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 2010-05-26

00:00:03linuxstbfunman: i.e. we're distributing an app with embedded data. Although I'm not 100% sure what the GPL says about that....
00:00:06Tornesince it seems to contain md5sums of the firmwares
00:00:09Torneer, bootloaders
00:00:23***Saving seen data "./dancer.seen"
00:01:24funmanlinuxstb: in the worst case we could download the OF from our servers in rbutil and do the patching on the person's computer
00:01:24funmansaratoga: does the guy on sansa forums answer your PMs ? (i didn't try to contact him)
00:02:14 Join Topy44 [0] (
00:02:50linuxstbTorne: That file doesn't appear to have been updated since 2007, so I'm guessing it's not needed
00:02:59Tornelinuxstb: hm
00:03:28*linuxstb looks towards bluebrother and domonoky1
00:04:16kisakfunman: if you're going to change your focus to include the nano 3g, I also have a classic 1g 160GB over here
00:04:48funmanis it similar to nano3g?
00:04:58bluebroIirc it is used for the iriver bootloaders
00:05:18kisakfunman: that's what I keep getting told
00:05:44linuxstbfunman: Yes, IIRC the 3g, classic 1g and 2g (and probably 3g) all use the same SoC.
00:05:56bluebroCan't check tonight, will try tomorrow
00:06:19linuxstbfunman: The advantage of the classic is a real disk - i.e. no FTL...
00:06:58funmanafaiu the nano2g FTL was reused in the 1st iphone
00:07:26linuxstbFrom the sounds of it, you don't want to go anywhere near the nano2g FTL.
00:08:07funmani mean, i hope TheSeven's work would work too on nano3g
00:08:41 Part Etu
00:08:50linuxstbYes, but from TheSeven's comments about it, it's not a clever system.
00:09:07linuxstbfunman: So do you own a nano3g?
00:09:23funmanmy brother does, so yes i do too :)
00:10:16 Quit bluebro (Quit: used jmIrc)
00:10:50 Quit halmi (Quit: halmi)
00:11:01TheSevenactually, i've seen a slightly improved version of that not-so-clever system on some iphones/itouches
00:11:06TheSevendon't remember which generations
00:11:51TheSevenso i guess some changes will be needed to make the ftl work on the 3g, but on the other hand, some of the really bad things might have been improved
00:12:32TheSevenit could of course be possible that the nano3g was release before those changes, and that it works mostly unmodified
00:12:56TheSevenanother improvement starting from the 3g is that it's unbrickable
00:13:20TheSeven(you can enter bootrom dfu mode by holding the reset combo for 10 seconds)
00:13:46Farthenwell, the 2g is virtually unbrickable, too. you might be forced to open it up but you can't brick it, too
00:14:02TheSeventhat dfu mode has already been exploited (and it's currently our only way to execute code in that device)
00:14:18TheSevenFarthen: actually you can easily brick it by opening it up
00:14:37TheSevenor at least make it fail shortly afterwards (hold switch etc.)
00:14:48TheSeventhose things really aren't easy to disassemble
00:15:01 Quit wincent (Ping timeout: 240 seconds)
00:15:30 Quit DerPapst (Quit: Leaving.)
00:15:34funmandisassembling is easy
00:15:47funmanmaking sure it can be reassembled is the hard part :p
00:16:35linuxstbfunman: Does your brother like his nano3g?
00:17:02funmanno, i can get it even without giving him a sansa
00:17:17funmanlast time i didn't take it because he didn't find the cable
00:17:43 Quit ender` (Quit: I think I remember an episode of MacGyver where he overthrew a violent dictator with a rubber band, 2 bottle caps, and some navel lint.)
00:18:48kisakdoes ender`'s client have a fortune cookie dispenser on it's back
00:19:06 Quit [CGL] (Remote host closed the connection)
00:21:18 Quit DataGhost (Ping timeout: 240 seconds)
00:24:09CIA-5New commit by bieber (r26292): Fixed some memory leaks in the theme editor
00:24:33TorneBagder: so, hm..
00:26:08TorneBagder: well, are you waiting on me to work out what to do?
00:26:19Bagderno, I think I'm done
00:26:33Bagdersorry, I should've said so
00:26:43TorneAh, right
00:26:53Torneoh, it's mirrored isn't it
00:27:13Bagderyeah, there's this fine delay...
00:27:46Tornethat's why i didn't know you were done ;)
00:28:20linuxstbbieber: Could you add a short README to your theme editor saying what's required to build it? Also, shouldn't main.cpp have a (C) header?
00:29:29bieberI'll ad a README shortly. It's not really fair to call it a theme editor yet, so far it's just a parser that can show you a debug view of the parse tree
00:29:44bieberI wouldn't imagine there should be a header, since it's going to be a standalone application?
00:31:45linuxstbbieber: I'm not sure what you mean - why wouldn't main.cpp need a (C) header?
00:32:38bieberOhhh, C as in copyright. You're right, I forgot to add one. I thought you meant a C-style header file
00:33:17*funman didn't understand © header as well
00:35:27bieberAre frequent commits frowned upon? i.e. will people be mad at me if I make a commit to add a README file and a (c) header?
00:36:05funmanpersonally i prefer numerous small commits to large commits
00:36:25funmanit's easier to bisect bugs, but perhaps for documentation it doesn't matter
00:36:48linuxstbAnd it's easier for others to review.
00:37:00 Quit BlakeJohnson86 (Ping timeout: 248 seconds)
00:37:31bieberMakes sense, thanks
00:39:10CIA-5New commit by bieber (r26293): Added a README and a missing copyright header to the theme editor files
00:40:21 Join Orbital_sFear [0] (~ldupin@
00:40:41 Part Orbital_sFear
00:45:33bieberWhen I start writing C++ code for the GUI app, would anyone object if I switch up capitalization style to match Qt's?
00:46:02gevaertsthe coding style we have is for C, so I wouldn't think so
00:46:11funmando you mean CamelCase?
00:46:12linuxstbbieber: I think you should follow rbutil's style, as that's set the example for C++/Qt in Rockbox.
00:46:37*linuxstb has no idea what that style is though...
00:47:10 Join guymann [0] (~charlie@
00:47:57bieberI'm taking a look
00:48:08 Join BlakeJohnson86 [0] (~bjohnson@2002:1876:a27b:0:227:13ff:fe65:1262)
00:48:35linuxstbbieber: I think bluebrother and domonoky1 will be the best person to talk to about coding style.
00:49:03bieberLooks like typical camelcase, uppercase for classes and lowercase for everything else
00:54:18funmanhm i found the cause for audio noise on clipv2
00:54:44funmanand also why the volume is lower on rockbox than OF
00:54:51funmanbut that means the volume doesn't go very low now
00:54:58 Quit efyx (Remote host closed the connection)
00:55:42funmanhmm no it's still there, just less hearable because the volume is higher :/
00:55:57funmani couldn't hear this noise in the OF playing silence
00:57:41 Quit stripwax (Quit:
01:03:33 Join fdinel [0] (
01:08:09 Quit evilnick_B (Quit: Page closed)
01:18:02 Quit fdinel (Ping timeout: 245 seconds)
01:20:54funmanhmm it's absent when playback is paused
01:23:02funmancould there be a background task which reads from storage regularly ? I assume no since else the HDD targets would need to always be spinning
01:23:25funmanregularly = several times per second, not each time the audio buffer empties
01:24:18funmani don't see the led anyway
01:24:57 Quit domonoky1 (Read error: Connection reset by peer)
01:26:58 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
01:28:36 Join saratoga [0] (~9803c6dd@gateway/web/freenode/x-puckxelmbeqjhtmh)
01:30:06funmanfwiw the noise is louder in the left headphone
01:31:29saratogafunman: I can ask about distributing the OF file, but i doubt they'll give me an answer
01:31:56saratogaany reason you want to? the current system seems to work quite nicely
01:32:36funmanperhaps people would stop asking a mkamsboot which worsk with newer OFs
01:32:51saratogai don't really care if they ask
01:33:08saratogai doubt they'll be that many more firmware updates for the ams players anyway
01:33:38funmanyeah there just have been a few this last month
01:34:07saratogai just delete the posts in the forums if people ask that in the dev thread
01:36:14saratogai'm pretty sure that noise on the clipv2 didn't used to be there
01:42:35funmanwhen the volume was higher (when we used DAC output to the phones instead of MIC/LINE/DAC summing) it was harder to hear
01:45:45 Part toffe82
01:48:04funmanit seems to be different when playing different files (with volume minimal = HP muted)
01:49:42 Quit bluebrother (Ping timeout: 240 seconds)
01:49:55 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
01:52:11 Quit ssorgatem (Remote host closed the connection)
01:52:20saratogawhich revision changed that?
01:52:59saratoganot r25732: as3525v2: fix volume setting (can now go very low)
01:53:49saratogaah i see
01:54:09funmanhm but r26045 should only affect FM
01:54:12funmanand rec
01:54:43funmanthe noise is still distinct before 25732 ?
01:57:42saratogai'm testing r26045 now
01:58:45funmanthe noise doesn't come from summing: disabling ADC/line/mic inputs in audioset3 -> still there
01:58:50 Quit bieber (Ping timeout: 264 seconds)
01:59:04 Join bieber [0] (
01:59:13saratogayeah its not r26045
02:00:26***Saving seen data "./dancer.seen"
02:02:39 Nick fxb is now known as fxb__ (
02:13:42saratogathe clicking was already there before write support, but its arguably less noticeable in r25732
02:13:46saratogafor example
02:14:20saratogain that version its more like a 0.5 Hz click
02:14:28saratogano idea what operates on that time scale
02:14:35funmanhm i noticed it changes with fclk
02:14:56saratogais anything but the CPU driven off fclk?
02:14:59funmanif you enable boosting it's very visible if you play with cpu frequency in the debug menu
02:15:04funmanpclk, but that should stay fixed
02:15:27funmani tried changing HZ but no result
02:15:49saratogain this version it only happens on boost
02:16:05saratogathe click perfectly aligns with teh cpu boosting
02:16:19saratogaalthough if i force it to stay boosted I get a much faster click anyway
02:16:56funmanfaster than with current svn ?
02:17:45saratogaits a lot worse with current svn
02:17:54saratogaif I use a format that doesn't boost like flac I can't hear it at all
02:18:44saratogai take that back i just barely hear it but i have to hold my breath and concentrate
02:19:11funmanhm yeah
02:19:29saratogaan amp would make this a lot easier to hear
02:19:39funmanof course we use maximum fclk now
02:19:49saratogaits weird that it doesn't depend on the amp gain
02:20:06saratogamakes it sound like its some kind of issue with the settings in the PMU rather then the DAC
02:21:51funmanhm yeah current use would scale with fclk i guess
02:23:55 Join anewuser [0] (anewuser@
02:24:02 Quit anewuser (Changing host)
02:24:02 Join anewuser [0] (anewuser@unaffiliated/anewuser)
02:25:49 Quit MethoS- (Remote host closed the connection)
02:26:46 Join j8048188 [0] (www-data@
02:27:15saratogar25890 is not bad either, at least unless it boosts
02:27:33saratogajust a faint warbling sound at the threshold of hear
02:27:53j8048188Hi all, One quick question about the e200r: Can you flash it with the standard firmware?
02:28:08saratogayeah i think so with e200tool
02:28:20j8048188so that's what e200tool does?
02:28:23saratogai'd check google, theres probably someone who has done it
02:28:27j8048188I can't find too much documentation.
02:29:15saratogai need an amp so i can more clearly hear this noise
02:30:36funmansince it doesn't happen when paused i wonder if this has to do with how we feed the DAC (dma)
02:30:43funmani tried with smaller bursts without effect
02:31:09j8048188Found something on ABI that might help.
02:31:20j8048188(at bottom of page, will try and report back)
02:32:12funmansaratoga: that's still more noise than with the OF I think
02:32:30saratogai can't hear anything with the OF
02:36:55saratogawe don't know what the as3543 registers are for the OF do we?
02:37:10funmannope that's what i was looking but it's complex
02:37:24funmanfor audio that is
02:37:26saratogathis could be as simple as having the VRM one notch too low
02:37:35saratogaor something stupid like that
02:37:43funmanfor PMU afaict it is all done at init but i copied them the other day and the sound was still there
02:37:50funmanVRM ?
02:37:55saratogavoltage regulator
02:38:24saratogathough i guess it would be wierd if DMA depended on having the VRM at a certain setting
02:38:37saratogaassuming the noise really is from a DMA transfer
02:38:46funmanDMA should work, but I was thinking of the pause between 2 transfers
02:39:12saratogahas anyone jtag'ed the clip+ yet? i thought it was pretty easy to do compared to clipv2
02:39:28funmanranma did
02:39:33funmanbut it's a dead clip+
02:40:10funmanwe transfer 8188 bytes at a time with DMA => 21.5 transfers per second
02:40:19funmanso it can't be that
02:40:25 Join mistertibs [0] (
02:40:52mistertibshow do I found out what my ipod's is at /dev/sd*?
02:41:21funmanmistertibs: run 'mount'
02:42:16 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
02:43:59funmansaratoga: the OF clipv2 inits PMU regs to: 0x1A-7 = 1, 0x17-7 = 0x30, 0x1A-2 = 0xCC, 0x19-2 = 0x41, 0x18-1 = 0x35, 0x18-7 = 0x31
02:44:29funman0x19-2 is the charger, but it also changes threshold for VBUS comparator (afaiu this is for USB detection)
02:55:26funmani still hear it with fclk@80MHz
02:58:07 Quit mistertibs (Quit: Leaving.)
02:58:30funmanat fclk@30MHz pclk@15MHz i'm not so sure, but there's still more noise than in OF
02:59:35 Quit funman (Quit: free(random());)
03:00:15 Join Forsaken_Boy [0] (~chatzilla@
03:04:28j8048188To change it to a regular sansa, I just flashed the other firmware on it and it worked like a charm.
03:06:40 Join krabador [0] (
03:06:57 Quit vaguerant (Quit: Java user signed off)
03:27:08 Quit S_a_i_n_t (Ping timeout: 248 seconds)
03:30:02 Quit krabador (Read error: Connection reset by peer)
03:35:55 Join S_a_i_n_t [0] (S_a_i_n_t@
03:39:01JdGordonbieber: ping
03:39:29JdGordonhey, im looking at the tag table and I'm a bit confused
03:40:19JdGordonmaybe that wiki page was confusing? the tags below mv.. you have them as |I
03:40:36JdGordonand the top three are *|.....
03:40:59JdGordonthe mv ones should be *I and top three *fIIIII
03:41:03bieberI thought the arguments were optional?
03:41:24bieberDoes it matter for mv? They only have one argument, so obviously that's either all or none
03:41:45JdGordondoes that make %mv illegal though?
03:41:50JdGordondoes it require the ()'s?
03:42:03JdGordonah ok
03:42:04bieberIf no arguments are given, it expects there to be no ()
03:42:17bieberI wasn't sure about a lot of them though, so do feel free to hack up the table
03:42:49bieberEspecially %Cl. Are l/c/r and t/c/b flags that get prepended to the width and height values?
03:43:16JdGordonalso, how do you feel about adding the WPS_TOKEN_ enum value to that table? it saves antother lookup on the next step of parsing
03:43:35S_a_i_n_tJdGordon: read upward a fair ways about your changes to checkwps (not sure if it hit SVN though, didn't see a commit if it did), does that mean I can just feed my themes to it and get the new syntax? That'd be awesome. I'm not looking forward to having to redo everything very much...
03:43:44JdGordonyes, though that is a bit nasty and we can change it to make it nicer
03:43:46S_a_i_n_ts/read/ I read/
03:44:03JdGordonS_a_i_n_t: yes
03:44:21bieberGo right ahead: I won't be using them, but it won't hurt anything to have them there
03:44:22JdGordonit took about 15min of actual coding to add that capability in :)
03:44:38S_a_i_n_tIs it in SVN yet? Or...will it be?
03:44:42bieberDo you want me to embed them in struct skin_element for tags, as well?
03:44:43JdGordonnot yet
03:44:52S_a_i_n_tsweet, I'll look out for the commit.
03:45:02JdGordonbieber: yeah
03:45:28bieberOkay, I'll make that change tomorrow before I do anything else
03:45:37S_a_i_n_tAnd thanks, that's a lifesaver. I understated how much I wasn't loking forward to the syntax change.
03:45:45S_a_i_n_t*looking even
03:46:08bieberTomorrow I'm going to try and focus on wrapping my parse tree in a Qt data model, and then I'll spend any time I have left just building up the parse table
03:46:24JdGordonbieber: I can build the table for you if you want
03:46:30bieberShould I make a local copy of whatever header defines the tokens in my themeeditor dir, or just link to it?
03:46:41bieberThat would be great, if you don't mind
03:47:10bieberIf no one has volunteered yet, btw, I can whip up a PHP script to do that conversion over the weekend
03:47:18bieberI won't have time for anything fancy, but it'll work
03:47:34JdGordonna, do the gsoc stuff first :) you dont want to run out of time with that
03:48:29S_a_i_n_tJdGordon cracks his whip ;)
03:48:32bieberYeah, that is probably the best plan
03:48:45JdGordonS_a_i_n_t: hehe no, I'm not mentor so is not my whip to crack
03:48:52*JdGordon just wants a good editor :)
03:49:13bieberHopefully it'll happen sooner rather than later
03:49:28S_a_i_n_tSame, but I'm afraid it'll put me out of work...well, at least give some competition ;)
03:49:35JdGordonhave you thought about how you will handle things like the sbs and wps bein on the screen together?
03:49:55S_a_i_n_t"script kiddies" will start churning out awesome themes now.
03:49:59bieberThe parser didn't take long to get together (especially if you're going to put the tag table together), and I did a lot of Qt work at the beginning of the Summer, so I can get running with that pretty quickly
03:50:46bieberWhat's the SBS? Sorry, I haven't delved much into the theme specifics yet, other than the WPS language
03:51:24*S_a_i_n_t wonders how to describe it...
03:51:30 Join mistertibs [0] (
03:51:36JdGordonthe sbs (Base Skin) is what describes the theme mostly...
03:51:41bieberOh, okay
03:51:49bieberIs that what provides the status bar and such?
03:52:04JdGordonit specifies a viewport which becomes the default viewport if another skin doesnt use the %wd (statusbar disable) tag
03:52:11JdGordonusually, yes
03:52:55bieberSo I guess just render that by default, and then render the WPS on top of it, or erase it if the %wd tag gets used
03:53:18JdGordonpretty much :)
03:53:51JdGordonit would be great if the editor could load a theme .zip, then let you choose to view the menus or wps or radio to edit
03:54:19mistertibsI have problem with ipod restore how to without itunes
03:55:16JdGordonis /dev/sdb2 the ipod?
03:55:20bieberJdGordon: That's the stuff I have scheduled to sort out after this week: before I get into rendering or anything complicated like that I'm going to try and get basic file handling nice and solid
03:55:48mistertibs/dev/sdb1 in mount it says it is that
03:56:40JdGordonso why are you trying to format /dev/sdb2?
03:57:16mistertibsbecause I am converting ipod to fat32
03:57:25mistertibsthe ipod is in mac os x format
03:57:32mistertibsI don't have any macs here
03:59:43JdGordondo ls /dev/sdb*
04:00:30***Saving seen data "./dancer.seen"
04:04:02 Quit n1s (Quit: Lämnar)
04:07:32JdGordonbieber: so last time :) | at the start with only one char lets you leave off the ()'s? * at the start lets you leave them off, or have all up to the first |?
04:07:43 Quit pixelma (Disconnected by services)
04:07:43 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:07:48 Quit amiconn (Disconnected by services)
04:07:50 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:07:50JdGordonand both allow empty ()'s?
04:08:03 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:08:12 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:09:34bieberJdGordon: Anything after a | is optional. If all the arguments are optional, then you can leave off the () altogether. If you have a * at the beginning, then you must either omit all the optional arguments or include all of them
04:10:32JdGordonok, just making sure
04:24:06JdGordonbieber: another fun edge case... how do you want to handle the differene between the lower case i tags (current track info) and upper cas I (next track info)?
04:24:31JdGordonin svn they are given the same token value and the parser sets a flag on the token struct to say if its for the next track
04:28:45bieberCan we just make them separate?
04:28:57bieberOtherwise I'll have to tear up the parser to make it work
04:31:54*S_a_i_n_t wonders what happened to the tracker.
04:35:03 Quit mistertibs (Quit: Leaving.)
04:36:16 Join chrisb [0] (
04:36:27 Join Buckwheat [0] (~chatzilla@
04:36:37BuckwheatWhats up everybody!
04:36:59JdGordonbieber: of use the same token and put the choice into the displayer?
04:37:02 Quit Buckwheat (Client Quit)
04:37:03chrisbwhat's the right disk format for microSD cards? vfat, ext2, fat?
04:37:16JdGordonthe origional text tag is going to be there in the editor right?
04:37:23chrisbmy debian complained FAT: incompatible with utf8
04:37:45JdGordonI mean it will use the skin_element tree directly?
04:37:57chrisband i run my shell, urxvt, in utf8
04:38:28chrisbnow the microdisk won't take some file names assigned by musicbrainz
04:39:12S_a_i_n_tdoesn't RB *need* everything in FAT32?
04:40:34krazykitchrisb, it has to be in fat (16 and 32 both work). as for your incompatible filenames, perhaps musicbrainz has a setting for fat-safe renaming
04:40:49 Quit bieber (Ping timeout: 264 seconds)
04:41:06chrisbS_a_i_n_t: thanks
04:41:16chrisbkrazykit: thanks, i'll check
04:41:23 Join Barahir [0] (
04:41:49S_a_i_n_theh... krazykit was the real brains there. Mine was almost a question also ;)
04:42:05 Quit anewuser (Quit: for SELL 2 by the price of 1 now!)
04:42:27 Quit Barahir_ (Read error: Operation timed out)
04:47:19 Join bieber [0] (
04:51:36 Join Tux2 [0] (
04:55:06 Quit j8048188 (Quit: CGI:IRC (EOF))
04:56:09 Join cool_walking_ [0] (~cb3b81c3@gateway/web/freenode/x-pawseocnjhxjnumu)
04:56:47 Quit TheSeven (Ping timeout: 240 seconds)
05:00:01*S_a_i_n_t notices the Nano MBRs in are all messed up and could lead to some confusion.
05:00:19S_a_i_n_tNoone seems to have thought of renaming them after the Nano2g was added.
05:00:42 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
05:02:59S_a_i_n_tmbr-nano2gb.bin and mbr-nano4gb.bin should be renamed to mbr-nano1g-2gb.bin and mbr-nano1g-4gb.bin
05:04:29S_a_i_n_tmbr-nano1g-2GB.bin and mbr-nano1g-4GB.bin rather, to avoid additional confusion.
05:05:46 Join Strife1989 [0] (
05:08:01 Quit chrisb (Ping timeout: 245 seconds)
05:09:08 Quit bieber (Ping timeout: 260 seconds)
05:09:23 Join bieber [0] (
05:10:17bieberJdGordon: Yes, I'll be using that directly
05:10:48bieberSo you could just give them the same token in the table, and have the display code check the "name" field when it encounters that token
05:11:03CIA-5New commit by Blue_Dude (r26294): Fix crash in talk.c logf line when outputting voice only clips, more comments in lang.h
05:12:37JdGordonbieber: yeah, and anyway I dont tihnk its so important to distinguish next and current tracks in the editor
05:13:33 Quit moparx (Quit: moparx)
05:13:49 Join CGL [0] (~CGL@
05:14:12 Nick CGL is now known as [CGL] (~CGL@
05:14:31CIA-5New commit by Blue_Dude (r26295): Oops, didnt mean to enable logf in talk.c
05:17:58 Join phanboy4 [0] (
05:21:07 Quit cool_walking_ (Quit: Page closed)
05:29:00JdGordonbieber: still there?
05:37:31 Quit Horscht (Quit: Verlassend)
05:43:08 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
05:45:47 Quit Topy44 (Ping timeout: 248 seconds)
05:46:56 Join Topy44 [0] (
05:51:14CIA-5New commit by Blue_Dude (r26296): lang.h voice comments were off by 1, fixed
05:53:08CIA-5New commit by jdgordon (r26297): add all the known tags to the table. Nothing is done about target specific tags yet
05:54:21 Quit Tux2 (Quit: Leaving)
06:00:33***Saving seen data "./dancer.seen"
06:11:27 Join Boldfilter [0] (
06:15:55S_a_i_n_tJdGordon: was it your intention to alphabetize the tag_table.c file?
06:17:17S_a_i_n_tOr is it some particular ordering that merely resembles alphabetic.
06:19:19 Join CaptainKwel [0] (
06:30:45 Quit liar (Ping timeout: 276 seconds)
06:38:26 Quit flatr0ze (Ping timeout: 245 seconds)
06:40:27S_a_i_n_tI suspect it doesn't actually matter, but I alphabetized tag_table.c. I kept the groupings the tags were in though, as I suspected that may be important.
06:44:04S_a_i_n_theh...messed it up. *Ahem*
06:51:26CIA-5New commit by jethead71 (r26298): SWCODEC: Wait for audio init to complete before starting radio or recording; hardware init might not be finished yet if those are the start screens.
06:54:07 Join einhirn [0] (
06:54:16 Join shai [0] (
06:59:41CIA-5New commit by jethead71 (r26299): Fix r26298 yellow: Somehow missed those warnings. Need playback.h.
07:00:26 Quit shai (Quit: Leaving)
07:10:47 Join JdGord [0] (~jd@
07:11:36JdGordS_ain't. I used the order from skin_parser.c to make sure o didn't miss any
07:12:20 Join shai [0] (
07:13:47CIA-5New commit by jethead71 (r26300): Gigabeat S: Turn off hardware effects (tone and 3d) when doing digital loopback for FM recording otherwise the signal and levels are poorly ...
07:24:03S_a_i_n_tJdGord: Well, now you didn;t miss any *and* they're in alphabetical order ;)
07:27:01 Nick fxb__ is now known as fxb (
07:28:23 Quit shai (Read error: Connection reset by peer)
07:28:44 Join shai [0] (
07:32:15 Quit CaptainKwel (Ping timeout: 252 seconds)
07:41:15 Quit rvvs89 (Ping timeout: 240 seconds)
07:44:01CIA-5New commit by pixelma (r26301): Give the radio screen setting(s) in the themes menu its own language strings so they don't simply reuse 'FM Radio' as in the main menu. Rename 'Remote ...
07:45:47 Quit bieber (Ping timeout: 248 seconds)
07:46:03 Join bieber [0] (
07:46:51 Join esperegu [0] (~quassel@
07:48:36leavittxWhat's wrong with tracker? Any of links like are down for me.
07:49:21S_a_i_n_tthe tracker is down.
07:49:26S_a_i_n_tDunno why though.
07:50:34pixelmaZagor didn't get it to work after the server move yet for some weird reason (works for him locally as far as I understood)
07:56:16 Nick fxb is now known as fxb__ (
07:58:49CIA-5New commit by pixelma (r26302): Change the hardcoded radio screen almost to how it was before - better use of screen estate, shorter lines, more info. Keep the 'progress' bar which ...
08:00:37***Saving seen data "./dancer.seen"
08:02:00JdGordonS_a_i_n_t: I'm not sure there is really any point making sure it is alphabetical just yet...
08:02:19JdGordonalso I see mistakes :)
08:02:38S_a_i_n_tIn the first or second one?
08:02:46S_a_i_n_tthe first was a shambles
08:03:42 Quit Boldfilter (Quit: Boldfilter)
08:04:47JdGordonthe 2nd
08:05:39S_a_i_n_tI think that may be because I thought I had to keep the groupings the tags were in.
08:06:21S_a_i_n_tnope, you're quite right.
08:06:25S_a_i_n_tI see it now too.
08:06:41S_a_i_n_tin the "p"s
08:07:39 Quit bieber (Ping timeout: 248 seconds)
08:07:52 Join bieber [0] (
08:10:15 Quit mc2739 (Ping timeout: 240 seconds)
08:10:52 Quit JdGord (Quit: Bye)
08:10:52 Join Flagoogleberries [0] (
08:17:53Flagoogleberrieshey there
08:18:05Flagoogleberrieshoping someone would be able to help me
08:18:16S_a_i_n_task your question.
08:18:21FlagoogleberriesI've got rockbox installed on my 5th gen 30gig ipod
08:18:22 Join shai_ [0] (
08:18:30Flagoogleberriesit's been running fine
08:18:46Flagoogleberriesbut i've recently put some .flac files on
08:19:17Flagoogleberriesmost of them work fine, but some have a kind of cycling scratchy noise on them that drown out the actual audio
08:19:34Flagoogleberriesthey work fine on my computer, so it's not a bad file
08:19:50Flagoogleberriesjust wondering what the cause might be and if there's anything I can do about it?
08:20:57pixelmaJdGordon: there was one more thing I would have wanted to do in the hwcodec radio skin which didn't seem possible yet - that would have needed %pb in a conditional and it didn't work. IIRC that's basically an old limitation but with all the work put into the bar widget I thought it could work now
08:21:30JdGordonI thought i fixed that?
08:21:33S_a_i_n_tpixelma: It's supposed to work.
08:21:42S_a_i_n_t%pb in a conditional I mean.
08:21:46 Quit shai (Ping timeout: 245 seconds)
08:21:47pixelmaJdGordon: %pm yes, but not %pb
08:22:00JdGordonpretty sure both should work
08:22:17S_a_i_n_tJdGordon: That might have something to do with that theme I got you to check out.
08:22:52 Nick shai_ is now known as shai (
08:22:53pixelmaok, have to try again. Hmm, could it be related to needing %pm twice even when not shown at the same time - is this possible?
08:23:17JdGordonwhy do you have pm twice?
08:23:33 Join flatr0ze [0] (
08:23:34JdGordonpeak meters arnt set up for multiple
08:25:53pixelmanow I remember... the old radio screen showed the prerecording time in the same line as the peak meter before and pushed the peak meter down one line. Currently I only do the former - if prerecording time is set then peak meter aren't shown at all anymore though
08:27:19pixelmaI wanted to replace the %pb below then with a %pm - but I also tried replacing %pb line with the prerecording time instead (which would only need one %pm then) but that didn't work either
08:27:31JdGordonwhen FS works again, open a task saying peak meters are in depserate need of some work. I tihnk they could be the last really outdated tag
08:27:38 Join merbanan [0] (
08:27:49pixelmaI'll try again though
08:33:16 Quit merbanan (Ping timeout: 276 seconds)
08:36:37Flagoogleberriesalright, guessing no one had any idea on my problem?
08:36:42Flagoogleberriescheers anyway
08:36:52Flagoogleberriesbye all
08:36:55 Quit Flagoogleberries (Quit: dum de dum)
08:36:56S_a_i_n_tpatience is a virtue
08:37:37 Quit esperegu (Read error: Operation timed out)
08:39:15CIA-5New commit by jdgordon (r26303): remove this branch
08:39:48saratogathat sounds like something worth opening a bug report for
08:40:41CIA-5New commit by jdgordon (r26304): recreat the branch. shouldnt need to keep this in sync with trunk as long as nothing major happens in the skin engine
08:43:23*JdGordon is looking for testers to play with that branch... unless you actually are testing for 3.6 rel
08:46:12 Join Guest6078 [0] (bjst@
08:46:13 Quit Guest6078 (Changing host)
08:46:13 Join Guest6078 [0] (bjst@rockbox/developer/Zagor)
08:46:27 Nick Guest6078 is now known as Zagor_ (bjst@rockbox/developer/Zagor)
08:51:49 Nick Zagor_ is now known as Zagor (bjst@rockbox/developer/Zagor)
08:53:14 Join wodz [0] (
08:56:54CIA-5New commit by jdgordon (r26305): make the parser accept the new format. skin_parser_old.c still accepts the old format (will be used for the updater checkwps)
08:57:43 Join esperegu [0] (~quassel@
08:58:49 Join ender` [0] (
08:59:44 Quit Strife89 (Quit: Reboot.)
09:01:25 Join flydutch [0] (
09:04:49 Quit elcan (Ping timeout: 264 seconds)
09:05:28 Join elcan [0] (
09:06:23 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
09:08:08pixelmaby the way, I found out a bit more about the crashes on my Ondio which sometimes happen when not replacing the full build. There is something fishy with language strings, they are sometimes completely messed up (I even saw strings that contained pieces of two different one) when going back even though I use English which should also be built in. If I have voice enabled then I get the crash
09:09:18S_a_i_n_tthat's....odd at best.
09:09:28pixelmait sounds like that shouldn't happen though at least with voice disabled and language set to English, but I'm not sure. I also wonder why the strings are so messed up in first place
09:10:12S_a_i_n_tsame crazy thing as "wrong builds in the wrong zips"?
09:11:21pixelmano, that's just replacing the ajbrec.ajz which should contain all the necessary parts on this target
09:11:46pixelmaas I said - at least with the English language
09:15:39 Quit [CGL] (Remote host closed the connection)
09:17:28 Join petur [0] (~petur@rockbox/developer/petur)
09:19:00 Quit kramer3d (Quit: Leaving)
09:20:19 Quit Forsaken_Boy (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
09:21:33 Quit esperegu (Read error: Connection reset by peer)
09:22:38 Join esperegu [0] (~quassel@
09:23:26 Join B4gder [0] (~daniel@rockbox/developer/bagder)
09:25:24 Join efyx [0] (
09:34:03CIA-5New commit by jdgordon (r26306): make checkwps work with the old format and be able to update to the new format
09:34:48JdGordonS_a_i_n_t: checkout the skinbreak branch, build checkwps and run your skins through it (one at a time) to update... then use a build from that branch to make sure it works :)
09:35:31 Quit GeekShadow (Read error: Connection reset by peer)
09:39:39 Quit flydutch (Quit: /* empty */)
09:41:04 Join flydutch [0] (
09:41:19 Quit xavieran (Ping timeout: 260 seconds)
09:44:07 Quit S_a_i_n_t (Ping timeout: 276 seconds)
09:45:33 Join S_a_i_n_t [0] (S_a_i_n_t@
09:45:50 Quit S_a_i_n_t (Client Quit)
09:46:01 Join lpereira [0] (
09:46:41 Join S_a_i_n_t [0] (S_a_i_n_t@
09:50:50 Join Rob2223 [0] (
09:51:48wodzwhat's wrong with such define: - do I need to escape something in this string?
09:53:52 Quit Rob2222 (Ping timeout: 245 seconds)
09:54:04Zagorwodz: how do you use it?
09:54:32 Join xavieran [0] (
09:54:33wodzinsert it in some array definition
09:55:02JdGordondoes anyone know how the cabbie images get into the zip file?
09:56:27S_a_i_n_tJdGordon: been wondering that myself for a while now...
09:56:41wodzZagor, it used like this:
09:57:15JdGordonwodz: can you not just put them in manually like the others?
09:58:17 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
09:58:55wodzJdGordon: this define will come from target tree
09:59:44JdGordonthose functions are in the target tree?
10:00:40***Saving seen data "./dancer.seen"
10:00:40JdGordonok, just just do #ifdef TARGET_DEBUG_MENU and hard code those options
10:03:32wodzJdGordon: You don't get the point - definitions can be different for different targets
10:04:31JdGordonyes but doing that is pretty messy...
10:05:06wodzleaving ifdefs as it is now is messy also
10:05:31JdGordonyes, but more understandable
10:05:56JdGordonis this perl regex wrong? /[\),]([^,\)]*?.bmp)\)/ <- trying to get it to match (filename.bmp) or ,filename.bmp)
10:06:03wodzI dissagree
10:06:53ZagorJdGordon: the first parenhesis is the wrong way
10:07:33JdGordonwoops.. but still no worky
10:08:12JdGordonhmm... it needs to match more, nute
10:08:44JdGordonyay :)
10:08:50Zagor /[\(,](.+?.bmp)\)/
10:09:06Zagoror even /[\(,](.+?.bmp)/
10:09:16Zagordon't require more than you want
10:09:22 Join DerPapst [0] (
10:09:35 Quit shai (Quit: Leaving)
10:10:12CIA-5New commit by jdgordon (r26307): make find the images in the new format
10:10:46JdGordonyeah I thought that bit before the * wasnt comepltyl right... but i wanted to change ti as little as possible
10:11:09JdGordonoh, no you pulled more than that off
10:12:45JdGordon/[\(,](.+?.bmp)/ didnt work.. it got too much at the start
10:13:27JdGordon%xl(A,lock-176x220x16.bmp,51,204,2) -> A,lock-176x220x16.bmp
10:13:59*JdGordon will stick with whats now in svn
10:14:25Zagorwell you did say the format was "(filename.bmp) or ,filename.bmp)"
10:15:25Zagorwith this more complex format you need your original regex
10:15:41JdGordonyeah, i forgot a few possibility :p
10:16:39 Join shai [0] (
10:22:36wodzhaha macro statement is perfectly ok, I was missing some includes
10:22:55 Join LinusN [0] (linus@
10:22:55 Quit LinusN (Changing host)
10:22:55 Join LinusN [0] (linus@rockbox/developer/LinusN)
10:42:36 Join Strife89 [0] (
10:47:21 Quit efyx (Remote host closed the connection)
10:47:55 Join efyx [0] (
10:48:05 Quit efyx (Read error: Connection reset by peer)
10:48:18 Join efyx [0] (
10:51:30 Quit DerPapst (Ping timeout: 272 seconds)
11:00:08 Quit shai (Quit: Leaving)
11:01:16 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
11:03:52 Quit phanboy4 (Ping timeout: 245 seconds)
11:07:03 Join mitk [0] (~mitk@
11:11:24 Join DerPapst [0] (
11:12:37 Quit JdGordon (Ping timeout: 245 seconds)
11:14:04 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
11:15:48mitkWhat is meaning of LANG_RECORD_TRIGGER_TYPE? Description in manual is empty.
11:16:11 Join shai [0] (
11:17:28S_a_i_n_tIt'll be for the recording triggers, like if volume goes over 'X'dB for 'X' period etc.
11:17:54S_a_i_n_tI'm pretty sure there's a description on recording triggers in the wiki
11:18:08wodzam I right that strncat is not included in rockbox libc?
11:19:34JdGordonuse strlcat
11:20:30 Quit esperegu (Read error: Connection reset by peer)
11:21:44 Join esperegu [0] (~quassel@
11:22:44wodzin strlcat size includes \0 or not?
11:23:35wodzhmm rather not
11:25:02 Quit wodz (Quit: Leaving)
11:28:00 Quit shai (Quit: Leaving)
11:32:17 Nick fxb__ is now known as fxb (
11:36:15 Join shai [0] (
11:42:34mitkYeah. Trigger mode is described indeed. But no trigger type. Is it action taken when recording is stopped by trigger in order to prepare to potential new recording? Available options are Stop, Pause, New File.
11:48:44 Quit DerPapst (Ping timeout: 260 seconds)
11:49:21 Join DerPapst [0] (
11:56:12 Quit esperegu (Read error: Connection reset by peer)
11:56:49 Join esperegu [0] (~quassel@
12:00:42***Saving seen data "./dancer.seen"
12:05:27 Join dfkt [0] (dfkt@unaffiliated/dfkt)
12:07:29 Join XXXStatik [0] (
12:10:04 Join watto [0] (~watto@
12:10:56 Quit GeekShadow (Quit: The cake is a lie !)
12:11:18 Quit XXStatik (Ping timeout: 240 seconds)
12:11:24 Quit S_a_i_n_t (Ping timeout: 248 seconds)
12:15:31 Join S_a_i_n_t [0] (S_a_i_n_t@
12:18:37 Join XXStatik [0] (
12:19:06 Join Strife89|Laptop [0] (
12:20:06 Join rvvs89 [0] (
12:20:41 Quit Strife1989 (Ping timeout: 240 seconds)
12:21:03 Quit Strife89 (Ping timeout: 260 seconds)
12:22:36 Quit XXXStatik (Ping timeout: 248 seconds)
12:27:51 Join funman [0] (~fun@rockbox/developer/funman)
12:40:24 Join Strife1989 [0] (
12:43:44 Quit BHSPitMonkey (Read error: Connection reset by peer)
12:50:20 Quit TheSeven (Ping timeout: 248 seconds)
12:50:44funmanjhMikeS: is it possible to change pcm_play_dma_get_peak_buffer like you did for recording ?
12:51:46 Quit shaggy-h (Ping timeout: 240 seconds)
12:51:50funmanbtw, is sleep(0) better than just yield() ? (i'm reading r26298)
12:52:05 Join ucchan [0] (
12:52:13 Quit shai (Quit: Leaving)
12:54:22 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
12:56:19funmanI guess sleep(0) can wait a bit longer since the number of ticks is adjusted by 1
13:00:07ucchanThere is a question. Though english.lang was updated in r26301, is this change included in 3.6?
13:00:41JdGordon3.6 wont be branched untill the end of the freeze
13:02:08ucchanJdGordon thansks. I will update japanese.lang.
13:04:11 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
13:07:51CIA-5New commit by funman (r26308): as3525: move audio loopback handling to audio_set_output_source
13:12:34 Join teru [0] (
13:13:48 Quit TheSeven (Ping timeout: 248 seconds)
13:14:15CIA-5New commit by funman (r26309): as3525v2: no need to set PMU_GATE bit (bit 3) in PMU_enable register when reading ...
13:14:22 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
13:15:42 Join shaggy-h [0] (
13:23:20 Quit DerPapst (Read error: Connection reset by peer)
13:25:58JdGordonI have the viewport colour being set outside of the viewport definition in a new tag. the problem with it is if you set a different colour mid way through, the text above the first colour change will be the colour the viewport ended with last time unless you set a colour before any text..
13:26:11JdGordonis that terrible?
13:26:36JdGordonor I fix it by remembering the start colour...
13:27:30pixelmahow do you set a new colour midway?
13:28:52CIA-5New commit by uchida (r26310): Japanese translation updates.
13:30:05JdGordonnew token
13:30:23*JdGordon changed it to remember the colour
13:32:43 Join DerPapst [0] (
13:34:17pixelmaZagor: with the tracker being down, how did it manage to get an "open bugs" mail out yesterday - or was it the one that actually should have been sent in the night to Satuday?
13:34:23pixelmaSaturday too
13:35:11Zagorit's only the tracker index.php page that is broken for some reason. the database lives, and so does the feeds.php page.
13:35:46Zagorindex.php works just fine when I run it from the prompt, but something makes it not work in apache
13:38:55 Nick YPSY is now known as Ypsy (
13:46:30 Quit storm` (Ping timeout: 240 seconds)
13:46:31jhMikeSfunman: 1) no, the play buffer needs bytes remaining (and should be live update too). 2) sleep(0) blocks to the end of the current tick, yield is runs another thread, if there is one to run
13:48:57*jhMikeS messed with the fft plugin and put the whole fft on COP (and made it not use 100% CPU all the time)
13:56:53 Quit esperegu (Read error: Connection reset by peer)
13:57:35S_a_i_n_tjhMikeS: Sweet! Does it show anything on spectograph/linear now? ;)
13:57:45S_a_i_n_tOn Nano 1/2g it doesn't
13:58:04 Join esperegu [0] (~quassel@ does. *just* enough to colour the very bottom pixel on the screen.
13:59:32JdGordonthe COP isnt drawing is it?
13:59:38jhMikeSS_a_i_n_t: didn't look at that
13:59:49jhMikeSJdGordon: and why shouldn't it? (but no it isn't)
14:00:22JdGordonjhMikeS: wierd shit will happen on the screen...
14:00:38JdGordonassuming the lcd driver is even thread safe
14:00:38jhMikeSI see spectrograph for log, not linear
14:00:46***Saving seen data "./dancer.seen"
14:00:46CIA-5New commit by funman (r26311): as3525: implement pcm_play_dma_get_peak_buffer correctly ...
14:01:11jhMikeSJdGordon: I know. Well, mpegplayer shares the screen between cores but it's carefully managed
14:01:35S_a_i_n_tjhMikeS: Meaning you see what I se? Or that you don;t see the option to change between logarithmic/linear?
14:01:36funmancowon d2 is dual core too
14:02:02jhMikeSS_a_i_n_t: I see linear as showing nothing, log looks ok
14:02:49S_a_i_n_tAha...then yes. You see what I do.
14:03:17jhMikeSI remember seeing it actually work before, did you?
14:03:26S_a_i_n_tThat used to happen with bars, but I expect that fixing that somehow had an effect on spectogram.
14:03:47S_a_i_n_tbefore the fix for bars not working in the same way
14:04:05jhMikeSodd that it should, it looks like all the different graphs are completely separate routines
14:04:35S_a_i_n_tbars used to show nothing...errr, but actually it showed as blank on logarithmic, not linear.
14:05:00*jhMikeS remembers (thought he still notes the sticky band 0 on bars)
14:05:52S_a_i_n_tI get the same here.
14:06:23S_a_i_n_t*sticky band 0 on bars.
14:07:26S_a_i_n_tswithing quickly between logarithmic/linear can (seemingly) fix that for a short while.
14:07:38jhMikeSdid you notice linear view shows really low-intensity bands better than log? (which seems like it should be the other way)
14:07:41S_a_i_n_tbut then it sticks when it maxes again next.
14:07:43 Quit flydutch (Ping timeout: 276 seconds)
14:08:39S_a_i_n_tyes, it does seem to do so.
14:09:23 Join flydutch [0] (
14:10:11jhMikeSI also made it mix left and right into mono before sending it off rather that just left channel
14:11:52 Join DataGhost [0] (~dataghost@unaffiliated/dataghost)
14:11:55 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
14:18:41 Join liar [0] (
14:25:27 Quit ucchan (Quit: Leaving...)
14:26:27S_a_i_n_tjhMikeS: Is this yet to be or never to be committed code?
14:33:11 Join froggymana [0] (www-data@
14:46:13 Quit S_a_i_n_t (Ping timeout: 252 seconds)
14:46:47 Join S_a_i_n_t [0] (S_a_i_n_t@
14:52:34 Quit steve|m (Ping timeout: 276 seconds)
14:56:00 Quit DerPapst (Quit: Leaving.)
14:57:28 Join robin0800 [0] (
14:58:42 Quit liar (Ping timeout: 258 seconds)
15:01:03 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
15:01:22 Join liar [0] (
15:04:58 Part LinusN
15:05:26 Join steve|m [0] (
15:06:01 Join XXXStatik [0] (
15:07:40 Quit flatr0ze (Ping timeout: 252 seconds)
15:09:42 Quit XXStatik (Ping timeout: 240 seconds)
15:13:52mitkDoes anyone here know what LANG_RECORD_TRIGGER_TYPE setting is? There is no description in manual & wiki.
15:14:51 Quit esperegu (Read error: Connection reset by peer)
15:15:30 Quit rvvs89 (Changing host)
15:15:30 Join rvvs89 [0] (ivo@pdpc/supporter/base/rvvs89)
15:16:02 Join esperegu [0] (~quassel@
15:18:34 Quit markun (Read error: Operation timed out)
15:21:45 Join panni_ [0] (
15:30:48 Join S_a_i_n_t_ [0] (S_a_i_n_t@
15:32:03 Quit S_a_i_n_t (Ping timeout: 258 seconds)
15:34:36 Nick S_a_i_n_t_ is now known as S_a_i_n_t (S_a_i_n_t@
15:46:31 Quit esperegu (Read error: Connection reset by peer)
15:47:40 Join esperegu [0] (~quassel@
15:49:42 Join PsYcHoHoLiC [0] (
15:50:50funmanmitk: i think it's the condition under which recording is stopped
15:52:13 Quit antil33t (Read error: Connection reset by peer)
15:52:19 Join antil33t [0] (
15:52:45 Quit froggymana (Quit: CGI:IRC)
15:53:29mitkfunman: But options are Stop, Pause, New File. It's not so clear for me.
15:53:54 Quit liar (Ping timeout: 258 seconds)
15:54:32 Join liar [0] (
15:54:38 Join XXStatik [0] (
15:55:49PsYcHoHoLiCsorry for the faq but is it stable on a v2 fuze?
15:58:13PsYcHoHoLiClooks like its not stable just yet...
15:58:25 Quit XXXStatik (Ping timeout: 260 seconds)
15:58:45 Quit mitk (Quit: Leaving)
15:58:58PsYcHoHoLiCbut last time i checked it was not even running progress :^)
16:00:36 Join kugel [0] (~kugel@rockbox/developer/kugel)
16:00:47***Saving seen data "./dancer.seen"
16:00:49kugelpamaury: ping
16:02:12kugelbieber: welcome \o/
16:04:17JdGordonbieber: hey. I'm trying out this git thing instead of the sv branch... jdgordon/rockbox/tree/skinbreak"> is setup for the new format (so you can test your parser against trunk or whatver)
16:04:19pamaurykugel: pong
16:04:35kugelpamaury: r26283 wasn't needed
16:04:52 Join Kitr88 [0] (~Kitar_st@
16:04:56kugelsee system-sdl.c in system_init()
16:06:04PsYcHoHoLiCthanks guys
16:06:11pamaurykugel: yes it is, SDL_QUIT message catcher don't call sys_poweroff
16:06:12kugelJdGordon: you could update the skinbreak branch too
16:06:14 Quit PsYcHoHoLiC (Quit: KVIrc 3.4.2 Shiny
16:06:32kugelpamaury: sure, atexit(sys_poweroff)
16:07:09JdGordonkugel: im working on the assumption that keping git in sync with changes is easier than svn (for which I've found is painful)
16:07:11pamauryah :) I didn't see that :) Then there must be a problem anyway because the sim doesn't close properly on my computer, let me check that
16:07:31kugelin fact, now we are calling SDL_Quit() twice, one if it being before the sdl objects in use are finalized
16:07:33 Quit Kitar|st (Ping timeout: 272 seconds)
16:07:41B4gder"[Mgx] RockBox Version 3.4 just got published" ...
16:07:57B4gderand a link to a forum that requires login
16:08:09pamaurykugel: ok, let me check once again without the one I added
16:08:10kugelpamaury: you were right about that it got forgotten in the target tree move, but I re-added it this way in a later commit. maybe you were not up-to-date?
16:08:13JdGordonspeaking of the sim.. does anyone know why the sim segfaults on exit?
16:08:28pamaurykugel: at you're right, perhaps my tree was not up to date
16:08:46kugelJdGordon: I think only the first svn merge is painful, after that it memorizes merge history
16:09:16 Quit Kitr88 (Ping timeout: 240 seconds)
16:09:19pamaurykugel: segfault :(
16:09:46kugelI got my segfaults fixed by atexit(sys_poweroff)
16:10:02pamaurylol, I'm rebuilding, just to be sure
16:10:06kugelwhich sim?
16:10:25pamauryit shouldn't change anything
16:10:52pamauryNo, it definetely segfault for me
16:11:28pamauryI think it's because SDL_Quit call sim_exit_irq_handler, which means that when sys_poweroff is called, there is not actived threads, perhaps it messing up the sim_* calls
16:12:01 Part Zagor
16:12:13kugelbut I assume it's because the event thread isn't terminated by sim_thread_shutdown()
16:12:56pamaurywhy would that cause a segfault ?
16:14:14kugelno idea
16:14:27 Join Kitar|st [0] (
16:14:54pamauryperhaps it would be safer to simulate a poweroff by sending a message, I think it's doable
16:15:59*kugel gets no segfault
16:16:03 Join XXXStatik [0] (
16:16:31kugeloh wait, now I do
16:20:20 Quit XXStatik (Ping timeout: 265 seconds)
16:20:27 Join flatr0ze [0] (
16:24:35funmanI found a hint in the OF that pclk upper limit is 80MHz on as3525v2
16:29:47 Join XXStatik [0] (
16:33:52 Quit XXXStatik (Ping timeout: 265 seconds)
16:34:06 Quit Zarggg (Ping timeout: 240 seconds)
16:35:44 Quit B4gder (Quit: It is time to say moo)
16:36:05 Quit esperegu (Read error: Connection reset by peer)
16:37:54kugelpamaury: freeze in sim_thread_shutdown() with the message
16:37:54pamauryI get a segfault in SDL_Quit, sim_thread_shutdown works for me
16:39:03 Join esperegu [0] (~quassel@
16:39:34kugelthis is my diff:
16:40:05 Join toffe82 [0] (~chatzilla@
16:41:23pamaurywith this you get a freeze in sim_thread_shutdown ?
16:41:44 Join solexx_ [0] (
16:42:41pamaurywait a min, I'll try to apply it to my tree
16:43:18pixelmadoes the forum have a setting for "send this PM to everyone" or everyone with some badge or so?
16:43:35pixelmajust wondering how one can PM so many people at once
16:44:28 Quit solexx (Ping timeout: 252 seconds)
16:44:40pamaurykugel: same for me
16:45:35pamaurykugel: it might be that some thread is not blocking at some point so the SDL_SemPost will never cause the thread to be closed, I'll try to list the thread close
16:46:01CIA-5New commit by funman (r26312): typo: 'nomasked' -> 'unmasked'
16:48:28Lloreanpixelma: I think you can PM a list of members by just typing all their names in separated by commas.
16:49:22 Join kugel_ [0] (
16:49:42 Quit kugel (Disconnected by services)
16:49:46 Nick kugel_ is now known as kugel (
16:49:50 Quit kugel (Changing host)
16:49:50 Join kugel [0] (~kugel@rockbox/developer/kugel)
16:51:40pixelmaLlorean: but even then creating such a list would take a while (still quicker than getting mail adresses one by one etc.)
16:51:59*GodEater sees a tweet about "Rockbox 3.4" just being published (on the motorola phones) :(
16:52:43pixelmaGodEater: is this what Bagder mentioned earlier?
16:53:08GodEatersorry - I missed it
16:53:20pamaurykugel: with your patch, no thread calls exit...
16:53:30*GodEater is in fact just an echo of Bagder
16:54:07kugelthere's an exit(0) in sys_poweroff()
16:54:29pamauryyet but sys_poweroff is called by which thread ?
16:55:07pamauryI think sys_poweroff is never called, otherwise it would close
16:55:18 Quit esperegu (Read error: Connection reset by peer)
16:56:00 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
16:56:14kugelpamaury: sys_poweroff() is called, otherwise sim_thread_shutdown() wouldn't be
16:56:23 Join XXXStatik [0] (
16:57:11pamauryit is not called on my computer
16:57:54 Join esperegu [0] (~quassel@
16:59:50 Quit teru (Quit: Quit)
17:00:07 Quit XXStatik (Ping timeout: 272 seconds)
17:00:20kugelpamaury: you said "same for me" when I said I get a freeze there
17:00:33pamauryyeah it freeze, but not at the same point it seems :)
17:01:21 Join XXStatik [0] (
17:01:31pamauryI don't get how sys_poweroff can get called
17:02:37kugelmaybe I'm wrong
17:02:48kugelI got it there before but not now anymore
17:03:43pamaurykugel: with grep, it can't be called in apps/ because it's not called from a button poweroff, it firmware/ it can't be called by powermgt.c nor by button.c
17:04:25pamauryin other targets, the sys_poweroff seems responsible for broadcasting SYS_POWEROFF, no SYS_POWEROFF will not call it :|
17:04:40kugelpamaury: ok, this exposes it:
17:05:11 Quit XXXStatik (Ping timeout: 264 seconds)
17:06:24pamauryok, now it freezes in sys_poweroff
17:09:23pamaurykugel: I have no idea currently, I must leave now but I'll be back and investigate this issue. The problem is that even when sim_thread_shutdown signal the semaphore, the threads don't close so probably the thread is not blocking or yielding
17:10:06 Join fyrestorm [0] (
17:12:25kugelpamaury: yea, there's strange code involving set/longjump. but I guess it worked for years so I have no idea why it shouldn't now
17:12:38 Part watto
17:15:59kugelthe segfault is in longjmp actually
17:18:33 Join DerPapst [0] (
17:19:06 Quit esperegu (Read error: Connection reset by peer)
17:19:58 Join esperegu [0] (~quassel@
17:20:22 Join einhirn [0] (
17:25:43 Join watto [0] (~watto@
17:30:57 Join XXXStatik [0] (
17:32:48 Join wincent [0] (
17:35:15 Quit XXStatik (Ping timeout: 260 seconds)
17:37:35 Quit wincent (Ping timeout: 264 seconds)
17:39:57 Quit lpereira (Quit: Leaving.)
17:40:55 Quit robin0800 (Remote host closed the connection)
17:45:15 Quit GeekShadow (Quit: The cake is a lie !)
17:54:18 Join flatrose [0] (
17:54:19 Quit flatr0ze (Ping timeout: 252 seconds)
17:58:07 Join Xerion [0] (~xerion@
17:58:46 Join CGL [0] (~CGL@
18:00:00 Join phanboy4 [0] (
18:00:01S_a_i_n_tsomeone may want to remove a rather confusing (and wrong) post from should the OP care to get around to reading bluebrother's instruction.
18:00:41S_a_i_n_tI thought about replying with "" but it seemed a bit tactless.
18:00:48***Saving seen data "./dancer.seen"
18:01:10 Quit einhirn (Read error: Connection reset by peer)
18:02:05S_a_i_n_tHmmm, there's a chance I'm parsing it wrong. It's confusing at best.
18:02:16 Join Zarggg [0] (
18:03:04CIA-5New commit by funman (r26313): as3525: write irq/fiq handlers in C ...
18:03:08 Quit Xerion (Ping timeout: 252 seconds)
18:05:02S_a_i_n_tNo, it is just wrong. The linked page doesn't even mention System Trace
18:06:59funmaninteresting bug caused by r26311: interrupts still happen after we masked them
18:07:45funmanbut the default isr (that we use for unhandled irq) is called
18:07:52 Quit petur (Quit: Leaving)
18:08:55funmanperhaps for vectored interrupts we can't use VIC_INT_EN_CLEAR but the bit 5 in the VIC_VECT_ADDRS entry
18:11:01 Join jgarvey [0] (
18:15:34funmannope it doesn't work.
18:16:02 Quit esperegu (Read error: Connection reset by peer)
18:16:37 Join esperegu [0] (~quassel@
18:17:38 Quit polobricolo__ (Ping timeout: 245 seconds)
18:19:20 Join wincent [0] (
18:22:29 Join polobricolo [0] (
18:23:14 Join Xerion [0] (~xerion@
18:26:07 Quit bluebrother (Disconnected by services)
18:26:11 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
18:29:17 Join MethoS- [0] (~clemens@
18:31:33pamaurykugel: the setjmp/longjmp is easy to understand: the thread run a wrapper, each time there is a blocking call (like switch/...), the code calls setjmp to memorize location and then longjmp to restore. If the longjmp fails, it's probably because the stack is invalid at this point
18:32:21kugelpamaury: no, set/longjmp is only usef for exiting threads
18:32:48pamauryAh my mistake then
18:32:55kugelthe jumpbuf's point to the return location of runtread which is wrapped around the actual entry points
18:33:18 Quit XXXStatik (Quit: Leaving)
18:33:25kugelthe segfault is that there's no proper jmpbuf for main thread which
18:33:37kugel(remove_thread is called for it)
18:34:09pamaurybut does it segfault for you ? With your last patch, it freezes on my computer
18:34:50 Join domonoky1 [0] (
18:34:54 Quit domonoky (Ping timeout: 258 seconds)
18:34:56kugelI meant the segfault you added SDL_Quit() for
18:35:19pamauryah :) But did you solve the freeze ?
18:36:15funmanlinux VIC code writes '32' to the default isr address register, any idea what that means?
18:38:13funmanthat would point 'just after the vectors'
18:38:41pamauryperhaps it dynamically writes code at this place
18:38:49pamauryI mean, at run time
18:39:24funmanalthough linux can be configured to have 'high' vectors (at 0xffff0000)
18:39:41funmani have a workaround although i'm not 100% sure if it's correct
18:40:01funmanin the undef irq handler, if no unmasked interrupt is active we just ack the interrupt and return
18:40:34funmanthe CPU would keep being interrupted until the interrupt is unmasked again, so the real isr would run and acknowledge the interrupt wrt the peripheral
18:40:55funmanbut would the CPU run or just keep being interrupted as soon as it returns from UIRQ() ?
18:41:37pamauryAsk a ARM specialist ? I know nothing about the ARM.
18:41:38funmanseems to work so far (testing with simple playback)
18:41:41*funman looks at Torne
18:42:34Tornei don't really get the problem :)
18:43:00funmando you know the ARM PL190 VIC ?
18:43:13funmanother VICs ?
18:43:39funmanwell i have my answer anyway: it doesn't work
18:43:44 Quit Strife1989 (Quit: Working in Windows for a bit.)
18:44:20funmanso in 3 words: we're fucked
18:44:40funmani can workaround the pcm/rec bugs but i don't know what to do for audio interrupts
18:51:46 Quit DerPapst (Quit: Leaving.)
18:52:16funmannope that should be fine
18:53:11pamaurykugel: do you really think the sim needs to gracefully ask the threads to close ? Can't it just send a SYS_POWEROFF message, wait and then let SDL close everything ?
18:53:32kugelyea, I think it should
18:54:57pamaurykugel: on my comtuer, closing the 'scroll' thread freeze. Looking at the code, the only opportunity for the thread to check whether it should close is the 'sleep' function but as it probably calls SDL_Delay, it means this thread can't be exited properly. But then I don't understand why it did work previously
18:55:59kugelI should tell you that the initial segfault can be fixed by avoiding to call remove_thread for the main thread, but I get another segfault in SDL_Quit()
18:56:59kugelthe main thread is no actual thread anyway (not created explicitely by SDL_CreateThread()) but it's what you get by starting the program
18:59:11 Quit Strife89|Laptop (Ping timeout: 272 seconds)
19:00:14pamauryindeed, this closing system for the sim is really strange
19:00:42*pamaury would go for a brutal solution, kill them and don't bother
19:02:05pamauryI seriously doubt this system ever worked in fact. the previous did at "atexit(SDL_Quit)" and did a exit(0); in the button handler no ?
19:02:11pamaury*previous code
19:02:13kugelpamaury: fixes the first segfault
19:03:09pamaurythe code did not remove the thread on exit ? Which thread(s) do exit actually ?
19:04:40kugelwell, sim_thread_shutdown closes all except the main and the event ones
19:04:54 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
19:05:12CIA-5New commit by learman (r26314): Updated Swedish translation (again).
19:05:19kugelthe event one could exit properly if we manage to make it via queue_broadcast, the main doesn't need to exit
19:05:53kugelpamaury: does the above patch work for you?
19:06:17kugelI don't get the other segfault in SDL_Qui() anymore after doing an ubuntu update (which updated libc6)
19:07:00pamauryworks smoothly, great job
19:08:20pamauryso the problem was just that the main thread is not a normal thread in fact
19:11:49funmani can't find a working locking scheme for INT_AUDIO(), perhaps DMAC is the only non maskable interrupt
19:12:26kugelpamaury: previously it was an explicit thread (the implicit main thread did read the button events), hence the thread code doesn't protect the main thread from removal
19:13:34pamauryThat explain things, I though the main thread protected it but I was wrong. The SDL simulator is really a delicate piece of software
19:13:58 Quit flydutch (Quit: /* empty */)
19:13:59pamauryI think you can commit it
19:14:32 Join TopyMobile [0] (
19:15:25kugelpamaury: it was only protected in sim_thread_shutdown (by the t->context.t != NULL check) but I overlooked the other way for the threads to exit
19:16:05pamauryThis should fix a entry in flyspray of the sim crashing on exit under Windows but I can't access FS anymore, that's a shame
19:16:21CIA-5New commit by kugel (r26315): Since the sdl to target tree move, the main thread cannot be removed anymore, ...
19:16:47kugelhm, I forgot to add that I also removed the (potentially dangerious) exit() in misc.c
19:16:56S_a_i_n_tpamaury: Don't suppose you remember which sim it was?
19:17:10S_a_i_n_tI don't ever recall seeing any crashes on exit.
19:17:34pamauryI think
19:17:48kugelerr, what's wrong with flyspray?
19:17:56pamauryperhaps it's unrelated, I think it was with an old revision, r24xxx
19:18:11pamauryWhen I go on flyspray I get an empty index.php
19:18:17*kugel too
19:18:38 Join shai [0] (
19:18:59S_a_i_n_tIt's technically running.
19:19:13 Join mt [0] (~mtee@rockbox/developer/mt)
19:19:35S_a_i_n_tthere was a comment earlier that explained it a bit better IIRC
19:19:55 Quit flatrose (Remote host closed the connection)
19:20:59 Join stripwax [0] (
19:21:00 Quit esperegu (Read error: Connection reset by peer)
19:21:28 Join Luca_S [0] (www-data@
19:21:55 Join esperegu [0] (~quassel@
19:22:26pixelmaabout flyspray - see logs 13:35
19:23:43 Quit rhodan (Remote host closed the connection)
19:24:13kugelpixelma: thanks
19:27:11 Join rhodan [0] (~Quassel@2001:1608:12:2::38)
19:32:54CIA-5New commit by funman (r26316): as3525: as revealed by r26311, it seems DMAC interrupt can't be masked ...
19:35:04 Quit wincent (Ping timeout: 260 seconds)
19:35:22 Join wincent [0] (
19:36:34 Quit Xerion (Quit: )
19:38:46 Quit CGL (Ping timeout: 240 seconds)
19:39:18 Join Stephen__ [0] (~S@
19:39:33 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
19:41:37 Quit shai (Quit: Leaving)
19:41:39Stephen__how hard would it be to have a soft hold switch like the clip+ but for the e220v2 ? my hold switch broke
19:43:37funmandead easy
19:43:50Stephen__for someone with no C experience ?
19:44:21*Stephen__ should look at the source and try make some sense of it
19:44:47kugelAlexP: I'm sorry I haven't yet managed to put up the list, I'll hopefully do it in the evening or tomorrow morning
19:46:38funmanStephen__: something like ?
19:47:11kugelAlexP: the list of pluginlib actions in the affected plugins
19:47:34Stephen__funman, if only I knew what half of tha tmeant! can i just patch that into a current svn ?
19:51:01 Nick fxb is now known as fxb__ (
19:53:37funmanStephen__: yep it should work
19:53:53*Stephen__ goes off to build
19:54:10 Join bmbl [0] (~Miranda@unaffiliated/bmbl)
19:57:51Stephen__it's asking wqhich file to patch ?
19:58:07funmanpatch -p1
19:58:11S_a_i_n_tchange the -p(whatever) +/-1
19:58:21Stephen__using <p1 failed at hunk i'll paste now hang on
19:58:57S_a_i_n_tor -p1?
19:59:51S_a_i_n_t"patch -p1 < patch.patch"
20:00:02 Quit kugel (Remote host closed the connection)
20:00:31Stephen__hunk #1 FAILED at 85
20:00:42 Join Strife89 [0] (
20:00:49***Saving seen data "./dancer.seen"
20:01:25S_a_i_n_t85 isn't too specific without the filename it failed in.
20:01:53Stephen__i removed it and patched again bu tno warning this time
20:02:31Stephen__it did say that button-e200v2-fuze.c and button-target.h were patched earlier
20:02:50Stephen__i gues snow just configure make build etc
20:02:53Stephen__and test
20:02:54S_a_i_n_tAwesome, compile and you should have exactly what you wanted.
20:06:34Stephen__thanks and thanks funman for the patch
20:13:31 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
20:13:39 Join CGL [0] (~CGL@
20:21:21 Join Strife1989 [0] (
20:24:43 Quit Strife89 (Ping timeout: 245 seconds)
20:30:38 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
20:30:39 Quit esperegu (Ping timeout: 264 seconds)
20:33:50 Join esperegu [0] (~quassel@
20:41:07 Quit stripwax (Quit:
20:46:04 Join stripwax [0] (
20:46:22 Quit stripwax (Client Quit)
20:53:34 Join Zagor_ [0] (
20:53:51Stephen__funman, sorry to be a pest but the patvch throws up the following error when compiling ?
20:55:21 Join shai [0] (
20:55:35 Join einhirn [0] (
20:55:42 Quit bertrik (Remote host closed the connection)
20:56:21 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
20:56:58funmantry adding this patch on top of your patched tree
20:59:50Stephen__ −− is that ok ?
21:04:57Stephen__ill try an see
21:10:21 Join petur [0] (~petur@rockbox/developer/petur)
21:11:28Stephen__still no worky funman,
21:13:21 Quit esperegu (Read error: Connection reset by peer)
21:13:34 Part watto
21:14:20 Join esperegu [0] (~quassel@
21:18:54 Join tchan1 [0] (
21:19:47 Quit tchan (Ping timeout: 240 seconds)
21:20:50 Join storm` [0] (
21:20:55funmanremove line 39 of kernel-as3525.c
21:21:25 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
21:23:48 Quit tchan1 (Ping timeout: 264 seconds)
21:28:25 Quit esperegu (Read error: Connection reset by peer)
21:28:29 Join Wraith_ [0] (
21:30:21 Join Conax [0] (
21:31:12 Join esperegu [0] (~quassel@
21:31:20 Join stripwax [0] (
21:31:35Conaxkommt hier jemand aus Deutschland?
21:31:39Wraith_Anyone used Rockbox on Sansa Clip Plus?
21:31:48evilnick_BConax: English only
21:31:51 Quit stripwax (Client Quit)
21:31:59ConaxI want to use rockbox on sansa fuze V2
21:32:22ThomasAHWraith_: many (including me) and it works fine
21:32:38Conaxdoes it run well?
21:33:19 Quit tchan (Ping timeout: 240 seconds)
21:34:09Wraith_are there any important problems with this?
21:34:20Wraith_maybe freezing?
21:34:30Conaxdon't know didn't installed rockbox yet
21:35:32Conaxwhen it doesn't work is there a chance to reset my mp3 player or is it bricked forever?
21:36:05 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
21:36:33evilnick_B"Even though the bootloader installation and rockbox itself are tested, there's always a risk of BRICKING your device PERMANENTLY. The Sansa AMS (except the e200v2, for which it involves opening the player) have no known method for recovery, so this risk will not go away, even in (if any) future stable releases."
21:36:34Luca_Sall the information you need is on the wiki, on the SansaAMS page. briefly, rockbox on fuzev2 and clip+ work quite well, with only sporadic crashes. the chance of bricking is always present, but AFAIK is quite small if you stick to the instructions.
21:38:53Conaxthx I read the articles but just aked here to be up to date
21:39:38Conaxbecause on the website the v2 ist in the stable category
21:40:15Conaxso I thought maybe there are some big problems
21:40:27funmanit's not been tested as much as v1 but it should work the same i think
21:40:38Conaxah ok
21:45:21 Join tchan1 [0] (
21:46:23 Quit tchan (Ping timeout: 245 seconds)
21:47:26 Join Blue_Dude [0] (~chatzilla@rockbox/developer/Blue-Dude)
21:48:38 Quit shai (Quit: Leaving)
21:51:13jhMikeSkugel: (for logs) all that crazy sim stuff jmpbuf came from your's truly. (easiest way to get threads back into their start routines so they can finish normally rather than be terminated forcefully)
21:53:38Stephen__funman, it comoiled but doesn't work :o/
21:54:39Conaxok thx guys I whish you a good night and when it works fine I give you a small report
21:58:43 Quit Conax (Remote host closed the connection)
21:59:11bieberDoes anyone have experience working with Qt's data model classes?
21:59:19funmanStephen__: sorry, perhaps it requires more work
22:00:05Stephen__ah no problems. i'll just live with it and us emy clip+ a bit more ;o) thanks for all the hard work.
22:00:53***Saving seen data "./dancer.seen"
22:01:34 Join ssorgatem [0] (~ssorgatem@
22:04:14ssorgatemfunman: ping?
22:06:07 Quit Blue_Dude (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539])
22:06:22ssorgatemwhere can I find an earlier working bootloader for fuzev2?
22:06:46 Join nik2 [0] (
22:06:58funmanStephen__: which key are you trying? softlock only works in the wps
22:07:18Stephen__in the wps trying rec + select
22:07:19 Part nik2
22:07:26funmanssorgatem: hm you need to build it from svn
22:07:54funmanStephen__: tried in different order? perhaps this combo is already used
22:08:18Stephen__nah its not used for anything and i tried different order and different buttons too
22:09:14funmanHAS_BUTTON_HOLD is removed from button-target.h ?
22:10:51funmanssorgatem: still white screen?
22:11:12Stephen__funman, yeah no sign of it in there.
22:12:16Stephen__just discovered there's no sound in recording mode fm radio works but not in recording
22:12:36Stephen__i'll try a current build from the website and check that
22:16:22 Quit bertrik (Remote host closed the connection)
22:16:37Stephen__refreshing database......
22:18:28Stephen__yeah it worked so it was in the modified build i made :ox
22:23:08 Nick tchan1 is now known as tchan (
22:23:17 Quit tchan (Changing host)
22:23:17 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
22:24:57ssorgatemfunman: hello
22:24:59Stephen__what was r26306 ?
22:25:10ssorgatemfunman: I'm currently posting it in the forum thread
22:25:35Stephen__broke some of my X5 themes
22:28:07 Join kugel [0] (~kugel@rockbox/developer/kugel)
22:28:17Stephen__ok found out what it was.
22:28:22Stephen__JdGordon, did it!
22:28:25kugeljhMikeS: "all that crazy sim stuff jmpbuf came from your's truly" ?
22:28:26 Nick Ypsy is now known as YPSY (
22:28:46Stephen__make checkwps work with the old format and be able to update to the new format
22:28:46Stephen__ −− seems to have broken them hmmm
22:28:53Stephen__must recheck them
22:28:56kugelI know it's the most simple way to terminate them, I was just saying it looks strange when you look at it first
22:29:56 Quit Wraith_ ()
22:33:41 Join Boldfilter [0] (
22:33:53ssorgatemfunman: I think the bootloader is the culprit
22:34:09 Part Boldfilter
22:34:32ssorgatemfunman: since it sometimes shows blank screen instead of "ver 2.0", and then shows correctly rockbox itself
22:35:07funmandid you try a bootloader built from svn ?
22:35:24ssorgatemfunman: nope
22:35:44ssorgatemfunman: only the one from the wiki
22:37:25 Quit bmbl (Quit: Bye!)
22:37:25pixelmaJdGordon: %pb in a conditional has the same problem you fixed with %pv. When used without parameters and followed by a | as a conditional separator the progressbar doesn't show
22:37:52jhMikeSkugel: yeah, when doing threads like the core from what it was
22:38:30funmanssorgatem: try building it from svn to see if it works better
22:39:44jhMikeSkugel: btw, why create a mutex and a cond for starting the event thread? (FYI: you don't lock the mutex before signalling the cond, which is why it would have problems)
22:40:00funmanor figure which revision your older bootloader uses and bissect between this revision and the revision of v2 bootloader
22:40:02pixelmaadding a space or something to my one line test wps helps (code: %?mh<%pb|test>)
22:40:08pixelmaJdGordon: ^
22:40:31kugeljhMikeS: the video system must be initialized before resuming
22:40:52kugeland the video system init must happen from the thread that polls the (keyboard) events
22:41:01kugelsdl conds requrie a mutex
22:41:31jhMikeSright, just poll a bool and make it true right before the message loop. (made a patch too)
22:41:33kugelI wrote some comments, are they unclear?
22:41:49jhMikeSyes, very clear
22:42:21kugela bool works too, but conds are made for that, aren't they?
22:42:46funmanssorgatem: v2 bootloader svn revision is r25903
22:43:19jhMikeSkugel: yes, but I was thinking it's one-timer thing that can just do a sleep loop (to make it simple and reliable)
22:44:08kugelI need to lock the mutex before signaling?
22:44:12jhMikeSthe mutex has to be locked before creating the thread, then the thread must lock it at the start and unlock just before doing the loop
22:44:16kugelit's locked all the time
22:44:39jhMikeSnot when you wait on the condition, doing that unlocks it
22:44:58kugeloh :)
22:45:14jhMikeSand cond isn't a sticky flag like a sem (which could also be used, needing just one object)
22:45:46kugelhow would a sem work?
22:46:29 Quit fyrestorm (Read error: Connection reset by peer)
22:47:19ssorgatemfunman: so far, no white screens
22:47:33jhMikeSkugel: create it with a zero count, wait on it. the event thread then ups the count just before the loop.
22:47:42ssorgatemfunman: i'll let it play all the night
22:49:12funmanssorgatem: i had made a working bootloader around march 28th but don't remember the exact revision
22:49:20kugelit appears I have misunderstood the concept of semaphores
22:52:00kugeljhMikeS: I don't mind if you change it :)
22:52:13ssorgatemfunman: ok
22:53:42jhMikeSkugel: ok. mostly my concern was the cond was not being used with a locked mutex.
22:54:19kugelthe sem solution sounds fine to me but if you prefer the busy bool polling, I don't feel strong about it
22:56:13funmani think busy waiting/polling are battery worst enemies on portable devices
22:57:10 Join Soap_Hotel [0] (
22:58:12Soap_Hotelshould the large drive support for iPods now being official be mentioned on the front page? (Project News)
22:59:50 Quit Stephen__ (Quit: Leaving)
23:00:44Soap_Hoteland S_a_i_n_t - I wouldn't chide the man for "reporting" Doom issues on his modified build - he's "discussing" them in the only place outside IRC one can - the forums. If he had submitted a bug to the tracker using his modified build I would view it differently.
23:00:45jhMikeSkugel: no need to create any objects then, but
23:01:34jhMikeS(got a bunch of other junk going, so building and checking would be inconvenient atm)
23:03:13 Quit esperegu (Read error: Connection reset by peer)
23:04:21kugeljhMikeS: seems to work
23:04:24 Join esperegu [0] (~quassel@
23:04:51jhMikeSI think I had a typo (SDL_Sem) right?
23:05:10jhMikeSpatch, line 30
23:05:35kugelyes indeed :)
23:06:43bluebroth3rbieber: (sorry for the late reply) the rbutil coding style should be pretty close to Qt style, at least for naming (including the setFoo() and foo() naming). At least my intention is to have that pretty similar :)
23:06:44kugelI also checked win32 build under wine (I added the SDL_AddTimer() thing because it would crash there)
23:07:02bluebroth3rfor indentation / braces we stick to Rockbox coding style.
23:07:12bieberbluebroth3r: Thanks
23:07:20bluebroth3rbasically, my intention is to use Qt style for C++, Rockbox style for C code.
23:07:40bieberYou wouldn't happen to know a good reference for working with the QAbstractItemModel interface off the top of your head, would you?
23:08:01bluebroth3rthe Qt documentation :)
23:08:13bluebroth3rbut apart from that, no. Sorry.
23:08:19bieberlol, that's what I'm trying to use
23:08:48bluebroth3rwanted to look into that myself (well, I did a while back trying to convert some of the views in rbutil to MVC) but time went out ...
23:08:59bluebroth3ras for too much Rockbox related lately :(
23:09:31bieberThis looks like it'll be might convenient once I get it down, it's just an awfully complicated interface to get a feel for. Although I'm sick today, so I'm hoping that all this will make a little more sense when my head clears up ;)
23:10:15 Quit einhirn (Read error: Connection reset by peer)
23:10:19bieberIf you just have linear data, it's pretty simple, but trees are giving me problems
23:11:42CIA-5New commit by jethead71 (r26317): UISimulator: cleaner startup using semaphore only to tell when event_thread is done initializing.
23:12:00bluebroth3rlinuxstb: from a quick check of the rbutil source it doesn't look like the bootloader-info file isn't used anymore. I'm pretty sure it was used before for h100 / h300 bootloaders, and IMO it should get used to verify all bootloader downloads
23:13:16CIA-5New commit by bieber (r26318): Trying to implement QAbstractItemModel for parse trees, haven't got it working yet (current state will spawn an empty treeview window)
23:13:47kugeljhMikeS: that looks a lot cleaner indeed, thanks :)
23:15:36jhMikeSkugel: welcome. btw, who calls sim_thread_shutdown such that the main thread would see threading code that would cause it to try to exit with jmpbufs?
23:20:08bluebroth3rbieber: just checked some code I did a while back, but I was using a QStandardItemModel :(
23:22:59kugeljhMikeS: whatever the atexit() hander calls, which is I guess the main thread
23:25:26kugelI'm not exactly aware of everything that happens at shutdown
23:25:33jhMikeSdoes SDL have an exit routine for the current thread?
23:25:47kugelit could also be the event thread which calls exit()
23:26:55 Quit esperegu (Read error: Connection reset by peer)
23:27:03kugeljhMikeS: nope, sdl doesn't have any way to cleanly terminate a thread (at least I couldn't find anything in the docs)
23:27:15kugelunlike pth which has that
23:28:12 Join esperegu [0] (~quassel@
23:28:53jhMikeSyou can get the ID, but not the SDL_Thread pointer :\...which is why I left it NULL in the main init
23:29:46kugelare you sure you did that?
23:30:04jhMikeSno :) did you do it?
23:30:32kugelbefore I moved it to the target tree the implicit main thread went to event polling and the rockbox main thread was created with SDL_CreateThread and thus had the pointer
23:31:04kugelyes, I did change it so the implicit main thread and the rockbox main thread are the same
23:31:34kugelI don't think the main thread has a SDL_thread pointer anyway, it's not created by sdl to begin with
23:32:04jhMikeSyeah, wonder it can't wrap an existing OS one...odd
23:33:47kugelit's not a thread until the threading engine is initialized but only a process, isn't it? I don't know if there are thread backends that can wrap the process into a thread
23:35:03jhMikeSkugel: does this work?
23:38:55kugeljhMikeS: looks like
23:40:25 Quit bluebroth3r (Quit: reboot)
23:40:51 Join Strife89 [0] (
23:41:16jhMikeSis anything relying on t being NULL? (rather than some garbage but non-NULL value)?
23:41:45jhMikeSother than the one function
23:42:28kugelsim_thread_shutdown(), but you could also start the for loop at 1
23:42:38kugelI don't think anything else does
23:42:44jhMikeSsim_thread_shutdown would just proceed because it sees a NULL t and must wait on t
23:43:02jhMikeSso really it can't be some arbitrary thing
23:46:53 Quit S_a_i_n_t (Ping timeout: 276 seconds)
23:50:09 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
23:54:23 Quit Luca_S (Quit: CGI:IRC (EOF))
23:54:52 Quit GeekShadow (Read error: Connection reset by peer)
23:55:43 Quit DataGhost (Ping timeout: 240 seconds)
23:58:29 Quit CGL (Quit:

Previous day | Next day