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 2005-12-10

00:00:10webguest76San: Have you tried searching for cygwin1.dll to see if you have multiple versions installed?
00:00:38Sanexcept for one
00:01:08amiconnThere's a filesystem for the good old amiga which allows this: It does write caching as long as data is written continuously, but if a (configurable) timeout expires after the last write operation, it starts flushing the cache to disk
00:01:16webguest76I don't know.. remove the devkit and install a full cygwin
00:02:58amiconnFor a removable device, just configure a short timeout, say 5 seconds, and you're almost as safe as with write caching disabled, but with better performance
00:03:06webguest76San: I trust you've tried restarting?
00:03:50webguest76amiconn: I'm sure lkml would welcome such a patch!
00:05:34Sanoh, wait, all the cygwin1.dll files are showing up now
00:05:42Sanwhat ones do i delete?
00:05:58webguest76Which ones do you have?
00:06:04webguest76And where's the devkit installed
00:06:21Santhe devkit in in C:\
00:06:56Sanand there are cygwin1.dll files showing up in system32 and \WINDOWS
00:07:23webguest76oh boy!
00:07:50Sangood or bad?
00:08:45webguest76Well.. is there one in c:\rockbox as well?
00:09:24amiconnCheck the versions
00:09:43webguest76Just delete the ones not in c:\rockbox, I'd say
00:10:18amiconnproperties, page 'version'
00:11:40 Quit linuxstb ("CGI:IRC")
00:13:51Sanw00t, works
00:13:53San:D gone?
00:19:16 Quit muesli_- (Read error: 110 (Connection timed out))
00:20:04RotAtoRwebguest76: works fine here...
00:20:35webguest76Ah, tehre we go
00:20:44webguest76for a while I got "no website configured at this address"
00:20:51webguest76a minute or so
00:22:54Sannow i get BASH: ../tools/configure no file or directory, there is!
00:22:58 Join DreamTactix291 [0] (
00:23:17webguest76try ls ../tools/configure
00:24:06webguest76not there?
00:24:16webguest76Well, you're in the wrong directory then
00:24:29webguest76Or maybe you just have the dir set up differently
00:24:31Sanit is there when i look for it in the browser
00:25:12webguest76Sounds like your dir structure is not how you expect
00:25:47webguest76Which dir are you in?
00:25:53webguest76c:\rockbox\build or something?
00:26:14webguest76do `pwd`
00:26:18webguest76and paste the result
00:26:20SanI am in guest
00:26:24Sani did do pwd
00:26:38webguest76And what's the path?
00:28:12webguest76That'd be /home/guest
00:28:13webguest76Where's the tools dir?
00:28:21Sanin guest
00:28:37webguest76Ah, you should create a build dir
00:28:40webguest76mkdir build
00:28:41webguest76cd build
00:28:43Sani did
00:28:54webguest76you should be in that dir when you configure
00:29:04Sangot it
00:29:14Sani didn't know you had to do it in the build dir
00:29:21Sani tought it was the guest dir
00:29:23Sanok, i get it
00:29:25Santhanks man
00:30:20San~/build> what do I do now?
00:31:15Sannever mond
00:38:57 Quit gromit` (Read error: 104 (Connection reset by peer))
00:42:42XavierGr_ok strange situation. What if a function calls a function that calls the previous one? What happens then?
00:44:55RotAtoRstuck in a infinite loop until a stack overflow maybe?
00:45:10RotAtoRbut i really don
00:45:25RotAtoR* but i really don't know much about these types of things
00:45:47XavierGr_Well you are close.
00:46:18XavierGr_My jpeg viewer will show a stack overflow error after ~12 jpeg changes.
00:47:27webguest76That sure does sound like "caught in an infinite loop until a stack overflow"
00:47:27 Join muesli_- [0] (
00:47:41webguest76Where infinite ~ 12
00:48:49XavierGr_wel I would assume that it calls the function before it finish its progress, but if I quit manually how all instances of main terminate?
00:49:10XavierGr_(maybe because plugin termination is called all instances of the viewer are lost I am not sure.)
00:49:41XavierGr_Thing is that after a few file changes I get the stkov error. It must be it.
00:50:28 Join Kohlrabi [0] (
01:09:01 Quit Moos ("Glory to Rockbox")
01:09:20 Join Kyl3 [0] (
01:12:13***Saving seen data "./dancer.seen"
01:12:38 Quit ender` (Read error: 104 (Connection reset by peer))
01:15:02 Quit dpassen1 ()
01:25:08 Quit webguest76 ("CGI:IRC")
01:36:48 Join Midgey34 [0] (
01:37:36Midgey34RotAtoR: I have some suggestions for colors on Bejeweled
01:38:37XavierGr_yes that was the culprit. No none of the functions call another before finishing and no more stkov errors :)
01:39:51 Join TCK [0] (
01:41:03 Quit Kohlrabi ("Leaving")
01:42:01Sani fixed pong
01:42:16Sani alligned the buttons better and easier
01:43:11RotAtoRMidgey34: sure, what kind of suggestions?
01:44:29Midgey34for COLOR_YELLOW I used LCD_RGBPACK(255,255,0)
01:44:43 Join Benacool [0] (
01:44:51Midgey34its a vibrant yellow as opposed to the current "gold" which actually looks rather ugly
01:44:58RotAtoRok, so something a little brighter
01:45:29Midgey34well, the gold on screen looks sort of puke green/brown
01:45:35Midgey34not very nice
01:45:40RotAtoRi don't have a h300 that I can test it on, so I just have to trust what others say about the color :)
01:45:43XavierGr_and please consider those who have dischromatopsia...
01:45:59Midgey34and pink I changed to a more dull color
01:46:14Midgey34so its easier to tell apart from purple
01:46:28Midgey34I'm using LCD_RGBPACK(255,136,205)
01:46:49Midgey34excuse me, 204 (not that much of a difference
01:47:07Midgey34that way its easier to pick out when scanning the board
01:47:08XavierGr_I remember 2-3 years ago playing Diablo 2. I couldn't recognize any difference between gold and green.
01:47:16XavierGr_Or in puzlle bubble
01:47:30RotAtoRok, I'll take a look at those
01:47:41Midgey34XavierGr: well, this new yellow is very different from green
01:47:58RotAtoRXavierGr_: are there any particular colors in Bejeweled that you have a hard time telling apart?
01:48:04Midgey34RotAtoR: try running them on the sim, it gives a decent representation of the color
01:48:24Midgey34at least here it does but I'm running an LCD monitor
01:48:49XavierGr_Sorry I don't own an H300 ;P Just saying this for the other users.
01:48:52RotAtoRfrom the pictures I've received of it actually running on an h300, the sim looks a fair amount different on my screen
01:49:02RotAtoRahh, ok ;)
01:49:06XavierGr_ah yes I can try the sim and see.
01:49:48Midgey34hmm, how does grayscale work? is it LCD_BRIGHTNESS or something else?
01:49:52RotAtoRspecifically the background looks much different, it's quite blue in the sim
01:50:07RotAtoRI'm not using any grayscale
01:50:17Midgey34right now I'm using LCD_BLACK, LCD_DARKGRAY
01:50:35Midgey34well, I'm trying to get grayscale working on rockblox
01:50:35RotAtoRwith a white background, light gray was too close to white and dark gray looked to close to black
01:51:03Midgey34and the current defines I'm using for grayscale cause a nasty bug
01:51:29Midgey34anything that's black is ignored by the method that check the lines to see if they're clear
01:51:36Midgey34color works perfectly though
01:52:03Sanyes it does
01:52:50Sani have got a nice color scheme going on :D
01:53:37Kyl3color messes with rockboy tho
01:53:42Midgey34for anyone interested, I've added scores to rockblox, dropping method (press a button, the block goes to the bottom), color/grayscale support (currently buggy) and a pause method. I plan on implementing highscore tables and saving games
01:54:02SanI am adding color to pong
01:54:18Midgey34someone requested snake2
01:54:22Midgey34but I haven't had time
01:54:25Sani will try that then
01:54:29Sani have loads of time
01:54:36Sankyl3, i got the pong buttons working better and easier
01:54:45Midgey34I got the playfield to be the correct size, that took like 30 sec
01:54:50Kyl3Is there anyone here who grants CVS access? I apply loads of patches that havent been in the builds yet and it would help if i had access.
01:54:52Midgey34but that's all
01:55:06Kyl3You did? awesome. I need to look around
01:55:14Midgey34Kyl3: write access?
01:55:22Midgey34and there's no patch yet
01:55:36Midgey34(for rockblox that is)
01:55:46Sancolor in snake?
01:55:47Midgey34hmm, I think you'd have to talk to linus
01:55:55Sani'll do that then
01:56:06Midgey34San: yes, NIghtCat requested it last night
01:56:09Kyl3what file?
01:56:33Midgey34that's pretty crazy
01:56:36Midgey34but I'm off
01:56:40Sankk, cya
01:56:43Midgey34I'll cya later
01:56:48 Quit Midgey34 ("Download Gaim:")
01:56:55Kyl3san does it mss with your rockboy or just me
01:57:19 Join cool [0] (
01:57:46Sanjust you
01:58:06Kyl3Maybe it was because I had black on white.
01:58:22Kyl3and what file to edit snake2/plugins?
01:59:40Sanwell, the new button allignment works in pong, but the color isn't
01:59:43*San tryed again
02:01:57Kyl3made it fit the screen
02:02:19Kyl3i think
02:02:43Kyl3lemmie check
02:04:45 Quit mirak_ (Connection timed out)
02:05:22Kyl3im gonna make pong better
02:09:04Sani did
02:09:10Sanwhat are you going to add?
02:09:38Kyl3I was gonna fix the buttons. But you did already. And my snake2 fix didnt work.
02:09:50*Kyl3 trys again
02:10:20Sanwhat are you doing? Making it wider?
02:10:34Kyl3trying to
02:10:46Sanfixed it
02:10:51 Quit Benacool (Read error: 110 (Connection timed out))
02:11:02Kyl3lol san you have to show me these things
02:11:13SanActually, I cant code for sh*t
02:11:18Sani never learned
02:11:21Kyl3neither can i
02:11:25SanI just look and make sense of it
02:11:33Kyl3I learned php 101 but
02:11:41 Quit cool (Read error: 104 (Connection reset by peer))
02:12:00 Join Benacool [0] (
02:12:35Kyl3aha I think I got it
02:13:12Santhink I got color in pong
02:14:38*San compiles
02:14:56Kyl3fixed Pong
02:15:13Sanwhat fix?
02:15:47Kyl3I made it to the 2 left control the left and the 2 right control the right.
02:15:53Santhats what I did
02:16:15Kyl3well actually then.. I dont know what the code name for A-B is
02:16:51Kyl3and if I were to set the right down to BUTTON_OFF then how would I quit?
02:17:08 Quit muesli_- (Read error: 110 (Connection timed out))
02:17:24Sanyes it would
02:17:36Sanunless you changed off to navi
02:17:54Kyl3i dont know the code name
02:19:05Kyl3BUTTON_NAVI lets try that
02:20:27Sanit is
02:20:51*Kyl3 compiles
02:22:14*San tests snake
02:23:13Kyl3lol i left pong running so i could see what button is what code name and the score is 50-52
02:24:26Kyl3grr its not BUTTON_NAVI
02:24:39 Join actionshrimp [0] (
02:24:52Kyl3BUTTON_SELECT maybe
02:28:13Sanhave a look at buttons.c
02:28:24Sando you know where plugins.h is?
02:28:32Kyl3yep. its BUTTON_SELECT
02:29:15Sanah, found it
02:29:33Kyl3whats in there?
02:30:07Sanit tells the .rocks what to do
02:32:02lostlogic*does a profile-to-profile same song, same CPU speed, same everything except IRAM placement comparison*
02:32:03 Quit Benacool (Read error: 104 (Connection reset by peer))
02:33:03Kyl3sanke2 is so screwed up youd laugh
02:33:31Kyl3pong is great tho
02:33:58Sanis that a bmp in snake 2?
02:34:45Kyl3what do u mean?
02:35:38Sanfigured it out
02:35:40Drumr|awayhey, do the VU meter, oscilliscope/oscillogram/ some other cool plugin stuff work in the h300?
02:35:58Sannever tryed
02:36:20Drumr|awayim curious, cause theyre included in the sim, but dont run in the sim
02:36:33 Join Benacool [0] (
02:36:40Kyl3theyre not in the plugin menu
02:36:40SanI could compile it for you
02:36:56Kyl3theyre in the source tho
02:38:16Sanwhat is the nevi button again kyl3?
02:38:52Kyl3trying snake2 again
02:40:30Kyl3I cant find where the resolutions are. If i change what I think they are to 220x176, it doesnt work
02:41:36SanI just took the graphic out all toghther
02:41:46Sanit is the graphic, not the resolutions
02:44:17Kyl3Oh! that works.
02:44:50Sanmine takes ages to compile
02:44:54Kyl3So I would take out anything that says #Define BMPHEIGHT/WIDTH
02:45:17Sansee all the crap that has x00100 and stuff
02:45:24Santhat is the graphic
02:45:31Sandelete all that
02:45:43Kyl3whats your CPU speed?
02:45:49Sanon PC?
02:46:20Sandont know
02:46:33Kyl3are you on XP
02:47:03Kyl3Start > My Computer > right click on it > Properties
02:48:14Kyl3mine is 3.4 tho.. mabye thats why it takes a bit longer for you
02:49:05*Kyl3 recompiles
02:49:15*San has an idea
02:49:24*Kyl3 asks what sans idea is
02:50:17*San isn't going to tell him till it work
02:51:34Sanwhat is that type of encoding for the images in the plugins?
02:52:15Kyl3Yes. And your gonna make your own I'm assuming
02:52:31Sanhow do I convert a bmp to hex?
02:52:54Kyl3Maybe open it in a text editor
02:53:01Kyl3hex editor*
02:54:01 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
02:54:35Kyl3hhmmm i deleted the 0x00 stuff (all of it not just the first part) and it wont compile
02:54:47Sandont compile
02:54:51Sandidn't work for me
02:55:14Kyl3what didnt work? what i just said?
02:55:35 Join webguest36 [0] (
02:55:45webguest36To convert bmp files, use tools/bmp2rb
02:56:48Kyl3wait... howd you know what we were talking about if you joined after the conversation about it?
02:58:57Kyl3ok. lol.
02:59:30XavierGr_ah webguest you ruined the suprise.
02:59:48XavierGr_you shouldn't have told them
02:59:54 Quit Benacool (Read error: 104 (Connection reset by peer))
03:00:23XavierGr_by the way fixing snake2 for H100 was my first step to get into rockbox code too.
03:00:31 Quit webguest36 (Client Quit)
03:00:53XavierGr_This picture around the screen is made by zezayer
03:01:03Kyl3I cant seem to make it wider
03:01:06XavierGr_and I did the hiscore and made it to fit the H100 screen.
03:01:39XavierGr_let me think If I can remember what I did back then
03:01:49XavierGr_but you will have to remove the pics to see.
03:02:13Kyl3its ok
03:02:45XavierGr_ah ok if you want to make it fit your screen you have to fiddle with the defined modifiers.
03:03:12XavierGr_Line 37
03:03:23XavierGr_Multiplier and Modifier 1 & 2
03:03:30Sanhow do you use bmp2rb?
03:03:48XavierGr_I introduced those values back then to make it fit larger screens.
03:03:49Kyl3type it when your in the tools dir
03:04:12Kyl3and go bmp2rb filename.bmp
03:04:20XavierGr_Though it will not fit exaclty (both height and width) except if the screen aspect ratio is the same with archos screens
03:04:39XavierGr_(that's why we put the little snakes images to fill the blank space)
03:04:39Kyl3and go bmp2rb -f4 filename.bmp*
03:04:49Kyl3oh thanks
03:05:41Sanok, i did
03:05:51Sannow how do i get the code, it came up on the screen
03:06:00Kyl3i would change the mutiplier to? 6?
03:06:14Kyl3good luck san... no C/P in the dev kit ;P
03:06:50XavierGr_there is copy paste.
03:07:11XavierGr_just right click on the rockbox shortcut
03:07:18XavierGr_and select properties
03:07:32XavierGr_goto options tab
03:07:53XavierGr_and tick quick edit mode and insert mode.
03:08:04Kyl3THANK you
03:08:30XavierGr_now hold click on the command line and select your test. right clcik to copy\
03:08:50XavierGr_right click to paste sorry
03:08:59XavierGr_and ctrl+c to copy
03:09:35Kyl3ctrl and c on the keyboard is copy
03:09:42SanI cant copy from the cmd thing
03:09:50Sanit doesn't highlight still
03:09:58Kyl3gotta restart it
03:10:23 Join ich [0] (
03:10:59Sanaaah, I was using the _RockBox.bat
03:11:28Sanaaaah, screen is too small
03:11:33Sanhow do I scroll up?
03:12:15***Saving seen data "./dancer.seen"
03:12:35XavierGr_you can set your cmd's screen size in the options I said before.
03:12:41XavierGr_play with the options to see.
03:13:11Sanah, ok, ty
03:13:33XavierGr_only thing I haven't found yet is to set a bigger buffer for the screen in order to scroll up and see the history. It will just not allow me to make it bigger dont know why
03:13:57 Quit ehntoo (Read error: 104 (Connection reset by peer))
03:13:58 Quit ich (Client Quit)
03:16:20Kyl3so... i fixed pong. and i edited my colors... and a left snake 2 alone... so now... ROCKBOY
03:16:38Sancolors in pong?
03:17:02Kyl3any luch with colors in pong?
03:17:04Sani did it before you though
03:17:08Sanonly tryed once
03:17:15Kyl3yes you did
03:17:41Sanit is 02.17 in the morning
03:17:55Sanand i will work on it more tomorrow
03:18:02Kyl39:17 here
03:18:07Kyl3cya san
03:18:16XavierGr_4.18am here
03:18:51 Nick San is now known as San||Away (
03:28:06 Nick _pill is now known as pill (
03:41:42 Join amiconn_ [0] (
03:41:59 Join burnhamd [0] (
03:47:05 Quit ReKleSS (
03:47:26NJoinReKleSS [0] (
03:53:18 Nick Lost-ash is now known as ashridah (n=ashridah@
03:54:44 Join NINJ4 [0] (
03:54:51 Nick NINJ4 is now known as Tandoc (
03:59:12 Quit amiconn (Read error: 110 (Connection timed out))
03:59:12 Nick amiconn_ is now known as amiconn (
04:02:47 Quit PaulJ_ (".")
04:11:52 Quit lostlogic (Remote closed the connection)
04:11:59 Join lostlogic [0] (
04:12:09 Nick _CoCoLUS is now known as CoCoLUS (
04:13:58 Join pengo [0] (i=xtofu@
04:23:26 Quit Tandoc ()
04:31:14 Join Jungti1234 [0] (n=jungti12@
04:35:12 Join ehntoo [0] (
04:40:02 Quit actionshrimp ("a bird in the bush is worth two in your house")
04:41:55 Quit Jungti1234 ("bye")
04:51:24 Join adamti91_ [0] (
04:53:50 Join adamti9999 [0] (
04:54:12adamti9999apparantley adamti91 is me,even though ive nevver been here before
04:54:22 Part adamti91_
04:54:26lostlogicI have written a perl program to compare two profiling runs, because I am cool that way.
04:54:28adamti9999anyone here?
04:54:59adamti9999my website has posted a VERY simple step by step tutorial on compiling rockbox for the h300 series if any one is interested
04:55:41Drumr|awaythats a really good compiling guide
04:55:48Drumr|awaynite all
04:56:00 Nick Drumr|away is now known as DrumRBoy|Away (
04:56:04adamti9999ill talk to you in MR
04:56:15adamti9999i was gonna give em my guide
04:56:39adamti9999super simple step by step guide to compiling rockbox for the iriver h300 series^^
04:57:01DrumRBoy|Awayyea thats the one
04:57:11DrumRBoy|AwayPLEASE disregard the first link
04:58:11 Join yngwi [0] (
04:58:32 Quit yngwi (Client Quit)
05:00:44 Quit adamti9999 ("CGI:IRC")
05:01:08 Join Rob2222_ [0] (
05:02:33 Quit ReKleSS (
05:04:45 Quit Kyl3 ("( :: NoNameScript 3.81 :: )")
05:07:14 Quit burnhamd ("Leaving")
05:07:27NJoinReKleSS [0] (
05:12:19***Saving seen data "./dancer.seen"
05:17:56 Quit Rob2222 (Read error: 110 (Connection timed out))
05:24:55 Join Jungti1234 [0] (n=jungti12@
05:25:54Jungti1234How about this player?
05:43:45 Quit RotAtoR ()
06:20:49 Join perplexity [0] (n=joust@
06:26:10lostlogicw00t I'm able to consistently cause reductions in ticks used.
06:29:30lostlogica whopping 1% improvement in ticks spent in codec during profiling
06:29:43lostlogicI'm pretty sure it translates into a greater real world impact in this case, but I can't swear to it
06:30:05DreamTactix291every little bit counts
06:30:46lostlogic(I saw what seemed to be a 5% reduction in boost usage on one song at 90mhz, but that could have been the difference between constant contention with other threads and just enough free time for other threads)
06:31:17 Quit Jungti1234 (Read error: 104 (Connection reset by peer))
06:33:54lostlogicalmost done with last profile run of the night and wishing it didn't take 2x song length+.
06:34:25DreamTactix291i kinda wish i knew more about this stuff to help you guys out
06:34:38DreamTactix291but i do what i can and lead the poor newbies on how to encode stuff that sounds good
06:34:45DreamTactix291that's my expertise i suppose
06:34:58lostlogicI didn't know much about a lot of this stuff before 2 weeks ago ;)
06:36:19DreamTactix291i've been researching audio codecs for like 2 years now
06:36:21DreamTactix291maybe longer
06:36:53DreamTactix291i just know nothing about programming
06:37:02DreamTactix291aside from being able to do simple stuff like "hello world"
06:37:09lostlogicI mean I have some schooling and experience with programming, just the specifics of rockbox and profiling I've learned lately
06:37:35lostlogicoops, last change was not better *scratch*
06:38:17lostlogicnow to build a non-profile build so I can play music and go to sleep, night!
07:10:20 Nick ashridah is now known as Lost-ash (n=ashridah@
07:12:23***Saving seen data "./dancer.seen"
07:50:31 Join ryan_ [0] (
07:51:35 Quit ryan_ (Client Quit)
09:12:26***Saving seen data "./dancer.seen"
09:25:29 Join Hooligan [0] (
09:58:28 Join TCK- [0] (
10:01:34 Join Vlad0man [0] (
10:01:56 Join Acksaw [0] (
10:02:03Acksawany developers here
10:02:41 Quit Vladoman (Read error: 110 (Connection timed out))
10:03:02Acksawor anyone?
10:04:18HooliganI'm here. Not a dev though
10:06:05Acksaw i wanna do this for my iriver
10:07:24HooliganI had wondered if such a thing were possible.
10:07:38HooliganWith Rockbox, I'm sure it'd be possible if someone wants to code it.
10:08:15HooliganIs Audioscrobbler open-source or not?
10:08:50Slasheriaudioscrobbler has specs on it's website how to communicate with the service
10:08:54Slasheriit's very simple
10:09:08HooliganAh, okay.
10:09:15Acksawwhat i wanna do is
10:09:25Acksawget all the songs i play on my iriver
10:09:40Acksawto, when connected via USB, upload to audioscrobbler
10:10:07Slasherithat is possible to do, but we cannot get the play time on units without a rtc
10:10:22Acksawplay time?
10:10:42XavierGr_Slasheri do you know how big is the plugin memory?
10:11:13XavierGr_for iriver and then archos targets.
10:11:16SlasheriXavierGr_: just a moment
10:11:50 Nick XavierGr_ is now known as XavierGr (
10:11:59 Quit TCK (Read error: 110 (Connection timed out))
10:12:14Acksawis what i want possible?
10:12:24DBUGEnqueued KICK XavierGr
10:12:24XavierGrAcksaw: I think yes
10:12:28Acksawthanks good!
10:12:40Acksawshould i make a topic on MR?>
10:12:43SlasheriXavierGr: ../firmware/export/config-h120.h:#define PLUGIN_BUFFER_SIZE 0xC0000
10:13:16Slasheriso 768 KiB
10:13:20XavierGrhm ~786kB
10:14:53***Alert Mode level 1
10:14:53XavierGrwoa only ~32kB for Archos targets!!!
10:14:59Slasherihehe :)
10:15:27***Alert Mode level 2
10:15:27XavierGrit would be very difficult to make the jpeg viewer not to have limits.
10:15:34***Alert Mode level 3
10:15:34XavierGrespecially sorting will be like hell.
10:15:55Slasheritrue.. but you can use the audio buffer
10:16:07Slasheribtw, why your text suddenly turned into red? :)
10:16:23***Alert Mode level 4
10:16:23XavierGrI want this as a last resort, because I will try to make it view and play if the pic can fit the plugin memory
10:16:51XavierGrops yes you are right but why, I didn't wrote anything for it.
10:16:57Slasheriah, that will be challenging then ;)
10:17:20XavierGrI had an idea to put the file list into a physical file instead of memory
10:17:35XavierGrbut the sorting of this list will be a hell if I dont use the audio buffer.
10:18:10XavierGrof I leave it unsorted it works pretty well.
10:18:17SlasheriHmm, but dir tree viewer has already the file names sorted, maybe you could use that
10:18:47XavierGrhmm yes you are right but how can I access these data from a plugin?
10:18:56Slasherihehe, i don't know :)
10:19:22XavierGrit would need many extra fn into the plugin api
10:20:10XavierGralso I have a problem with the list, when I try to read a line from the file.
10:20:16XavierGrI can use 2 ways.
10:20:52XavierGrread_line() and I dont care about the lenght of each filename, con is that I have to run from the start of the file to go to preffered line.
10:21:47XavierGrI can use fixed lenght filenames (padded with 0) and then use lseek to goto a specific line, then no loops.
10:22:17XavierGrBut is there a way to delete aa specific character from a file?
10:22:27 Quit Maxime` ()
10:23:52XavierGr(this is needed for pictures that the viewer can't open, I will mark each line with a letter, but then If I add a letter to the line I will break the fixed length structure)
10:26:24***Alert Mode OFF
10:26:31 Join Maxime [0] (
10:28:47perplexityXavierGr: Why not use a heap for the file names.. with variable length, and just have a fixed size table with pointers to each name ? How many names are you looking at?
10:30:35perplexityOnce you have the file names cached in memory then your sorting and organising is limited to shuffling the pointers around in the table.. the filenames never move..
10:33:00perplexityIf you had a pointer table for 5000 files, you are only looking at 20kb of ram for the pointers, the rest could be used as filename storage heap.. (on the iriver that is a huge amount of files)
10:34:20 Join TCK [0] (
10:37:02 Join Membrillo [0] (
10:38:41perplexityUsing a list configured like that it would also be dead easy to sort in place using a quicksort or even a bubble sort.. no shuffling required, just swaping 32bit pointers
10:40:50 Join Gacel [0] (
10:40:57Gacelgood morning
10:41:22MembrilloYo Brosef
10:41:24 Join muesli_- [0] (
10:44:10XavierGrperplexity I need a static buffer for that
10:44:24XavierGrthen I am limited again.
10:44:25perplexityYep.. so why not allocate one?
10:44:45XavierGr1000 file * 30 letters = 30KB
10:45:03XavierGrI did that before and I was very limited on archos models
10:45:18perplexityWhy do you need to define a filename length?
10:45:39perplexitySome may be 10 letters, some may be 50.. with a heap buffer and pointer list it does not matter how long they are
10:45:49XavierGrwell how can I allocate a free buffer without using get audio buffer?
10:46:06perplexitySorry, I thought you were going to use the plugin buffer
10:46:25XavierGrI am going to use the plugin buffer.
10:46:30 Join LinusN [0] (
10:46:46LinusNperplexity: the plugin buffer is 32kbytes on archos
10:46:47perplexityIn which case I'm confused (not suprising really)
10:46:49 Quit TCK- (Read error: 110 (Connection timed out))
10:47:03XavierGrperplexity: I want to use the audio buffer only if the picture is very big
10:47:11perplexityyes, I did see that LinusN.. thanks..
10:47:25LinusNXavierGr: you can forget about loading jpegs in the plugin buffer on archos
10:47:30perplexityXavierGr: So if you want to view a big picture you stop playback, but otherwise playback continues ?
10:47:49XavierGrpreplexity: yes something like that.
10:47:51perplexityHow many levels of grayscale does the Archos handle?
10:48:03LinusN30 something
10:48:19perplexitySounds like an inconsistent interface XavierGr.. Like "why did my music just stop when I loaded the 10th picture, but the 1st 9 were all ok"
10:48:59XavierGrwell it is better than having explicitly to stop music. We can ofcourse ask before this happens.
10:49:13Gacelwhy not decode the jpeg directly to the screen? you just need a small buffer for every line
10:49:45XavierGrI don't have any knowledge of the JPEG decoding proccess...
10:49:58XavierGrI don't know if this can be done. Linus?
10:50:03perplexityI'm just wondering about the utility of loading a jpeg to a 30 something grayscale screen..
10:50:14Gaceli do ;) i did somthing like that on a palm once
10:50:18perplexityAre you scaling them on the fly XavierGr ?
10:50:43LinusNAcksaw: yes
10:50:52AcksawDid you see the thread i made on MR/
10:50:53Gacelscaling would require a few lines, but still not the full image
10:50:58XavierGrperplexity: I am doing this for fun and educational purposes. I am certain that I am not going to use the jpeg viewer much, But I would like to make it work
10:51:16LinusNAcksaw: better use my nick, else i won't be notified by my irc client
10:51:33perplexityGacel: I agree completely.. you can just decode it on the fly (provided it's not an interlaced jpeg)
10:51:43AcksawLinusN : did you see the thread i made on MR?
10:51:48LinusNAcksaw: what thread?
10:52:12LinusNGacel: XavierGr didn't write the jpeg decoder
10:52:38AcksawLinusn: i'll just find it
10:52:49Gaceli didn either, just used the offical one (which provides a line by lien mode)
10:53:02Gacelof course this will probably a different one
10:53:16perplexityHow does that go for jpegs that are progressively encoded Gacel ?
10:53:44Gacelgood question... it was a a vew yeras ago...
10:54:04Gacelcant remember
10:54:46XavierGrit could be way slow to decode line by line though...
10:55:17LinusNthe thing is that the jpeg viewer can zoom
10:55:30LinusNso it's not exactly trivial
10:55:46Gacelits challenging ;)
10:56:02LinusNit can still be done...
10:56:12perplexityanything is possible :)
10:56:25LinusNAcksaw: ah, that one
10:56:35Gacelthere are jpeg viewers that work acceptable on a 33mhz palm with 64k memory segment size...
10:57:11Gacelof course quite slow
10:57:40perplexityI don't see it as too much of a hassle Gacel, but the zooming is more of a challenge.. especially with the 32k plugin buffer on the Archos
10:58:44LinusNAcksaw: rockbox can gather runtime data
10:59:16AcksawLinusN: so thats a yes?
10:59:25LinusNyes, of course it is possible
10:59:28LinusNi gotta go
10:59:31LinusNcu later
10:59:33 Part LinusN
11:00:28AcksawNow just to find someone who can do it..
11:00:32XavierGrI will finish my "loading from disk" approach to see performance.
11:00:38XavierGrYou can do it
11:00:51perplexityI think given the limitations you face with memory, thats a pretty good way to do it XavierGr
11:01:16 Quit San||Away (Read error: 110 (Connection timed out))
11:01:40XavierGrThe performance bottleneck may come when sorting. But I will try to make it fast.
11:01:52perplexityThe only thing I could suggest to speed it up is a similar index approach.. once you have the file written out you could generate an index for it so you know exactly where to lseek to for each record number, and could sort using that list instead
11:01:57 Quit Membrillo ()
11:02:44XavierGrlseek could be nice but also a dead end on my plan...
11:03:56XavierGrundortunately the jpeg viewer can't load many different formats so every time it founds a non compatible jpeg it justs exits. In my plan I want to mark those instances so the first encounter the user is noticed and then the program skips those entries
11:04:16XavierGrto do that I will add a specific byte to the line the file is in the list.
11:04:43XavierGrTo use lseek I must write the filenames in a constant lenght (E.g padded with 0)
11:05:12XavierGrbut when I will have to insert that specific byte the whole const lenght structure will broke.
11:05:17XavierGrny ideas?
11:05:27perplexityNo.. you write out all file names sequentially terminated with /0 and then you build an index to all the names.. if a file is not compatible you remove it from the index.. but leave it in the file on disk
11:06:01perplexityBut if you must to it your way, then just write the file name out with a preceeding byte.. "a/file/name/here"
11:06:12perplexityand if it's not compatible change it to "b/file/name/here"
11:06:17XavierGrah so you mean to make a int array which will have the byte position of each entry?
11:06:25perplexityexactly :)
11:06:38XavierGrthat seems good!
11:06:44 Quit Hooligan (Read error: 104 (Connection reset by peer))
11:06:45 Join ender` [0] (i=ychat@
11:06:49perplexityYou could create it on the fly, or you could scan the file after you have created it and do it fomr htere..
11:07:08perplexityIf you use the array, you then just sort the pointers in the array, no messing about with the on disk file anymore..
11:07:38XavierGryup that sounds great thanks.
11:07:46perplexitywelcome :)
11:07:52perplexityI have to run now.. bbl
11:09:26 Quit Cassandra (Read error: 104 (Connection reset by peer))
11:11:37 Join Hooligan [0] (
11:12:27***Saving seen data "./dancer.seen"
11:13:46 Join Kohlrabi [0] (
11:16:59 Quit Acksaw ()
11:32:25 Join Jungti1234 [0] (
11:33:40 Quit Jungti1234 (Client Quit)
11:43:59 Quit Rob2222_ (Read error: 110 (Connection timed out))
11:46:47 Join mirak [0] (
11:57:05 Join markun [0] (
12:09:30 Quit muesli_- (Read error: 110 (Connection timed out))
12:23:06 Join muesli_- [0] (
12:29:27 Join linuxstb [0] (
12:35:12 Join DangerousDan [0] (
12:38:17 Quit muesli_- (Read error: 104 (Connection reset by peer))
12:40:21 Quit Hooligan (Read error: 104 (Connection reset by peer))
12:41:36 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
12:42:30 Join San [0] (n=Test@
12:44:28 Join NicoFR [0] (
12:46:04 Join psych1 [0] (
12:47:07 Join cannard [0] (
12:47:14 Join KN|stiff [0] (
12:47:53 Join Febs [0] (
12:49:47 Part psych1
13:03:03 Quit San (Read error: 110 (Connection timed out))
13:04:23 Join Acksaw [0] (
13:04:39Acksawis it just me or can i see the screen with the backlight off again on my h300?
13:04:42Acksawwith rockbox
13:05:29 Nick NibbIer is now known as Nibbler (
13:12:30***Saving seen data "./dancer.seen"
13:26:31 Join muesli_- [0] (
13:28:02 Join webguest42 [0] (
13:30:01webguest42hi @ll, I am looking for a guide or some tips to upgrade my gmini XS 200 with a 40gb hdd . But I am already stuck with the ZIF interface and the way to connect it my pc to clone the orinigal hdd content. Anyone knows anything?
13:38:06 Join Zak1392 [0] (
13:38:21Zak1392hey guys
13:39:15Acksawhey zak
13:39:46Acksawhows the family?
13:40:17Zak1392is there any docs out there that shows me how to compile rocks?
13:40:32Zak1392what family ;)
13:40:41Acksawyeh there is
13:40:44XavierGrto compile or to program?
13:41:00Acksawits called eli_sherar
13:41:35XavierGrand /docs/plugin_api for the interface
13:44:08XavierGrdownload the source
13:44:31XavierGrand then you will find a folder called docs
13:44:38XavierGrthere are some files with various information
13:44:49XavierGrthough some are obsolete.
13:45:25Zak1392downloading now
13:45:38Zak139229secs remaining
13:47:23Zak1392i don't see plugin_api...
13:48:37Zak1392i won't worry about it then
13:49:05 Quit edx ()
13:54:00 Join muesli- [0] (
14:00:27XavierGrHelp!! Why if I call plugin_get_buffer twice my first buffer is lost?
14:00:51XavierGrI thought it overwrites FREE memory not one that I use.
14:01:20 Quit muesli_- (Read error: 110 (Connection timed out))
14:02:48Zak1392[Bill Clinton voice] Maybe because you're so fat! Hahahahaha!
14:02:58Zak1392but really, i wouldn't know
14:08:11linuxstbXavierGr: plugin_get_buffer() just returns a pointer to the memory in the plugin buffer not used by the plugin itself. If you call it twice, it will give you the same value.
14:09:07linuxstbYou can read the implementation in apps/plugin.c
14:09:36XavierGrwell I call it first on a function to get an int buffer, then on another function it is called to get a char buffer. Isn't there a way to make the function allocate space on different adresses?
14:10:26linuxstbNo - you would have to write your own simple malloc system - i.e. call plugin_get_buffer() once at the start of your plugin, and then allocate memory from there inside your plugin.
14:11:05XavierGrwhat if I add at the return value the range of adress that the previous buffer uses?
14:11:14SlasheriHmm, tagcache is already capable browsing artists :) But i am afraid that it wont replace tagdb because it requires too much processing power on archos :/
14:11:44XavierGrSlasheri: Optimize it :\
14:12:11linuxstbXavierGr: The buffer_size parameter passed to plugin_get_buffer() is only used to return the amount of memory available. You don't pass a buffer size to the function.
14:13:06XavierGrwait I thought that the parameter was to ask for a specific amount of bytes.
14:13:24linuxstbRead the source to that function in apps/plugin.c
14:13:44linuxstbYou will see that buffer_size is only written to, never read.
14:14:15XavierGrbut I don't get it. Let's say I call it with a value of 10. What happens?
14:14:28XavierGrwhat's the return value.
14:14:40XavierGrI will look there to see if I can get it.
14:14:58linuxstbThe return value of the function is a pointer, and *buffer_size is set to the amount of memory in the plugin buffer not used by the plugin itself.
14:15:29linuxstbAs I said, *buffer_size is never used - it's just set to the size of the buffer the function returns.
14:16:45XavierGrconfusing, I will read the source then maybe I will bother you a little more if I don't get it.
14:16:51 Join muesli_- [0] (
14:17:41linuxstbJust think of the function as not taking any parameters and returning two values - a pointer the buffer, and an integer containing the size of that buffer.
14:18:03linuxstbIf you call the function 10 times, you will always get the same two values returned.
14:18:27XavierGreven if I change the parameter?
14:18:32 Join webguest25 [0] (
14:18:36linuxstbYes. The parameter is ignored
14:18:44 Join Membrillo [0] (
14:19:18XavierGrThat could be tricky then
14:19:32linuxstbAs I said, you will have to manage that buffer yourself inside your plugin.
14:19:41XavierGrI though that it was like a malloc, but then I was very confused that there was no free.
14:19:41webguest25Slasheri: perhaps its worthwhile just optimising the present tagdb apps, *shrugs* I uae filetree personally
14:19:54linuxstbXavierGr: No, it's not like malloc at all.
14:20:17Slasheriwebguest25: converting it to ram version would be *extremely* painful
14:20:29linuxstbWhen your plugin is loaded, there is some spare memory at the end of the plugin buffer - that function just returns a pointer to that buffer, and tells you how big it is.
14:20:52Slasherithe ram version is now already working pretty well but the disk only access and cache building on slow cpus needs still some designing
14:20:58XavierGrah that more clear now.
14:21:42webguest25Slasheri: will your db app replace the others then ?
14:21:57Slasheriwebguest25: i haven't think about that yet
14:22:22linuxstbSlasheri: Is "tagcache" code completely independent from "tagdb" then?
14:22:22Slasheriat least it doesn't need any external tools to build the db
14:22:28webguest25I dont understand the benefit of your app ?
14:22:34MembrilloZak1392: you think the newer version is better?
14:22:54Slasheriwebguest25: the database is created on the player itself and loaded into ram (if enough ram is available)
14:23:09XavierGrok then my previous statement was right. If it returns a pointer to the start of the trailing (unused) plugin buffer, then, If I add the previous address range to the pointer I can use the rest of the buffer.
14:23:17Slasheriit allows very fast database searches and boot time, without needing pc apps to update the database
14:23:27webguest25will it support full artist/album/year etc browsing ?
14:24:05Slasheriwebguest25: currently artist/album/genre/song browsing is being implemented
14:24:21Slasheriand that will be as fast as browsing directories with dircache enabled
14:24:23 Quit muesli- (Read error: 110 (Connection timed out))
14:24:33webguest25sounds like it'll be very comprehensive then
14:24:54linuxstbXavierGr: Yes, I think so. That's what I meant by managing the buffer yourself. You will need to keep track of how much of the buffer you have allocated to different variables.
14:26:04XavierGryes, fortunately the first call is a fixed buffer of integers called only once at the start of the program.
14:26:48 Quit Zak1392 ()
14:27:14linuxstbXavierGr: If it's fixed, then you can just allocate it statically in your plugin - i.e. int myarray[ARRAY_SIZE];
14:30:01XavierGrlinuxstb:1) Yay it's working like a charm. 2) No I can't. The size of the array is given by how many jpeg the user has in the current folder. And that cannot be clculated at compile time.
14:30:59linuxstbXavierGr: In which case, it isn't "fixed". But I'm glad you've got it working.
14:31:21XavierGr(yeah sorry, fixed I mean per running time)
14:32:03linuxstbIs the iriver's 768KB plugin buffer big enough for most of your jpegs?
14:32:25 Join San [0] (n=Test@
14:33:30XavierGrlinuxstb: Yes I haven't encounter yet a jpeg that will not fit into memory.
14:33:48XavierGrBut I will add a dialogue to stop music if the picture is very big.
14:34:05Sanjpeg viewer?
14:34:38Sanhows it going?
14:34:54XavierGr(and I have the most versatile right now, and limitless, if targets memory can handle it)
14:35:06XavierGrSan: there is already one.
14:35:25XavierGrI just made some modifications to be able to change files on the fly.
14:35:27Sanbut, does it work?
14:35:43XavierGr33 shades of grey :)
14:36:15NicoFRhow hard would it be to adapt it for the H300 ?
14:36:54 Part webguest25
14:38:29 Quit steveb ("Reconnecting")
14:38:30XavierGrNicoFR: Very hard, at least for me. I don't know a thing about jpeg pictures.
14:39:04 Join steveb [0] (n=dave@
14:40:45Sananyone wana make a simple B&W PacMan and then I will finish it off?
14:41:34linuxstbSan: I would like to see a pacman emulator which runs the original pacman arcade ROMs. But I haven't found a good starting point.
14:41:58Sani have the source of a very basic pacman
14:43:04Sani am going to try and port it as a plugin
14:43:21 Quit Membrillo ()
14:44:22Sanwhat is a .bas file
14:46:08SlasheriXavierGr: in fact now i got an idea how to make it work on archos too =) But that would limit maximum artist/album/genre count to 2000-3000 because archos has a small plugin buffer. But that limit should be enough for users
14:46:59Slasheri(that limit applies only to building the db on the player itself)
14:47:13XavierGr3000 artist is enough I think. I have merely 4000 tracks on my 60gb disk
14:47:28Slasherihehe, nice :)
14:47:55XavierGrSlasheri: Then if the only limitation is the OTG build it won't be a problem.
14:48:10XavierGrThose that don't like it should buy more RAM. :)
14:48:11Slasheriyes, there is no other limitations
14:48:17Slasherihehe :)
14:49:29 Join linuxstb_ [0] (
14:49:41 Quit linuxstb (Nick collision from services.)
14:49:44 Nick linuxstb_ is now known as linuxstb (
14:56:13 Join willows_mine [0] (
14:56:17willows_minehey guys
14:56:23willows_mineusing rockbox on h140
14:56:57willows_minehaving trouble figuring out how to split mp3s with it... does the h140 version even have the splitter? Or do I need to update it or something
14:58:17 Join Membrillo [0] (
14:59:10linuxstbwillows_mine: No - the splitter isn't implemented on the 140
15:00:03XavierGrplugins are very time consuming in means of porting progress.
15:00:12XavierGreach one has to be modified for every new target!
15:00:15willows_minedamn... I have like an hour long mp3 file that I want to split and I don't know what the best way to do it would be... any good programs you know of?
15:00:37XavierGryes more simple
15:00:45linuxstbWill soundforge re-encode the MP3?
15:01:03XavierGrif you have the mp3 plugin
15:01:18 Join actionshrimp [0] (
15:01:42willows_minethanks man! x
15:01:47linuxstbmp3directcut doesn't re-encode, so is probably the best choice.
15:01:58willows_mineok I'll check it out
15:02:33 Join muesli- [0] (
15:03:49XavierGrlinuxstb: I think that soundforge will not reencode if it is the same file. (E.g if you don't cut and copy somewhere else) Anyway just details, if you can find a liitle free program that the best
15:03:52 Part XavierGr
15:08:53 Quit Acksaw ()
15:12:33***Saving seen data "./dancer.seen"
15:16:24willows_mineXavier... lol... it seems a bit complex for me. I just wanna go to certain points and divide the track into individual tracks... fancy giving me some tips? lol
15:17:44 Join Rob2222 [0] (
15:20:35 Quit muesli_- (Read error: 110 (Connection timed out))
15:20:42linuxstbI've changed the RTC driver a little for the Archos (I've tested it, and it still works fine on my recv1) to make it easier to add RTC support for the new targets. Does anyone object to these changes? (amiconn, LinusN?)
15:20:45linuxstbIt's here:
15:30:06Membrilloanyone here have windows 2000 sp4?
15:30:21MaximeI had Membrillo, why ?
15:30:32Membrillodo you have it now
15:30:41Maximeno sry
15:30:45Membrilloah ok
15:30:45lostlogic*goes for some more agressive tremor opts*
15:32:06 Quit linuxstb (Read error: 104 (Connection reset by peer))
15:32:16 Quit NicoFR (Read error: 104 (Connection reset by peer))
15:34:15Membrillodoes anyone here have access to any windows builds other than XP?
15:37:31Gaceldefine acces? i am working un w2k right now
15:38:33lostlogicwith latest opts, it almost plays while profiling! I hope that means similar improvements in practice!
15:39:58 Quit San (Read error: 110 (Connection timed out))
15:41:44 Join linuxstb [0] (
15:43:45 Part willows_mine
15:45:17 Quit muesli- (Read error: 110 (Connection timed out))
15:45:28 Join webguest53 [0] (
15:45:49 Join Jungti1234 [0] (n=jungti12@
15:46:11 Join San [0] (
15:46:37webguest53Slasheri: have you seen this
15:47:09 Part webguest53
15:49:47lostlogicI'm sure that this is overestimated by the reduction in calls profiled, but total time in codec recorded before: 04814886 and after: 02250481
15:50:42Slasheriyes, i just tried that dircache thing but it seems to be a bug elsewhere. Even with dircache disabled, rockbox crashes with an address error on long file names in root
15:52:18 Quit Jungti1234 ("bye")
15:58:49preglowwoah, 768kb for plugins?
15:58:52preglowthat's quite a lot
15:58:56 Join akaidiot [0] (n=nope@
15:59:08crwlyes, 640 kB should be enough for everybody
15:59:35 Quit KN|stiff (Read error: 104 (Connection reset by peer))
15:59:58 Join KN|stiff [0] (
16:01:20lostlogicI'm going to be so sick of listening to my test song by the time I'm done optimizing Tremor.
16:02:17preglowyou'll be sick of optimising tremor as well
16:02:35lostlogicpreglow: btw, I wrote a profile output comparator in perl, so now I generate the output files in Java and compare them in perl, just to drive you insane.
16:03:18preglowthe clever thing to do would be writing the comperator in c#
16:03:25preglowor no, visual basic!
16:03:34lostlogic <−− shows differences between two profiling runs per funciton tick / call diffs (sorted by total ticks and total calls)
16:03:48lostlogicpreglow: :-P yes yes, I'm thinking about rehacking the parser into perl unless you beat me to it.
16:04:07lostlogicwow, plays without skipping while profiling now, I think I'm winning.
16:04:19preglowthat's pretty nice
16:05:05linuxstblostlogic: Does that mean your profiling is getting more efficient, or Tremor?
16:05:10lostlogicI'll crank CPU freq back down to 90mhz and test my torture song again momentarily to see if I've killed the 90mhz skips and stuff
16:05:14lostlogiclinuxstb: combination in this case
16:05:57lostlogiclinuxstb: started playing at 1/2 speed yesterday when I stopped profiling inlines, and today I inlined 2 more often called functions, which got me down to a few skips, then I made one more IRAM change and got to zero skips.
16:06:31linuxstbSounds good. I may even start encoding some of my FLACs to Vorbis soon.
16:06:46 Join muesli_- [0] (
16:06:54lostlogicI'll let you know my qualitative 90mhz torture test results.
16:07:45 Quit San (Read error: 110 (Connection timed out))
16:07:51 Quit Membrillo ()
16:08:03preglowi have only vorbis music
16:08:18preglowso this is right up my alley indeed
16:09:21 Join XavierGr [0] (
16:09:37XavierGrHi again
16:11:04lostlogicgah, too damn cold in my house, can't code with frozen fingers
16:11:23XavierGrcan't say the same
16:11:45XavierGrmy window is open at this time.
16:15:01 Join Lear [0] (
16:15:22CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
16:15:22*linuxstb remembers his student life in Scotland - coding with gloves on
16:15:48lostlogicdamn, still 3 skips on the whistles in this song, but the boost ratio's fallen another few percent
16:16:11XavierGrwhy on the whistles?
16:16:28lostlogicapplause + whistles are large on transient noises which are the hardest to encode.
16:16:47preglowdoes the iriver firmware handle all oggs without skips?
16:16:58lostlogicI've never heard a skip on iRiver, certainly not on this song
16:17:14lostlogicYES! I've hit 90% boost on this song, previous low was 94!
16:17:28markunlostlogic: did you replace the mdct by fft and sorting/rotating?
16:17:28XavierGrI think it has some problems with low bitrate ones.
16:17:39lostlogicmarkun: no, all I've done is move things in IRAM and inline a few things
16:18:14preglowmarkun: someone has started working on replacing the imdct
16:18:19preglowi believe it was perplexity
16:18:49preglowhe's already replaced it with a naive approach, and it works
16:19:04Learlostlogic: what did you move?
16:20:45lostlogicLear: I've re-arranged pretty much everything −− I moved all data to IRAM except those lookup tables you said are not used by current libvorbis, and then took all functions out, and moved them back in based primarily on how much cache thrashing they'd cause.
16:21:23lostlogicLear: after that I merged codebook.c and bitwise.c so I could inline oggpack_look and oggpack_adv because they are called way too much to remain function calls.
16:23:32lostlogicI think the best thing that I moved to IRAM was the arrays of pointers that the codec moves around and calls through a lot
16:24:29preglowso, vorbis will be one of the few codecs we have which does not keeps its output buffers in iram
16:24:38preglowi wonder how that'll turn out when we do more dsp
16:24:55lostlogicI thought the buffers were in IRAM?
16:25:16lostlogicsynthesis.c:static ogg_int32_t *ipcm_vect[CHANNELS] IBSS_ATTR;
16:25:16lostlogicsynthesis.c:static ogg_int32_t ipcm_buff[CHANNELS*IRAM_PCM_END] IBSS_ATTR LINE_ATTR;
16:25:52preglowgimme a sec...
16:25:54 Join tucoz [0] (
16:26:04preglowi distinctly remember having done that only in a local copy
16:26:09preglowi must have been really pissed or something
16:26:32XavierGrnewb question: what's an inline function?
16:26:38markunI can't drink and code at the same time
16:26:41preglowXavierGr: do you know how macros work?
16:26:47lostlogiccomplete torture song, 89-90% boost, down from 99 when I started this whole project... still 4 skips though :(
16:26:57tucozhey, preglow and linuxstb. thanks for the support. How do you think we should arrange this? (LaTeX)
16:27:08preglowmarkun: oh, me neither, but i do try sometimes
16:27:23preglowtucoz: well, nothing big, you've already got stuff done, right?
16:27:28XavierGrpreglow: Faintly...
16:27:33linuxstbtucoz: Basically, one file per chapter sounds sensible - and I think you've done that already.
16:27:35 Join San [0] (
16:27:52linuxstbWe need to work out how to produce the different manuals for the different targets though.
16:28:01preglowXavierGr: the function isn't called at all, the code of the function is instead moved to every place it is called
16:28:01markunpreglow: I work in a very small company where it's normal to drink beer after 17:00, but every time I try I have to stop coding after a few minutes :)
16:28:20preglowit doesn't exactly do wonders for the concentration
16:28:27tucozhmm, it should be possible with a pre-processor.
16:28:41preglowi say let's not worry about this now
16:28:50preglowa solution will present itself, and i sincerely doubt it'll break what we're doing
16:29:02tucozBut, maybe we should have one file per chapter, and do it from within LaTeX
16:29:04XavierGrpreglow: and what is there much gain in performance with this?
16:29:21preglowXavierGr: you don't have to call the function anymore, and that takes a bit of time if the function is called often
16:29:22XavierGr^(remove what)
16:29:24 Join petur [0] (
16:29:36preglowtucoz: what do you mean?
16:29:50XavierGrhmm and are there any drawbacks?
16:30:14tucozpreglow, instead of having a iriver install chapter, one could have a install chapter and hopefully there are conditionals within latex
16:30:36lostlogicXavierGr: larger code size, because the code is duplicated wherever it was 'called from'
16:30:36linuxstbXavierGr: The drawback is that the code is duplicated - increasing the size of your program.
16:30:53preglowtucoz: yes
16:30:58preglowtucoz: there are, btw
16:31:04tucozpreglow, oh, great :)
16:31:05preglowtucoz: and if not, we'll just use the preprocessor
16:31:29lostlogicfortunately, Tremor is actually quite small in code size, so that doesn't look like it'll cause a problem, plus the functions I inlined only had a couple of call sites (3 and 2 I believe).
16:31:30preglowtucoz: doing it any other way would not be productive
16:31:48tucozpreglow, that might be true, yes.
16:32:05preglowtucoz: the chapter files for some platforms would be more updated than others, etc
16:32:35preglowso yes, one file per chapter, with some parts differing from platform to platform
16:32:49tucozok, sounds good.
16:33:32preglowindeed it does
16:34:03tucozthat will be somewhat similar to rockbox code as well ;)
16:34:07linuxstbtucoz: Do you have CVS access?
16:34:11tucozlinuxstb, no
16:34:27preglowbut ok
16:34:31preglowdo you have this lying around somewhere?
16:34:33preglowi'd like a quick look
16:34:58tucozit's a link in the forums.
16:35:01tucozI think
16:36:00preglowwhere did you get the makefile?
16:36:28tucozIt is something we use at school. I am writing my master's thesis in latex
16:36:39tucozThe LaTeX-rules file?
16:36:49linuxstbMaybe a useful approach would be to have separate rockbox-h100.tex, rockbox-player.tex, rockbox-recorder.tex etc files which are the main LaTeX files for each manual. These would specify which chapters are applicable to each target, and could define some target-specific macros.
16:36:53 Quit Shaikoten (Read error: 110 (Connection timed out))
16:36:57lostlogicLear: are you interested in telling me what data I've moved to IRAM that was just plain foolish to move there? I don't actually know things about the vorbis encode/decode process really, and you do.
16:36:59preglowtucoz: is it ok to redistrute it?
16:37:09tucozlet me check
16:37:24Learlostlogic: I'm not sure I know that much more than you, I must say... :)
16:37:42lostlogichehe, well how did you fidn out that some of the lookup tables aren't used in current libvorbis encoding?
16:38:12LearI played around with adding most/all stuff to iram, and inlining the most heavily called functions in bitwise.c, but didn't manage to see much difference...
16:38:45LearDownload all releases of libvorbis and see which block sizes they use.
16:39:05lostlogicnice nice −− having profiling results to compare definitely helps ;)
16:39:11preglowlinuxstb: well, that would pretty much only make sense if we were to include different files for different targets, and i don't think we'll be doing that
16:39:33linuxstbSome chapters may only be applicable for some targets - such as FM radio or recording.
16:39:52linuxstbAnd we need a way to define macros for different targets.
16:39:55tucozpreglow, . I think that is the origin of the file
16:40:08preglowlinuxstb: so, did you find out how to do that from within latex?
16:40:45linuxstbNo, but I remember being able to define macros.
16:40:51linuxstbAnyway, I've got to run now, bbl.
16:41:11preglowyup, but if we end up having to use the c preprocessor, then that approach won't work well
16:41:14preglowso we need to figure that out first
16:41:15lostlogicLear: so I basically just did the next logical step from what you'd done by inlining most of bitwise.c into codebook.c, and moving more data to IRAM and applying cache trashing logic to the code IRAM strategy.
16:46:11preglowoh, and btw
16:46:16preglowtucoz: what charset to use?
16:46:23lostlogic*sigh* still haven't managed to break below the 50% boost level for the torture album at 124mhz.
16:47:11tucozhmm, not sure. when I copied and pasted from the webpage I hade to set that to something. iso-latin1 I thing
16:47:14 Join keilj [0] (
16:47:32 Quit San (Read error: 110 (Connection timed out))
16:47:36keiljlo people
16:47:37 Join NicoFR [0] (
16:47:44LearThat torture album, what quality isn't encoded at?
16:47:49tucozno-escape also I believe.
16:47:51Lear*is it
16:48:41lostlogicLear: just 7, it's Dave Matthews Band - Live At Red Rocks −− lots of transients, and lots of applause
16:50:05keiljwell this is chan is moving slow...
16:50:19LearWell, current version uses about 50-55% for q6 files, I think.
16:50:55preglowi wish they'd come up with vorbis 2 soon
16:51:08lostlogicwhat's going to be so great about vorbis?
16:51:12LearBut using an FFT could make quite a difference, if the FFT is good. I tried the patch for the lowmem branch that does that, and it was around 30% faster...
16:51:25XavierGrwhat? Is there a plan for vorbis 2 codec?
16:51:41LearBeen talked about for years, actually. :)
16:51:50preglowthis time around i'm willing to bet it wont be so embedded unfriendly
16:52:33Learremove some things that didn't turn out too well in practice, and try some new stuff too.
16:53:26lostlogicwell I'll submit my current optimizations to patch tracker, because they are definitely better by some percentage... it'll be up to y'all with CVS if they are worth including.
16:53:39keiljim off to bed, just like to say that everyone working on rockbox has done such a great job, im looking forward to seeing what becomed of it
16:54:18 Quit keilj ()
16:55:01 Join webguest78 [0] (
16:56:03tucozpreglow, do you think it looks ok?
16:56:21tucoznot the content, but the setup.
16:56:50preglowyeah, looks decent
16:58:09tucozgood. I got to go, see you
16:58:35 Part tucoz ("Leaving")
16:59:55webguest78Slasheri: Look at yesterdays log at 19.44.35, this might be the reason for the directorycache failing at long filenames
17:00:39 Join San [0] (n=Test@
17:00:43 Join muesli- [0] (
17:00:45Sanif I made a plugin, how do I compile it so It is included in the H300 plugins?
17:03:15LearDrop it in the plugins folder, adjust sources, make. I think...
17:06:26XavierGrput the c file into plugins
17:06:37XavierGrthen add the filename of the c file in SOURCES
17:09:32 Quit akaidiot (Read error: 110 (Connection timed out))
17:12:36***Saving seen data "./dancer.seen"
17:17:00 Quit muesli_- (Read error: 104 (Connection reset by peer))
17:24:22peturAm I correct that the current way of manipulating the LCD doesn't use a buffer? That is: writing text somewhere erases that part of the LCD and then writes the new stuff.
17:25:46 Quit webguest42 ("CGI:IRC")
17:29:31 Quit pengo ("Curious? Open binary with Qubero")
17:40:01markunCan a li-ion charger also charge li-poly batteries?
17:40:42XavierGrI think yes
17:40:50XavierGrit desn't matter.
17:41:01lostlogicsame voltage/current curve IIRC
17:41:07mirakhey what the V3 bootloaders brings ?
17:41:27XavierGrAmpage and voltage (and polarity) matters not chargers
17:42:44peturmirak: H3xx? fix for bootproblems, LCD off when backlight off (afaik)
17:42:54markunI was asking because the Gigabeat has a LTC4055 li-ion charger chip and I hope the battery can be replaced by a higher capacity li-poly battery in future.
17:42:57miraklcd remote you mean ?
17:43:11peturmain lcd
17:43:37mirakthere is a distinction between LCD screen and backlight on H300 ?
17:43:44mirakI though there was no backlight
17:43:53lostlogic... no, it's a backlit color LCD
17:44:13lostlogicyou just can't see the LCD at all without direct sunlight and squinting if the backlight is off
17:44:31XavierGrmarkun: I can't be sure if you can do that. But be sure to check first.
17:44:37preglowi think you can
17:44:48preglowbut like xavier says, check first
17:45:46miraklostlogic: don't know, on the original firmware when light is off you can't see nothing
17:46:25 Quit Nibbler (Connection timed out)
17:46:32preglowthey turn off the screen
17:46:39preglowsince it's pretty unusable without backlight
17:46:45preglowrockbox doesn't do that yet
17:47:34preglowat least that's what i've heard
17:47:53peturdidn't Linus fix that already?
17:47:57Slasheriwebguest78: Ah, hmm. In what timestamp?
17:48:04Slasheri*time zone
17:48:23preglowSlasheri: point is that MAX_PATH is 255, which almost certainly not enough anymore
17:48:26preglowSlasheri: utf8, etc, etc
17:48:40Slasheripreglow: Ah, that's true with utf8
17:48:42preglowjust search for MAX_PATH, and you'll see it
17:48:46Slasheriso we should increase it
17:48:51preglowin the logs, that is, amiconn meantioned it
17:49:10preglowthe biggest utf8 character converted from ucs2 (which vfat uses), is three bytes
17:49:14preglowso worst case is probably MAX_PATH*3
17:49:30Slasheriuh, that's bad.. we really should increase MAX_PATH then..
17:50:24XavierGrMAX_PATH *3 OMG
17:50:34 Join _FireFly_ [0] (
17:51:11XavierGrI am happy though, with my apprach I don't have to worry about it (jpeg file scrolling)
17:51:52preglowSlasheri: did hardeep fix the queue file thing?
17:52:10Slasheripreglow: i don't know, probably not yet
17:55:57mirakI made a new If I want to upgrade rockbox, I need to erase .rockbox ? Can't we keep some settings ?
17:57:10 Quit NicoFR (Read error: 110 (Connection timed out))
17:57:14lostlogicmirak: can just install new rockbox over old, and not lose setting
17:57:14preglowyou don't need to erase it
17:57:31lostlogicor you can save a settings file separately and then reload the settings after upload (required for config layout changes)
17:57:35preglowjust beware that sometimes rockbox will delete your settings
17:57:38preglowso save them
17:58:59 Join RotAtoR [0] (
18:03:03Learslasheri: better be a bit careful about increasing max_path. Buffers with that size are often allocated on the stack, afaik...
18:03:35 Quit muesli- (Read error: 110 (Connection timed out))
18:03:38Slasheriyep, that's correct. And archos already has little memory..
18:04:07Slasheriprobably somebody with archos unit, could try increasing it
18:04:28XavierGrRockbox is starting to put the DAPs to the limit...
18:06:08peturbuild non-unicode for targets with low mem?
18:08:29XavierGrcan a pointer take a negative value?
18:08:58peturnot that I know
18:09:12Gacelnoormally not
18:09:25Slasherino, pointers are unsigned
18:09:34XavierGrbecause the very first element of my buffer points to 0. But NULL is also zero
18:09:39miraklostlogic: hum the plugins and fonts are built into rockbox ?
18:09:51XavierGrwhen I am searching for bad pointers this must be excluded
18:09:52peturthen use 0xFFFFFFFF?
18:09:57lostlogicnot builtin, loaded at runtime, they are just in the directory
18:11:11 Join dropandho [0] (
18:11:14miraklostlogic: ok
18:11:55XavierGrso ptr = -1 will create an error, right?
18:12:15dropandhohey xavier- have you heard if linus had any luck with the ticking unit?
18:12:31petursigned-unsigned mismatch warning
18:12:44preglowXavierGr: no, ptr = -1 is the same as writing ptr = 0xffffffff
18:13:04preglowsome compilers might whine, some might not
18:13:13XavierGrTruth is that he has much in his mind, and I think that he didn't had the time to check it.
18:13:16preglowbut then again, you'll very, very seldon need to set an address expliocitely
18:14:33XavierGrDebuggin this sort of bechaviour was very difficult.
18:14:52XavierGrImagine that the array holds seek positions into a file.
18:15:32XavierGrthen if a file is invalid I made that pointer NULL (0) then when I reached that pointer I could get in an infinite loop under certain circumstances.
18:16:05peturthen use 0xFFFFFFFF....
18:16:18peturunless that is still a valid position
18:16:31petur(I would imagine not)
18:16:36XavierGrphew our compiler will not complain with -1
18:18:32 Join Dima202 [0] (
18:19:01XavierGrnow I have to sort that HUGE list. It will be a beast.
18:19:40dropandhoxavier- what cha workin on?
18:21:30XavierGra jpeg viewer with file change capabilities without haveing to quit first
18:21:52XavierGrI have already made that but with serious limitations
18:22:01XavierGrso I started again to make it better.
18:22:21dropandhogreat- good luck!
18:22:33dropandhofor 120/140?
18:24:31 Quit San (Read error: 110 (Connection timed out))
18:25:57 Quit dropandho ()
18:26:02mirakI am watching that cube, that's nice on H300. I am wondering if the 10fps limitation to watch xvid on H300 is not because of remanence of the lcd screen
18:26:10miraklatency of the lcd
18:29:12 Join Nibbler [0] (
18:31:41 Join webguest55 [0] (
18:32:01webguest55Looks like the wrong MAX_PATH is already being a problem:
18:33:38preglowwebguest55: yeah, know
18:33:54preglowdepends on how many non-latin chars he's got
18:36:30webguest55It's a likely cause though
18:36:31 Quit DreamTactix291 (Read error: 104 (Connection reset by peer))
18:37:03 Join DreamTactix291 [0] (
18:43:15 Join NicoFR [0] (
18:45:23*_FireFly_ listen currently to Gregorian-Masters of Chant IV :)
18:48:49XavierGrChoral music?
18:49:41XavierGrwhich copmoser?
18:52:45XavierGrI thought that you were listening Chants from the Gregorian Era 1200+
18:59:14lostlogicanyway to insert a measured number of pixels of space into a WPS?
19:01:28 Join dpassen1 [0] (
19:04:04_FireFly_lostlogic: ??
19:06:04lostlogicI want to position a line of text right against the bottom of the screen... the font I'm using doesn't lign it up properly, I need to go down like 5px, not 11...
19:10:29_FireFly_Lost-ash: currently not possible
19:10:54lostlogic*nod* *sad*
19:11:10_FireFly_ehm lostlogic i mean :)
19:12:38***Saving seen data "./dancer.seen"
19:15:36mirakhem where is the page with the bootloader ...
19:16:47mirakwhere ? ...
19:17:31mirakgot it
19:17:37mirakI mess doc and wiki
19:18:29 Nick paugh is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
19:19:05 Join muesli_- [0] (
19:28:49XavierGranyone here that knows very well doings of qsort?
19:29:14_FireFly_ XavierGr doings ??
19:29:38 Join Cassandra [0] (
19:29:42_FireFly_do you mean how quicksort works ??
19:29:44XavierGrwell, doings I mean proccess
19:30:45XavierGrI want someone to help me understand how to program the "compare" support function that qsort uses.
19:31:12XavierGrIt is full with double pointers which confuse me.
19:32:02_FireFly_?? double pointer
19:32:20_FireFly_my manpage say only one pointer for the compare fn
19:32:49XavierGr*(char **)p1;
19:32:58XavierGrwhat is that?
19:33:14XavierGrit makes my head hurt
19:33:15dpassen1a dereferenced pointer to pointer to char
19:33:23_FireFly_the two parameters of the compare fn gives an pointer to two objects which should be compared
19:33:27 Quit webguest78 ("CGI:IRC (EOF)")
19:33:39 Quit AliasCoffee ("Leaving")
19:33:46_FireFly_that's the qsort in my manpage; void qsort(void *base, size_t nmemb, size_t size,
19:33:46_FireFly_ int (*compar)(const void *, const void *))
19:34:08XavierGryes I got that and I am trying to do that. Thing is that the objects to be compared are in a file
19:34:49XavierGrp1 holds the seek position of the string that is to be compared
19:35:40_FireFly_that is a simple int afaik
19:35:59lostlogic:( hold detection not work on h300 wps?
19:36:00_FireFly_why then the *(char**)p1
19:36:01XavierGrbut it is past as a const void*
19:36:32_FireFly_the simply do int var = *((int*)p1)
19:36:36XavierGr(that was with my last attempt which had the strings in memory)
19:36:49_FireFly_ah ok
19:37:39XavierGr_FireFly_: It compiled lets see if it works now.
19:37:41_FireFly_both parameters holds the position of two strings in the file am i right ??
19:37:51XavierGrThe other time it just hanged.
19:38:00_FireFly_hiho Cassandra
19:38:04CassandraWho's working on the iPod port?
19:39:39_FireFly_XavierGr: how looks like the var of the base-parameter
19:39:58_FireFly_is it an array of ints which holds the seek-pos
19:40:10_FireFly_of the strings in the file
19:40:13 Quit NicoFR ()
19:40:23XavierGr_FireFly_: It runs though, it will not sort the list. And now opening a list of 850 files will take like 3 seconds to open.
19:40:47XavierGrso in compare I included code to open the file and read to that position.
19:41:24XavierGrand then *s1 = buf1 and *s2 = buf2 (bufs are from read line)
19:41:46 Join gromit` [0] (n=gromit`
19:41:46_FireFly_strcmp ??
19:42:27XavierGrafter that a strcmp between s1, s2
19:42:36XavierGrlet me paste it in paste bin.
19:42:48_FireFly_strcmp(buf1 ,buf2); should also do it ;)
19:43:09XavierGrnow that you mention it yes
19:43:32XavierGrI was so confused with it that I tried to change it as little as I could
19:45:46XavierGrforgot to remove s1 s2 but anyway.
19:46:29XavierGrand I should remove the O_CREAT too
19:46:33_FireFly_why O_CREATE ?
19:47:19_FireFly_and then check the return value
19:47:38XavierGrwhat it should be?
19:47:50XavierGrah of the fd you mean
19:47:58_FireFly_what's this for: pos[*((int*)p1)] .. yeap
19:48:29XavierGrI dunno you told me so. :)
19:48:46XavierGrpos[x] has the seek position of the string.
19:49:13_FireFly_and what has then the p1 as value i thought the seek-pos
19:49:26XavierGrso I need to tell it: I want the value of p1,p2 into that pos[]
19:49:42XavierGr(between the brackets)
19:50:00_FireFly_if you had called qsort like this qsort(pos,10,sizeof(int))
19:50:25_FireFly_then p1 and p2 holds already the position and not the array index
19:50:56XavierGrso just put pos[p1] pos[p2]?
19:51:27_FireFly_ *((int*)p1) and *((int*)p2)
19:51:46_FireFly_p1 and p2 holds the seek-point itself and not the array index
19:52:27_FireFly_because qsort walks through the given "array" and gives the values to the compare-fn
19:52:48_FireFly_which the fn is currently "working" on
19:53:10XavierGrso can it work out, or it is a dead end?
19:53:57_FireFly_i bet in your current version lseek will return an failure or at least the read_line fn will fill the buffers with garbage
19:54:18XavierGrarhg that way it seems that it will not sort it at all. It leaves the array index as is
19:54:50XavierGrlet me check that with a splash() (and then a reset. haha
19:54:57_FireFly_qsort sorts the values in the array not the index ;)
19:55:40_FireFly_e.g. if you have an array that looks like this: {100,120,140,160,180,200}
19:56:09XavierGrso it sorts the positions.
19:56:20 Join San [0] (n=Test@
19:56:38_FireFly_yepp you will get back the position of the sorted strings ;)
19:57:00_FireFly_ehm the position in the array will represent the sorted strings
19:57:08_FireFly_after qsort
19:57:51XavierGrbut it should work because we comapre strings not the positions
19:58:20_FireFly_you uses the string but you give qsort only an array of seek-positions
19:58:55_FireFly_then he will sort the seek-pos according to your compare-fn-result
19:59:19saa[b_r]iderhello everyone, I just made my first guide in the wiki
19:59:28XavierGrOkay lets make an example: 5, 4, 2, 3 is the list
19:59:40XavierGrthe array is 0, 1, 2, 3
19:59:58saa[b_r]idershould I create a link to it in the wiki's webhome?
20:00:32XavierGrfirst step it will comapre 5 with 4. it will swap 0 with 1. right?
20:01:05_FireFly_if you return that 4 is lower then 5 :)
20:01:47XavierGrthen if the positions are swept then the list will be sorted if I read from the position array. (the list in the file will not be sorted, but I use the array)
20:01:52 Join yngwi [0] (
20:02:41_FireFly_yepp if you uses the rearranged array
20:02:57XavierGrso it should work.
20:03:06XavierGrIt must be something else.
20:03:08petursaa[b_r]ider: nice introduction!
20:03:19_FireFly_no you will have to create a new file when you want also the strings be sorted in the file
20:03:38XavierGrno I don't want the list to be sorted in the file
20:03:39lostlogicpreglow: got a bug in my parser −− it can't tell the difference between two symbols which have the same offset from two different parts of RAM, because nm says same offset either way, and the .map file doesn't contain them.
20:03:53XavierGri want the array to be sorted. That's what I am after.
20:03:54_FireFly_the use simple the rearragned array
20:04:16_FireFly_then you will get the string in sort-order
20:04:24XavierGrI do, that's why I say there is something else
20:05:22 Join Kyl3 [0] (
20:05:41_FireFly_XavierGr: have you checked the return values of the fns used in your compare-fn and have you also checked if the array has really the "sorted"-seekpos
20:05:50_FireFly_after qsort
20:06:55 Quit muesli_- (Read error: 110 (Connection timed out))
20:07:24XavierGrnot yet, I will try to look that now. Thing is that even if I get this right, it will create a vast delay (fortunately only first time) when the first pic is opening (folder with more than 800 pictures)
20:08:09XavierGrI bet all the file related fn in compare do that.
20:08:16 Join NicoFR [0] (
20:09:02linuxstbCassandra: preglow and I are working on the iPod.
20:09:22XavierGr_FireFly_: Thank you were very helpfull.
20:09:33saa[b_r]iderpeter: thanks :)
20:09:33XavierGradd/, you
20:09:44saa[b_r]idersorry petur
20:10:15_FireFly_XavierGr: np :)
20:10:47 Quit NicoFR (Client Quit)
20:11:14petursaa[b_r]ider: np, I got it - it's my real name anyway :)
20:11:24 Join NicoFR [0] (
20:11:27XavierGrGod so many hours for this. I will have a break I can't code anymore.
20:11:32 Quit Lear (Remote closed the connection)
20:11:39_FireFly_XavierGr: ;)
20:12:03 Join NightCat [0] (
20:12:33saa[b_r]iderpetur: :P it's just that people here are from all over the world, didn't want to say your name wrong or something, just in case :)
20:13:31NightCatCould someone help me with setting up a cygwin and other on win32?
20:13:32 Join muesli_- [0] (
20:13:46peturwhat's your problem?
20:13:53NicoFRI can try to halp you on cygwin
20:14:15NightCatI tried to do configure -> make -> error...
20:14:47petur;) you typed 'error' ?
20:15:13NightCatFunny course not :)
20:15:16mirakhem I am trying to patch 1.29kr with bootloader V3 and mkboot doesn't create an output patched file
20:15:51NicoFRwhat sort of build did you ask for in configure ?
20:15:53NightCatOk, if someone could and want to help me, then go to private channel?
20:16:10petur<NightCat> I tried to do configure -> make -> error...
20:16:10petur<NightCat> I tried to do configure -> make -> error...
20:16:10peturNightCat: what's the error?
20:16:13linuxstbWhat's the error message you are getting?
20:16:14peturNightCat: what's the error?
20:16:46*petur slaps his IRC client around
20:16:58 Join hshah [0] (
20:17:07 Quit hshah (Client Quit)
20:17:14NightCatHere it is:
20:17:17NightCatsrc/it/itread.c: In function `it_load_sigdata':
20:17:17NightCatsrc/it/itread.c:1161: error: insn does not satisfy its constraints:
20:17:17NightCat(insn 2275 1148 1149 77 0x0 (set (reg:QI 8 %a0)
20:17:17DBUGEnqueued KICK NightCat
20:17:17NightCat (mem/s/j:QI (plus:SI (reg/f:SI 8 %a0 [354])
20:17:17NightCat (const_int -288 [0xfffffee0])) [0 mididata S1 A8])) 37 {*
20:17:18***Alert Mode level 1
20:17:18NightCat:1060} (nil)
20:17:20NightCat (nil))
20:17:22NightCatsrc/it/itread.c:1161: internal compiler error: in reload_cse_simplify_operands,
20:17:24NightCatat reload1.c:8378
20:17:26NightCatPlease submit a full bug report,
20:17:28NightCatwith preprocessed source if appropriate.
20:17:30NightCatSee <URL:> for instructions.
20:17:32NightCatmake[4]: *** [/home/NightCat/rockbox-devel/build/apps/codecs/dumb/itread.o] Erro
20:17:34NightCatr 1
20:17:36NightCatmake[3]: *** [all] Error 2
20:17:38NightCatmake[2]: *** [dumb] Error 2
20:17:40NightCatmake[1]: *** [build-codecs] Error 2
20:17:42NightCatmake: *** [all] Error 2
20:17:44NightCatThis is an output of make command...
20:17:48NightCatThe end of this output...
20:17:53linuxstbYou need gcc 3.4.x for the iriver builds
20:18:14peturor 4.0.x
20:18:16linuxstbThose errors normally come with using gcc 3.3.x
20:18:38NightCatI setting up my cigwin with this:
20:18:40linuxstb4.0.x should work, but it seems to generate slower code for most codecs compared to 3.4.x
20:19:07peturworks here, but I'm only using mp3...
20:19:07NightCatIs it wrong way?
20:19:59NicoFRI get the same message
20:20:20NicoFRbut I still have a rockbox.iriver in my build dir at the end of the process...
20:20:28 Quit Gacel (Read error: 110 (Connection timed out))
20:20:31peturthat page does say 3.4.x or later, earlier versions fail to compile some of the codecs
20:20:37linuxstbNightCat: Which version of m68k-elf-gcc do you have installed? Type "m68k-elf-gcc −−version" to find out
20:20:46NightCatYes there is a rockbox.iriver
20:21:34NightCat$ m68k-elf-gcc −−version
20:21:35NightCatm68k-elf-gcc (GCC) 3.3.5
20:21:35NightCatCopyright (C) 2003 Free Software Foundation, Inc.
20:21:35***Alert Mode level 2
20:21:35NightCatThis is free software; see the source for copying conditions. There is NO
20:21:35***Alert Mode level 3
20:21:35NightCatwarranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
20:22:08linuxstbThat's the problem - you should have installed 3.4.x.
20:22:10_FireFly_NightCat: you need gcc-3.4
20:22:31NightCatWhat shell I do to have it?
20:22:45NightCatHow can I get it I mean...
20:22:50linuxstbHow did you install m68k-elf-gcc originally?
20:23:29NightCatI have a rockbox.iriver file and other after make command, bu the command make install does not work at all after make...
20:24:08peturbecause not everything got build
20:24:17 Join San||Away [0] (
20:24:31_FireFly_there is no make install when compiledn rb for an target
20:24:34XavierGrlmao _FireFly_: you were right buffers inside compare never return something usefull.
20:24:36NightCatI install it:
20:24:48_FireFly_XavierGr: ;)
20:25:25_FireFly_XavierGr: on an pc you had gotten an sigsev
20:25:41petur_FireFly_: doesn't MAKE INSTALL create the zip?
20:25:42_FireFly_when the programm trys to access memory beyond the array bounds ;)
20:26:09_FireFly_petur: could be but i had only used make zip
20:26:10NightCatHm, then for what I need make install?
20:26:14NicoFRit's make zip
20:26:25peturyes of course... sorry
20:26:26 Quit San (Read error: 110 (Connection timed out))
20:26:37XavierGrmake install installs a rockbox folder to the sim
20:26:38_FireFly_NightCat: nothing
20:26:48XavierGrmake install will output an error on normal builds
20:26:50_FireFly_if you don't uses the sim
20:27:12NightCatThen it all means that I install and set up my cigwin properly?
20:28:06_FireFly_but what is with the error you had posted ??
20:28:49NightCatI don't know...
20:28:51_FireFly_and you need an gcc version 3.4.c or higher for the iriver as target
20:29:12linuxstbNightCat: If you run the Cygwin setup.exe again, does it give you more than one different version of m68k-elf-gcc to install?
20:29:14NightCatAsk again: how can I get it?
20:29:30NightCatA second...\
20:29:54linuxstbIf not, then you should uninstall that package, and you will need to compile the cross-compiler yourself from source.
20:30:19NightCatWhat mirror shell I use to download the gcc?
20:30:24_FireFly_i have just discovered an strange behaviour
20:30:27linuxstbEric's mirror.
20:30:41NightCatA second :)
20:30:49linuxstbI know nothing about Eric's mirror, by the way, I'm just following the Wiki instructions.
20:31:15linuxstbIf Eric's mirror doesn't offer arm-elf-gcc 3.4.x, then we need to correct the Wiki.
20:31:32_FireFly_when you are on wps and paused the current playing song and then press rew to get to the start of the song the "time-scrollbar" didn't get updated
20:31:36***Alert Mode OFF
20:32:12_FireFly_but if you unpaused the song the bar is updated correctly
20:32:35NightCatlinuxstb: it gives 3.4.4
20:32:45NightCatlinuxstb: no 3.4.3
20:32:52 Join Paul_The_Nerd [0] (
20:32:54linuxstbAnd is that the only version?
20:33:18NicoFRthat's for gcc ?
20:33:25NicoFRor m68k-elf-gcc ?
20:33:31NightCatlinuxstb: it gives 3.3.5 or 3.4.3
20:33:44NightCatThat's for m68k-elf-gcc
20:33:53linuxstbIn which case, you should unselect 3.3.5 and select 3.4.3
20:34:49NightCatOk... I need some time...
20:35:37_FireFly_i mean the progress-bar
20:36:36NightCatAnd for what i need this m68k-elf-gcc 3.4.3? I don't understand...
20:36:46linuxstbYou need that to compile Rockbox :)
20:36:56NightCat(It's downloading now...)
20:36:59_FireFly_for iriver-targets
20:37:10linuxstbYou had an older version of m68k-elf-gcc installed before - that was causing your problem.
20:37:16NightCatAnd what I was doing?
20:37:28NicoFRNightCat: where did you find it ?
20:38:03NightCatI mean if i have the rockbox.iriver file?
20:38:18NightCatNicoFR: Find what?
20:38:33NightCatSorry for my english...
20:38:41NicoFRm68k-elf-gcc version 3.4.3
20:38:57NicoFRthe latest version I see is 3.3.5
20:39:24NicoFRusing the cygwin installer or the website ?
20:39:28petur_FireFly_: what do you mean with 'on wps'?
20:39:36NightCatthe cygwin installer
20:39:50_FireFly_petur: wps = while playing-screen ;)
20:40:18NightCatlinuxstb: Are you here? :)
20:40:43peturjust tried and apart from some jumping around the progress bar is ok
20:40:44NicoFRoh right, I was doing it but didn't see the new version :)
20:40:48NightCatAnd I just build a simulator and it works :)
20:41:04CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
20:41:04*Cassandra is still looking for people working on the iPod port.
20:41:13_FireFly_NightCat: the sim uses the "normal"-gcc
20:41:25preglowCassandra: i think i qualify
20:41:43NicoFRwell last time I tryed to build the sim it didn't work
20:41:47_FireFly_Cassandra: linuxstb and preglow
20:41:56_FireFly_as linuxstb has said to you
20:42:05NightCatOk, when I'll download the new version and install it... What shell I do?
20:42:06CassandraOoops. Missed that.
20:42:51Cassandralinuxstb, my friend changed her mind, so I now have an iPod Nano to develop on. What's free?
20:42:53linuxstbNightCat: Delete your build directory, create a new one, and then run configure and make
20:43:00NightCatI mean how can I jump into coding? :)
20:43:06_FireFly_linuxstb: or create a other one
20:43:20preglowCassandra: almost anything is free! :-)
20:43:28preglowCassandra: i'm working on interrupts and timer tick, currently
20:43:36preglowand am hoping to have some code worth commiting soon
20:43:44CassandraOK if I have a go at the button driver. That's probably a nice simple place to start.
20:43:46linuxstb_FireFly_: NightCat currently has a half-working build directory created with the old gcc - which is why I'm suggesting it gets deleted
20:43:54CassandraI assume there's a Linux driver available?
20:43:57preglowCassandra: yes... perhaps...
20:43:58 Quit Rob2222 (Read error: 104 (Connection reset by peer))
20:44:06preglowCassandra: sure, that's all we have in way of documentation
20:44:16_FireFly_linuxstb: ah ok
20:44:16linuxstbI also have some small ipod commits pending - cleaning up the i2c code a little and reading the RTC.
20:44:38CassandraThought it might be.
20:44:47CassandraI'll hold off until you commit then.
20:44:53NightCatAh... 50 % downloaded...
20:45:12 Join Rob2222 [0] (
20:45:56linuxstbCassandra: There are always plugins that need adapting to the ipod buttons. The win32 sim is working.
20:46:14linuxstbAt the moment, all the plugins are disabled for the ipod.
20:46:51linuxstbCassandra: Have you tried installing Rockbox on your Nano yet?
20:47:28 Quit yngwi ("Chatzilla [Firefox 1.5/undefined]")
20:47:30preglowlinuxstb: will cvs work?
20:47:47preglowlinuxstb: perhaps i should commit my thread.c with the sleep loop commented out
20:47:51linuxstbGood point.
20:48:05preglowi'll do that now
20:48:09preglowthen cvs should at least boot
20:49:03Cassandralinuxstb, building the cross compiler now.
20:49:10linuxstbCassandra: Are you on Linux or Windows? Installing the ipod bootloader from Windows will be a challenge.
20:49:23CassandraFortunately I have one of each.
20:49:36CassandraMy development I tend to do in Linux.
20:49:51linuxstbGood. You can install rockbox itself from either, but the bootloader requires you to use "dd" to access the raw partition.
20:49:53preglowyou'll be wanting to deal with bootloaders from there, at least
20:50:20CassandraYeah. I'm following the instructions on the wiki.
20:56:35 Join b0br [0] (
20:57:12peturlooking at the (H3xx) LCD: is lcd_blit() going to be used (I find no references)?
20:57:43NightCatpetur: you working on H3X0 LCD?
20:58:33peturwhat's working? I'm looking at the code... If I understand it enough, I may start working on it ;)
20:58:56NightCatpetur: it's a good new for me :)
20:59:13peturI'm bothered by the slow speed and blinking...
20:59:53NicoFRNightCat: you said you could build the sim right ?
20:59:53NightCatpetur: I have not much knowledges to do this...
21:00:06NightCatNicoFR: Yes :)
21:00:12NicoFRon win32 ?
21:00:19 Quit San||Away (Read error: 110 (Connection timed out))
21:00:39NicoFRwhich sim was it ? win32 h300 ?
21:00:50NightCatNicoFR Yes and Yes :)
21:01:02NicoFRI'm trying to build this exact sim and I can't
21:01:21NightCatpetur: I want unblinking and color WPSs :)
21:01:30NicoFRI get a gcc error message teling me it can't find 'cc1'
21:01:32NightCatpetur: Go on!
21:01:48 Join PaulJ [0] (
21:01:54NightCatNicoFR: tsrange...
21:02:00NightCatNicoFR: strange...
21:02:06NicoFR"installation problem"
21:02:12NightCatА по русски кто-нибудь говорит?
21:02:14linuxstbpetur: No, I don't there is a plan to implement lcd_blit() - do you need it?
21:02:43NightCatNicoFR on what stage?
21:02:44peturnope - just saw it wasn't implemented...
21:02:56linuxstbIt's only needed if you want to write to the LCD without using lcd_framebuffer - it's used for the video player on the Archos targets.
21:03:09NightCatJust built my RB without errors :)
21:03:13linuxstbOr maybe it's the greyscale library...
21:03:16NicoFRNightCat: it prints "CC button.c" then immediatly an error message
21:03:38NicoFRbtw the normal h300 build works fine now thanks to you :)
21:04:16NightCatCould it be an old versions bug?
21:04:29peturI see the LCD driver (hardware) supports some kind of window - may be nice to use one day
21:04:36NightCatNicoFR: I mean could you download all new versions?
21:04:36NicoFRit looks more like something is missing
21:04:52NicoFRI think I have all the latest versions
21:05:19NightCatNightCat: When I have such problems I starts from the beginning of installation :)
21:05:31linuxstbpetur: Someone sent a patch to the rockbox-dev mailing list which uses the window update functions - but no-one has tried it.
21:05:46linuxstbThe person that wrote the patch didn't have a device to test it on...
21:06:04peturI only joined the list recently
21:06:13peturIs it available somewhere?
21:06:20peturI could have a look at the code
21:06:22linuxstbYes - look at
21:06:37linuxstbProbably about 4-6 weeks ago.
21:06:45NightCatNicoFR: Did you do all 4 steps when installing?
21:07:10NicoFRI did
21:08:03NightCatInclude adding PATH=${PATH}:/opt/m68k/bin string?
21:09:07linuxstbpetur: It was posted to rockbox-dev on November 21 - subject "Iriver h300 lcd code."
21:09:09NightCatNicoFR: What path you have to your files (I have this: C:\RockBox\home\NightCat\)
21:09:16 Join solexx___ [0] (
21:09:43petur:) I just found it as well...
21:10:08NightCatI just have some problems when I have this path C:\RockBox\home\Night Cat\...
21:10:26 Join Acksaw [0] (
21:10:28NightCatThen I reinstall my windows user to have a C:\RockBox\home\NightCat\...
21:10:47NightCatpetur: Keep working! :)
21:10:48NicoFRyeah I made sure to have no spaces in the path
21:11:19 Part Acksaw
21:11:43NightCatThen only one different thing: drive c: intead of f:
21:12:29NightCatNicoFR: Are you a programmist (I mean have you С exp.?)
21:12:34NicoFRI don't think it can be that
21:12:41***Saving seen data "./dancer.seen"
21:12:59 Join Acksaw [0] (
21:13:12NightCatIf You need in sim very very then try to reinstall all from beginning...
21:13:59CassandraWhere do the built tools get put these days?
21:14:13NightCatNicoFR: When you select what to install in the cigwin installer did you remove some puncts wich was in installer bu was not on wiki page?
21:14:24linuxstbCassandra: The same place - tools/
21:14:43CassandraI thought they were platform dependent now.
21:15:00NicoFRI don't think so
21:15:23linuxstbOnly the tools required for a particular platform are compiled. But they are still compiled in the same place.
21:16:05NightCatCause I only mark what was said in wiki and leave what was marked.
21:16:39NicoFRI did that too
21:17:16NightCatNicoFR: Any ideas?
21:17:49NightCatCould someone said what can I do for my first time near the RB developing?
21:17:53NicoFRin your cygwin bin dir, do you have a 'cc1.exe' file ?
21:18:24NightCatOnly cc.exe
21:18:42linuxstbNightCat: Anything you want to :) Changing a plugin or developing a new one is a relatively easy way to start.
21:19:02linuxstbOr look at the feature requests and see if you could do any of them.
21:19:36webguest55Cassandra: seen ?
21:19:38NightCatlinuxstb: Actually I mean something like "Hello World!" you know ;)
21:20:04linuxstbNightCat: There is already a "hello world" plugin
21:20:36NightCatlinuxstb: And I don't know С :) I shoud have some practising...
21:20:38linuxstbCan you now create a that you can install on your iriver?
21:20:43Cassandrawebguest: Yes.
21:20:58 Quit solexx_ (Read error: 110 (Connection timed out))
21:20:59NightCatlinuxstb: Why
21:21:17linuxstbThat contains everything you need - rockbox.iriver, the codecs, plugins etc
21:21:26NightCatlinuxstb: Why not .rockbox folder and rockbox.iriver file?
21:22:13NightCatlinuxstb: How can I do it?
21:22:33NightCatlinuxstb: (I already have an rockbox.iriver file)
21:23:25NicoFRmake zip
21:23:47 Join linuxstb_ [0] (
21:23:53 Quit linuxstb (Nick collision from services.)
21:23:57 Nick linuxstb_ is now known as linuxstb (
21:24:28NightCatlinuxstb: Done. I have an
21:24:39NightCatlinuxstb: The next step?
21:24:57linuxstbInstall it on your iriver and check it's working.
21:25:02linuxstbAfter that, it's up to you.
21:25:12linuxstbIs there anything you want to change in particular?
21:25:46NightCatYes but it's all to hard I think...
21:26:48NightCatSome questions: 1. Have many times per day/week will be good to update the source from cvs?
21:26:56NightCatHave = How :)
21:26:59preglowjust see how often stuff happens
21:27:09preglowif something has happened in cvs, it's wise to update
21:27:23linuxstbI always update before I start work on something.
21:27:34linuxstbYou can't update too often.
21:28:01NightCatThe second what I have to done for ex. if i change 2 strings in some files I must to compile simulator again to check it?
21:28:13 Join Moos [0] (
21:28:51 Quit muesli_- (Read error: 110 (Connection timed out))
21:29:15linuxstbYou can check it on the simulator or your player - it's up to you. Some developers use the simulator a lot, others (like me) don't use it at all.
21:29:41NightCatBut anyway a have to rebuild it all the time?
21:29:49 Join xand [0] (n=irc@tor/session/x-eb5f4eddeabe2c53)
21:30:06peturmake will rebuild what needs to be rebuild
21:30:40NightCatOh I see...
21:32:09NightCatThen I change something -> run "make" -> run "make zip" -> uploads files to DAP -> check it. Im I right?
21:32:11 Join muesli_- [0] (
21:33:02linuxstbOr the alternative is to change something -> make (in your simulator build directory) -> make install -> test it in the simulator.
21:34:03NightCatlinuxstb: good, any eiser way to test?
21:34:31NightCatlinuxstb: What code editors or something could I use?
21:34:41linuxstbThose are the only two ways to test your code.
21:34:52NightCatlinuxstb: I mean my workspace :)
21:34:52linuxstbI have no idea about code editors - I just use emacs or vi
21:35:23NightCatYou mean simple text editors?
21:35:58webguest55Simple is hardly something I'd call emacs...
21:36:04NightCatAnd last what language is it? Perl or C?
21:36:10webguest55But yes, a text-editor
21:38:25NightCatThen I going to use Notepad++ (
21:38:28peturIn case somebody here knows: are X and Y swapped on the H300 LCD driver?
21:39:03NicoFRthat's the one I use too :)
21:39:29NightCatNicoFR: Cool :)
21:40:12dwihnoSimple is hardly something I'd call vi... ;)
21:40:55preglowoh, hooray
21:41:01 Quit b0br ("CGI:IRC")
21:41:04preglowprocess windows doesn't allow me to kill
21:41:05 Join b0br [0] (
21:41:08 Quit b0br (Client Quit)
21:41:37NightCatGuys you'r rock :)
21:43:05lostlogicpreglow: if you are going to check out the profiling stuff, I just posted a mildly updated version of the patch to sourceforge. added the NO_PROF_ATTR attribute define, and used it in Tremor, fixed the bug I mentioned in my parser by using objdump instead of nm to locate symbol names, and added
21:43:48preglowwill check out later
21:44:01lostlogicyep yep, just wanted to mention, I'm off for the day probably.
21:45:21preglowdeveloping in windows is such a fucking dream when it doesn't even allow me to close my crashed programs
21:45:40*lostlogic hugs linux
21:45:56*xand hugs debian
21:46:17peturpreglow: haven't seen that yet... and have crashed many
21:46:49preglowi've seen it bunches of time
21:47:28preglowperhaps i should allow this thing a reboot
21:48:17 Join Guest229461 [0] (
21:51:12 Quit NicoFR ()
21:55:39NightCatAnd another question: could i use cvs and building mechanism to install new RB build wersions on my DAP instead of using the wiki daily builds page?
21:56:35NightCatAnd if I do some changes in code and then run cvs update -dP will it delete my changes?
21:57:59XavierGrno it will merge them
21:58:13XavierGrbut yes you can use cvs to compile your own OR latest builds.
21:58:35XavierGrIf you want to play and have a clean CVS just copy the whole CVS to another folder.
21:58:57XavierGrI have rockbox-devel and update it once in a while, but never fiddle with it.
21:58:57NightCatHm... Then how can I stay up to date and doing changes together?
21:59:23XavierGrwhen I want to make new code I copy this folder and start making changes, then you can update in that new instance or make patches too.
21:59:41NightCatOk, I see.
21:59:42XavierGrif you want to stay uptodate and make changes
21:59:42preglowi just plain and simple use one folder
21:59:54preglowit's completely recent and contains all my changes
21:59:57XavierGrthen just do your changes and update
21:59:57preglowas long as i haven't broken anything
22:00:24 Quit linuxstb (Read error: 110 (Connection timed out))
22:00:31XavierGrbut preglow sometimes your changes will make a conflict with update.
22:00:41NightCatpreglow: what in the code you have already change?
22:00:41preglowyes, and then i just correct the conflicts
22:00:48preglowNightCat: oh, quite a bit
22:00:59preglowlots of musepack stuff, lots of system stuff
22:01:01 Join linuxstb [0] (
22:01:20preglowit doesn't matter
22:01:24preglowas long as you're careful, it works fine
22:01:29XavierGrpreglow: because sometimes maybe I will do strange things(or evil) I just copy the cvs and leave the main folder clean.
22:01:39preglowit's up to you
22:01:41NightCatThen if I want to adjust a game to fit my DAP screen where shell I go?
22:01:45XavierGrthat way if I mess it hard I can always delete the new cvs and run with a clean one.
22:01:55preglowif you keep everything in one folder, you just need to be careful when you commit
22:02:04peturwindows devs: I can recommend TortoiseCVS (
22:02:36XavierGrpreglow: I just don't have access to commit :)
22:02:38NightCatWhat is it?
22:02:41webguest55NightCat: if you're just changing a plugin, definately just go with a single folder
22:03:13NightCatOk I'll try :)
22:03:51peturit's a CVS client for windows, it integrates in the explorer. You can right-click a folder and do 'update',...
22:03:53XavierGrif you mess that plugin just delete it and update
22:03:54 Quit ender` (Read error: 110 (Connection timed out))
22:04:38XavierGrand can we combine it with cygwin?
22:05:02NightCatXavierGr: I'll want to try it :)
22:05:27 Quit Acksaw ()
22:05:34NightCatI have a problem with CVS cause I have an Proxy internet...
22:05:44XavierGrNightCat: I mean if you made serious mistakes and you can't revert manually, of course you want to try your changes
22:05:57 Quit KN|stiff (Read error: 110 (Connection timed out))
22:06:05NightCatAnd I have no idea how to set up cvs for proxy...
22:06:35preglowyou've only got port 80?
22:06:43preglowand those standard ones
22:06:50NightCat"I'll want to try it" I meant TortoiseCVS :)
22:07:25NightCatI have to set up and my login and password to have internet :)
22:07:32XavierGranyway, time to hit the sack.
22:07:35XavierGrlater all.
22:07:51*petur crosses his fingers as he tries the lcd_update_rect() code from Kaggen
22:08:43NightCatHoly s... TortoiseCVS is 5 mb in weight :)
22:09:37preglowyeah, but it's very nice :-)
22:10:33NightCatDownloaded it now try...
22:11:17peturKaggens code seems to work OK or it never gets called :)
22:11:48peturWill need to find a way to compare update speed
22:27:18 Quit _FireFly_ ("Leaving")
22:27:38NightCatpetur: What you need to wind that way?
22:27:53NightCatpetur: Can I help somehow?
22:29:16NightCatpreglow: it seem not to update my files (I mean TortoiseCVS)...
22:29:23NightCatpreglow: Why?
22:29:48preglowdon't ask me
22:30:17peturNightCat: do you get those green overlays on the icons of your files?
22:31:20NightCatNo, but it updates (i think)... How can I test it?
22:31:39peturif it updates, you get the output in a window
22:31:50peturjust like normal cvs
22:32:10NightCatIt writes that update success.
22:32:57NightCatOh, yes know it works :)
22:34:34peturto do diffs or merges, you can use winmerge (, it integrates in TortoiseCVS
22:35:48peturto create patches, richtclick on the file you changed, choose CVS and then 'Make Patch'
22:36:29 Nick Lynx_awy is now known as Lynx_ (
22:36:32NightCatDownloading... ;)
22:38:50 Quit muesli_- (Read error: 104 (Connection reset by peer))
22:41:32 Join San [0] (
22:41:46SanI am making a plugin that changes the color of the Player
22:42:11Sanhow do i coompile it, where do i put it in the sources file?
22:43:25^BeN^san, are you Sanitarium from mistic river?
22:44:07^BeN^you can try to anser my question?
22:44:45Sanwho are you?
22:45:00^BeN^im Paprica
22:45:04Sanoh, ok
22:45:06Santhats easy
22:45:14Sanhave you got the dev kit?
22:45:31Sanok, open it and then type "pw"
22:46:01Sanis the image a BMP?
22:47:01^BeN^which image?
22:47:23Santhe one you want to be in the rb code like 0x001 and stuff
22:48:09Sanok, put that in the /home/guest/tools folder
22:49:11^BeN^ok now?
22:49:28Sanin the dev toolkit type cd tools
22:49:35San"cd tools"
22:50:49Sannow type "bmp2rb filename.bmp"
22:51:44^BeN^10x =]
22:51:53^BeN^its will be in color?
22:51:58Sani dont know
22:52:02Sannever tryed color
22:52:08 Quit Paul_The_Nerd ("Chatzilla [Firefox 1.5/undefined]")
22:52:19Sanjust put #include "lcd.h" at the top and it might be
22:52:49 Join KN|stiff [0] (
22:52:56 Join Noel_sad_song [0] (
22:53:19Sanwhat are you making?
22:53:46^BeN^i wont to make a larger pictures and numbers for the solitare
22:54:13Sanah, cool
22:54:53^BeN^mm thereis some way to invert the numbers to bmp picture?
22:55:22Sandont think so
22:56:49 Join _FireFly_ [0] (
22:57:30 Quit Guest229461 (Read error: 110 (Connection timed out))
22:57:49_FireFly_maybe i have found out why the images on the remote blinks/flickers if they shouldn't do that
22:57:50^BeN^mm do you know how to compile only one file? because "make" takes 2-3 minutes
22:59:59NightCatSan: I want to do some experiments with colors and lcd... But I'm a NB... Can u help me?
23:01:11 Quit Noel_sad_song ("CGI:IRC")
23:02:42NightCatHow cam I for ex. chage colors of screen?
23:02:52NightCat(Like you do...)
23:03:29Sanhave you get the dev kit?
23:04:02NightCatI have the cigwin installed and setuped :)
23:04:20NightCatAnd I already build the sim :)
23:04:26Sanok, good
23:04:31NightCat(Many times :) )
23:04:41Sanopen up firmware/export/lcd.h
23:04:44 Join Owner [0] (
23:05:18NightCatY :)
23:05:33Sanlook for "rockbox Blue"
23:05:33NightCatIt's a simple step :)
23:06:04NightCatLine 167 :)
23:06:23Sansee where is says (039, 394, 830)?
23:06:24NightCatIts an RGB color?
23:06:33Sanjust change them
23:06:48NightCatOk it's quite easy :)
23:07:03Sanyes it is
23:07:08Santhen save it and compile it
23:07:26^BeN^[23:57:53] <^BeN^> mm do you know how to compile only one file? because "make" takes 2-3 minutes
23:07:29NightCatOk i'll try it, what else?
23:07:41CassandraHmmm. This apple_os.bin file. Is it made using the standard Rockbox bootloader build systems?
23:08:11preglow^BeN^: lemme guess, you're using cygwin?
23:08:14Sanyou cant compile one file
23:08:20 Nick Owner is now known as slawson (
23:08:26 Join g33 [0] (n=g33@
23:08:33Sanbut it you keep the window open after a full compile, it is really fast
23:08:38Sanhey g333
23:09:08Sanif you keep the program open after a full compile, and then compile again, it only compiles the files that need to be compiles
23:09:11g33just installed rockbox on my h140.. and ive noticed that one of my songs (actually the first and only one ive tried) doesnt sound as good as with original firmware
23:09:28preglowCassandra: isn't that the extracted apple firmware?
23:09:28g33the treble is all sharp and almost noisy
23:09:39linuxstbpreglow: Yes.
23:09:53g33its a 24khz 56kbps mp3 file.. sounds a lot smoother with iriver firmware
23:10:14DBUGEnqueued KICK San
23:10:14*San Skinny Dips in a sea of Dream Theater - Live At Budokan Disc 1 - As I Am
23:10:25dpassen1that's because we have a very simple resampler, currently
23:10:33amiconng33: The rockbox resampler doesn't work well with such low sample rates
23:10:38amiconnAsk preglow ;)
23:10:58Sananyone wana help me make a color changer for rockbox?
23:11:12NightCatI want!!!
23:11:14NightCat :)
23:11:28Sanbut can you code it?
23:11:28g33so whats the "minimum" for it to sound as good as original firmware?
23:11:37San128 maby
23:11:41NightCatI interested in All that connected to getting RB into colors :)
23:11:56Sanbut can you help me code it?
23:12:02NightCat<San> but can you code it? Interesting question ;)
23:12:03merbananamiconn: who is in charge of the resampler ?
23:12:21dpassen144.1 KHz should sound perfectly, all other sampling rates may sound a bit off
23:12:42g33whats the diff between stereo width and crossfade?
23:12:44***Saving seen data "./dancer.seen"
23:12:45 Quit _FireFly_ ("Leaving")
23:12:48Sanbut, how can a song like that sound good at all?
23:13:06NightCatg33: Maybe you just resemple your song on PC?
23:13:18g33well it doesnt sound very good, but it doesnt sound disturbingly bad on the original firmware
23:13:23NightCatSan: You'r right...
23:13:39 Join Rincewind [0] (
23:13:45g33that doesnt give me any better quality, just bigger file
23:14:18 Join linuxstb_ [0] (
23:14:24NightCatg33: But it give you a better quality on RB :)
23:14:32NightCatIm I right?
23:14:50g33other than that, RB seems awesome
23:14:54 Join _slawson_ [0] (
23:15:11g33smaller fonts, bookmarks, fading backlight, crossfade, gapless.. omfg
23:15:15NightCatAnd with 20 gigs it's not a big problem :)
23:15:26g33does the optical out work in RB?
23:15:31g3340gb :D
23:15:31NightCatYou'r right :)
23:15:57NightCat(Correct me if I it's not...)
23:16:23g33good.. cause ive just ordered a $270 DAC from china :)
23:16:35peturLCD updates using the window function of the hardware doesn't give the speed increase I was looking for :(
23:16:47peturWill add it to the tracker anyway.
23:16:58Cassandrapreglow, Oh, sorry, yes. I meant this rockboot.bin. Make seems to go all recursive on me if I try to build the iPod bootloader.
23:17:04 Join Rincewind_12 [0] (
23:18:34linuxstb_Cassandra: Which step on the WIki page are you up to?
23:19:39CassandraTowards the end of step 1 - building the bootloader.
23:20:24Cassandra../configure seems to work, make doesn't.
23:20:38g33whats the voice thing do?
23:21:03 Join webguest18 [0] (
23:21:09Bagderis xargs in the "devkit" ?
23:21:15 Join muesli_- [0] (
23:21:37 Join _FireFly_ [0] (
23:21:38 Quit _slawson_ ("Chatzilla 0.9.67+ [SeaMonkey 1.5a/2005120909]")
23:22:02_FireFly_woot i have found the flickering problem with images on the remote-wps
23:22:15Bagderno devkit user around?
23:22:30Cassandramake -C /u/christi/work/rockbox-devel/rockbox/tools CC=gcc rdf2binary convbdf scramble ipod_fw bmp2rb codepages
23:22:48Cassandramake keeps trying to execute that over and over again.
23:23:09linuxstb_That's odd. I've just tried it and it builds fine.
23:24:20CassandraI don't think I've messed up building the compiler, because it seems to build a binary OK.
23:25:22BagderCassandra: what if you run that command line manually?
23:25:23NightCatg33: it's voice menu :)
23:25:33Bagderthe make -C one I mean
23:25:39CassandraOh, hold on. Now that's doing the same.
23:25:41NightCatAll menu commands player says :)
23:25:46CtcpIgnored 2 channel CTCP requests in 0 seconds at the last flood
23:25:46*Cassandra looks bemused
23:26:26BagderCassandra: then cd to the tools dir and run 'make CC=gcc rdf2binary convbdf scramble ipod_fw bmp2rb codepages'
23:26:45NightCat!!! g33 !!!
23:26:45 Quit webguest18 ("CGI:IRC (EOF)")
23:27:22NightCatI answered your question :)
23:27:29CassandraJust tells me everything is up to date.
23:27:39g33still not sure what that means :)
23:27:51g33is it like a computer voice that reads the names of the files? :)
23:28:06Cassandramake seems to be calling itself recursively.
23:28:19linuxstb_Cassandra: Can you build other things - e.g. the h120 bootloader?
23:28:27g33hmm... this 192kbit mp3 sounds sharper too
23:28:35NightCatFor ex. you select General Options and DAP says: "General Options" in your earphones :)
23:28:50Cassandrairiver firmware seems to be building OK.
23:29:05CassandraI can try the bootloader if you like.
23:29:25g33really? awesome
23:29:36CassandraDamn focus
23:29:40NightCatYes :)
23:29:41linuxstb_Can you build Rockbox itself for the iPod? I can just send you the bootloader.bin if you want.
23:29:50 Quit slawson (Read error: 110 (Connection timed out))
23:30:04Cassandralinuxstb, nope that's failing with the same error now.
23:30:09g33so what does the option .talk mp3 mean? :)
23:30:17CassandraIt seemed to be working earlier.
23:30:30linuxstb_Obvious question - is arm-elf-gcc in your path?
23:30:35NightCatA moment...
23:30:46g33i dont hear anything
23:30:58 Join webguest80 [0] (
23:30:59CassandraOh, hold on, spurious symlink.
23:31:07 Nick Lynx_ is now known as Lynx_awy (
23:31:24CassandraNow build bootloader is giving me make: *** /u/christi/work/rockbox-devel/rockbox/bootloader: No such file or directory. Stop.
23:31:49Bagderit sure sounds like some kind of path-related problem
23:31:49NightCatg33: ofcourse :)
23:31:58Bagderpath or dir
23:32:11CassandraI think you're right.
23:32:17NightCatg33: Here the Voice HowTo:
23:32:22linuxstb_Do you have the "bootloader" directory? It's at the same level as apps, tools etc
23:32:35g33cool thanks
23:32:46*Cassandra slaps her forehead.
23:32:55NightCatg33: You need a voice file to do this :)
23:33:01Cassandrait would appear not. Just checked it out.
23:33:42preglowoh well
23:33:47_FireFly_YES! ;) i have a solution for the image flickering problem and have also found an unneeded call to wps-display-images
23:33:47preglowi've gotta go
23:33:52preglowlater, all
23:34:05CassandraSeeya preglow.
23:34:59webguest55Bagder: yes, devkit has xargs
23:35:02 Join Caydel [0] (
23:35:09Bagderwebguest55: thanks
23:35:18petur_FireFly_: what is it? mod coming up that uses xargs
23:35:40 Quit webguest80 (Client Quit)
23:37:02_FireFly_petur: if have some text(e.g. the volume- or battery-value) and some images(e.g. play-mode,battery-symbol) on the same line then the images starts flickering on the remote
23:37:16_FireFly_s/if/if you
23:37:57 Part Caydel ("Leaving")
23:38:16peturnot only on the remote :(
23:39:03peturis your fix only for the remote?
23:39:37_FireFly_i see it only on the remote (h120-user) :) but the fix should isn't only for the remote because the code which gets fixed is used for alle displays
23:40:14 Join JazzBone [0] (
23:40:19 Quit Rincewind (Read error: 110 (Connection timed out))
23:43:07_FireFly_petur: are you h3xx user ??
23:43:39_FireFly_can you test this fix pleas ??
23:44:42NightCat_FireFly_: I also can!!!
23:45:05Rincewind_12I have a suggestion for a few changes in the options menu
23:45:20 Quit KN|stiff (Read error: 110 (Connection timed out))
23:45:36Rincewind_12is there anybody here who is willing to work on the menus?
23:45:54_FireFly_NightCat: why not :)
23:46:01NightCat_FireFly_: Say how to apply the pach?
23:46:06NightCatWhat command?
23:46:33_FireFly_patch -p0 < </path/to/patch-file/patch-file>
23:46:38_FireFly_in the root of the sources
23:47:11_FireFly_e.g. patch -p0 <flicker-fix.patch (when the file is in the same directory)
23:47:18peturhmmm. it's in gwps-common.c, I bet that won't do much on my main display?
23:47:37CassandraSuccess of a sort. I've installed the bootlloader, but it's no longer booting the original OS. Just hangs on "booting original firmware"
23:47:38_FireFly_petur: this fix on for the wps
23:48:20_FireFly_i have it mentioned haven't i ??
23:49:02NightCatHm, the command do nothing...
23:49:11peturoh - was mixing gwps with rwps :D
23:49:21linuxstb_Cassandra: Have you tried installing Rockbox itself?
23:49:30_FireFly_petur: ;)
23:49:33CassandraJust building it now.
23:49:34 Quit Rincewind_12 (Read error: 104 (Connection reset by peer))
23:49:50petursorry, still quite new to RB (and even my H340)
23:50:22_FireFly_NightCat: to which directory have you downloaded the patch ??
23:50:25NightCatWhen I use command patch it just freeze...
23:50:37_FireFly_NightCat: ouch
23:50:38NightCatTo my rockbox-devel
23:50:50_FireFly_but this isn't a problem from my patch ;)
23:51:05linuxstb_Exactly what command did you type?
23:51:11NightCatI mean the cygwin freeze.
23:51:15 Join qweasd [0] (
23:51:22NightCatand all
23:51:28_FireFly_only patch ??
23:51:31NightCatalso I try patch -p0 flicker-fix.patch
23:51:46NightCatThe same result.
23:51:53qweasdcan someone help me?
23:51:54_FireFly_no it's wrong try this: patch -p0 <flicker-fix.patch
23:51:56linuxstb_You should type "patch -p0 < flicker-fix.patch"
23:52:08qweasdi want to learn C progamming
23:52:14qweasdwhere should i start
23:52:26linuxstb_Chapter One
23:52:30qweasdany particular tutorial?
23:52:30linuxstb_of any C book
23:52:44qweasdalso what program should i use?
23:52:51linuxstb_:) Do you have any programming experience?
23:53:02qweasdno :(
23:53:22qweasdwell ive compiled my own rockbox using the tutorial on the rockbox site
23:53:37 Join KN|stiff [0] (
23:53:46NightCat_FireFly_ ... make -> error
23:53:52petursame here...
23:54:00qweasdand i know the printf, if, functions
23:54:00Sanqweased, Just studie the source and try to make sense of it? Thats what I did.
23:54:08_FireFly_petur: which error
23:54:12 Quit muesli_- (Read error: 104 (Connection reset by peer))
23:54:14peturerror: 'struct gui_img' has no member named 'stride'
23:54:21NightCat$ make
23:54:21NightCatmake -C /home/NightCat/rockbox-devel/tools CC=gcc rdf2binary convbdf scramble de
23:54:21NightCatscramble mkboot bmp2rb codepages
23:54:21NightCatmake[1]: `rdf2binary' is up to date.
23:54:21NightCatmake[1]: `convbdf' is up to date.
23:54:22***Alert Mode level 1
23:54:22NightCatmake[1]: `scramble' is up to date.
23:54:24NightCatmake[1]: `descramble' is up to date.
23:54:26NightCatmake[1]: `mkboot' is up to date.
23:54:26peturneed to patch the .h file as well?
23:54:28NightCatmake[1]: `bmp2rb' is up to date.
23:54:28qweasdSan how long did it take you
23:54:30NightCatmake[1]: `codepages' is up to date.
23:54:32NightCatCC thread.c
23:54:34NightCatAR+RANLIB /home/NightCat/rockbox-devel/build/librockbox.a
23:54:36NightCatCC gui/gwps-common.c
23:54:38NightCatgui/gwps-common.c: In function `wps_draw_image':
23:54:40NightCatgui/gwps-common.c:1207: error: structure has no member named `stride'
23:54:42NightCatgui/gwps-common.c:1209: error: structure has no member named `bitmap_part'
23:54:44NightCatgui/gwps-common.c:1209: error: structure has no member named `src_x'
23:54:46NightCatgui/gwps-common.c:1210: error: structure has no member named `src_y'
23:54:47peturNightCat, don't post your complete output here :)
23:54:48NightCatgui/gwps-common.c:1210: error: structure has no member named `stride'
23:54:50NightCatmake[1]: *** [/home/NightCat/rockbox-devel/build/apps/gui/gwps-common.o] Error 1
23:54:51qweasd San how long did it take you
23:54:52NightCatmake: *** [all] Error 2
23:54:56Bagderstop that flooding please
23:55:08NightCatpetur: Why it's funny :)
23:55:15Sanstarted today, and am making color in plugins now
23:55:24_FireFly_there are some parts of my combined -bitmap-support laying around
23:55:34SanBTW, anyone know what this meens:
23:55:36peturok :( then.
23:55:37_FireFly_i will make a clean patch
23:55:39SanDamn, wait
23:55:47Sanlogo.c:188: error: `LOGO_WIDTH' undeclared (first use in this function)
23:55:47Sanlogo.c:188: error: (Each undeclared identifier is reported only once
23:55:47Sanlogo.c:188: error: for each function it appears in.)
23:55:47Sanlogo.c:189: error: `LOGO_HEIGHT' undeclared (first use in this function)
23:55:47Sanlogo.c:214: error: `LOGO' undeclared (first use in this function)
23:55:48***Alert Mode level 2
23:55:48Sanmake[2]: *** [/home/guest/build/apps/plugins/logo.o] Error 1
23:55:50Sanmake[1]: *** [rocks] Error 2
23:55:52Sanmake: *** [all] Error 2
23:55:54Sanwhat does that meen?
23:56:01Sanoh, wait...
23:56:35Cassandralinuxstb, I think I'm in trouble. Can't get the iPod to register as a USB device any more, so no way to install Rockbox on it.
23:56:54linuxstb_Cassandra: Don't worry :)
23:57:01peturSan: looks like seriously overdriven input (distorted)
23:57:07Sani know
23:57:11Santhat was a while ago
23:57:18Sanan accident
23:57:20qweasdSan, where did you start learning the source?
23:57:20linuxstb_Cassandra: Are you in "disk mode" ?
23:57:25Sani didn't meen to post that
23:57:38qweasdSan, where did you start learning the source?
23:57:43qweasdcan u give me a link?
23:57:54_FireFly_petur: ok updated
23:57:59Cassandralinuxstb, nope. I have no idea how to use this thing at all, I'm afraid. All I got was the base unit. I'm borrowing my housemate's cabel.
23:58:07Sani just looked at all the files
23:58:16Sanand tryed to figure out how they worked
23:58:27linuxstb_Cassandra: Follow the instructions at the very bottom of the IpodBoot wiki page
23:58:27Sanit only took me about 5 hours of messing around
23:58:32Santhen i got it
23:58:52qweasd5 hours!!!
23:58:53linuxstb_Cassandra: i.e. hold down MENU and SELECT until it reboots, then hold down SELECT and PLAY until it enters disk mode
23:58:58qweasdive been doing this tutorial

Previous day | Next day