-build from 20051001 and 20051002 ??
12:20:44arkaschaok, oggs working again after building from source, thanx for the tip
12:21:07arkaschaanother question: I thought about creating a plugin for vcard files. Has anyone thought about this yet?
12:24:46arkaschasorry update: the oggs DON'T work with a daily. The track starts but stops after 2 secs with a codec failure...
12:25:20_FireFly_can you use ogginfo ??
12:26:10arkaschathe oggs are ok, I can play them on my pc and with the iriver firmware and the older 2.4, but I'll check
12:26:24_FireFly_i had the same failure
12:26:38markunarkascha: last time I got codec failures I had to rm everything in the build dir, then do configure and make again. Maybe it helps.
12:26:41_FireFly_ogginfo said that there were a hole in the file
12:27:17_FireFly_after retagging it the failure was gone.. i think the failure cames, when id3tags are used in ogg files
12:28:10 Join einhirn [0] (
12:28:58arkaschaogginfo shows nothing special except for the usual 'no upper and lower bitrate set'
12:30:02_FireFly_then test the tip from markun
12:30:03guillaumearkascha: it's weird, i have the latest daily builf for h120 and oggs work well, but i've had a problem with one mp3
12:31:16arkaschanope, fresh build shows codec failure as well, hmmm
12:31:56ghode|afkarkascha: did you overwrite the files or delete the rbx dir first?
12:32:25_FireFly_what files did you copy from your build ??
12:32:47arkaschaI removed everything from my build dir and copied the rockbox.iriver
12:32:52_FireFly_mybe you must also copy the *.codec files into .rockbox/codecs/ on the iriver
12:32:53linuxstbarkascha: Did you do a "make zip" and then uncompress the zip to your iriver?
12:32:56arkaschaok, maybe I have to copy anything else?
12:33:08ghode|afkyes ;p
12:33:08linuxstbarkascha: Yes - do a "make zip".
12:33:15arkaschano I did not zip, it's a single file
12:33:34ghode|afkyou need to replace the whole dir + .iriver file when updating
12:33:38_FireFly_try make zip and copy the files in the zip onto you iriver
12:34:15linuxstbUnder Linux, I simply do "make zip" followed by "unzip -d /iriver/" where /iriver/ is the mount point for my H140.
12:34:30arkaschaahem, which dir does have to be replaced? includes your rockbox.iriver, plus the files in the .rockbox directory.
12:35:31arkascha?? the build dir contains all sort of files I *never* copied...
12:35:35_FireFly_what about the daily-builds ?? why it doesn't gave an daily-build from yesterday ??
12:35:40_FireFly_no arkasha
12:35:49_FireFly_type make zip in your build dir
12:36:00linuxstb_FireFly_: Not sure, but other strange things are happening on the server - such as missing IRC logs.
12:36:12_FireFly_and copy the files in the created zip-file to your iriver
12:36:52_FireFly_this zip-file includes the rockbox.iriver and all needed files which resists in the .rockbox dir on the iriver
12:38:16arkaschaok, looks like oggs work again now, thanx. But I never copied the complete build dir to my iriver. Does it say so in the compile instructions?
12:38:33ghode|afkif it doesn't, it should
12:38:48_FireFly_the problem is, that the rockbox.iriver file doesn't includes the codecs :)
12:39:05arkaschasure I see that point :-)
12:39:08_FireFly_the codec are seperate file which are in .rockbox/codecs
12:40:45arkaschaok, so once again... and comment about the vcard idea?
12:41:53 Quit guillaume (Remote closed the connection)
12:43:48 Join preglow [0] (
12:44:25solexxfrom a user's pov: i wouldn't use (if you think of it as some kind of address book)
12:44:38solexxI handle such stuff with my mobile phone
12:45:02solexxthe mein reason would be the lack of a decent input mechanism (keyboard)
12:45:41solexxand I carry my mobile phone more often with me than my iriver
12:46:08arkaschaok, but I hate mobiles and have a collection of vcds from my address book I used to import as text files to the iriver. that works but files are ugly to read, so I thought about beautifying their look...
12:46:25arkaschawe could stick to the standard format this way
12:46:46arkaschaI mean instead of translating all addresses to native formats for mobiles
12:46:56solexxif you need it: i say do it
12:53:35 Quit t0mas (Read error: 104 (Connection reset by peer))
13:05:09 Join amiconn [0] (
13:05:14 Quit Paul_The_Nerd (Read error: 104 (Connection reset by peer))
13:05:54solexxhi amiconn
13:06:17 Quit webguest52 ("CGI:IRC (EOF)")
13:06:24solexxare you familiar with how the wps is parsed?
13:06:31solexx(the cfg file)
13:07:24solexxpaul_the_nerd and I talked about it yesterday
13:07:34amiconnno, not really
13:07:35solexxand we found some reproducable bugs
13:08:30 Join webguest95 [0] (
13:15:00***Saving seen data "./dancer.seen"
13:15:36 Quit Maxime ()
13:19:04linuxstbamiconn: Am I right in saying that lcd_mono_bitmap_part() isn't currently used in Rockbox? I can't find it using grep.
13:19:17amiconnIt is used lcd_mono_bitmap()
13:20:36linuxstbBut it sounds like you want to use it for the WPS quite soon?
13:20:43amiconn...and it would be used in wps if my idea gets implemented (clip regions from a larger bitmap)
13:21:36_FireFly_a simple solution might be when additional to the tag the coords and the size of the bitmap part will be provided
13:21:56_FireFly_in the wps
13:23:22 Join Maxime [0] (
13:23:24amiconnHowever, I'd like to find a way that shortens the wps code
13:23:48amiconnI mean the wps definition language code
13:24:36_FireFly_hmm we could extends the %xl tag so it can optional have the coords and size of the part of an bitmap
13:25:47_FireFly_or n additional tag which defines a kind of virual additional bitmaps for the bitmap in a combined bitmap
13:26:24_FireFly_so %x loades the bitmap complete and the additonal tag defines only the parts
13:26:48amiconn...perhaps where a left-out coordinate means 'use the last value'
13:27:41amiconnThat should help keeping the .wps file small, because often either the x or the y coordinate are identical, as the individual images are below each other, or side by side
13:27:59 Quit preglow ("leaving")
13:28:16amiconn..and height and/or width are also the same for same-type images (like multiple battery levels etc)
13:28:33_FireFly_you are right
13:30:41_FireFly_an example might be: %xl|a|<bitmap-file>|0|0| %xp|a|b|<x>|<y>|<width>|<height>
13:30:59_FireFly_can be the coords in the %xl tag left off ??
13:31:46amiconnWould mean the same as %xp|c|a|0|8|20|8
13:32:34amiconnThe destination coords should be part of the display tag
13:33:32amiconnAs it is now, I would need to load the same .bmp twice if I want to display it in 2 different places
13:34:37_FireFly_this extension means that the IMG struct has an additonal var which indicates if this image is image itself or a part of an image
13:35:48 Quit arkascha (Read error: 104 (Connection reset by peer))
13:35:57_FireFly_and the ptr var has only the pointer to the original combined bitmap
13:36:19_FireFly_if the var (e.g. bool bpart) is true
13:37:40amiconnIt doesn't need a bool to indicate whether it's a part.
13:38:03amiconnIt just needs the pointer to the bitmap, and the x/y/w/h clip region variables
13:38:25amiconnThey would equal 0/0/width/height/ for a full image
13:38:48amiconn...set by the .bmp loader
13:39:27amiconn%xp would just fill another struct without loading a new .bmp, but instead copying the aliased struct contents and replace x/y/w/h
13:39:30_FireFly_but then we have to change the call lcd_mono_bitmap so it also have an x,y parameter
13:39:40_FireFly_or use lcd_mono_bitmap_part
13:39:53amiconnThat's what lcd_mono_bitmap_part() is for
13:40:20linuxstband lcd_mono_bitmap() is already just a wrapper around lcd_mono_bitmap_part()
13:40:27_FireFly_yo i know
13:40:44_FireFly_so lcd_mono_bitmap would be obsolete
13:40:49amiconnlinuxstb: Yes, in order to save some code space as less parameters need to be passed
13:41:04amiconnlcd_mono_bitmap() is called quite often
13:41:09_FireFly_because we have to pass always the x and y coords
13:41:29amiconn_FireFly_: Only in the wps. Many other places will still use lcd_mono_bitmap()
13:42:06_FireFly_hmm thats true
13:42:38amiconnHmm, in fact we need to store one additional value, and that's the stride parameter
13:47:41_FireFly_is the stride paramter a kind of offset ??
14:11:20_FireFly_i will try to create a plugin for testing combined-images
14:25:06 Join arkascha [0] (
14:26:12 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
14:26:13 Join Lost-ash [0] (
14:35:10 Quit arkascha (Remote closed the connection)
14:44:49 Join muesli- [0] (
14:46:59 Quit ashridah (Read error: 110 (Connection timed out))
14:47:49 Join |Lupin| [0] (
14:47:53|Lupin|Hello, folks.
14:48:32|Lupin|I was just wondering: is there any prefered version of gcc / binutils to build RockBox for iRiver players ?
14:49:13|Lupin|Debian comes with facilities to build crossed versions for binutils-2.15 and gcc-3.4.3. Is this ok ?
14:50:27_FireFly_look at
14:50:36linuxstbI think gcc-3.4.3 should be fine, but binutils-2.16 is recommended.
14:51:27|Lupin|_FireFly_: linuxstb: thans to both of you...
14:52:18|Lupin|hmm I have absolu!ely no idea how to build a crossed version of binutils-2.16...
14:52:19_FireFly_is there an howto how to compile/make a plugin for rockbox ?? i become some link failures
14:53:15_FireFly_read the whole site
14:53:23_FireFly_a page i mean
14:57:02|Lupin|Yesthere is one, in the documentation section.
14:57:08linuxstb|Lupin|: Just follow the instructions at the link _FireFly_ gave you. You simply download the source, untar it, and then do the normal configure, make, make install (but with special parameters to the configure).
14:57:26|Lupin|_FireFly_: You can also download the sources from CVS (or tarballs) and have a look to docs/README
14:58:06|Lupin|linuxstb: Is it really riskyto useThe binutils-2.15 Debian provides ?
14:58:09 Quit muesli- (Read error: 104 (Connection reset by peer))
14:58:51_FireFly_i got it
14:59:11linuxstb|Lupin|: I don't think it's "risky" - just try it and see what happens.
14:59:38_FireFly_i had called some functions(e.g. splash) without using the api-pointer
14:59:46_FireFly_to call the functions
15:00:52|Lupin|linuxstb: What I meant with risky was: Do you think that binutils-2.15 insteadof 2.16 can damage definitely a player ?
15:02:20amiconnMost likely 2.15 won't work
15:02:27amiconn(at build time)
15:02:58|Lupin|amiconn: ok, thanks.
15:03:05amiconnIirc, binutils 2.15 don't know about the emac unit, so assembling rockbox code will fail
15:04:44|Lupin|amiconn: Just out of curiosity, may IAsk whatthe emac unit is, please ?
15:06:29amiconnemac = enhanced multiply-accumulate unit, a unit supporting dsp-like operations
15:07:43|Lupin|ahah, ok, thanks.
15:10:17linuxstbHas anyone else noticed that running the X11 sim under Linux breaks the X11 keyboard repeat - i.e. after running the sim, keyboard repeat is off for every X application.
15:10:30linuxstbOr is it just my installation...
15:11:29_FireFly_i have noticed it also
15:11:40linuxstbI have to type "xset r on" after running the sim to turn repeat back on.
15:11:59amiconnlinuxstb: It shouldn't. If it does, something within X11 is broken
15:12:01 Join Paul_The_Nerd [0] (
15:12:44amiconnThe sim needs to disable keyboard repeat for the button simulation to work properly, but only does so when its window is active.
15:13:13amiconnIt should reenable keyboard repeat on exit and when its window is deactivated
15:13:27|Lupin|Would it be difficult To write a text-mode UI simulator, guys ?
15:15:03***Saving seen data "./dancer.seen"
15:15:27 Nick Lost-ash is now known as ashridah (
15:15:32linuxstbamiconn: I've just found the problem - it segfaults on exit, and doesn't get as far as calling the restore function.
15:15:39linuxstbI didn't notice that before.
15:16:21amiconnSo there's something to fix...
15:17:31*amiconn summons LInusN
15:17:38 Join Febs [0] (
15:17:56linuxstbIt seems to be something to do with voice UI.
15:18:04linuxstb(this is an iriver sim)
15:18:13 Join XavierGr [0] (
15:19:37 Join Raxus [0] (
15:20:35XavierGrRaxus hello if you want to PM you have to register your nick
15:21:32RaxusXavierGr...xaderfos edw
15:21:56Raxuskala...den epitrepontai ta PM edw?
15:22:20XavierGrspeak to english
15:23:07XavierGrDo you know how to register?
15:23:07RaxusSo i have to register before u send me the file?
15:24:25amiconnXavierGr: You can allow privmsgs from unregistered users
15:24:30RaxusYes I see your PM
15:25:27amiconn. /msg nickserv set unfiltered on
15:25:35amiconn(w/o the dot)
15:26:13RaxusXavierGr...I hadd DCC Ignore for .exe files...I just disabled it, try to send it again plz
15:26:14amiconnThis does only work if you're registered and identified yourself, and is a permanent setting
15:26:24amiconn(until you reset it again)
15:28:34 Join webguest93 [0] (
15:29:08webguest93in the chat logs recently I've seen references to root.m3u - is that produced automatically?
15:30:01solexxwebguest93: playlist options -> create playlist
15:30:31*solexx still wants a menu entry which creates an empty playlist
15:30:45webguest93ahh right - I know you can create a recursive playlist.. I just wasn't sure if "root.m3u" was a new feature
15:31:02amiconnshweet :)
15:31:19amiconnNow I have my iriver running at 45/124 MHz instead of 48/120
15:31:37amiconn(preparation for a better timer handling at cpu frequency changes)
15:36:40webguest93playlists are just standard m3u playlists right? It would be kinda neat if you could associated extra settings with them like....
15:36:53webguest93I want this playlist to start shuffled
15:37:08webguest93I want the playlist re-shuffled everytime I load it
15:37:11webguest93and so on
15:38:40Paul_The_NerdYeah, I was reading mention of ideas like that in a thread at MR.
15:40:46webguest93I don't actually know anything about the format of an m3U file - does it have comments info like that could be stored in?
15:41:08Paul_The_NerdWell, if you were gonna do that, you may as well just define a custom rockbox playlist format anyway.
15:41:23webguest93that I think would get a lot of opposition
15:41:30webguest93and probably rightly so
15:41:50webguest93people crete playlists with outher software and copy them over
15:42:06Paul_The_NerdBut you don't have to remove .m3u support
15:42:25Paul_The_NerdYou can just have an "improved" m3u-based format.
15:42:53Paul_The_NerdWhich would essentially be an m3u with a line at the start containing playback mode information, and a different extension on the file.
15:44:01webguest93true... but then if you start with a standard one and change it to include extra info - you end up with 2 copies old and new format - not a biggy i know but irritating
15:44:55Paul_The_NerdWell, you can delete files on the fly
15:45:10Paul_The_NerdOnce the user's updated the playlist, they can just delete the .m3u version.
15:46:12 Quit ashridah ("Leaving")
15:46:13webguest93yep - like I said - not a big issue
15:47:34webguest93it seems that m3u "extended format" files contain extra info - but there's no space for custom info
15:47:52webguest93ah well just an idea - and a low priority one.
15:47:58webguest93goota go
15:48:02 Quit webguest93 ("CGI:IRC")
15:49:13Paul_The_NerdHrm... extended m3u looked kinda pointless to me. =/
15:51:53 Join gromit` [0] (n=gromit`
15:52:30 Part amiconn
15:56:59_FireFly_ok my plugin works :)
16:01:04Paul_The_NerdFor the purpose of?
16:01:04XavierGrwhat plug-in?
16:01:29_FireFly_a test plugin for combined bitmaps
16:01:51_FireFly_it can be found + source here
16:02:26_FireFly_i have used a combined bmp from the iriver firmware
16:04:51Paul_The_NerdY'know... with combined bitmaps you could also pretty easily add support for animations then.
16:05:32_FireFly_thats up to you how to use the posibility of using combined images ;)
16:06:24_FireFly_this is only a test it, amicon and i was thinking how to extend wps to display parts of combined bitmaps
16:07:20Paul_The_NerdAll that *really* does is increase the number of images you can use by allowing you to combine multiple similar ones into one file though, practically speaking right?
16:08:03_FireFly_yepp and reduce the size of the wps
16:08:37Paul_The_NerdStill, a very cool feature.
16:08:50_FireFly_so that you have only one bitmap to load and than you create(over a seperate wps-tag) the parts
16:14:22_FireFly_the number of images or parts of images will be restricted to 52 because valid image id will be at the moment only a-z or A-Z
16:14:48 Quit Raxus ()
16:15:53Paul_The_NerdWouldn't it make sense to have that number be 52^2
16:16:02Paul_The_NerdYou have 52 images, with 52 parts each?
16:16:45_FireFly_for this the id must be extended to two chars
16:16:47Paul_The_Nerd%xdaa through %xdaZ whith those all being parts of a
16:17:06_FireFly_and i don't know if this will be accepted
16:18:10 Join muesli- [0] (
16:18:45Paul_The_NerdWell, what's the benefit if parting then?
16:19:35_FireFly_mainly to reduce the size of the wps
16:19:36 Join DangerousDan [0] (
16:19:40_FireFly_a little bit
16:19:54Paul_The_NerdDon't you have to define the parts themselves though?
16:28:49Paul_The_NerdIs there a defined number of maximum sublines?
16:34:12_FireFly_Paul_The_Nerd, you are right this won't be really reduce the size of wps
16:34:29Paul_The_NerdI think it does simply things though.
16:34:47Paul_The_NerdOne option is to completely redefine the way wps images are handled.
16:34:52linuxstbOne advantage is that it reduces the number of .bmp files you need on the disk.
16:35:28Paul_The_NerdOne could just pair .wps and .bmp files, and in the future have it only handled by parting? (Just as an example alternative)
16:35:44Paul_The_NerdThe wps always loads the .bmp with the same filename, and that's that.
16:36:15Paul_The_NerdIt'd be a bit of an overhaul, and break WPS screens worldwide though. :(
16:36:24_FireFly_no the file will be only once loaded
16:37:00_FireFly_and becomes a id
16:37:32_FireFly_the parts will be generated by giving this id to the wps-tag which will be used to generate the parts
16:38:17_FireFly_i have updated my zip file because the bmp was in grayscale
16:38:26_FireFly_not 1 bit color depth
16:46:53Paul_The_NerdDoes the alternating sublines have a limit at 12?
16:55:59 Quit muesli- (Read error: 110 (Connection timed out))
16:58:54_FireFly_Paul_The_Nerd, ups i missunderstood you
16:59:29_FireFly_in conjunction with the "The wps always loads the .bmo files.."
16:59:39Paul_The_NerdOh, right
16:59:52Paul_The_NerdYeah, that was just my mind wandering to random ideas.
17:00:26Paul_The_NerdI think the WPS code needs to be audited by someone who knows it... as it is, there's some extra characters getting appended I think.
17:01:10 Join muesli- [0] (
17:02:26muesli-hehe..always the same discussion about 're'
17:02:35muesli-re= back
17:03:42 Join matsl [0] (
17:03:46Paul_The_NerdFair enough
17:13:26 Quit _FireFly_ ("Leaving")
17:15:04***Saving seen data "./dancer.seen"
17:20:32 Quit solexx ("leaving")
17:21:27 Join solexx [0] (
17:27:49 Join dpassen1 [0] (
17:28:44 Join bagawk [0] (n=lee@unaffiliated/bagawk)
17:33:25 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
17:53:26 Quit DangerousDan (Read error: 104 (Connection reset by peer))
17:57:11 Join DangerousDan [0] (
18:03:32 Quit ender` (Read error: 113 (No route to host))
18:25:29 Join bagawk_ [0] (n=lee@unaffiliated/bagawk)
18:27:32 Quit bagawk (Read error: 110 (Connection timed out))
18:37:55 Join ender` [0] (i=ychat@
18:43:41 Join _FireFly_ [0] (
18:53:12 Join amiconn [0] (
18:57:02_FireFly_hi amiconn
18:57:58_FireFly_amiconn a test plugin for combined_bitmaps can be found on
19:15:07***Saving seen data "./dancer.seen"
19:15:21 Quit muesli- (Read error: 110 (Connection timed out))
19:16:15amiconn_FireFly_: As I already said, you need to store the stride value in the structure
19:17:05_FireFly_you are right or we must define how the bitmaps have to created
19:17:06amiconnIn your demo case it will work without it, because the partial bitmaps all have the same width (30)
19:17:30_FireFly_no i had to
19:17:41_FireFly_look at the source which is also in the zip-file
19:18:03amiconn...and you added a manual correction for it, as the .bmp width is 32 pixels
19:18:23_FireFly_for the included bmp a stride must be width of bitmap plus 2
19:18:47amiconnYes, as the partial images are 30 pixels and the total width is 32 pixels
19:19:06amiconnStride must always be the width of the complete bitmap
19:20:12amiconnWhen the partial images have different widths, such a manual correction will no longer work
19:21:54_FireFly_also for combined bitmaps, where the parts not beneath each other but side by side ??
19:22:06_FireFly_must be the stride the width of the bitmap ??
19:22:39_FireFly_ok now i know for what the stride is :)
19:23:01_FireFly_it indicates where the next bitmap line begins
19:23:55_FireFly_ok when i back to linux i will test it if i set stride to the with from img[0] which is the complete bitmap
19:24:55 Quit bagawk_ ("Leaving")
19:28:51 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
19:34:42 Quit matsl (Read error: 110 (Connection timed out))
19:35:54 Quit _FireFly_ ("IceChat - Its what Cool People use")
19:36:11 Join matsl [0] (
19:37:24 Join webguest20 [0] (
19:38:19 Join _FireFly_ [0] (
19:40:30webguest20Hi guys. Odd behaviour in build 051001-1258 in H120. Power up, press A-B, select FM Radio, Press Play - blank screen. Now cursor up for repeated "sound settings" menu option over and over.
19:45:50 Part webguest20
19:52:12solexxI can confirm that (build from sep-30 with remote patch)
19:53:29_FireFly_hmm i cannot confirm that(latest cvs with remote and dirchache patch)
19:54:00_FireFly_when you press play in FM-Radio the preset menu will be shown up
19:54:20_FireFly_but if no preset are defined the display is blank
19:58:44solexx_FireFly_: do you also have a downloadable with remote *and* dircache patch?
19:58:53solexxdownloadable build
20:04:16_FireFly_in a few minutes :)
20:04:53*solexx thinks rockbox should have branches in its cvs
20:06:02 Join hd [0] (
20:06:48_FireFly_it can be found on the file is
20:07:00_FireFly_it's e complete build with all necessary files
20:07:04_FireFly_-e a
20:08:14solexxgot it
20:10:19 Quit goa ("Client suicide")
20:10:20 Quit hd (Remote closed the connection)
20:10:36 Join goa [0] (
20:11:09XavierGramiconn: any idea why cpu_idle_mode function can have a bad tuning effect on the FM radio for the iriver?
20:13:58solexxis there a way to trigger (re-)generation of the dir cache?
20:17:25XavierGrwhy do that?
20:20:00 Quit Maxime (Read error: 104 (Connection reset by peer))
20:20:09 Join Maxime [0] (
20:23:01solexxi just installed _FireFly_'s build with dircache and i don't think the cache is already there
20:24:07solexxi expect the disk not to spin up when using the dir browser
20:24:10solexxam i wrong?
20:25:33solexxon first boot, rockbox said "dirbuffer full" or sth like that
20:25:57|Lupin|Guys: does the cross-gcc we need to compile rockbox have to support C++, or is C enough ?
20:27:09XavierGrsolexx did you enabled the dircache?
20:27:41 Join Slyck [0] (
20:27:50XavierGralso if it said dirbuffer full it means that you have in a folder more than 400 subfolders.
20:28:10_FireFly_|Lupin|, only C
20:30:15solexxthen i increased the limits and the message went away
20:30:29|Lupin|_FireFly_: Thanks.
20:31:58 Join solexx_ [0] (
20:32:14solexx_(sorry for my lagging responses, my wifi is currently broken)
20:32:54 Part amiconn
20:33:15solexx_XavierGr: thanks for the hint. didn't know i had to enable it and didn't find it on the first quick search
20:33:19solexx_now it's on
20:33:46XavierGrisn't it awesome? Now you can browse and the HD will never spin on.
20:34:02XavierGrAlso no you can set the spin of the disk to a lower value.
20:34:03dpassen1how much does it add to boot time?
20:34:18XavierGr2-3 seconds but only in the first boot.
20:34:31dpassen1sounds nice
20:34:39XavierGrIn other boots it will just spin the disk for a little longer.
20:34:52XavierGrbut it is transparent for the user.
20:35:04dpassen1excellent, i believe i will give it a shot
20:35:34XavierGrsolexx: Also now you can set the spin of the disk to a lower value. Because browsing will not set the disk on.
20:35:52 Join phaedrus961 [0] (
20:35:53solexx_XavierGr: this is awesome!
20:35:54|Lupin|Someone using Debian, here, please ?
20:36:07XavierGrI had it to 10 seconds before the patch. To be able to re think If I wanted to select something else.
20:36:10linuxstb|Lupin|: Yes.
20:36:21solexx_feels even better than i expected
20:36:48|Lupin|linuxstb: May I ask how youset up the xgcc, please ?
20:37:50linuxstbI just followed the Wiki instructions - downloaded the source to binutils and gcc and cross-compiled them myself.
20:38:15linuxstbI mean "compiled it as a cross-compiler" myself"
20:38:37|Lupin|linuxstb: ahah. So you didn't follow Debian's instruction, and probably youwere right...
20:38:55linuxstbNo - there is nothing to be gained by doing it the "debian way".
20:39:07|Lupin|linuxstb: And jcd you have to give special options to gcc's configure, so that it finds the right version of binutils ?
20:39:37|Lupin|linuxstb: It's what I didn't know, precisely.
20:40:06linuxstbFirst you compile and install the binutils compiled for m68k-elf. You then need to add the bin directory containing those binutils to your path, and compile gcc for m68k-elf.
20:41:11|Lupin|linuxstb: ok.
20:41:16 Quit Maxime (Read error: 104 (Connection reset by peer))
20:41:22|Lupin|linuxstb: Did you use stow to install the packages ?
20:41:38 Join Maxime [0] (
20:41:41linuxstb|Lupin|: No. Just trust the WIki instructions and ignore Debian :).
20:42:07linuxstbCompile with a prefix like /home/lupin/m68k/ and everything you install will go under that directory.
20:42:11|Lupin|linuxstb: Stow is not debian, is it ?
20:42:24linuxstb|Lupin|: I've never heard of Stow.
20:42:51linuxstbAh GNU Stow - just googled for it.
20:42:59|Lupin|linuxstb: give it a try, if you havetime. It's a nice tools when you wantTo install something in a clean wau.
20:43:05 Join Lear [0] (
20:47:50 Quit solexx (Read error: 110 (Connection timed out))
20:56:06Slyckwhere can i go to find info on the porting of rockbox to the h300 iRiver? cant find anything on the main rockbox site
20:57:07linuxstbSlyck: Start here:
20:57:19linuxstbAnd here:
20:58:24Slycknice to see the h100 fw is almost completed
20:58:28Slycklooks awesome
20:59:31Slyckwhats "Multi-codec Architecture"?
21:00:32Slyckquite a bit done to the h30 aswell didnt think you guys had done anything but the wiggler
21:04:30 Quit Rick (Read error: 104 (Connection reset by peer))
21:05:40 Join Rick [0] (
21:07:35linuxstbSlyck: "Multi-codec Architecture" is just the name give to the part of Rockbox that deals with audio playback. It can handle an unlimited number of codecs via codec "plugins".
21:11:55Slyckvery nice
21:15:09***Saving seen data "./dancer.seen"
21:17:03 Join thedude02 [0] (n=thedude@
21:18:01 Part thedude02
21:18:37|Lupin|bye there
21:18:41|Lupin|thanks a lot for helping
21:18:48 Quit |Lupin| ("leaving")
21:26:49 Join linuxstb_ [0] (
21:30:37 Join linuxstb__ [0] (
21:43:37 Quit linuxstb (Read error: 110 (Connection timed out))
21:45:19 Join |Lupin| [0] (
21:45:22|Lupin|Ri again, folks.
21:45:45|Lupin|Sorry to bother you again. I get an error when compiling the firmware for an iRiver 140.
21:46:06fuzziewhat error?
21:46:11|Lupin|Thereis a CONVBDF which make doesn't know how to interprete.
21:46:23_FireFly_you have to do a make in tools first
21:46:28fuzziego into tools, run make there first
21:46:34|Lupin|(I use a cvs version, and I just did an update)
21:46:53|Lupin|stupid error...
21:47:20 Quit linuxstb_ (Read error: 110 (Connection timed out))
21:48:03|Lupin|seems to compile much better now, thanks guys.
21:48:15|Lupin|And sorry for not having re-read the doc just before compiling.
21:50:55|Lupin|And now the file which mut be copied on the player is rockbox.iriver, and then just play it like a normal mp3. Right ?
21:51:34LearNo, you need a bunch of files, easily created with "make zip".
21:52:07LearAnd you have flashed your unit?
21:52:26Lear(But you can play an .iriver file, which will cause a restart with the new firmware)
21:52:28|Lupin|Lear: no, it is not yet here...
21:53:09_FireFly_you have patch the original firmware with the rockbox bootloader and update the player with the patched version
21:53:30|Lupin|not yet, as I said.
21:53:57|Lupin|I'll readThat.
21:54:09_FireFly_without the bootloader rockbox won't work ;)
21:54:35|Lupin|bye, this time it's t!ue :-)
21:54:41 Quit |Lupin| ("leaving")
21:55:02Slyckim off
21:55:40 Quit Slyck ()
21:58:13_FireFly_ok i have updated my plugin now you can through an config-file(cb.cfg) specifiy which bitmap should be loaded an wich parts should be generated from the bitmap and where the parts should be displayed
22:01:14_FireFly_max images is restrictet to 10(bitmap+9parts)
22:01:30_FireFly_it can be found + source on
22:01:51_FireFly_this is a test plugin for combined_bitmaps
22:02:50_FireFly_the source bitmap will be displayed on position 0,0
22:08:19 Quit matsl (Read error: 104 (Connection reset by peer))
22:08:46 Join matsl [0] (
22:14:35 Nick TiMiD[farAway] is now known as TiMiD (n=TiMiD[
22:14:35DBUGEnqueued KICK TiMiD
22:14:43_FireFly_hi TiMiD
22:15:07TiMiDhi :)
22:15:37TiMiDI've a question (prepare yourself, it's a dumb question ^^)
22:15:50TiMiDwhat is exacly "id3db" in rb ?
22:16:31_FireFly_i think it's similar to the db thing in the original firmware
22:17:22TiMiDI think so (but I never saw it running (don't find how to activate it in the menus) )
22:17:30_FireFly_afaik which this db you can sort our files depending on id3 tags
22:18:11TiMiDthe pbl is that it's somehow handle in the code that also handles file tree view
22:18:21TiMiD(and it's a mess )
22:18:27linuxstb__It's described here:
22:18:32TiMiDok ty
22:18:34_FireFly_o to slow ;)
22:19:50_FireFly_TiMiD, if you want to know i have written a test plugin for combined bitmaps
22:20:22TiMiDI saw it on the forum :)
22:20:26TiMiDgreat ;)
22:21:12_FireFly_i have an updated version which can be controled through a config-file
22:21:48_FireFly_it can be found on the same place (
22:22:13TiMiDI look
22:29:05TiMiD_FireFly_: how can I test it ?
22:29:24_FireFly_in the zip file there is a test file included
22:29:48_FireFly_simply copy the .rockbox dir in the zip on your player
22:30:57TiMiDyes, that what I did
22:31:05TiMiDI have a .cfg and a .bmp
22:31:18_FireFly_and the plugin is in the plugins dir
22:32:24_FireFly_in the cfg file can you define the bitmap to be loaded and up to 9 parts
22:32:26TiMiDI don't see it (maybe because you put in plugins/ and rb only looks into rocks/
22:32:51_FireFly_my failure
22:33:01_FireFly_it should be in rocks/ ;)
22:33:03TiMiDI ran it manually :p
22:34:01_FireFly_ok the zip is corrected
22:34:45TiMiDok so it takes a ingle bmp and extracts parts that can be displayed where you want ?
22:35:03TiMiDlike sprites (for example in rpg maker :D)
22:35:27_FireFly_it usese the lcd_mono_bitmap_part function
22:35:58 Join mrelwood [0] (
22:36:08_FireFly_i have extended the used image struct in the plugin so i holds now the source coords and the strade value
22:36:17_FireFly_it holds
22:36:39mrelwoodis it common knowledge that the pcm recording in iriver has left and right channels swapped?
22:38:03TiMiD_FireFly_: what you could do is to create an id for each part so that a part can be accessed (eventually displayed) by specifying its id
22:38:51TiMiDthen it could become used in wps ^^
22:39:03TiMiD(instead of loading tons of bmp
22:39:43mrelwoodare people using the pcm recording, or is it just for reference?
22:39:43_FireFly_the plugin is an result from the thinking about using combined bitmaps in wps with amiconn
22:39:51TiMiDnext step : animated gray_scale bmps :D
22:40:39_FireFly_that should be no problems insteed lcd_mono_bitmap_part use lcd_bitmap_part ;)
22:40:53_FireFly_but then it works only for the main-lcd
22:41:00_FireFly_the remote is monchrome
22:41:13TiMiDtechnically you can do gray-scale on remote display
22:41:22_FireFly_really ??
22:41:41TiMiD(I saw it when testing my lists, it was going so fast and it was gray :p)
22:41:46mrelwoodare you planning display plugins for iriver h300?
22:42:02TiMiDmrelwood: h300 is not even supported :(
22:42:09_FireFly_no for the iriver H1xx serie currently
22:42:20_FireFly_a extension to the wps
22:42:23mrelwoodokay. cool :)
22:42:26TiMiDbut when it will it will have the same functionnalities as the other players
22:42:42mrelwoodany of you worked with the recording?
22:42:49 Join einhirn [0] (
22:42:57_FireFly_i think most of the code for H1XX could be used also for the H3xx
22:43:15TiMiDyes, exceped display code -____-
22:43:40_FireFly_yepp that is the only thing that must be written from scrach due the color display ;)
22:43:47TiMiDI think (but it's my personnal opinion) that the code needs some cleaning
22:44:09TiMiDI'm working on tree.c and it's hell :(
22:44:32TiMiD(oops I hope the guy who wrote it isn't connected XD)
22:44:52TiMiDit's the same with wps I suppose ?
22:45:11mrelwoodi guess the recording feature is not at the top of the priority list?
22:45:48_FireFly_i think the complexity of these codes is grow by each new device which is supported by rockbox
22:45:49TiMiDmrelwood: I don't know ... if you want you can submit patches for thsi :)
22:46:10TiMiDand there are too many global variables :(
22:46:32mrelwoodi would've started working on it 1.5 years ago if I knew a slightest bit about coding...
22:46:52TiMiDalso to handle id3db and file browsing in the same file was a bad idea :D
22:47:33_FireFly_the only positive of this might be that no code must be written double
22:47:45TiMiDmrelwood: then strike a dev until he wants to code it ^^
22:48:18TiMiD_FireFly_: yes, the reason was that id3db display and filetree display was the same code
22:48:31_FireFly_was ??
22:48:36TiMiDbut with the things I coded, it could be separated
22:48:41mrelwoodis dev a person, or "developer"? (newbie :P)
22:48:53_FireFly_mrelwood: yes
22:48:57TiMiDmrelwood: dev's are not human ^^
22:49:14_FireFly_mrelwood: dev = developer
22:49:48TiMiDall devs are crazy (at least, it's how it is in my school ^^)
22:50:15TiMiDcomputer science is dangerous for your brain
22:50:30mrelwoodok. is there a specific "dev" that is fighting the recording feature?
22:51:43TiMiDnothing was commited since Sat Aug 13
22:52:02TiMiDbut it doesn't mean noone is working on it (I hope)
22:52:38TiMiD_FireFly_: if there wasn't the id3db, I would rewrite the file tree from scratch I think
22:53:11mrelwoodTiMiD, the latest is included in the daily builds, right?
22:54:40_FireFly_hmm what about this: we have two function(filetree, id3db) which gaves the string to display at a specifig position
22:54:44mrelwoodthen i've been using the latest since aug 13. :o) Atleast the Rockbox recording doesn't have the iRiver glitch.
22:55:05_FireFly_yepp afaik
22:56:53TiMiD_FireFly_: since the list uses callback functions, this is how it must work :p
22:57:14TiMiDI did a callback fn for the filetree
22:57:33TiMiDbut it works over the whole filetree arch that also handles id3db
22:57:39TiMiDso it's not clean
22:57:52_FireFly_hmm then you have to split it
22:58:06_FireFly_if possible
22:58:08TiMiDit unsplittable :D
22:59:00TiMiDif I split I broke everything
22:59:13_FireFly_hmm then a complete rewrite might be necessary
22:59:25TiMiD(becauuse there are also others parts of the fw taht uses things exported by the filetree)
22:59:53TiMiDyeah let's rewrite rockbox ^^
23:00:17_FireFly_i meant mainly the filetree :)
23:00:45TiMiDit's like a card castle (I don't know if this is how you say it in english)
23:01:16TiMiDif I rewrite smth I must rewrite everything that's linked to it and sicne everything is linked ...
23:01:28_FireFly_i know i have expressed it a little bit bad ;)
23:02:37TiMiDno, you are right
23:02:48_FireFly_maybe we could generate function which gaves the needed information to this parts of the fw which needed information from the filetree
23:04:04TiMiDnot temporary :p
23:04:20TiMiDif another part of rb needs some informations then it should have it
23:04:41TiMiDthis is to be included in the filetree api
23:04:41_FireFly_ok then not temporary
23:04:45 Join amiconn [0] (
23:05:06_FireFly_hi amiconn have you read the logs or should i say it again about my test plugin
23:05:07TiMiD(but hey it's crazy to rewrite everything !!!)
23:05:10TiMiDhi amiconn
23:07:05CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:07:05*amiconn is checking the log
23:12:13amiconn_FireFly_: The problem with greyscale bitmaps in wps is not the change in function
23:12:56amiconnThe real problem is that while b&w bitmaps have a notion of foreground & background, greyscale and colour bitmaps are opaque
23:13:42amiconnWhile this isn't a problem for elements which don't overlay text, it will be a problem for a background bitmap
23:14:29linuxstb__amiconn: Is the solution to have a certain colour act as transparent?
23:14:30_FireFly_hmm then it should be if possible first drawn(the background image)
23:15:10***Saving seen data "./dancer.seen"
23:15:59amiconnlinuxstb__: That would be very difficult at least. These bitmaps are in native lcd format
23:16:44amiconn_FireFly_: Yes, the wps would need two layers: background and content
23:16:47_FireFly_amiconn: hmm what about when defining which color should be transparent on loading the bitmap
23:17:12linuxstb__amiconn: Are you thinking more about greyscale than colour. Or both?
23:17:30_FireFly_and then this color is left of by converting the bitmap in nativ lcd-format
23:17:30amiconnIt's the same in respect to the problem
23:17:55_FireFly_so that these parts are "empty"
23:18:11amiconnI think supporting greyscale for the small element bitmaps would be too difficult
23:18:32amiconnMy idea is that a greyscale or colour background bitmap should be allowed
23:18:48_FireFly_with the widget system, which TiMiD is working on, it might be a less problem to have a back- and foreground for wps
23:18:59 Quit ansivirus (Read error: 110 (Connection timed out))
23:19:00amiconnPreload/display bitmaps should always be monochrome, but with a selectable foreground shade/colour
23:19:04linuxstb__The "blit" function is simpler for colour - at least for 16-bit colour. In which case it is easier to check for a transparent pixel. We also have lots of colour values to choose a tranparent one from.
23:19:28linuxstb__If we add transparency to greyscale, we effectively have 5 colours.
23:19:32TiMiD_FireFly_: uh oO ?
23:19:51linuxstb__Which is obviously a problem.
23:20:13amiconnlinuxstb__: Bitmaps in rockbox use the native lcd format (or the closest match for monochrome bitmaps and non-monochrome displays)
23:21:17TiMiDamiconn: what about a separated boolean table which would say for each pixel of the bitmap if it must be dran or not
23:21:34TiMiDlike that you could send only the needed pixels to the screen
23:21:35_FireFly_TiMiD: should i haven't said that ?? :)
23:21:53TiMiD_FireFly_: I'm not working on grayscale :p
23:22:34TiMiDamiconn: (but I only looked from far to the grayscale code since it's assembly and I don't want to understand)
23:22:35linuxstb__amiconn: I agree with regards to low depth LCDs. But with 16-bit, it becomes easier to assign one of the 65536 values to transparency, and check for it when blitting.
23:23:21TiMiD16 bits on monochrome displays oO
23:23:38_FireFly_i think he mean on color displays ;)
23:24:33TiMiDeven with 8 bits you could use 254 for transparency for example (since it's only 33 shades)
23:24:41amiconnTiMiD: We're not talking about the grayscale lib here, that's definitely code not meant to be used in the core
23:25:28amiconnIt draws too much cpu power for everyday hours-long use.
23:26:00amiconnThe H1x0 lcd allows 4 shades of grey, and the lcd driver already supports that
23:26:28amiconnOnly very few parts of rockbox already use it, core parts being only splash() and the rockbox logo
23:26:54linuxstb__and Sudoku. :)
23:27:00 Quit Lear ("Chatzilla [Firefox 1.4.1/undefined]")
23:27:09amiconnlinuxstb__: That's not a core part
23:27:24amiconnPlugins use it as well, that's correct
23:27:31_FireFly_then the only possible solution which i currently see is that we have to implement an background and foreground
23:27:31linuxstb__Sorry - just read the first half of your sentence.
23:27:37amiconn(sudoku + minesweeper + cube)
23:27:45TiMiDI thought everything was handled by grayscale lib
23:28:14amiconnNo, the grayscale lib handles things with >4 shades - up to 33
23:28:19linuxstb__grayscale lib is for displaying more grayscales than the LCD natively supports.
23:28:20amiconn(on archos and H1x0)
23:28:44amiconnThere are 3 plugins using it - grayscale, mandelbrot, and the jpeg viewer
23:29:19amiconnI'm still planning to use it for porting the solid grey cube mode to archos
23:29:19TiMiDI used it also in plasma and fire plugins (but they are not official ^^)
23:29:56TiMiDbtw, I noticed that the 8bit grayscale is not linear on iriver display
23:30:14TiMiDor at least doesn(t seems
23:30:30amiconnThese LCDs aren't really made for greyscale
23:31:03TiMiDyou wrote the grayscale lib ?
23:31:27amiconnI'm currently working on removing the need to boost the CPU all the time
23:31:36TiMiDwouldn't it be possible to add sime kind of palette correction specific to the display ?
23:31:40amiconnIt already works, it just needs some more cleanup
23:31:47TiMiDamiconn: seems hard :/
23:32:03amiconnNot really, once you know how to do it
23:32:24TiMiDI forgot that multitask on rb doens't switch if you don't tell it to
23:32:37TiMiDso it can't switch when writing a pixel ;)
23:32:39amiconnI also want to include the timer tick interrupt into my new CPU frequency transition handling for timers
23:33:38amiconnIn order to work best, the new mechanism has only one precondition - the higher frequencies need to be integer multiples of the base frequency
23:34:40amiconnSo I shifted the 48 MHz mode to 45 MHz, and 120 MHz will become 124 MHz
23:35:02TiMiDyou can control the cpu freq mhz by mhz ?
23:35:41amiconnYou set the frequency via 2 dividers which control the PLL
23:35:50amiconnThis allows rather fine control
23:38:13 Join tucoz [0] (
23:38:27TiMiDhi tucoz
23:38:47tucozamiconn: did you see that guy that managed to get blue on his h120?
23:38:56tucozand what do you think caused it?
23:39:14TiMiDblue oO
23:39:17amiconntucoz: It's a quite normal effect for that type of LCD. I also got that once on my archos
23:39:23tucozhehe, ok.
23:39:31TiMiDhow is it possible ?
23:39:36amiconnIt's just displaying black and the contrast is way higher than normal
23:39:47amiconnCall it 'ultra-black'
23:39:53TiMiDha :p
23:39:54Moostempetature way
23:40:02tucozah, I see.
23:40:56tucozshort visit, got to go. see you
23:40:58 Part tucoz
23:42:37TiMiDhe should have put rockbox in his link section
23:44:09linuxstb__His work would be useful if he had chosen a H300 to play with.
23:44:27linuxstb__But I don't understand his motivation at all with a H140.
23:44:53TiMiDI can somehow understand
23:44:54 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
23:44:57TiMiDit's fun :)
23:45:33TiMiDsome time ago, I seriously thought of trying to port uClinux to iriver
23:45:51TiMiD(but this is too much for me, I'm too lazy)
23:46:09linuxstb__That would be fun because it is new and you would be the first person to do it. What he's trying to do has already been done by Rockbox.
23:46:39TiMiDalternatives are good :)
23:47:19linuxstb__Someone should send him a H300 :)
23:47:38TiMiDno ! i want it supported in rb :p
23:48:21linuxstb__I meant that he could help the port to the H300.
23:48:35linuxstb__But you're right - he is enjoying himself.
23:49:09TiMiDIf his code is gpled, then it should be good for everyone
23:49:31amiconnHmpf, I have a problem :(
23:49:40TiMiDwe can't blame him if he doesn't want to contribute to rb
23:49:40linuxstb__Can anyone help?
23:50:13amiconnIf I convert the timer tick to use my new mechanism, I can't do that for the gmini code, so the change will break it
23:50:57_FireFly_than you nee a kind of fallback
23:51:17_FireFly_maybe in source-code through a define
23:51:23amiconnThe gmini build is currently unmaintained, but I would like to avoid that
23:51:50_FireFly_so that the old code is used for gmini and your new code for the rest
23:52:20linuxstb__Is the reason you can't do it on the gimini because you don't know how to, or because it can't physically be done?
23:52:20amiconnThe problem is that it's a whole new mechanism, I'm even planning to mode the tick timer code to timer.c
23:52:41amiconnI don't know how to as I don't know the hardware
23:53:11linuxstb__How advanced was the gmini port? Was it working?
23:53:39*linuxstb__ reads wiki
23:53:43amiconnIt was partially working (disk access, browsing), but not playing music
23:56:01linuxstb__You obviously have three choices - don't make your changes, make your changes and break gmini, or add lots of #ifdefs and try and keep the gmini alive with obsolete code.
23:57:10TiMiDI think I would choose #2 :)
23:57:24 Join Aison [0] (
23:57:39linuxstb__Do you think it will be hard for a gmini dev to fix the problems you are planning to cause?
23:57:54amiconnThat depends on how the timer works
23:58:13 Quit ender` (Read error: 113 (No route to host))
23:58:13linuxstb__I see your problem.
23:58:40amiconnOn iriver it is simple, because there's a separate prescaler which allows any factor between 1 and 256, not just powers of 2
23:59:15amiconnI just calculate the timer count for base frequency and prescaler == 1 (or a power of 2 <= 16)
23:59:53amiconnFor the higher frequencies I then multiply this prescaler value with the frequency multiplier - that's why it should be an integer number

Previous day | Next day