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 2005-12-21

00:00:23mirakthe divide happen after the mult
00:00:30mirakI guess
00:00:48amiconnNo guesswork involved
00:01:02*mirak don't want to open eclipse tonight
00:01:06amiconnApart from that, LCD_RGBPACK() happens at build time, not at runtime
00:01:30mirakI am not sure what it changes, can you explain ?
00:02:52linuxstbMy understanding is that 0->0, 255 -> 31 and everything else is proportional in between,.
00:03:18 Quit Jungti1234 ("bye")
00:03:59amiconnlostlogic: Do you check/set the correct EMAC mode?
00:05:07miraklinuxstb: it's the "mostly done at compile time that I don't get". So there is an array that map the precalculated values instead of computing.
00:05:59amiconnNo. All the calculation within the macro is performed by the compiler (if you use constants for r, g and b, that is)
00:06:07linuxstbLGB_PACK is a macro, so when you write LCD_RGBPACK(255,255,255) in the source, the compiler will calculate it at compile time and replace it with 0xffff
00:06:22amiconnThe compiled code does only contain the final result (a 16 bit unsigned integer)
00:08:08mirakok, I am not familiar with that
00:08:47linuxstbIt's like saying "x=3+5;" - the compiler will simplify that to just be "x=8;"
00:09:33mirakbut this can't work for loaded images
00:09:38mirakor xvid for exemple
00:09:42mirakright ?
00:10:18mirakI can understand that it works for wps
00:10:20LinusNmaybe we should raise the normal frequency from 45MHz to 56MHz, to compensate for the increased cpu load due to the larger lcd?
00:10:28LinusNon the h300
00:10:35linuxstbThat's a different issue, and you possibly wouldn't use the LCD_RGBPACK macro.
00:10:46amiconnsurely not
00:10:53linuxstbxvidcore for example has its own internal routines for converting to rgb565
00:11:10linuxstbIt's a yuv2rgb conversion
00:11:11markunffmpeg also has a routine for it.
00:11:52 Quit webguest77 ("CGI:IRC (EOF)")
00:12:01amiconnEither you'd use these routines if they're fast enough, use fast-but-imprecise bit cutting, or use an optimised version of the formula in LCD_RGBPACK()
00:12:09amiconnThat's what the grayscale library does
00:12:41*petur discovers belgian beer doesn't help to do unsigned arithmetic
00:12:53LinusNpetur: wow
00:13:20peturliving in belgium doesn't help ;)
00:13:24LinusNpetur: you're nobel prize material for discovering that :-)
00:13:25*Bagder writes that down
00:13:33amiconnIn the grayscale library, I use a simple trick to optimise away the division.
00:13:40*linuxstb puts down his Guinness to write that down
00:13:46*LinusN tries with english beer
00:14:27amiconnI use the fact that 1/255 = 1/256 + (1/256)/256 + ((1/256)/256)/256 + ...
00:14:38markunGrolsch helps!
00:15:10amiconnThe first two terms deliver sufficient precision, and /256 can be replaced by >> 8
00:15:43mirakpetur: you speak flaman or french ?
00:15:43amiconnSee apps/plugins/lib/gray_draw.c lines 510,511
00:16:21peturand a bit of french, german and english
00:17:00mirakamiconn: ok
00:17:18nathanhthough any modern compiler will replace /256 with >> 8 for you
00:17:33nathanhin fact, gcc will even convert /255 into the crazy combo of shifts and adds
00:17:48amiconngray_ub_gray_bitmap_part() does the same, but coded in asm
00:17:54 Join Kohlrabi [0] (
00:24:16 Join memmem [0] (
00:24:42memmemGCC may even turn the division into multiplication.
00:27:14 Quit Moos ("Rockbox?... The best ;)")
00:30:22peturthis is useless, will have another go at it tomorrow... damn beer ;)
00:30:35 Quit petur ("here today, gone tomorrow")
00:37:32 Quit memmem ("ERC Version 5.0.4 $Revision: 1.726.2.19 $ (IRC client for Emacs)")
00:39:59linuxstbOK, who broke the ipod builds? :)
00:40:05lostlogicx2.14% overall reduction in time used to decode OGG from assembly optimizing render_line and render_point
00:40:21 Quit gromit` (Read error: 104 (Connection reset by peer))
00:40:32linuxstb(and the Ondio...)
00:42:50 Quit Rob2222 ()
00:42:50 Quit ender` (Read error: 104 (Connection reset by peer))
00:49:35 Join Rob2222 [0] (
00:52:54amiconnLinusN: I see RED :/
00:53:04LinusNamiconn: me too :-)
00:58:23lostlogicxpreglow: you were right to look in floor1.c for optimizations for high bitrates. got at least 1 full percent boost (looks like two kinda) improvement by optimizing render_point and render_line on my bitrate spike test song.
00:59:05linuxstbLinusN: Do you want me to fix solitaire for the ipod?
00:59:15linuxstbAgain.... :)
00:59:23LinusNwait until after my commit
01:01:08linuxstblostlogicx: How much faster is Tremor now compared to CVS?
01:01:25linuxstbYou seem to be report 1% increases every few hours
01:02:03lostlogicxunfortunately I've also found that the total is less than the sum of the parts :(
01:03:10lostlogicx8% probably
01:03:28linuxstbSo what is a typical boost ratio now?
01:04:23lostlogicxon my higher bitrate test song 50-51 (q7, 239kbps avg)
01:05:10 Join DJDD_ [0] (
01:05:28lostlogicxon my lower bitrate song *checks* looks like 44-46
01:05:45linuxstbThat's comfortable.
01:06:12linuxstbAre you going to move onto faad next? :)
01:06:16lostlogicxyeah, it's getting there... still not nearly as fast as MAD
01:06:31lostlogicxtbh, I'm probably done after Tremor, because I only have Oggs in my collection :-\
01:07:07linuxstbWell, your profiling work will definitely be use for whoever tackles faad.
01:07:13 Quit yngwi ("Chatzilla [Firefox 1.5/undefined]")
01:08:14lostlogicxI hope so. There are definitely limitations with profiling because of the extra jumps involved causing slightly different code paths, but at least it's something.
01:08:48lostlogicx217kbps average song is ending the full song at 43% boost.
01:12:30ratpack91lcd patch worked :)
01:17:07 Quit linuxstb ("CGI:IRC")
01:23:10 Join elinenbe [0] (
01:29:41 Quit ratpack91 (Read error: 104 (Connection reset by peer))
01:30:49 Join ratpack91 [0] (
01:31:30 Quit San||Away (Read error: 110 (Connection timed out))
01:40:43 Join linuxstb [0] (
01:43:54 Join YouCeyE [0] (
01:47:09***Saving seen data "./dancer.seen"
01:49:17 Quit Kohlrabi ("Leaving")
01:51:10 Quit miner49er (Remote closed the connection)
01:54:39elinenbeLinusN: now with the faster screen updates, up the max scroll speed ;)
01:56:34nathanhis the faster screen updates due to the inlining of lcd funcs?
01:57:06nathanhwoo, i recall some discussion about that
01:57:25nathanhsome timing issues... i guess they were resolved
01:57:41LinusNyeah, but it still isn't as fast as it could be
01:57:55LinusNsince the iriver engineers screwed up again
01:58:14*nathanh expresses complete shock
01:58:30nathanhwhat did they do wrong?
01:58:59LinusNthey connected the Register Select pin to the A1 address pin
01:59:13nathanhhar har har
01:59:39LinusNif they had connected it to a higher address pin, like A4, we could have made burst accesses
01:59:55lostlogicstupid iRiver engineers.
02:00:02nathanhaye, i understand the implications :-)
02:00:14LinusNbecause of this, we can't use the high speed ram mode in the lcd
02:00:35nathanhhrm, hardware hack + rockbox...
02:04:08lostlogicnew battery day.
02:04:27YouCeyEhi LinusN
02:04:59YouCeyEany docs on programming embedded systems?
02:05:11YouCeyEu guys do lot of work.. is it documented to learn about it
02:05:18YouCeyEtutorial kind of thing
02:05:24LinusNYouCeyE: i dunno
02:05:41ratpack91ee degree helps :)
02:05:44PaulJI think the LCD brightness patch for H300 needs to be modified to work with the most recent source
02:05:57LinusNi believe there is an o'reilly book on the subject
02:06:15ratpack91c an c++ in embedded systems
02:06:26LinusNnever read it though
02:06:26lostlogicLinusN: as far as coding standards −− I'm puting a lot of assembly into Tremor to make it suck less, how should I format this? Having #ifdefs all over is ugly, but keeps the assembly near what it replaces which is good, thoughts?
02:06:47LinusNlostlogic: it's really up to you
02:07:04lostlogicLinusN: alright −− I'll try to make it Not Ugly (TM)
02:07:38YouCeyEratpack91, i am in comp sci
02:07:41LinusNlostlogic: we try not to restructure the codec source code, to be able to update them
02:07:49YouCeyEbut didnt work much in programming embed sys
02:07:56YouCeyEc and c++ is fine..
02:08:03YouCeyEhardware side is the problem
02:08:32LinusNYouCeyE: hacking a c64 for years helps too
02:08:46*LinusN has no degree whatsoever
02:08:51lostlogicLinusN: Hmm, that would be quite difficult to avoid when replacing things like ++ and −− with macros, but I'll keep it as close as possible so that patch rejects are obvious.
02:09:02lostlogicLinusN: you're a pimp though :)
02:09:43YouCeyEi believe one simple document explaining one feature in rockbox can help starters
02:09:52LinusNnext mtv show: "pimp my player"
02:09:54YouCeyEmissing the link b/w hardware and s/w
02:09:59YouCeyElol LinusN
02:10:21YouCeyEcoat the dap with gold and diamonds?
02:10:31lostlogicYouCeyE: just read some code :)
02:10:34LinusNbejeweled :-)
02:10:52YouCeyEwill check it out lostlogic
02:11:04YouCeyEall source in cvs, right?
02:11:19LinusNYouCeyE: the trick is to understand how the hardware works, but you don't always need that
02:11:41YouCeyELinusN, yes.. but starting is the problem
02:11:47LinusNembedded programming is much about knowing the limits
02:11:50YouCeyEneed some cheap thing to test programming
02:12:02nathanhan h300 with spinners and hydraulics
02:12:22YouCeyEcant afford to play it on my iriver..
02:12:26ratpack91you can get some evaluation boards
02:12:44YouCeyEratpack91, yes.. i had one board.. c166
02:12:55nathanhyouceye: rockbox programming is safe, only the bootloader is dangerous
02:12:59YouCeyEshld find it where i kept it
02:13:10lostlogicstupid coldfire needs separate add instructions for dest Ax vs dest Dx.
02:13:20YouCeyEso.. rockbox code is same for all players..
02:13:30nathanhcode is, binaries are different for each
02:13:31YouCeyEonly boot loader is diff for players..
02:13:47nathanhboot loader is also diff for players, some players dont even need a boot loader
02:13:50YouCeyEdo u guys use cross compiling stuff?
02:14:24LinusNwe have to, since the cpu in the iriver isn't the same as in the pc
02:14:30 Quit DJDD_ (Read error: 110 (Connection timed out))
02:14:43YouCeyEwaht are good ides in linux? like visual studio
02:14:57LinusNi have no idea
02:15:13LinusNi use emacs and a bash shell
02:15:14YouCeyEnathanh, i use vim but want to try some ide
02:15:24LinusNsome say eclipse is good
02:15:30YouCeyEisnt it for java?
02:15:43nathanheclipse is for any language
02:15:47LinusNit is configurable
02:15:50nathanhthere are plugins for pretty much everything
02:16:04nathanhi wouldnt bother, stick with vim if you know vim
02:16:08YouCeyEcool.. initially it was for java.. i thought.. it was a few years ago..
02:16:15YouCeyEnathanh, i am good with vim..
02:16:32YouCeyEjust for debugging stuff..
02:16:40YouCeyEi use ddd sometimes
02:17:09AciDYouCeyE > you can try kdevelop or anjuta
02:17:35YouCeyEanjuta.. checking it out
02:17:36AciDand to be honest, the c/c++ plugin for eclipse is very, very bad
02:18:40YouCeyEu guys inspire me to learn some embedde sys programming..
02:18:47*YouCeyE searching for my c166 board
02:18:50nathanhgo for it, get the precompiled bootloader installed
02:18:59nathanhthen follow the instructions to download the cvs and compile a rockbox.iriver binary
02:19:26nathanhonce the bootloader is installed, its pretty much impossible to brick the player with rockbox.iriver
02:19:35nathanhso you can do whatever changes you like
02:20:48AciDI wish I could easily get my hands on an h3x0..
02:27:41YouCeyEk.. found my board.. phytec c166
02:28:57 Quit ratpack91 (Read error: 104 (Connection reset by peer))
02:30:24 Join ratpack91 [0] (
02:37:05YouCeyEwhat is the best realtime OS on linux?
02:37:13YouCeyEused to use vxworks in windows
02:37:37YouCeyEi know that in linux there are alternatives like tinyos ..
02:38:23LinusNi don't know
02:38:37 Join Hooligan [0] (
02:38:56LinusNeverytime i have used a commercial realtime os on a platform, my customer has always used windows as a dev platform
02:41:43HooliganForgive me for asking, but are there currently any immediate benefits to the H300 LCD speed increase, or will that help for future additions
02:45:05LinusNslightly less lag when scrolling, longer battery life
02:45:14LinusNfaster rockboy
02:45:42HooliganOh, thought Rockboy's problems lied primarily in it's lack of optimization.
02:45:55LinusNthat too
02:46:41elinenbeLinusN: what is the theoretical speed increase for the high speed RAM mode −− if iriver had done it properly?
02:48:02elinenbealso seeing that I ONLY use my iriver for rockboy (just kidding), what's the speedup with the new LCD code?
02:49:14LinusNthe rockboy speedup will probably not be substantial
02:50:15LinusNthe theoretical speedup if we could use highspeed mode and burst access...i'd guess roughly 50%
02:50:33LinusNi haven't calculated though
02:50:54LinusNperhaps even more
02:50:56HooliganLinusN: Wow. Time to rewire my H320.
02:51:51mirakAciD: I agree the plugin suck. But I still use it, there is nothing else valuable. I just disable the indexing otherwise it's unusable.
02:52:08mirakI mean the DCT C c++ plugin for eclipse
02:52:14AciDah yes
02:52:19HooliganLinusN: Would it be possible to add a function to RB that grabs the filename of the currently playing track, searches for a text file in the directory with the same name, and displays it? I know enough C to know that it's not hard, but not quite enough to code it yet...
02:52:45mirakAciD: I tried anjuta it just crashes. Kdevelop is horrible
02:52:51LinusNyou can do it with a plugin
02:53:33AciDmirak > you just discovered the awful truth, there is no good IDE for c/c++ on unix...
02:53:35mirakHooligan: it would be nice to have the album cover in background :)
02:53:54mirakAciD: I shot everyone that says vi
02:54:00mirakbang bang :)
02:54:10AciDnor emacs
02:54:20mirakxemacs as well
02:55:07mirakis there a patch around or maybe it's done, to have a key press to just enable backlight without activating a function
02:55:44mirakactually I use old key, but well the behavior of the original firmware is nice in this regard
02:56:39Hooliganmirak: I don't know, but I've just been turning key lock on and off to turn on the backlight.
02:56:57mirakHooligan: yes that's the hold key, not old
02:57:32Hooliganmirak: Well, that makes a lot more sense then.
02:57:58mirakI could have say the "ancient" key
02:58:03mirakas well
02:58:22HooliganNo need to hurt the key's feelings. It feels elderly enough.
02:59:09ratpack91hold switch's feelings are already hurt after have been called a key
02:59:30HooliganFair enough.
03:01:30LinusNtime to sleep
03:01:40LinusNnite all
03:01:48 Part LinusN
03:04:47 Quit mirak ("Ex-Chat")
03:05:35 Join actionshrimp [0] (
03:06:49 Join amiconn_ [0] (
03:06:54 Quit ratpack91 ("Leaving.")
03:08:43 Quit XavierGr (Read error: 104 (Connection reset by peer))
03:09:05 Join San [0] (
03:23:17 Quit amiconn (Read error: 110 (Connection timed out))
03:23:17 Nick amiconn_ is now known as amiconn (
03:40:35 Quit San (Read error: 110 (Connection timed out))
03:47:03 Quit Hooligan (Read error: 104 (Connection reset by peer))
03:47:13***Saving seen data "./dancer.seen"
04:09:18 Quit phaedrus961 ("Leaving")
04:12:12 Join phaedrus961 [0] (
04:12:13 Join Mongey| [0] (
04:16:26Mongey|woah go Linus
04:23:37 Quit `3nergy (Read error: 104 (Connection reset by peer))
04:24:00 Join z35 [0] (
04:25:31 Join `3nergy [0] (
04:26:17 Join XavierGr [0] (
04:30:12 Join Rob- [0] (
04:54:17 Quit TCK (Read error: 110 (Connection timed out))
04:54:19 Nick paugh is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
05:11:11 Join ehntoo [0] (
05:11:52 Quit Mongey| ()
05:15:24 Join San [0] (
05:25:33 Quit Rob2222 (Read error: 113 (No route to host))
05:39:05Rob-Does anyone here have an iriver h300 who could give me the sequence of keys for flashing? I am blind and would like to install rockbox.
05:39:31lostlogicRob-: sure, give me a minute to boot the stock firmware.
05:41:46lostlogicRob-: Long press menu, short press menu, down 4 times, short press menu, left, short press menu then wait about 60s for the flash process to complete (I believe you hear the hdd spin down when it's complete)
05:42:41lostlogicRob-: then you should be able to boot rockbox by just pressing play, or boot the original firmware by holding record and pressing and holding play for about ten seconds
05:42:53Rob-which is menu?
05:43:04lostlogiccenter button is menu
05:43:11nathanhdont forget to put rockbox.iriver file and .rockbox directory onto the H300 before flashing
05:43:39Rob-ok, I'll give it a go, thanks!
05:43:44lostlogicgood luck!
05:45:50YouCeyEcan i put rockbox.iriver file in .rockbox directory?
05:46:01YouCeyEwill it work?
05:46:22YouCeyEthen i think it makes sense to include everything in .rockbox folder
05:46:32YouCeyEi unrared it separately
05:46:38YouCeyEand missed to include bootloader
05:46:57nathanhrockbox.iriver is not the bootloader
05:46:58lostlogicYouCeyE: bootloader is part of firmware, rockbox.iriver is firmware file
05:47:08lostlogicermh yeah, ignore me, I speak with ass.
05:47:16***Saving seen data "./dancer.seen"
05:47:48YouCeyEyes.. but its easier to missout that boot loader file in the first try
05:48:11YouCeyEbeing a user perspective, i didnt know details initially
05:48:27nathanhits not really ready for users, its still in heavy development
05:48:38YouCeyEk :)
05:49:28YouCeyEi wish iriver had clock support
05:49:37 Quit actionshrimp ("a bird in the bush is worth two in your house")
05:54:04 Quit San (Read error: 110 (Connection timed out))
05:58:56 Quit nathanh ("CGI:IRC (EOF)")
06:06:48 Quit z35 (Connection timed out)
06:07:41 Join z35 [0] (
06:34:10 Join TCK [0] (
06:41:35 Join nathanh [0] (
07:06:54 Quit DreamTactix291 (Read error: 110 (Connection timed out))
07:21:29 Join DreamTactix291 [0] (
07:23:22 Quit TCK (Read error: 104 (Connection reset by peer))
07:27:37 Join DT291 [0] (
07:28:22 Nick DT291 is now known as DreamTactix91 (
07:28:34 Quit DreamTactix291 (Nick collision from services.)
07:28:37 Nick DreamTactix91 is now known as DreamTactix291 (
07:32:53 Join San [0] (
07:41:44 Join Paul_The_Nerd [0] (
07:47:20***Saving seen data "./dancer.seen"
07:50:40markunmorning Bger
07:59:40Bgermarkun i need a lil' bit help ...
08:00:45BgerLD rockbox.elf
08:00:58Bger/home/Bager/Rockbox/lcdh300/rockbox-devel/build/apps/plugin.o:(.rodata+0x78): u'
08:01:08Bgercollect2: ld returned 1 exit status
08:01:08Bgermake[1]: *** [/home/Bager/Rockbox/lcdh300/rockbox-devel/build/apps/rockbox.elf]1make: *** [all] Error 2
08:02:39Bgeri really don't get what's the problem
08:07:42markunhm, me neither..
08:08:17 Quit RotAtoR ("zzzzzzz")
08:08:59Bgeri have this problem for 2nd time ...
08:09:36Bgeramiconn ?
08:10:43Bgermarkun at least do you know how can i tell LD to be more verbose ?
08:20:39markunnot even that, but I'm a bit busy with packing because I have to catch a plane.
08:21:04BagderBger: make V=1
08:21:17Bagderthat at least shows the full command line
08:22:28 Quit Nibbler (Read error: 113 (No route to host))
08:26:10Bger10x, Bagder
08:26:23Bgerhm, i see that linus has commited my lcd-turn-off patch ...
08:27:00Bgerbut it needs more work...
08:33:36 Quit ze (Read error: 104 (Connection reset by peer))
08:34:44 Join ze [0] (
08:35:02 Join ender` [0] (i=ychat@
08:48:42 Quit Paul_The_Nerd ("Leaving.")
08:49:43 Join webguest49 [0] (
08:49:46 Join _FireFly_ [0] (
08:49:47 Join vHs_ [0] (
08:51:09 Quit webguest49 (Client Quit)
08:51:14 Join webguest68 [0] (
08:52:11 Quit webguest68 (Client Quit)
08:53:00 Join bytie [0] (
08:55:12 Quit bytie (Remote closed the connection)
09:00:19nathanhbger: why does it need more work?
09:00:42Bgerbecause i want to make lcd_init_device shorter
09:01:08Bgerand also we don't need to reset the LCD on power-on
09:01:22Bger(this saves 3 or 4 * sleep(1))
09:02:47 Quit vHs (Read error: 110 (Connection timed out))
09:04:45 Join Rob2222 [0] (
09:07:06 Join tvelocity [0] (
09:17:12 Quit markun ("off to Italy :)")
09:20:15 Join Nibbler [0] (
09:21:42Bagdertoday is the year's shortest day
09:21:45Bagderup here
09:23:14Bagdersunrise 08:47 sunset 14:47
09:23:49 Join saratoga [0] (
09:23:50tvelocityi'm usually not even awake at 14:47
09:24:20tvelocitythat's kinda depressive
09:24:23Bagderthen I'm in Stockholm, further north they have even shorter days
09:24:39tvelocityfinding that it's allready night when you wake up
09:24:49tvelocityruins your day
09:24:58Bagdergoing to work: darkness, going home from work: darkness
09:25:02tvelocitywell, night, even...
09:26:21Bger14:45 ...
09:29:05 Quit saratoga ("CGI:IRC (EOF)")
09:32:58perplexityOk.. some ADC values for the h300-non-lcd-remote.. 0x00 == HOLD ; 0x18 == |>> (skip forward) ; 0x33 == <<| (skip back) ; 0xC8 == Vol- ; 0xE4 == Vol+ ; 0xFF == Not inserted or idle
09:33:59perplexityWith the remote selection patch I have to select H300LCD remote, then H300-Non-LCD remote to get it to work.. else the ADC just always reads 0x00 and no buttons besides play work.
09:34:35 Join warewolf [0] (
09:34:58warewolfhey, anyone know if it's possible to charge via USB w/ rockbox on the H3xx iRivers?
09:35:06warewolfer.. while still playing music
09:35:12DreamTactix291i don't think so
09:35:19warewolfsoftware limitation or HW?
09:35:36perplexityno it's not possible..
09:35:39warewolfthe original iriver firmware didn't let you :/
09:35:56ashridahwarewolf: as we've seen that's not really a good indicator :)
09:36:01perplexityYou only have 500ma available on the usb port, which is likely not enough to properly charge the battery and run the player
09:36:10DreamTactix291not sure if its a hardware limitation or not
09:36:11warewolfperplexity: ah, point.
09:36:15DreamTactix291but perplexity brings up a good point
09:36:16ashridahbut yeah, that'd be my assumption too
09:36:45Bagderbut it could extend the playing time
09:37:25warewolfBagder: possibly... It totally bugged me that the unit itself refused to operate as a USB powered HD and had to use its internal battery when acting as a USB HD
09:37:41warewolfactually, could rockbox do that?
09:37:54warewolfhave the unit run power off the usb when it's just in disk mode?
09:38:12warewolf(I've been really, *really* impressed with rockbox on my H340, thanks guys)
09:38:21BagderI thought it did that already
09:38:49warewolfBagder: for the iRiver firmware I'm positive it ran off the internal battery when acting as a USB hard drive to a PC
09:39:00warewolfBagder: I have no idea how rockbox does it
09:39:12ashridahyou're only really garunteed 500mA. you could probably get more from a powered hub, but doing it from a motherboard would be risky
09:39:13Bagderit always runs off the batteries
09:39:20Bagderthe question is if it charges or not
09:39:24ashridah(and even then, garunteed isn't a garuntee :) )
09:40:10ashridahif only we'd all standardised on firewire
09:40:24warewolfspeaking of
09:40:31warewolfwtf is up with apple removing firewire from ipods
09:40:37warewolfdidn't they help make the damn spec?
09:41:09BagderI bet there were a ton of companies involved
09:41:12ashridahyeah, and then, like betamax, it got canned because no-one adopted it, even if it sucked less
09:41:20ashridahlots of laptops have the gear
09:41:36*Bagder recalls Apple-and-SCSI
09:41:36ashridahbut it didn't penetrate into the pc market weel
09:41:38ashridahwell even
09:41:58warewolfBagder: oh yeah, the big [ ] square connector that had A SINGLE GROUND LINE. *groan*
09:42:20zefirewire didn't? i think all our pc's have firewire ports
09:42:30zethough i think i removed it from mine cause i didn't really use it and needed the free irq's
09:42:31DreamTactix291mine doesn't
09:42:33Bagderports yes, devices no
09:42:37 Quit AliasCoffee ("Leaving")
09:42:53 Join Aramil [0] (
09:42:55DreamTactix291i have a card
09:43:00DreamTactix291but no open PCI slot right now
09:43:01zemy parents just got a dvdr/vcr combo with a firewire port to record from their digital handicam
09:43:06DreamTactix291i really need a new mobo
09:43:24Bagdercameras are the only things I know people actually use fireware for
09:43:26*warewolf makes some tea
09:43:32Bagdervideo cameras
09:43:48*Bagder types like a pro
09:43:53DreamTactix291i'd love to use my HDDs in their enclosures with firewire instead of USB
09:44:41zehmm video cameras, yeah...
09:44:54zei'm pretty sure i've seen plenty of firewire HD's about
09:45:08Bagderyes, but I don't know anyone who use them
09:45:11DreamTactix291mine can be either way
09:45:13zeand there's more firewire audio interfaces coming out
09:45:18DreamTactix291no card in my PC right now
09:45:19zewhich USB really sucks for
09:45:23DreamTactix291no firewire for me
09:45:55zeafaik anyway
09:46:08DreamTactix291i like how you can chain firewire devices
09:46:21zeyou're supposed to be able to chain usb
09:46:28zeup to 128 devices in a chain
09:46:30zeor somesuch
09:46:33DreamTactix291doesn't happen though
09:46:42zei haven't see much of anything with a port for chaining though
09:46:51warewolfUSB is a spoke topology
09:46:57warewolfor so I thought
09:47:05zeexcept for usb hubs of course
09:47:11zeand mac keyboards
09:47:17DreamTactix291and my keyboard
09:47:22***Saving seen data "./dancer.seen"
09:47:23DreamTactix291which isn't a mac keyboard
09:47:24warewolfand some monitors
09:47:31warewolfDreamTactix291: pfucia?
09:47:37zethere's usb monitors?
09:47:43DreamTactix291a really cheap no name one
09:47:47DreamTactix291but it has two USB ports on it
09:47:48warewolfze- some monitors have usb hubs in them
09:49:20warewolf'm just going to ahve to bite the bullet and buy another desk charger for my H340
09:50:36warewolfso has anyone reverse engineered the H3xx firmware to see how the iRiver firmware does the EQ?
09:50:44warewolfI noticed that I get much better bass response with rockbox firmware
09:50:49warewolfwhat'd youall do RIGHT?
09:51:31Bagderwell, we don't do any EQ at all atm, we just set the DAC
09:51:42 Quit San (Read error: 110 (Connection timed out))
09:51:48nathanhiriver with WOW is pretty good
09:51:49Bagderthe bass/treble settings I mean
09:51:56warewolfnathanh: you're no audiophile
09:51:57nathanhthough i suspect they're just using companding
09:52:06nathanhdamn right im not audiophile, audiophiles are all idiots
09:52:15warewolf"wow", "trubass", "srs" are all crap
09:52:21warewolfjust give me a damn eq :)
09:52:28warewolfI'll turn it into a smiley face \_/
09:52:39Bagderdr preglow has one in the pipe
09:53:00warewolfBagder: well the settings on the dac /now/ are impressive enough re: bass
09:53:44warewolfoh, anyone know what the battery miliwatthour rating is on the H3xx units?
09:53:47warewolfis it 1500mw?
09:54:14*warewolf googles
09:55:05Bagdermah is milliamperehours
09:55:29warewolfer, you know what I mean
09:55:45warewolfI know just enough to be dangerous
09:56:03*warewolf groans
09:56:05warewolfThanks iRiver!
09:56:06warewolfInternal Rechargeable Li-polymer Battery
09:56:45*warewolf downloads manual
09:57:08nathanhyeah, you know that \_/ is the height of audiophilia
09:57:33warewolfnathanh: the srs/wow stuff just .. does evil things to the sound
09:58:01 Quit perplexity (Read error: 113 (No route to host))
09:58:08warewolfI can't describe it, it just changes the loudness of certian ranges of the sound that is supposed to make it sound good on cheap headphones
09:58:17warewolfI don't have cheap headphones made in 1980
09:58:33warewolfI have my sony earbuds, or my sony MDR-V60s
09:59:31*nathanh shrugs, my akg405s cant compete with sony earbuds
09:59:33warewolfwhich reminds me I need to buy a new pair of V60s, the left speaker is buzzing
10:00:07 Quit tvelocity (Success)
10:00:11warewolf(they lasted something like 4+ years, and the pleather has all come apart and is fuzzy now :)
10:00:39 Quit Vladoman (Read error: 110 (Connection timed out))
10:01:37warewolfSoB! the english iRiver manual for the h340 doesn't say what the battery is inside
10:01:52 Join Vladoman [0] (
10:02:03Bgerwarewolf i think it's the same as h100's one
10:02:29warewolfBger: you appear to be losing characters in your nick.
10:02:42*warewolf hands badger an a nd a d
10:02:54Bgerno, i'm not Bagder :)
10:03:03*Bagder points to Bger. We are two
10:03:17warewolfWe are Borg.
10:04:25 Quit akaidiot (Read error: 110 (Connection timed out))
10:06:46Bgerwarewolf in fact i'm Bager ... but i changed my nick to be more different than Bagder's one
10:09:30Bagderand I am in fact a bot
10:09:49nathanhexplains why he's here 24x7
10:10:29 Join lamed [0] (
10:10:45warewolfIRC: where the men are men, the women are men, and the little girls are FBI agents.
10:11:12Bgergood one
10:12:11lameddon't you think ipoders might want to pgdown using (BUTTON_SELECT | BUTTON_SCROLL_FWD)?
10:13:56warewolfholy moses
10:14:05warewolfrockbox beats iriver firmware hands down re: supported audio codecs
10:14:53BagderI bet Rockbox beats just about all firmwares for any portable player
10:14:59 Join Zagor [0] (
10:15:53Bgerexcept old archoses with MAS ? ;)
10:16:16Bgerthat is, codec-wise
10:16:50warewolfAOLIM is broke again
10:17:16warewolfI successfully authenticate against their login servers, then their login servers redirect me to a DEAD chat server
10:17:23warewolfseriously, -ECONNREFUSED
10:19:33*Bagder transfers many files from localhost to localhost
10:24:50lamedanyone seen t0mas lately?
10:25:41linuxstblamed: No, "page down" isn't needed with the wheel - you just move the wheel fast.
10:26:27Bagdert0mas ( signed off 3 days and 0 hours ago
10:28:39lamedbagder - idle command?
10:28:50Bagder /MSG logbot seen t0mas
10:29:15lamedlinuxstb - is it fast enough? -don't have an ipod
10:29:20lamedbagder - 10x
10:29:46***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
10:30:00***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
10:30:01 Join Kohlrabi [0] (
10:30:41***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
10:30:58***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
10:31:18***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
10:31:54linuxstblamed: Yes. That's the selling point of the wheel - it makes scrolling through long lists easy.
10:38:07warewolfwhy, on the last day before I go on holiday does one of our secondary servers have to suddenly have a raid mirror go bad
10:39:19Bgerwarewolf murphy's law :)
10:40:33warewolfthe sad thing is I was going to tinker with the box too
10:40:41warewolfand I come to find out that it is sick
10:40:48lamedlinuxstb: I remember a long, long time ago, I've scrolled with an ipod. I was so angry of my h100 player that day.
10:41:08warewolfmaybe it saw me coming and decided to throw a hissy fit.
10:46:45***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
10:46:51***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
10:47:19lamedbagder - why's the logbot ignoring me? ( /MSG logbot seen t0mas)
10:47:49Bagderprobably because you haven't registered at freenode?
10:48:15warewolfI was going to say something about lameness, but I decided to be nice.
10:48:22*warewolf laughs
10:50:06lamedno, i'm registered
10:50:19warewolfholy honking horseshit is freenode's IRCD chock full of crap
10:50:45warewolfand I thought the +abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ1234567890 modes opers get was bad already
10:50:50***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
10:51:06*warewolf knows some freenode staff, or atleast ex-staff
10:51:15 Join tvelocity [0] (
10:51:47***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
11:08:25 Join Polo_o [0] (
11:08:35lamedbtw warewolf, lamed is my real hebrew nickname. it is saied more like laa-med, but nevermind that you can always pick on me.
11:08:58***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
11:09:01 Quit Aramil (No route to host)
11:09:40 Join ratpack91 [0] (
11:11:40warewolflamed: don't worry about it, I didn't know how to spell "werewolf" when I picked this nick.
11:12:40warewolflamed: so feel free to pick on me too. I'm sure you'll find some choice-quality misspellings of mine (which I am notorious for) that will amuse you.
11:20:40lamedwell I still can't get the logbot to answer me... ta-ta
11:20:42 Quit lamed ("CGI:IRC")
11:29:40 Join San||Away [0] (n=test@
11:47:26***Saving seen data "./dancer.seen"
11:50:32 Quit ratpack91 ("Leaving.")
11:50:37 Join mirak [0] (
11:52:23 Join einhirn [0] (
11:52:47 Join Jungti1234 [0] (n=jungti12@
11:54:50 Quit _FireFly_ ("Make it idiot proof and someone will make a better idiot.")
11:58:42 Join webguest80 [0] (
11:59:23webguest80hi all
11:59:35webguest80I'm looking for a MAS assembler
11:59:42webguest80may you help me?
12:00:30VladomanYou don't want it :-)
12:02:56Bgerwebguest80 i don't think there is such thing for free ...
12:03:05webguest80k I see
12:03:33webguest80it seems kind of nasty to write one
12:03:46Bgerbut do you have the specs ?
12:04:22webguest80I've already did a disassembler
12:04:27Bgerwhat ?
12:04:30Bgerfor which MAS ?
12:04:42webguest80the one inside AV300
12:04:53Bgerwhere did you get info from ? :)
12:05:03webguest80still don't know the meaning of registers
12:06:05Bgeramiconn ?
12:06:12webguest80I'm looking in bookmarks
12:06:13 Join dwihno [0] (n=dw@
12:06:14webguest80just wait
12:08:01 Join Aramil [0] (
12:09:19 Quit tvelocity (Connection timed out)
12:10:18amiconnBger, webguest80: There is a MAS assembler. It's not free, but leaked on the net somewhere
12:11:15Bgeramiconn i just wanted to attract your attention...
12:11:48webguest80wait for the site as this is not my PC so I'vent got it right now
12:11:54webguest80instead look at my disass
12:12:02webguest80in linav cvs
12:13:02webguest80it's not the current version as they still have to upload the last I've sent
12:25:32webguest80ok... can't find the site
12:25:35 Join Sando [0] (
12:25:38 Part linuxstb ("Leaving")
12:25:51webguest80anyway I've got at least two interesting pdf
12:25:58 Part Sando
12:26:14webguest80one describes the instruction set
12:26:21webguest80and the pipeline
12:26:48webguest80the latter describes (only partially) the interfaces
12:27:26webguest80if you give me an address I'll be happy to send you infos
12:29:02webguest80the disassembler in cvs can correctly decode the uploaded code (2 blocks of memory used for mp3 decoding)
12:29:53Bgeramiconn are you interested in this ?
12:29:57webguest80unluckily there's no way to get access to the program ROM so we can't disassemble the firmware
12:30:17Bgeraha :(
12:30:36webguest80no way about it
12:30:57webguest80the dsp has distinguish 3 different memory areas
12:31:02webguest80D0 (Data 0)
12:31:06webguest80D1 (Data 1)
12:31:12webguest80P (Program code)
12:31:38webguest80Program code is not readable, even with dsp code
12:32:12webguest80it's only used for instruction decoding
12:33:19Bgerthey've protected themselves
12:35:04webguest80we still have 4k words of RAM
12:35:11webguest80that can be used for code
12:35:51webguest80My target was being able to use them for emulating old sound chip
12:35:52 Quit thegeek (Read error: 104 (Connection reset by peer))
12:36:14webguest80such as 6581 of c64 or similar (NES chip etc..)
12:36:24webguest80for emulation purpose
12:37:12 Join petur [0] (
12:37:17Bgerheh, this is good, but no vorbis for example ...
12:37:21Bgerpetur, hi :)
12:37:48peturI saw your LCD patch was committed
12:37:51webguest80well, why not?
12:37:59Bgeryes, i saw that too
12:38:04webguest80dunno if it can stretched in 4k
12:38:13peturdidn't get time to try it yet
12:38:15webguest80but maybe it can be written...
12:38:30webguest80I'm not much in codec writing anyway so I dunno
12:38:34Bgerwebguest80 that first and 2) maybe it doesn't have enough power
12:38:35 Join thegeek [0] (
12:39:06Bgerpetur i'm looking at making it to power on quicker...
12:39:24nathanhbger: your lcd patch is awesome... longer battery life!
12:39:50peturI hope to get my patch out of the door tonight, then I'll adapt the brightness patch
12:39:54Bgernathanh, yes, i felt that too, but didn't make any measurements
12:40:02webguest80that's all anyway...
12:40:11webguest80the beast is kind of strange to program
12:40:22Bgerwebguest80 good info
12:40:24peturlast night was bad, no programming after contact with beer ;)
12:40:28webguest80as you must pay attention to the pipeline
12:40:31Bgerhehehe petur :)
12:40:58peturBger: only online during the day?
12:41:07Bgerwebguest80 i think people that could be interested in your work aren't here atm :(
12:41:11 Join edx__ [0] (
12:41:11Bgerpetur yeah :(
12:41:24webguest80hmm ok
12:41:33webguest80one these days I'll bne back
12:41:41nathanhim interested :-)
12:41:44webguest80and I'll give the address
12:41:57Bgervery good
12:41:59webguest80of the site where I got the inofs
12:42:44webguest80fyi... about the strange way it works...
12:42:55Bgeris there any chance of older MASes being compatible with this one ?
12:43:15webguest80all 35xx seems to be code compatible
12:43:31webguest80the only changing thing
12:43:32Bgerolder MASes = ones in the players rockbox runs on
12:43:40webguest80seems to be the registers
12:44:06webguest80every 35xx dsp has 256 registers
12:44:14webguest80they are used to setup the chip
12:44:25Bgerah, yes, they are all 35xx series
12:44:40webguest80and only some of them are fix in all version
12:45:00webguest80still have to figure out the meaning of lot of them
12:45:27 Quit San||Away (Read error: 110 (Connection timed out))
12:45:44Bgerthis will be hard work...
12:46:00amiconnI guess the MAS performance would be sufficient to decode vorbis, as it even manages to *encode* mp3 realtime
12:46:14webguest80I agree
12:46:15amiconnHowever, 4K words RAM aren't enough
12:46:27webguest80we're speaking of instruction of 20 bit
12:46:37webguest80a word is 20 bit in dsp
12:46:42amiconnyes I know
12:46:51webguest80so we have 4096 instructions... not bat
12:47:10webguest80look in commodore scene what they can do in 4k :D
12:47:11nathanhwait, so the iriver stock firmware uses the MAS, and rockbox is using the CPU solely?
12:47:18Bgernathanh nooo
12:47:23Bgerthere's no MAS in irivers
12:47:32nathanhokie dokie
12:48:30amiconnwebguest80: Vorbis needs large tables which won't fit. Same with WMA
12:48:43webguest80ok I see
12:48:46webguest80I didn't know
12:49:04webguest80this can be a big problem
12:49:27webguest80the dsp hasn't got room for great things
12:50:15webguest80but it's great in fast calculations
12:50:27webguest80it's even able to do MAC
12:52:05mirakamiconn: the mp3 even in 320kbps is really poor quality
12:52:14mirakI mean in record mode
12:52:58amiconnThe MAS only generates frames up to 192kbps, but the quality is rather good
12:53:41webguest80ok... that's all
12:53:45webguest80gotta go now
12:54:21 Quit webguest80 ("CGI:IRC (EOF)")
12:54:22 Join webguest08 [0] (
12:54:44webguest08I've fallen
12:54:58webguest08I was saying
12:55:17webguest08that maybe I've found someone who wants to help to write a MAS assembler?!?
12:55:26mirakwhat's the MAS ?
12:55:52 Quit webguest08 (Client Quit)
12:56:44Bagderthe mp3 decoder/encoder chip of the Archos models
12:59:10 Quit edx (Read error: 110 (Connection timed out))
12:59:15nathanhmirak: its like a mini cpu, with its own ram (4kB), and you feed it data and it spits data back out
12:59:22Jungti1234I will come again
12:59:22 Quit Jungti1234 ("bye")
13:09:02 Join Jungti1234 [0] (n=jungti12@
13:11:06 Quit Aramil ("Leaving")
13:13:30 Join amiconn_ [0] (
13:14:34 Join Sando [0] (
13:15:12 Part Sando
13:18:41 Join LinusN [0] (
13:24:07 Join perplexity [0] (n=joust@
13:25:12 Quit ehntoo (Read error: 110 (Connection timed out))
13:25:49 Join [IDC]Dragon [0] (
13:27:56*[IDC]Dragon saw some MAS discussion
13:30:16Jungti1234PLAYTIME 150 Hrs
13:31:18 Quit amiconn (Read error: 110 (Connection timed out))
13:31:18 Nick amiconn_ is now known as amiconn (
13:34:08 Part LinusN
13:44:53 Join Sando [0] (
13:45:35 Part Sando
13:45:40 Join Sando [0] (
13:45:54 Part Sando
13:46:33 Join Sando [0] (
13:47:25 Quit nathanh ("Leaving")
13:47:30***Saving seen data "./dancer.seen"
13:53:11*amiconn spots [IDC]Dragon
13:53:21amiconnWb, ltnirc
13:57:47 Join nathanh [0] (
13:59:05 Quit perplexity (Read error: 113 (No route to host))
14:08:11 Join apgap [0] (
14:12:17[IDC]Dragonhi amiconn
14:13:47Bgerhow should i read this (apps/settings.c)
14:13:49Bger /* how many bits within the bitfield (1-32), MSB set if value is signed */
14:13:51Bger unsigned char bit_size; /* min 6+1 bit */
14:14:19Bgerthis is part of the definition of struct bit_entry
14:14:42*[IDC]Dragon recognizes his code
14:14:45Bgerit says "min 6+1 bit" ?
14:15:02Bgerbut there are many settings which use 1 bit for example
14:15:54[IDC]Dragonwithout looking at it, this is probably the min width for the member
14:16:15[IDC]Dragon6 bit=0..63 plus sign
14:16:46Bgerso if i write something like {1,S_O(val),...
14:16:55Bgerit'll be min 6 bit on the disk ?
14:18:05*[IDC]Dragon looks
14:19:08[IDC]Dragonthe comment meant the sizes of the strict member itself
14:19:35Bgerbut if i say 1 bit, it'll be 1 bit on the disk ?
14:19:51[IDC]Dragonwe want bit_size to be able to express 1...32 and have an extra bit to mark it as signed
14:20:11Bgerah, okay
14:20:13Bger10x :)
14:20:42[IDC]DragonI wanted that struct as compact as possible
14:20:55[IDC]Dragonsince the array is large
14:21:41[IDC]DragonS_O is just a convenience macro
14:22:02[IDC]Dragonexpanding into 2 parameters, size and offset
14:22:24Bgeryes, i saw this already, 10x :)
14:24:52Bgerwhat happens if i have #if CONFIG_REMOTE_KEYPAD==H100_REMOTE)
14:25:04Bgerwhat happens if i have #if (CONFIG_REMOTE_KEYPAD==H100_REMOTE)
14:25:13Bgerand CONFIG_REMOTE_KEYPAD is undefined ?
14:25:43Bgerno preproc. error ?
14:25:48Bagderyou should not do that comparison if it is undefined
14:26:06Bagdergcc will treat the undefined ones as zero in #if expressions
14:26:16Bger10x :)
14:26:16Bagderso #if undef == undef will be true
14:26:56Bagderuse -Wundef :-)
14:27:02Bagder"Warn if an undefined identifier is evaluated in an #if directive."
14:27:25peturor put #ifdef around it
14:28:44Bger#if (defined) && (...) ...
14:28:49 Join San [0] (n=test@
14:28:53Bgeroh, #if defined(...)
14:31:03 Nick Lynx_awy is now known as Lynx_ (
14:43:07 Join dropandho [0] (
14:46:12 Join Mongey| [0] (
14:46:13XavierGrNow isn't this great? I just fixed my linus account (I had changed the home dir to rrot and couldn't login) now I destroyed the root account (I cahnged the ownership of all folders to my account and now the root can't log). And all this to be able to view the contents of my folders in Ubuntu.
14:46:31dropandhohey all!
14:46:33XavierGropps s/Linus/Linux
14:46:53dropandhoanyone happen to be working on more remote screen porting- seems that has lost some steam
14:47:53dropandhonot much advancing on that format at all these days really
14:48:19XavierGrwell only the recording screen is left I think....
14:48:35XavierGroh and the virtual keyboard.
14:48:45crwlXavierGr, after that ownership fuckup, you might as well reinstall
14:49:09XavierGrbut truth is that TiMiD can't be found these days, though with his work it could be easy to port the rest of the screeens.
14:49:34dropandhoyeah- and that seems to be a piece is missing for sleep timer
14:49:43dropandhogot it- was wondering who did the work on that
14:49:55XavierGrcrwl: Well I thought that was Linux all about. Not to have to re-install as frequently as Windows.
14:50:08Bagderit is
14:50:18BagderXavierGr: but then you shouldn't mess it all up
14:50:53dropandhoany ideas what is missing for the 1st official irver release?
14:51:02dropandho(for 120/40 that is)
14:51:17peturdropandho: pre-recording and a fix for the recording glitch is coming soon (both H3xx and H1xx)
14:51:33dropandhoohh nice
14:51:42XavierGrBagder: With all respect to Linux, I can't see how it can be easy for PC newbies to run Linux. I have quite some experience with PCs and I have quite a hard time to do what I want.
14:51:43dropandhoand then maybe an official release of it all?
14:51:57BagderXavierGr: so?
14:52:14BagderLinux just rocks, I don't care if newbies use it
14:52:40crwlyou just should have avoided using root privileges (except when installing programs) and you would have been fine
14:52:58BagderXavierGr: you can mess up your windows install too
14:53:01XavierGrI thought that Linux was moving more to the newbie side to be able to expand its userbase.
14:53:03crwlit's not possible to screw up system file permissions as normal user :P
14:53:18BagderXavierGr: it is a lot more user-friendly now than ever before
14:53:26Bagderand it'll be so even more in the future
14:53:41Bagderusing Linux is _your_ benefit, not some else's
14:54:09XavierGrI didn't meant the opposite. It is free after all.
14:54:27XavierGrAnd I really want to use Linux one day and get rid of Microsoft
14:54:30amiconnLinux == user-friendly? Hmm, somehow I doubt that
14:54:42amiconnIt's really flexible, yes, but user friendly?
14:54:45Bagderit is user-friendly
14:54:49BagderI am a user
14:55:28Bagderjust trying to put things in perspective
14:56:06BagderI want all these powers and possibilites that windows won't let me
14:56:50 Quit z35 ("Leaving")
14:57:06XavierGrLinux is a whole different beast than Windows, while I know many things about windows still I can't do nothing in Linux after 3-4 hours of reading and fiddling with a terminal.
14:57:26Bagderit is quite different, yes
14:59:24*amiconn wonders what exotic things Bagder wants to do that aren't possible in Windows
14:59:47XavierGrAnd that packet manager seems really annoying to me. I want to be free of installing a program. That is a major reason why I dislike Windows, only to find out (until now) that when I want a program I have to install it via the packer.
14:59:47Bagdersimple things like configuring my windows manager
15:00:29Bagderspeedy rockbox builds! ;-)
15:00:42BagderXavierGr: that is not true
15:00:43amiconnThat's not Window's fault
15:01:38*petur slaps cygwin around
15:01:47peturslow as hell
15:01:57Bagderstill, using Linux fixes it
15:02:04Bagderas so many other things are better in linux
15:02:22Bagderimho of course
15:02:42peturbuilding stuff in MS DevStudio is also very fast, does that make Windows better?
15:03:01Bagderwhat a comparison
15:03:08XavierGrThe only real drawback of converting right now is the major loss of my software and some minor loss of hardware.
15:03:12Bagderls is faster than dir!
15:03:18Bagderpetur: nope
15:03:25Slasherihehe, i only get a headache if i have to use windows.. It's so unpleasant when you have to do everything using the mouse..
15:03:51Bagderyes, back to the window manager argument
15:03:52XavierGrSlasheri and what if you don't know the commands?
15:04:23SlasheriXavierGr: well, i can't say but i have found man pages and google for bigger problems helpful
15:04:32*petur hugs his right-click context menus
15:04:58SlasheriXavierGr: and of course with a window manager (such as kde for example), you can do (if you want) most of the things with the mouse
15:05:33XavierGrSlasheri that's what I am talking you need major training (even doing this alone) before one is capable of using Linux.
15:06:00SlasheriXavierGr: that might be true.. but i am not probably the right person to say that :)
15:06:11crwlone needs major training before being capable of using windows too
15:06:13Bagderwell you've trained for windows too
15:06:17Slasherianyway, man pages are useful because they include the "see also" section
15:06:23crwland it seems you're confusing administrating and using
15:06:35Slasheriso often it's enough to remember one command and look at the see also section
15:07:05 Quit nathanh ("not this shit again")
15:07:09XavierGrBagder: The training process in Windows was quite different. The user can do many things with the GUI and then he can explore various options as he goes.
15:07:22BagderXavierGr: and that is different?
15:07:58Bagderand I didn't say the learning would be the same
15:08:01XavierGrWell I have to use the terminal and copy paste things from a FAQ that I can't really understand (not to mention remember)
15:08:11Bagderjust that you've spent lots of time learning windows too
15:08:26Bagderyou don't "have to" use the terminal in most cases
15:08:51XavierGrI have the opposite experiance.
15:09:09XavierGrAll examples in the Ubuntu starter guide needs a terminal command
15:09:10Bagderwell I of course prefer everything command line
15:09:26Bagderand I don't use ubuntu ;-)
15:09:53XavierGrYes I know what you mean, once you remember all those commands you are a full fledged user bat what until then?
15:10:15XavierGrBagder: What's your disto?
15:10:26Bagderdebian unstable
15:10:47XavierGrwell Ubuntu I have heard that is quite similar to debian....
15:10:58Bagderyes, it is based on debian
15:11:09Slasheriubuntu has the same package manager, but the installation procedure should be simpler
15:11:17Jungti1234hehe bye
15:11:23 Quit Jungti1234 ("bye")
15:11:30crwlubuntu is mostly debian with a release schedule
15:11:36SlasheriXavierGr: have you tried kubuntu? It should come kde installed as default iiuc
15:11:37crwltherefore, a very good distro
15:11:44XavierGrinstalling Ubuntu was quite simple, yes I agree on that.
15:12:06crwlkubuntu is of course even better, because gnome sucks :)
15:12:14Slasheriyeah :)
15:12:21XavierGrSlasheri: No I haven't, I will fiddle around some more with Ubuntu and then maybe I will consider kUbuntu.
15:12:22Bagderand emacs beats vi! ;-)
15:12:34SlasheriXavierGr: probably you should try that then
15:12:43Slasherikde much much more user-friendlier than gnome
15:12:57crwlBagder, now that's bullshit ;)
15:13:09Bagderactually, I think the most of the Linux vs Windows is a state of mind
15:13:20crwlSlasheri, i hear they call gnome user-friendly because you can't do much with it
15:13:23Bagderif you decide you love windows ways, you'll never like Linux ways
15:13:27crwlespecially configure it to suit your liking
15:13:34Slashericrwl: hehe :D
15:13:41XavierGrThough one thing is for sure. While it made my life harder and I keep swearing at Linux (because I can do almost nothing) I don't regret it because I have earned valuable experience.
15:14:19[IDC]DragonOT: does somebody know a simple way of 1:1 chatting with somebody over http?
15:16:03[IDC]Dragon(without installing a client, registering, be firewalled, etc)
15:16:23XavierGrAlso on another note: Where does the GRUB bootloader files are hanging? I thought that they will be in the boot priority disk, but I can't find them. If I load an old ghost image of Windows I will loose the bootloader and I won't be able to boot to linux. How can I run GRUB again? (without Linux)
15:17:14mirakXavierGr: you gave up ?
15:17:30mirakXavierGr: they are in /boot/grub/
15:17:31XavierGrno I didn't. (yet)
15:18:04BgerXavierGr don't do it!! :)
15:18:05XavierGrbut these files aren't on the boot HD?
15:18:11mirakXavierGr: depends if you ghost image is of the whole disk or only the window partition
15:18:50mirakXavierGr: no, but there is a binary, it's called grub, and it needs some files in /boot
15:18:51XavierGrwell the ghost image is has the windows partition of the priority boot HD.
15:19:25mirakXavierGr: you can do a boot floppy
15:19:42mirakand there is mini isos of grub, with only grub
15:19:46XavierGrMy question is how GRUB is executed if there are no binarys in the boot disk. (My installation of Linux is competely saperated from the boot disk but it loads nevertheless)
15:20:09mirakwhat do you mean by separated ?
15:20:15XavierGrok I should look for them then.
15:20:33mirakwhat's your problem, what do you want to do and why ?
15:21:40 Quit San (Read error: 110 (Connection timed out))
15:21:43mirakXavierGr: I don't think one need major training, just to use it. But if you want to do some weird things then yes.
15:21:58XavierGrI have 2 HDs with 2 partitions each. 1 has windows and it is selected as bootable. The second HD has Linux but it is not bootable. How grub can boot Linux from a disk that hasn't got boot priority?
15:22:02mirakfor exemple a fresh linux install on a clean hard drive with only linux is not that hard to use
15:22:18mirakbut when you need to mount partitions etcetera it became harder
15:22:57mirakXavierGr: what I did is windows only on one drive and linux only on the other. What you should have done is install linux with only the linux drive as master
15:23:21mirakbecause linux certainly installed grub in the masterboot record of the primary drive
15:23:43XavierGrbut I can't see any physical files there?
15:23:57mirakno it's a bootloader
15:24:06mirakit's binary code
15:24:08XavierGrand I did the opposite unfortunately Linux is on the slave disk and windows on the primary.
15:24:11mirakin the first sector
15:24:15mirakI think
15:24:34XavierGrso when I ghost another image I will loose the bootloader for sure.
15:24:53mirakdepends if your ghost is of the full disk or just one partition
15:25:25XavierGrit is one partition but it's the primary partition that windows boot.
15:25:30SlasheriXavierGr: grub doesn't use that partition boot flag at all
15:25:44SlasheriXavierGr: grub is normally installed on the hdd master boot record (mbr) sector
15:25:58Slasheriand the grub binary files are located on the /boot partition
15:26:41XavierGrSlasheri: If this sector is somehow deleted and loose the bootlaoder is there any way to configure it again externally(with a floopy or something)
15:27:04mirakso in clear what I did is this : I installed windows with only the master disk in it. then I installed linux but I removed the windows drive. What you need then is to edit the /boot/grub/menu.lst to add a windows boot entry with the line map (hd0) (hd1) and map(hd1) (hd0)
15:27:12SlasheriXavierGr: yes, you need to reinstall the bootloader to boot again
15:27:41mirakXavierGr: if you delete the first sector you lose the partition map, you lose everything
15:27:48XavierGrMaybe I will have to do the brute force method. I will load my ghost image and see for myself. :D
15:28:10mirakXavierGr: it will not break the bootsector
15:28:15Slasherihmm, partition map shouldn't be on the first sector
15:28:29XavierGrmirak the windows ghost image is pre-;inux created.
15:28:48mirakXavierGr: the problem however is that you will not be able to boot without having the linux drive because grub needs some files in /boot/
15:28:48amiconnThe partition table *is* part of the MBR
15:28:55amiconn(well, except extended partitions of course)
15:28:57XavierGrif grub wrote something on that partition it will be lost after I load the ghost image.
15:29:04Slasheriamiconn: hmm, really? interesting..
15:29:18mirakXavierGr: grub writes nothing
15:29:32XavierGrmirak: no I wont loose the Linux partition.
15:29:48Slasheriok, that probably true as i have only experienced the mbr by creating a bootable floppy using assembler (so no partition map)
15:29:53mirakXavierGr: I think you need to fix first your windows installation and run fixmbr or something like that. Then do what I said
15:30:04XavierGr(it is on a different disk so the ghost image has nothing to do with it. The ghost image will only affect the primary boot partition of the primary HD.
15:31:06XavierGrwhere windows lies, Linux is on the primary partition of the slave HD
15:32:04amiconnSlasheri: A floppy disk doesn't have a MBR
15:32:20Slasheriamiconn: yes, true. It only has the bootable sector
15:32:32mirakyes but actually you *can't* boot windows without having your slave disk connected !!!
15:33:41XavierGrha so if I remove or delete the Linux partition on the 2nd HD I will loose my windows installation?
15:33:42mirakbefore or after ghosting it will be the same
15:33:59mirakXavierGr: no you will just not be able to boot to it
15:34:07XavierGryes that
15:34:13XavierGrboot to it I mean
15:34:13mirakyou will then need to run fixmbr from the rescue cd
15:34:29XavierGrof Windows you mean?
15:34:39peturfdisk /mbr
15:34:45XavierGrah yes I remember that
15:34:59XavierGrI don't use it because I have my valuable ghost images. :D
15:35:56 Join leftright [0] (
15:36:48leftrightSlasheri, Hi there
15:37:32Slasherileftright: hi, sorry but i need to run to a bus in few minutes
15:37:54 Quit leftright (Client Quit)
15:38:23 Join leftright [0] (
15:38:41Slasherileftright: what do you wanted to say? :)
15:39:10leftrightjust wanted to chat about artifacts which I hear during track transitions
15:39:19leftrightbut we can chat later
15:39:32XavierGrah yes known problem :)
15:39:37Slasheriah, lets check that later today :)
15:39:44leftrightdont miss your bus :)
15:39:55Slasherihehe, i go there now ;) ->
15:47:34***Saving seen data "./dancer.seen"
15:50:26Mongey|anyone know any programs to replace m$ publisher
15:50:51leftrightopen office ?
15:51:30Mongey|doent do publisher
15:52:31leftrighterr, google :->
16:22:13 Join perplexity [0] (
16:25:53perplexitylostlogic you here?
16:29:42lostlogicxperplexity: yeah
16:30:37perplexitycan you send me, or put a patch up for the retarded things gcc is doing with the coldfire code?
16:31:03perplexityI'm off for a week as of tomorrow night and I'm planning on spending some time seeing about improving the coldfire generator in gcc and submitting a patch
16:31:55lostlogicxperplexity: /msg me your email and after this meeting I'll send you some info about the stupid things I've seen.
16:32:37perplexityno hurry.. I'll be checking it while I'm away in any case
16:38:52apgappetur: good work with with your prerecording patch! have you tried to change the range to 3 minutes (should be possible) ?
16:40:18peturapgap: I'll look into this ;)
16:40:40peturMy patch isn't available yet
16:40:54 Quit Zagor ("Client exiting")
16:41:12peturI rewrote parts of it last night and this morning, one small bug left.
16:41:40petur3 minutes will be a bit much, don't know how much memory the H1xx has?
16:42:10Bagderthe h100 has 16
16:42:11petursame as H3xx then...
16:42:53Bagderstill, you should probably consider a system to adjust to what the target has
16:43:11Bagderwe might see >32mb targets in the future
16:43:16peturah yes, more work :)
16:44:10 Quit mirak (Success)
16:44:10peturI'll probably release a working version first, then look into increasing the time
16:44:16Bagdersounds wise
16:44:35peturotherwise it 'll never get ready ;)
16:44:55Bagderright, adding a basic function and then expanding from there is a good strategy
16:45:13Bagdergetting it into CVS also allows others to join in and help easier
16:45:26solexx_ mentions WavPak *en*coding support for Iriver
16:45:31solexx_where is it?
16:47:43BgerBagder are there any objections in making new option "Remote type" for irivers ?
16:48:00Bgertill we find out how to detect it
16:48:08perplexityplease please please :)
16:48:10 Join mirak [0] (
16:48:22BagderI of course prefer to auto-detect it, but until then I think its fine to have an option for it
16:48:27petursolexx_: H1xx and H3xx recording only writes WAV files atm...
16:48:30Bgerperplexity there is a patch about this on the tracker
16:48:37perplexityGuys... I think I figured out how to detect it this morning..
16:48:41 Join San [0] (
16:48:42Bgerbut i'm merging it and mine ...
16:49:01Bgerperplexity how ???
16:49:13solexx_petur: that's what I thought, too
16:49:27perplexitygive me a couple of tics to hit the source code and get back to you Bger :)
16:50:06 Quit Mongey| ()
16:50:11 Join akaidiot [0] (
16:50:49Bgerheh this will make last hours of reading settings.c and settings_menu.c nearly meaningless :)
16:50:59perplexityOk.. how do we detect the LCD remotes ??
16:51:06Bgerat least i know how to add settings in the future
16:51:33Bgerperplexity : yes, please ? ;)
16:51:59perplexitythe NON-LCD remote simply uses the same play button gpio as all the others.. but does nothing else other than feed values to the ADC_REMOTE.. I thought the LCD remotes tripped a remote detect pin.. checking now
16:53:30perplexityGPIO_READ & 0x40000000 to detect the presence of a normal LCD remote ?
16:53:47 Join webguest89 [0] (
16:54:15webguest89petur: regarding prerecordig - h110 has 16mb ram
16:54:31webguest89oh, Bagder said that
16:54:31Bgerperplexity yes, it does this
16:54:35 Quit webguest89 (Client Quit)
16:54:49perplexityOk.. with the NON-lcd remote.. that bit is not set..
16:54:58Bgerare you sure ?
16:55:07perplexityso if you see ACD values != 0xFF and that bit is not set, you have a non-lcd remote
16:55:17Bgeroki, good :)
16:55:37Bgerand between 2 LCD remotes ?
16:55:55perplexityOn that I have no idea.. I only have the NON-LCD remote here
16:55:57Bgerbetween THE
16:56:13 Part Sando
16:56:18 Join Sando [0] (
16:56:19perplexityI posted the ADC values for that to the thread on the patch tracker this morning, and also to here
16:56:30Bgerah, yes
16:56:54Bgerperplexity could you add them to ?
16:57:01Bgerat the end of that page ...
16:57:05perplexityUmm.. yep.. sure
16:57:23Bger10x in advance
16:59:33Bgerlast time i tryed my NON-lcd remote rockbox didn't read anything except the play/pause/on button
16:59:46Bgerthat is, adc_read(ADC_REMOTE)
17:00:53perplexityyes.. me also.. but using the patch on the tracker.. I switched to H300-lcd remote, then to h300-non-lcd remote and the ADC starts reading.. dunno why it does not normally.. have not investigated that far yet
17:02:02 Part Sando
17:02:09Bgerwhat, what ?
17:02:22Bgerhm, strange
17:05:10 Join San||Away [0] (
17:05:39San||Awayserver -m
17:06:48amiconn3 minutes prerecording is impossible even with 32MB ram
17:07:24 Quit San (Read error: 110 (Connection timed out))
17:08:16perplexitychanges done Bger
17:10:05perplexityNow.. I have no idea why the ADC_REMOTE does not read until I change those remote types.. must find that little bug
17:15:07 Join muesli- [0] (n=muesli_t@
17:19:15 Join Mongey| [0] (
17:20:53perplexityDamn.. it *does* trigger the remote detect bit :(.. how could I have missed that
17:21:13mirakis there something special to do to write in a file when using the write method ? I end up having nothing in the file I created with open
17:23:17Bgerperplexity no worries :)
17:23:36preglowwhat a life
17:23:53apgapamiconn: 32000000 / (44100 * 2 (stereo) * 2 ( 16 bit)) ~= 3 minutes
17:23:56perplexityI wonder if the 300 LCD and non-LCD remotes use the same ADC values.. and the 100 remote uses different ones..
17:24:00Mongey|Bger; when can you get that remote patch?
17:24:10BgerMongey| it's not ready yet
17:24:10perplexitywhich would be why iriver states the h100 remote won't work on the h300
17:24:14apgapamiconn: but we need some for codec and plugs and recording marginal
17:24:28 Quit San||Away (Read error: 110 (Connection timed out))
17:24:35Bgerperplexity they have never said "it won't work"
17:24:44apgapapgap: but 2.5 should be possible
17:24:46Bgerthey said "we don't promise it will work"
17:25:17 Quit muesli- ("ich will Kühe!!!")
17:25:26perplexityyeah.. And I see from the wiki that the adc values are different in any case
17:26:02 Quit Mongey| (Read error: 104 (Connection reset by peer))
17:28:24 Part Polo_o
17:29:17perplexityI can't see where remote_type gets loaded on boot with the patch on the tracker.. that may be why I have to toggle the settings to get it to work
17:30:56 Join San [0] (
17:31:19peturapgap: there is also a bit of that buffer that you can't use as recording needs to continue while you're writing to disk
17:33:01lostlogicxugh, some change I made has my player freezing.
17:33:33peturapgap: currently that limit is 2MB
17:33:40perplexityI see it now Bger ... the debug screen uses adc_read which only displays the buffered results.. if the remote is not operational, selected or held.. then the adc_scan for the remote is not triggered, so the values never update :)
17:34:14perplexitywith the default code.. the h300 thinks the non-lcd remote is on hold all the time and it never reads the adc hardware
17:35:13perplexitywith the patch on the patch tracker for remote selection.. it never sets the remote type when the settings are read from disk.. only sets it when you use the menu.. so the remote will never work until you manually select it in the menu..
17:36:36 Join JazzBone [0] (
17:38:17perplexityTook a long time for my low watt ideas bulb to light.. but it got there
17:38:32Bgerhehe mine is lower :P
17:38:44preglowlostlogicx: you still interested in that asm snippet?
17:38:55Bgerok, i'm trying the new remote patch :)
17:38:58lostlogicxpreglow: which?
17:39:01preglowlostlogicx: on the ml
17:39:27lostlogicxpreglow: as my last post says, I found a faster solution without messing with the MAC which didn't introduce artifacts.
17:39:44preglow'cause you're using the mac wrongly anyway :)
17:39:54lostlogicxpreglow: oh? do tell!
17:40:32preglowwell, we use the emac unit in fractional mode, which means the accumulator move instructions return the top 32 bits of the multiplications
17:40:41preglownot the bottom 32 bits, which is usual
17:41:04preglowso you've got to switch accumulator modes before you try something like that
17:41:06preglowbut yeah
17:41:12preglowi don't think that solution is very fast anyway
17:41:24lostlogicx*nod* I'll keep that in mind if I find another place to take advantage of it.
17:41:58preglowbut yeah, using just good old add, sub and muls there is probably more than good enoughj
17:42:05preglowthe muls instruction isn't so slow on our coldfire
17:42:11preglowsince that actually uses the emac unit
17:43:03lostlogicxyep, and I use 2 less registers, and (much more importantly) don't use a jbsr to divide compared to GCC.
17:43:24perplexityoops.. plugging usb in while in the display IO debug menu.. *boom*
17:44:03 Quit YouCeyE (Remote closed the connection)
17:44:09lostlogicxpreglow: right now I'm just frustrated by this freezing when I try to play any ogg thing. Started randomly last night, and I feel like I've tried reverting every change I made last night to no avail.
17:44:50preglowahh, that's fun
17:44:56preglowthat's where i'm at with musepack right now
17:45:15preglowgot a really nice speedup going, but something makes it crash
17:45:15lostlogicx*nod* most frustrating thing in programming to me.
17:45:30preglowand refuse to play low bitrate files...
17:47:38***Saving seen data "./dancer.seen"
17:47:55perplexitythis is where a decent sim would be great.. or serial port and gdb stub :)
17:48:12peturor ICE
17:48:26perplexityw00t h300 non-lcd remote with hold working :)
17:50:44Bgerok, i don't understand something
17:52:25Bgerhm, maybe my light bulb is getting hotter
17:52:42perplexityI think mine just burned out.. time for a scotch
17:54:32Bgerwhere should i set firmware layer variable, which is stored on the disk?
17:55:20perplexityin the patch on the tracker it is a simple setting.. works great here
17:55:24apgapbah, the ipodlinux project is so boring without a CVS checkin log.. hard to follow what is happening there
17:56:10Bgerperplexity ?
17:56:18perplexityyes Bger ?
17:56:43Bgerhm, i'm looking at that patch atm
17:57:08perplexitywant me to send you the complete patch I'm using here for your perusal?
17:58:06perplexitygive me a tic to generate it
18:01:37peturgot to go, bbl
18:01:48 Quit petur ("CGI:IRC 0.5.7 (2005/06/19)")
18:02:37 Quit [IDC]Dragon ("CGI:IRC")
18:05:06 Quit apgap ()
18:06:42 Quit dropandho ()
18:10:36 Join Sando [0] (
18:12:54 Join RotAtoR [0] (
18:13:23Bgerare all functions like static bool contrast(void) (apps/settings_menu.c) called when loading settings ?
18:14:38Bgerok, i'm dumb
18:17:12perplexity@@ -874,6 +875,7 @@
18:17:12perplexity lcd_remote_scroll_speed(global_settings.scroll_speed);
18:17:12perplexity lcd_remote_scroll_step(global_settings.scroll_step);
18:17:12DBUGEnqueued KICK perplexity
18:17:12perplexity lcd_remote_scroll_delay(global_settings.scroll_delay * (HZ/10));
18:17:12perplexity+ set_remote_type(global_settings.remote_type);
18:17:13***Alert Mode level 1
18:17:13perplexity #ifdef HAVE_REMOTE_LCD_TICKING
18:17:16perplexity lcd_remote_emireduce(global_settings.remote_reduce_ticking);
18:17:17perplexity #endif
18:17:19amiconnBger: App layer code usually checks global_settings.whatever itself. Firmware layer code usually has a function interface to set parameters, so any setting of that kind needs to be added to settings_apply()
18:17:27perplexityThat hunk in settings.c is what is needed to reload the settings on boot
18:18:11lostlogicxpreglow: well I got it playing... garbled, but playing, which is easier to deal with than freezing.
18:18:17perplexityI have a straaaange problem though.. I removed all key definitions for the remote volume..but it still works.. gotta find that ghost somewhere :)
18:18:45 Join Moos [0] (
18:19:29Bgeramiconn : just realised your last sentence :)
18:20:01CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
18:20:01*amiconn has got an idea for vkeyboard and remote support, and also for morse display on archos
18:20:25Bgerhm, after a little i would like someone with h100 remote to test things ...
18:21:54amiconnThe problem is that we have small and large LCDs. On H1x0 and H3x0 the whole vkeyboard fits on the main display. For the remote we would need several pages the same way as for archos (currently having 3 pages)
18:22:53amiconnNow, if we don't use several pages but rather have all characters on one virtual page, which is scrolled as the cursor moves around so that the cursor always stays visible...
18:23:18 Join muesli- [0] (n=muesli_t@
18:23:30amiconnThe same could be done for the morse input display, which is actually just a help screen. Scroll list up& down
18:26:13amiconnThe list could even be reformatted dynamically, allowing to use the UI font
18:27:14***Alert Mode OFF
18:29:24*Bger is looking for test "mice" :)
18:29:40Bgeramiconn good idea
18:30:48amiconnNow if only TiMiD were around :/
18:33:21preglowwhere's he gone anyway?
18:33:24preglowhaven't seen him for yonks
18:33:35amiconnThe vkeyboard strings could be made a part of the language file, that would allow to have chinese, japanese, hebrew...
18:33:43Bgerreally, someone here who wants to test the last remote patch ? i'm searching for people with h100 remotes (both h1x0 and h3x0 users)
18:33:59 Join Lear [0] (
18:34:00amiconnBasic ascii should be built-in
18:34:33Bgeramiconn : do you have remote ?
18:34:48amiconnPerhaps it would make more sense to couple the vkeyboard to the default codepage setting
18:34:50Bgerpreglow you ?
18:34:52amiconnBger: yes
18:35:04amiconn(almost never use it)
18:35:07Bgerthen do you have little time now ? :)
18:36:33preglowi don't have my remote with me
18:36:37preglowi never use it
18:36:38Bgeri want to see 1) does it work (tm); and 2) whether the new button mappings are OK
18:36:53Bgerah i understand
18:37:17Bgerbtw, i'm wearing my h300 on my belt, so i'm using the remote nearly all the time
18:37:43perplexitythis is just perverse.. this button logic is going to blow my brains apart :\
18:37:59Bgerperplexity :))
18:38:14Bgerperplexity you don't have LCD remote, do you ?
18:38:22amiconnperplexity: Huh?
18:38:49perplexityI'm working with the remote selection patch from the tracker and it's doing my head in
18:39:41mirakBger: wow you got style
18:39:41Bgerperplexity the patch from the tracker isn't complete ...
18:39:44mirakBger: :)
18:39:53mirakI don't trust the belt clip
18:39:54Bgermirak : what ?:)
18:40:00mirakI prefer have it in my pocket
18:40:25mirakBger: don't know I always found the cell phone on belt a bit ringard
18:40:25Bgermirak : do you have LCD remote ?
18:40:28perplexitywell it is _so_ close to almost working here.. but this nested if statement for remote button detection is just killing me.. it should be working..
18:40:28Bgerh100's one
18:40:30mirakI mean sucky
18:40:35perplexitywhat is missing from it Bger ?
18:40:39mirakBger: no, I wich I had one
18:41:02Bgerperplexity the button mappings aren't correct
18:41:07Bgerin the tree ...
18:41:07mirakhowever I have a case with a glass
18:41:10Bgerin the menu ...
18:41:30perplexityOh yes.. I'm just trying to get it working properly in the wps
18:41:45Bgerbtw, it's working correctly in the wps ...
18:41:56Bgerat least here
18:42:09Bgerbut i didn't tryed the other patch
18:42:39Bgersomeone here with h100 lcd remote ?
18:42:45perplexityHere my volume buttons are inverted.. and there is something really ugly with the adc values....
18:43:46Bgerhm, i don't see anything weird ...
18:44:10Bgeru just go from the top to the bottom starting from the least ADC reading
18:46:17 Join tvelocity [0] (
18:47:40perplexityso it would appear.. but something wierd is playing with my button values
18:48:36 Quit leftright ("CGI:IRC (EOF)")
18:50:12 Quit muesli- ("ich will Kühe!!!")
18:50:47 Join actionshrimp [0] (
18:51:53 Quit XavierGr ("Trillian (")
18:52:24 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:56:53Bgerok, updated patch on tracker (No 1367824)
18:58:14 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
18:58:53 Join linuxstb [0] (
18:59:11mirakhey how can I adapt the timers to rockbox ?
18:59:22mirakwhat's the concept general concept of time in rockbox ?
18:59:50mirakit's all based on HZ ?
18:59:55linuxstbThere is a current_tick variable which is incremented HZ (100) timers per second.
19:00:11Bgercoldfire has 2 timers
19:00:14Bgerone is used by rockbox
19:00:22Bgerand the other is available to plugins
19:01:01Bgerthe second is used also for dimming the h100's main display's backlight
19:01:11Bgerso this is disabled if the timer is in use
19:01:22linuxstbpreglow: Have you looked at the ipod audio code at all?
19:01:22Bgerthe second timer, even
19:02:11mirakso how we can't have under a second precision ?
19:03:05linuxstbYes - current_tick will give you 1/100th of a second precision. I don't know about the timers.
19:03:08amiconnHmm, there seems to be no firmware update file for IGP-100 :(
19:03:27mirakwhat method do you use to know if the optimisation is good ? You calculate over a minute for a sample ?
19:03:54mirakI mean I am wondering if it's worth keeping any reference to time in the xvid code
19:04:50preglowlinuxstb: no, got distracted yesterday and now i'm hung over
19:05:25linuxstbmirak: There are two things you want to time - 1) The average time it takes to decode a frame (i.e. the number of frames per second you can decode) and 2) The amount of time the library spends inside each function - so you know where it is worthwhile spending time optimising.
19:05:54preglowi'm going away for a little while tomorrow too
19:06:04linuxstb2) is where lostlogic's profiling code will probably be useful.
19:06:48linuxstbpreglow: I'm going away for a couple of weeks on Saturday. I'm hoping to finish the audio before then, but not sure if I'll have time.
19:06:56preglowi wont have the time, that's for sure
19:06:59preglowi'm leaving tomorrow morning
19:07:29preglowi could try doing some stuff now, but it's hard trying to use this particular brain today
19:08:20miraklinuxstb: yep, but if we don't have a precision bigger than one second, I don't see how to caculate the time spent in one function
19:08:25mirakhow do you do that ?
19:08:40preglowwe have precision bigger than one second
19:09:25preglowlinuxstb: btw, seems david bryant's getting a nano
19:09:27mirak<linuxstb> There is a current_tick variable which is incremented HZ (100) timers per second.
19:09:40mirakhe lied to me :)
19:10:01mirakhow to get more precision ?
19:10:01preglowhe says flat out that you've got precision bigger than one second right there
19:10:11preglowyou have 100 ticks per second
19:11:00mirakthat's left 4 ticks for a frame decoding
19:11:09miraksince of course we will run 25fps without a problem
19:11:47preglowyou have higher resolution timers than that too
19:11:50linuxstbpreglow: Do you know if David Bryant is interested in working on anything else apart from wavpack on his Nano?
19:12:01preglowlinuxstb: sincerely doubt it, he says he doesn't have much time as it is
19:12:33preglowbut he has picked on other stuff from time to time
19:12:33 Join NicoFR [0] (
19:12:34preglowwe'll see
19:18:01 Join Bluechip [0] (
19:21:06 Quit Kohlrabi (Read error: 104 (Connection reset by peer))
19:21:23preglowi'd really love to find out what this cache issue is all about
19:21:46preglowit's unusable without the cache
19:21:58linuxstbAnd unusable with the cache...
19:22:17preglowwell, sporadically usable
19:22:46 Quit NicoFR (
19:22:46 Quit akaidiot (
19:22:46 Quit amiconn (
19:22:46 Quit warewolf (
19:22:46 Quit elinenbe (
19:22:46 Quit mbr (
19:22:46 Quit pill (
19:22:46 Quit Nilisco (
19:22:46 Quit Ismo (
19:23:04NJoinNilisco [0] (
19:23:15 Join elinenbe [0] (
19:23:17***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
19:23:18NJoinakaidiot [0] (
19:23:25 Join pill [0] (i=pill@
19:23:26***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
19:23:29 Join amiconn [0] (n=jens@
19:27:47NJoinwarewolf [0] (
19:27:57 Join Ismo [0] (i=laitinei@
19:27:58***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
19:28:16 Join mbr [0] (n=mb@
19:28:18***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
19:34:43 Quit San (Read error: 110 (Connection timed out))
19:38:29 Join mofoGX [0] (n=mofo@
19:47:39***Saving seen data "./dancer.seen"
19:58:39 Quit tvelocity (Connection timed out)
20:01:50MoosHello Slasheri :)
20:04:48SlasheriMoos :)
20:05:29 Join NicoFR [0] (
20:05:38preglowso, how fares tag cache?
20:06:28PaulJpreglow: read the topic ;)
20:09:32preglowlucky me that i just asked how it's going, then
20:09:36Slasheripreglow: now searching works too (artists, albums and songs) :)
20:09:54Slasheriand the engine itself is even quite stable now
20:10:24Slasheristill need to add settings and test more the operation without additional ram to load the into
20:10:47 Join KN|stiff [0] (
20:11:03 Quit KN|stiff (Client Quit)
20:11:16 Join _FireFly_ [0] (
20:15:24PaulJsorry. i should learn english before posting silly comments
20:16:00 Join linuxstb_ [0] (
20:16:13mirakhem a declared fuction in a header doesn't need to really exit right ?
20:17:31_FireFly_mirak: no it doesn't need to really exist if the function is never be used
20:17:45 Quit linuxstb (Read error: 110 (Connection timed out))
20:20:18 Quit mofoGX (Read error: 110 (Connection timed out))
20:21:09 Join LinusN [0] (
20:23:14preglowx86 is such a crappy arch
20:26:54 Quit NicoFR ()
20:27:15preglowtry coding assembler for it without breaking out in tears all the time
20:28:17*_FireFly_ would like to have an cell-chip-based system to play with it a bit ;)
20:29:45preglowcell looks like a standard dsp design
20:31:19_FireFly_could be but the best is that one cell-chip has 8 cpu-cores arranged as an cell on it (one master an 7-slave)
20:31:37preglowit'll be hell programming for cell
20:31:48_FireFly_that is a really multi-core cpu and has an ppc alike arch implemented
20:31:49preglowi'm not convinced cell is a good plan
20:32:11preglowports from/to ps3 is not going to be straightforward
20:33:08_FireFly_multi-core using games are on their very start
20:34:10_FireFly_but you will need in future more cpu's to get better graphic-/physic-effects because the CPU-clock frequenz is nearly on it max
20:34:49preglowit's not just being multicore
20:35:08preglowthe cell processors are good at primarily one thing: crunching numbers
20:35:26preglowso you can't just put a generic thread onto a cell core and think it'll just blaze ahead
20:35:35preglowanywho, we'll see
20:35:35_FireFly_that's true
20:35:53preglowyou can do very much cool stuff with just crunching numbers anyway
20:37:07_FireFly_the biggest problem will be when threads are used in games is to have a good communication design between the several used threads
20:37:12mirakif a function is in a .c but never called, is it built or not into the binary ?
20:37:20_FireFly_mirak: it is build
20:37:33_FireFly_and linked into the binary afaik
20:37:47LinusNyes it is
20:38:21_FireFly_LinusN: to which topic is your answer ?? :)
20:38:22LinusNyou can use a special functionality in gcc/ld to weed out unused functions by putting all functions in a separate segment
20:38:58LinusNand let the linker remove unreferenced segments
20:39:12perplexityw00t finally a properly working non-lcd remote on the h3x0 :)
20:40:49LinusNnow we need to find a way to autodetect the remote type
20:40:52_FireFly_afaik a non-lcd one is easier then a lcd-one ;)
20:41:19 Join Gibbed [0] (
20:41:21 Quit Rick (Nick collision from services.)
20:41:23perplexityI just spent 3 hours beating my head against the patch on the tracker.. I'd debate the word "easy"
20:41:33 Nick Gibbed is now known as Rick (
20:41:38perplexitybut yes.. I still have no idea how to tell them apart :(
20:41:45 Join bger_cgiirc [0] (
20:42:15perplexityIf we can't tell them apart, I'm not sure how to integrate the hold switch function into the bootloader :(
20:42:17_FireFly_hmm the GPIOs are the same when an remote is connected ??
20:42:25bger_cgiircLinusN: if you have time, could you try how are the button mappings of h100's remote with my last patch ?
20:42:48_FireFly_bger_cgiirc: what have you changed ??
20:42:50perplexitybger_cgiirc I uploaded that patch for you also
20:43:00LinusNbger_cgiirc: i won't have much time, but i'll see what i can do
20:43:21bger_cgiircperplexity: 10x
20:43:52bger_cgiircLinusN: i've made h300 and h100 button mappings the same
20:44:04bger_cgiircuf :(
20:44:28_FireFly_i hope you didn't changed the volume and ff/rew mapping
20:44:45bger_cgiirc_FireFly_: for h100 remote ?
20:44:58bger_cgiirci changed it ...
20:45:01_FireFly_i like the current one
20:45:16bger_cgiirc_FireFly_: could you try the new one ?
20:45:34_FireFly_i could if you have the patch somewhere online
20:45:43bger_cgiircit's on the patch tracker
20:46:06perplexityta for the info lostlogicx I can't reply coz I'm not registered :)
20:46:35 Join muesli__ [0] (
20:49:42_FireFly_bger_cgiirc: now i have to switch the fingers when navigating into an sub-dir/-menu :(
20:51:36 Join petur [0] (
20:52:28 Join shebb [0] (
20:52:56 Quit bger_cgiirc ("CGI:IRC")
20:54:15shebbI am trying to compile the simulator build. What do I have to modify to get it linking properly? /bin/ld cannot find -lwinmm
20:54:59 Join bger_cgiirc [0] (
20:55:05bger_cgiirc_FireFly_: bad :(
20:56:11Learshebb: Interesting... winmm isn't always available... One way would be to not define ROCKBOX_HAS_SIMSOUND. :)
20:56:19_FireFly_i haven't it yet tested on the device this statement was made while looking into the patch
20:56:20Learshebb: using devkit or cygwin?
20:56:38bger_cgiircin fact i don't have idea how to easily solve this
20:57:24bger_cgiircthat is, runtime remote swithching and better button mapping
20:57:45_FireFly_why was the reasign needed ??
20:58:02bger_cgiirc_FireFly_: all these are defines
20:58:15shebblear: devkit
20:58:29bger_cgiirch300 lcd remote acts just like the main unit
20:58:42bger_cgiircit has joystick
20:59:07Learshebb: suspected that... :) doesn't include the needed file then.
20:59:07bger_cgiircup - vol up; left - REW; right - FF; down - vol down; joy push - navi/menu
20:59:14_FireFly_for this the new one are great i think but not for the h100-remte
20:59:33shebbWell, I can try undefining that... I do not need sound for my current tests
20:59:37bger_cgiircyes, but the problem is that they are compile-time defined ...
20:59:47_FireFly_i know
21:00:56shebblear: autoconf.h already undefines this
21:01:55LearHm... Yes, the lib is unconditionally included in the makefile, I just realized.
21:01:56bger_cgiircany better suggestions than changing switch(button) to if-s or if-ing on remote type ...
21:02:08 Join mofoGX [0] (n=mofo@
21:02:16LearSo you could comment out that line in apps/Makefile, as a quick fix.
21:02:19_FireFly_bger_cgiirc: a other one on the h100-remote i push instinctively vol-up to move the cursor down
21:02:45bger_cgiircand it moves it up ?
21:03:07_FireFly_yepp that is confusing
21:03:16_FireFly_for me
21:03:20bger_cgiirc_FireFly_: any ideas ?
21:04:50shebbdo you mean: LDOPTS += -lwinmm
21:04:51_FireFly_and another one now you need maybe both hands to use it proper .... no sorry
21:05:01shebbCommenting out that gives me just other problems
21:05:12Learshebb: yes.
21:05:57shebbundef ref to CreateCompatibleDC
21:06:25shebbin uisw32.c amongst other errors
21:06:46_FireFly_shebb: it seams there a some dev-libs missing
21:06:50LearBut the win32 sim was buildable before I added that line...
21:07:02_FireFly_shebb: on your system
21:07:14shebbthat could be the case...
21:07:35_FireFly_because CreateCompatibleDC is an "standard"-fn under windows
21:07:35shebbI may be a little behind the times when it comes to the dev kit
21:08:43shebbwell, I can try reinstalling things and just getting by debugging on my h140
21:09:05bger_cgiirchm, maybe a mapping function ....
21:09:35_FireFly_.. which returns the mapping for the selected remote
21:09:42_FireFly_as an array or
21:09:53_FireFly_something else
21:10:06bger_cgiircthis should be the way ...
21:10:09bger_cgiircbtw, _FireFly_
21:10:20_FireFly_yes ??
21:10:25bger_cgiircwhat's the situation with h100 remote and original fw ?
21:10:34bger_cgiirchow does it behave
21:11:25_FireFly_ff/rew is used to navigate up/down
21:12:06_FireFly_navi/menu enter a sub-menu/option and stop to go one level up
21:12:49_FireFly_when you are in the settings or file-tree
21:13:13bger_cgiircaha ...
21:13:40 Join San [0] (n=test@
21:14:56_FireFly_and vol up/down, rec a-b/mode and source/bitrade has no functions when in file-tree /settings
21:15:25_FireFly_ups a-b/mode is also for one level up
21:15:34bger_cgiirchm, isn't the original rockbox mapping better?
21:16:51_FireFly_compared to the iriver-fw one ?? then yes because you have an page-wise scrolling via source/bitrate
21:18:01bger_cgiircah, yes
21:19:02 Join miner49er [0] (
21:19:34miner49erCan someone answer a compiler/c question for me please?
21:19:43Sanwhat is it?
21:20:18miner49erhi Sans. Here's my question:
21:21:00miner49erin this piece of code: if( test1() && test2() )
21:21:01_FireFly_but if we use en get_remote-mapping fn then we can't use switch for the button handling
21:21:13miner49eris test2 evaluated if test1() returns false?
21:21:18_FireFly_because we can't use vars in an case-statement
21:21:29_FireFly_miner49er: no
21:21:31 Quit muesli__ (Read error: 104 (Connection reset by peer))
21:22:10_FireFly_if the first to test value in an AND statement gets false then the complete statement is false regardless of the other values
21:22:45miner49erso, the other test (if it's a function call) isn't run?
21:22:56 Quit JazzBone ("Miranda IM! Smaller, Faster, Easier.")
21:23:00miner49er...there's an easy way to find out I suppose...
21:23:38miner49erso the compiler is clever that way?
21:23:41_FireFly_yeah two functions post something to the console(when under linux or is an windows-cmd-app) or uses an message-box
21:23:55_FireFly_miner49er: yepp
21:24:12LearNot really, part of the language specification, afaik.
21:24:24_FireFly_Lear yepp
21:24:57 Join JazzBone [0] (
21:25:12miner49erFireFly: cheers for your help. I work with high-level c# at work and I know for a fact that does it but wasn't sure about c, especially GCC
21:25:16_FireFly_because why should app check the other values in an AND statement when one just tested value is false
21:25:34_FireFly_s/app/a app
21:25:41 Quit mofoGX (Read error: 104 (Connection reset by peer))
21:25:51miner49eryeah, it makes total sense, but I wonder if there may be some cases where the programmer might want both functions run
21:26:12Bagderminer49er: then you write the program differently
21:26:47_FireFly_save the return values in vars and check then the vars when the two fns should be always called
21:26:48miner49erinteresting stuff
21:27:20 Join tvelocity [0] (
21:27:21_FireFly_in this case it's logically afaik
21:27:29miner49erBadger: you'd only write it differently if you were aware of what the compiler was doing to the code!
21:27:38bger_cgiirc_FireFly_: better suggestion
21:27:44Bagderthat is how the language C works
21:27:53_FireFly_bger_cgiirc: tell it :)
21:27:55Bagderit is not compiler specific
21:28:10_FireFly_and afaik all other programming languages also
21:28:22bger_cgiircswitch on (__get_button) ; __get_button -> get_button() on archos targets etc
21:28:24_FireFly_at least all high-level-languages
21:28:28miner49ernearly finished my new plugin anyway...wish I could get the iRiver Sim to work though, works a treat on my Archos...
21:28:56_FireFly_bger_cgiirc: ??
21:29:02miner49ersorry, subject change :-)
21:29:11bger_cgiircand on iriver __get_button is our function, that makes the mapping
21:30:16_FireFly_ah __get_button is a define for get_button on archos on iriver it is a real fn
21:30:28bger_cgiircsomething like this
21:30:47bger_cgiirci'll see tomorrpw
21:31:05_FireFly_this wouldn't break the current used switch-statements
21:31:15bger_cgiircyes, exactly
21:31:16miner49eroh, could someone with cvs write access make a little change and move the hiscore saving code from Snake2 into the hiscore lib (and pass a char* parameter)? Or should I raise a change request?
21:31:39_FireFly_miner49er: make a patch and post it on tracker
21:31:41miner49er...i suppose I should figure out how to make diff files really...
21:32:24_FireFly_if you have done your changes in a rb-source-tree which you had checked-out via cvs ...
21:32:44_FireFly_then you can type simply cvs diff -u > patch-file in the root-dir of the source-tree
21:33:30miner49erFirefly: okay, I will give it a bash, cheers.
21:34:53_FireFly_miner49er: otherwise you need a seperate unmodified source-tree
21:35:10 Quit bger_cgiirc ("CGI:IRC")
21:35:59miner49erI'm using cvs, so it should be straightforward...
21:36:40_FireFly_and then you make patches with this command: diff -uR <unmodified-source-tree-dir> <modified-source-tree-dir> > patch-file
21:37:37_FireFly_miner49er: you need the second methode if you have added new files or new directorys(at least for new directorys) to your local source-dir
21:37:45 Part shebb
21:38:38_FireFly_and then the needed additional diff option-parameter is -N (new-files)
21:39:05_FireFly_i have made a simple shell-script for the second-methode :)
21:40:26miner49eroh, right, so i need a whole 2nd copy of the source tree? so first I have to ensure I'm totally in-sync, make a duplicate, add my changes then run diff command?
21:40:41_FireFly_miner49er: yepp
21:41:39_FireFly_you can check-out the rb-sources into a seperate dir which you then can also easily keep in sync with the latest cvs on the server
21:41:57miner49erFireFly: that is quite hardcore - what are the chances of f*cking things up - i.e. making a pointless diff file? how can it easily be checked before submission?
21:42:15 Join DocterD [0] (
21:42:42_FireFly_miner49er: the first check you can look into the patch file
21:42:58_FireFly_because it is a text-file
21:43:01 Quit Lear ("Chatzilla 0.9.69 [Firefox 1.5/0000000000]")
21:43:07miner49eri'll figure it out...
21:43:35_FireFly_and/or you can patch with this new created diff-file a clean source-tree
21:43:45_FireFly_and test it then
21:44:03miner49errightm yeah, that makes sense!
21:44:25miner49erjust create a fresh tree, add the patch to that.
21:45:00_FireFly_you know when you run cvs -dPC then your already checked-out local copy will be converted to a clean one
21:45:18_FireFly_if you have done changes in this tree
21:47:02_FireFly_cvs co -dPC i mean
21:47:35 Join muesli__ [0] (
21:47:42***Saving seen data "./dancer.seen"
21:49:12_FireFly_argh "cvs update -dPC" now it is the correct command ;)
21:56:48lostlogicxdwihno: nice work on that tremor patch, much cleaner than merging the two files like I did.
21:58:20BagderI think you mean lear
21:58:47lostlogicxMagnus == dwihno which is who did the commit afaics
21:58:58Bagdernope, Lear is that Magnus
21:59:06Bagderdwihno is another Magnus
21:59:17lostlogicxyikes, then yes, I meant lear :)
22:00:14peturI see current_tick being used as unsigned long, long, unsigned int, int. Should it not be unsigned long?
22:00:37BagderI believe it is signed
22:00:48Bagderand int and longs are of the same size on all our actual targets
22:01:15peturso long should be OK
22:01:27mirakwhen there is only the header of a function does the binary have some trace of it ?
22:01:37Bagdermirak: no
22:02:31_FireFly_Bagder: then is why use int when long uses the same space in ram
22:02:51Bagderbecause not all targets are the same
22:03:02Bagderhence my use of "actual" up there
22:03:04_FireFly_hmmm right i'm stupid ;)
22:03:19Bagderthe Archos Gmini uses a CPU with 16bits ints
22:04:02Bagdernot that I think we'll ever run on the Gmini, but hey...
22:04:20 Quit Rob2222 (Read error: 104 (Connection reset by peer))
22:04:20preglowdifficult platform, that
22:04:24preglowin more ways than just that
22:04:41_FireFly_but other targets in the future will have different size for int and long
22:04:56Bagder_FireFly_: possibly, yes
22:05:00 Join Rob2222 [0] (
22:05:00 Quit DocterD ("Leaving")
22:05:29peturdoing a search for current_tick doesn't show a nice result :( even kernel.c mixes int and long in sleep()
22:06:56*petur excuses himself for being so critical, has been doing software review this week ;)
22:07:11_FireFly_but current_tick is defined as long in kernel.h
22:07:20peturit is yes
22:07:50_FireFly_i think this mixup should be corrected
22:10:05mirakhum why there is not the graphical volume anymore ?
22:10:18_FireFly_mirak: ??
22:10:45mirakdon't know I only have values in db
22:10:58_FireFly_i have a graphical volume display in the statusbar on my device
22:11:11mirakI should update rock
22:11:16_FireFly_maybe you had the settings changed ??
22:11:48 Join solexx [0] (
22:12:03_FireFly_the graphical volume wasn't removed in any official-builds(daily or bleeding edge)
22:12:51peturimho a tick counter should be unsigned anyway....
22:13:06Bagderbecause 240 days is not enough?
22:13:20 Quit JazzBone ("Miranda IM! Smaller, Faster, Easier.")
22:13:33_FireFly_Bagder: i want at least 420 days ;)
22:13:46Bagderiirc, it is signed for easier comparisons/wrap cheks
22:13:59_FireFly_that is a good point
22:14:01*petur remembers Win98 crashing after 40 days because of somebody saying 'nah, 40 days is enough'
22:14:17Bagderwell, it shouldn't crash due to wrap
22:14:29Bagderbut I doubt anyone has experienced a wrap
22:14:45BagderI mean without it being forced
22:15:08_FireFly_240 days can only be reached when the device is in mains
22:16:30_FireFly_afaik no battery in this size can hold enough power to run a device for 240 days maybe only in a very deep powersaving-mode
22:18:01peturwell, audio_recorded_time() will remain unsigned long or I have to change a lot of other places that have nothing to do with my patch...
22:22:19 Quit solexx_ (Read error: 104 (Connection reset by peer))
22:24:04 Quit muesli__ (Read error: 110 (Connection timed out))
22:31:10_FireFly_the build of the x11-sim is broken due the undefined settings-apply and set_remote_type fns
22:32:29peturpre-recording for iRiver is here:
22:32:44_FireFly_ups only set_remote_type fn
22:32:47peturthis patch makes patch 1384031 obsolete
22:34:25_FireFly_petur: you can delete your own patch-tracker entries when you was logged in with an account while submitting a patch
22:35:46peturme has just learned something new about sf
22:36:13*petur needs to learn to type ;)
22:37:08_FireFly_petur: set the type to deleted
22:37:22_FireFly_ehm status i mean of the patch
22:37:27peturI did, it worked...
22:47:48mirakshould I remove all post processing stuffs in xvid ?
22:49:21preglowi don't think we'll have enough cpu to use it
22:49:23preglowbut i'm not sure
22:49:51mirakbecause I am removing as much code I can to reduce the plugin
22:50:19mirakwell if one day somebody want postproc he will have to reimplement it
22:50:27miraklike if he want encoding :)
22:50:36mirakin several years
22:50:58mirakwell anyway in several years we will just have to run linux
22:51:16mirakif we can record in mpeg4 directly from a dap cam whatever
22:52:13mirakI reduced from 950 to 750 ko
22:55:44 Quit Rob- (
22:56:15NJoinRob- [0] (
23:04:32_FireFly_now i have another patch pending ;)
23:06:24mirakextern VLC sprite_trajectory_code[32768];
23:06:32mirakthat's an array ...
23:07:00_FireFly_yepp an 32Kb big array if VLC is a one byte var
23:07:23miraknope it's 0x40000 !
23:07:34mirakit's an array of a struct
23:07:38mirakI must remove that
23:07:53_FireFly_the 32768* sizeof the struct
23:09:38mirakcool it can be removed
23:09:38miner49erIs camelCase really frowned upon by the rockboxers? I'll need some serious search and replace if it is!
23:09:47mirakincluding the function that use it
23:24:18 Quit _FireFly_ (
23:24:24preglowvariables are lower case
23:24:28preglowit's a rockbox rule
23:24:31NJoin_FireFly_ [0] (
23:24:39_FireFly_preglow: ??
23:24:48mirak480036 bytes for the plugin
23:24:51mirakit gets smaller
23:27:54 Part Sando
23:29:20 Quit tvelocity ("Leaving")
23:32:36 Join nathanh [0] (
23:33:29miner49erpreglow: What about structure names?
23:34:20 Join Sando [0] (
23:34:37miner49erhow do you distinguish between functions and variables? or_is_it_all_underscore_spaced?
23:35:28_FireFly_miner49er: due the fact that no typedefs are allowd in the rb-sources you have to type struct <structname> var
23:35:55_FireFly_or what do you mean ??
23:38:15miner49erFireFly: okay, what if I have a structure called apple and I want lots of apples, what would the naming convention be? I would choose Apple for the struct name and apple for the 'instance' of the structure.
23:38:45miner49ercrap example, I know.
23:38:48preglowyou don't typedef
23:38:51preglowahh, right
23:38:59preglowfirst, no mixed case in rockbox
23:38:59_FireFly_preglow: ;)
23:39:13preglowi suggest you just read some source
23:39:18miner49erhuh? So, I can't do it? I can't make an array of structures?
23:39:31preglowthe few rules we have are outlined in docs/CONTRIBUTING
23:39:46peturwhat is the place to set the brightness at startup (depending on the saved setting)?
23:39:47miner49erI don't geddit
23:40:12_FireFly_miner49er: you can do an array of an struct
23:40:22preglowwhy can't you have an array of structs?
23:40:29_FireFly_e.g. struct <structname> var[100]
23:40:32miner49erFirefly: that's okay then :-)
23:40:33preglowstruct mystruct omglol[45];
23:41:10miner49eryeah, I know how to do it, I just wanted to know what the naming conventions were
23:41:12_FireFly_typedef is only a tool to omit the struct when defining an "instance" of an struct
23:41:35_FireFly_to type struct i mean
23:41:36miner49erit's no problem placing struct in front of var name...
23:42:05miner49erwhat rockboxable product do you have FireFly?
23:42:31miner49erwhat mp3 player I mean...
23:42:35peturis apps/main.c init() the place?
23:42:56 Join Jungti1234 [0] (n=jungti12@
23:43:38miner49ercool, I reckon my plugin will be good on that
23:43:49miner49er...on the H120 that is.
23:44:12petursettings help anyone?
23:44:45_FireFly_petur: afaik in the settings_apply-fn
23:44:55miner49er...mind you, you already have gameboy working, so this is just another game and therefore boring
23:45:17_FireFly_miner49er: i don*t use the dap as an gameboy replacement ;)
23:45:35 Quit akaidiot ("( :: NoNameScript 3.81 :: )")
23:45:46petur_FireFly_: thanks, looks like the place
23:47:13miner49erFireFly: fair enough, I know it's fairly pointless, I just like the challange of getting something working.
23:47:44***Saving seen data "./dancer.seen"
23:48:15Mooswhat kind of game will be your plugin? by curiosity
23:49:34 Join Paul_The_Nerd [0] (
23:50:12_FireFly_time for bed good night everybody
23:50:24Moosgood night mate
23:50:27 Quit _FireFly_ ("Leaving")
23:51:03miner49erhow do I register my nick for private messages?
23:51:21Bagder /msg nickserv help register
23:51:58miner49ernight FireFly, cheers for you help
23:52:21miner49erBadger: What the heck does that mean? (I'm pretty new to IRC)
23:52:37 Quit San (Read error: 104 (Connection reset by peer))
23:52:42Bagdernickserv is an automated "service" that takes care of such things
23:52:42ender`try typing that
23:52:53miner49erso I just type that lot in?
23:53:08Bagder /msg is how you send a private message
23:53:12miner49eraaah, something happened!
23:57:17 Join akaidiot [0] (
23:58:34 Join Mongey| [0] (

Previous day | Next day