Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2011-07-28

00:08:22 Join TheLemonMan [0] (
00:13:35 Quit mshathlonxp ()
00:18:32 Join krazykit [0] (~krazykit@
00:19:51 Join merbanan [0] (
00:21:45 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
00:22:34 Nick dfkt_ is now known as dfkt (~dfkt@unaffiliated/dfkt)
00:22:43 Quit efyx (Remote host closed the connection)
00:24:37*rasher wonders how FS #10057 isn't a bug
00:24:38fs-bluebot album art fails to display (bugs, closed)
00:24:58rasherHaving to have album art for every artist is pretty silly
00:26:55rasherhm, nevermind
00:27:07 Quit kevku (Quit: KVIrc 4.0.4 Insomnia
00:27:17rasherStill think it's a decent improvement
00:27:52AlexPrasher: Especially as embedded doesn't work in pictureflow, so the suggested work around a) makes you use embedded and b) doesn't work in all places
00:28:13rasherI did miss a part of it
00:28:14AlexPAnd still does nothing to make tracks you bought jut work
00:28:33rasherI thought Rockbox wasn't using album artist at all
00:28:46rasherRather the problem is that it *always* uses it
00:29:06AlexPyeah, but one then the other would seem to be sensible
00:29:11AlexPas you suggested
00:30:16***Saving seen data "./dancer.seen"
00:30:27rasherThere's even a patch!
00:30:45AlexPI don't quite understand it being closed
00:30:53AlexPAnd embedded isn't an answer
00:31:15 Join alexbobp [0] (~alex@
00:31:34rasherI mean, it's quite possible (likely, even) that the patch doesn't apply, but it's not like it'd be any work to re-do it
00:33:19saratogaFS #11231 is definitely a regression
00:33:20fs-bluebot Skipping and stuttering in low bitrate .WMA files (bugs, unconfirmed)
00:34:39saratogamy original test driver for wma can properly handle VLC init on that file (but can't actually decode the file because it doesn't support some other features)
00:35:02saratogasomewhere between now and 2007 someone broke something
00:38:44rasherWell that should narrow it down
00:41:25*rasher suggest we just throw out the desc field in languages
00:41:29rasherit's practically useless
00:42:21 Quit ender` (Quit: I went to a general store. They wouldn't let me buy anything specifically.)
00:42:47saratogawtf i can't compile a sim for clean svn check out of r19504
00:42:58saratogadid somethign change since then that i'm forgetting?
00:48:13 Quit gbl08ma (Quit: Saindo)
00:48:30saratogahuh its make -j that breaks, plain make works
00:48:36rasherAh yes
00:51:04rasherHm, I think I had this information at one point when we were doing the binsize graphs, but we ended up just always using -j 1
00:51:42gevaertssaratoga: the old makefile setup didn't handle -j well
00:51:58saratogaweird i used to use it
00:52:04saratogaalthough i had fewer processors in those days
00:53:39saratogar25783 "works" although it still doesn't decode the file right
00:53:47saratogait does at least get past init without corrupting itself
00:55:20saratogaits weird how ssvn up randomly freezes on some files while flying by other at an incredible rate
00:56:19 Quit ReimuHakurei (Read error: Connection reset by peer)
00:56:23 Join ReimuHakurei [0] (
00:59:30 Join Scromple [0] (
01:00:53 Quit leavittx (Read error: Operation timed out)
01:01:47saratoga27591 works
01:01:51saratogabah have to go
01:01:56saratogawill track this down later
01:06:52 Join Unknown_Monkey [0] (
01:07:25Unknown_Monkeyhey has anyone ever installed rockbox on a ipod nano 2g in linux
01:09:08 Quit Unknown_Monkey (Client Quit)
01:18:08 Quit mudd1 (Quit: Ex-Chat)
01:34:09 Join lebellium_ [0] (
01:34:39 Quit lebellium (Ping timeout: 246 seconds)
01:34:50 Nick lebellium_ is now known as lebellium (
01:40:37 Join sideral [0] (~sideral@rockbox/developer/sideral)
01:46:04 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
02:01:48 Quit Rob2222 (Read error: Connection reset by peer)
02:02:03 Join Rob2222 [0] (
02:02:33 Join Guinness [0] (
02:02:41 Join _jhMikeS_ [0] (
02:02:41 Quit _jhMikeS_ (Changing host)
02:02:41 Join _jhMikeS_ [0] (~jethead71@rockbox/developer/jhMikeS)
02:02:41 Quit jhMikeS (Disconnected by services)
02:02:42 Nick _jhMikeS_ is now known as jhMikeS (~jethead71@rockbox/developer/jhMikeS)
02:04:56 Quit fdinel (Ping timeout: 255 seconds)
02:05:23 Quit Guinness` (Ping timeout: 255 seconds)
02:07:07 Quit TheLemonMan (Quit: Ex-Chat)
02:10:01 Quit d3remark (Ping timeout: 250 seconds)
02:15:40 Quit liar (Ping timeout: 258 seconds)
02:20:56 Join fdinel [0] (
02:30:17***Saving seen data "./dancer.seen"
02:32:19 Quit lebellium (Quit: ChatZilla 0.9.87 [Firefox 6.0/20110721152715])
02:38:20JdGordonrasher: the hueristic for finding the external AA for a track is pretty simple, if you come up with a better one to catch more files then we can probably get it commited easily
02:39:09rasherJdGordon: Well I did. It was in the task Buschel closed
02:40:32JdGordonrepoen it and commit the patch?
02:40:35JdGordonlooks fine
02:43:37rasherShould probably switch the order though
02:44:02rashersince Artist-Album is the most likely case to succeed
02:44:28JdGordonyou know the way I tihnk would be best? parse a text file which lets the user have 100% control over the filename format
02:45:00JdGordonuse the skin tags to keep it simple (duh :) )
02:45:08JdGordonthen they could orxder it however they want
02:45:10 Quit mc2739 (Quit: leaving)
02:45:16JdGordonone line per file scan
02:46:46 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
02:47:23 Quit Thra11__ (Remote host closed the connection)
02:51:43*sideral found and hopefully fixed a buffer overrun in load_tagcache
02:51:49 Join Topy44 [0] (
02:52:38 Quit T44 (Ping timeout: 255 seconds)
02:58:58 Quit evilnick (Read error: Connection reset by peer)
03:06:00JdGordonhmm... OK actually using the new playlist catalogue stuff... the WPS context menu is a bit wierd
03:06:21JdGordon"Playlist" there (which should probably be changed to "current playlist" gives you the options to save the current playlist
03:06:43JdGordon"Playlist catalogue" there adds the *current track* to a catalogue playlist... not the whole playlist
03:07:33JdGordonnot sure what to do about that
03:16:16 Quit Topy44 (Read error: Connection reset by peer)
03:24:58 Join mystica555_ [0] (~mike@
03:36:49Lloreansoap: Actually, I really like the idea of a chapter of the manual wholly dedicated to listing all the OF features, and where to find them in Rockbox. Too bad that's almost as much work as writing a whole new manual, for each device (if not more)
03:47:41 Join Topy44 [0] (
03:51:20 Quit sideral (Ping timeout: 276 seconds)
04:04:02DEBUGEOF from server (Connection reset by peer) (snapshot: netstuff.c line 545)
04:04:02***Saving seen data "./dancer.seen"
04:04:04***Started Dancer V4.16
04:04:04***Connected to on port 6667
04:04:04***Logfile for #rockbox started
04:04:06Mode"logbot :+i" by logbot
04:04:10***Server message 501: 'logbot :Unknown MODE flag'
04:04:11 Join logbot [0] (
04:04:11 Join Topy44 [0] (
04:04:11 Join mystica555_ [0] (~mike@
04:04:11 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
04:04:11 Join fdinel [0] (
04:04:11 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
04:04:11 Join Guinness [0] (
04:04:11 Join Rob2222 [0] (
04:04:11 Join Scromple [0] (
04:04:11 Join ReimuHakurei [0] (
04:04:11 Join alexbobp [0] (~alex@
04:04:11 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
04:04:11 Join merbanan [0] (
04:04:11 Join krazykit [0] (~krazykit@
04:04:11 Join bieber [0] (
04:04:11 Join Horschti [0] (~Horscht@xbmc/user/horscht)
04:04:11 Join Zarggg [0] (
04:04:11 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
04:04:11 Join alouette [0] (
04:04:11 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
04:04:11 Join [Saint] [0] (
04:04:11 Join n17ikh [0] (
04:04:11 Join user890104 [0] (~Venci@
04:04:11 Join tmzt_ [0] (
04:04:11 Join SagaciousKJB [0] (
04:04:11 Join soap [0] (~soap@rockbox/staff/soap)
04:04:11 Join pjm0616 [0] (~user@
04:04:11 Join parafin [0] (
04:04:11 Join fs-bluebot [0] (
04:04:11 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
04:04:11 Join utanapischti [0] (
04:04:11 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
04:04:11 Join bluefoxx [0] (
04:04:11 Join D-Strife89 [0] (
04:04:11 Join mystica555 [0] (
04:04:11 Join pixelma [0] (quassel@rockbox/staff/pixelma)
04:04:11 Join amiconn [0] (quassel@rockbox/developer/amiconn)
04:04:11 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
04:04:11 Join Hadaka [0] (
04:04:11 Join skx` [0] (
04:04:11 Join eintopf [0] (
04:04:11 Join kugel [0] (~kugel@rockbox/developer/kugel)
04:04:11 Join saratoga [0] (9803ec71@gateway/web/freenode/ip.
04:04:11 Join FOAD [0] (~dok@
04:04:11 Join jfc [0] (
04:04:11 Join belak [0] (belak@subtle/user/belak)
04:04:11 Join swilde [0] (
04:04:11 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
04:04:11 Join pireau [0] (1000@pdpc/supporter/student/pireau)
04:04:11 Join Farthen [0] (
04:04:11 Join Elfish [0] (amba@2a01:4f8:100:90a1:abc:abc:abc:abc)
04:04:11 Join simonlnu [0] (JXdfdzRo1P@unaffiliated/simonrvn)
04:04:11 Join AlexP [0] (~alex@rockbox/staff/AlexP)
04:04:11 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
04:04:11 Join ender| [0] (
04:04:11 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
04:04:11 Join FoH [0] (
04:04:11 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
04:04:11 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
04:04:11 Join fyre^OS [0] (
04:04:11 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
04:04:11 Join K3yn5t4 [0] (
04:04:11 Join Bagder [241] (~daniel@rockbox/developer/bagder)
04:04:11 Join dre [0] (dre@gateway/shell/
04:04:11 Join ukleinek [0] (
04:04:11 Join bthomson [0] (
04:04:11 Join mikroflops [0] (
04:04:11 Join yosafbridge [0] (
04:04:11 Join Staphylo [0] (
04:04:11 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
04:04:11 Join niekie [0] (~niek@CAcert/Assurer/niekie)
04:04:11 Join aevin [0] (eivindsy@unaffiliated/aevin)
04:04:11 Join Xerion [0] (
04:04:11 Join knittl [0] (~knittl@unaffiliated/knittl)
04:04:11 Join Espreon [0] (~espreon@wesnoth/developer/espreon)
04:04:11 Join ack` [0] (
04:04:11 Join guymann [0] (
04:04:11 Join Utchy [0] (
04:04:11 Join preglow [0] (thomj@rockbox/developer/preglow)
04:04:11 Join scorche [0] (~scorche@rockbox/administrator/scorche)
04:04:11 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
04:04:11 Join neferty [0] (~andor@
04:04:11 Join [fred] [0] (
04:04:11 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
04:04:11 Join rasher [0] (~rasher@rockbox/developer/rasher)
04:04:11 Join Vimk [0] (
04:04:11 Join bzed [0] (
04:04:11 Join crwl [0] (
04:04:11 Join TBCOOL [0] (
04:04:11 Join ranmachan [0] (ranma@2a01:4f8:130:9321::2)
04:04:11 Join avacore [0] (
04:04:11 Join Torne [0] (~torne@rockbox/developer/Torne)
04:04:11 Join jae [0] (
04:04:11 Join iq_ [0] (
04:04:11 Join ThomasAH [0] (
04:04:11 Join balintx [0] (
04:04:11 Join simabeis [0] (
04:04:11 Join kkit|sh [0] (
04:04:11 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
04:04:11 Join CIA-14 [0] (
04:04:11 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
04:04:11 Join zu [0] (
04:04:11 Join Rondom [0] (~rondom@2a01:488:66:1000:b24d:4f2f:0:1)
04:04:11 Join @ChanServ [0] (ChanServ@services.)
04:04:11 Join maraz [0] (
04:04:11 Join desowin [0] (~desowin@atheme/member/desowin)
04:04:11 Join feisar-_ [0] (
04:04:11 Join scorche|sh [0] (~scorche@rockbox/administrator/scorche)
04:04:11 Join ved [0] (
04:04:11 Join Galois [0] (
04:04:11 Join Barahir [0] (
04:04:11 Join rudi_s [0] (
04:05:11 Join Zambezi [0] (Zulu@
04:18:45 Quit mystica555_ (Ping timeout: 258 seconds)
04:22:47 Join Keripo [0] (
04:42:20 Join Toes [0] (~Toes@unaffiliated/toes/x-000000001)
04:43:16 Join mystica555_ [0] (~mike@
04:51:18 Quit pixelma (Disconnected by services)
04:51:20 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:51:23 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:51:33 Quit amiconn (Disconnected by services)
04:51:34 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:51:52 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
05:05:29 Quit mystica555_ (Ping timeout: 276 seconds)
05:07:03 Quit GeekShadow (Quit: The cake is a lie !)
05:13:22 Join factor [0] (~factor@
05:19:47 Join mystica555_ [0] (~mike@
05:28:58 Quit mystica555_ (Ping timeout: 250 seconds)
05:37:40 Join mystica555_ [0] (~mike@
05:41:54 Quit Horschti (Quit: Verlassend)
05:47:02 Join Rob2223 [0] (
05:49:35 Join Strife1989 [0] (
05:50:32 Quit Rob2222 (Ping timeout: 252 seconds)
05:53:04 Quit D-Strife89 (Ping timeout: 260 seconds)
06:04:07***Saving seen data "./dancer.seen"
06:11:55 Quit simonlnu (Remote host closed the connection)
06:30:09 Quit Scromple (Ping timeout: 264 seconds)
06:36:47 Quit mystica555_ (Ping timeout: 255 seconds)
06:37:07 Join mystica555_ [0] (~mike@
06:40:39 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
06:45:39 Quit Topy44 (Read error: Connection reset by peer)
06:48:25 Join Topy44 [0] (
06:51:38 Join stoffel [0] (
06:54:25[Saint]HaimN: Lets hope you check logs...
06:54:56[Saint]Regarding your PM, no, not really...I really don't have the time to take on another project presently.
06:55:04[Saint]However, I am happy to mentor you.
06:56:46[Saint]If, by some fluke I get through what I want to in the next <arbitrary_time_period>, and within that time you are still no further my attitude may be changed. But at this point I really need to do the final push for my RaaA theme.
06:57:30[Saint]*theme(s) [I have 3 ports to complete still (~2.25ish really ;))]
06:57:45JdGordon[Saint]: really you shuold get it into svn as it is now and work on updates once it is in
06:59:34[Saint]WHat do we do regarding the font? Convert it on demand with convttf (each font size increment needs a different command to convttf to achieve a consistent result), or just don't bother having a .ttf source in there and full the converted .fnt as needed?
07:00:04[Saint]*s/and full/and pull/
07:01:31[Saint]I'd like to convert from .ttf as needed, but convttf is such a "moody" tool that each font conversion needs to be rewieved in the debug output and increment/decrement added to ascent/descent to achieve the correct height.
07:01:39JdGordonwe/you/someone fixes convttf so it isnt stupid
07:01:51JdGordonat the very least file a bug
07:02:35[Saint]Its impossible to make an 18, 36 or 48 (all those sizes are used) without half a dozen extra flags passed to convttf per font.
07:02:44[Saint]its always +/- 1~2px
07:03:42[Saint]I've had a look at the convttf sources, but its a bit overwhelming for me. I'm not sure /what/ the f**k is happening in several places.
07:05:17[Saint]just doing "./convttf -p fontoutputheight fontname.ttf" *works*, but you can't guarantee that fontoutputheight is actually what you passed to it.
07:05:26[Saint]98% of the time it isn't ;)
07:05:42JdGordonwithout really looking into it that much (i.e about 2 min), I'm guessing you should be able to force a hieght in the code fairly easily in the main loop around "for(row=0; row < glyph_height; row++)"
07:06:03JdGordonjust vertically center/whatever the row
07:06:51[Saint]I might have a better look this evening, after not having battled with the tool for a few hours beforehand to blind me with rage. :P
07:08:55JdGordonI still dont entirely undersand your issue... file a proper bug or quite whinging
07:11:29[Saint]with a theme you need a font to be the exact height you want it, telling convttf to make a font of a specific height was an extremely high chance of making the font (sometimes quite drastically) highor or lower than the expected output size, convbdf does the same thing. So if you need a font of <height X> you need to be able to review the output in debug and add ascent/descent to it.
07:11:35[Saint]What's not to understand there?
07:12:04JdGordonare either of those tools supposed to resize fonts?
07:13:08[Saint]Yes. ./convttf/bdf outputheight fontname.ttf/bdf" where outputheight is the required font size.
07:13:27JdGordonobviosuly it isnt working... FILE A BUG
07:13:52JdGordon+bug the relevant people who actually coded it
07:21:39 Quit Keripo (Quit: Leaving.)
07:28:06 Quit mystica555_ (Ping timeout: 258 seconds)
07:30:20 Join Keripo [0] (
07:33:09 Quit stoffel (Remote host closed the connection)
07:36:01 Join mudd1 [0] (
07:42:04[Saint]JdGordon ...what d'ya think about a "soft lock" for the theme, like the guy in the Android thread is (kinda, I think he wants a setting but I _should_ be able to do it with a touch area) talking about.
07:42:13[Saint]It'd be a lot of work, worth it?
07:47:44 Join mystica555_ [0] (
07:53:24 Quit mystica555_ (Ping timeout: 260 seconds)
07:58:05 Join JdGord [0] (
07:58:21JdGordTouch softlock is simple to implement
07:58:21 Join mystica555_ [0] (~mike@
07:58:42JdGordAlso simple for themers to fuck up and make it unlockable
07:59:49[Saint]That sounds like you're not confident I can pull it off ;)
08:00:40 Join Buschel [0] (
08:00:40JdGordIt would need two touch regions (maybe only 1) and a conditioal for any touch (maybe already there) to show a splash?
08:01:21JdGordWe could of couse do exactly like the current player swlock and splash internally
08:01:47JdGordYou can also implement softlock with skin vars in svn
08:02:40 Join simonlnu [0] (e75CfdMLdA@unaffiliated/simonrvn)
08:02:49JdGordHmm, maybe not, that might need a way to disable a region outside of conditional viewports
08:04:10***Saving seen data "./dancer.seen"
08:04:23[Saint]I was just thinking two conditional viewports with touch areas that set a var of 1/0 and when pressed and toggling the other touch areas in the theme by checking that value, also using the 1/0 variable to check whether to display the locking or unlocking viewport.
08:04:33JdGordIll have a look tonight maybe
08:05:22JdGordI'm happy to add the softlock code to make it work simply
08:05:25Buschelrasher, AlexP, JdGord: I do not mind if you want to change the algorithm to find AA (FS #10057). Just re-open it, if you like to work on this again. I just had the impression that there was nobody really caring about this (no updates since >2 years) and that this was not really a bug but a limitation in our code. And I still think embedded AA is an alternative −− even though you may not like it
08:05:26fs-bluebot album art fails to display (bugs, closed)
08:05:49[Saint]unless I wanted to have a shitload of conditional viewports, it would be easier to pull the touch areas out of the viewports they are in and into the fullscreen viewport for all the items that are staticall displayed to make it easier to toggle them.
08:07:23JdGordWould we want to allow *any* regions to fire when locked?
08:07:40[Saint]I could do it with skin vars, but, if it is possible to add "softlock lock/unlock" touch areas that toggled all other touch areas except itself when true it would be easier.
08:08:04[Saint]and yes, you'd want to be able to unlock it with a touch region, I assume ;)
08:08:54JdGordWelll obviously lock will work when locked :)
08:09:24JdGordI was thin.king about adding an option to make certain specific regions also work when locked
08:09:27[Saint]something similar to the way the "mute" touch region works, but instead of muting/un-muting, nullifying all other touch areas barring itself when locked and activating them again when unlucked.
08:09:47JdGordThe whole patch should be half a dozen lines
08:10:37[Saint]If its possible to do that with a single touch area it would be awesome, the code needed to do it with skin variables would be far far more that ~12 lines :D
08:11:05 Join antil33t [0] (
08:11:18[Saint]I won't be heartbroken if it can't be done though, it'll just mean it will take me longer and carving up the current theme a bit.
08:12:59JdGordonanother way to do it would be having a region change the party-mode setting
08:13:05 Quit JdGord (Quit: Bye)
08:13:26[Saint]you could still change the volume them too,
08:14:01JdGordonauto-lock after a timeout is not something we should add by default
08:14:18JdGordonim also not sure how that could be implemented
08:15:11 Join einhirn [0] (
08:15:41[Saint]If I'm going to do a softlock, whether its done with an addition to the touch regions in the skin code, or I do it with intricate skin variables and conditional touch areas, I want to block all touch areas apart from the lock/unlock area itself (obviously).
08:16:20[Saint]Still being able to change volume via the HW keys if present.
08:17:53[Saint]that way you could still use HW keys to back out of the WPS (if they're present), and still be able to use the touchscreen in those menus/screens and then re-enter the WPS and have it still be "soft locked".
08:18:06JdGordonthe way I'll implement it is with a single 'lock' region which will disable all non-'lock' regions when locked, but i will also add an option to make a region workable while locked
08:18:22JdGordonso you could still get to the menu, or popups while locked
08:19:43[Saint]Oh, another flag similar (and working alongside, hopefully) the * and & touch area flags.
08:19:56JdGordonthough really, cabbie should be in feature freeze by now and all phones turn the screen off for a lock
08:20:00[Saint]*s/the/to the/
08:20:50[Saint]You think my cabbieis geting too much of a case of the feature creeps?
08:20:56[Saint]*cabbie is
08:21:09 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 3.6.18/20110614230723])
08:25:53 Join Scromple [0] (
08:27:33JdGordon[Saint]: yes and no, more that I tinhk you will never get it commited if you dont freeze it
08:27:40JdGordonby all means add more once it is in svn though
08:28:22 Quit n17ikh (Ping timeout: 255 seconds)
08:29:07 Quit fyre^OS (Quit: quit)
08:32:46JdGordon[Saint]: really more important is putting high res images into svn :)
08:32:54*JdGordon also wants a 1024*600 cabbie already :)
08:35:57[Saint]In most cases I reused existing cabbie graphics from the source or slightly modified them, when I do the first tablet port my intention is to make all the images very large-scale SVGs and replace all the other cabbie images in SVN with images scaled from these master images.
08:36:27[Saint]the color/brightness variation in svn cabbie through the different ports is terrible.
08:36:46 Quit Scromple (Quit: Leaving)
08:36:49[Saint]even I can see it and I'm colorblind :D
08:42:43JdGordonthis work really should be in a seperate git repo untill we move to git, and then we import it like we would external libraries
08:42:50 Join T44 [0] (
08:43:13 Quit Topy44 (Ping timeout: 255 seconds)
08:44:32 Quit mystica555_ (Ping timeout: 264 seconds)
08:48:17pixelma[Saint]: greyscale and monochrome icons shouldn't be done from an SVG though. Scaling from the large colour version looked bad enough and they needed carefully setting each pixel
08:49:05[Saint]pixelma: Yes, I was going to ammend that statement to say just the colour targets specifically. Sorry.
08:49:49[Saint]If any greyscale or monochrome work needs doing, rest assured (or not :P) I will call on you.
08:50:03pixelmaalso - even though I see the colour differences too, I always thought that they may have been created on purpose to even out differences in the actual targets' displays
08:51:14[Saint]I considered that, but I came to the conclusion it was poor imaging software mixing the colours up during scaling.
08:51:48pixelmaon the other hand... I'm not sure if any of the original port authors had more than one target
08:52:08[Saint]Even if it is to make up for color differences, I still think it would be better for the source images to be made from the one "master" image set. Brightness can be used on the targets if it is too dark/light.
08:55:01pixelmadisplays can look differently in actual colours too but then that can also happen on the same target. If I have the opportuntity sometime I should take a picture of amiconn's and mine c200s showing the same theme with default settings or so. There's quite some difference
08:55:23pixelmaadmittedly, the c200 display is crap anyway
08:55:44 Join n17ikh [0] (
08:56:23 Join ender` [0] (
08:56:43 Join webguest44 [0] (
08:57:30 Join mystica555_ [0] (~mike@
09:02:12 Quit webguest44 (Quit: CGI:IRC (EOF))
09:13:01 Join petur [0] (~petur@rockbox/developer/petur)
09:13:07 Quit mystica555_ (Read error: Connection reset by peer)
09:18:59 Join mystica555_ [0] (~mike@
09:22:40 Quit mudd1 (Ping timeout: 276 seconds)
09:24:34 Join d3remark [0] (
09:29:00 Quit utanapischti (Quit: WeeChat 0.3.5)
09:29:27 Join sasquatch [0] (
09:33:39 Quit mystica555_ (Ping timeout: 240 seconds)
09:37:06 Join wodz [0] (
09:39:16 Join mystica555_ [0] (~mike@
09:40:39 Quit d3remark (Ping timeout: 240 seconds)
09:41:09 Join d3remark [0] (
09:50:32 Quit mystica555_ (Ping timeout: 240 seconds)
09:52:18wodzukleinek: I see you are working on rk27xx, what's the outcome of your work?
10:01:47ukleinekwodz: nice you're pinging me
10:02:09wodzI have short brake in my vacation :-)
10:02:16ukleinekwodz: nothing valueable yet. I started to look what you did so far
10:02:21 Quit bluebrother (Disconnected by services)
10:02:22 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
10:02:31ukleinekwodz: not everything that you describe works for me
10:02:42wodzwhat specificaly?
10:02:43ukleinekwodz: e.g. I cannot open the system partition
10:02:59wodzhow are you opening it?
10:03:33ukleinekwhen I touch RKLDSET.FLW the device ends up in DFU mode
10:03:50ukleinekI have an odys neo x66 btw
10:03:54wodzthat's pretty cool I would say :-)
10:04:12***Saving seen data "./dancer.seen"
10:04:15ukleinekwhat is cool?
10:04:25wodzIf you can go stright to DFU
10:04:40 Quit fs-bluebot (Ping timeout: 255 seconds)
10:04:43ukleinekyour usbtool works for me, too
10:05:09wodzyou mean custom scsi command to open partition?
10:05:09 Quit neferty (Ping timeout: 240 seconds)
10:05:14ukleinekwodz: yes
10:05:32wodznice, it didn't work as expected for me
10:05:38ukleinekthe custom scsi command to enter dfu though
10:05:48 Join FoolOnHill [0] (
10:05:51 Join fs-bluebot [0] (
10:05:54ukleinekI cannot open systemmode via the scsi command
10:06:13ukleinek... and I wonder if the magics are right
10:06:31wodzto judge why is that you have to dump nand bootloader and dissasemble
10:06:52ukleinekother than that I started to examine the Rock27Boot.bin I found on the net
10:07:03wodzthere are a few of thouse
10:07:14 Join neferty [0] (~andor@
10:07:20ukleinekup to now I checked rkboot_s1.bin (which are all equal for the bins I found)
10:07:37wodzrkboot_s1.bin is sdarm controller init code
10:07:37ukleinekyou make some things different in your bootloader
10:08:00ukleinekyeah, there is a NAND command in it and modifying the PLL, too
10:08:05wodzmy stage1 is plain wrong but I had not much time to work on that
10:08:49 Quit FoH (Ping timeout: 276 seconds)
10:09:12wodzI pretty much understand what rkboot_s1.bin do but I had no time formalize this into working code
10:09:28wodzIt's easier to use this blob directly anyway
10:09:40ukleinekI wonder if the sdram initialization is robust though
10:09:51wodzwhat you mean?
10:09:57ukleinekthey seem to try a few modes and use the first that seems to work.
10:10:26wodzthey try a few different sdram organizations. That's correct
10:10:40ukleinekI remember the hardware guys at the company I used to work for, did the same and had some issues
10:10:51ukleinek(on a different SOC, though)
10:10:54wodzAnyway I never saw something different than 8x16 in this rk27xx devices
10:13:04wodzAre you interested in some particular area of development?
10:13:21ukleinekwodz: yeah, getting my x66 up :-)
10:14:22wodzIf you have wiki account please write down a small section about your device in rk27xx port page
10:14:56ukleinekwodz: ok
10:15:11wodzdoes lcd work as advertised?
10:15:26ukleinekdidn't try yet to put custom code on the player
10:16:45ukleinekdidn't found out yet how to test it, and I wasn't sure if I this would destroy the original firmware
10:17:58wodzIf you can enter DFU you can upload code without touching nand so OF is safe
10:18:14ukleinekwodz: I wonder if the names of the devices you have would be sensible to add, too.
10:19:04ukleinekwodz: depends on the code I upload, I guess :-)
10:19:11wodzI have one device which is "noname" but seems to strict reference design for rk27xx.
10:19:36ukleinekwhere do the pictures come from?
10:20:50wodzthis were on ebay auctions. I contacted sellers to confirm this devices are rk27xx based
10:21:29ukleinekwodz: which documentation do you have?
10:22:00ukleinekif have a "RK27xx Product Datasheet A1.1"
10:22:28ukleinekthis doesn't cover nand and rtc though, the latter should be easy enough though.
10:25:07 Join kramer3d_ [0] (~kramer@unaffiliated/kramer3d)
10:28:14ukleinekdo I just need to give the bootloader to the rk27load tool
10:28:50 Join mystica555_ [0] (~mike@
10:28:58 Quit d3remark (Ping timeout: 255 seconds)
10:29:35wodzyou have to provide 2 or 3 images. First is sdram init, Second is loader (or custom code if you fit within codesize limit), Third is without size restrictions
10:29:58JdGordon[Saint]: <- untested but pretty surprised if it doesnt work, use 'lock' for the region to lock/unlock. if the *first* character of the action is ^ the region will work when locked (not needed for lock).. it *must* be fre the * or & symbol
10:30:21ukleinekso I stick to rkboot_s1.bin and rkboot_s2.bin and use your code as third image
10:30:33wodzthat wouldn't work
10:30:59wodzrk27load is designed to work with our implementation of second stage loader
10:31:34*ukleinek rereads the README
10:33:20 Quit mystica555_ (Ping timeout: 240 seconds)
10:33:47[Saint]JdGordon: so..."%T(0,0,100,100,^&volup)" for exapmple?
10:34:15[Saint]Ok, cool...I'll have a play with it this evening.
10:35:17JdGordonif you have a better suggestion for the ^ or label let me know
10:35:19 Quit [Saint] (Quit: Imagination is for turbo-nerds who can't handle how kick-butt reality is. I'm a kick-butt reality master! I would rather die, than be imaginative. I mean that.)
10:36:12 Join sideral [0] (~sideral@rockbox/developer/sideral)
10:37:46 Join [Saint] [0] (~Saint]
10:40:40 Join mystica555_ [0] (~mike@
10:42:23 Quit neferty (Ping timeout: 260 seconds)
10:43:54 Quit mystica555_ (Client Quit)
10:44:18 Join neferty [0] (~andor@
10:47:49 Quit antil33t ()
10:48:48 Join leavittx [0] (~lev@
10:48:48 Quit neferty (Ping timeout: 260 seconds)
10:49:04sideralSlasheri, kugel: I found the bug that crashed my simulator after the DB refresh yesterday (where the display_formats got trashed)
10:49:10 Quit swilde (Remote host closed the connection)
10:50:10 Quit wodz (Quit: Leaving)
10:50:21 Join mystica555_ [0] (~mike@
10:50:25 Join neferty [0] (~andor@
10:51:23 Join TheLemonMan [0] (
10:51:56sideralUnfortunately it's not the crash in retrieve() I see occasionally on target. And I don't know yet whether the fix also fixes the filename trashing.
10:52:41 Join swilde [0] (
10:54:39 Quit user890104 (Read error: Connection reset by peer)
10:54:39 Join user890104_ [0] (~Venci@
10:57:45 Quit user890104_ (Read error: Connection reset by peer)
10:57:48 Join user890104 [0] (~Venci@
10:58:57 Quit sideral (Disconnected by services)
10:58:58 Join sideral1 [0] (~sideral@rockbox/developer/sideral)
10:59:16 Nick sideral1 is now known as sideral (~sideral@rockbox/developer/sideral)
11:00:23 Join z180 [0] (
11:00:58sideralIf you'd like to have a look:
11:01:12sideralI'll commit this next week unless I run into any issues with it
11:07:47 Quit mystica555_ (Ping timeout: 255 seconds)
11:08:24 Quit sideral (Remote host closed the connection)
11:08:49z180the nano 3g doesnt apear to be very different technically
11:09:07 Join sideral [0] (~sideral@rockbox/developer/sideral)
11:12:17[Saint]z180: The Nano2G and the iPod Classic share quite a few similarities if I'm not mistaken.
11:12:41[Saint]*Nano3G, rather.
11:13:04[Saint]I believe it was a Nano3G exploit that got the foot in the door for the Classic port.
11:14:54 Join lebellium [0] (
11:18:39 Quit sideral (Remote host closed the connection)
11:20:42 Join sideral [0] (~sideral@rockbox/developer/sideral)
11:28:37 Join pamaury [0] (
11:28:38 Quit pamaury (Changing host)
11:28:38 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
11:36:38 Join wodz [0] (
11:38:32z180a port would be easier than to nano 5g or my Sony a840?
11:39:31TorneNobody can say until they try both :)
11:40:58wodzI believe the main barier on nano3g is FTL code
11:41:23wodzbesides that nobody seem to work on the port
11:43:08Tornethe nano 3g was not a very popular device
11:44:57 Join wodz_ [0] (
11:46:18AlexPbuschel (logs) I have nothing against embedded album art - it is just that a) To have to add it isn't a solution when the alternative solution of remove the empty but present album artist tags was rejected as having to fiddle with tags and b) it doesn't work with pictureflow which some people care about.
11:48:09 Quit wodz (Ping timeout: 255 seconds)
11:53:42 Quit SagaciousKJB (Ping timeout: 240 seconds)
12:04:16***Saving seen data "./dancer.seen"
12:05:19 Join wodz__ [0] (
12:06:33 Join lebellium_ [0] (
12:07:48 Quit sideral (Ping timeout: 246 seconds)
12:07:56 Join SagaciousKJB [0] (
12:08:51 Quit wodz_ (Ping timeout: 255 seconds)
12:08:54 Quit lebellium (Ping timeout: 250 seconds)
12:08:58 Nick lebellium_ is now known as lebellium (
12:20:01 Join sideral [0] (~sideral@rockbox/developer/sideral)
12:23:42 Join jordan` [0] (~bip@2001:660:3302:2826:225:90ff:fe20:d9a8)
12:26:30 Quit jordan` (Client Quit)
12:26:47 Join jordan` [0] (~bip@2001:660:3302:2826:225:90ff:fe20:d9a8)
12:40:22 Quit sideral (Quit: Leaving.)
12:40:53 Join sideral [0] (~sideral@rockbox/developer/sideral)
12:42:10 Join liar [0] (
12:44:16 Join casainho [0] (
12:51:45JdGordondoesnt the sdl raaa put themes/etc in ~/.config/
12:53:36 Join japc [0] (~japc@
12:55:33 Quit petur (Quit: *plop*)
12:57:30JdGordonah bloody hell... /usr/local/share/rockbox
12:59:49JdGordonwhy is −−debugwps so bloody useless? :p
12:59:54JdGordonsomeone really should fix that!
13:00:01*JdGordon glares at JdGordon
13:02:28JdGordon[Saint]: does the fact you cant use *,& with the volume or bar touch tags cause problems?
13:02:36*JdGordon guesses not as he hasnt been bugged about it
13:05:39 Join wodz [0] (
13:05:55[Saint]Not really, I can't imagine using volume on a hold area....someone might want to do it, though.
13:06:18[Saint]"none" not using & or * is an issue, though.
13:06:34[Saint]Not a huge one, but one nonetheless.
13:06:46JdGordonI'm also thinking that those 2 chars are stupid and i want to replace them with proper options
13:07:07[Saint]ESKINBREAK! :P
13:07:14JdGordonI know... I know...
13:07:33JdGordonthis is an easy one, and i dont tinhk many people have touched touchscreen stuff
13:07:44 Quit wodz__ (Ping timeout: 240 seconds)
13:07:45JdGordonmaybe that backwards compat hack stuff is a good idea :p
13:07:47[Saint]This is true.
13:08:08JdGordonright now the order of those chars is important and they are meaningless
13:08:12JdGordonlock first
13:09:41[Saint]Yeah, sorry...things at home have been....interesting. Haven't had a chance to play.
13:09:48JdGordonyeah, thats fine
13:09:54[Saint]the patch *applies*, but that's as far as I've gotten.
13:10:25JdGordondo I make the current "hold?" tag work or need to add a new one?
13:10:31JdGordonI'd like to not add another one
13:13:04JdGordonhehe i cant unlock :p
13:24:45 Quit Keripo (Quit: Leaving.)
13:24:56[Saint]You'd need to add a new one, no?
13:25:13[Saint]"hold" is entirely different, and applies to all screens.
13:25:36 Join Keripo [0] (
13:26:03CIA-14New commit by jdgordon (r30218): skin engine softlock support for touchscreens: ...
13:26:13JdGordonoh dammit... i forgot part of the message
13:26:32JdGordon%mh returns 't' if the touchscreen is locked so you can know the difference
13:26:34[Saint]we also need to be careful that this tag only is relevant for the wps, or the screen it was invoked from (if it works in the .sbs too)
13:26:45JdGordonshouldnt be a problem as i cant imagine any touch targets will also have a button lock
13:26:58JdGordonno we dont
13:26:58[Saint]you wouyldn't want to have to go back to the wps just to unlock the screen if you navigated out of the WPS
13:27:08JdGordonoh yes, it is per-skin
13:27:15[Saint]Ah, right.
13:27:37JdGordonI was going to put it on a shared data struct but decided against that
13:29:02[Saint]so what does the hold condition look like now? %?mh<true hw|true software|false>"?
13:29:23 Join robin0800 [0] (
13:29:50JdGordonno, no change
13:30:15JdGordonyou have to use %?if(%mh, =, "t")<> if you care about hardware/touch differences
13:40:17JdGordonwheres the ciabot gone?
13:40:38JdGordonstill building?!
13:41:48 Quit Keripo (Quit: Leaving.)
13:44:22*[Saint] kicks CIA-14
13:45:09 Join wodz_ [0] (
13:46:32 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
13:47:35 Quit z180 (Ping timeout: 258 seconds)
13:47:58 Quit wodz (Ping timeout: 258 seconds)
13:49:58 Join lorenzo92 [0] (~chatzilla@
13:50:15 Quit robin0800 (Quit: Leaving)
13:53:25 Quit lorenzo92 (Client Quit)
14:04:11 Join FoH [0] (
14:04:20***Saving seen data "./dancer.seen"
14:05:45 Quit FoolOnHill (Ping timeout: 246 seconds)
14:11:10 Quit FoH (Quit: ¡ooʇ ‘ǝןdoǝd ǝɹɐ sʇɐq)
14:12:40 Quit sideral (Quit: Leaving.)
14:14:15*[Saint] wonders where the build result went.
14:14:29[Saint]surely the round is finished...?
14:15:14*[Saint] kicks CIA-14 again
14:16:44 Quit alouette (Ping timeout: 240 seconds)
14:16:57JdGordonserver must have died or something
14:17:05JdGordonI thought it was supposed ot commit suicide after 20min?
14:17:30[Saint]I was under a similar impression.
14:21:29JdGordonTorne: for git migration it would be great if the tools stuff built into a new dir instead of tools to make .ignore easier
14:22:32TorneA new dir where?
14:22:38Torneand i'm not sure why it makes ignoring it easier
14:22:55Tornethe point of it not going into your build dir is they are not target-dependent, so there's no need to compile them again
14:24:35*Torne is not opposing changing it, just saying it's not obvious why it's related to the migration.
14:24:38gevaertsJdGordon: that's not really a VCS issue
14:24:51JdGordonoh, i know
14:25:04Tornethe .gitignore is just going to be a copy of what's svn:ignore'd now
14:25:20JdGordonwhat is svn:ignored?
14:25:29JdGordoni got a svn checkout to try and find out but couldnt
14:25:30 Join wodz [0] (
14:25:31Tornethe tools builds :)
14:25:40Tornevarious utils builds
14:25:45Torneand directories called build*
14:25:47JdGordonhardcoded filenames?
14:26:05JdGordonbuild* is good enough for me then, i wanted to make sure build/ wouldnt be in git status :)
14:26:08Torneif you have a svn checkout, go to the root and do svn pg -R svn:ignore
14:26:11Torneit will show them all
14:26:45TorneYeah. In the root we ignore build* and output
14:26:58Torneand then we ignore various target files under tools/ and utils/
14:27:04 Join kramer3d__ [0] (
14:27:04Tornethat's basically it.
14:27:53JdGordonthe android folder probably has some folders to skip also
14:27:54 Quit wodz_ (Ping timeout: 255 seconds)
14:28:07TorneThere aren't any set there
14:28:30TorneAnyway, if there's something not ignored that should be feel free to update svn:ignore properties for it
14:30:09Tornei haven't converted it yet, so it's not a problem to add more :)
14:30:35 Quit kramer3d_ (Ping timeout: 250 seconds)
14:33:24*JdGordon cries at doing a manual change
14:34:19 Join sideral [0] (~sideral@rockbox/developer/sideral)
14:35:04 Quit krazykit (Ping timeout: 240 seconds)
14:35:49 Join wodz_ [0] (
14:36:12 Join d3remark [0] (
14:37:49JdGordon[Saint]: ping?
14:37:50 Quit kramer3d__ (Ping timeout: 258 seconds)
14:38:37 Quit wodz (Ping timeout: 260 seconds)
14:38:51 Quit factor (Quit: Leaving)
14:38:53JdGordonhow do you feel about these optoins to replace the single chars... "allow_while_locked", "long_press", "repeat", "reverse_bar"
14:39:38[Saint]I was unaware there was a single-char flag to reverse bars...
14:40:18[Saint]also...I'm not sure those work.
14:40:41[Saint]its possible presently to do a repeating long press, how would this be worded in that scheme?
14:40:52JdGordonadd both
14:41:37 Quit sideral (Remote host closed the connection)
14:42:27 Join sideral [0] (~sideral@rockbox/developer/sideral)
14:42:32[Saint]So just "long press" is the "fire once" option, and long_press, repeat" is the "keep firing" option?
14:42:55JdGordonrepeat is the old &
14:43:01JdGordonlong_press is *
14:43:27JdGordonyou dont need both if you havnt in the past
14:43:45[Saint]Aha, right. Yeah...I can dig it.
14:44:05JdGordonok, so im almost ready to commit, so if you dont like those option names now is the time to speak up :)
14:44:11JdGordondont want to break thems again
14:44:51[Saint]I'd go for "repeat_press" so as to not mix it up with repeat.
14:45:12[Saint]the rest are fine.
14:46:59 Quit d3remark (Ping timeout: 258 seconds)
14:49:09[Saint]So, I take it these new touch flags require comma seperation? ie. "%T(0,0,-,-,allow_while_locked,repeat_press,volup)"?
14:49:28JdGordonaction must go before the options
14:49:42[Saint]Oh, that's different...
14:49:58[Saint]but still comma seperated flags?
14:50:08JdGordonyeah, same as every other tag with params
14:50:45[Saint]right, that's cleaner...brings them more into line with all the other optional param use cases.
14:51:52 Quit sideral (Remote host closed the connection)
14:52:59 Join sideral [0] (~sideral@rockbox/developer/sideral)
14:53:24CIA-14New commit by jdgordon (r30219): Change the way the %Tl() (touch region) tag is done to remove dodgey 1-char settings. ...
14:53:38JdGordonto the wiki!
14:56:03*[Saint] removes an l from the commit message...
14:56:40[Saint]%Tl is (touched last) and doesn't accept any of those params.
14:57:53[Saint]Oh...btw JdGordon these new params work with "none" now? Or is thata "try it and see"? ;)
14:58:41[Saint]as it stands prior to this commit (and perhaps still) none did not accept (or did not respect) * or &
14:59:54CIA-14r30218 build result: 1 errors, 0 warnings (jdgordon committed)
15:00:31saratogaso the change that breaks that wma file is changing a single global variable to "static"
15:01:45JdGordon[Saint]: none will parse with the params, wheather they work or not is a different story
15:01:53JdGordonand doh to the commit message being wrong
15:02:06amiconnTorne: Imo it would be a good idea to build the tools in a host architecture specific sub-dir of tools/
15:02:24[Saint]JdGordon: that's ok re: none. There's no loss of functionality thee compared to prior to the commit.
15:02:44[Saint]it doesn't break anything, jsut wondering if it fixed it ;)
15:02:48amiconnThis way a working copy could be shared between build hosts (e.g. through nfs)
15:02:49saratogai thought making something static would generate a compiler error if you did something with it outside of the file, how can changing to static actually change the result of calculations?
15:03:24Tornesaratoga: that sounds like an optimiser bug or something
15:03:35 Join Poodlemastah [0] (
15:03:36Torneamiconn: possibly. someone is welcome to do that ;)
15:05:50*amiconn will probably look into this if no one beats him to it
15:06:05saratogatorne: different -O levels and different gcc versions don't seem to matter
15:06:07CIA-14r30219 build result: All green
15:06:26[Saint]yay for hugely delayed build results!
15:06:38saratogathe weird thing is that the variable isn't actually used in the calculation that fails
15:06:47saratogaso something else must be screwed up
15:07:13[Saint]23:25, commit. 00:59, build result...nice ;)
15:07:25 Join gbl08ma [0] (~gbl08ma@
15:07:46JdGordonthe previous one was 93min
15:07:53Tornesaratoga: hrm.
15:08:32sideralTorne: I'm thinking about how to migrate my git repo (w/ lotsa branches) to be based off the Gerrit one. I'm wondering if I could arrive at a Gerrit-compatible repo be rewriting my existing one with the user-name mapping you've used. Any comments or ideas?
15:08:41saratogaFWIW this is the commit that changed things to static:
15:08:54saratogaafter that change one of the huffman tables ends up containing garbage
15:09:11saratogawith predictable results on decoding
15:09:23[Saint]JdGordon: you dodn't write any actual code for "And for the setting_inc/dec/set action the setting name must follow the action name, *then* the options" did you...?
15:09:27[Saint]it'd never get used.
15:09:39saratogabah i'm going to get coffee, its too early for this
15:09:41amiconnPredictable garbage??
15:09:45[Saint]the vars can only be used with %Tl, and these params only with %T
15:10:18sideralsaratoga: there's also some IRAM attrib changes in that diff
15:10:23[Saint]and %Tl uses an identifier anyway, not a touch region action.
15:10:27[Saint]JdGordon: : ^
15:11:25*JdGordon is confused
15:11:39JdGordonmy commit messages should have been %T() not %TL(), everything else is fine
15:12:00[Saint]not regarding set/increment/decrement it isn't.
15:12:09[Saint]those are explicitely for %Tl
15:12:11JdGordonsetting_inc/dec is config settings, not skin vars
15:12:22[Saint]Oh...right. heh, whoops.
15:12:34JdGordonI was wondering why there is no action to set/inc/dec a skin var?
15:13:07[Saint]because its not needed.
15:13:23[Saint]you check if the region is touched and set/inc/dec that way.
15:13:55gevaertsI suspect the slow build may have something to do with a lack of android build clients and maybe buggy handling of that situation
15:14:02[Saint]a touch region that handled that would just give another way to do the same thing.
15:14:57 Quit Torne (Ping timeout: 276 seconds)
15:16:58JdGordon[Saint]: sure, but i tihnk the direct route is actually cleaner
15:16:58JdGordonbut not tonight
15:18:36 Join Torne [0] (~torne@rockbox/developer/Torne)
15:21:09Tornesideral: i think a graft will work better
15:21:31Torneif you filter your branch to match what i did that will fix all the SHAs but that will take fucking aeons
15:21:37[Saint]ie. "%Tl(identifier,timeout)<%vs(identifier,set,value)>" and "%T(0,0,-,-,touch_action,set,identifer,value)" are essentially the same thing.
15:21:55TorneYou can probably just update all your branches to HEAD of the old repo, then add a graft so it thinks you have already merged everything
15:22:22[Saint]the former has the advantage of being able to be used with multiple touch areas for less code by sharing touch area identifiers.
15:23:21[Saint]"%T(0,0,-,-,touch_action,set,identifer,value)" is my fictional vision of the syntax of your proposal, btw.
15:23:24sideralTorne: that would be a solution, yes. But I'd like to avoid rebasing by branches. Some of them are for archival only.
15:23:31[Saint](doen't actually exist, yet)
15:23:43sideralTorne: I don't think it takes much longer than one or two hours (did it once before)
15:23:51Torneyou don't need to rebase them
15:24:06saratogaamiconn: yes its always the same garbage, even on different machines
15:24:13TorneYou don't need to merge them at all, but if you don't you'll need a graft at each branch base
15:24:33sideralYeah that could work
15:24:41gevaertssaratoga: maybe some aliasing with other stuff elsewhere with the same name?
15:24:49saratogasideral: I bisected the contents of that commit, its the line "static uint16_t levtab_big[1336] MEM_ALIGN_ATTR;" that does it
15:25:07sideralTorne: I've done a rewrite of the whole tree before, didn't take that long.
15:25:08saratogayeah that or maybe some random buffer overflow else where
15:25:15Tornesideral: Alternatively you can just rebase the branches onto the right points on the new tree
15:25:17sideralsaratoga: I see
15:25:28Tornethis is trivial for branches with no merges :)
15:25:38Tornetrickier with merges, but still technically possible.
15:25:47TorneWell, feel free to rewrite the whole thing if you want
15:25:55gevaertssaratoga: right. I'd bet on a buffer overflow. Mapping this stuff in the address space changes
15:25:57 Quit gbl08ma (Ping timeout: 255 seconds)
15:26:05ukleinekTorne: if the trees are identical it's easy, too
15:26:19sideralTorne: yeah... still kinda tedious :) What would the rewrite script look like? did you change anything besides chopping of some roots and branches and changing the email addresses?
15:26:34TorneI moved a lot of branch heads and tags around
15:26:40Tornebut the old repo only has trunk so that doesn't matter.
15:26:50TorneAll trunk commits should be identical except their authors
15:27:18sideralOK. Could you the email-address mapping available to me?
15:27:58saratogayeah its got to be overflow, looking carefully, the end of the buffer is correct, just the first half is overwritten with something else
15:28:26sideralsaratoga: How are you looking at it −− with a sim?
15:28:48 Join FoH [0] (
15:28:55saratogasideral: yeah sim and also the codec lib test program
15:29:10sideralthen you can set a hardware watchpoint in the debugger
15:29:22Tornesideral: yep
15:29:32Tornesideral: gimme a sec, just updating the gerrit mirror while i think of it
15:29:35sideralsaratoga: that's how I found the load_tagcache buffer overrun last night :)
15:29:37Torne(and there's a new tag so i have to intervene)
15:30:01saratogai don't actually know how to use gdb, was just going to bisect the code until i find the point where that table gets overwrriten
15:30:40sideralsaratoga: this is a good time to learn it, it's rather simple. I could step you through it tonight if you wish
15:31:17Tornei can't tell you how to rewrite your tree though
15:31:22*user890104 have just submitted FS #12209 & FS #12210
15:31:22fs-bluebot iPod 6g/classic usb product id (patches, unconfirmed)
15:31:23fs-bluebot Bulgarian translation update (patches, unconfirmed)
15:31:23TorneI didn't do it by rewriting :)
15:31:25Slasherisideral: hi! you found something?
15:31:34Tornegit-svn just mapped them over for me
15:31:42sideralthanks Torne! got it
15:32:49Torneif you do do it successfully please share it :)
15:32:51sideralSlasheri: Hi there! yes, I found the cause of the display_format trashing.
15:33:09Slasherisideral: thats great!
15:33:23sideralTorne: will do!
15:34:51Slasherisideral: so tagcache_load could cause a buffer overflow? that could explain the internal tagcache structure corruption i saw earlier
15:35:39sideralyep, that's right. although typically it overflows into stuff allocated by tagtree_init, that is, the formats
15:36:01saratogas->lsp_pow_m_table1 = &vlcbuf3[0];
15:36:17saratogasvn blame
15:36:46 Quit Poodlemastah (Read error: Connection reset by peer)
15:37:20saratogaoh wait no, thats vlcbuf4 thats messed up
15:37:25sideralSlasheri: I saw the DEBUG comment you added, by crash backtrace scraped past it very closely ;)
15:37:39Slasherisideral: nice!
15:38:22 Join gbl08ma [0] (~gbl08ma@
15:41:17saratogahmm wait thats definitely wrong, just not sure why yet
15:43:17 Quit casainho (Quit: ChatZilla 0.9.87 [Firefox 4.0.1/20110518051943])
15:43:37 Nick tmzt_ is now known as tmzt (
15:44:29sideralSlasheri: unfortunately this is a different bug to the one I was hunting. My repro scenario was adding too many files at once (hence the overflow), which always yields the DB in nonready state.
15:44:55sideralSlasheri: the bugs I see happen when the commit and reload actually have succeeded
15:45:43sideralTorne: do you expect any more changes to the git history on Gerrit?
15:46:26 Join keyb_gr [0] (
15:47:02TorneEverything appears to be fine
15:47:22TorneI'm updating it when I remember (not automatically, it still requires some manual intervention)
15:47:31sideralTorne: OK, thanks!
15:47:34Tornebut I haven't changed history since it was published and I don't intend to.
15:47:41sideralSlasheri: Do you have a comment regarding the patch?
15:48:27TorneSo yeah, I wouldn't assume you can use it to actually do development
15:48:27Torneunless you want to pester me about updates a lot ;)
15:48:42sideralTorne: yeah, understood. I just thought about it last night for the purpose of getting this patch ^^ reviewed
15:51:07sideralSlasheri: I'd guess that if the tagcache header structure is corrupted, it would have to be an overrun by memory allocated before the tagcache; possibly the dircache
15:52:04Tornesideral: You can't currently upload anything for review
15:52:13TorneThe repository is entirely read only except to me
15:52:21TorneOnly the sandbox repos are open for business
15:52:30sideraloh, OK
15:52:36saratogaok stupid question
15:53:12saratoga&vlcbuf3[VLCBUF3SIZE] where vlcbuf3 is defined at int16_t vlcbuf3[VLCBUF3SIZE][2]
15:53:39saratogadoes that pointer point to 2*VLCBUF3SIZE bytes into the array or 4*VLCBUF3SIZE bytes
15:53:47Tornesideral: Technically I could allow reviews
15:53:55Tornethen you'd ahve to submit it manually via svn
15:54:07Torneas long as you preserved the Commit-Id line gerrit would notice it had happened and close the review.
15:54:25Tornei just haven't gotten that far yet :0
15:55:22 Join wodz__ [0] (
15:56:54sideralI can't preserve the commit-ID line because git-svn adds it only when actually committing
15:56:57 Quit swilde (Ping timeout: 252 seconds)
15:57:10sideralnot when I post stuff for review
15:58:07 Quit wodz_ (Ping timeout: 260 seconds)
15:58:21sideralsaratoga: Code like this should be forbidden :)
15:58:26Tornei think you are mistaken..
15:58:29Tornenot the git-svn-id
15:58:34TorneGerrit's commit-id hook
15:58:44TorneIt's just a line of text, it's nothing to do with git-svn
15:58:46sideralah, I see what you mean. Right
15:58:48Torneor git, in fact.
15:59:09TorneYeah. gerrit is perfectly happy with things getting pushed directly to branches even if a review exists, it just closes the review as [MERGED]
15:59:18Torneas long as it can identify which review it was, by commit-id
15:59:23Tornechange-id even
15:59:27Tornethat's why i am confusing you (and myself)
15:59:39Tornecommit != change
15:59:39saratogathis is the commit that did it:
16:00:09sideralTorne: OK
16:00:17saratoga s->lsp_pow_m_table2 = &vlcbuf3[VLCBUF3SIZE];
16:00:27Tornesideral: anyway, hopefulyl the swedes will be around to bug soon and I can start moving gerrit over to its real home
16:00:29saratogaapparently that points to the end of VLCbuf3 rather then the middle as I would have expected
16:00:42Tornethen eventually we can kill svn :)
16:00:43saratoga(since its a 2D array)
16:01:00sideralsaratoga: I'd guess that pointer points beyond the end of the entire array
16:01:26saratogasideral, well its a 2D array of dimension VLCBUF3SIZE x 2
16:01:45sideralyeah, saw that.
16:02:02saratogaso I think the assumption was that it would give *vlcbuf3+VLCBUF3SIZE === vlcbuf3[VLCBUF3SIZE/2][0]
16:02:08saratogaactually i'm not sure why that doesn't work
16:02:21sideralbut this goes at the end of the first dimension, irrespective of the second dimension
16:02:35sideralThat why this code should be forbidden :)
16:03:15saratoga2D array in c always impress me because they manage to be even more confusing then raw pointer arithmetic
16:03:30sideralyou can have as many int16s in the second dimension as you like
16:03:38*[Saint] wonders if the build output could be named "<target> <revision> <date>.ext" instead of just so that building into the same directory made it easier to roll back to your previous builds by not overwriting the last one.
16:04:24***Saving seen data "./dancer.seen"
16:04:39[Saint]...for all builds, not just binaries.
16:05:12[Saint]Hmmm...wouldn't work with voicefiles, though.
16:06:05sideralsaratoga: The C FAQ explains this rather well:
16:06:22sideralthe first given dimension is the outer dimension
16:06:47sideralit's consistent with the inside-out type construction rules of C
16:08:00saratogayes I understand that
16:08:48[Saint]...whoops, "<target> <revision> <date> <time>.extension", rather.
16:09:01sideralso, as array indices start counting at 0 and end at dimension-1, &vlcbuf3[VLCBUF3SIZE] points beyond the (outer) array
16:09:05[Saint]to me that seems much saner than "rockbox.extension"
16:09:10saratogai just assumed that since the array is stored [[0 1] [0 1] ...] in memory that giving a single array index would treat it as a single 1D array that was composed of all the sub elements concatinated together
16:10:24sideralsaratoga: that's seems to be what the author of that patch thought as well, so you're at least not alone :)
16:11:46sideralThe type of vlcbuf3[VLCBUF3SIZE] is int16_t type[2]
16:12:35CIA-14New commit by saratoga (r30220): Fix a bug introduced in r27463 that caused the line spectral pair look up tables to overwrite the noise coding VLC tables. Should fix decoding of ...
16:14:46 Join evilnick [0] (
16:14:46 Quit evilnick (Changing host)
16:14:46 Join evilnick [0] (~evilnick@rockbox/staff/evilnick)
16:15:04 Quit wodz__ (Quit: Leaving)
16:15:06saratogawhile that fixed the crash on this WMA file, the decoder still doesn't work since apparently there are more bugs :(
16:16:28CIA-14r30220 build result: All green
16:21:07 Quit leavittx (Ping timeout: 240 seconds)
16:39:18 Quit liar (Ping timeout: 258 seconds)
16:44:28LloreanjhMikeS: According to a post in the forum, the Gigabeat F panic might be related to the backlight code, somehow.
16:56:57 Join d3remark [0] (
17:05:28 Quit Xerion (Quit: )
17:09:17 Join stoffel [0] (
17:14:18 Join piotrekm [0] (~piotrek@unaffiliated/piotrekm)
17:14:40 Quit piotrekm (Client Quit)
17:16:01 Join Xerion [0] (
17:16:50 Join Thra11 [0] (~thrall@
17:21:09 Join liar [0] (
17:26:50 Quit liar (Ping timeout: 258 seconds)
17:28:12 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
17:30:16 Quit saratoga (Changing host)
17:30:16 Join saratoga [0] (9803ec71@rockbox/developer/saratoga)
17:34:30 Quit ruskie (Excess Flood)
17:34:37 Join ruskie [0] (
17:34:38 Quit ruskie (Changing host)
17:34:38 Join ruskie [0] (ruskie@sourcemage/mage/ruskie)
17:34:40 Quit Thra11 (Ping timeout: 246 seconds)
17:35:33 Join Thra11 [0] (
17:45:13 Join alouette [0] (
17:54:01 Join liar [0] (
18:04:21CIA-14New commit by rasher (r30221): New language: Basque by Asier Arsuaga (FS #12205)
18:04:23CIA-14New commit by rasher (r30222): Add Basque to translate site
18:04:25***Saving seen data "./dancer.seen"
18:08:07CIA-14r30221 build result: All green
18:22:44 Join Poodlemastah [0] (
18:24:08 Quit gbl08ma (Ping timeout: 255 seconds)
18:28:23 Nick SagaciousKJB is now known as GigaBrick (
18:31:15 Quit sideral (Quit: Leaving.)
18:37:38 Join gbl08ma [0] (
18:46:11*[Saint] has converted 36 individual .ttf fonts (each needing its own slight tweaks) and is now quite ready to punch convttf in the data-balls.
18:46:25 Join mudd1 [0] (
18:47:14amiconnsaratoga: Maybe that table is one element short? ;)
18:47:41 Join kevku [0] (x@2001:470:28:773:babe:feed:dead:beef)
18:47:56[Saint]10pt~50pt in 2pt increments, in bold, regular and light should be sufficient for most of the coming future ports of this theme, though. Enough to not need to go through this for a while, anyway.
18:48:11[Saint]Unless we start seeing *super* high res devices.
18:48:16 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
18:48:30saratogaamiconn: which table?
18:49:14 Quit gbl08ma (Quit: Saindo)
19:00:41 Join robin0800 [0] (
19:01:31amiconn[Saint]: Do you have something to test? Specifically for 800*480 (landscape)
19:05:16[Saint]Not for landscape, no. Just 480x800 portrait and 240x320 portrait.
19:06:14[Saint]Thanks for reminding me I need to make a landscape 800x480 for you slideout keyboard guys, though.
19:07:46[Saint]I'm currently working with the 320x480 port, and then (as I promised a certain skin engine developer ;)) I'll be doing a 1280x800 tablet port but the 800x480 landscape should just be a case of repositioning elements already developed.
19:09:36[Saint]someone made a landscape 800x480 iiuc...but its approximately the same quality as the other RaaA themes, so...basically unusable IMO
19:09:59[Saint]well, usable. but not terribly functional.
19:10:36*[Saint] needs to find a recruit to mentor...
19:11:48saratogawhy do you think its one element short? IIRC I computed those sizes back during the original port of the codec
19:13:04bertrikI remember people talking about the ipod nano 1g USB transfer being quite slow on linux and really slow on windows
19:13:30bertrikOn linux, there is a quirk enabled for the ipod nano 1g USB VID/PID, I wonder if that's related to slowness
19:13:50bertrikapparently READ_CAPACITY returns an incorrect value
19:16:12[Saint]my Nano1Gs *crawl* with transfer on Windows.
19:16:22[Saint]speed is vastly better on Ubuntu
19:16:38[Saint]Windows == 10~20kbps :-S
19:16:51[Saint]Ubuntu == 1~1.2MB/s
19:20:48bertrikWe can try to give it a different USB VID/PID (one without a known quirk) and see if it gets better/different
19:21:13bertrikI guess rockbox does the right thing regarding READ_CAPACITY
19:23:37amiconn[Saint]: There is 800*480 cabbie, but it
19:23:56amiconn's rather buggy
19:25:08 Quit simonlnu (Ping timeout: 240 seconds)
19:25:13 Join fdinel [0] (
19:25:27[Saint]Yeah, afaik it uses the same base code as the other RaaA themes, but with modified positioning values.
19:25:34[Saint]So it shares the same quirks.
19:25:56[Saint]ie. ffwd/rwd popping up if the screen is touched...*anywhere*
19:26:48 Quit Poodlemastah (Quit: ChatZilla 0.9.87 [Pale Moon 5.0-x64/20110624114230])
19:26:56 Join simonlnu [0] (JH6yXSOv92@unaffiliated/simonrvn)
19:27:28amiconnRegarding shuffle and repeat - personally I don't need them at all. Both are always off on my devices; if I want to play something shuffled, I use "insert shuffled" only
19:27:54amiconnYes, that popup bug is definitely there
19:30:08 Join Horscht [0] (
19:30:09 Quit Horscht (Changing host)
19:30:09 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:30:29 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
19:36:56 Join Strife89 [0] (~Strife89@
19:38:17*[Saint] despises hospitals even more when he's in them but not the patient.
19:39:26bertrikI changed the ipod nano 1g USB VID/PID and it doesn't seem to make much difference on USB transfer speed on linux (about 2.5 MB/s)
19:39:38bertrikI'll try windows now
19:40:08[Saint]Its so bullshit, we've been here for ages, and the nurses keep popping in and asking silly questions about the pain the GF is in, and each different one goes off and checks what medications they are able to give the GF...but as a Dr has not reviewd her yet there are no medicines charted.
19:40:17[Saint]So all they can offer is paracetamol.
19:40:29AlexP[Saint]: Wrong channel
19:40:39[Saint] it is. ;)
19:45:17bertrikI get about 1.2 MB/s on windows 7 when writing music over USB to the rockboxed ipod nano 1g (with different USB VID/PID)
19:45:19 Quit simonlnu (Remote host closed the connection)
19:45:22 Quit d3remark (Ping timeout: 252 seconds)
19:47:52bertrik(but I haven't measured *before* the USB VID/PID change ...)
19:48:10[Saint]have you got OF and pre-V/PID change numbers?
19:48:16[Saint]oh...dammit. too slow.
19:51:00 Join Buschel [0] (
19:52:53 Quit japc (Ping timeout: 264 seconds)
19:54:26 Join simonlnu [0] (kGUM2nV1kq@unaffiliated/simonrvn)
19:54:51 Quit stoffel (Remote host closed the connection)
19:57:19 Quit ReimuHakurei (Quit: If I use this, I will disappear, and Shana-tan will remain...)
19:57:35 Join ReimuHakurei [0] (
20:04:26***Saving seen data "./dancer.seen"
20:07:49pamaurydoes rockbox support multiple audio outputs ?
20:08:10pamauryI have a mio device which has headphones + speaker and also support bluetooth headphones
20:08:57pamauryI see there is a HAVE_SPEAKER define but I don't know how the speakers works for example
20:10:29saratogalots of devices have headphone and line out
20:10:47saratogagenerally we try to detect which is actually in use and only power up that one
20:11:07saratogasome of the irivers also have digital out
20:11:43pamauryso each driver is responsible for managing this ? There is no device independent code for it ?
20:12:04saratogathe audio drivers usually handle it
20:12:15saratogasince generally its just muxing between outputs on a DAC chip
20:12:42saratogais the concern figuring out when to power up the bluetooth?
20:13:49pamaurythe concern is "is there a function to select output ?". If not, I'll have to do something special for bluetooth but it's not quite near to being working :)
20:14:06bertrikthe ipod nano 1g VID/PID doesn't seem to make much of a difference on USB write speed on windows 7
20:14:20 Quit keyb_gr (Ping timeout: 258 seconds)
20:14:48saratogausually we haven't needed to select the output, its done automatically based on whats connected
20:14:57saratogabut i'm not sure what the iriver does
20:15:50saratogaso the iriver at least has an option for it
20:17:17pamauryI think there two functions audiohw_enable_lineout and audiohw_enable_speaker
20:20:32 Join keyb_gr [0] (
20:25:12pixelmasignature spam in the forums once more by creamXX39 - already caught by akismet but no-one nuked it yet it seems
20:27:14 Join Poodlemastah [0] (
20:27:25 Quit Poodlemastah (Client Quit)
20:37:09saratogawe should just disable signatures
20:38:18 Join hskf [0] (~hskf@
20:45:00 Join deegroovey [0] (
20:46:07deegrooveyplease excuse me for bothering you, I had a quick question that probably has a quick answer
20:46:49saratogaDon't "ask to ask". If you come in and ask your question, people will see it and help if they can. If you come in and ask, "can I ask a question?", nobody will want to bother asking what the real question is and you may have just lost important assistance. The channel is open to questions and you do not need to get permission to ask them.
20:47:24saratogabasically, just ask and if someone knows they'll help you
20:48:07deegrooveyI've already installed rockbox once, successfully. But I need to reinstall the actual files. Can I just delete the rockbox files and unzip a new copy? Or do I need to run the installer util again after doing that?
20:48:47saratogayou don't need to do either
20:48:58saratogajust unzip the build you want to your player, overwriting files as you do
20:49:35deegrooveyso the firmware doesn't point to particular sectors on my ipod's tiny harddrive?
20:51:09deegrooveythat's a relief. thank you
20:52:03saratogaFWIW this is the kind of thing that usually faster to find in the manual then on IRC
20:53:18deegrooveyoh, but the reason I had to delete it first was that I originally copied the files over to my ipod using linux, and it did such a horrible job of unnecessarily fragmenting every file that rockbox lags between every menu item. but for some reason windows defrag considers those files "unmovable"
20:53:55deegrooveyso I was going to delete them altogether and copy them over using windows
20:54:40saratogafragmentation isn't going to matter to the rockbox menus
20:55:03deegrooveysaratoga, for some reason the manual won't open. It says the pdf file is damaged
20:55:27saratogayou can use the HTML version until you get your PDF viewer working
20:56:08deegrooveyoh, my pdf viewer works, just not with the manual's particular pdf file
20:56:39deegrooveyit says something about the dictionary being missing or something
20:57:10deegrooveybut bye, have to go. thanks
20:57:16 Part deegroovey ("Leaving")
21:09:45 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
21:10:17 Join wodz [0] (
21:11:32 Quit mc2739 (Ping timeout: 252 seconds)
21:13:27 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
21:18:51 Quit alouette (Quit: On me voit, on me voit plus !)
21:26:59*amiconn wonders whether bertrik has just hit the general slowness of rockbox software usb
21:27:58 Quit FoH (Quit: ¡ooʇ ‘ǝןdoǝd ǝɹɐ sʇɐq)
21:29:54 Join mystica555_ [0] (
21:30:36 Quit Strife89 (Ping timeout: 240 seconds)
21:30:57bertrikamiconn, some people get only a few 10s of kB/s rockbox USB transfer speed with windows, but this doesn't seem to affect me
21:31:40bertrikfor me, USB transfer with windows are quite a bit slower than on linux, but not that dramatic
21:32:52amiconnWindows does no write caching on removable devices, which should explain the difference
21:34:53amiconnIndependent of that, rockbox usb is significantly slower than OF usb for me on all sw usb devices where this kind of comparison is possible (i.e. excluding the beast where the OF doesn't do ums)
21:36:09*amiconn often uses OF usb or edm because of this, when transferring larger amounts of data
21:37:50bertrik[Saint], can you try changing the USB VID/PID on your ipod nano 1g and see if it makes a difference?
21:38:23bertrikChange #define USB_PRODUCT_ID 0x120a in firmware/export/config/ipodnano1g.h into #define USB_PRODUCT_ID 0x120b for example
21:39:08bertrikor perhaps something completely different, like 0x8888
21:41:26 Quit Thra11 (Quit: kthxbai)
21:49:16 Quit advcomp2019 (Ping timeout: 260 seconds)
21:50:46 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
21:54:27 Join bieber_ [0] (
21:55:35 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 3.6.18/20110614230723])
21:59:51pamaurythere is no reason for the rockbox usb stack to be particularly slow afaik
22:00:28 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
22:04:29***Saving seen data "./dancer.seen"
22:11:39 Join sideral [0] (~sideral@rockbox/developer/sideral)
22:12:00 Quit bertrik (Read error: Connection timed out)
22:12:48 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
22:23:12 Join Thra11 [0] (~thrall@
22:24:54 Quit Stummi (Quit: Bye!)
22:32:20 Quit liar (Ping timeout: 258 seconds)
22:32:52 Join liar [0] (
22:34:34 Quit bertrik (Read error: Connection timed out)
22:35:21 Join bertrik [0] (
22:35:21 Quit bertrik (Changing host)
22:35:21 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
22:39:42 Quit [Saint] (Read error: Connection reset by peer)
22:41:08 Join [Saint] [0] (
22:42:36 Quit thomasjfox (Remote host closed the connection)
22:44:10 Quit wodz (Quit: Leaving)
22:47:22 Quit bieber_ (Remote host closed the connection)
22:56:10 Quit ReimuHakurei (Quit: If I use this, I will disappear, and Shana-tan will remain...)
22:56:38 Join ReimuHakurei [0] (
22:57:58 Quit hskf (Quit: Bye)
23:16:11 Quit mystica555_ (Ping timeout: 260 seconds)
23:26:13 Join Horschti [0] (~Horscht@xbmc/user/horscht)
23:27:24 Quit Thra11 (Ping timeout: 255 seconds)
23:28:52 Quit Horscht (Ping timeout: 264 seconds)
23:29:21 Join Thra11 [0] (
23:32:38amiconnpamaury: It might be the usb stack or the storage driver - the net result is the same
23:35:03*amiconn can only really compare usb speed on pp targets (PP502x ipods, iriver h10 and c200v1)
23:35:51bertrikamiconn, what was the maximum length again for the "Building DB %d found" phrase on archosplayer?
23:36:07 Quit Thra11 (Ping timeout: 250 seconds)
23:36:24amiconnThe display is 2 lines x 11 chars
23:37:00amiconnSplashf does word wrap, which has to be taken into account
23:38:01amiconnI assume %d to be 5 digits max.
23:38:45bertrikQuite a few of the recent language updates use rather long text for this particular phrase
23:39:20amiconnThe English version fits exactly. "Building DB" is 11 chars, "99999 found" another 11
23:39:57amiconnI've seen one or two with failto fit the purpose
23:41:30amiconnIf just "found" gets cut off it's not a big loss - but the number should stay visible (that's why I introduced the shortened version in the first place) because otherwise you have no progress indicator
23:41:51pamauryamiconn: the storage driver is good enough and the usb stack is ok. It might be the usb driver. Do not understimate windows too :)
23:42:27amiconnThat won't explain why it's faster when using edm
23:43:25amiconnI can't measure atm, but iirc on my mini g2 edm is about twice as fast as rockbox usb
23:43:34pamaurywhat is edm ?
23:43:37saratogaisn't USB essentially just feeding you packets? how does that end up being faster or slower on the same hardware?
23:43:53saratogathe bootloader disk mode on ipods
23:44:01saratogaerr Apple ROM disk mode rather
23:44:02amiconnEmergency disk mode (the usb mode built into the rom)
23:44:04saratoganot our bootloader
23:44:19pamauryis it windows which is particularly slow ? or linux too ?
23:44:36bertrikamiconn, unfortunately on some language an excessively long string slipped through (basque, chinese-simp, italiano, polski, romaneste)
23:45:36amiconnchinese-simp doesn't matter at all. It's excluded altogether since it's not latin, greek or cyrillic
23:47:40amiconnpamaury: My experience is with windows. I rarely use usb with linux (the exception being my archos Player, because of no x64 windows driver)
23:47:43bertrikRegarding USB speed: on windows 7, USB write transfer is about two times slower on my ipod nano 1g (1.2 MB/s) than on linux (ubuntu 11.04)
23:48:07 Join Thra11 [0] (~thrall@
23:48:31bertrikI tried to change the USB VID/PID to avoid any usb storage quirks modes, but found no effect
23:48:32*amiconn should perform measurements when he's back home
23:50:06pamauryI should do some measurement too, but I don't have a usb capable player right now (except ongoing fuze+). If I have time I'll investigate this issue, at the time as the clip+ one
23:50:32amiconnbertrik: I kind of hoped that translators would deduce why 'database' had been abbreviated, and do the same in their language
23:51:36 Quit mudd1 (Ping timeout: 240 seconds)
23:51:51amiconnFor German this was easy - using the same two letters
23:51:52pamaurycan't there be comments on lang files ?
23:53:02AlexPpamaury: There is a description field for each phrase
23:53:10AlexPBut everyone seems to ignore it :)
23:53:31AlexPHaving said that, I don't know if the description is good here
23:54:13 Join mudd1 [0] (
23:56:26sideralSlasheri: still around?
23:57:52 Join anewuser [0] (becd8193@gateway/web/freenode/ip.
23:57:55 Part anewuser
23:58:15 Quit Thra11 (Quit: kthxbai)
23:58:33 Join Thra11 [0] (~thrall@

Previous day | Next day