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 2004-06-26

00:06:04 Quit elinenbe (" HydraIRC -> <- Chicks dig it")
00:07:36***Saving seen data "./dancer.seen"
00:08:10 Quit iSheep ()
00:21:13 Join elinenbe [0] (
00:26:08 Quit AciD (Read error: 104 (Connection reset by peer))
00:53:28 Quit c0utta (Read error: 54 (Connection reset by peer))
01:02:44 Quit top_bloke (Read error: 60 (Operation timed out))
01:10:13 Part amiconn
01:36:31 Join Nibbler [0] (
01:57:58 Nick midk|movie is now known as midk (
01:59:11 Quit midk ("just STOP it arspy")
02:00:58 Join midk [0] (
02:07:40***Saving seen data "./dancer.seen"
02:22:52 Quit Nibbler (Read error: 54 (Connection reset by peer))
02:25:50 Join AciD [0] (
03:30:50 Nick midk is now known as midk|game (
04:01:05 Join Nibbler [0] (
04:07:44***Saving seen data "./dancer.seen"
04:13:52 Nick midk|game is now known as midk (
05:09:17 Nick midk is now known as midk|here-but-go (
05:09:17 Quit Nibbler (Read error: 54 (Connection reset by peer))
05:09:28 Nick midk|here-but-go is now known as midk (
06:07:46***Saving seen data "./dancer.seen"
06:12:51 Join michaelolsen [0] (
06:13:10*michaelolsen looks around
06:13:42michaelolsenhmmm i think the only nick i remember in her is logbot lol
06:13:50 Part michaelolsen
06:13:59 Join Nibbler [0] (
06:38:31 Quit Nibbler (Read error: 54 (Connection reset by peer))
07:52:55 Quit AciD (""Linux ? Ya moins bien, mais c'est plus cher"")
08:07:50***Saving seen data "./dancer.seen"
08:14:11 Join Nibbler [0] (
08:31:59 Quit Nibbler (Read error: 54 (Connection reset by peer))
09:47:04midki figgered it out
09:47:12midkwow −− i have a free button on my clock!
09:47:26midk*is usually buttonless and attempts to map functions to key combos
09:47:35midkf2... very accessible
09:47:38midkhow to utilize it.
09:50:11midkah crap
09:51:04BCfound a luverly bug in in gcc yesterday
09:51:09midkthat is?
09:51:31midki hate having to compress text to fit it on the lcd
09:51:36midk*hard pressed for lcd space
09:51:41BCit cannot handle bools in a struct
09:51:43midk rb->lcd_puts(0, 3, "[UP] Poweroff ON");
09:51:43midk rb->lcd_puts(0, 4, "[DWN]Poweroff OFF");
09:51:53midkhopefully not applying to rockbox.
09:52:00BCwtf is "poweroff on"
09:52:04midki know!
09:52:10midki cant convey it with like 15chars
09:52:15midkidle poweroff activated
09:52:37midkok i'll have to do a sort of
09:52:43BChow about AUTO-OFF: Yes/No
09:53:15midk rb->lcd_puts(0, 3, "[UP] AUTO-OFF on");
09:53:15midk rb->lcd_puts(0, 4, "[DWN]AUTO-OFF off");
09:53:46BChow about a tik box?
09:54:10BC[x] off-timer
09:57:01midkno room
09:57:23BCroom for "off" no room for "X"
09:57:32midkyah that's about it
09:57:49BCwell dude good luck, think laterally
09:57:57midkjust btw.. this is the help screen
09:58:04midki meant.. i have no room at the options screen :)
10:04:57midkbc did you see my creditroll code?
10:05:12BCno spam it ove
10:05:58midk*bites nails
10:06:37midkdcc or aim?
10:07:27midk credits.txt
10:07:44midklogbot, you'll yell at me for it
10:07:53***Saving seen data "./dancer.seen"
10:07:55BCdcc or aim it ove then I can just dsave and compile
10:08:04midkit's integrated into the clock
10:08:13midkit doesn't work on its own
10:08:26BCi'll wait until I can see it running ;)
10:08:33midkit's in the patches page, bc.
10:08:47midki am seriously afraid to touch it...
10:08:54midklast time it totally screwed it up
10:09:23midkbut it's at that link if you'd like to see
10:11:38BCrb->sleep(HZ/2) sleeps for 1/2 or .5 of a second
10:12:17BCif (btn != BUTTON_NONE && !(btn & BUTTON_REL))
10:12:22BCwatch your braces dude
10:12:28midkwait what
10:12:33midki messed up a comment
10:13:10midksee... some of this hurts my brain to comprehend.. so i just say.. if it works, it's fine
10:13:19midkif you have a suggestion, please tell me! :)
10:13:52BCwhat is the value of BUTTON_NONE ??
10:14:46BCif ( (!btn) && (!(btn&BUTTON_REL)) )
10:14:58midkif that will work, i shall use it.
10:15:10BCif your info is correct, it will work
10:15:22midk*saves old code just in case
10:15:42DBUGEnqueued KICK BC
10:15:42BC if ( (!(btn==BUTTON_NONE)) && (!(btn&BUTTON_REL)) )
10:15:58midkif you slip me a twenty now i can bump you up a bit in the credits list
10:16:39BCimho: either go alphabetical, or most help
10:16:44 Join Nibbler [0] (
10:17:19BCfalls a poor second to amiconn
10:17:20midk*checks to make sure bc's name is near top of list before going on
10:17:45BCyou could do reversa alphabet - that would please Zagor
10:18:13midkno, i think i will do it in order of most help
10:18:42midk*wonders if dstar5 would kill him if he put him last
10:23:03BCdeath comes to us all in the end
10:23:31midkOKIES YAYYYY
10:23:36midk*puts dstar5 last
10:27:42BCdeath appraoches
10:28:19midk*replaces dstar5 at top of list
10:28:40midk*places arsepie at the end of the list
10:28:44midkend end
10:28:48midkafter the blank spaces and black holes
10:29:28BCyou wanna ut a copper bar or something over the top of the credits :)
10:30:09midkyou mean, just under your and my names bc
10:30:13midkwe belong together
10:30:17midkwell not that way
10:30:25BCoh evil fate how wicked thou art
10:30:40midkyou don't like me?
10:30:45midk*puts away leather and whips set
10:31:14BCEver checked up to see what your mates buy off ebay?
10:31:38midkdo i want to know?
10:31:40BCmy mate buught a cat-o-9 ...geez we gave him grief down the pub for that ;)
10:31:56midkhow much??!
10:33:02BCdont recall
10:33:55midklet's get this bc
10:34:04midki'll split the cost, but i get to use it wednesdays and sundays
10:35:13midkwil you get it for me bc
10:37:32 Join amiconn [0] (
10:39:16amiconnhi bc!
10:39:41midkhey amiconn. hi bcbc.
10:40:15amiconnBC: Did you get your function callback to work?
10:40:42midktime for me to sleep
10:40:49amiconnnite midk
10:40:51BCyes, thanks dude, would you believe my two (or so) hours of insanity were down to a MISSING SEMI COLON - ARGH
10:40:52midkso good nite bcbc and amiconn.
10:40:56BCnite mk
10:41:00midkGOOD WORK BC
10:41:15amiconnBC: lol
10:41:52 Nick midk is now known as midk|sleeeeeepz (
10:42:01BCgotta update my cvs and double check the grey scale mod
10:42:09amiconnBC: I've seen you have a problem with bools in structs?
10:42:32BCyes, nasty stuff
10:42:45BCwas okay before I split it into a lib
10:42:52amiconnI wonder why is that, since both Jörg's alpine_cdc plugin and a test plugin of mine use that (for thread status)
10:43:13BCarray of struct with bool in a lib the issue
10:43:47BCthe array indexer counts the bool as 1 sometimes and 4 at others
10:44:36BCmaybe I will strip it back to a demo once I have finished the real work - we might be able to trace it in a simple example
10:45:48amiconnIirc bool == char in sh1 gcc, so 1 byte is correct. However, the compiler sometimes has to pad fields to get proper alignment.
10:46:05BCthis was all in sim, not seen a target yet
10:46:52BCin sh-1 a bool is 1byte
10:47:06BCwhen I save my .vol options on the pc a bool is 4
10:47:28BCI think the problem is most likely in stdbool.h
10:47:51BCi bet bool is enummed
10:49:11amiconnBool may perfectly have different sizes on different architectures.
10:49:39amiconnCode should not rely on specific lengths of datatypes
10:49:47BCindeed, my issue is that it has different sizes between lib modules both under simulator conditions
10:50:37BCone proc accesses the data one way, another proc accesses the data another way
10:50:47BCseen it before with gcc
10:51:42BCit seems that (internally) sizeof(struct) does not return the sizeof(all-entries-in-struct)
10:52:36BCby made up example: struct { bool b;} s; sizeof(s)=4 sizeof(s.b)=1
10:53:11BCso once you put it in an array all hell breaks loose
10:53:39amiconnImho this is correct behaviour - the compiler pads to longword.
10:54:21BCso when you do b1=s[1].b you would expect to receive elements 4,5,6 AND 7 into your bool yes?
10:54:59BCthen you would also experience the problem I had
10:55:10BCcos that is what it does
10:55:34amiconnIn your example, if you do struct { bool b;} s[10]; the array would be 40 bytes in size.
10:55:49BCthe size of the array is irrelevant
10:56:12BCthe relevant point is that SOME procedures believe the struct to be one size based a calculation of a bool being 4 bytes...
10:56:12amiconns[0].b should give back byte 0, s[1].b byte 4 etc.
10:56:30BCand OTHER procedures will believe the struct to be calculated on a bool size of 1
10:56:39BCdo you want to see it in action?
10:57:23BCit's all okay until you have an array of srtuct
10:57:35amiconnThis may happen if you use different struct packing settings for compilation of the two procedures
10:58:15amiconnIt did work before you broke up your lib?
10:58:34BCbefore it was no-lib-at-all
10:58:35amiconnAll source code was in one file back then?
10:58:43BCklondike runs fine
10:58:51BCbreak into lib... engine busted
10:59:01BCyes, all in one file
10:59:23BCcuppa T ...brb
10:59:52amiconnNow you have a plugin under /apps/plugins and a number of library files in /apps/plugins/lib and it doesn't work anymore?
11:20:13BCthat's correct
11:20:13 Quit Nibbler (Read error: 54 (Connection reset by peer))
11:20:31BC(sorry about the delay, found ants - bwwwaaahahahahahahahaaaaaaa)
11:21:26amiconnThen I suspect the Makefiles in /apps/plugins and /apps/plugins/lib call gcc with different options, at least for simulator builds
11:22:24BCi declare a static array in main() ...then pass the pointer to lib_init()
11:23:25amiconnI guess you have a header file defining your struct?
11:28:23amiconnThe makefile for the plugins does include a hell lot more -m and -D for gcc than that for the lib..
11:29:18BCthe libs don't actually compile under sim anyway
11:29:40BCbut as we are the only lib authors and yours don't do sim anyway, it only effects me - lol
11:31:18amiconnIf the libs don't even compile - how do you experience your problem?
11:31:30BCloads of bodges to make it work
11:32:45BCthat's probably why I need an extra bodge to get bools working :)
11:36:45amiconnThe compiler flags are very different for compiling the lib and the plugins for the sim...
11:37:53BCYes, it is still quite in it's infancy, but sadly I just have to sit back and wait for others to fix the core code - so in the mean time, I just work around all the bugs
11:38:01amiconnMost notably: While for the plugins it does -DWIN32 -DNOCYGWIN and -mno-cygwin (the latter even twice -lol) for the lib it doesn't do any of these
11:38:28 Join acathla_ [0] (
11:38:56BCyes, they are certainly reqquired to get things going ....did you see my email "libplugin blues"?
11:39:59amiconnI will compare the lines with their counterparts when compiling for the target and see what I can do
11:40:25acathla_does anyone have a backup of the hardware mod : Nick Robinson's Sony Remote Joystick RM-X4S interface. The link is broken
11:40:35BCsorry, no
11:41:02BCchecking the "libplugin blues" email will probably tell you most of what you need to know
11:41:16BC(sorry, that was aimed at amiconn)
11:44:16BCacathla_ ...can check the url from here if you like?
11:45:01BCami ....for the first line, I now use
11:45:01BC#if defined(SIMULATOR) && !defined(__MINGW32__)
11:48:58amiconnI now think I found the flag difference that most likely causes your problem ...
11:49:25amiconnFor the target, both plugins and lib are compiled with -O (makes sense)
11:49:55amiconnBut for the sim, the lib is still compiled with -O, while the plugins are compiled without. Argh!
11:52:22BCahhaaaaaa - righty
11:59:58amiconnI try to understand the Makefile mess - it seems while building for the target uses the individual Makefiles within the source dirs, the (win32) simulator build is entirely driven by /uisimulator/win32/Makefile ...
12:00:26amiconn..except that it also uses the target's makefile for the plugin lib...
12:01:01BCrespect ...I must be honest, makefiles still confuse me, I really should take the time to work them out one day ...not the syntax, just the real-life use of them
12:02:08amiconnI don't really know much about Makefiles either - maybe this is why I'm confused
12:03:27BCit's the way they all daisy chain that does my head in ....then at some point they sew back into "configure" which is bash script ...another point of inexpereience
12:03:43 Join LinusN [200] (
12:04:05BCMornin' cap'n
12:04:06amiconnhi LinusN
12:04:15LinusNyo d00dz
12:04:46amiconnDid you read our conversation about Makefiles, plugins, libs and simulators?
12:06:45LinusNthe simulator build is a kludge
12:06:51amiconnI don't really get how all this is supposed to work - am I right that the Makefile in /uisimulator/win32 controls the sim build almost completely, except the plugin lib
12:07:20LinusNas you said, the uisimulator Makefile does everything except for the libs
12:07:55***Saving seen data "./dancer.seen"
12:07:59amiconnShouldn't the plugin lib build then be added to it also?
12:09:41amiconnI wonder why it is done that way - wouldn't it be more straightforward to use the individual Makefiles for both target and simulator builds?
12:10:22BCI guess it was an easy starting point?
12:10:31LinusNyes, but we ddidn't want to "pollute" the target makefiles with sim stuff
12:11:12LinusNthe target build is the "main" build
12:12:25amiconnAgreed. Then the libplugin should be added to the sim Makefile, and everything related to simulator should be taken out of the libplugin Makefile, correct?
12:14:34amiconnIf I would only know how to do that...
12:16:59amiconnDifferent topic: If I build a dev environment with cygwin according to the Howto or use BC's devkit, I get sh-gcc 3.3.1
12:17:42amiconnApparently the "official" compilation still uses sh-gcc 3.0.4, which sometimes behaves differently.
12:18:12amiconn(See yesterday's discussion on the grayscale bug with 3.3.1
12:18:58amiconnWhat is your recommendation, are there other known problems with gcc 3.3.1?
12:20:10LinusNi have no idea
12:20:17LinusNi use 3.2 myself
12:20:49LinusNnot at home, it seems
12:22:40amiconnThe binaries in the daily build packages are identical to those I get when compiling with sh-elf-gcc 3.0.04 on cygwin.
12:23:07amiconnWith gcc 3.3.1 I get smaller binaries (2..3 %)
12:24:15LinusNfeel free to do the needed changes to make it compile with all gcc versions
12:27:15amiconnSo far I didn't find problems with the binaries built with sh-elf-gcc 3.3.1, other than that with grayscale I did already fix (I was relying on compiler behaviour there, which changed)
12:28:10amiconnajbrec.ajz: 187 KB -> 180 KB
12:29:46amiconnThere is a slight chance that the gdb stub may cause problems when compiled with 3.3.1, but I can't test thtis
12:31:09LinusNhow could it cause problems?
12:32:13amiconnGcc 3.0.4 used to save the pr register at the start of all functions and restore it at the end.
12:32:37amiconnGcc 3.3.1 optimizes this away if no other function is called from within this one
12:33:22amiconnIf you call a subroutine from within an asm block, gcc does not notice that
12:34:00amiconnThe gdb stub uses jsr/bsr...
12:34:39LinusNi'll try that someday
12:35:11amiconnI could compile a debug build for you, but as said I can't test it myself...
12:40:38amiconnArgh! Debug build is completely broken (at least for recorders)
12:44:30amiconnmain.c: In function `init':
12:44:30amiconnmain.c:189: error: `rb' undeclared (first use in this function)
12:44:30DBUGEnqueued KICK amiconn
12:44:30amiconnmain.c:189: error: (Each undeclared identifier is reported only once
12:44:30amiconnmain.c:189: error: for each function it appears in.)
12:44:30***Alert Mode level 1
12:44:30amiconnmake[1]: *** [/home/Administrator/rb-patched/debug-recorder/main.o] Error 1
12:44:31***Alert Mode level 2
12:44:31amiconnmake: *** [apps] Error 2
12:45:20amiconnA few warnings show up before that...
12:45:34BCwtf - main.c using the plugin interface?
12:50:40amiconnApparently it does have something to do with the DEBUGF macro...
12:54:32***Alert Mode OFF
12:56:33amiconnThe thing is that main.c includes plugin.h, which redefines that macro if PLUGIN is defined. Why is that?
12:58:58amiconnOr rather, why are the files within apps compiled with -DPLUGIN ?
12:59:59 Join Nibbler [0] (
13:06:50BCsecond q ...gotta be a mistake
13:07:10BCfirst q ...plugins will need to call-back
13:07:40BCfor a plugin to output to debug it will use rb-> prefix on the debug call
13:08:13amiconnYes, but then the files in apps should be compiled without -DPLUGIN=1
13:09:06BCi agree
13:23:14amiconnLinusN: are you reading?
13:58:28 Part amiconn
14:07:59***Saving seen data "./dancer.seen"
14:21:00 Quit Nibbler (Read error: 54 (Connection reset by peer))
14:28:49 Join c0utta [0] (
15:11:15 Part acathla_ ("Byebye")
15:33:39BClol - watched them pluggin that this morning - seems like a good deal
15:34:13BCwill remote control your cable box
15:34:38BCspare battery ...must mean it can be swapped easily
15:35:13elinenbeBC: enough battery for 24 hours of use!
15:35:25elinenbeI think it comes with 3 - 8 hour batteries
15:35:40BC2 rechargeable lithium-ion batteries for up to 24 hours of continuous use
15:35:41 Join AciD [0] (
15:35:52BCcertainly not 12hrs of video playback - lol
15:36:27BCCF slot has to be the winner
15:36:39BCnow to find £400
15:37:43BCmight be nice to throw a copy of that link at the ml
15:58:34 Join Nibbler [0] (
16:08:00***Saving seen data "./dancer.seen"
17:12:47 Quit Nibbler (Read error: 54 (Connection reset by peer))
17:13:41 Join lImbus [0] (
17:14:06lImbusHi All
17:14:13BChey hey limbo-man
17:17:34lImbusBC, do you know if there is space within the status bar left ?
17:17:52BClet me look...
17:18:20lImbusI see battery, plug sign, SPACE, volume, Stop/Pause/Play-Symbol, Repeat, Shuffle, SPACE, Time
17:18:34lImbusdunno what's supposed to be shown on the SPACES
17:18:41BCthere seem to be gaps a) before time and b) after battery
17:19:02BC"plug" yes forgot that one
17:19:27BCi agree with your finding
17:19:28lImbusyup. the problem is following: I try to teach my mates (the ones who are NOT developers) to not turn off the unit before the hdd has spun down
17:20:04BCspinning disc icon approaching?
17:20:17lImbuswhould be need, eh ?
17:20:22lImbusneat, of course
17:20:37BCi think ON+F1 or whatever to rtigger an orderly shut-dwon
17:21:07lImbusit seems hard for them. even the trick to set disk spindown and baclight timeout BOTH to 10 seconds (so backlight shown disk spinning in 90 % of case) does not fit to the mind.
17:21:09BCdid you notice 3.02 went up?
17:21:39BCi never bother to wait in honesty
17:21:52lImbusthey just don't think about it. so a sign of more warning character would show hem how to treat.
17:22:15BCif the backlight wont help them ,then an icon will make little odds
17:22:42lImbusamiconn showed me a pdf yesterday (see log for url) that explains pretty well how the emergency parking of the hdd-heads is done, and how much it decreases your hdd-life
17:23:26lImbusthe backlight is not a solution because they want it longer than 10 seconds. and more than 10 seconds of hard disc does not make sense and consumes too much power
17:23:32BCi know, i'm just lazy - maybe I will write an ordely shut-down for PLAY=2S or ON+f1
17:24:19BChow about off-off-off
17:24:47lImbusoff off off by the meaning of 3 seconds-off ?
17:24:55lImbuspressing off tree times does not work
17:25:03BCno, tripple click == orderly shutdown
17:25:19BCmy v1 (like linus') also shuts down in about 1S
17:25:21lImbusit's too often the way you WANT to go somewhere else, eh ?
17:25:55lImbusahhm no, ok.
17:25:57lImbusmy mistake
17:26:02BCI know it does not work, but if you are to write an orderly shut-dwon you can tie it to whatever you want
17:26:14lImbusI thought one could go one level up in dir browsing or menues by off
17:26:47BCON+F1 is currently unused, but it's not very one handed
17:27:44lImbusoff off off would be as well as on2 secs.
17:28:22lImbusI just see in the code there is the keylock symbol somewhere else
17:29:06lImbuson-f1 is the id3-tag info while in WPS
17:29:30BCON+F1 for 2S
17:29:40lImbusthe keylock is just the space between shuffle and timie
17:30:11BCit really doesn't matter about the key, my suggestion is importantly that an orderly shut-dwon should exist
17:30:58lImbusfor me, as I am too impatient and for the noobs that do not want to bother.
17:31:12lImbustechnical newbies I mean
17:31:24BCtoo impatient for what?
17:31:51lImbusfor waiting 10 secs, then switching off.
17:31:59lImbusif in a hurry, it kills me
17:32:11BCwell pick a keystroke and go 4 it :)
17:32:57lImbusthat would be fine.
17:33:13lImbusI would not even need the icon anymore
17:45:17BCwould still be cute though
17:46:31lImbusalready hacking on it.
17:58:47lImbusBC: there is not enough space in the statusbar
17:58:51BCI reckon
17:59:00lImbus2 pixels between plug sign and volume.
17:59:16lImbusI thought it was more, then connected the dc after seeing the code
18:00:05BCspinning icq flower next to the clock would be my choice
18:05:55BChow about an animation around/behind the clock
18:08:02***Saving seen data "./dancer.seen"
18:08:45lImbusjup, that is my next intend
18:09:01lImbusjust flipping background, it that works
18:57:01 Join Nibbler [0] (
19:21:24 Nick midk|sleeeeeepz is now known as midk (
19:21:24 Quit Nibbler (Read error: 54 (Connection reset by peer))
19:22:39midkhi bcbc
19:24:45 Quit BC (Read error: 54 (Connection reset by peer))
19:38:40lImbusto BC who will certainly be reading the log as soon as he notices his connection had been reset by that silly boy called peer:
19:39:18lImbussorry, I was for pizza. Now I will be off. try to bug me on monday. EMail is not a good idea as the network is beeing reconfigured these days.
19:39:45 Quit lImbus (" HydraIRC -> <- IRC has never been so cool")
20:03:46 Join pfavr [0] (
20:08:04***Saving seen data "./dancer.seen"
20:41:25 Quit pfavr ("ChatZilla 0.9.61 [Mozilla rv:1.7/20040624]")
20:57:13 Join Nibbler [0] (
21:26:44 Join telliott [0] (
21:26:45 Quit Nibbler (Read error: 54 (Connection reset by peer))
21:41:51telliottI wish this show was syndicated across the country so I could listen without being tied to my computer.
21:44:38telliottsorry. Wrong window.
22:01:30 Join dstar5 [0] (
22:08:05***Saving seen data "./dancer.seen"
22:32:26 Part telliott
23:03:10 Join Nibbler [0] (
23:42:16 Quit dstar5 ("Leaving")
23:42:16 Quit Nibbler (Read error: 54 (Connection reset by peer))

Previous day | Next day