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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2008-09-01

00:00:16Casainhoanother question - anyone knows a free svn online hostng projects? like have svn or cvs
00:03:23Casainhoso, anyone knows why I am getting that errors?
00:06:47bertrik_dionoea, you wrote md5sum right? what do you think about the patch in FS #9209?
00:08:19dionoeabertrik_: looks good. (In fact it might be wise to not compile that plugin in default builds ... kind of like the test plugins)
00:09:19*bluebrother spots a typedef struct in this header :/
00:09:58gevaertsCasainho: again, where is MATRIX_MCFG2 defined?
00:10:38pixelmadionoea: you also were the one who introduced the "live moving" in the playlist viewer, right?
00:11:03Casainhogevaerts: AT91_REG MATRIX_MCFG2; // Master Configuration Register 2 (hperiphs)
00:11:07bluebrotherCasainho: is that the exact casing of the error you get?
00:11:14Casainhogevaerts: line 129 of that file
00:11:35bluebrotherthe line you pasted has AT91_REG, while the error you posted at91_reg
00:12:36Casainhowell, I get errors on each line of that file :-(
00:12:42*bluebrother considers this header file ugly as hell
00:13:02bluebrotherah −− so you simply picked one randomly?
00:13:14Casainhoexport/AT91SAM9260.h:1041: Error: bad instruction `at91_reg AIC_FFDR'
00:13:16Casainhoexport/AT91SAM9260.h:1042: Error: bad instruction `at91_reg AIC_FFSR'
00:13:18Casainhoexport/AT91SAM9260.h:1043: Error: junk at end of line, first unrecognized character is `}'
00:13:19Casainhoexport/AT91SAM9260.h:1085: Error: bad instruction `typedef struct _AT91S_PIO{'
00:13:20Casainhoexport/AT91SAM9260.h:1086: Error: bad instruction `at91_reg PIO_PER'
00:13:22Casainhoexport/AT91SAM9260.h:1087: Error: bad instruction `at91_reg PIO_PDR'
00:13:24Casainhoexport/AT91SAM9260.h:1088: Error: bad instruction `at91_reg PIO_PSR'
00:13:26Casainhoexport/AT91SAM9260.h:1089: Error: bad instruction `at91_reg Reserved0[1]'
00:13:27gevaertsCasainho: please don't do that
00:13:28Casainhoexport/AT91SAM9260.h:1090: Error: bad instruction `at91_reg PIO_OER'
00:13:29Casainhoexport/AT91SAM9260.h:1091: Error: bad instruction `at91_reg PIO_ODR'
00:13:31Casainhoexport/AT91SAM9260.h:1092: Error: bad instruction `at91_reg PIO_OSR'
00:13:32kugelstop it
00:13:39Casainhoexport/AT91SAM9260.h:1093: Error: bad instruction `at91_reg Reserved1[1]'
00:13:40Casainhosorry - should be pastbin
00:13:45pixelmayou've been told once
00:13:46kugelhell yes
00:14:42Casainhosorry - forgot
00:14:51Casainhothe erros are here on pastbin:
00:15:43bluebrotherCasainho: IMO these struct stuff is a bad idea anyway −− you should take the datasheet and create the register defines yourself IMO
00:16:15 Quit n1s ()
00:17:17linuxstbCasainho: Look at the other files (e.g. something like mcf5249.h) in firmware/export/ for what bluebrother means
00:17:40kugel"export/AT91SAM9260.h:2674: Error: junk at end of line, first unrecognized character is `}'"
00:18:03kugelthat's several times in the error messages
00:18:08linuxstbCasainho: Are you compiling in Windows or Linux? That's a DOS text file...
00:18:17Casainhohmmm - maybe is this: #ifndef __ASSEMBLY__
00:18:19Casainhotypedef struct _AT91S_SYS {
00:18:42kugeldos2unix is your man
00:18:51gevaertsThat's against the rockbox coding style, but I don't see how that's related to the problem
00:18:58Casainhobut I am tryingto build the crt0.S file...
00:19:13linuxstbThat would explain it then, you can't write C code in a .S file...
00:19:42dionoeapixelma: yep, that was me. Is it broken?
00:19:55bluebrotherI have the impression that it has issues with the typedeffing of foo and *foo
00:20:05bluebrotherthe same time using the same name that is
00:20:06pixelmadionoea: slightly (sometimes), see for details
00:20:17gevaertsIndeed. That's why that #ifndef __ASSEMBLY__ is there. Just define the right symbols...
00:20:30_emptypedef makes things ugly really fast
00:20:34bluebrotherok, I take that back. Overlooked the somewhat hidden P ...
00:20:58 Quit ompaul (Client Quit)
00:21:07pixelmadionoea: it's a bit weird though
00:21:09bluebrotherisn't it nice to hide information? You'll always will ask yourself what exactly this type means ...
00:21:56dionoeapixelma: i'll try and debug it tomorrow. Just remind me to do so
00:21:58gevaertsbluebrother: but surely it's intuitively clear that AT91_REG means volatile unsigned int..
00:22:05dionoeaTime to go sleep now. Good night.
00:22:14pixelmadionoea: nice, will do ;)
00:22:25bluebrothergevaerts: true. :|
00:22:57Casainhookok - I did resolved by doing "#define __ASSEMBLY__" before include that header file :-) −− thank you all :-)
00:23:06kugelCasainho: have you made sure you're using the correct line endings (LF for linux)
00:23:26*bluebrother frequently gets annoyed by those guys at work typedeffing all stuff all the time and even not knowing why doing so :/
00:23:52Casainhokugel: don't know, but that same header file builds perfectly on another project, builded with the same tools here on GNU/Linux
00:24:25kugelCasainho: Ok. But if you want to be sure, run dos2unix on that fule
00:24:34_empI can download the current build source in one file, or is subversion the only way to get the source?
00:24:36gevaertsYou should fix the line endings (i.e. set the correct svn properties), but gcc should handle it fine
00:24:49Casainhokugel: thanks :-)
00:25:02kugel_emp: there's the source archive at the current build page
00:25:11bluebrotherCasainho: a header file doesn't build. It gets included, thus the compile unit including it needs to build
00:25:43bluebrotherdepending on the compile unit this might fail, but gcc will still show the error occuring in the header
00:26:07Casainho:-) −− I am learning a lot :-)
00:26:40_empkugel, how often are those updated?
00:26:56kugel_emp: on every source code change
00:27:02gevaerts_emp: every commit
00:27:07kugelit's just as current as the current builds
00:27:23_empis there a release or stable branch that doesn't change as often?
00:27:40BigBambinot for lots of years
00:27:48gevaertsThere should be one soon
00:27:51kugelwell, rockbox 2.5, but that's heavily updated. You'll want to wait for 3.0
00:27:53bluebrotherno. But I was thinking if we should create tarballs for rbutil releases
00:27:58BigBambialthough we are in feature freeze for a release at the moment
00:28:15kugeloutdated rather
00:28:21_empit would be far easier to maintain a port, if I didn't have to worry about the source changing nightly
00:28:32bluebrotheras for rbutil, we have tags for each release
00:28:34 Quit mf0102 ("Ex-Chat")
00:28:48BigBambi_emp: So just work of a specific revision
00:29:08 Join krazykit [0] (
00:30:03_empand there is a file I can download for the release build of rbutil?
00:30:28BigBambiEach revision is tagged in svn and built from that
00:30:32bluebrotherno. At least not for the releases made
00:31:57_empideally, I would just download a release, open the archive, patch it, build it, and install it.
00:32:23 Quit bluebrother (Nick collision from services.)
00:32:28 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
00:32:31BigBambiWhat are you doing (out of interest)?
00:32:41bluebrotherbut IMO we should consider offering tarballs in the future. Doesn't even need the full sources, just strip everything not needed by rbutil
00:32:44 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
00:32:55bluebrotherwould also make it much easier for other distro packages
00:33:19_empwell, I have a rough working port of rockbox on openbsd, so now, I want to start down the road of being able to submit for inclusion in the ports tree. This requires a lot more work.
00:33:38BigBambiport in what sense?
00:34:10gevaertsBigBambi: they call packaging "porting" to make it sound more impressive :)
00:34:14krazykita "ports" package
00:34:14BigBambiah, packaging rbutil
00:34:36_empwell, I need a port to build a package
00:34:39BigBambigevaerts: indeed, packaging I understand - porting to me means something quite different :)
00:35:17*bluebrother did rpm packages some years ago, but that's much less impressive :)
00:35:32BigBambiI also got confused about the talk of doing it to rockbox, not rbutil
00:35:43bluebrotheranyway, releasing a tarball would be a good thing imo
00:35:45gevaertsHow will we be expected to handle distribution packages that may be older versions?
00:35:59_empwell, here's the checklist I am going to start going down:
00:36:17 Quit bughunter2 ("bye")
00:36:19bluebrothergevaerts: I want to make rbutil check its version automatically on startup and display a warning anyway
00:36:28_empthen, I can get some help from the openBSD community to test this some more
00:36:38gevaertsbluebrother: in non-offline mode I guess :)
00:36:53bluebrothereven in offline mode! :D
00:37:01*bluebrother spots the clock
00:37:18*gevaerts clearly hit bluebrother on the head too hard earlier
00:37:19 Join tvelocity_ [0] (
00:37:42bluebrotherdang, got to get some sleep now. Got work to do tomorrow errr ... later
00:37:58 Quit bluebrother ("sleep")
00:38:46 Nick tvelocity_ is now known as tvelocity[a] (
00:41:34CasainhoI am trying to build a port that don't have yet any display - just 2 LEDs for now - can I build RB bootloader without the need to build the display code?
00:43:32 Quit bertrik_ ("Leaving")
00:43:57gevaertsSure, why not?
00:44:40 Quit waldo ("Konversation terminated!")
00:47:21Casainhoin my targe-config file I have all comment about the display, however I am getting this error
00:47:32 Quit tvelocity[a] ("Αποχώρησε")
00:47:42CasainhoIn file included from common.c:21:
00:47:43Casainho/home/cas/Documentos/RockboxPlayer/rockbox_port/rockbox/firmware/export/lcd.h:38:5: warning: "LCD_DEPTH" is not defined
00:48:31Casainhosorry, thats not the error :-) −− is this one:
00:48:34Casainhocommon.c: In function ‘printf’:
00:48:36Casainhocommon.c:74: error: ‘LCD_HEIGHT’ undeclared (first use in this function)
00:48:54 Nick Kopfgeldjaeger is now known as Kopfi|offline (
00:50:00kugelCasainho: So? Can't you imagine what this error means?
00:50:31Casainhoyes I can, but how can I build without any thing about the display?
00:51:09kugelstop the functions from accessing the display, e.g. by putting some #if 0 ... #endif around
00:52:01Casainhookok... tought that maybe could be another easy way :-) - thanks :-)
00:52:01kugelor, rather define something like NO_DISPLAY, and put #ifndef NO_DISPLAY ... #endif around, that's slightly cleaner
00:52:25kugelwell, rockbox isn't made to run without display
00:53:03kugelyou could also create a parser, so that every access to the display will not display anything, but rather make the LEDs flash or something
00:53:41Casainhomy task is to make the leds flash using RB kernel :-)
00:53:58kugelsee, that's what I proposed
00:54:35kugelso, you actually don't want to remove the LCD code, but rather create an led api
00:54:55 Quit ender` (" The Japanese really love nature. For instance, the art of bonsai. That's where they take a little baby tree and torture it")
00:55:26Casainhokugel: well, I am a beguinner, I don't understand somethings you are saying :-)
00:55:30 Quit tvelocity (Read error: 110 (Connection timed out))
00:55:35Casainhoparser? :-)
00:55:50kugeli meant wrapper
00:56:35kugelyou make a wrapper, which redirects LCD functions to the LEDs
00:56:49kugelthat's what i would do
00:57:34CasainhoI think I understand...
00:57:38kugelbut i wonder what's the goal of this. why not using the LCD in the first place?
00:57:44Casainhobut I am in bootloader stage...
00:58:06kugelCasainho: the rockbox bootloader is a mini-rockbox, that doesn't change anything
00:58:07Casainhobecause I don't have drivers for LCD yet
00:58:47Casainhoand would like to first use the LEDs, much esiear than making LCD drivers...
00:59:41kugelwell, that you need to re-implement functions like printf()
01:00:50Casainhokugel: okok, after... now I would like to build the RB bootloader even if it's is empty
01:01:53kugelthat won't helpyou
01:02:14kugelhow are you planning to see if it works?
01:02:59Casainhoyes, I need to have a base to start...
01:03:28Casainhoa problem here: I can't find the common.c file, but I am getting an error on it...
01:03:41Casainhowhere is common.c file?
01:04:24Casainhoit is not listed on SOURCES file....
01:04:56kugelis this even a rockbox file?
01:05:18kugelI don't know where it is
01:05:43 Quit krazykit ("Connection reset by beer")
01:05:45Casainhooh, I got it −− bootloader/common.c
01:07:48 Quit tessarakt ("Client exiting")
01:08:05 Nick oofus is now known as oofus[away] (
01:08:14 Nick oofus[away] is now known as oofus (
01:08:41 Quit oofus (Read error: 104 (Connection reset by peer))
01:13:01Casainhoand now I am in the linking stage, for the first time :-) −− but gettin g this error:
01:13:44Casainhousr/local/arm9/lib/gcc/arm-elf/4.1.1/../../../../arm-elf/bin/ld: ERROR: /usr/local/arm9/lib/gcc/arm-elf/4.1.1/libgcc.a(_udivsi3.o) uses FPA instructions, whereas /home/cas/Documentos/RockboxPlayer/rockbox_port/rockbox/build/bootloader/bootloader.elf does not
01:18:51linuxstbUse the Rockbox recommended arm-elf-gcc (i.e. the one built by
01:18:58rasheramiconn: what to do about the characters with ENCODING -1?
01:19:19_empthe works great
01:19:28scorche|shCasainho: we have gone over this before...
01:20:25Casainhookok - so I will build arm-elf with :-)
01:21:27rasheramiconn: ah, seems they're all blank, at least in the cronyx fonts. No harm in discarding them then.
01:24:30 Quit super (Read error: 60 (Operation timed out))
01:25:24***Saving seen data "./dancer.seen"
01:27:23 Quit MethoS- (Remote closed the connection)
01:28:18 Quit faemir ("Leaving")
01:32:23 Quit saratoga ("CGI:IRC")
01:54:58 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
01:57:57saratogaso are we branching 3.0 off today?
01:58:30 Join safetydan [0] (n=evertond@rockbox/developer/safetydan)
01:59:13 Quit Casainho (Read error: 60 (Operation timed out))
02:07:47 Quit sbhsu (Read error: 110 (Connection timed out))
02:13:58 Quit culture (Success)
02:21:35 Quit saratoga ("CGI:IRC (EOF)")
02:28:17 Quit XavierGr ()
02:31:15 Quit dabujo ("( :: NoNameScript 4.2 :: )")
02:32:39 Quit |Marco| ("leaving")
02:35:42 Quit nplus (Remote closed the connection)
02:39:21 Join Casainho [0] (
02:39:44Casainhohello again :-)
02:40:10CasainhoI did build the RB bootloader however I am getting a bootloader bin file of 0 bytes...
02:41:16Casainhoon the bootloader C file I just get one warning: "rockboxplayerlittle.c:411: warning: function declaration isn’t a prototype" −− that the main(), why that error?
02:42:09Casainhoand the other warning at linker stage is this one: "/usr/local/arm-elf/lib/gcc/arm-elf/4.0.3/../../../../arm-elf/bin/ld: warning: cannot find entry symbol reset_handler; defaulting to 00000000"
02:42:43Casainhoreset_handler is refered on file .lds...
02:43:07Casainhodoes anyone have clues about why bootloader file is 0 bytes?
02:49:29 Quit bd_ (Read error: 104 (Connection reset by peer))
02:49:54 Join ranok [0] (
02:58:57ranokthat's rather annoying
02:59:51 Join MU{lappy} [0] (
03:00:29Casainhobye bye
03:00:32 Quit Casainho ("ChatZilla 0.9.83 [Firefox 3.0.1/2008072820]")
03:03:41 Join wpyh [0] (
03:04:05 Quit markun (Read error: 104 (Connection reset by peer))
03:04:14 Join markun [0] (
03:05:25 Quit Thundercloud__ (Remote closed the connection)
03:05:50 Join Thundercloud [0] (
03:05:51 Quit ranok ("Peace and chow")
03:15:56 Nick JdGordon|zzz is now known as JdGordon (n=jonno@rockbox/developer/JdGordon)
03:16:03*JdGordon kicks scorche|sh back
03:25:28***Saving seen data "./dancer.seen"
03:34:52kugelJdGordon: hey, how's your flu? Do you feel better?
03:35:47 Quit DerDome (Nick collision from services.)
03:35:48 Join DerDome1 [0] (
03:36:00 Nick DerDome1 is now known as DerDome (
03:39:26 Part pixelma
03:49:24 Join reacocard [0] (n=reacocar@WL-311.CINE.HMC.Edu)
03:52:01 Quit kugel ("ChatZilla 0.9.83 [Firefox 3.0.1/2008070208]")
04:03:14 Join mazling [0] (
04:04:26 Quit Thundercloud (Remote closed the connection)
04:09:40 Quit phinze (Read error: 110 (Connection timed out))
04:16:01 Quit advcomp2019 (Read error: 104 (Connection reset by peer))
04:16:15 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
04:19:43 Nick beta2k_ is now known as beta2k (
04:35:07 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
04:52:39 Part wpyh
05:00:34DEBUGReceived signal 15 (SIGTERM), terminating (snapshot: fplrun.c line 385)
05:00:34***Saving seen data "./dancer.seen"
05:10:41***Started Dancer V4.16
05:10:41***Connected to on port 6667
05:10:41***Logfile for #rockbox started
05:10:46CtcpVersion from freenode-connect!freenode@freenode/bot/connect
05:10:47***Server message 501: 'logbot :Unknown MODE flag'
05:10:47Mode"logbot :+i" by logbot
05:10:47 Join logbot [0] (n=bjst@gateway/web/cgi-irc/
05:10:47 Join perrikwp [0] (i=4aa794a0@gateway/web/ajax/
05:10:47 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
05:10:47 Join mazling [0] (
05:10:47 Join reacocard [0] (n=reacocar@WL-311.CINE.HMC.Edu)
05:10:47 Join DerDome [0] (
05:10:47 Join markun [0] (
05:10:47 Join MU{lappy} [0] (
05:10:47 Join safetydan [0] (n=evertond@rockbox/developer/safetydan)
05:10:47 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
05:10:47 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
05:10:47 Join Bger [0] (n=bager@rockbox/developer/Bger)
05:10:47 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
05:10:47 Join Zarggg [0] (
05:10:47 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
05:10:47 Join jon-kha [0] (
05:10:47 Join toffe82 [0] (
05:10:47 Join desowin [0] (n=desowin@atheme/member/desowin)
05:10:47 Join goffa [0] (n=goffa@
05:10:47 Join nobodyspecial [0] (
05:10:47 Join daurnimator [0] (
05:10:47 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
05:10:47 Join fdinel [0] (
05:10:47 Join avis [0] (n=ident@pdpc/supporter/student/avis)
05:10:47 Join rasher [50] (n=rasher@rockbox/developer/rasher)
05:10:47 Join Rob2223 [0] (
05:10:47 Join nima [0] (
05:10:47 Join Galois [0] (
05:10:47 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
05:10:47 Join BlakeJohnson86 [0] (
05:10:47 Join HellDragon [0] (n=jd@Wikipedia/HellDragon)
05:10:47 Join fyrestorm [0] (
05:10:47 Join swimmer [0] (
05:10:47 Join jfc [0] (
05:10:47 Join UncleRemus [0] (
05:10:47 Join Acksaw [0] (
05:10:47 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
05:10:47 Join Jorsik [0] (
05:10:47 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
05:10:47 Join midkay [0] (n=midkay@rockbox/developer/midkay)
05:10:47 Join Dhraakellian [0] (
05:10:47 Join _emp [0] (
05:10:47 Join lastebil [0] (
05:10:47 Join Soap [50] (n=soap@rockbox/staff/soap)
05:10:47 Join Hadaka [0] (
05:10:47 Join crashd_ [0] (
05:10:47 Join ComradeHaz [0] (i=Haz@
05:10:47 Join feisar [0] (
05:10:47 Join fxb__ [0] (
05:10:47 Join wurscht [0] (
05:10:47 Join crwl [0] (
05:10:47 Join ch4os [0] (n=ch4os@unaffiliated/ch4os/x-059673)
05:10:47 Join topher|away [0] (n=topher@epiar/founder/topher)
05:10:47 Join Slasheri [0] (
05:10:47 Join beta2k [0] (
05:10:47 Join parafin [0] (
05:10:47 Join Strogg [0] (n=jean@unaffiliated/strogg)
05:10:47 Join DaCapn [0] (
05:10:47 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
05:10:47 Join Bjoern-Erik [0] (
05:10:47 Join thegeek [0] (
05:10:47 Join neddy [0] (n=john@nat/sun/x-3546bed5da8e335f)
05:10:47 Join gromit` [0] (
05:10:47 Join Zom [0] (n=zom@reactos/tester/Zom)
05:10:47 Join GodEater [0] (i=c2cbc962@rockbox/staff/GodEater)
05:10:47 Join freqmod_qu [0] (n=quassel@2001:700:300:1800:213:d3ff:fee9:5ed0)
05:10:47 Join vort3x [0] (n=vortex@unaffiliated/dfa001)
05:10:47 Join suom1 [0] (
05:10:47 Join axionix [0] (
05:10:47 Join scorche|sh [50] (n=scorche@rockbox/administrator/scorche)
05:10:47 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
05:10:47 Join agaffney [0] (n=agaffney@gentoo/developer/
05:10:47 Join num1 [0] (n=brian@unaffiliated/num1)
05:10:47 Join DogBoy [0] (n=john@unaffiliated/dogboy)
05:10:47 Join dionoea [0] (n=dionoea@videolan/developer/dionoea)
05:10:47 Join shodanX [0] (
05:10:47 Join rvvs89 [0] (
05:10:47 Join Battousai [0] (n=bryan@gentoo/developer/battousai)
05:10:47 Join desowin_ [0] (n=desowin@atheme/member/desowin)
05:10:47 Join tarbo [0] (n=me@unaffiliated/tarbo)
05:10:47 Join pabs [0] (
05:10:47 Join Kopfi|offline [0] (
05:10:47 Join Tuplanolla [0] (n=jani@unaffiliated/tuplanolla)
05:10:47 Join courtc [0] (n=court@unaffiliated/courtc)
05:10:47 Join nicktastic [0] (n=nick@unaffiliated/nicktastic)
05:10:47 Join basti [0] (n=basti@
05:10:47 Join maddler [0] (
05:10:47 Join Bagder [241] (n=daniel@rockbox/developer/bagder)
05:10:47 Join preglow [0] (
05:10:47 Join SUSaiyan [0] (
05:10:47 Join HEx_ [0] (
05:10:47 Join joshin [0] (n=joshin@unaffiliated/joshin)
05:10:47 Join @ChanServ [0] (ChanServ@services.)
05:10:47 Join jmspeex [0] (
05:10:47 Join Llorean [0] (n=DarkkOne@rockbox/administrator/Llorean)
05:10:47 Join lostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
05:10:47 Join Zambezi [0] (
05:10:47 Join delt_away [0] (i=1000@
05:10:47 Join ze [0] (
05:10:47 Join iamben [0] (
05:10:47 Join maraz [0] (
05:10:47 Join w0rd54 [0] (
05:10:47 Join idshark [0] (
05:10:47 Join unpaidbill [0] (
05:10:47 Join FOAD [0] (
05:10:47 Join blithe [0] (
05:10:47 Join Kohlrabi [0] (n=Kohlrabi@
05:10:48***Server message 477: 'logbot #rockbox :[freenode-info] why register and identify? your IRC nick is how people know you.'
05:16:59 Join Horschti [0] (
05:17:45 Quit Horscht (Nick collision from services.)
05:18:20 Quit Horschti (Client Quit)
05:20:12 Join plus [0] (
05:20:42plusHi, I have a 5G 60GB iPod with iPL installed, and I just upgraded to the latest build of Rockbox for my iPod, but it no longer loads from the boot menu
05:21:20 Join EspeonEefi [0] (i=espeonee@STRATTON-FOUR-TWENTY-THREE.MIT.EDU)
05:24:27scorche|shplus: that is likely related to how it looks for the rockbox binary, however we do not support the ipl bootloader here
05:28:32 Join webguest80 [0] (n=18226480@gateway/web/cgi-irc/
05:30:25 Join RastaUdu [0] (n=18226480@gateway/web/cgi-irc/
05:32:39 Quit reacocard (Read error: 110 (Connection timed out))
05:33:32 Quit webguest80 ("CGI:IRC (Ping timeout)")
05:36:07RastaUdui have a question
05:36:20RastaUdui installed rockbox on my 30gb ipod...
05:36:37RastaUdui cant find a way to make fonts bigger.... is it possible?
05:37:44scorche|shyou need to change your font...this is described in the manual
05:38:09RastaUdui cannot find
05:38:51scorche|shlook harder ;)
05:38:52RastaUduthe page
05:38:56RastaUducan u help me
05:39:03 Join reacocard [0] (n=reacocar@WL-311.CINE.HMC.Edu)
05:39:59scorche|shi just did...the channel guidelines state that you should read the manual before asking for help in here also seem to be violating a few other ones...(the link to the channel guidelines are in the topic)
05:40:06advcomp2019there is a manual link on the homepage
05:40:56RastaUdui know that
05:41:11RastaUduwhat guide i follow the loading fonts?
05:41:29scorche|shRastaUdu: there is a section in the manual on that
05:42:36 Join Lambdumb [0] (
05:42:47RastaUduwhat section do i have to look for
05:42:50RastaUdui cant find it
05:43:29scorche|share you looking at the html manual or the pdf version?
05:43:46RastaUduthe online
05:43:56scorche|shthe pdf function typically has a "search" functionality, so you can usualy juts search for "font"
05:44:32scorche|shyou dont see anything in here relating to fonts?
05:44:58RastaUduloading fonts
05:45:10RastaUdubut i dont want to load fonts i want to make fonts bigger
05:45:30 Join Lambdugh [0] (
05:45:38scorche|sh20:37 < scorche|sh> you need to change your font
05:45:47scorche|shi told you earlier that you need to...
05:46:00RastaUduu jsut playing around with words
05:46:04RastaUduu not helping me at all
05:46:32DogBoyhe's suggesting that you help yourself
05:46:34scorche|shstp using "u" as said in the you not understand that by me saying that you need to change your font that you need to load a new font?
05:46:50nobodyspecialRastaUdu: all the time ou spen arguing with scorche|sh you could have searched the manual and found the answer
05:47:15advcomp2019RastaUdu, you need to choose a different font to get it larger
05:47:16RastaUduok... wel i went to the manual....and it says load fonts
05:47:27RastaUdui dont need to load fonts i need to make the fonts bigger
05:47:31RastaUduand i dont know how
05:47:44RastaUduand in the manual i cant see anything about font resizing
05:47:51RastaUduso i came here to ask how to do it
05:47:54advcomp2019RastaUdu, you need to choose a different font to get it larger
05:48:05advcomp2019that is the only way
05:48:08RastaUduand i find this asshole that tells me useless stuff
05:48:13scorche|shRastaUdu: for the third time, you need to change the font to get it the fonts pack, many fonts come in different sizes
05:48:22RastaUduok thans advcomp
05:48:33scorche|shRastaUdu: keep it civil...
05:49:06RastaUdushut up
05:49:09RastaUdui cant stand u
05:49:26scorche|shyou still are using "u"...
05:49:53RastaUduso what
05:49:55advcomp2019RastaUdu, read the guidelines please
05:50:19scorche|shRastaUdu: they are linked in the topic of this channel...they list the guidelines that one should follow while in this channel
05:50:39RastaUduthis is dumb
05:51:02scorche|shone should also follow them if one expects to receive help here..
05:51:15advcomp2019RastaUdu, irc guidelines are not dumb
05:51:24RastaUdubut scorche is
05:51:35scorche|shi am what?
05:51:47RastaUdusince i asked for help cause i dint find anything on the guide and he started Bitching
05:52:14scorche|shRastaUdu: this is your last warning...keep the topic to rockbox and follow the guidelines..
05:52:17advcomp2019he has not
05:52:41RastaUduthanx for your help scorche
05:52:56RastaUduYOU are really helpful
05:55:32nobodyspecialdoes the source build properly under freebsd?
05:55:32 Quit RastaUdu ("CGI:IRC (EOF)")
05:55:55scorche|shnobodyspecial: it should provided you are using the correct crosscompilers
05:56:23nobodyspecial*so I need to build the linux toolchain?
05:57:21scorche| should work there
05:58:05 Join wpyh [0] (
06:02:02nobodyspecialalright, Thanks
06:03:14 Quit Lambdumb (Read error: 110 (Connection timed out))
06:11:52 Quit Zarggg ()
06:20:49 Join iLoVeRocknRoll [0] (n=18226480@gateway/web/cgi-irc/
06:21:08iLoVeRocknRollhey everyone
06:21:31iLoVeRocknRollanyone here?
06:21:54advcomp2019iLoVeRocknRoll, yes
06:22:01 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
06:22:06wpyhyes, someone is here
06:22:18iLoVeRocknRolli was wondering if it is possible somehow to use the ipod while i am recharging it
06:22:44wpyhyes, it's possible
06:22:48wpyhare you using RB?
06:23:26scorche|shiLoVeRocknRoll: hold the menu button while inserting the cable...this is all described in the manual though
06:23:54iLoVeRocknRollwill it charge anyways
06:24:07iLoVeRocknRolli mean it wont screw my ipod
06:24:52iLoVeRocknRollok i jsut tried it works.. but no background light?
06:25:42scorche|sheh?...the backlight should work as normal
06:25:53iLoVeRocknRollno it doesnt work
06:26:00scorche|shif it would screw up your ipod, do you think we would allow it? ;)
06:26:05iLoVeRocknRolli mean everything works when plugged in but no light
06:26:43scorche|share you running the latest build?
06:26:55iLoVeRocknRolldownloaded like a couple min ago
06:27:36iLoVeRocknRolloh i wentr on options and i turned on
06:27:41iLoVeRocknRollidk why it was off
06:27:54scorche|shprobably because you turned it off at some point ;)
06:28:17iLoVeRocknRolli see there is a gameboy emulator too in it... i read the guide... it says i can put the games enywhere...
06:28:32iLoVeRocknRolli did put the game in the ipod but it desnt load it
06:28:38iLoVeRocknRolli mean i cant even see the file
06:28:59scorche|shthen you may have to check your file view options
06:29:15iLoVeRocknRollwhere is it located
06:29:29scorche|shin the menu somewhere...ceck the manual
06:29:41iLoVeRocknRolli am in the file view
06:29:50iLoVeRocknRollwhat option am i fixing?
06:30:31scorche|shshow files
06:31:04iLoVeRocknRolllemme try hold on
06:32:14iLoVeRocknRollnice it froze
06:33:56iLoVeRocknRollawesome lmao
06:35:33iLoVeRocknRollthis stinks
06:36:02 Quit iLoVeRocknRoll ("CGI:IRC")
06:43:09wpyhscorche: was that some kind of 1user?
06:45:09scorche|shif you are refering to "luser", then likely
06:45:23wpyhhm... ok
06:45:28*wpyh didn't know
06:48:52 Join massiveH [0] (
06:56:16 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
06:57:27 Quit Genre9mp3 (Client Quit)
07:05:47 Quit gevaerts (Nick collision from services.)
07:05:59 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
07:07:40 Join xxbatman94xx [0] (i=d1ad5e65@gateway/web/ajax/
07:08:02xxbatman94xxheyy umm i need help with making voice files for my ipod
07:08:22xxbatman94xxim using the rockbox utility
07:10:44***Saving seen data "./dancer.seen"
07:11:21 Quit xxbatman94xx (Client Quit)
07:49:14 Quit DaCapn ()
07:51:25 Part toffe82
07:56:46 Join bughunter2 [0] (n=Jelle@
08:00:15 Join nplus [0] (n=nplus@141.25.Globcom.Net)
08:00:36 Join austriancoder [0] (n=austrian@rockbox/developer/austriancoder)
08:01:35 Part safetydan
08:02:19 Quit massiveH ("Leaving")
08:04:30 Quit JdGordon ("Konversation terminated!")
08:07:20 Join JdGordon [0] (
08:19:47wpyhgevaerts: please consider this (small) patch for meizu_dfu.c
08:21:57 Nick num1 is now known as num1_ (n=brian@unaffiliated/num1)
08:25:01 Join ShockValue [0] (
08:25:32ShockValuehi all.. i found a couple threads on rockbox being implemented on the Rio Karma.. is that still in the works or a dead end?
08:26:37GodEatergood question
08:26:48GodEaterafaik there's no work in progress currently
08:27:02GodEaterbut the guy who was looking at it still mentions wanting to get back to it
08:28:42austriancoderCan somebody tell me the state of gsoc "rockbox as application"?
08:29:35scorche|shgood question ;)
08:30:16 Quit bughunter2 ("bye")
08:33:21 Join stoffel_ [0] (
08:37:19 Quit BigBambi (Read error: 113 (No route to host))
08:37:53GodEateraustriancoder: it's currently quantum - the waveform has yet to collapse
08:37:58num1_austriancoder: All I can tell you is that at the moment, rockbox is not an application
08:40:55 Join Bagderr [0] (n=daniel@gateway/web/cgi-irc/
08:44:50 Nick Bagderr is now known as B4gder (n=daniel@gateway/web/cgi-irc/
08:46:47 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:48:36B4gderis -fPIC really always right for the sim gcc options?
08:49:10LinusNi doubt it
08:49:15B4gderme too
08:49:27B4gderr18376 makes it so
08:51:28*GodEater goes to look up what -fPIC does
08:51:53 Join ender` [0] (
08:52:53*GodEater is enlightened
08:54:40JdGordonB4gder: sorry I didnt reply to your email.... somewhat forgot about it untill this morning :p
08:54:51JdGordoncan you pass me the log from -gsoc?
08:56:13B4gderI don't have a log from there
08:57:19JdGordonok, we'll have to live with some of it being repeated then :p
08:57:29JdGordonim not so sure we should fail it...
08:57:46B4gderlet's continue in #rockbox-gsoc
08:57:47*GodEater thinks this should probably go back to -gsoc ?
08:57:57*JdGordon isnt sure why... but ok
08:57:59*scorche|sh deletes his similar line
08:58:20scorche|shJdGordon: to prevent blown egos/drama, i would say
08:58:21 Join Rob2222 [0] (
08:58:24*GodEater prods at LinusN too
09:04:24austriancoderB4gder: we can use -fpic or -fPIC to get rid of the warning..
09:05:06B4gderit doesn't just silence the warning
09:05:29austriancoderit fixes the warning
09:05:47B4gderyes, but does it produce a funtional output all over?
09:05:54austriancoderjust google about "warning: creating a DT_TEXTREL in object."
09:06:02B4gderyou're missing my point
09:06:50B4gder-fPIC is a pretty big change
09:07:06B4gderand I'm doubting that it works unconditionally like that
09:07:12austriancoderis there an other solution?
09:07:38B4gderfirst keyword being "unconditonally", I don't think everyone gets that warning to start with
09:08:45B4gderdo you think we all get it?
09:09:34B4gderis it ld that warns?
09:09:50B4gderI don't see any warnings in the build table
09:10:25B4gderor perhaps, what does it take to get the warning?
09:10:32austriancoderyes... its ld
09:10:46***Saving seen data "./dancer.seen"
09:11:07B4gderso, what ld do you have on what arch?
09:11:38B4gderthe ld for the plugins/codecs and not for the main rockboxui?
09:11:50austriancoder1) remove -fPIC from configure script. 2) mkdir sim; cd sim 3) ../tools/configure 50 Sim 4) make
09:12:07austriancoderGNU ld (GNU Binutils) 2.18
09:12:18B4gderso why is there no warnigns in the build table? because it doesn't detect them?
09:12:39B4gderI don't have -fsim and I have no warnings
09:12:44B4gdernot a single one
09:12:58B4gderso, what ld do you have on what arch?
09:13:56austriancoderx86 - GNU ld (GNU Binutils) 2.18
09:14:18austriancoderand yes.. i get these warnings for plugins/codecs
09:16:12 Quit Rob2223 (Read error: 110 (Connection timed out))
09:16:36B4gdernot a single one of the build server sims get these warnings
09:17:00GodEatertalking of which - mine will be offline this week - it's been moved while the room it's in is painted
09:17:00austriancoderwhat ld version is used on them?
09:17:16amiconnNo such warnings here either. x86_64 needs -fPIC though, and has it set for quite some time now
09:17:40B4gderright, that's the thing I recalled
09:18:00amiconnAdding it for all sims is most probably wrong
09:18:17amiconnI'd expect trouble with crosscompiled sims at least
09:19:07B4gderwe don't doubt you getting these, we just question the fix
09:19:44 Join nplus_ [0] (n=nplus@141.25.Globcom.Net)
09:19:45amiconnAnd it's a warning that doesn't result in an error. That's fundamentally different from x86_64
09:19:58austriancoderfeel free to revert my commit
09:20:06B4gderamiconn: ah true
09:20:06amiconnThe latter would also throw that warning without -fPIC, but then linking actually fails
09:20:46B4gder"This means that you are including an object in the shared library which was compiled without the -fpic option."
09:20:58B4gdermakes no sense...
09:22:03B4gderand it is odd that austriancoder gets it but nobody else
09:22:15*GodEater just checked, and he also didn't get it
09:23:28austriancodergcc -v:
09:23:50B4gderhm gentoo...
09:24:26*GodEater remembers getting odd build errors occasionally on gentoo
09:25:10B4gderyeah, we've seen it before...
09:26:11 Join funman [0] (
09:27:28funmanladies? and gentlemen, please fasten your seat belt, the Sansav2 port is about to take off !
09:27:55*austriancoder does a sim compile test on an other gentoo machine
09:28:45 Quit nplus (Read error: 110 (Connection timed out))
09:29:01amiconnDid you try crosscompiling and testing a win32 sim with your change?
09:29:28B4gderfunman: rock!
09:29:40 Join t0mas [0] (n=tomas@rockbox/developer/t0mas)
09:29:43*B4gder buckles up
09:30:46*GodEater gets into the RockboxRollerCoaster and buckles up too
09:30:48funmanwe have access to 3 buttons and the keypad led on one of the models (Clip)
09:31:16austriancoderamiconn: not yet.. but I can try my luck... never did it before
09:31:18*scorche|sh actually ordered a Clip yesterday
09:31:37funmanbut that doesn't necessarily mean we are safe from bricking it
09:31:56funmanwe need some help from people with e200, because this model has a real recovery mode
09:32:48linuxstbfunman: How are you injecting code - have you improved upon the "mkamsboot" code I wrote?
09:33:32funmanlinuxstb: no we use the same technique, and this is why we need experiments with e200
09:33:48funmanI have a few ideas where we could put code, but not sure if we can really access it
09:34:16funmanonly testing will tell, and I don't want to brick another Clip :P
09:34:55 Join webguest26 [0] (n=577b4bd7@gateway/web/cgi-irc/
09:34:55B4gderfunman: but if it has a recovery mode, how do you brick it?
09:35:06funmanB4gder: the Clip hasn't
09:35:20B4gderah, I'm confusing matters...
09:35:30funmanthe m200 hasn't, the e200 has
09:35:40funmanI don't know about Fuze
09:35:48 Quit midkay ("Leaving")
09:35:49t0masgood morning :)
09:36:03B4gdert0mas: hey ho you long lost soul! ;-)
09:36:17t0mashehe.. sorry :)
09:36:31t0masreally busy job..
09:36:57t0masI do read the rockbox news ;-)
09:37:13GodEaterwhich changes once in a blue moon ;)
09:38:38t0masindeed.. takes no time at al ;)
09:40:07 Quit Lambdugh (Connection timed out)
09:41:31 Quit webguest26 ("CGI:IRC (EOF)")
09:48:02 Join scotty [0] (
09:49:10 Join midkay [0] (n=midkay@rockbox/developer/midkay)
09:52:02 Join Lambduh [0] (
09:52:08 Part scotty
09:54:10 Join scotty_007 [0] (
09:58:25rasheramiconn: See FS #9347 for font cleanup. I think it should be post-3.0 stuff to make sure we don't break anything.
09:59:25 Quit EspeonEefi (Read error: 104 (Connection reset by peer))
10:00:16 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
10:01:19pondlifeThat last commit is giving me lots of warnings when I build - "warning: -fPIC ignored for target (all code is position independent)"
10:02:04B4gderpondlife: right, we've been discussing the not so greatness of that commit this morning
10:02:10pondlifeI see
10:02:31*pondlife should finish reading logs before speaking
10:02:41B4gderregarding 3.0, I would prefer a release branch done... just about now
10:04:36rasherB4gder: I don't see anything speaking against it either
10:05:05pondlifeShould the -FPIC be reverted/fixed first?
10:05:11pondlifei.e. so no warnings at least
10:05:11*linuxstb wonders how the merging back to trunk will happen, and what should go in the release branch (e.g. rbutil/manual changes...)
10:05:22 Quit num1_ ("God does not believe in Atheists, therefore they do not exist.")
10:05:33B4gderI think most things will just get committed to trunk as usual
10:05:45B4gderand only selected release-fixes go into the branch (as well)
10:05:55JdGordonwill anything be backported to the release?
10:06:06pondlifeI'd say fixes but not features
10:06:12linuxstbAre rbutil and the manual close to release-ready? Or are we not worried about those?
10:06:27JdGordonand the new theme site?
10:06:35B4gderwell, as I see it we have no choice but to do this
10:06:36austriancoderpondlife: feel free to revert... i think i can live it all these warnings
10:06:44B4gderor to scrap the release plans (again)
10:06:47linuxstbThe theme site is outside the main rockbox svn
10:06:52B4gderwe can't just continue having a freeze
10:07:01pondlifeB4gder: Nooo, release time is now!
10:07:06B4gderI think so too
10:07:34B4gderI don't think trying to perfect things will be succesful
10:07:55pondlifeI see it as "3.0 is effectively in feature freeze forever, but trunk becomes unfrozen."
10:08:04B4gderI agree
10:08:41JdGordonthen bugger it... hows tonight for release?
10:08:42pondlifeaustriancoder: Do you get many sim warnings without -fPIC?
10:08:54JdGordonhas nayone actually got any bugs which they are working on almost ready?
10:09:03B4gderJdGordon: I think we should branch first and give some days to cleanup on the branch
10:09:09LinusNany 3.0 showstoppers left?
10:09:14JdGordonI doubt there is any point
10:09:25JdGordonsure branch, but why wait unless someone is actually working on it?
10:09:28pondlifeJdGordon: Nothing stopping you adding fixes to the 3.0 branch
10:09:30B4gderlet's wait for Llorean for some input on that, he's our unofficial release-guru
10:09:48 Join pixelma [50] (i=pixelma@rockbox/staff/pixelma)
10:09:48pondlifeJdGordon: you, or me, I mean!
10:09:56JdGordonLinusN: we went into the freeze with all showstoppers fixed
10:10:04JdGordonpondlife: :)
10:10:09B4gdernot really a wait, but some time to work on the release packages and make them nice
10:10:18JdGordonoh sure
10:10:21LinusNhow are the manuals?
10:10:36B4gderwho reads manuals anyway? B-]
10:10:38*B4gder ducks
10:11:04pondlifeThere are 28 bugs open marked as "3.0", but some are fixed in SVN and awaiting new bootloaders.
10:11:29pondlifeFurthermore, some are on non-3.0 targets
10:11:39pondlife(Assuming we're being picky there.)
10:12:03rasherThere's also a vague hope that manual, rbutil and translations might get a few fixes if we branch for a week's time
10:12:41pondlifeFS #8999 is probably the worst one remaining as far as I can tell - CPU stays boosted.
10:12:42linuxstbSo which targets are we including? I think it will be simpler to just include everything, but there were some people saying the 1g/2g ipod for example wasn't good enough (no shutdown/sleep)
10:13:11pondlifelinuxstb: 1G also suffers from scrollwheel failure -
10:13:31pondlife(Or rather, some do.)
10:13:55JdGordonid say the wheel issue is biugger than cpu staying bootsed... but neither are showstoppers
10:14:11austriancoderpondlife: - without -fPIC
10:14:13LinusNthe h300 bootloader has issues... :-(
10:14:18 Join Mathiasdm [0] (
10:14:57pondlifeLinusN: I implore you and petur to put some of your life aside for that task... ;)
10:15:09pondlifeMy H380 is quite happy, sadly.
10:15:10JdGordondevcon 09!
10:15:21LinusNpondlife: mine too
10:15:54pondlifeLinusN: But yours also has a workiing USB bootloader mode with v5, right? Mine does not...
10:16:15LinusNthat's interesting
10:16:23pixelmaI'm not sure the manuals are up for a release (though I can't tell from the top of my head which important things could be missing, but I know of places with outdated information). And I listened to parts of the DevCon meeting records yesterday... it's been said that a manual release doesn't have to happen with the release - and a manual (or general documentation) cleanup week was suggested
10:17:15pondlifeI suspect an ATA (or something else) init timing issue causing both the USB bootloader issue and petur's issue. Having reordered things since v5, the USB bootloader now works, but we don't really know why...
10:18:50JdGordonpixelma: as long as they dont have wrong install instrucstions...
10:18:59JdGordonthats the only big daanger with the manual
10:21:45pixelmathose could be a lot better for some targets, e.g. H100/300 or the Iaudios don't mention the things you hve to do after putting the bootloader on in the "automated installation" part which comes first (rbutil shows some instructions there but they can easily be overlooked)
10:22:08pixelmarasher: just had a quick look at FS #9347 - why are you suggesting to rename snap tp snap-8? If I remember correctly, it needs 10 pixels of space, or is it just an example?
10:22:54rasherpixelma: Then it should be snap-10, yes. It was just a quick example
10:23:29pondlifeHmm, silly question - how do I redirect both stdout and stderr to the same file (i.e. make >t 2>t)..?
10:23:47funman> 2>&1
10:29:41 Join mf0102 [0] (
10:29:55 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
10:31:22pixelmarasher: ok, there are also fonts which state a wrong number in the name, at least compared to what I would expect. E.g. helvR10 needs 12 pixels, helvR12 needs 14, mix12 needs 13...
10:32:13rasherpixelma: This should also be fixed. Post-3.0 obviously, as set in the "Due in version" field.
10:32:18rasherIdeally as soon as possible.
10:34:17pixelmaI've been wondering whether it would be good or not to do all those renaming at once
10:34:20linuxstbWouldn't that be nice to fix before release? So that themes can be fixed?
10:36:02pixelmaand a suggestion for the helvR* specifically - rename to something slightly different than helvR (maybe "helvetica" or so) to avoid confusion when talking about helvR12
10:37:01funmanI have some trouble with mixed labels in asm:
10:37:29funmanld sees a function into a function
10:37:42 Join krz_ [0] (
10:37:54rasherlinuxstb: I think it should just *after* release, so themes can get fixed before next release
10:38:04pondlifeI agree
10:38:20linuxstbIf you do that, then you'll get a load of themes that work with trunk, but not with 3.0
10:38:46rasherThat's always the risk
10:39:00linuxstbIt's not a "risk" - you seem to be saying you want to do that.
10:39:09 Join Nibbler [0] (
10:39:10linuxstbWhat's the problem with renaming now?
10:39:26rasherlinuxstb: Breaking almost every theme *just* before release is not nice
10:39:56pondlifeaustriancoder: A small subset of the sim build I get:
10:40:19pondlifeI'm using gcc (GCC) 3.4.4 (cygming special, gdc 0.12, using dmd 0.125)
10:40:46amiconnrasher: The svn themes should of course be fixed in parallel with the rename
10:40:48B4gderaustriancoder: I'd say you should revert that commmit and research further why you get the warning
10:40:59linuxstbrasher: I agree it's not nice, but I think it's preferable to the alternative.
10:40:59pondlifeThat's a lot of warnings, makes it hard to see any real problems...
10:41:19LinusNi think release time is just the right time to break things
10:41:34pondlifeLinusN: Immediately post-release, you mean?
10:42:18LinusNbecause you can expect the user to do a clean install for the release
10:43:06rasherSo when 3.0 releases, almost no themes in the wild works? Ew
10:43:28LinusNhmm, that's a point
10:43:38linuxstbrasher: And the alternative is that shortly after 3.0 is released, themes don't work any more with it...
10:43:53rasherlinuxstb: Only if we overwrite the current themes
10:43:59B4gderwell, the 3rd option is to not rename fonts at all...
10:44:22linuxstbOr a 4th option - leave the old fonts there with the broken names, and do "svn cp" to create new ones with new names
10:44:24B4gderuntil we break something else in the wps syntax or similar
10:44:25rasherI don't see why themes couldn't be kept as "3.0"
10:44:39pixelmaI tend to agree with linuxstb (not strongly though) and it also depends on whether we really will have a release more often
10:45:09 Join Mathiasdm2 [0] (
10:45:15 Quit amiconn (Nick collision from services.)
10:45:17rasherThe themes site should offer releases as "required build", I think
10:45:21 Join amiconn [50] (n=jens@rockbox/developer/amiconn)
10:47:45*amiconn would just rename the fonts, fix the svn themes and leave the rest to the theme authors
10:47:54pixelmasomehow I don't like option 3 (because of inconsistent and confusing names) or 4 (beause the font folder will get a lot larger)
10:48:01amiconnThe theme site is something that seems to take forever
10:48:05linuxstbrasher: I just think 3.0 will be around much longer than it will take to fix the themes, and that's the cleaner of the two options.
10:48:20 Join cfp [0] (
10:49:49linuxstbpixelma: Do you know how many fonts are we talking about?
10:49:50pixelmasome themes are still broken anyways
10:50:08rasherI guess as long as we don't have the themes sites, maybe it is smarter to rename now, since authors will probably jsut overwrite their (currently working) themes
10:50:16pixelmalinuxstb: unfortunately no
10:50:57linuxstbI also think most users will have an old .rockbox directory there anyway (with the old fonts), so for them, the themes won't break anyway (if we choose non-conflicting new names)
10:51:08pixelmabut rasher's patch names some and I already mentioned 3 more earlier
10:51:26scotty_007are only the themes affected by the name changes?
10:55:10cfpcouldn't we add, just in release 3, a switch in wps parser to replace new names with old ones ?
10:55:41rashercfp: Sounds really hackish
10:56:05 Join Nico_P [0] (i=53915df2@gateway/web/ajax/
10:57:54cfpbut it could allow every themes to be fixed for trunk
10:58:32rasherI'm pretty sure that's lowest on the list of ways to go. That would be a rather nasty binsize hit
10:59:56*linuxstb proposes a quick vote - 1) Rename prior to release; 2) Rename after release; 3) Don't rename at all; 4) Include copies of fonts with both names; 5) Hack Rockbox to rename-on-load
11:00:20*linuxstb goes for 1
11:00:28*B4gder votes ... 3
11:00:46rasherI'm going to say 1
11:00:51B4gder(with the rename postponed until we do other things)
11:01:26*B4gder still has scars from soname bump fights ;-)
11:01:45 Quit Mathiasdm (Read error: 110 (Connection timed out))
11:01:58cfpi'd say 5
11:02:02rasherB4gder: this is probably less troublesome, but yeah, that option is my next choice
11:02:24*amiconn votes 1
11:02:34linuxstbSo rename a decent amount of time prior to 3.1, and hope we have a themes site that would enable us to do it by script?
11:02:48linuxstbi.e. not immediately after 3.0?
11:03:19*linuxstb would probably still prefer 1.
11:03:21cfpqounds good too
11:04:23*pixelma leans towards option 1 too
11:04:39B4gdermy postpone idea would be to combine the font rename with some other future theme/wps syntax breakage
11:05:00B4gderbut I seem to be in a minority anyway ;-)
11:05:07*gevaerts agrees with B4gder
11:05:27cfpbut if names are changed before the release, it means that we have to ask themes designers to modify their themes just before the release
11:05:30*pixelma suggests calling for an RSB decision ;)
11:05:31linuxstbI think I would prefer to just get it done
11:05:39rasherB4gder: it seems cleaner, yes, but then we might get to wait for ages. It doesn't happen that often, actually
11:06:00rashercfp: Many of the themes can be fixed by anyone, and it's a quite simple fix
11:06:03B4gderright, but I think we should be (more) careful about break themes
11:06:08cfpwe could add a tag for themes working with 3.0
11:06:32B4gderand imo the font rename isn't terribly important
11:06:42rasherB4gder: This is very minor breakage though, and now that we've got the WPS Galleries sorted, it's easier to fix
11:07:03B4gderhm, and a themes site could fix them on the fly ;-)
11:07:27funmanlinuxstb: do you remember why you used ld in mkamsboot Makefile ?
11:07:43gevaertsSo we first finish the themes site, then rename, and then release?
11:07:45*gevaerts hides
11:07:48cfpbut what about adding a "Ready for rockbox v3.0" to complying themes
11:08:03cfpis the themes site so close to being ready ?
11:08:14cfpwouldn't that delay the release a lot ?
11:08:52funmanlinuxstb: please have a look at this patch
11:08:55gevaertsIt would. That's why I needed to hide :)
11:09:17B4gderwe've had the fonts named like this for ages, I think having functional themes is more important than to all of a sudden make the names "right"
11:10:41gevaertsIf we had a "bug" of similar magnitude in the wps syntax (i.e. a wps tag that doesn't have the right name according to some naming scheme), I think it would clearly be post-3.0
11:10:42*rasher ponders doing something with bash, curl and sed to fix the themes in the WPS GAlleries
11:10:47***Saving seen data "./dancer.seen"
11:11:48cfpi guess I agree with gevaerts
11:12:30rashergevaerts: but what then, wait until next release?
11:12:35rashergevaerts: When to fix it?
11:12:47pondlifeRename fonts as a 3.1 todo?
11:12:56gevaertsSo my proposal is : release fonts as they are, split theme site and/or wiki in release and svn parts, and fix
11:13:00B4gderand when we have a themes site preferably
11:13:29 Join gregzx [0] (
11:13:48rasherThat really should be a 3.1 goal.
11:13:54rasherThe themes site, not the fonts.
11:14:02B4gderin fact, once we have such a site we could always try to convert themes to the version the users uses, as long as it knows all the quirks
11:14:19gevaertsWe will probably have some wps changes in svn during the 3.0 lifetime, so we need to somehow support two theme branches anyway
11:15:15 Join avis- [0] (
11:25:51 Join create [0] (
11:26:07linuxstbIMO we should either do it today, or wait until either we have a themes site, or another major wps/theme change that breaks compatibility. Avoiding the need to split the themes gallery should be the #1 priority.
11:26:57 Quit stoffel_ ("Lost terminal")
11:27:17linuxstbAnd perhaps we should only break compatibility just prior to releases - to ensure themes always work with the most recent release.
11:29:29linuxstbfunman: ld is needed to ensure the code is linked to run from the correct address. Things didn't work correctly without it.
11:29:44 Quit avis (Read error: 110 (Connection timed out))
11:29:51 Nick avis- is now known as avis (
11:30:02 Nick create is now known as cfp- (
11:30:34cfp-linuxstb: then we should wait
11:30:34gevaertslinuxstb: that's a nice idea in theory, but would it work in practice? It means that something like viewports-in-wps can't have a decent testing period before it's in a release
11:30:42 Quit cfp (Read error: 110 (Connection timed out))
11:30:49 Nick cfp- is now known as cfp (
11:31:46linuxstbgevaerts: Things are tested before going in SVN anyway - by people either compiling the patch, or unsupported builds.
11:32:12linuxstbAnd "new feature" mostly doesn't mean "breaks compatibility"
11:32:32rasherlinuxstb: I think as soon as we have a themes site, keeping themes for specific releases should get trivial
11:32:42 Quit pondlife (Read error: 110 (Connection timed out))
11:32:47*cfp votes for not breaking anything and leaves
11:32:52 Quit cfp ()
11:33:38gevaertslinuxstb: true.
11:34:21*gevaerts thinks we can't spring a huge font-renaming on theme authors a week before the release without prior warning
11:35:14linuxstbI don't see it as springing it on theme authors - anyone and everyone can edit one line in the theme.cfg file and reupload the theme.
11:36:05amiconnBagder: Imo it'd be a good thing to rename the fonts before 3.0: (1) It was discussed at devocn that the fonts folder needs cleanup
11:36:25amiconn(2) Several fonts are *unusable* atm anyway. The rename would go in parallel with the fix
11:37:01B4gderI disagree
11:37:08B4gderremoving silly fonts we can do at once
11:37:31gevaertslinuxstb: if that "anyone" gets a bit more concrete it could be acceptable. Just relying on anyone without knowing a name however doesn't sound right to me
11:38:31amiconnAnd (3) postponing things isn't good in general. The postponing tends to go forever...
11:43:30rasherB4gder: Is it possible to edit files directly in twiki without breaking anything?
11:43:41rasherB4gder: for someoen with server access, I mean
11:44:11B4gderit's bascially RCS behind the scenes
11:44:23B4gderon individual files I mean
11:45:16linuxstbHmm, so we could run a script on the server to fix themes...
11:52:26 Quit Nibbler (Read error: 113 (No route to host))
11:53:17 Quit Nico_P (" ajax IRC Client")
11:55:49 Join PaulJam [0] (
12:01:35 Quit swimmer (Client Quit)
12:04:35 Join swimmer [0] (
12:09:33 Nick fxb__ is now known as fxb (
12:12:58 Join culture [0] (
12:15:28 Join swima [0] (
12:16:14 Quit swimmer (Read error: 60 (Operation timed out))
12:16:45 Nick swima is now known as swimmer (
12:17:53 Quit gregzx ("ChatZilla 0.9.83 [Firefox 3.0.1/2008070208]")
12:19:18 Join dabujo [0] (
12:35:40 Join BigBambi [0] (i=86ceaf40@rockbox/staff/BigBambi)
12:38:26BigBambiIs anyone else unable to reach the forums?
12:39:02gevaertsI can reach them
12:39:20BigBambiBugger. I wonder if work has done something sneaky
12:39:45BigBambigevaerts: Could you give me the IP?
12:40:45BigBambigevaerts: That appears to be a picture of a plastic lizzard in a chair
12:44:54gevaertsBigBambi: it's a name-based virtual host
12:45:19*gevaerts blames scorche|sh for that lizzard
12:45:43funmanld gives me trouble mistaking labels with functions:
12:45:59 Join moos [0] (
12:46:01BigBambigevaerts: ah, damn. Thanks anyway
12:46:19gevaertsBigBambi: time for local /etc/hosts editing?
12:47:09BigBambiah good plan
12:47:26BigBambiexcept the windows version (at work :°°
12:48:42funmanis there some syntax I can use in my .S files to tell ld not to consider said label as a function ?
12:58:36 Join Nico_P [0] (i=53915df2@gateway/web/ajax/
13:01:43linuxstbfunman: Can you post your entire test.S file?
13:04:48funmanif I use label: instruction instead of label:\ninstruction it's better
13:05:40funmanno, the problem comes if I put a comment right after the label
13:07:09linuxstbfunman: You didn't close the comment at line 33...
13:07:45funmanah damn !
13:07:58funmanthanks for your snake eyes
13:09:59linuxstbBTW, the Rockbox style is to write assembly instructions in lower-case, so you may want to get used to that style...
13:10:39linuxstbupper-case is reserved (in general) for pre-processor symbols
13:10:49***Saving seen data "./dancer.seen"
13:12:41 Quit mazling (Read error: 110 (Connection timed out))
13:18:34 Join swima [0] (
13:18:41 Nick Kopfi|offline is now known as Kopfgeldjaeger (
13:21:14 Join Thundercloud [0] (
13:22:39 Part swima ("[IRSSI] Should I do my BOBBIE VINTON medley?")
13:24:03 Join Schmogel [0] (
13:33:09 Quit swimmer (Read error: 110 (Connection timed out))
13:34:14 Quit BigBambi (" ajax IRC Client")
13:39:28 Join swimmer [0] (
13:46:58 Nick Kopfgeldjaeger is now known as Kopfi|offline (
13:47:39 Join faemir [0] (
13:49:53 Join JordanG [0] (
13:53:36JordanGHey all, random (probably not rockbox's fault) question: I lost a hard drive, so I'm trying to copy my music off my ipod, but it seems lots of files get corrupted (they skip around, make clicking sounds, etc) although these errors aren't present when I play straight from the ipod. Any ideas?
13:55:03 Join fdinel [0] (
13:55:48GodEaterso they sound bad when played back from the "new" hard drive ?
13:56:28JordanGGodEater: indeedy.
13:56:50JordanGThough they sound fine playing from the standalone ipod, or when i play them on my computer w/ the ipod hooked up via usb.
14:00:29JordanGweird, huh?
14:01:19gevaertsJordanG: try comparing files on the ipod with their copy on the PC
14:01:21 Join miepchen^schlaf [0] (
14:02:29JordanGgevaerts: compare how?
14:02:57gevaertsdepends on your os
14:03:45JordanGI'm running linux
14:04:07GodEaterwould be a good method
14:04:16gevaertsOK. md5sum on both sides then (md5sum /media/ipod/file /media/non-ipod/file)
14:04:18GodEateror some other hash
14:05:12JordanGah, gotcha
14:05:36JordanGgimme a min to find a file w/ problems
14:07:37JordanGhmm. hold on, something has made darn near everything freeze. which is... odd...
14:07:51*gevaerts suspects the new hard drive
14:08:33*GodEater does too
14:09:14JordanGi don't think so, i had this problem on two different machines.
14:09:46gevaertsGodEater: which svn revision are you using in your FS #8943 tests?
14:10:14 Quit JordanG (Remote closed the connection)
14:10:16GodEatergevaerts: last test was with 18355
14:11:11gevaertsGodEater: can you try with r18370 or later?
14:11:55GodEaterI can - but not right now
14:11:59 Join JordanG [0] (
14:12:32GodEaterwhat's the difference ?
14:12:40gevaertsBasically my sansa "fixes" broke UMS for everyone else.
14:14:34JordanGhmm. My computer is really not pleased with my ipod now for whatever reason... It just locks up when I try to access the thing.
14:14:38 Join zakk_ [0] (
14:15:18rasherJordanG: This is not really Rockbox related.
14:16:25JordanGrasher: Probably not, just thought people 'round here had some expertise in the subject, and I've been scratching my head about it for a while. I can bugger off, if you'd prefer.
14:16:39 Join krz__ [0] (
14:16:46GodEatergevaerts: broke it in what way? With the same symptoms ?
14:16:55rasherJordanG: Well this channel is strict about being on topic. There is #rockbox-community for off-topic chatter
14:17:07GodEaterwhere similar people hang out ;)
14:17:31gevaertsGodEater: it eats filesystems for lunch
14:17:49GodEaterwith or without sauce ? :)
14:17:50gevaertsI noticed on ipod (both mini and video)
14:17:53JordanGrasher: gotcha, thanks.
14:19:53LinusNGodEater: "same symptoms"? like "completely hosed"? i love detailed bug reports :-)
14:20:21 Part JordanG ("ERC Version 5.3 (IRC client for Emacs)")
14:22:15GodEaterLinusN: I could have dumped the disk for you I suppose
14:22:19GodEatershould I do that next time ? :)
14:22:37GodEaterlinux wouldn't even recognise it as VFAT any more
14:22:57LinusNwhich is worse btw, "completely hosed" or "eats filesystems for lunch"? :-P
14:22:58GodEaterI had to force it to mount with -t vfat, and when I did, the mess on there was pretty indescribable
14:23:28GodEaterthere was no trace of a directory structure, and the file names that were there were using characters I don't have buttons for on my keyboard
14:23:40GodEateris that better ?
14:23:59LinusNthere you go. your filesystem has been eaten alive :-)
14:24:15gevaertsAnyway, I'm pretty sure that it was my fault, not yours :)
14:24:34*B4gder stands up to take the blame
14:24:38B4gderme me!
14:24:44*LinusN blames bagder
14:24:51*GodEater blames perl
14:24:58B4gderI'm sorry, I'll do better next time!
14:25:25gevaertsYou should have made an svn pre-commit script that rejects buggy code!
14:25:25LinusNyeah, or we will revoke your svn access!
14:25:44B4gderyou mean I'm free to go now?
14:26:29 Quit Mathiasdm2 ("Invisible Internet Project:")
14:29:47GodEaterB4gder: hold on sonny - you don't get out of here that easily
14:30:17 Join fragilematter [0] (n=barbu_do@
14:31:29B4gderaustriancoder: so how about that revert?
14:32:24*GodEater still needs to get http access to the git repo working
14:32:38GodEaters/still needs/still needs B4gder/
14:32:51 Quit krz_ (Read error: 110 (Connection timed out))
14:32:52B4gderor even the svn repo over http, right?
14:33:00Nico_Pthat would be nice
14:34:30GodEaterwhile I would applaud both, I can only use GIT
14:34:44GodEaterour proxy doesn't understand the PROPFIND methods that svn-over-http requires
14:35:21GodEaterat least, it didn't last time I tried
14:35:30GodEaterit did the electronic equivalent of "huh?"
14:35:55Nico_Psame here IIRC
14:35:57B4gderyeah, older/stricter/sillier proxies do that...
14:36:11GodEaterI think mine falls into the two latter classes
14:36:13Nico_PGodEater: I think the git repo can be cloned fine over http
14:37:40GodEaterindeed, hence my request ;)
14:37:59GodEaterunless you're saying it already can be
14:38:02GodEaterand I'm just being dull ?
14:38:15Nico_PI think it already works
14:38:23GodEaterI mislaid the repo url
14:38:26GodEaterdo you have it to hand ?
14:39:17Nico_PI don't remember it either... B4gder?
14:39:30GodEaterhehe - I asked him last week ;)
14:39:55B4gderits not over http iirc
14:40:17B4gderwe run git-daemon from xinetd for it
14:40:34 Join Zarggg [0] (
14:40:42Nico_Poh then maybe I'm confusing it with the one I have at
14:45:31 Part fragilematter
14:51:45 Join cfp [0] (
14:52:41B4gderI'm thinking we talk to llorean before we branch
14:52:49 Nick Kopfi|offline is now known as Kopfgeldjaeger (
14:53:07B4gderhe being unofficial release-person
14:53:27B4gderother than that, I see nothing in the way
15:01:32funmanoops, my code overflows by 2 instructions
15:02:04linuxstbYes, hacking the v2 brings a new dimension to binsize...
15:03:03funmanI could check other versions of the firmware to see if I can put more than 0xa4 bytes
15:03:53linuxstb0xa4 bytes should be enough for anyone
15:03:53austriancoderB4gder: revert done
15:06:30 Join borelg_ [0] (
15:07:48 Quit dabujo ("( :: NoNameScript 4.2 :: )")
15:08:05 Part cfp
15:08:42borelg_please can you answer me?
15:08:58linuxstbborelg_: Did you ask a question?
15:09:42borelg_yes please
15:09:53borelg_no sorry
15:10:11borelg_ok my question is this
15:10:39borelg_how can I understand if my mp3 player is compatible with another custum firmware?
15:10:47borelg_sorry for my english....
15:10:54***Saving seen data "./dancer.seen"
15:11:40linuxstbThe mp3 players that Rockbox works on are listed at the top of this page -
15:12:38borelg_yes I've read this but a person said that his mp3 (that it doesn't match in the list) can be modified
15:13:10 Join BingoJ [0] (
15:13:35borelg_because its firmware was the firmware of another mp3 that you have modified
15:13:38B4gderborelg_: that might be true, but it doesn't run rockbox
15:14:24borelg_and there isn't an howto to modified the firmware?
15:14:33borelg_i know that is very very difficult
15:15:59borelg_my mp3 is from mediacom... there isn't it in the list....
15:16:09borelg_there is a solution?
15:16:13B4gderyes, see that url
15:16:17B4gderand start working
15:16:21borelg_ok thanks
15:16:33B4gderand gather some skilled friends too, this is going to take time!
15:16:52borelg_ok but what url?
15:17:06borelg_ah sorry
15:17:23borelg_what could I do?
15:17:35B4gderyou mean you're done already?
15:18:02borelg_What should I do?
15:18:08borelg_for my english
15:18:52desowineverything what's described in that wiki page
15:18:53 Join LambdaCalculus37 [0] (
15:19:43borelg_but where?
15:19:57desowinthe link B4gder posted twice
15:20:03B4gderI think its a bot
15:20:09B4gderit loops already
15:20:12borelg_What do I have to search
15:20:33borelg_yes i've understand
15:20:33B4gderborelg_: go read that page and stop being stupid
15:20:43borelg_and I'm not a bot....
15:20:57gevaertsB4gder: we're all bots
15:21:14borelg_ok I've understand that I must go to that page
15:21:25borelg_but in that page there are many links
15:21:39*B4gder stops talking and just looks at borelg_
15:21:40borelg_what link do I must read?
15:22:26B4gderyou must read the entire page
15:22:28B4gder3 times
15:22:34B4gderthen a forth time
15:22:41B4gderthen ask us again if you need clarifications
15:22:45 Join vitja [0] (
15:23:01borelg_ok thank you very much!! :-)
15:23:08borelg_now I go to read
15:23:19borelg_bye bye and good afternoon
15:23:34 Quit borelg_ ("Ex-Chat")
15:24:08scotty_007Badger: stop making jokes, please - got tears in my eyes already :-)
15:26:14*LambdaCalculus37 goes to the logs to see what all of this was about
15:31:41 Join Mathiasdm [0] (
15:36:00 Join n1s [0] (n=nils@rockbox/developer/n1s)
15:39:42 Part scotty_007
15:42:42 Join kugel [0] (n=chatzill@unaffiliated/kugel)
15:43:14 Nick MU{lappy} is now known as Unhelpful (
15:47:42 Quit austriancoder ("Lost terminal")
15:49:48 Join J-23 [0] (
15:56:36 Quit LambdaCalculus37 ("Do quit now, there's a demon around the corner!")
15:57:22 Join JohanC [0] (n=51f206a8@gateway/web/cgi-irc/
16:02:11 Join oofus [0] (
16:14:39kugellinuxstb, rasher: Most themes won't break, as most bring the fonts they are using with them
16:15:24gevaertskugel: in that case they will break other stuff...
16:15:33gevaertsSo they need fixing anyway
16:16:23kugelgevaerts: sure they need fixing, but they still will work. They will just create a larger font folder
16:16:40gevaertsAs some of the renames only fix the size, I guess there could very well be collisions then
16:16:50*gevaerts didn't actually check this
16:17:09gevaertsOf course we could do the rename so that this doesn't happen
16:18:37kugelgevaerts: As far as I read the renaming wouldn't only fix the size, but rather replace the the short cryptic abbreviations with something usefull
16:20:00kugelSo, if a theme uses helvR12 (which is included in the theme package), and the rockbox version was renamed to helvetica14, the theme would copy the old helvR12 to the font folder and still work
16:23:22Jorsika question
16:23:44Jorsikcan i take d2 official firmware and modify it somehow?
16:26:15B4gderJorsik: sure go ahead
16:27:33Jorsikwith what do i open it?
16:28:18B4gdersince this is rockbox, let's assume we talk rockbox and then you don't do that at all actually. We write a new firmware for it
16:28:54Jorsikthere is no rockbox for d2
16:29:05B4gderthere is ongoing work to put rockbox on the d2
16:29:15Jorsikhow far is it
16:29:38B4gderit runs rockbox, but not really stable for every-man use
16:29:49kugelfunman: would it be helpful if i get a fuze?
16:29:56B4gdersee the dedicated forum thread for details
16:30:15 Quit vitja (Read error: 110 (Connection timed out))
16:30:28Jorsiki only need to assign functions to buttons
16:30:36funmankugel: sure, nobody has tested it so far afaik
16:31:04GodEaterJorsik: and why would you assume that that's simple ?
16:31:13Jorsiki dont
16:31:30Jorsikbut i wanna try
16:31:46kugelfunman: I need to get a new device soon anyway. I'm thinking of either a fuze or a sony a818 (sony provided source code of the firmware (linux kernel) and bootloader (u-boot))
16:31:46GodEaterwell please go ahead
16:31:49B4gderJorsik: then read up on the forum thread and join in
16:32:12 Join tvelocity [0] (n=tony@
16:32:15kugelboth cost almost the same, 8GB for ~100€
16:32:15B4gderkugel: don't count on those sources to be any fun
16:32:21funmankugel: if there is nothing hidden it's not funny :)
16:32:25B4gderthey all release "sources"
16:32:38B4gderwith all the fun being binary modules anyway
16:32:40 Part LinusN
16:32:57B4gderso the sources are stock linux and uboot tarballs you can download from anywhere
16:33:10kugelB4gder: No, they're modified
16:33:22B4gderbut are there hw-specific drivers too?
16:33:39B4gderthen things are better than average! ;-)
16:33:45kugelI compared the sony sources with stock sources
16:33:54 Quit Nico_P (" ajax IRC Client")
16:34:02 Join Nico_P [0] (i=53915df2@gateway/web/ajax/
16:34:10linuxstbkugel: Do you have that diff available?
16:34:40kugelthere's a whole mp200 (the soc the sony uses) dir which contains a lot, which isn't found in the stock source
16:34:58kugellinuxstb: yea
16:35:16linuxstbIs a datasheet available for that SoC?
16:35:36kugelI assume "diff -ru dir1 dir2" is the way to go?
16:36:02B4gderkugel: yeps
16:36:15linuxstbMaybe with -N?
16:36:21B4gderright -N too
16:36:33kugellinuxstb: yes and no, sony released a very detailed service manual, but nec (the producer of the mp200) didn't reveal much
16:36:47kugelthere's a thread in the new ports forum
16:36:54B4gderand hopefully those changes are hw-changes too, and not just platform setup stuff etc
16:37:41 Quit mf0102 ("Ex-Chat")
16:40:12kugellinuxstb: the diff is pretty big, I recommend using a tool like beyond compare
16:43:48kugellinuxstb: have you had a look at the service manual
16:46:48 Quit Schmogel ("Miranda IM! Smaller, Faster, Easier.")
16:48:40 Quit swimmer (Read error: 104 (Connection reset by peer))
16:49:19 Join swimmer [0] (
16:50:24 Quit zakk_ (Remote closed the connection)
16:55:35 Join Thundercloud_ [0] (
16:55:37kugelfunman: Have you found infos about the ams3525?
16:56:26funmanwhat kind of infos ? did you follow the forum thread ?
16:56:43kugelfunman: like MHz, components
16:56:52funmanI have the datasheet
16:57:05kugelyes I followed the forum thread, but I can't remember every single post
16:57:16funmanbut most of the info is cryptic for me
16:57:19kugelfunman: do you have a link?
16:57:32 Quit JohanC ("CGI:IRC (Ping timeout)")
16:57:33funmanit's not public but ams give it away for free
16:58:09kugelSo, it's impossible to share?
16:58:29kugelif I don't make it public?
16:58:48funman it's 266MHz
16:59:30B4gderthose sony diffs contain a million unrelated changes
17:00:36B4gderwhich explains why the diff is so enormous
17:00:41kugelB4gder: Yes, but also many related changes. As I said, using a compare tool like beyond compare is more useful than looking at the plain diffs
17:01:09B4gderI'm not arguing about its usefulness
17:01:22B4gderI'm just saying a vast portion of the diffs are unrelated
17:01:34B4gderno matter what diff tool you use
17:02:31B4gder"Linux kernel version: 2.6.10-argonplusevb_psram_default-makena_dev" ;-)
17:03:17B4gderI think also that the sony sources come from various linux versions
17:03:23B4gderwhich explains the diffs all over
17:04:28kugelB4gder: yea, I also saw some Nokia copyright notices
17:04:44B4gderyes, but in omap
17:05:22B4gderthe mp200 code is (c) NEC
17:05:56B4gderand montavista it seems
17:06:35kugelbut it's still modified for the mp200 (e.g. the arch/arm/mach-mp200 folder is entirely new)
17:06:44B4gderand some lovely asian comments!
17:07:27B4gderuhm, s/some/many
17:08:51kugelthe service manual provided detailed info, e.g. block diagrams, pin assignments
17:09:00B4gderthat's nice
17:09:24 Quit Thundercloud (Read error: 110 (Connection timed out))
17:10:16kugelso, the question is still: sansa fuze or sony? :)
17:10:56***Saving seen data "./dancer.seen"
17:11:18kugelwith both eyes looking at porting rockbox to it
17:11:32 Quit Lambduh ("Leaving")
17:12:08 Quit BingoJ ("Leaving.")
17:17:16JdGordonin a similar line of thinking to renaming fonts... anyone made a descision on when to remove the deprecated lang strings and bump the lang version?
17:17:44 Join bughunter2 [0] (n=Jelle@
17:23:19 Quit Zarggg ()
17:25:40rasherkugel: yuck
17:26:11 Join virtuoso015 [0] (n=vinay@
17:26:56 Quit miepchen^schlaf ()
17:31:11 Quit JdGordon (Read error: 104 (Connection reset by peer))
17:32:03beta2kRemove them
17:32:08beta2kDescision made :)
17:33:42 Join cfp [0] (
17:33:43 Part cfp
17:34:41rasherkugel: So even if we fix the fonts, the user might still end up with broken fonts because themes include them? That's really awful
17:35:27kugelrasher: yes, and depending on how the renaming looks like, the themes will even break rockbox, e.g. when helvR12 is renamed to helvR14
17:36:03kugelthemes including a helvR14 would overwrite the rockbox version
17:37:24 Join toffe82 [0] (
17:38:18amiconnNo renamed font will have the name of a previously existing font
17:38:48 Join herrwaldo [0] (
17:41:10 Quit oofus (Read error: 113 (No route to host))
17:41:22rasheramiconn: By virtue of the current names being so stupid
17:41:34rasherAlso careful planning!
17:44:00 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
17:45:10 Join oofus [0] (
17:45:42pondlifeMaybe the new font names could be formalised to allow them to be listed in size order (i.e. all the 12-point fonts together)?
17:46:05 Part B4gder
17:46:12pondlifefamily.size.fnt or somesuch?
17:46:28LloreanI'm going to be out all day today, at least. It's been a holiday weekend, and I've been busier with family stuff than expected, but am just dropping in to say I think we should go ahead and branch.
17:47:37pondlifeWho will actually do the branch? Someone who knows more about SVN that me, I hope...
17:49:19LloreanYeah, I'd certainly bugger it up entirely. It's a slightly ceremonial thing, so I'd put Bagder forward as them an who should do it, but as long as it gets done in the next day or so, and it gets done the "right" way, I'm happy. :)
17:49:38pondlifeBagder's not around though... so maybe tomorrow?
17:49:55pondlifeAh, sorry - misread your words
17:50:18 Join sbhsu [0] (
17:50:52Llorean"as the man"
17:50:57*Llorean has spacebar issues.
17:52:45 Quit Nico_P (" ajax IRC Client")
17:52:57 Quit funman ("leaving")
17:53:40pixelmapondlife: if you rename the fonts so that all 12-point ones are together etc., fonts which exist in more than one size will be apart. Not the nicest either...
17:54:34pondlifepixelma: By default they should be alphabetical still, but I'd like the ability to organise by point size in future.
17:55:04LloreanI'd prefer point size.
17:55:07kugelpondlife: Yea, I thought of this too. Should be doable
17:55:08pondlifeNothing urgent, but maybe worth including this sort of thing in a new naming convention.
17:55:14LloreanI don't need to see "what different sizes are Nimbus available in"
17:55:23LloreanI need to see "which fonts won't break the vertical spacing of my current theme"
17:56:43kugelI want to have both
17:56:52 Join cfp [0] (
17:57:28 Part J-23
17:57:36Lloreankugel: Since we can't name files both ways, the simplest solution is to have the files named in such a way that the default sort handles it either as one or the other.
17:58:26LloreanAnd I feel a much saner default is by font size, since the size or size range is going to determine which fonts a person is willing to use.
17:58:52LloreanIt's rare that someone says "I want to use Courier, and I don't care at all what size it is, 8-point is fine, whatever"
17:59:47kugelLlorean: afaik the size is in the header of a fnt file, so we could have a extra menu, sorting the fonts not by filename, but by what's written in the header
18:00:07cfphey guys !
18:00:08LloreanAnd that would require opening and scanning every font file, every time.
18:00:14cfpsorry to drop in
18:00:27cfpbut has anything been decided about the fonts being renamed ?
18:00:31kugel"USHORT height 2 font height in pixels "
18:00:38cfpor is it still being discussed ?
18:00:52Lloreankugel: That doesn't address it being horribly inefficient compared to filenames.
18:00:58kugelLlorean: or building a database out of it once.
18:01:12Lloreankugel: Ah, yes, the "let's just keep it all in RAM and waste more of it" excuse.
18:01:24Lloreankugel: Why is "sort by name of font" preferable?
18:01:46*amiconn also prefers sorting by name
18:02:01kugelI really see a point in sorting by font familiar, but I'd prefer size sorting. If it's too complex and not worth it to have both, then rather size sorting
18:02:11SoapPerhaps a "theme maintainer" badge, Llorean.
18:02:22SoapGet some people signed up to maintain themes.
18:02:28LloreanSoap: Good idea!
18:02:45rasherWhy not just include the font size first in the name?
18:02:46 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:02:47SoapAnother one of those things non-coders can do.
18:02:49Lloreankugel, amiconn: Under what conditions is "family" going to be a higher priority than "size"?
18:02:57Lloreanrasher: That's what I'm in favour of.
18:03:29amiconnChoosing another font size of the same style...
18:03:34kugelLlorean: I said I prefer size over family, but I most prefer most
18:03:36SoapThe only clean way, IMHO, to address the font size issue is to just bite the bullet and fix it now. I can't think of a scenario which won't cause someone some grief if a clean sweep isn't just made.
18:03:38 Join Nibbler [0] (
18:03:47amiconnYou know, I'm used to non-graphical themes which aren't locked in to a single font size
18:04:02 Quit domonoky (Client Quit)
18:04:04Lloreanamiconn: But often are limited by the screen size, so you need "X or less" for the size anyway.
18:04:29SoapIt would be nice if the clean sweep could be done AFTER the new theme site is up, though. That way RBUtil could be updated not to use and that site could be completely removed from usage.
18:04:33Lloreankugel: Yes, you prefer "attempt to have your cake and eat it too", and I've tried to explain to you several times that sometimes we need to pick between two things rather than shoehorn them both in.
18:04:36 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:04:45amiconnPlus, fonts of the same family tend to have the same or similar unicode coverage. Not so for different fonts of the same size
18:05:12SoapFor I don't see how any method of fixing the font size can be accomplished so long as is still pointed to by rbutil.
18:05:31LloreanSoap: We should just disable fonts in RBUtil in the next build, and until we have a real site.
18:05:36LloreanSoap: Er, disable themes
18:06:18 Quit cfp ()
18:06:34kugelWell, isn't it possible to have two font folders: 1 which contains the fonts sorted by size, and 1 which contains links the the fonts sorted by family
18:06:52kugels/the the/to the
18:07:42SoapWhile I think disabling themes in rbutil is _a_ solution, Llorean, I don't think it is a good one. People won't upgrade, and the satellite forums (such as ABI) will just tell people to use an older version of rbutil if functionality (no matter how broken) is removed. (This is my crystal ball speaking).
18:07:42kugeland creating the folder with the links could also be done at compiling time
18:07:53kugelhow does that sound?
18:08:04rasherI still think that font size is far more likely to be important than family for most users, most of the time
18:08:17LloreanSoap: If nearly every single theme in RBUtil doesn't work anyway, I don't see why the satellite forums would do that.
18:08:58Lloreankugel: Very hackish and annoying, since people will try to add their own fonts for themes and be confused by the fact that it shows up in one sort and not the other, and so on.
18:09:02rasherSo having the size first in the name seems the most sensible option to me.
18:09:52Lloreankugel: Use of themes is basically dependent on a maximum font size in nearly all cases, so the majority of cases need sorting by size anyway to make it easier to find a font you can use for your situation.
18:10:02kugelLlorean: then we have to tell them to either not include fonts at all, or create both
18:10:25Lloreankugel: Or we could pick a single sorting method and be done with it...
18:10:45Lloreankugel: Tell me, when exactly is sorting by family preferable? Under what circumstances?
18:10:55LloreanWhat conditions make size irrelevant?
18:11:11*domonoky just wants the new theme site up and running... the you dont need to disable theme installing in rbutil :-)
18:11:39Soapthat was kind of my point.
18:11:53 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
18:12:01kugelLlorean: both are relevant...
18:12:10domonokybut as it looks like we wont have the new site till the 3.0 release, we should just disable it for the rbutil release version...
18:12:10amiconnLlorean: I still prefer sorting by family, also because the list would look less strange
18:12:17LloreanSoap: I'm just assuming it won't be up for 3.0, and for 3.0 we'll want a new RBUtil build with the latest fixes, so why not disable themes in it for now?
18:12:32amiconnHowever, if we're going to sort by size, we need zero-padding for sizes <10 pixels
18:12:54amiconnOtherwise the ordering would be really strange
18:12:54Lloreankugel: Yes, but under what conditions is Family important to the point where you don't need to check size first?
18:13:04 Join funman [0] (
18:13:06Lloreanamiconn: Since we control the filenames, that's no problem.
18:13:27amiconnLlorean: Sometimes you want a specific familiy but don't need the exact size
18:13:56amiconnE.g. I'd prefer a 12px font when it looks nicer than a 13px font (which would be best-match size wise)
18:14:01Soapthemes could be updated in a weekend with 5 people I'm guessing.
18:14:11domonokyamiconn: but for most of our wps you need a specific size, and not a specific familiy, or the wps layout will break..
18:14:19rasheramiconn: Please understand that you're not the typical user..
18:14:27amiconnI know...
18:14:38kugelLlorean: with my solution you can have both. I don't see why you are rejecting this with the excuse "users will...."
18:14:39Lloreanamiconn: Yes, but even then, you know 14px+ is going to push information off the screen, so you can scroll to 13px, see what fonts are available from there downward (upward?)
18:15:09rasheramiconn: I understand your point perfectly, I just think sorting by size would benefit more users
18:15:25Lloreankugel: I'm rejecting it with the excuse "it requires making things more complicated, not less, and creates additional code complication"
18:15:38Lloreankugel: You need to create code to manage these links to the other fonts, to know what to do with them.
18:16:06kugelI don't have a problem with creating code
18:16:07 Quit Thundercloud_ (Read error: 110 (Connection timed out))
18:16:19Lloreankugel: You really need to learn that we favour trying to keep things simple where possible. Having two font folders, one with links, is the opposite of simplicity. Having the files named in a way that is expected to benefit the most users, is simple.
18:16:22amiconnOtoh, sorting by size would have the advantage that we could exclude extremely large fonts for small LCDs (e.g. >16px for archos) and extremely small fonts for fine-pitched LCDs (e.g. tinyprop* for G5)
18:16:25pixelmaI'm torn myself (usually I pick from font size) but I can't imagine this in a list (I can't actually say why but it feels unnatural, maybe it's because what one is used to from a PC). Also you could see which font size is missing from a certain font family you like... "simple solution" all fonts come in all sizes ;)
18:16:37amiconnI mean exclude them while packaging
18:16:49Lloreanpixelma: If all fonts came in all sizes, we could have a separate font size menu. :)
18:17:17amiconnIn fact we could do that now, but going by name makes it easier
18:17:30kugelSoap: if you need help with maintaining themes, I'd volunteer
18:17:52SoapI know I know I know about not giving timelines - but relatively how far away is the replacement site? I don't even know what issues still remain to be resolved.
18:18:15SoapIMHO 3.0 should wait for the replacement theme site if it is anywhere close to complete.
18:18:39amiconnSo, SS-Family-Variant is the way to go?
18:18:42Soap3.0 RC1 could be next week, but 3.0 waits until theme site populated with working themes and rbutil is out.
18:18:52amiconn(e.g. 12-Nimbus-Regular)
18:19:06Lloreanamiconn: That would be my personal preference.
18:20:02amiconnI'd go for the full names instead of weird abbreviations, btw.
18:20:27 Join fragilematter [0] (n=barbu_do@
18:20:36 Join krazykit [0] (
18:20:38 Quit funman ("leaving")
18:20:42amiconnHelv->Helvetica, Ter->Terminus etc
18:20:45*pixelma wonders whether there would be a way for different sorting methods to have both, though can't imagine one
18:20:50domonokySoap: the code for the new theme site is in SVN, dont know whats missing from it..
18:21:59 Join bertrik [0] (
18:22:02kugelpixelma: I proposed to look into the header of a fnt to sort by size, but it's to complex
18:22:12kugeltoo rather
18:22:51 Join Ubuntuxer [0] (
18:23:12Lloreankugel: It's much simpler to have the filenames consistent and just parse them anyway, if you're gonna do any operations on data.
18:23:24LloreanSince you already have the filenames available since you're listing the files.
18:23:28rasheramiconn: sounds wonderful
18:24:01kugelLlorean: of course. I was just replying to pixelma that it's possible in theory
18:24:34kugelshe wondered if there's a way
18:25:10 Join Lambduh [0] (
18:26:16Lloreankugel: Well, the obvious way would be to use the information from the filenames though. If they're named consistently, you can just ignore the first three characters to get the family name.
18:26:25LloreanWell family-variant
18:26:44*Llorean doesn't see why you'd ever peek into the file's header at all.
18:27:16 Quit krz__ ("ChatZilla 0.9.83 [Firefox]")
18:27:23kugelLlorean: I realized that this way isn't good long ago. I don't know why you are still arguing against it
18:29:22pixelmaI thought filenames too, like sort ba last 2 characters of the filename but I guess even if that would work it comes with a penalty too that is not worth it...
18:29:24Lloreankugel: Since you never said you realized it was bad, and the last thing you said about it was still in favour of it, I had no way of knowing this. You merely started offering other alternatives, never actually acknowledging you realized it was bad, and brought it back up as an "option" to other people, so I had to assume you thought it was still bad.
18:29:27Lloreaner still good
18:30:41 Quit jhMikeS (Nick collision from services.)
18:30:47 Join jhMikeS [50] (n=jethead7@rockbox/developer/jhMikeS)
18:31:25pixelmaI hoped someone had an ingenious new idea...
18:34:43 Quit agaffney (Read error: 104 (Connection reset by peer))
18:34:54 Join agaffney [0] (n=agaffney@gentoo/developer/
18:35:51 Quit PaulJam (".")
18:37:43 Join Seed [0] (
18:37:49 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:39:43 Part Ubuntuxer
18:44:27 Quit Mathiasdm ("Invisible Internet Project:")
18:44:51 Join miepchen^schlaf [0] (
18:45:34 Join mf0102 [0] (
18:46:55 Quit Seed ("cu, Andre")
18:47:05*linuxstb reads a long discussion about font name conventions in the logs, but didn't notice a concensus about when to do the renaming...
18:47:25*rasher thought SS-Family-Variant.bdf
18:48:02 Nick oofus is now known as oofus[away] (
18:48:12 Part Llorean
18:48:49 Quit krazykit (Read error: 110 (Connection timed out))
18:48:51linuxstbI think SS-Family-Variant will look odd (Family is usually written first), but agree it's probably the most useful way to sort them, given how Rockbox works at the moment.
18:51:25 Quit bertrik (Read error: 60 (Operation timed out))
18:52:18 Part virtuoso015
18:56:27*domonoky li,e
18:56:38*domonoky likes 12-Nimbus-Regular :-)
18:56:42 Join Arathis [0] (
18:56:50 Join Xerion [0] (
18:57:27linuxstb08-Nimbus-Regular looks strange though...
18:58:12linuxstbIs "Regular" the accepted term, not "Normal" ?
18:58:21rasherDo we need the regular?
18:58:45rasherI agree to -Bold, but won't regular be assumed?
18:58:55*gevaerts votes for SS-Variant-Family.bdf. If you invert things, invert them properly
18:59:10 Join Schmogel [0] (
18:59:19rashergevaerts: Not inverting, just ordering by importance.
18:59:27linuxstbrasher: That has the advantage of "12-Nimbus.fnt" sorting first
18:59:32wpyhfirst, some background on what we're talking about on -community:
18:59:44dionoeaWhy use SS- as a prefix? That's kind of counter intuitive. People are used to have fonts listed by name (not by size).
18:59:56dionoeaof course it's easier if you want to try all the small fonts or something
18:59:59linuxstbdionoea: Read back in the logs for the discussion...
19:00:07rasherdionoea: Because size is usually the most important for a user trying to pick a font
19:00:08wpyhgevaerts: could you try appending 0xa5a5a5a5 0x5a5a5a5a twice to the end of rockboot.ebn?
19:01:19gevaertswpyh: trying now
19:02:09gevaertsDoesn't help
19:02:17gevaertsActually that doesn't surprise me
19:02:28rasherI wonder how to name the "Hebrew-*" fonts... they probably should be merged into some of the other fonts, or just deleted
19:02:41wpyhgevaerts: does sending the original M6SL.ebn work?
19:02:43gevaertsThe device gets the first 2048 bytes, so the end bytes don't matter here
19:03:18gevaertsM6SL.ebn still works
19:03:25wpyhso there's something wrong
19:03:32wpyhah, ok
19:03:49wpyhI think you should add some space after CUFD (at 0x20)
19:04:07*wpyh has to go soon
19:04:24wpyhgevaerts: try comparing the output (from the usbtracer)
19:04:30wpyhwith the original m6sl.ebn
19:04:36wpyhthere are some differences I think
19:04:42wpyhunfortunately I have to go soon
19:04:49wpyhplease post here, I'm still online
19:06:36 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
19:09:47 Join DeSnajpa_VI [0] (
19:09:50 Join globi [0] (n=Drk@2001:470:1f12:d3:0:0:0:2)
19:10:01 Part globi
19:10:02 Join Thundercloud [0] (
19:10:06rasherErk, the FONTBOUNDINGBOX isn't very reliable, it seems
19:10:10 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
19:11:00***Saving seen data "./dancer.seen"
19:11:41 Join AhtiK [0] (n=ahti@
19:14:21 Join Zarggg [0] (
19:15:04rasherWhat's the difference between ter-u14b.bdf and ter-u14v.bdf?
19:15:39pixelmadionoea: since you seem around now, may I remind you of the moving track in playlist viewer bug? ;)
19:16:01dionoeapixelma: ah ... right
19:16:18pixelmarasher: amiconn found one but thinks one of them can be dropped, unfortunately I didn't remember which
19:20:22 Nick oofus[away] is now known as oofus (
19:21:50 Quit AhtiK (Read error: 60 (Operation timed out))
19:22:38 Part DeSnajpa_VI
19:26:21 Part fragilematter
19:30:02 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
19:30:53 Quit Arathis ("Bye, bye")
19:32:39rasherpixelma: posted a preliminary list of renames to the flyspray tracker
19:32:50rasheractually it's a shellscript, but the list is there in plain sight
19:35:16rasherWhat happened in the "when" discussion?
19:35:29rasherWait until we have a themes site so we can keep working-for-3.0 themes around?
19:36:05 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:36:10rasheramiconn, pixelma: Thoughts on FS #2918 (other than it being possibly outdated)
19:36:40 Join Yondering [0] (
19:39:12pixelmarasher: sure, I'm all for it and I even thought that this was already the convention (and fixed snap in r14280)
19:39:45 Quit Horscht (Client Quit)
19:40:21pixelmathe chicago font is not in SVN anymore
19:40:33rasherpixelma: Great, I think it should be the convention indeed.
19:41:07 Join saratoga [0] (n=9803c6dd@gateway/web/cgi-irc/
19:41:10rasherWe could possibly even check this automaticly, once all fonts are the same encoding
19:41:53saratogaLear: when you get a chance, send me the source for that Tremor IMDCT you mentioned, I couldn't find anything on google for it
19:44:26Learsaratoga: Sure, just say how (tried DCC, but that didn't work).
19:45:17 Quit tvelocity ("Αποχώρησε")
19:45:26saratogaLear: sorry, using the webclient, email would be fine
19:45:36saratogacan you open a private chat?
19:47:26 Join Horscht [0] (n=Horscht@xbmc/user/horscht)
19:52:18 Join borelg_ [0] (
19:52:39borelg_but I've lost the wiki url
19:52:48borelg_con you give me it please....?
19:53:18BigBambiclick on the wiki link on every page of
19:54:17borelg_no non wiki general url
19:54:42borelg_I want to use a custum firmware on my mp3 mediacom
19:54:52BigBambiRockbox doesn't run on that
19:54:56borelg_sorry but I'm not english
19:55:02BigBambiAll the players it works on are listed on
19:55:34borelg_and there isn't a way to put rockbox on my mp3?
19:55:42 Join AhtiK [0] (
19:55:55BigBambiIt would mean a whole new port:
19:56:06borelg_yes this is the link
19:56:15borelg_but what does mean?
19:56:21borelg_what does it mean?
19:56:27BigBambiIt is instructions for how to do a new port
19:56:40borelg_and waht is it?
19:56:40BigBambiDo you have skills in electronics and low level programming?
19:56:48BigBambiwhat do you mean what is it?
19:56:49borelg_yes a bit...
19:56:52saratogamore of an outline of how one could develop a new port really
19:56:58borelg_what is it a new port?
19:57:10BigBambiIt is converting Rockbox to run on a new player
19:57:27borelg_ah ok thank you very much
19:57:30BigBambiYou need to identify the hardware, work out a way to run custom code, write drivers,....
19:57:39borelg_if I will be have a problem
19:57:46borelg_can I ask help here?
19:57:53borelg_ok thanks
19:58:02amiconnrasher: The ter-xxxv versions are the "EGA/VGA bold" versions, as opposed to the "normal bold" versions
19:58:03borelg_now I go to work on new port
19:58:08amiconnI think we can drop them
19:58:20BigBambiborelg_: Good luck!
19:58:57amiconnrasher: One final question though - shall we rename rockbox_default.bdf to its actual name?
19:59:24 Join scotty_007 [0] (
20:00:10amiconnThis happens to be "Clean", i.e. the same as clR6x8-8859-2, but with a different encoding, i.e. also a different glyph coverage
20:00:34amiconnWe should merge them, and limit the builtin font to iso8859-1 (via convbdf option)
20:01:24amiconnIdeally we find the unicode versions of those fonts which we currently only provide in limited encodings
20:02:09 Quit borelg_ ("Ex-Chat")
20:02:13 Join mirak [0] (
20:03:06 Quit mirak (SendQ exceeded)
20:03:56 Join mirak [0] (
20:04:53 Join echelon [0] (
20:05:04amiconnrasher: Regarding the renaming - I think it should be done as "svn rename" in order to keep the history
20:05:23echelonhi, what audio drivers are used for the rockbox?
20:05:53rasheramiconn: ah yes of course
20:05:54gevaertsechelon: drivers for the correct hardware...
20:06:03echelonwhat's the process?
20:06:12echeloni have an embedded system i'm trying to hack
20:06:14gevaertsWhat do you mean?
20:06:27echelonjust needed a few pointers for where i should look
20:07:31echelonit's through i2c
20:07:47saratogawhy do you want to look at rockbox drivers?
20:07:59Bagderechelon: so you have similar hw like one of the rockbox targets?
20:08:08amiconnrasher: I am wondering a bit about the names you suggested. If we want SS-Family-Variant, some names need their extra dash removed, and others are missing the variant
20:08:18 Quit XavierGr (Nick collision from services.)
20:08:18amiconnOr do we leave out the variant if there's only one?
20:08:29echelonnot necessarily, i figured the procedure for writing drivers was similar
20:08:30 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
20:08:55rasheramiconn: I don't think there'a ny point including the variant if it's "-regular" or there's only one
20:10:19amiconnI'd leave out the '00-' from 00-Starmap
20:10:27amiconnThis thing just looks odd
20:10:44rasherBut that's apparently its name :\
20:11:26amiconnRegarding #2918 - this is clearly wanted, and it's necessary to check which of those fonts still need it
20:11:48amiconnsnap is already fixed, and chicago12 is gone. That leaves 2
20:12:00 Part echelon
20:13:12rasherAnd possibly others
20:13:15amiconnThe name is "00 Starmap" btw, not "00-Starmap". That's already a deviation...
20:14:23rasherDidn't I rename it to that?
20:14:42 Quit Yondering ("Leaving")
20:14:51amiconnoh, indeed you did
20:15:00*amiconn wonders how svn will cope with the space
20:15:15rasherI should hope it'd be fine
20:15:22*amiconn suggests to do it in steps.
20:15:31amiconn(1) Drop fonts we want to drop
20:15:33rasherSome build scripts might break though
20:15:55amiconn(2) Do the unicode conversion, and merging. This includes the build system fix for sysfont
20:16:00amiconn(3) Do the rename
20:22:44gevaertswpyh: ping
20:22:47gevaertsmarkun: ping
20:23:21 Quit Nibbler (Read error: 113 (No route to host))
20:27:19 Join showstopr1 [0] (
20:27:26 Join bertrik [0] (
20:31:29amiconnrasher: Do you have an opinion on uwe_prop, proportional, and rockfont-8?
20:32:07amiconnBoth 'proportional' and 'rockfont-8' are based on uwe_prop, but there are slight differences
20:32:37amiconnrockfont-8 has the best unicode coverage of those (and is my work ...)
20:33:13Dhraakellianiirc, I switched away from uwe_prop because of its 4 being easy to mistake for 1
20:33:34amiconnuwe_prop and proportional are designed with the characters being as narrow as possible, while rockfont-8 is a bit more relaxed in order to improve readability
20:34:21rasheramiconn: no real opinion, no
20:35:17amiconnAlso, the first 2 have their baseline right at the bottom, so there's no room for descenders at all. rockfont-8 has the characters shifted up by 1 pixel to allow for somewhat better look
20:36:44 Quit XavierGr (Nick collision from services.)
20:36:53amiconnWhile the first 2 are striving for the narrowest possible characters in general, uwe_prop goofs on this both for some symbols as well as for some Latin-1 characters. Some of them are even wider than in rockfont-8
20:36:55 Join XavierGr [0] (n=xavier@rockbox/staff/XavierGr)
20:37:03amiconnproportional fixes some of them, but not all
20:37:52 Nick oofus is now known as oofus[away] (
20:38:11rasherSeems like at least one of them could be removed
20:39:58amiconnAll 3 fonts were made by rockboxers. uwe_prop by Uwe Freese, proportional by Jrg Hohensohn, and rockfont-8 by me
20:40:22 Quit AhtiK (Read error: 110 (Connection timed out))
20:41:34 Join bluebrother [0] (n=dom@rockbox/staff/bluebrother)
20:42:04amiconnAlso, is there away to (automatically) check themes in the wiki for the fonts they use?
20:43:06amiconnThose we're renaming and those we're dropping because of brokenness shouldn't be a problem, but if we want to drop a functional font (Hebrew* as you suggested), it might be in use by a theme
20:45:51rasherI don't really see what purpose the Hebrew-* ones serve. They're just Schumacher-Clean :\
20:47:48amiconnTo me it looks like someone used an aritrary font as a template, and didn't both editing the properties properly
20:48:20amiconnI may be wrong though
20:49:03rasheroh, could be
20:49:11rasherI'm just looking at the source
20:51:47 Quit scotty_007 ("Leaving.")
20:52:03mcuelenaerewpyh: in what way?
20:52:15wpyhum, it just sits there
20:52:21wpyhthe lcd didn't even turn on
20:52:32 Join Thundercloud_ [0] (
20:52:39mcuelenaerehmm probably the backlighting is differently wired
20:52:55gevaertswpyh: do you have information about the .ebn headers (for m6sl) ?
20:52:58mcuelenaeredid you try running the Rockbox binary?
20:53:13wpyhmcuelenaere: running the Rockbox binary as in?
20:53:18wpyhgevaerts: um, no
20:53:26wpyhI mostly compare the outputs and guess
20:53:30mcuelenaerewpyh: compiling it, copying it as onda.bin etc
20:53:42mcuelenaerethe Rockbox bootloader
20:53:45mcuelenaerefor the vx747
20:53:46wpyhgevaerts: btw, what version did you use when you used the hardware tracer?
20:54:03gevaertswpyh: version of what? The OF?
20:54:06wpyhmcuelenaere: yes, that's what I did. I compiled the bootloader, then used the restore tool, and used it as "onda.bin"
20:54:11wpyhgevaerts: yes, the OF version
20:54:14*gevaerts doesn't remember
20:54:22wpyhgevaerts: is it an English version?
20:54:38mcuelenaerewpyh: ah so you were trying to run the Rockbox bootloader?
20:54:45wpyhmcuelenaere: yeah..
20:54:48mcuelenaereand what happens when you try to run the OF files?
20:54:57wpyhgevaerts: where did you get that OF from?
20:55:27wpyhmcuelenaere: running the original bin's will erase the nand (just like the restore process)
20:55:32gevaertsThe meizu website. The problem is that I downloaded more than one version, and I don't remember which one I traced
20:55:40mcuelenaerewpyh: ah yes, forgot about that part :)
20:55:52mcuelenaerebut when you did it last time, did it showed something on the screen?
20:55:54wpyhgevaerts: ok, I'll try to find out which one
20:56:18wpyhmcuelenaere: it's ok, now I have my videos on the micro sd card :D
20:56:57wpyhmcuelenaere: yes, the last time I tried it showed garbage on the screen (but at least it was on)
20:57:41mcuelenaerehmm weird
20:57:52gevaertswpyh: the version doesn't really matter. The header contains things like total size and maybe CPU type. I need to find out what exactly...
20:57:59mcuelenaereso probably it's eroring somewhere before backlight_init()
20:58:16mcuelenaerewpyh: do you have some time, your vx767 and a compile environment?
20:58:25wpyhmcuelenaere: yes, I do have it
20:58:28*wpyh is multitasking
20:58:36mcuelenaereI see :)
20:58:49 Quit fyrestorm (Read error: 104 (Connection reset by peer))
20:58:51wpyhgevaerts: I actually think the version matters... I'm trying to compare the sent data with the original file
20:58:56mcuelenaereok, edit firmware/target/mips/ingenic_jz47xx/system-jz4740.c
20:59:06mcuelenaereand uncomment the sti() call
20:59:13mcuelenaeredo the same in bootloader/vx747.c
20:59:23mcuelenaereeh recomment it rather
20:59:49wpyhsee, on the M3/M6TP/M6SP, the whole file size is written to the 4 bytes after CUFD
21:02:41 Join AhtiK [0] (
21:03:55wpyhmcuelenaere: ah... I seem to have modified bootloader/ondavx747.c :p
21:04:16mcuelenaereah yes, try first doing svn up
21:04:37wpyhok :)
21:04:58wpyhactually I added a "return 0;" before sti() in bootloader/ondavx747.c
21:05:20mcuelenaerethen of course, init_backlight() won't work
21:05:25mcuelenaereeh backlight_init()
21:06:58*gevaerts managed to "brick" his M6SL \o/
21:08:02 Quit Thundercloud (Read error: 110 (Connection timed out))
21:09:11wpyhmcuelenaere: yeah :p
21:09:18wpyhgevaerts: great :D
21:10:13wpyhmcuelenaere: it still doesn't work...
21:10:16wpyhI'll try reverting it
21:10:31gevaertsI just copied the first 64 butes from M6SL.ebn and filled in the correct size
21:10:49mcuelenaerewpyh: ok, then try putting backlight_init() higher
21:10:55mcuelenaereie just after main()
21:11:02***Saving seen data "./dancer.seen"
21:11:03wpyhmcuelenaere: ok
21:11:10wpyhgevaerts: so it was the header, eh?
21:11:25mcuelenaereand if that doesn't work, try putting it just after system_main() in system-jz4740.c
21:11:34mcuelenaere(that's the first C function executed after crt0.S)
21:14:05 Quit showstopr1 (Read error: 110 (Connection timed out))
21:15:26kugelDid someone delete my post in the v2 thread?
21:16:29wpyhgevaerts: ok, it's the english version of 2.004.6. please compare the two −− AFAICS, the 32-bit int right after "CUFD" is different
21:16:31gevaertskugel: apparently
21:16:31 Join Thundercloud [0] (
21:16:35linuxstbkugel: Looks like someone did. I don't know who, but those kinds of posts don't last...
21:16:40wpyhmcuelenaere: still didn't work, I'll try again
21:16:49 Join fyrestorm [0] (
21:17:14mcuelenaerewpyh: which step exactly did you try?
21:17:35wpyhI undid everything I changed, then compiled
21:17:47wpyhI'll try moving the backlight_init() call upwards
21:17:57scorche|shkugel: twas soap and he was quite right to...
21:18:12gevaertswpyh: well, yes. That's the size :)
21:18:35*_emp almost has a working port built. But needs help.
21:18:52kugelOk, I thought it would be usefull if they know, that a Fuze owner is going to help them (they don't have a Fuze owner yet, and I talked to funman, who seemed to be quite happy about that)
21:18:52SoapI did. Sorry I didn't PM. I thought it was obvious the post was ripe for deletion as it offered nothing in the development thread. In the future I will be sure to PM you if a similar situation arises.
21:19:20SoapYou are not a Fuze owner. You might be in the future. That is the difference.
21:19:22wpyhgevaerts: can you please try making a 16-byte hole right after CUFD?
21:19:30wpyhthen write the size there
21:19:31SoapSave your message for when you have the hardware in hand.
21:20:38wpyhmcuelenaere: still, nothing
21:20:44domonoky_emp: then ask...
21:20:46wpyhmaybe my utility is broken
21:21:11mcuelenaerewpyh: weird, because what crt0.S initialises is generic for the jz47xx CPU's ...
21:21:50wpyhmcuelenaere: ah, my tool is broken :(
21:22:01wpyhI wasn't using usbtool :p
21:22:11wpyhsorry :p
21:22:18wpyhlet me fix it first ;)
21:23:54gevaertswpyh: it "works" with 8 zero bytes following the size
21:24:16 Quit moos ("Rockbox rules the DAP world")
21:25:03 Join massiveH [0] (
21:25:32_empdomonoky, I'm running into two current issues. 1) the file I am pulling down keeps changing, so I cannot create a checksum that lasts more than a day, 2) the root keeps changing from rockbox-(some number) to rockbox-(someother number), which I can regex around, but makes it difficult to maintain. Any chance you could have a stable version and current version, where the current version continues development, and the stable is locked down w
21:25:32_empith patches for reliability and security?
21:26:04wpyhgevaerts: hm... just 8? it seems to have 24 "unknown" bytes after the size
21:26:22 Join jernejovc [0] (n=matej@
21:26:43amiconnAnyone have an idea about how to check the themes for the fonts used?
21:27:03gevaertswpyh: Possibly, but just clearing those 8 made it accept it
21:27:07wpyhfrom what I see from M6SL.ebn, there is a header of 32 bytes (probably the basic irqs), then an unknown 32 bytes (containing CUFD, size, and 24 unknown bytes), then the rest of the irq table
21:27:18 Part Dhraakellian ("Meddle not in the affairs of ircops, for they are (not so) subtle, and quick to anger.")
21:27:21gevaertswpyh: I'm pretty sure the 4 bytes at 0x38 are also size-related
21:28:28wpyhah, yes
21:28:29linuxstbamiconn: You mean just get some overall statistics on what fonts are used?
21:28:34wpyhthose are probably the size of the main firmware
21:28:36wpyhlet's see
21:28:49gevaertsIndeed. That should help to find out where that is
21:28:53domonoky_emp: about 1) which files do you want to checksum ? about 2) the development version is in trunk... every release has its own branch, so it depends on what you want.. getting the latest release is difficult.. true..
21:29:40wpyhgevaerts: the int at 0x24 is 1451BC, the int at 0x38 is 13D1BC, while 1451BC-13D1BC=32768
21:29:49wpyh32768 is the size of the nand loader block
21:30:12 Quit AhtiK (Connection timed out)
21:30:23_empI need to checksum the archive, so in this case: rockbox.7z; this is to help prevent against MITM
21:30:26amiconnI was listening to the devcon recordings; there's a section where we're talking about fonts (part2, starting at 1:00:00)
21:30:50wpyhthen, when it is written by the dfu tool, the int at 0x24 is overwritten by 1451CC
21:31:02wpyhgevaerts: I was talking about 2.004.5, sorry :p
21:31:08gevaertswpyh: and 32768 is also the content of 0x30. Coincidence?
21:31:14wpyhor 2.004.4 :p
21:31:18_empI need a stable archive file.
21:31:36wpyhgevaerts: nah, not a coincidence to me :p
21:31:38 Quit Thundercloud_ (Read error: 110 (Connection timed out))
21:32:33domonoky_emp: there is no stable release for rockbox at moment (onyl the very ancient 2.5 release)...
21:32:46wpyhif you look at M6SL.ebn for 2.004.6, then it's 0x1463B8-0x13E3B8=32768, and when written by the dfu tool, the int at 0x24 will become 0x1463C8 (because the data size is the file size + the dfu suffix of 16 bytes)
21:32:50amiconnJudging from that, we should drop some more fonts than those currently suggested
21:33:01_empdomonoky, any plans to create a stable archive in the future?
21:33:22wpyhgevaerts: but what about the ints at 0x28 and 0x2c?
21:33:33wpyhthey are always 0x00000002 and 0x00000014
21:33:37scorche|sh_emp: we are currently in feature freeze for the upcoming 3.0 release
21:34:21 Quit perrikwp (" ajax IRC Client")
21:35:51_empscorche!sh, yeah, I say that. I'm really interested in a stable rbutil since it pulls downs the latest stuff anyway. I don't expect users to want to build and maintain cross compilers.
21:36:15BigBambi_emp: So it is rbutil you want, and not rockbox
21:36:43_empBigBambi, yes, but the only archive I have found on the site is rockbox.7z. Is there an alternative?
21:36:53scorche|sh_emp: oh...this is about your custom build?
21:37:07domonoky_emp: thats not rbutil..
21:37:11BigBambiNo, but I can't imagine it would be very difficult to create one for each rbutil release
21:37:30BigBambi_emp: You have been confusing everyone by talking about Rockbox when you mean rbutil
21:37:36domonoky_emp: so you want a source archive of rbutils release ?
21:37:47_empdomonoky, yep; sorry.
21:38:01_empI didn't think of them separately.
21:38:02domonoky_emp: why not pull them from svn ?
21:39:41_empdomonky, let me check to see if that is a viable option in the port system. That is what I have done locally myself.
21:40:09linuxstbdomonoky: For the next release, I think a source tarball will be nice. It will also be useful for the ebuild for gentoo.
21:40:43amiconnrasher:"> seems to be broken. FF complains about a redirect loop
21:40:58domonokyjup.. and that reminds me of a bug to fix in the rbspeex Makefile.. there are conflicts with the use ARCH define :-)
21:41:42rasheramiconn: odd
21:41:46rasheramiconn: fixed now anyway
21:44:11amiconnrasher: Is there a reason why you have clR6x12, but only that single size?
21:44:47_emphere's a work in progress Makefile for the port:
21:45:06rasheramiconn: No idea, I think I just have whatever's in xorg's cvs
21:48:45*amiconn also wonders why his X11 doesn't seem to have any .bdf
21:49:00 Quit Xerion (Read error: 104 (Connection reset by peer))
21:49:18wpyhmcuelenaere: I'm still trying to figure out what I did wrong...
21:49:36mcuelenaerewpyh: is usbtool working?
21:49:45wpyhhaven't tried yet :p
21:50:23 Join Xerion [0] (
21:50:56*wpyh is using usbmon to look
21:51:40 Quit Lear ("ChatZilla 0.9.83 [Firefox 3.0.1/2008070208]")
21:53:41wpyhmcuelenaere: I forgot to say, that usbtool doesn't compile now
21:53:51mcuelenaerehow so?
21:54:00wpyhthere is an undeclared variable "size" in nand_dump() and rom_dump()
21:55:08mcuelenaerehmm weird you're right
21:56:19mcuelenaerewpyh: svn up
21:57:26 Join AhtiK [0] (
21:57:37wpyhit compiles now :)
21:59:26gevaertswpyh: could 0x40 and further be real code?
21:59:43*gevaerts probably asks silly questions
22:00:01wpyhgevaerts: I think so...
22:00:16wpyhbut what about 0x00-0x1f?
22:00:32wpyhmcuelenaere: my tool acts funny :(
22:01:23mcuelenaereyou mean yours or mine?
22:01:30wpyhit sends different data...
22:01:34wpyhmaybe something's wrong
22:01:54wpyhI'll check again
22:01:55 Quit saratoga ("CGI:IRC (EOF)")
22:04:12gevaertswpyh: the 0x14 could be an offset to the code start. It's at 0x2c, and 0x2c+0x14==0x40
22:04:55wpyhmcuelenaere: my checksum seems to be wrong. I'll try again
22:05:09wpyhgevaerts: right
22:05:26mcuelenaerewpyh: checksum?
22:05:29gevaertsI wouldn't expect such an offset to be relative though
22:05:40wpyhyeah, the vx767 checksums the VX767_1.0.dl :p
22:05:53wpyhgevaerts: yes, and what about the first 32 bytes of the file?
22:06:00wpyhit looks like a table... no?
22:06:05*gevaerts can't do all bytes at once :)
22:06:12 Nick oofus[away] is now known as oofus (
22:07:40 Join matsl [0] (
22:10:00wpyhmcuelenaere: it almost works...
22:10:10 Quit BigBambi ("Leaving")
22:11:42 Join swima [0] (
22:14:25 Part swima ("[IRSSI] -- Artemus Ward")
22:16:10 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
22:17:45wpyhmcuelenaere: I give up −− I'll try usbtool
22:19:14*gevaerts doesn't have a clue about what those first 16 bytes mean
22:21:38amiconnrasher: Seems like clR6x12 was the only one which was extended to unicode :/
22:21:55amiconnSo all we have are our ISO8859-1 and -2 versions we shoul dmerge
22:22:55amiconn(I mean for clR6x8)
22:23:31amiconnlinuxstb: Are you working on something regarding the fonts used in themes?
22:26:01wpyhmcuelenaere: usbtool works
22:26:20mcuelenaerewpyh: and does the firmware works?
22:26:24mcuelenaeredoes the backlight goes on?
22:26:34wpyhmcuelenaere: I've only tried with the OF :p
22:26:40wpyhwait 5 minutes :)
22:27:53wpyhI forgot the definition of the checksum :(
22:28:01 Quit AhtiK (Read error: 110 (Connection timed out))
22:28:51wpyhhm... seems not
22:29:40mcuelenaereseems not even if you place backlight_init() in system_main() ?
22:30:51wpyhno, I mean I didn't get the checksum wrong −− I got another thing wrong, but I don't know what it is
22:30:59*wpyh needs to calm down
22:32:21linuxstbamiconn: Sorry, no - I have to work on other (non-rockbox) things...
22:32:27bluebrotherdomonoky: do you think renaming ARCH was a good idea? What if a user explicitly wants to use a specific ARCH value?
22:32:42 Join Casainho [0] (
22:33:20domonokythen he should use RBARCH=xxx make ?
22:33:41domonokyand we really only need this ARCH thing for macosX..
22:33:41bluebrotherIMO the Makefile should cope with an external value of ARCH (which it currently doesn't)
22:34:03bluebrotherhow about if you want to cross compile on 64bit for 32bit or similar?
22:35:04*bluebrother needs to read up on the make manual
22:35:55wpyhmcuelenaere: the bins are packed for vx767
22:36:07mcuelenaerein what format?
22:36:30Casainhohello RB people :-)
22:36:37wpyhlook at the end
22:36:42wpyhit's quite difficult to explain :p
22:37:06mcuelenaereahh ok
22:37:11*bluebrother hides
22:37:12mcuelenaereit's just some kind of tar file
22:37:20mcuelenaereor concatenated file
22:37:45domonokybluebrother: the problem with the ARCH is, that it is set on some targets, where we dont want it.. and we want to set it on mac for both builds ..
22:37:46CasainhoI am trying to build a new port... I am trying to build the bootloader, I get one file for bootloader but is a 0 bytes size :-( −− can anyone guess why is a 0 bytes bootloader files? −− here is the pastbin of build time:
22:38:01wpyhit's like, it has VX767_1.0.dl appended to it
22:38:41mcuelenaereCasainho: check your lds files and see if it matches the sections you've compiled
22:38:41wpyhok, now I got the OF working, but rockboot.vx747 still doesn't work
22:38:42bluebrotherdomonoky: do we really want it to be unset on i.e. i386? The problem is that we expect it to be unset to construct the Makefile targets
22:39:02mcuelenaerewpyh: are you running linux or windows?
22:39:03BagderCasainho: make V=1 may offer some more clues
22:39:14Casainhomcuelenaere: thanks
22:39:20Bagderdoes bootloader.elf look fine?
22:39:23domonokybluebrother: yes, the ARCH define was not intended for normal crosscompiling... we built it specific for the mac problem..
22:39:59wpyhmcuelenaere: I think we are missing something in rockboot.vx747: the signature at the start (0x78563412) and the file size after that
22:40:01bluebrotherdomonoky: I know ... I did it ;-)
22:40:02wpyhI'm running linux
22:40:08*wpyh has to eat lunch soon
22:40:24mcuelenaerewpyh: I ignored that because only the original code used that AFAIK
22:40:25CasainhoBagder: bootloader elf have 35kb...
22:40:30bluebrotherstill, in that case the Makefile should cope with setting ARCH.
22:40:44domonokybluebrother: if someone really wants to crosscompile it, he shoud fix the makefile :-)
22:40:47 Join fml [0] (n=4fd3c4b2@gateway/web/cgi-irc/
22:40:48BagderCasainho: then the bootloader command or objcopy seem to be the "bad guy" here
22:40:50wpyhmcuelenaere: ok
22:40:52mcuelenaerewpyh: (original code as in what used to be 10.bin)
22:41:15mcuelenaerewpyh: can you easily see a usb device disconnect on linux?
22:41:22*bluebrother doesn't like half-baked solutions
22:41:26gevaertsmcuelenaere: dmesg
22:41:30mcuelenaereyes I know
22:41:35wpyhmcuelenaere: I don't understand what you mean by "easily see"?
22:41:35mcuelenaerebut I meant 'easily'
22:41:42bluebrothertail -f /var/log/messages :)
22:41:48wpyhwhile sleep 0.1; do dmesg; done?
22:41:50mcuelenaereWindows generates a nice bell sound
22:41:52domonokybluebrother: and at least the lib wil have the wrong name for rbutil if you use it this way...
22:42:10gevaertsmcuelenaere: you can tweak udev to do that :)
22:42:15mcuelenaerewell never mind, try fiddling with the USB register manipulation in bootloader/vx747.c
22:42:16CasainhoBagder: is there a way to see if there is code on elf file?
22:42:22mcuelenaeremove it up until you get the usb disconnect
22:42:23fmlHello. How long shold I hold down the Select button (on Sansa e200) to fire a 'long press'? My dad can't tap it quickly enough. Couldn't we introduce a setting for the long press threshold?
22:42:28BagderCasainho: readelf perhaps?
22:42:43*wpyh has to eat lunch
22:42:50bluebrotherdomonoky: well, that's part of the problem.
22:43:06wpyhmcuelenaere: I don't quite understand though.. please keep posting, I'll take a look later
22:43:23fmlAs it's done on PC where you can set the time interval between two mouse clicks that should be a double click.
22:43:41mcuelenaerewpyh: there's this line: 'REG8(USB_REG_POWER) &= ~USB_POWER_SOFTCONN;
22:43:48mcuelenaereit generates a USB disconnect
22:43:59mcuelenaerewhich is the only debugging output if you can't use the LCD
22:44:10mcuelenaeretry to use it to look where the code crashes
22:44:19mcuelenaereie moving it up and down until you find the bug
22:44:55mcuelenaerecode flow is like this: crt0.S -> system_main() in firmware/target/mips/ingenic_jz47xx/system-jz4740.c -> main() in bootloader/vx747.c
22:45:02domonokybluebrother: feel free to make this makefile nicer.. i dont really know a bettter solution without changing more of the Makefile (and without braking other OSs) ..:-)
22:45:18CasainhoBagder: here the readelf -a :
22:46:12BagderCasainho: but what does your bootloader command do?
22:46:21 Quit miepchen^schlaf ()
22:46:33Bagder"boottool" in the configure script
22:46:47Casainhowell, the bootlaoder C file just flash a LED using a timer....
22:46:59mcuelenaereCasainho: in tools/configure
22:47:04BagderCasainho: run make V=1 please
22:47:11Bagderthen check the output what commands that are used
22:47:33Bagderthe elf is converted to bin with objdump
22:47:46Bagderthen the bin is converted to a bootloader by a "boottool" of your choice
22:48:19Bagdersomewhere around there it goes wrong in your build
22:48:43Bagderas usual, if you can provide a patch against the rockbox svn we could help out...
22:48:58Casainhohmmm - boottool, don't know what is this....
22:49:14BagderCasainho: then read tools/configure again
22:49:21Bagderand search for "boottool"
22:49:22CasainhoBagder: I don't know how to make patchs...
22:49:30mcuelenaereCasainho: svn diff
22:49:39BagderCasainho: I figure this is a good time to learn!
22:51:14 Join ompaul_ [0] (n=ompaul@gnewsense/friend/ompaul)
22:52:07Casainhookok - I must learn about SVN - I was waiting to use to host the port I am trying to do... - because now we are 2 persons working on it... but there is problems with for us :-(
22:52:17Casainhohere the configure: #boottool="cp " #??
22:52:19Casainho bootoutput="rockbox.playerlittle_bootloader"
22:52:32kugelCasainho: before "svn diff" you need to "svn add" the files you create
22:52:39BagderCasainho: you'll find that using svn will be a lot easier, since then not only 2 people can help out, but hundreds of rockbox hackers
22:52:54BagderCasainho: and the make V=1 output shows that used as... ?
22:52:54 Quit fml ("CGI:IRC (EOF)")
22:53:11Casainhoyes, yes, I want to use SVN but I need to find a free server...
22:53:23Bagdercheck out rockbox
22:53:27Bagdermake your changes
22:53:30Bagderpost the patches
22:53:42Casainhoactualy I am already using svn for other project, we are 4 people working there :-)
22:53:44Bagderwell, at least I think so
22:54:19Bagderyou're only creating barriers by not doing the development like all the other rockbox hackers do it
22:54:22CasainhoBagder: okok, I will try to make the patch −− but, what it that boottool?
22:54:38Bagderok a third time:
22:54:45BagderCasainho: and the make V=1 output shows that used as... ?
22:55:03Bagdermake V=1 is your friend
22:55:07Bagdermake V=1 can be helpful
22:55:44Bagderboottool as I described already above, is what converts a bin to a bootloader for your target
22:55:58Bagderin your case, I don't think you need to do anything with the bin file
22:56:18Bagderso it could just do cp
22:56:22Bagderor possibly even mv
22:56:26 Quit ompaul (Read error: 60 (Operation timed out))
22:57:23 Join LambdaCalculus37 [0] (
22:58:01CasainhoBagder: I send the output of make V=1 to a text file and I can't find the word "boottool" there...
22:58:15Bagderof course not
22:58:21Bagderboottool is set in configure
22:58:29*kugel facepalms
22:58:38Bagderand then configure puts that tool in the generated Makefile
22:58:40Casainhoso I will uncoment the boottool on configure to be the "cp" command :-)
22:59:55 Join dabujo [0] (
23:00:15Casainhodid that and did run configure and make all, the same result, bootloader file with 0 bytes
23:00:32Bagderthen show us the make V=1 output
23:03:33Casainhook, here:
23:04:10Bagderso bootloader.bin is zero bytes too?
23:04:50bluebrotherIMO it doesn't make sense to use rockbox.blah if it's in fact a bin file ...
23:05:19Bagderbluebrother: true, but configure is made to kind of assume that so it is easiest to go with that for starters
23:05:46bluebrothercan't you just make configure to use "bin" as extension for the result?
23:05:48Casainhojust saw now that bootloader.bin is also 0 bytes...
23:06:05BagderCasainho: then is your next research area
23:06:12 Nick oofus is now known as oofus[away] (
23:06:24mcuelenaerewpyh: I'll read the logs for your results
23:06:28 Quit mcuelenaere ("Zzzz")
23:06:51Bagderbluebrother: yes that's indeed possible, and when Casainho gets further we can bug him on such details :-)
23:07:08CasainhoBagder: bootloader.mpa here:
23:07:34Bagderoh there's no code
23:07:50bluebrother.text is empty
23:08:41Casainhooh, I must be missing something...
23:09:44linuxstbCasainho: Have you actually written a bootloader? i.e. a .c file in bootloader/ and added it to bootloader/SOURCES ?
23:09:59Casainholinuxstb: yes, I did....
23:10:09CasainhoI am trying to make the patch now...
23:10:30linuxstbI can't see it being compiled in your "make V=1" output, but maybe I missed it...
23:11:05***Saving seen data "./dancer.seen"
23:12:11BagderCasainho: and your specifies the crt object file in STARTUP() ?
23:13:10CasainhoBagder: ??? ehe - I think no... just did copy flash LED code crt0.S and lds file, as bootloader files...
23:13:24 Join tvelocity [0] (n=tony@
23:13:32Bagderyou need to make a in similar style like the other rockbox ones
23:14:22 Join einhirn [0] (
23:14:22Bagderbut do work on creating a patch, it'll be a win in the longer run
23:14:31Bagderand make sure you can repeat the procedure easily
23:14:34 Quit domonoky (Read error: 104 (Connection reset by peer))
23:14:38Casainhookok :-)
23:15:13kugelSlasheri: ping
23:15:36 Join n17ikh [0] (
23:16:30 Join miepchen^schlaf [0] (
23:21:22 Nick oofus[away] is now known as oofus (
23:21:47amiconnrasher: Your pixel size values are quite a bit off...
23:22:18rasheramiconn: Yeah, I based it losely on FONTBOUNDINGBOX, and some of them are quite obviously wrong, I'm sure more of them are subtly wrong
23:22:31*amiconn is building a list
23:22:56CasainhoBagder: I should do a diff, diff to base or a diff to head?
23:23:56Bagderto head I guess
23:25:56 Quit pondlife (Read error: 110 (Connection timed out))
23:28:26 Join Mathiasdm [0] (
23:30:07 Quit bertrik ("Leaving")
23:32:12 Quit Schmogel (Read error: 104 (Connection reset by peer))
23:32:46CasainhoBagder: nothing did happen −− what should I see? in file patch file?
23:33:06CasainhoI am using a GUI svn program, the "RapidSVN"...
23:33:12BagderI've never used that
23:34:03kugelCasainho: just execute "svn diff > myfile.diff" in the root directory
23:34:14kugelfrom the terminal
23:34:28kugelor shell or whatever you call it
23:34:31Bagderand remember to 'svn add' all new files first
23:35:46Casainhowell, there are a few files over few directories... I must do svn add for each one?
23:36:23kugelCasainho: you can do "svn -R add" directories
23:36:45kugelwith -R being the recursive option. adding files already version'd doesn't matter
23:36:58 Join nuonguy [0] (
23:38:30 Join AhtiK [0] (
23:38:30bluebrotherwell, you can also svn diff > myfile.diff; diff -u /dev/null my/added/file >> myfile.diff
23:38:53*bluebrother sees recursive adds and shudders
23:40:47bluebrotherrecursively adding should be avoided, unless you know exactly what you're doing (and checking twice before diff / ci) or you might face major hurts ...
23:41:15 Quit ompaul_ (Client Quit)
23:45:04wpyhmcuelenaere: if I treat the device as a vx747 (using vx747's bin files), then it will disconnect (which should means that it runs our code)
23:45:20wpyhmcuelenaere: if I treat it as a vx767 though, it doesn't
23:46:46 Quit matsl (Remote closed the connection)
23:48:29Casainhookok, here is the diff file :-)
23:49:58linuxstbCasainho: You don't include rockboxplayerlittle.c in bootloader/SOURCES
23:50:16linuxstbPlus you appear to have some unrelated changes in that diff - have you done a "svn update" recently?
23:51:01 Join safetydan [0] (n=evertond@rockbox/developer/safetydan)
23:51:34Casainholinuxstb: maybe didt'nt the svn update... I were working on other svn GUI program, I made there...
23:51:43linuxstbPlus what are the files in bootloader/rockboxplayerlittle-bootstrap/ - are they needed?
23:52:18linuxstbYou should really only have one new .c file in bootloader/ with all other changes in firmware/
23:53:45 Join num1 [0] (n=brian@unaffiliated/num1)
23:54:04Casainholinuxstb: okok, all that files on bootloader can later be removed...
23:54:07num1what is Super bass?
23:54:17num1and what is MDB?
23:54:36linuxstbCasainho: If you want others to look at your patch, it would be much easier if you moved them out of the way now - they make the patch much harder to read.
23:55:00linuxstbYou also appear to have made a mistake on the first line of tools/configure - adding the word "little" before #!/bin/sh
23:55:01bluebrotheryou should really clean up that registers define mess and unrelated stuff. It's a pain to look through that diff
23:55:17amiconnrasher: Hmm, seems our naming scheme was a bit too simplistic :/
23:55:34amiconnThe 'Fixed' family comes both in weight and density variants
23:56:16 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
23:56:55*linuxstb thought that "svn diff" would make a diff between the current code and the revision of the svn checkout, so wonders why there are all these reversals
23:57:07linuxstbCasainho: So you used "RapidSVN" to create that diff?
23:57:26*linuxstb suspects "diff to head" was wrong, and "diff" was the right option
23:57:41Casainholinuxstb: no, I did "svn add *"
23:58:03*bluebrother didn't even understood the question about "head" and "base" −− what's "base" on a svn checkout?
23:58:18Casainho"diff > rbplayer.diff"
23:58:28linuxstbI'm not sure about "base", but I'm guessing "head" means a diff between the (modified) checkout and the latest revision

Previous day | Next day