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 2005-05-07

00:02:48*LinusN forgot to log off irc when he left
00:03:12amiconnhi LinusN
00:03:43[IDC]DragonHi there
00:03:53Tanghi Linus
00:04:54[IDC]DragonLinusN, do you mind -ffunction-sections -fdata-sections for th iriver bootloader?
00:07:01*rasher updates ircstats
00:07:06*[IDC]Dragon is sleepy + says goodnight
00:07:22 Quit [IDC]Dragon ()
00:08:10 Join dwihno [0] (~dw@
00:17:09amiconnLinusN: Did you read about my findings of unused functions?
00:17:58amiconnI did a compile with splitting functions and data into separate sections, and then linked with garbage collection
00:18:16amiconnFor recorder v1, there are 17 unused functions and 2 unused variables
00:18:31amiconn(and I found a bug while doing that)
00:19:11LinusNthe ata_device variable?
00:19:39amiconnThe garbage-collected version crashed on 'View HW info' because of that
00:20:02amiconnCPUAdrEr from trying to access the unaligned char as an int
00:21:04oddholy crap:
00:29:17***Saving seen data "./dancer.seen"
00:31:48LinusNgotta get some sleep
00:31:50LinusNnite all
00:32:06MoosCamarogood night
00:32:10 Part LinusN
00:36:21Tangbye all
00:36:52 Quit Tang ("Chatzilla 0.9.68a [Firefox 1.0.3/20050414]")
01:11:48*HCl isn't from usa so can't help, unfortunately.
01:22:58 Quit lodesi_ (Read error: 110 (Connection timed out))
01:25:50 Part MoosCamaro
01:26:30 Join XShocK [0] (
01:39:02 Join einhirn_ [0] (
01:45:49 Join Howman [0] (
01:45:49 Quit XShocK (Connection reset by peer)
01:45:55 Nick Howman is now known as XShocK (
01:47:26 Quit einhirn (Read error: 145 (Connection timed out))
01:49:24 Quit XShocK (Read error: 104 (Connection reset by peer))
01:50:11 Join XShocK [0] (
02:03:13 Join DMJC [0] (
02:29:19***Saving seen data "./dancer.seen"
02:29:49*Bagder does Makefile cleanups
02:30:03HClany news XShocK ?
02:30:10*HCl should work on a gui for the searchengine
02:30:17HCli'll prolly finish it tomorrow.
02:46:21*amiconn found more unused code...
02:47:58*Bagder scratches head
02:49:39amiconnThere's an ugly clash as well - int debug_fm_detection gets defined twice
02:49:58amiconnOnly the linker garbage collector will find that
02:50:32amiconnThere's a problem - one definition is in firmware/tuner_samsung.c, the other in apps/recorder/radio.c
02:50:48amiconnBest solution is to completely remove it, imho
02:51:22Bagdersince the firmware one is in a lib, the apps one would override it on link, won't it?
02:51:41amiconnHmm, then the debug menu option doesn't make much sense any more
02:52:07amiconnThe problem is that this is an apps-firmware cross dependency
02:52:30HCldoes anyone know how i can make vim autoindent code with 4 spaces rather than 8 ?
02:52:37amiconnif I remove it from radio.c, then compiling for a target with a philips tuner only would fail
02:52:59amiconnHCl: Sorry, no vim here
02:53:06Bagderno vim here either
02:54:46amiconnBagder: I think you're right about the override on link, but when I link with −−gc-sections, linking fails
02:54:51rasherHCl: hang on..
02:55:09Bagderamiconn: yes, and it is still not very nice
02:55:18Bagderunless intended
02:55:36Bagderbut it it was intended, it should've been mentioned
02:55:53amiconnYup. The question is whether the fm debug item should stay
02:56:05amiconnFM is not yet working on iriver...
02:56:08Bagderno opinion from me really
02:56:58HClamiconn: australianperson was working on that recently.
02:57:08rasherset ts=4
02:57:09rasherset expandtab
02:57:12rasheris what I have
02:57:23amiconnHCl: I know... austriancoder btw
02:59:04HClrasher: odd, i have that, i think...
02:59:09amiconnThe point is that the debug menu item displays 2 things. (1) whether fm hardware was detected and (2) a value which is meant for debugging (1), but only has a meaning for the samsung tuner chip
02:59:47HClah well..
02:59:48amiconnI will remove (2), but don't know whether I should leave (1) in place
03:00:10HCli think austriancoder had 1 working on iriver
03:00:23amiconn..and (2) displayed as 0, yes
03:00:51rasherHCl: hrm, I also have (on another computer) sts=4
03:00:58rashernot sure what that does :-\
03:01:03amiconn(that's correct, as it has no meaning for the philips tuner)
03:03:11HClmaybe its short for set..?
03:03:16HCli'll try to add that.
03:03:51HCldoesn't work..
03:03:53rasherno, set sts=4
03:03:54HClsomething with tags
03:04:18HClnope :/ still 8...
03:04:23HClwhenever i do a { and press enter
03:04:26HClit'll add 8 spaces
03:04:28HClrather than 4...
03:04:46rasherah yes..
03:04:48rasherbugger, that
03:04:57rasherI think that's mandated by the syntax file or something
03:04:58*HCl asks in vim
03:05:04HCli mean
03:07:03HCl03:02 < strull> HCl: :he 'sw
03:07:29Bagderand people say emacs is strange ;-)
03:08:02HClBagder: my thoughts exactly
03:08:03HCl :P
03:08:08HClits set sw=4
03:08:10*HCl goes to try
03:08:27rasherhe might as well have said ":help shiftwidth"
03:08:33rasherlazy bugger
03:08:47HCl'shiftwidth' 'sw' number (default 8)
03:08:47HCl local to buffer
03:08:47HCl Number of spaces to use for each step of (auto)indent. Used for
03:08:47HCl |'cindent'|, |>>|, |<<|, etc.
03:08:50HClyay, much better.
03:09:12rasheryes indeed
03:09:42Bagdercrt0.o is not found in librockbox.a
03:09:51rasheroh, I actually had shiftwidth in one of my .vimrcs
03:09:55Bagderit is simply picked from the file with that name
03:10:18Bagdereven thought it _is_ in the lib
03:11:13BagderI'm making the build put the objects in a dir hierarchy
03:11:35rasherHCl: "filetype indent on" is also nice
03:13:51Bagder...and it even seems to work
03:15:37XShocKHCl: i am sorry. wasn't here
03:15:47XShocKi Actually didn't do anything at all
03:16:08rasherBagder: digest is weirded out again btw :-\
03:16:21rasheror rather, hasn't been built.. did you remove it from the automaticness?
03:24:10 Quit XShocK (Read error: 104 (Connection reset by peer))
03:30:13HClfonts are drawn downwards from the xy you put in putsxy, right?
03:34:27Bagderyes, afair
03:37:38HClyup, my word wrapping code works fine.
03:37:57*HCl will continue with it tomorrow to create a very crude search query editor
03:38:32rashermore word wrapping code o.O
03:38:41rasherHCl: did you look at the viewer plugin?
03:38:56rasherThere's wrodwrapping as well
03:39:06HCli need an implementation that can tell me exactly where a word is located on the screen
03:39:17HCli sort of assumed there wouldn't be an implementation that allows that yet.
03:39:28rasherah, probably correct
03:39:29 Join XShocK [0] (
03:45:06amiconnMeh, I see red :(
03:45:56HClit wasn't meh
03:46:02amiconnThe config variables are really a bit messy :(
03:46:03*HCl rolls over and falls asleep
03:46:11amiconnHCl: No, it was me :(
03:46:46amiconnWth is HAVE_CHARGING undefined for the simulators?
03:46:56 Quit XShocK (Read error: 104 (Connection reset by peer))
03:46:57 Join XShocK [0] (
04:05:08amiconnThe changelogs in the cvs compile status table are somewhat off...
04:05:38 Join QT_ [0] (as@area51.users.madwifi)
04:10:18 Quit QT (Read error: 148 (No route to host))
04:16:42 Join StrathAFK [0] (
04:22:03 Quit XShocK (Read error: 104 (Connection reset by peer))
04:22:25 Quit DMJC (Read error: 110 (Connection timed out))
04:22:35 Join XShocK [0] (
04:23:45 Join DMJC [0] (
04:29:23***Saving seen data "./dancer.seen"
04:32:22 Quit Strath (Read error: 110 (Connection timed out))
04:40:33 Join ashridah [0] (
04:47:25 Quit cYmen_ (Read error: 113 (No route to host))
04:58:49 Quit ashridah ("Leaving")
04:59:01 Join kaltek [0] (
04:59:34kaltekcan i get some help with my jukebox studio 20?
05:00:53kalteki have a nforce board and my jukebox isnt reconized
05:18:27 Quit kaltek ("CGI:IRC (EOF)")
05:45:46 Join asdsd____ [0] (
05:45:49 Part asdsd____
05:53:18 Quit XShocK (Read error: 104 (Connection reset by peer))
06:29:25***Saving seen data "./dancer.seen"
06:44:56 Join ashridah [0] (
06:46:07 Join Strath [0] (
07:00:15 Join HeavyHelmet [0] (
07:01:33HeavyHelmetanyone in?
07:01:34 Quit HeavyHelmet (Client Quit)
07:01:36 Join HeavyHelmet [0] (
07:02:22 Quit HeavyHelmet (Client Quit)
07:05:14 Quit StrathAFK (Read error: 110 (Connection timed out))
07:10:08 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
07:38:29 Join StrathAFK [0] (
07:43:20 Quit StrathAFK (Success)
07:44:08 Join StrathAFK [0] (
07:45:15 Quit Strath (Nick collision from services.)
07:45:18 Nick StrathAFK is now known as Strath (
08:29:26***Saving seen data "./dancer.seen"
09:00:36 Quit Seed (Nick collision from services.)
09:00:43 Join Seed [0] (
10:29:27***Saving seen data "./dancer.seen"
10:59:39 Join amiconn_ [0] (
11:00:23 Quit amiconn (Nick collision from services.)
11:00:24 Nick amiconn_ is now known as amiconn (
12:29:31***Saving seen data "./dancer.seen"
12:40:44 Join MoosCamaro [0] (
12:40:55MoosCamaroHey all
13:09:07 Join StrathAFK [0] (
13:23:43 Join austriancoder [0] (~austrianc@
13:25:51 Join Chamois [0] (
13:26:15Chamoisaustriancoder : news about fm radio and remote control ?
13:28:17 Quit austriancoder (Client Quit)
13:33:27ashridahyou scared him!
13:34:37Chamoisi am sorry
13:35:25 Quit Strath (Read error: 110 (Connection timed out))
14:29:33***Saving seen data "./dancer.seen"
14:32:42dwihnoAnyone know what's causing the "cooking" effect of mp3s?
14:33:03dwihnoor whatever that decook utilities do
14:43:45 Join TCK- [0] (
14:43:45 Quit TCK (Read error: 104 (Connection reset by peer))
14:59:21 Join Aditya [0] (
15:03:38AdityaSo I found this like 2 days ago.. now I am all crazy about learning about firmware :P
15:07:58Adityaany of ou guys know some good books for low level stuff?
15:22:58ashridaha good deal of rockbox isn't really all that low level.
15:23:22ashridahit's often just a matter of sticking to the minimal c library implementation rockbox provides
15:24:09DMJCheh pity there's no magical set workign to 1
15:24:19SlasheriHmm, I am still wondering why a sleep call in debug_menu.c (the part that plays the sample.wav in while loop) crashes rockbox kernel but yeild will not.
15:24:42ashridahDMJC: haha. a 'do_what_i_mean() function? :)
15:24:56DMJCthink windows registry key
15:25:13DMJCcrashes: value=0
15:25:43ashridahDMJC: that would imply that rockbox was in any way finished, just disabled.
15:25:46DMJCor in rockbox's case... just playback my damn files = 1
15:26:36ashridahof course, unless you'd like to take over from where Linus left off, learn coldfire assembly, and start working on the complicated task of buffer management and watermarking...
15:27:00DMJCI'm working on the trident dri 3d gfx driver for linux/X
15:27:16DMJCI think that'll keep me busy enough for a while
15:29:08Adityathe 2 mins I am away, people start talking
15:29:26Adityaso rockbox has some sort of "api" already?
15:29:32DMJCif I work on rockbox, at best I get some features going
15:29:41DMJCif I get my laptop's videocard doing 3d..
15:29:47DMJCwell.. hehehe
15:30:49AdityaI am just waiting for my higher level engineering courses
15:31:02AdityaI am pretty good on the software side, just need to get my hardware side up to speed
15:31:27 Part amiconn
15:32:01DMJCthey made me learn java at uni
15:32:10DMJCno C
15:32:18DMJCof all the useless courses..
15:32:24ashridahAditya: yes. it's got a full api, including the ability to load modules at runtime.
15:32:31DMJCwould've been better off enrolling in engineering
15:32:45ashridahit's fairly extensive, it just doesn't have much in the way of a cohesive playback system, since the original platforms did hardware decoding
15:33:37ashridahDMJC: engineering tends to focus on c++ and java, i find. computer science tends towards C (plus some other object oriented language, usually)
15:33:50DMJCcomputing science did java here
15:34:00DMJCand that's why UNISA sucks
15:34:03Adityacause yeah.. I was reading some of the source files.. and the all had random calls to functions like backlit_display_do_something()
15:34:05DMJCsame with adelaide
15:34:08ashridahit started off with java here, but adds C the next year.
15:34:18AdityaCS is not about language
15:34:58AdityaCS at my school starts with JAVA to get total newbies used to programming.. then they go to C, then assembly, pascal and w/e else there is. Some courses actually use multiple languages
15:35:14 Join TCK [0] (
15:35:16DMJCwhich school?
15:35:20Adityabut I am a Comp Engineering major.. so looking forward to the engineering part
15:35:31Adityauniversity of maryland,college park
15:35:46DMJCfriend of mine went to stanford...
15:35:52DMJCI mean berkeley
15:35:54ashridahgood lord. places still teach pascal?!
15:35:58DMJChe's at stanford doing PHd
15:36:09Adityayou'd be surprised ash
15:36:17DMJCberkeley had such a large language choice I was shocked when I saw it
15:36:29Adityabut basically, its supposed to teach you how to think in different environments
15:36:34DMJCfrom pascal and cobol, to C/C++/C#/.net
15:36:39DMJCit's pretty crazy
15:36:40ashridahAditya: i agree, that's a good idea.
15:37:00AdityaI dont think our school does C#.. but thats JAVA basically anyways..
15:37:16ashridahi'm just surprised that anyone sees pascal as relevent, except as an exercise in either a) maintaing old code or b) reading old code for conversion.
15:37:17AdityaI am waiting for like digital logic design or microprocessors (!!!)
15:37:21ashridahi probably shouldn't be that surprised
15:37:30DMJC1.01ghz ppc g4
15:37:35ashridahi just used a paper on binary tree rebalancing which had code in pascal that i had to convert to C
15:37:36DMJChmm.. how did that clock increase..
15:37:39Adityabrb.. switching back to PC
15:37:51 Quit Aditya ("Download Gaim:")
15:39:26 Join Aditya [0] (
15:40:23Adityabah.. stupid gaim.. wont connect to freenode on PC.. for some reason..
15:41:23 Quit TCK- (Read error: 110 (Connection timed out))
15:43:50 Quit thegeek (Read error: 131 (Connection reset by peer))
15:44:46 Quit Aditya ("CGI:IRC")
15:47:03 Nick QT_ is now known as QT (as@area51.users.madwifi)
15:48:16 Join courtc [0] (
15:50:17 Join thegeek [0] (
15:51:43 Join Aditya [0] (
15:52:02Adityastupid mcafee and its port blocking
15:52:32*DMJC returns to stabbing osx tiger in the face
15:52:55DMJCtrying to run it in pearpc
15:53:06DMJChas some... issues
15:53:07AdityaI don't like OSX in general
15:53:14Adityabut I won't go there
15:53:23DMJCtheir CLI tools suck
15:53:35DMJCand their gui leaves some to be desired
15:53:49Adityawell if you ask me.. there are many things in OSX that are there just to be different.. which annoys me..
15:54:12DMJCthe apple menu is as old as apple itself
15:54:23Adityapeople have this stupid idea that using other's interface ideas is somehow bad.. yet people always talk about open source or w/e..
15:54:59Adityaxp's start menu is the best I have seen.. I find it easier to use that .. even more so than the desktop
15:55:07DMJCI dunno
15:55:13DMJCgnome 2.10 has a damn nice menu system
15:55:32DMJCthe places menu has made nautilus a lot more appealing for me
15:55:47DMJCI'm starting to use it when I would normally use CLI for similar tasks
15:56:32Adityaadobe and macromedia need to make a linux version of photoshop and flash
15:56:36Adityathen I will move over to linux
15:56:44Adityabut right now.. I am missing those two
15:56:50Adityaon linux that is
15:56:55DMJCwell... that's adobe now
15:57:01DMJCthey bought out macromedia
15:57:13DMJCmacromedia said that they were porting flash
15:57:23DMJCand adobe has updated reader 7
15:57:26ashridahAditya: whoever made XP's start menu goofed. it should never have had a way to cover itself with a submenu.
15:57:31DMJCit no longer sucks, has proper desktop
15:58:11Bagderadobe reader 7 on linux even uses libcurl! ;-)
15:58:19Bagderauthored by yours truly
15:58:35DMJCheh cool
15:59:14Adityaash, what do you mean by submenu?
15:59:22Adityadont the submenus expand to the right?
15:59:34Adityaoh you mean the all programs thing?
15:59:43DMJCif you like winxp so much
16:00:06DMJC kind of disturbing
16:00:27Adityathats very cool
16:00:35Adityaits not disturbing..
16:00:42DMJCstill waiting for a proper gtk version of notepad
16:00:51Adityabecause people can now easily convince family to use linux
16:00:51DMJCand a wordpad port damnit
16:01:11Adityacause I know I wouldn't be able to do it for a thousand years if I used gnome and asked my bro or dad to start using linux
16:01:30DMJCyou could skin gnome to look very close to winxp
16:01:47Adityayeah. but I would have to fiddle with all the menus and such
16:01:50Adityathis is much easier :P
16:02:08DMJCa full explorer.exe remake on linux would be kind of interesting
16:02:33AdityaI aminly like the alt-tabbing they managed to reproduce
16:02:38DMJCbut that's what reactos is for
16:03:05DMJCwhat linux is to unix
16:03:08DMJCreactos is to windows
16:04:10Adityahah cool
16:04:15ashridahexcept reactos is supposed to be binary-compatible.
16:04:18ashridahwhich linux is not
16:04:18AdityaI prob wouldn't use it yet though
16:06:35DMJCreactos is getting there but it's quite a long time off being useful
16:08:18 Quit ashridah ("Leaving")
16:12:08DMJCwhat the hell
16:12:17DMJCwine outperforms windows in some areas?
16:12:28DMJCscroll to near bottom
16:17:25AdityaI am thinking because of page file stuff
16:17:31Adityalinux might have a better way to do that
16:18:40DMJCheh welcome to swap partitions
16:18:55DMJCmore efficient caching of hdd
16:22:10AdityaI am going to get a gig of ram over the summer
16:22:15Adityawhich will put me at 1.7 gigs..
16:22:21Adityashouldn't need a pagefile then :P
16:22:34Bagderwindows will swap anyway
16:22:40Bagderlinux doesn't
16:23:54Bagder(at least not until you actually start using most of it)
16:24:36DMJCI'm saving for another 1gb of ddr 400
16:24:55Bagderwhat ARE you doing with all that ram?
16:24:55DMJCbattlefield 2 wants 1gb as recommended
16:25:04Bagderah games...
16:25:09DMJCcompiling code, games
16:25:22DMJCopen office is 1.5gb
16:25:22BagderI have 512KB, it is more than enough for code
16:25:39DMJCmainly games/media
16:25:43Bagder1.5gb ram when compiling?
16:25:51DMJCoh yeah... I have apple shake 3
16:25:56DMJCand maya 6.0
16:26:05DMJCmaya I can actually use..
16:29:18AdityaI can run maya on 512 mb
16:29:20Adityaworks just fine
16:29:34***Saving seen data "./dancer.seen"
16:29:35Adityaand no.. you can turn off page files
16:29:46Adityawindows wont use them until you run out of most of ur memory anyways
16:30:29DMJCcan run maya on 512 mb
16:30:34DMJCbut you don't really want to
16:30:46DMJChell the more ram maya gets the better
16:31:23DMJCram, cpus, and gfx cards are all the same... you always want faster and larger capacities of all
16:32:12Adityayeah I know
16:32:19AdityaI have 768 mb of ram right now
16:32:24Adityaand a 128 mb ATI AIW
16:32:29Adityait runs fine
16:33:34DMJCthat's roughly the same
16:33:45DMJCgeforce 4 ti 4800se 128mb, and 1gb ram
16:35:18Adityaeh.. geforce ti..
16:35:44AdityaI am obligated to fight you to the death as you are using nvidia
16:36:04DMJCmy drivers work on bsd yours don't I = win
16:36:54DMJCnice and simple how it breaks down to platforms it runs on and performance
16:38:03AdityaI use slack
16:38:06Adityaand ATI works on there
16:38:18DMJCrunning gentoo atm, but i dabbled in bsd
16:38:27DMJCfreebsd isn't too bad
16:38:39Adityano real motivation for me to try it
16:38:55DMJCyeah if your existing setup works, don't change it
16:39:30DMJCI'm one of those crazy tweaker people who's always trying new platforms and crazy emulation schemes
16:43:11RickConsider emulating the human brain
16:44:12DMJCheh too many registers..
16:44:36DMJChuman brain basically isn't possible to emulate on a pc
16:49:59 Join Sando [0] (lolsteam@
16:53:46 Quit Sando (Read error: 54 (Connection reset by peer))
17:08:02zethats like emulating a supercomputer on a c64
17:09:19Adityadamn rage against the machine has a bad ass guitarist
17:09:33zethough if you got everyone on the internet to cooperate with a distributed computing emulation
17:09:44zethen you might get somewhere
17:10:02Adityaze: the human brain has lots of relational stuff that really can't be programmed into a comp very well now
17:10:12zeAditya: true
17:12:20zereally many elements... its analog, physically self-reorganizing, ...
17:12:33zeoh, and hi.
17:13:59 Join asdsd____ [0] (
17:14:04 Part asdsd____
17:16:34 Join muesli- [0] (
17:47:18 Quit muesli- (Read error: 145 (Connection timed out))
17:56:13 Quit Chamois (" HydraIRC -> <- IRC for those that like to be different")
18:03:14 Join muesli- [0] (
18:12:00 Quit DMJC ("Leaving")
18:14:10 Join DMJC-L [0] (
18:20:24 Quit muesli- (Read error: 148 (No route to host))
18:29:38***Saving seen data "./dancer.seen"
18:34:34 Quit StrathAFK (Read error: 110 (Connection timed out))
18:39:24 Join muesli- [0] (
18:40:42 Join Sucka [0] (
18:45:19 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
19:02:52 Join cYmen [0] (
19:09:12 Quit DMJC-L ("Leaving")
19:17:46 Quit muesli- (Read error: 148 (No route to host))
19:19:49 Join Shagnar [0] (
19:25:43 Quit Shagnar ("( :: NoNameScript 3.81 :: )")
19:30:30rasherooh, lcd details for h300
19:36:35 Quit webmind (Remote closed the connection)
19:41:49 Join webmind [0] (
19:48:05thegeekhm, I can't remember if this is where I found it, what is the url to that irc-topic-image-collector?
19:51:18HClwhats a good ftp client for windows..
19:53:10HClmrf :/
19:53:19thegeekI think it's very good;)
19:53:40HClactually, my ftp server is rejecting my connection
19:53:43thegeekand not just because it has "leet" "warez" "scene" site-to-site transfer abilites
19:53:44HClthats why lftp isn't working.
19:54:37HClodd. why is it refusing my connection o.o
19:59:02 Join lodesi [0] (
20:01:20HCl >:/
20:01:25*HCl bites the isp of his parents.
20:01:35HClthey blocked ftp for certain ranges, including my university.
20:01:43HClit still works on port 2121
20:10:39courtcheh, lftp is good :) filezilla is decent for winwin.. and oss
20:17:25 Join preglow [0] (
20:29:41***Saving seen data "./dancer.seen"
21:04:57HClthere's a port of lftp for windows
21:18:31 Join muesli- [0] (
21:22:41 Quit lodesi (Read error: 104 (Connection reset by peer))
21:38:35 Join mrelwood [0] (
21:38:54mrelwoodany news on the iriver recording glitch?
21:39:09 Join ehntoo [0] (
21:42:40 Join cYmen_ [0] (
21:45:43 Quit cYmen (Read error: 104 (Connection reset by peer))
21:47:13 Quit muesli- (Read error: 145 (Connection timed out))
21:52:10 Quit mrelwood ("CGI:IRC (EOF)")
21:52:29 Quit webmind (Read error: 131 (Connection reset by peer))
21:55:26 Join webmind [0] (
22:29:45***Saving seen data "./dancer.seen"
22:55:50 Join amiconn [0] (
23:11:51Bagderanyone wanna try my upcoming build patch?
23:14:45Bagder (for the log)
23:28:42HClmerf, my invert word doesn't work...
23:28:45*HCl goes to bed..
23:32:49 Join [IDC]Dragon [0] (
23:33:57[IDC]Dragonhi out there!
23:34:05Bagdergood evening
23:36:06[IDC]Dragonamiconn, do you read?
23:36:37amiconnAh, hi
23:36:52*amiconn was busylooking at code...
23:37:12[IDC]DragonI just tried yesterdays sectioned compilation, doesn't work
23:37:23[IDC]Dragonthe vector table is all crap
23:37:40amiconnHuh? My sectioned builds of rockbox work just fine
23:37:58[IDC]DragonI didn't try real rockbox, but bootbox
23:38:08amiconnI made a comprehensive table of unused functions in the various builds (SH platform only)
23:38:44[IDC]Dragondo I need the .comment part in the lds file?
23:38:44amiconnI already found this time setting stuff, which saved almost 2 KB by disabling it
23:38:57amiconnI didn't need that
23:39:02HCltime setting stuff?
23:39:04 Quit courtc (Read error: 110 (Connection timed out))
23:39:21 Join courtc [0] (
23:39:47amiconnHCl: A screen for setting the current time, and a helper function to voice the time. That was dead code on platforms with no rtc (Ondio)
23:40:03HClwhen i start implementing the runtime db, i'll need to add some code to rockbox to give some sense of passage of time
23:40:08HCleven on systems without rtc
23:40:44HClbut yea, setting current time isn't relevant on units without rtc
23:42:25[IDC]Dragonamiconn: does your .lds contain any KEEP directive, other than KEEP(*(.vectors)) ?
23:42:36amiconn[IDC]Dragon: Here's the table (openoffice format):">
23:42:50*HCl is tired. yup.
23:42:53[IDC]DragonI'm more interested in you mods
23:42:55*HCl read .exe at first :/
23:43:02[IDC]Dragonto makefile and .lds
23:43:10amiconnNo, there's only KEEP(*(.vectors))
23:43:33[IDC]Dragonno other entry point?
23:44:00amiconnI did the sectioned compilation as a quick hack, adding the compiler flags to the configure-generated Makefile
23:44:20amiconnEXTRA_DEFINES=-ffunction-sections -fdata-sections
23:44:27[IDC]Dragonsame here, but I changed the script, too
23:44:41[IDC]Dragonyes, got that
23:44:50amiconnThen I changed the .lds files for the section name wildcards
23:45:02amiconn( and
23:45:16[IDC]DragonI only needed
23:46:26amiconn*(.text*), *(.rodata*), *(.data*), *(.bss*)
23:46:40amiconn...and the KEEP(*(.vectors))
23:46:43[IDC]Dragonyes, same here
23:47:15[IDC]Dragonthe vector table I get is like a conter
23:47:34amiconnFinal thing was to add -Wl,−−gc-sections to the 2 places where -elfs are created in apps/Makefile
23:48:24amiconnThis is not a complete transition; when I run the build up to the end I get 150-MB-plugins
23:48:52amiconn...because and the corresponding Makefile would need similar treatment
23:49:19[IDC]DragonI build no plugins
23:49:44amiconnThat was all that was required to get this to work...
23:49:54[IDC]Dragonso one place for the linker was ok for me
23:50:09[IDC]Dragonlooks all the same here
23:50:23amiconnI'm currently fiddling with another problem. I try to finally clean up the i2c driver
23:50:43[IDC]Dragonto do what?
23:50:57amiconn(you probably know, the PB5 handling that has nothing to do with i2c)
23:51:15 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
23:51:21*[IDC]Dragon forgot, or never knew
23:51:22amiconnI moved the PB5 handling to mas_reset() for the player, as it is mas_enable there
23:51:46amiconnFor the v1 recorder PB5 is charger control, so I added this to power_init()
23:52:02amiconnHowever, I'm not sure where to put it for fm/v2/Ondio
23:52:17amiconnPB5 is "hold power" there
23:53:16amiconnIt might suffice to simply don't care, as either the archos firmware or the bootloader should already have initialised it
23:53:51[IDC]Dragonsound like the power code isa good place
23:54:11[IDC]Dragonsounds ... is a
23:55:11[IDC]DragonI'd prefer not to rely on previous code
23:55:47[IDC]Dragonbtw, I think we should enable charging by default
23:56:00amiconnI don't agree
23:57:08[IDC]Dragonfor flat battery problems
23:57:14amiconnIf we do, we have no way to get a usable battery state
23:57:41[IDC]DragonI was about to ask that to Jerry
23:58:07[IDC]Dragonthe boxes with hardware charging also are facing this

Previous day | Next day