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 2009-10-14

00:00:26Torneender: if your actual question is "it takes too long to start playing when i select a track, how do i make it quicker" then there are some suggestions :)
00:00:35Torneturn dircache on if you don't have it on already
00:00:37pixelmalinuxstb: the "specimg" is used to get the path to some images, so e.g. specimg-front for the player images. In this example if specimg is ipodnano2g is specimg it'll pick an ipodnano2g-front.png for it etc.
00:00:44Tornedon't use album art
00:00:55Torne(or use album art that's already the right size)
00:02:00pixelmaso I want to know to get the name of the svg/png/pdf correctly
00:02:02enderTorne: it takes about 20 seconds before it starts playing. i'm not using any album art, and you're right - it probably isn't loading (as the disk LED isn't on)
00:02:09*kugel notices a skin is zero'd 3 times upon loading
00:02:29Torneender: what player, and do you have the directory cache enabled?
00:02:38linuxstbpixelma: Then yes - "ipodnano2g" will be the name.
00:02:41enderiRiver H120, cache enabled
00:03:01kugel4 times even!
00:03:12enderi also noticed that the Loading % updates much slower between ~85% and 95%
00:03:19pixelmalinuxstb: thanks
00:03:29Torneloading %?
00:04:02Torneooh. i've just paniced my ipod
00:04:10Torneupdating size on empty dir entry 79
00:04:20enderit shows "Loading... XY% done (STOP to abort)"
00:04:30Tornethat's not the track loading, that's the playlist being created
00:04:37ender(in a gray box at the center of screen)
00:04:41 Quit TheSeven (Read error: 60 (Operation timed out))
00:04:54Tornewhere are you starting playback from? a directory, or the database?
00:05:02linuxstbTorne: I've seen that a few times on my Nano2G, but thought it was just FTL flakiness...
00:05:13Tornelinuxstb: this might be ATA DMA flakiness, potentially ;)
00:05:20Tornesince my build has that..
00:05:23Tornebut it's been super stable for months
00:05:25enderi just resume playback
00:05:36Torneender: how long is the playlist you are resuming, though? :)
00:06:33endermight be pretty long - currently at 1300 entries
00:06:41TorneEr, then that's why, probably
00:07:00pixelmaDgby714: is that ok colour wise"> ?
00:07:07Torneit takes a while to load a playlist with 1300 entries
00:07:14Tornego and play just one directory
00:07:25Torneyou should find the splash with "Loading" pops up for a very brief time indeed
00:07:29Tornepossibly so short you can't even read it
00:07:53enderhm, you're right, that starts instantly
00:08:06 Join intrados [0] (
00:08:30 Join HBK [0] (
00:09:21linuxstbpixelma: The hold switch isn't quite right. It's white, and should start further to the left, and also be slightly longer, and slightly thinner...
00:10:38JdGordon|but apart from that... perfectly accurate!
00:10:46JdGordon|why the the word forward abbreviated?
00:10:50linuxstbYes, it looks great
00:10:58pixelmaI always drew the hold switch a bit thicker so it's visible at all
00:11:37linuxstbOK, but I think it should start a lot further to the left - about where the curve starts
00:11:50linuxstbAnd be about 50% wider
00:11:54 Quit bertrik_ (Remote closed the connection)
00:12:09pixelmaJdGordon|: that's how the manual will refer to the button - and seeing that it takes enouigh space as fwd, e.g. in button tables...
00:12:19Torneender: so yes, it's the length of your playlist that's the issue.
00:12:20linuxstbOn my monitor, it's almst exactly the same size/shape as my actual Nano2G
00:12:33enderthanks for the help
00:12:38 Quit jgarvey ("Leaving")
00:12:59Torneno problem, just next time don't disagree with people who have read the code ;)
00:13:11pixelmaJdGordon|: s/button/wheel rotation whatever
00:13:34Dgby714hers is a bit wider then mine bit it my just be my rez
00:13:34linuxstbpixelma: Although shouldn't the screen have a black background now?
00:13:45amiconnTorne: Thing is, rockbox shouldn't need that long for loading playlists that long
00:14:13linuxstbpixelma: I see some of the newer targets have a black screen (beast, D2...)
00:15:02Torneamiconn: hm. i wouldn't know..
00:15:05amiconnThe rockbox playlist system was developed for a CPU that's significantly slower than the H120's
00:15:22Torneender: so maybe there *is* something wrong regardless, but it'snot what you thought ;)
00:15:23enderbtw, i suggest that the splash colour is changed to white text on dark grey background, because (at least for me) it's pretty much unreadable with black text on light grey background (at least on my H120 light gray never was very light)
00:15:30linuxstbIsn't the problem that the database inserts tracks individually, rather than as "queries" ?
00:15:34amiconnAnd it was developed to outperform the OF's (Archos) playlist loading - by far
00:16:19Tornelinuxstb: ender said he was resuming playback, though
00:16:24JdGordon|linuxstb: that is definatly *a* problem... missed the begning of this convo though so dunno if its the same problem or not :)
00:16:45linuxstbJdGordon|: Yes, I wasn't paying attention either...
00:16:49enderlinuxstb: looks like just my playlist took a very long time to load
00:17:06linuxstbender: Was it created from the database?
00:17:23amiconnHmm, database could indeed be the problem
00:17:43enderno, just from me adding stuff from a few directories and stored .m3u files
00:18:05amiconnYet another entry for Mr. Someone's todo list - fix playlist creation from database
00:18:15*JdGordon| has never seen "loading xx%" splash for playlists
00:18:24JdGordon|amiconn: pretty sure thats there already
00:18:35 Quit DerPapst1 ("Leaving.")
00:19:16JdGordon|its a more general "make the database more like a psueo-filesystem" issue though
00:20:04pixelmalinuxstb: is this one correct ? And by the way, no pictures for you this evening then anymore... ;)
00:20:34enderyup, LCD_WHITE text on LCD_DARKGRAY is much more readable than LCD_BLACK text on LCD_LIGHTGRAY background on my H120
00:20:38kugelis a static global struct guaranteed to be zero'd at boot (with or without BSS_ATTR)?
00:21:27amiconnThere is no explicit BSS_ATTR
00:21:32Dgby714pixelma: heh my ipod lol
00:21:46amiconnAnd both bss and ibss are zeroed
00:22:01JdGordon|thats an assumption its better to not rely on
00:22:16kugelthat's why I put the word guaranteed into it
00:22:16linuxstbpixelma: That looks subtly different to my ipod... My top is thinner, and the hold button is also thinner. But I think I'm being far too fussy ;)
00:22:21amiconnIt's not an assumption
00:22:29 Join ifonefox [0] (
00:22:53pixelmalinuxstb: alrigjt
00:22:59pixelmaalright too
00:23:05Dgby714linuxstb: yeah the top is just a bit to thick in that pic
00:23:32linuxstbAlso, the bottom is the wrong way round - the dock connector is on the left, headphone on the right
00:23:39 Quit ifonefox (Remote closed the connection)
00:23:41kugelamiconn: so I can drop an init function that does nothing more than memset'ing it to zero?
00:24:04 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
00:24:34***Saving seen data "./dancer.seen"
00:24:37linuxstbkugel: You sure it will never need to be re-inited?
00:24:58kugelthere's a different function for that, which is still used
00:25:20amiconnJdGordon|, kugel: Our crt0.S zeroes bss and ibss on all targets, and there is a reason why it has to: gcc relies on it
00:25:53amiconngcc puts zero-initialized data into .bss instead of .data by default
00:26:22amiconnAnd that makes sense, as it saves binsize (not only on rockbox)
00:27:36kugelI saw our crt0s do that, I just wasn't sure if any static data is by default in bss or not
00:28:15amiconnIt depends on whether it's (nonzero) initialized or not
00:28:27amiconnIf you want to make it obvious that the code relies on that struct being zeroed, add a static initializer
00:28:37linuxstbkugel: If in doubt, look in
00:28:41amiconnThis won't change anything in the generated code
00:29:32amiconnJust (static) struct foo bar = {0};
00:29:58 Join mc2739 [0] (n=mc2739@rockbox/developer/mc2739)
00:30:02kugelthe other fields are zeroed, by standard, right?
00:30:43kugelI'll do that then
00:31:26 Join z35 [0] (
00:34:08froggymanis anybody aware that the picture for the iPod Nano 2G and Onda VX747 dont have pictures on the theme, and build pages?
00:34:09kugelamiconn: does that work for arrays of structs equally (static struct foo bar[MAX] = {0};)?
00:34:40froggymanis anybody aware that the picture for the iPod Nano 2G and Onda VX747 dont have pictures on the theme, and build pages?
00:34:51linuxstbfroggyman: Yes. (no need to ask twice...)
00:34:59froggymansorry for that
00:35:07*froggyman didnt do that on purpose
00:37:32 Quit ender (" I planted some bird seed. A bird came up. Now I don't know what to feed it.")
00:42:16amiconnHmm, good question. I think it will work though.
00:43:27kugelhrm, I somehow get a warning with that, although I did that trick before just fine
00:43:33 Quit intrados (Read error: 110 (Connection timed out))
00:43:59pixelmalinuxstb: I updated the png - same link
00:44:14 Join intrados [0] (n=intrados@
00:44:40 Quit HBK (Read error: 104 (Connection reset by peer))
00:45:03 Quit Thundercloud (Remote closed the connection)
00:45:56 Quit tomers (Read error: 113 (No route to host))
00:45:58 Join HBK [0] (i=hbk@
00:51:18 Quit HBK ()
00:51:53linuxstbpixelma: That looks great to me. Commit!
00:52:01kugelamiconn: is that incorrect? "static struct wps_state wps_state = { 0 };" ?
00:52:20kugelit must be something too obvious, I've been doing this before without problems
00:52:38 Quit Dgby714 (Read error: 113 (No route to host))
00:54:25 Join HBK [0] (
00:55:12*amiconn starts looking for the crystal ball
00:55:37kugelstatic struct wps_state wps_state = {.paused=0}; seems to work :/
00:55:44 Quit HBK (Read error: 104 (Connection reset by peer))
00:55:52kugelmaybe I did that with an array before :/
00:56:16Tornefor a struct you need to specify a value for each member
00:56:30Tornebeing able to initialise by name is a gnu extension
00:56:46Torneyou don't with an array (it fills in other elements after what you specify with 0)
00:57:00 Join HBK [0] (
00:57:34amiconnA struct initializer also fills missing elements with 0
00:59:10amiconnkugel: What warning do you get without specifying a member?
00:59:44kugelmissing initializer
01:03:17 Join fdinel [0] (
01:03:24 Join TheSeven [0] (n=theseven@rockbox/developer/TheSeven)
01:03:29pixelmalinuxstb: I'm too tired and try to commit in the morning - if not then tomorrow in the evening
01:03:56linuxstbpixelma: Very wise.
01:04:27linuxstbpixelma: Will you also be able to create the small image for the download pages, and the UI sim image?
01:06:50kugelbuggers, the sim segfaults (unmodified svn)
01:07:11*kugel blames the new gcc 4.4.1 that got installed today
01:08:04 Quit bertrik (Read error: 113 (No route to host))
01:08:14pixelmalinuxstb: small image is already done here. A sim image shouldn't be too hard because the labels are on a separate layer, you only have to use an overall image size so that the display will be correct
01:14:35 Quit TheSeven (Nick collision from services.)
01:14:52 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
01:15:04 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
01:17:25kugelor is it sdl ?!
01:18:04kugelarg, I can't compile a sim anymore
01:20:00 Part toffe82
01:20:35 Join Thundercloud [0] (
01:21:33 Quit linuxstb (Read error: 113 (No route to host))
01:22:02 Join linuxstb [0] (
01:24:58amiconnSim builds seem to be broken
01:26:20amiconnAt least crosscompiled win32 sims are. "/usr/bin/ld: cannot find -lmingw32"
01:26:29amiconnThis is the wrong linker...
01:26:35linuxstbMust be rasher's commit...
01:27:19 Quit polobricolo (Remote closed the connection)
01:27:28 Quit barrywardell ()
01:28:02kugelI'm compiling on linux
01:28:52amiconnMe too, but I'm building win32 sims
01:29:00 Quit Thundercloud (Remote closed the connection)
01:29:40amiconnlinuxstb: Actually it might not be broken when configuring new, but 'make reconf' breaks
01:29:59linuxstbI was about to ask if you had re-run configure...
01:30:19amiconnWell, 'make reconf' does re-run configure...
01:30:34amiconn(reconfiguring from a Makefile created before that commit)
01:30:45*amiconn faces reconfiguring 15 sim builds :\\
01:31:00linuxstbI assume that's not enough, when the method of configuring has changed...
01:31:52kugelI deleted the whole build dir already, no success
01:32:25amiconnA clear warning in the commit message would have been helpful...
01:33:30 Quit robin0800 (Read error: 110 (Connection timed out))
01:34:12 Join Strath [0] (
01:35:33elinenbe'ello all...
01:35:48n1sif people rely on that reconf thing we should introduce a version scheme for generated makefiles and refuse to run with an incompatible version instead of pretending all is fine
01:37:56 Quit TheSeven (Read error: 110 (Connection timed out))
01:38:31 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
01:39:18kugelhrm, the other sim build couldn't include SDL.h even after make dep
01:39:30kugelthere's something fishy on linux also
01:39:49amiconnSeems to work after manual reconfiguration
01:40:01linuxstbkugel: Build system doesn't seem to complain...
01:40:36kugellinuxstb: my first sime build didn't complain about SDL.h
01:41:17kugelmaybe it doesn't link correctly
01:42:23kugelit picks the wrong sdl on my system
01:44:40 Quit Strath ("Leaving")
01:44:52 Quit n1s ("Lämnar")
01:45:16 Quit stephen_ ("Leaving")
01:45:24kugelit picks ~/libsdl/bin, which isn'T even in my path!
01:45:47kugelthe correct libsdl is in ~/.libsdl/bin, as shown by 'which sdl-config'
01:45:59 Join Strath [0] (
01:46:44kugelstrange thing
01:47:21 Quit HBK (Read error: 54 (Connection reset by peer))
01:48:28 Join HBK [0] (
01:49:49StrathIt seems that obo's work on the Sansa View port of Rockbox has stalled... I would like to pick up the task but may need a guiding hand to get started. Would someone have a few minutes to spare?
01:51:59 Quit MethoS- (Remote closed the connection)
01:55:30 Quit matsl (Read error: 104 (Connection reset by peer))
01:55:32 Quit kugel (Remote closed the connection)
01:55:47StrathI've got the device, have read the wiki, have disassembled the firmware, have an svn copy of the Rockbox source and am currently trying to build the bit of Sansa View code that exists.
01:56:36 Quit Horscht (Read error: 110 (Connection timed out))
01:57:14 Quit rphillips (Client Quit)
02:02:49 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
02:04:53 Join TheSeven [0] (n=theseven@rockbox/developer/TheSeven)
02:05:43 Quit HBK (Read error: 54 (Connection reset by peer))
02:06:58 Join HBK [0] (i=hbk@
02:08:35 Quit intrados (Read error: 145 (Connection timed out))
02:14:03 Quit undersys (Read error: 110 (Connection timed out))
02:15:12 Join AndyIL [0] (n=pasha_in@
02:16:51 Join JdGordon1 [0] (
02:17:01 Nick fxb is now known as fxb__ (n=felixbru@
02:18:44JdGordon1arg, bugger..... the line number parser being off isnt what i tohght it was :(
02:20:16JdGordon1topik: i might need that bad wps...
02:21:51JdGordon1the %pb tag might make line nuumbers one too many
02:22:01 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
02:23:25 Join Mouser_X [0] (
02:24:35***Saving seen data "./dancer.seen"
02:27:07 Quit AndyI (Read error: 110 (Connection timed out))
02:29:20JdGordon1bugger! apparetly line counting is totally fubar... or wc -l lies
02:29:52 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
02:30:04 Quit HBK ()
02:33:27 Quit efyx_ (Remote closed the connection)
02:36:51StrathI modified tools/configure to remove the "not yet supported" exit. I chose (N)ormal and got errors about missing header files. I made a new build directory and chose (B)ootloader. it completed and I've got an firmware.mi4 and rockbox-info.txt file. I guess the next step is to somehow load that file on the target device but I would rather not brick it.
02:41:02midgeyStrath: i believe you copy the file down in Sansa USB mode and it should recognize it as a firmware "upgrade". If successfully, when you restarted the screen will display odd patterns or lock up.
02:41:25Strathand recovery?
02:41:44midgeyYou'll have to put the original Sansa firmware back on in recovery mode if you want to make it useable again
02:41:47 Join intrados [0] (
02:41:51JdGordon1which sansa?
02:42:39 Quit Llorean ("Leaving.")
02:47:53 Quit Mouser_X (Read error: 110 (Connection timed out))
02:50:39 Join saratoga [0] (i=98039f25@gateway/web/freenode/x-iehamqmuplsijjcc)
02:51:20JdGordonStrath: do you intend to actually try getting it working? or just want to test if for shits-n-giggles?
02:53:39saratogadid Obo get e200tool working on the View?
02:53:43saratogai don't think he did
02:54:00saratogathat might be a good place to start since you might need e200tool if you break your player
02:55:13 Join Strife89 [0] (
02:58:20 Quit amiconn (Nick collision from services.)
02:58:22 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
02:58:23 Quit pixelma (Nick collision from services.)
02:58:24 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
02:58:42 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
02:58:44 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
03:00:23midgeywhat did obo end up doing to his view?
03:00:59 Quit intrados (Read error: 110 (Connection timed out))
03:02:53Strathya... i'll have to check into e200tool... afk a bit
03:04:25 Join rphillips [0] (
03:05:09 Quit JdGordon1 (Read error: 110 (Connection timed out))
03:11:13 Quit kkurbjunW (Remote closed the connection)
03:11:31 Join kkurbjunW [0] (n=karlk@
03:15:04 Join hd [0] (
03:15:30 Quit kkurbjunW (Remote closed the connection)
03:15:51 Join kkurbjunW [0] (n=karlk@
03:16:29 Join HBK [0] (
03:21:33Strathintend to get it working...
03:23:02 Quit thegeek (Read error: 54 (Connection reset by peer))
03:25:12 Quit HellDragon (Read error: 110 (Connection timed out))
03:25:21StrathI had worked on porting to the Archos Gmini a few years ago (we got to the point of booting, lcd, and buttons, then mine got bricked and the product was discontinued. (and never sold enough units to build a mind share beyond the three or four working on it)
03:26:33saratogai vaguely remember hearing about that port
03:26:42saratogawell before my time anyway
03:27:16saratogathe view is a tough port though because its more undocumented PP hardware
03:29:02Strathat least with the View, the cpu is already supported by gcc and binutils
03:29:22Strathhad to start with an undocumented instruction set
03:30:48saratogathat is certainly a plus
03:39:25 Join intrados [0] (
03:41:05 Quit lennyk ()
03:43:00 Join perfectdrug [0] (
03:44:13perfectdrugI just made the svg for the nano2g FS #10675 :)
03:45:51perfectdrugthis was an easy one I hope it was still necessary
03:49:54 Quit midgey ()
03:54:11 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
03:59:59 Join goat [0] (
04:00:35 Nick goat is now known as Guest91961 (
04:01:12Guest91961seems rather empty
04:01:22Guest91961and it is!
04:01:23 Quit Guest91961 (Client Quit)
04:03:16 Quit perfectdrug ("CGI:IRC")
04:03:54 Join intrados1 [0] (
04:04:20 Quit saratoga ("Page closed")
04:05:31 Quit TheSeven (Nick collision from services.)
04:05:51 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
04:06:03 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
04:06:36 Part froggyman
04:14:30 Quit panni_ ("( :: NoNameScript 3.81 :: )")
04:15:55 Join dmb [0] (n=Dmb@unaffiliated/dmb)
04:16:38 Quit intrados (Read error: 110 (Connection timed out))
04:17:09 Join midgey [0] (
04:19:18 Join FOAD_ [0] (
04:20:41 Quit FOAD (Read error: 60 (Operation timed out))
04:20:41 Nick FOAD_ is now known as FOAD (
04:24:39***Saving seen data "./dancer.seen"
04:26:05LambdaCalculus37 <−− Rockbox running on the 2nd gen nano if any devs want to see.
04:29:25 Quit HBK (Read error: 110 (Connection timed out))
04:32:27JdGordonbad time to bring this up... but someone really needs to get non-hardware type #defines from target configs.... e.g HAVE_JPEG or HAVE_PITCHSCREEN
04:33:14LambdaCalculus37JdGordon: At least that task has been logged now. :)
04:34:08LambdaCalculus37Bagder: (for the logs) Feel free to blog and link to my photos.
04:34:20 Join tomers [0] (n=chatzill@
04:38:13 Quit Rondom (Nick collision from services.)
04:38:23 Join Rondom [0] (
04:44:16 Join tomers_ [0] (
04:50:35 Quit Strife89 ("Bed.")
04:51:03 Quit tomers_ ("ChatZilla 0.9.85 [Firefox 3.0.14/2009090216]")
04:52:42 Quit tomers (Read error: 145 (Connection timed out))
05:12:51 Join Llorean [0] (
05:12:53 Quit dmb (Read error: 104 (Connection reset by peer))
05:13:06 Join dmb [0] (n=Dmb@unaffiliated/dmb)
05:13:47 Quit LambdaCalculus37 ("Fwump")
05:25:42 Quit Strath ("Leaving")
05:26:08 Join HBK [0] (
05:27:42 Quit intrados1 (Read error: 110 (Connection timed out))
05:30:46 Join avacore^ [0] (
05:41:46 Quit avacore (Read error: 110 (Connection timed out))
05:45:18 Join KBH [0] (n=hbk@
05:46:31 Quit HBK (Read error: 60 (Operation timed out))
05:51:15 Quit Horschti ("Verlassend")
05:52:05 Quit dmb (Read error: 131 (Connection reset by peer))
05:52:59 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
05:55:29 Nick hd is now known as HellDragon (n=jd@Wikipedia/HellDragon)
06:08:24 Join intrados1 [0] (
06:17:39 Join FOAD_ [0] (
06:24:43***Saving seen data "./dancer.seen"
06:30:05FlynDicebertrik: kugel: (for the logs) thanks for your help with the callback idea, when I actually looked at it it made perfect sense. Still didn't solve my problem though...:(
06:30:51 Join Mouser_X [0] (
06:32:17 Part Mouser_X ("Leaving")
06:33:12 Quit FOAD (Read error: 110 (Connection timed out))
06:33:12 Nick FOAD_ is now known as FOAD (
07:11:44*pixelma reads about perfectdrug doing a Nano2G svg too...
07:13:07 Quit kkurbjunW (Remote closed the connection)
07:14:47 Join kkurbjunW [0] (n=karlk@
07:17:10JdGordonhey kkurbjunW, can you have a look at FS #10627 to make sure it doesnt break the beat f/x?
07:21:56 Join Grahack [0] (
07:24:02CIA-85New commit by pixelma (r23166): Full set of 2nd gen Nano images for a manual (svg, png and pdf).
07:36:53 Join docgnome [0] (n=docgnome@
07:36:55docgnomehey, what's a good media player that i can put rockbox on that is powerful enough to play ogg?
07:43:32midgeypretty much every player on the front page except the Archoses (which can only do MPEG audio and WAV)
07:50:23 Join FOAD_ [0] (
07:50:26docgnomeyeha i have an archos
07:50:29docgnomewhich sucks
07:50:37docgnomeits a giant brick and also kinna broken
08:01:16 Quit FOAD (Read error: 110 (Connection timed out))
08:01:16 Nick FOAD_ is now known as FOAD (
08:08:51 Join hd [0] (
08:19:27 Quit BHSPitMonkey ("Ex-Chat")
08:19:56 Quit HellDragon (Connection timed out)
08:21:13 Join gitster [0] (
08:24:27 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:24:28 Join Utchybann [0] (
08:24:44***Saving seen data "./dancer.seen"
08:25:49 Quit JdGordon ("Leaving.")
08:26:44 Join kugel [0] (n=kugel@rockbox/developer/kugel)
08:28:02 Join fyrestorm [0] (
08:29:33 Join Rob2222 [0] (
08:29:33 Join esperegu [0] (n=quassel@
08:34:04 Join ender` [0] (
08:38:08 Join JdGordon [0] (
08:40:11kugelamiconn: I read more about the warning yesterday, it apparently can be disabled by removing -pedantic (
08:40:29 Join gitster` [0] (
08:41:36JdGordonkugel: you get your mini2440 yet?
08:41:46 Quit fyrestorm (Read error: 104 (Connection reset by peer))
08:42:02 Join fyrestorm [0] (
08:42:07kugelJdGordon: no, the seller just cancelled it again
08:42:12 Quit gitster` (Client Quit)
08:42:18kugelI now order it for the third time :/
08:42:20JdGordonhaha lame
08:42:34JdGordonim not having much luck getting code running on it
08:42:41JdGordonor copying files onto it more accuratly
08:43:33kugelrockbox code or any code?
08:43:48JdGordonthe windows apps to upload code isnt great
08:44:39 Join FOAD_ [0] (
08:45:12gitsterHi, I have a fairly trivial fix for displaying %Sp in wps; it was incorrectly reporting that a normal playback pitch was 1000%. I threw it in to the FlySpray tracker as a patch (#10673). What do I do next? Just wait?
08:45:58 Quit FOAD (Read error: 60 (Operation timed out))
08:45:59 Nick FOAD_ is now known as FOAD (
08:46:49kugelamiconn: adding -Wno-missing-field-initializers seems to shut gcc of
08:47:07 Quit Rob2223 (Read error: 110 (Connection timed out))
08:47:34CIA-85New commit by dave (r23167): Nano2g image by Marianne Arnold
08:47:35 Join flydutch [0] (
08:47:49kugelJdGordon: :( we might want to ask bob how to do it
08:48:14kugelhe already runs rockbox on it ( Bob_C: ping )
08:49:34kugelJdGordon: here's the mutli aa patch kugel-rb.git?a=commitdiff;h=a0b2a207c4f2e1d3f9bfa8f9b93e4c8334461536;hp=288363b2389d5d965700a7018661605367c346db">;h=a0b2a207c4f2e1d3f9bfa8f9b93e4c8334461536;hp=288363b2389d5d965700a7018661605367c346db
08:49:54JdGordonput it on the tracker somewhere... Unhelpful said he'd have a looky
08:52:20 Quit JdGordon (Read error: 104 (Connection reset by peer))
08:55:28amiconnkugel: Hmm, disabling that warning in general is probably not a good idea
08:56:13linuxstbgevaerts: Could you add the Nano2G manual to the build system? It's now building.
08:56:13 Join JdGordon [0] (
08:56:27Bob_Ckugel: hello
08:56:34JdGordonBob_C: hey
08:56:35amiconnIf it doesn't warn if you just specify one (arbitrary) struct member, I think that is the better way
08:56:52JdGordonI got my mini2440 today... can you help me get the bootloader running on it?
08:57:16JdGordonpreferably in lniux...
08:57:19*amiconn wonders where else in rockbox this is used, as he's pretty sure it is
08:57:35kugelamiconn: not sure if using a gnu extension to avoid warnings on valid code is really the better way
08:58:00amiconnDesignated initializers aren't a gnu extension. Iirc they're just C99
08:58:36kugel{ .member = 0 } is a gnu extension
08:59:31 Quit midgey ()
08:59:38*linuxstb also thought it was C99
08:59:56Bob_Cjdgordon: Do you have a JTAG connection to the board?
08:59:57gitsterIt is C99
09:00:09kugelok, my mind is wrong then
09:00:17kugelor memory, rather :p
09:00:24JdGordonBob_C: yes, i was hoping to go over serial/usb though...
09:00:47gitsterIs RockBox gcc only, by the way?
09:00:57JdGordonthe jtag cable isnt long enough to be usable :p
09:01:58linuxstbgitster: Yes, I'm sure it is.
09:02:12gitsteroh, then it does not really matter if it is gcc extension or c99 ;-)
09:02:18linuxstbgitster: Although as this discussion shows, we try and avoid gcc extensions...
09:02:30amiconnAh, gcc doesn't warn if you do this for arrays (initializing not all elements), even without using designated initializers
09:02:42*amiconn remembered doing this somewhere
09:02:44 Join Thundercloud [0] (i=thunderc@
09:03:47amiconnlcd-player.c has static char icon_mirror[11] = {0}; and that obviously compiles without warning
09:04:35kugelamiconn: yeah, that confused me yesterday too
09:05:27Bob_Cjdgordon: I have the jtag adaptor on a 25 way cable
09:06:05 Join maruk [0] (
09:06:10Bob_Cjdgordon: The board should come with uboot in the NOR flash, this talks on serial @115k and can be used to flash the NAND
09:06:41Bob_CBut I overwrite uboot with the rockbox bootloader
09:06:55JdGordonhow do i connect to uboot in linux?
09:07:08 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
09:07:19Bob_CThe supplied JTAG adaptor conforms to a "Triton" model, this is supported by openocd
09:08:31JdGordonsounds like i need to go buy an extension cable for the parallel cable
09:08:53Bob_CI have been using windows, so I'm not sure, just need a terminal program to talk to com port
09:12:35tmztJdGordon: minicom should work
09:12:37tmztalso need lzrz or whatever it's called
09:17:39Bob_CI use the serial port for debugging, so its useful, but the bootloader has to be loaded via JTAG unless we find a way to use uboot
09:18:18tmztwhy won't uboot work?
09:23:05Bob_CIt was easier for me just to flash the rockbox bootloader into NOR and overwriting uboot
09:24:03Bob_CIt's a question of creating a rockbox build that will run from NAND, it's probably simple, I just didn't look into it
09:26:42linuxstbHow is the NAND attached? Will you need a software FTL?
09:27:44Bob_CThe s3c2440 has a NAND controller, it can boot from NAND by copying 4K into internal RAM
09:28:35Bob_CWe discussed NAND here and I was persuaded not to bother with it ;)
09:32:14 Quit kugel (Remote closed the connection)
09:34:13 Nick hd is now known as HellDragon (n=jd@Wikipedia/HellDragon)
09:36:10 Quit JackWinter2 (Read error: 54 (Connection reset by peer))
09:36:26 Join JackWinter2 [0] (
09:39:15 Quit gitster (Read error: 104 (Connection reset by peer))
09:39:28 Join gitster [0] (
09:41:32 Join lucent [0] (i=lucent@unaffiliated/shadows)
09:44:35 Quit Thundercloud (Remote closed the connection)
09:45:22 Quit fyrestorm (Read error: 54 (Connection reset by peer))
09:45:22 Quit advcomp2019 (Read error: 54 (Connection reset by peer))
09:45:31 Quit ender` (Read error: 54 (Connection reset by peer))
09:45:33 Join fyrestorm [0] (
09:45:40 Quit DataGhost (Nick collision from services.)
09:45:42 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
09:45:48 Join DataGhost [0] (i=dataghos@unaffiliated/dataghost)
09:46:04 Join ender` [0] (
09:52:59 Quit w1ll14m|_ (Read error: 60 (Operation timed out))
09:54:51 Join w1ll14m| [0] (
09:56:36 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
09:57:12linuxstbZagor: Hi. Could you update www? There's now a thumbnail image for the nano2g in svn.
09:57:39Zagordone even :)
09:58:35linuxstbWhat's needed for daily builds for the nano2g?
09:59:10linuxstbAnd I also noticecd that an old is in the build server's download directory (it was renamed to
09:59:20linuxstbAnd the fuze link on the daily builds looks broken...
09:59:25*linuxstb hands Zagor some coffee
09:59:25Zagorlinuxstb: add it to www/
10:00:56CIA-85New commit by dave (r23168): Add ipodnano2g
10:01:16Zagorand there it is!
10:01:32Zagoruh no, I looked at the mini
10:01:53Zagorit is present in the current builds table though
10:02:11linuxstbThere's no zip there either. I guess that needs to wait until tomorrow?
10:02:28linuxstbAh no, I got the URL wrong - it's there.
10:02:54Zagorright, the current is there. the daily is updated early morning.
10:03:22linuxstbNo, I mean the daily is there -
10:03:31Zagorwww/ is updated to add daily builds. (yes this needs to be untangled)
10:03:48*linuxstb shrugs -
10:04:03linuxstbCould it simply be using "builds" now?
10:04:48linuxstbSo I should add ipodnano2g to www/ as well?
10:05:16CIA-85New commit by dave (r23169): Add ipodnano2g
10:05:27Zagor"builds" doesn't say which builds should be in which table. we need another list for that. or possibly a single master list that replaces all others.
10:05:49Zagorthere we go, daily nano2g
10:06:47CIA-85New commit by dave (r23170): Correct the Fuze entry - it was renamed to sansafuze
10:08:28linuxstbWhat about the manual build for the nano2g? It now (as of about 3 hours ago) builds.
10:08:35linuxstb(for the nano2g)
10:09:55ZagorI've added manual and voices to the nightly build
10:12:29linuxstbOK, thanks.
10:13:16linuxstbAll we need now is a new rbutil release...
10:15:04 Quit fyrestorm (Read error: 110 (Connection timed out))
10:16:49 Join pyro_maniac1 [0] (
10:17:59 Quit KBH (Read error: 131 (Connection reset by peer))
10:23:26linuxstbrasher: I see the theme site is now working for the nano2g - thanks. Are you able to change the names of the two Nanos to "iPod Nano 1G" and "iPod Nano 2G" ? (or change all ipods to use "1st gen" etc) ?
10:24:46***Saving seen data "./dancer.seen"
10:27:44 Join n1s [0] (n=n1s@rockbox/developer/n1s)
10:27:45 Join petur [50] (n=petur@rockbox/developer/petur)
10:27:58 Join einhirn [0] (
10:37:03 Join lifeless_ [0] (n=lifeless@
10:40:08 Quit lifeless_ (Client Quit)
10:42:15 Join thegeek [0] (
10:55:12topiknice that the nano2g theme page now only says 'current build' and not 3.4
11:00:30 Join robin0800 [0] (
11:07:02 Quit TheSeven (Nick collision from services.)
11:07:19 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
11:07:30 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
11:07:49 Quit daurn| (Read error: 148 (No route to host))
11:08:45 Quit TheSeven (Nick collision from services.)
11:09:04 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
11:09:16 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
11:14:34 Quit kkurbjunW (Remote closed the connection)
11:15:10 Join kkurbjunW [0] (n=karlk@
11:24:33 Quit Sajber^ (Connection reset by peer)
11:25:23 Join Sajber^ [0] (
11:29:51 Join Dgby714 [0] (
11:51:06 Join gtkspert [0] (
11:52:11 Quit TheSeven (Nick collision from services.)
11:52:28 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
11:52:40 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
12:02:57 Quit lifeless__ (Remote closed the connection)
12:03:19 Join lifeless_ [0] (n=lifeless@
12:04:27 Join esperegu_ [0] (n=quassel@
12:06:02 Quit gtkspert_ (Read error: 101 (Network is unreachable))
12:08:37*linuxstb reads and wonders if there are any projects we could "partner" with - i.e. promote their PC app on our site and recommend it for use with Rockbox, and they add good support for Rockbox to their app, and recommend people run Rockbox on their DAPs...
12:08:45linuxstbi.e. compete with the "itunes + ipod" culture...
12:10:14*linuxstb has never heard of it
12:10:18tmztwrong name
12:10:34tmztit's .net with coming windows port
12:11:13topikwouldn't any app that supports msc work with any rockboxed player that has usb storage?
12:11:16*linuxstb thinks there's going to be a contradiction though - rockbox aims to be lean and mean, PC-based media players don't...
12:11:47linuxstbtopik: Sure. But the point is to make things easier for a user - in the same way Apple do with ipod+itunes integration.
12:12:32linuxstbi.e. a "supported" combination of PC-app and Rockbox.
12:12:58topikyou surely don't want that pc app update rockbox's database and such
12:13:15linuxstbOf course - that's long been a desired feature.
12:13:23GodEaterwe surely *do* want that
12:13:29linuxstbBut the difference is that the PC app is completely optional.
12:14:08GodEaterthe problem I see with going down this route is the massive choice there is in PC apps, and the fact that everyone will have their own favourite.
12:14:17GodEaterI personally can't stand banshee for example :)
12:14:40topikall pc apps i know (except itunes) will remember what it synched on its own. unrelated to the player/firmware used
12:15:52linuxstbtopik: But there are some Rockbox-specific features, such as knowing what audio formats the device can play.
12:16:03n1sGodEater: that is easily narrowed down if we require the media player to run on linux, windows and macos
12:17:07topiktrue. but a fair amount of players don't have usb without rebooting outside of rockbox
12:17:44n1stopik: that's not true
12:17:54topikall i have do
12:17:57topikfor now
12:18:07n1swhich do you have then?
12:18:09topikperhaps it's the immaturity of those targets
12:18:21topikboth very fresh
12:18:27n1syeah, those are very new
12:18:29tmztof course I took the selection criteria to be iTunes-like
12:18:42tmztand Banshee is probably the closest
12:19:01topikif you want it user-friendly, perhaps MTP support would be nice
12:19:07linuxstbtopik: What does the USB mode have to do with anything? A PC app can just look to see if Rockbox is installed on the device, and act accordingly.
12:19:08tmztwith tocuhscreen comes more consumer focus? (excepting the accessible version for a moment)
12:19:35n1sall the "stable" ports have usb in rockbox, except for release versions for ipods in which it's disabled because of the charging thing
12:19:54topiklinuxstb, i am biased by the limited exposure to mature targets
12:19:56 Quit esperegu (Read error: 113 (No route to host))
12:20:11tmztand Banshee would allow plugins, if the interface was fully documented others could create their own plugins
12:20:26 Join Topy [0] (
12:20:40n1salso early ipods don't have usb at all so it's impossible on them :)
12:20:43topikare there any stats on which targets gets the most downloads?
12:20:46linuxstbtopik: But the "reboot into disk mode" doesn't really change much. You insert the USB cable, you enter disk mode. You remove the cable, you're back in Rockbox. That's how the Nano2G works for me.
12:21:22tmztand MTP makes copy and unplug doable, storage will always require a flush and eject
12:21:36topiktrue linuxstb, but then the pc-app would 'detect' by looking for a .rockbox dir or something?
12:21:54linuxstbtopik: Yes. It would have to do that on most targets anyway (USB is hardware-controlled).
12:22:01tmztactually, it's possible to support virtual media change/eject
12:22:26tmztpress select when you are done copying
12:23:19gevaertsvirtual media change/eject?
12:23:23tmztand the OS will see eject request and unmount cleanly, assuming mass storage supports tge scsi commands required
12:23:25topikwhen i look at the official sansa (fuze) forum, it amazes me how many people use software like rhapsody and wmp
12:23:50*linuxstb thinks about it, and concludes that he would quite like MrSomeone to integrate Rockbox's playback engine into rbutil...
12:24:19tmztyeah, I'm sure thing like jaz drives needed it
12:24:47***Saving seen data "./dancer.seen"
12:24:49tmztI guess it would only work on soft usb targets
12:26:29topiklinuxstb, would that come with a rbutil simulator too?
12:26:35 Nick fxb__ is now known as fxb (n=felixbru@
12:26:51*gevaerts still has no idea what tmzt is talking about
12:28:13topikunplugging/replugging through software instead of physically
12:29:25tmztwell, nand is obviously not removable media
12:29:41tmztbut there's no reason it couldn't be made to appear to be
12:29:49gevaertsyes there is
12:30:01tmztthink of it as a non-optical drive with eject button
12:30:04n1slinuxstb: to use as a mediaplayer or just for fun? :)
12:30:05gevaertsif you do that, multiple partitions don't work in windows
12:30:26linuxstbn1s: Yes, to use as a media player... I've yet to find one I like.
12:30:34tmztoh, that includes on sd card readers?
12:30:53gevaertsI suspect so, yes
12:31:13n1slinuxstb: it will be a strange beast indeed :)
12:31:18linuxstbn1s: Although to be honest, I think I would probably prefer a command-line player - I mainly just want to be able to do "play *.flac" in a directory, but have the ability to seek/pause/skip etc.
12:31:33tmztI'm quite sure I have used partitioned sd cards with usb readers
12:31:37linuxstbSomething like the Player sim ported to curses...
12:31:46tmzteven multilun devices
12:31:55gevaertsmultilun has nothing to do with this
12:32:12tmzteven on multi lun devices
12:32:18tmztyes, I know
12:32:30gevaertsactually, rockbox presents itself as a removable-media device on players where we don't need multiple partitions
12:32:33tmztwhat is the issue? is there a kb document or something?
12:32:42gevaertstry it
12:33:11gevaertsdid that card reader show itself as removable?
12:33:20tmzthow exactly?
12:33:37tmztnot sure, I think eject worked (not remove device)
12:33:55gevaertsin rockbox, by editing usb_storage.c
12:34:05tmztof course since led shut off I have no idea if firmware shut itself down
12:34:12tmztdon't have usb target currently
12:34:21tmztonly Fuze
12:34:55tmztok, on those devices, any reason why "click select to safely remove" wouldn't work?
12:35:27 Join Topy44 [0] (
12:35:58preglowZagor: still no mails here, btw
12:37:42Zagorpreglow: we get odd bounces from you. I have cleared the bounce flags on your account several times
12:37:55 Quit T44 (Read error: 110 (Connection timed out))
12:38:24tmztgevaerts: interesting
12:38:48tmztit's only microsoft management console that breaks?
12:43:28 Nick YPSY is now known as Ypsy (n=ypsy@
12:50:51 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
12:52:38 Quit Topy (Read error: 110 (Connection timed out))
13:04:31 Join T44 [0] (
13:09:24 Join sansauser [0] (
13:09:45sansauserhey guys, any known issues with rockbox utility under win7?
13:10:38sansauserive got a sansa c200 v1 here, and i cant install the bootloader, and rbutil doesnt find it, yes its in MSC/UMS mode
13:10:47 Join pamaury [0] (n=pamaury@
13:11:00sansauseri cant seem to start it in recovery mode either, c200 has no record button
13:11:57pamaurySlasheri: ping
13:12:12gevaertssansauser: the c200 definitely does have a record button
13:13:22sansauserwell i cant find it on it, theres volume buttons, play, next, rewind, context?, and menu/power, and a hold on the underside
13:14:12 Quit avacore^ (Read error: 60 (Operation timed out))
13:15:03 Join avacore [0] (
13:15:12sansauserope, no button at that place on mine O_o only the hold switch
13:15:43linuxstbsansauser: Are you sure it's a c200 then? If Rockbox doesn't install, and it doesn't match the description in the manual...
13:15:47sansausermine has a ring and not a square
13:16:17 Join polobricolo [0] (
13:16:48 Quit Topy44 (Read error: 104 (Connection reset by peer))
13:17:00linuxstbsansauser: Does your player look like ?
13:17:54sansauser the one on the right
13:19:09 Join DerPapst [0] (
13:19:40linuxstbThat looks like a Sansa c100
13:19:53pixelmathat's a c100, the c200 would be the one in the middle
13:20:31sansauserah well
13:20:51sansausergot it for free anyway
13:20:55*pixelma wonder where MarcGuay went
13:20:59pixelmawonders too
13:21:01sansausermy e200 has died
13:21:22sansauserhardware failure i think
13:21:35sansauserwont even boot up, not recognisable on the pc at all
13:21:51gevaertscould be a loose flash board
13:22:06sansausernah, checked
13:22:19linuxstbHave you tried entering manufacturing mode?
13:22:26sansauseropened it up alot, needed to resolder the headphone jack, lol
13:22:39Dgby714sansauser: ive got no problems in Win7 except when closing the program.
13:23:22 Quit elinenbe (" HydraIRC -> <-")
13:23:25linuxstbDgby714: BTW, does the themes site work OK with your rbutil from svn?
13:23:29linuxstb(for the nano2g)
13:23:34Dgby714ill check
13:24:20linuxstbAnd from tomorrow (I think) the manual should be working.
13:25:35Dgby714yeah it works just had to clear local cache
13:25:58gevaertsCan a theme site admin have a look at the Spartan Black theme for e200? It's not actually an e200 theme
13:26:32 Quit lifeless_ (Remote closed the connection)
13:26:35 Join lifeless__ [0] (n=lifeless@
13:26:38 Quit avacore (Read error: 60 (Operation timed out))
13:27:27sansauserwell atleast i have a mp3 to go, so thanks guys
13:27:27 Join avacore [0] (
13:27:42pixelmagevaerts: it's not only included there falsely
13:27:44sansauserill return f i feel like trying to get the e200 back to life later
13:28:29 Quit sansauser ("CGI:IRC (EOF)")
13:29:36pixelmagevaerts: and the most weird thing it is sometimes wrong and on other pages of targets with the same screen size it's not (e.g. e200 and e200v2) maybe it's related to whether it was already there when the subpage was created or not
13:31:41gevaertspixelma: maybe checkwps is involved? Spartan Black is said to work with 3.2 or current on e200, not with 3.4
13:35:24 Quit pamaury (Remote closed the connection)
13:43:32linuxstbgevaerts: The theme itself is stored in a "220x176" directory, so would have thought the theme site shouldn't have linked to it for a 176x132 target, regardless of checkwps result.
13:50:02 Join pamaury [0] (n=pamaury@
13:50:21 Quit T44 (Read error: 110 (Connection timed out))
13:52:15Unhelpfulkugel: couldn't i just pull one of your branches? i assume multi_aa_2?
13:54:39 Quit DerPapst ("Leaving.")
13:54:52 Quit evilnick (Read error: 113 (No route to host))
13:56:13 Join syrius [0] (n=syrius@unaffiliated/syrius)
13:56:24syriuswill rockbox ever support a port to the psp?
13:56:54LinusNsyrius: well, rockbox-as-an-app could certainly work on the psp, if someone did the work
13:58:19 Join barrywardell [0] (n=barrywar@rockbox/developer/barrywardell)
13:58:30syriusit would be cool if rockbox did so
13:58:54 Quit JackWinter2 (Remote closed the connection)
13:59:05LinusNbut in the psp case, perhaps it would be better to borrow selected parts of the playback engine+codecs and make a psp-specific media player instead
13:59:39LinusNi mean, many of the emulators and stuff already have better native psp versions
14:00:22LinusNrevamp lightmp3 and have it use the rockbox playback engine
14:00:49Torneer, part of the point of lightmp3 is that it can use the second proc for decoding using sony's codecs though
14:01:23 Join JackWinter [0] (
14:01:35Torneusing the PSP efficiently requires that you work the way the SDK expects you to, mostly, so I dunno how applicable our codecs/playback code would be...
14:05:00 Quit Rob2222 ()
14:05:55 Quit JackWinter (Read error: 104 (Connection reset by peer))
14:06:14LinusNTorne: i'm afraid so
14:06:49LinusNTorne: however, didn't lightmp3 support flac as well?
14:07:01LinusNor was it aac?
14:07:03Torneyup, the ME can't play every codec. it does those the traditional way
14:07:12Tornei.e. leaves the ME idling
14:07:18Torneand does it on th emain cpu, sucking up way mor epower.
14:07:20preglowZagor: weird, what kind of bounces?
14:07:40Torneif you wanted everything to work nicely you *would* need to port rockbox as a firmware
14:07:55Tornebecause within the PSP OS you can't run arbitrary code on the ME, but in general it's pretty much just another MIPS core
14:08:29 Join JackWinter [0] (
14:09:50syriusTorne, is there any advantage to working to make the psp ported rockbox by being on a mips arch?
14:10:08syriusthere is a cheap box you can get for that 240 dollars
14:10:18syrius190 usd + 50 for shipping
14:10:20Torneer, not sure quite how to parse that
14:10:57syriuswell I mean if you where developing the rockbox firmware on a mips arch computer Torne
14:10:59tmztwhat is ME?
14:11:05syriuswindows ME?
14:11:09Tornetmzt: media engine
14:11:14Tornethe PSP's second processor
14:11:17tmztwhich is?
14:11:26tmzta second mips or something else?
14:11:30Torneit's a dual core MIPS similar to the way that PP is dual core arm7
14:11:45Tornebut the PSP OS uses it solely to run Sony's codecs for audio/video
14:11:55Tornesyrius: portalplayer, the chip in the older ipods and various other DAPs
14:12:11syriusyes I know that the ipod uses arm arch
14:12:21syriusI didn't know the psp used mips though
14:12:29tmztbut it can draw on framebuffer or you have to shuffle screens somehow to the main processor application?
14:12:29Tornesyrius: anyway, no, there is no need to be using a MIPS computer to develop for PSP
14:12:32 Join barrywardell_ [0] (
14:12:36syriusit seems they all use all different kinds of processor types
14:12:50syriuslike ps3 and xbox 360 use powerpc
14:12:52Tornetmzt: I don't know exactly what the limits are. All software for the PSP is basically developed for the existing Sony kernel
14:13:04 Quit flydutch ("/* empty */")
14:13:05Zagorpreglow: I'm not sure. the bounce notification I get strangely shows the outgoing mail as bounce notice
14:13:05tmztyes, but those aren't handheld consoles
14:13:08Tornetmzt: which doe snot let you run code on the second core at all, other than the code Sony provide
14:13:12tmztand ps2 used mips I thought
14:13:39syriuswell I know you can develop it on a non mips computer I was asking if there was an advantage to doing it on one Torne
14:13:44Tornesyrius: no.
14:13:50tmztis the dsp image (for lack of a better term) uploaded at run time?
14:14:12syriusI know somethings about compiling
14:14:12Tornetmzt: i believe it's just part of the firmware, i donm't think it's uploaded anywhere
14:14:16tmztis it a smp system where both cores boot the same firmware loader?
14:14:18Tornei don't think the ME has its own memory or anything.
14:14:36Tornei'v eonly done a tiny amount of PSP development though, so i can't really tell you any of these details
14:14:38syriusthat is what makes linux and free/open source software is that you can make it for many processor types
14:14:50Tornewhat's *visible* from the SDK is that the ME is a black box accessed solely via OS calls
14:14:50tmztif it doesn't have it's own (protected) memory, what's the issue?
14:15:05Tornetmzt: the issue is thta the OS expects to be managing it
14:15:18Tornemaking it do something else probably requires ditching the OS
14:15:23tmztthrough mailboxes?
14:15:30Bob_CThere is a #define called GIGABEAT_F and one called TOSHIBA_GIGABEAT_F, are these interchangeable?
14:15:30Tornei have no idea.
14:15:45tmztI haven't either, only worked with qualcomm msm phones with similar concept
14:15:50Tornebut there's been homebrew development for many years on the PSP
14:15:54syriuslike with debian if it has the source in the repository you can get the source the debian way and set it to compile for ur arch
14:16:11Torneand very few apps use the ME for anything other than mp3 decoding
14:16:20Torneand the methods involved are not clear or documented :)
14:16:26tmztsyrius: right, and rockbox is mostly c and already supports multiple architcecthures
14:17:14tmztif View ever gets very far it will be interesting to do dsp stuff on it, I had one for a while but had to return it
14:17:49syriusis that for text files
14:17:52tmztof course, you are saying psp just has a second mips core, not a dsp
14:18:00tmztSansa View
14:18:06syriuswhat is that?
14:18:24Torneyeah, it is just a second MIPS, but I'm not sure whether it's really SMP or whether the second one has different peripheral access.
14:18:40Tornealmost all homebrew dev is done with an open SDK that mirrors the capabilities of the official one
14:18:46Torneso there's very little easy to find docs on these details
14:19:45tmztsyrius: it's an mp3 player only in New Ports currently and barely running code
14:19:49 Quit barrywardell (Read error: 145 (Connection timed out))
14:19:49 Nick barrywardell_ is now known as barrywardell (n=barrywar@rockbox/developer/barrywardell)
14:20:32 Quit antil33t (Read error: 131 (Connection reset by peer))
14:20:38 Join antil33t [0] (n=Mudkips@
14:21:33 Join flydutch [0] (
14:22:27preglowZagor: i use a redirect, nothing wrong with those of late?
14:22:39preglowif not, i have no idea what is wrong, my ordinary mail works just fine
14:23:31Zagoraha, looks like the aliases are broken somehow
14:24:08 Join midgey [0] (
14:24:49***Saving seen data "./dancer.seen"
14:29:11pamaurySlasheri: ping
14:29:23 Join linuxguy4 [0] (
14:30:46Zagorpreglow: fixed now. sorry about that.
14:30:50preglowZagor: no worries
14:32:41 Quit linuxguy3 (Read error: 145 (Connection timed out))
14:35:32 Quit Sajber^ (Read error: 104 (Connection reset by peer))
14:52:39Slasheripamaury: hello
14:53:16pamaurySlasheri: hello, I have some questions about dircache, do you have time to discuss about it ?
14:53:34Slasheripamaury: please ask :) i just came from work
14:54:43 Join Omlet [0] (n=omlet05@
14:54:52pamauryok, that's about the implementation of dircache_remove. I don't understand why it's implemented like it currently is. This is because I'm working with the MTP protocol and directly accessed the dircache structure.
14:54:57 Join Omlet05 [0] (n=omlet05@
14:55:47pamauryCurrently the implementation only set d_namelen to 0 without changing any pointer and without resetting all other field. Which means that from nearly all points of view, the entry remains valid and this is a pain.
14:56:49linuxstbSlasheri: BTW, related question, but I noticed that bootloaders (or at least the nano2g bootloader) has a 150KB "opendirs" buffer. Any idea what that's for?
14:57:11Slasheripamaury: hmm.. that was probably the simplest way to "remove" an entry from the cache, because the entry has to remain there until the cache is rebuilt
14:58:03Slasheripamaury: i don't remember how it was implemented excactly, but if those pointers don't refer to anything, i don't see why it wouldn't be possible to destruct those also
14:58:28Slasherilinuxstb: hmm, that was dircache related also?
14:59:25linuxstbIt's directory-related... ;)
14:59:26pamauryI agree it's the simplest because only single linked list are implemented bu anyway it should be possible to skip the element by walking though the list. anyway I think that declaring an element is invalid just because "d_namelen=0" is quite error-prone, no ?
15:00:25Slasheripamaury: hmm.. i will check how that was implemented, just a moment :)
15:01:36Slasheriah, ok.. indeed, that is quite simple implementation
15:01:52 Quit syrius (Read error: 110 (Connection timed out))
15:02:18Slasherii don't remember if those pointers should remain valid
15:03:58Slasherione way would be to extend the list to two directional linked list (and increase ram consumption a bit)
15:04:08Slasherithen it would be simple to invalidate those pointers
15:04:15pamauryIf I'm correct, dircahe is only a tree in a fixed length memory block so modifying the pointers shouldn't be a problem. The only thing is tht deleting an element is more complicated but it's doable
15:04:33Slasherinope, it's not fixed length
15:04:53pamauryit's allocated at boot time
15:05:02pamauryno ?
15:05:28Slasheriyes, but filename stored in the middle of those pointers is variable length allocated
15:05:59Slasheriah yes, the whole dircache block is fixed length indeed
15:06:23 Join DerPapst [0] (
15:13:05pamauryso, do you think it's possible/allowed/not dangerous to modifiy the pointers ? Or should the code stay as it is ? It's because it's weird to have such a convention on valid entris and having it written nowhere.
15:13:05 Join Topy44 [0] (
15:16:33 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
15:16:39 Quit kkurbjunW (Remote closed the connection)
15:17:06 Quit dfkt (Nick collision from services.)
15:17:10 Nick dfkt_ is now known as dfkt (i=dfkt@unaffiliated/dfkt)
15:17:17 Join kkurbjunW [0] (n=karlk@
15:18:06 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
15:19:25Slasheripamaury: at least database stores those pointers internally so you should check it could handle those modifications
15:20:31 Join jgarvey [0] (
15:22:59 Quit midgey ()
15:25:10 Quit DerPapst ("Leaving.")
15:29:44 Join T44 [0] (n=Topy44@
15:34:37 Join mick3 [0] (
15:34:50 Quit mick3 (Client Quit)
15:36:20 Join dfkt__ [0] (i=dfkt@unaffiliated/dfkt)
15:36:27 Quit TheSeven ("ChatZilla 0.9.85 [Firefox 3.5.3/20090824101458]")
15:36:30 Quit dfkt (Read error: 110 (Connection timed out))
15:36:49 Quit dfkt_ (Nick collision from services.)
15:36:56 Nick dfkt__ is now known as dfkt (i=dfkt@unaffiliated/dfkt)
15:37:35 Quit Omlet ("( :: NoNameScript 4.22 :: )")
15:37:43 Quit Omlet05 ("( :: NoNameScript 4.22 :: )")
15:48:26 Quit Topy44 (Read error: 110 (Connection timed out))
15:48:32 Join kyle6513 [0] (n=kyle6513@
15:52:59 Join evilnick [0] (i=0c140464@rockbox/staff/evilnick)
15:58:03 Join DerPapst [0] (
16:01:56 Quit barrywardell ()
16:02:57 Join dmb [0] (n=Dmb@unaffiliated/dmb)
16:10:04 Join barrywardell [0] (
16:10:42 Join faemir [0] (n=faemir@
16:11:59 Join barrywardell_ [0] (
16:12:00 Quit barrywardell_ (Client Quit)
16:16:56 Quit antil33t ()
16:18:24 Join funman [0] (n=fun@rockbox/developer/funman)
16:24:50***Saving seen data "./dancer.seen"
16:28:44pyro_maniac1funman: i have to bother you again. there are still some things missing for the samsung devices
16:29:27pyro_maniac1funman: there are no themes for yh920 and yh820
16:30:41 Quit barrywardell (Read error: 113 (No route to host))
16:32:47 Join toffe82 [0] (n=chatzill@
16:32:52 Join enthdegree [0] (
16:33:17funmanpyro_maniac1: the screen is the same (size, depth, format) than h100, h120 and iaudiom5 so all the themes are compatible
16:33:33funman(i only checked yh920
16:34:57pyro_maniac1i didn't know that. will rbutil select from screen size automaticaly?
16:35:00 Join MethoS- [0] (n=clemens@
16:35:14funmanno idea
16:35:48 Join Rob2222 [0] (
16:35:53enthdegreeWhat framerate can a c200v1 playback videos?
16:36:16funmanenthdegree: it depends on the resolution and bitrate i would say
16:37:09funmanpyro_maniac1: the yh820 lcd is unique however
16:37:19enthdegreeWill rockbox ever be able to play other video formats? (say, something like mp4?)
16:37:48 Part LinusN
16:38:03funmanif you write a decoder for it, sure
16:38:09pyro_maniac1enthdegree: only if somebody spend some time and make it work.
16:38:50enthdegreeBut the devices do have enough processing power, right?
16:39:09Tornesome of them may do
16:39:14Tornesome of them definately do not
16:39:35LloreanMost of them are unlikely to be fast enough to get much useful speed out of h.264 (which is one of the most common formats in an MP4 container)
16:39:42Torneonly way to know for sure is to implement it and test it.
16:39:58funmansome even have a DSP to decode H264 (like ipod video) but this hardware isn't documented at all
16:40:14 Nick fxb is now known as fxb__ (n=felixbru@
16:40:17enthdegreewow, that stinks.
16:40:37Torneyah, some of hte OFs rely on hardware video decoding
16:41:19funmanpyro_maniac1: you should ask to domonoky or bluebrother about rbutil
16:42:30pyro_maniac1i will have a closer look later and maybe bother somebody later ;-)
16:43:00funmanand for yh820 you should ask the artists ;)
16:45:43 Part pyro_maniac1 ("Leaving.")
16:46:22enthdegreeWould h.264 be harder to decode (take more processing power) than mpeg-1/2?
16:47:49funmanenthdegree: yes
16:49:48 Quit Horscht ("Verlassend")
16:50:11Torneit's far harder for the same resolution
16:50:22Torneand h264 stuff tends ot be at a higher resolution as well to make it worse :)
16:50:37 Quit kyle6513 (Read error: 110 (Connection timed out))
16:51:13enthdegreeMaybe it'll make a good code project for me in the future. {c:
16:53:25 Join saratoga [0] (i=9803c264@gateway/web/freenode/x-wuesxjchaiehyyat)
16:55:28 Quit polobricolo (Remote closed the connection)
16:55:55LloreanTorne: re track order in the database - while it should be inserted in the order displayed, I *think* if you choose a category and "insert" it you get a default sort rather than the sort it would display if you enter it (I know it had this problem in the past for me, I'm not sure these days)
16:57:53 Join TheSeven [0] (n=theseven@rockbox/developer/TheSeven)
16:59:22 Quit bubsy ("Party time, EXCELLENT! *wring wring wring!!* Party on, Wayne! Party on, Garth!")
17:00:15 Join mcuelenaere [0] (
17:08:27 Nick janne_oksanen is now known as new_york_times (n=janne@
17:09:47 Nick new_york_times is now known as janne_oksanen (n=janne@
17:13:47 Quit w1ll14m| (Read error: 110 (Connection timed out))
17:16:42Unhelpfulbeast almost certainly could, especially if we can make use of the IPU for some of the filtering work
17:18:18 Quit DerPapst ("Leaving.")
17:18:47funmanI believe some mpegplayer code could be factorized to be reused by other video decoders
17:20:03 Join ballubasch [0] (
17:20:38ballubaschhi there
17:21:15 Quit liar_ (Remote closed the connection)
17:21:45ballubaschwell I just post my question, maybe anyone can help..
17:21:53Unhelpfulkugel: why size_func? why not a struct dim?
17:23:15 Join HBK [0] (
17:24:01Unhelpfulother than that, i don't have any problems with the parts of this that are in areas i know well. i would suggest trying a struct with dimensions in place of a get-size callback, this *might* give you a smaller binsize hit and will reduce call depth by one :)
17:24:27ballubaschi'm trying to flash my IRiver H120 to bootloader 7pre4 to use a CF card, because my Harddisk broke... did it all like the Wiki said, bit the flashing seems to be stuck with "ROLO... Executing"
17:24:58ballubaschhow long should this take, it's about 15 mins here
17:28:44funmanI don't have a h120 but I expect it to be almost instant
17:29:11 Join Lss [0] (
17:29:17ballubaschthank you... I guess i shouldn't turn it off now..?
17:29:25funmanso there is a problem, not sure what you should do (I don't know if irivers can be bricked or not)
17:30:36rasherThey can
17:31:03ballubaschit can be, unfortunately, I read in the forums that IRivers being stuck, and people don't dare to turn it off for hours
17:31:20ballubaschbut I'm not shure if this was during bootloader update
17:31:25Slasheriballubasch: you shouldn't "run" the flash image (i.e., short press)
17:31:35Slasheriballubasch: long press on the flash image and then select iriver_flash
17:32:22ballubaschyes, I knew that, strangely the "open with" menu didn't show up
17:32:25Slasheriif it says "ROLO..", just reset the thing
17:32:47Slasheriyou have to press the joustick longer
17:32:47ballubaschaahhh... Thanks
17:33:20ballubaschmaybe I slipped off and pressed again, wondered myself
17:33:32Slasherijust try again :)
17:33:34ballubaschok, I'll try a reset, thx a lot
17:36:28*TheSeven just found a deadlock issue in the USB driver framework
17:36:30 Quit MethoS- (Read error: 104 (Connection reset by peer))
17:37:39 Join ender [0] (
17:37:40ballubasch[success] :-D Thanx Slashery, I already was concerned
17:38:54 Join MethoS- [0] (n=clemens@
17:41:17Slasheriballubasch: great :)
17:41:31ballubaschswapping to CF now :-)
17:41:53 Join blitzwing [0] (
17:42:49 Join blitzwing1 [0] (
17:44:36 Quit ender` (Read error: 60 (Operation timed out))
17:46:54 Quit blitzwing ("CGI:IRC (Ping timeout)")
17:48:11 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
17:49:13CIA-85New commit by kugel (r23171): Remove wps_reset(), and rename skin_data_init to skin_data_reset() and extend it to zero most of the wps_data members. ...
17:54:53 Quit Rondom (Read error: 104 (Connection reset by peer))
17:55:30 Join Rondom [0] (
18:00:54CIA-85New commit by kugel (r23172): Whoops, I had that warning turned off locally. ...
18:01:11 Join liar_ [0] (n=liar@
18:04:06 Join tomers [0] (
18:06:36 Quit tomers (Remote closed the connection)
18:15:22 Quit petur ("work->home")
18:23:17TheSevenwow. device descriptor read/64 succeeded!
18:23:53TheSevensomething wrong with "set address"
18:24:32 Quit pamaury ("exit(*(int *)0 / 0);")
18:24:42 Quit bluebrother (Nick collision from services.)
18:24:44 Join tomers [0] (
18:24:45 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
18:24:53***Saving seen data "./dancer.seen"
18:25:12 Quit funman ("free(random());")
18:25:54tomersTheSeven: I'm interested to know what are you working on (USB stuff)
18:26:25 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
18:26:37 Quit dfkt (Nick collision from services.)
18:26:41 Nick dfkt_ is now known as dfkt (i=dfkt@unaffiliated/dfkt)
18:26:44TheSevenI'm currently trying to make the USB OTG controller (in function mode) of the ipod nano 2g work
18:26:55TheSevenit's the one described in the s3c6400x datasheet
18:28:24tomersTheSeven: Nice! Can you please provide me with a link to the datasheet? Are you committing your code, or putting it on FS?
18:28:50TheSevendepends on how nice it will run, i think...
18:29:14 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
18:30:04tomersTheSeven: I'm thinking about working on the Fuze's USB controller. I guess I could use you and gevaert's experience by then. But we got no datasheet yet.
18:30:08 Quit avacore (Read error: 60 (Operation timed out))
18:30:20 Quit dfkt (Nick collision from services.)
18:30:22 Nick dfkt_ is now known as dfkt (i=dfkt@unaffiliated/dfkt)
18:31:02 Join avacore [0] (
18:31:08tomerslinuxstb: Could you please check whether r23157 fixed "FS #10666 - Rockbox software USB doesn't connect with OS X 10.4" ?
18:31:23 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:33:23 Quit TheSeven (Read error: 104 (Connection reset by peer))
18:33:39 Join JdGordon1 [0] (
18:34:35 Join TheSeven [0] (n=theseven@rockbox/developer/TheSeven)
18:39:52 Quit maruk ("Leaving.")
18:40:15topik1190 pages ... pop quiz at 11?
18:40:22 Quit Rondom (Nick collision from services.)
18:40:32 Join Rondom [0] (
18:41:58tomersTheSeven: Thanks!
18:46:24 Join DerPapst [0] (
18:47:07mcuelenaereTheSeven: I can give you rev1.2 of that if you want to
18:47:54TheSevencan't hurt, but I don't think that will have much relevance for usb...
18:48:22mcuelenaereUSB OTG is mentioned in the changelog, but I don't think it'll be anything big
18:48:44TheSevensomething about a reg 0x44 describing endpoint directions?
18:48:49*mcuelenaere can't find it on google anymore, will upload it
18:48:54 Quit intrados1 (Read error: 60 (Operation timed out))
18:49:24mcuelenaerenope sorry, looks like 0x44 isn't mentioned..
18:49:45mcuelenaereyou do know of the Linux sources for this chip, do you?
18:51:02 Quit ballubasch ("CGI:IRC")
18:51:06 Quit liar_ (Client Quit)
18:52:13 Part Grahack
18:59:58 Join midgey [0] (
19:00:19 Join ruj1 [0] (
19:00:20 Join intrados1 [0] (
19:03:22 Join avacore^ [0] (
19:05:10 Quit Rob2222 ()
19:05:56 Join bertrik [0] (
19:11:21 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:13:42 Quit kkurbjunW (Remote closed the connection)
19:14:24 Quit avacore (Connection timed out)
19:15:06 Join kkurbjunW [0] (n=karlk@
19:15:19 Join JdGordon| [0] (n=Miranda@
19:18:55 Quit saratoga ("Page closed")
19:24:47 Join liar|pc [0] (n=liar@
19:28:14 Join efyx_ [0] (
19:28:52 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
19:29:35 Join bubsy [0] (n=Bubsy@
19:35:31 Quit JdGordon1 (Read error: 110 (Connection timed out))
19:35:52 Join Strife89 [0] (n=michael@
19:38:42 Join Dege [0] (i=Dege@
19:41:36 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
19:43:17linuxstb_tomers: As I said in the bug report, I haven't experienced that bug. (I don't have OS X)
19:43:49gevaertsTheSeven: what is this deadlock?
19:46:21 Quit dfkt (Read error: 110 (Connection timed out))
19:49:34 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:49:49TheSevengevaerts: if drv_send waits by just yielding, not sleeping or waiting for a wakeup, the whole thing will lock up, probably because that thread has a too-high priority
19:50:06TheSeven(or is this thing called with interrupts disabled? but a yield should reenable them.)
19:50:16TheSevenanyways, it causes the USB intterupts to get lost
19:50:27TheSeven(and thus the send function to wait forever)
19:50:37gevaertsTheSeven: ah, yes. That wait has to be done right
19:51:58 Quit intrados1 (Read error: 60 (Operation timed out))
20:03:20 Quit Horscht ("Verlassend")
20:05:39TheSevengevaerts: can you imagine any reason for the usb core logf'ing "SET_ADR xx" but not actually calling set_address?
20:06:24gevaertsTheSeven: it seems to do that if usb_core_ack_control() fails
20:06:25TheSevenfrom a quick glance at a code, it's probably usb_core_ack_control that's failing
20:07:31TheSevenhrm, it's locking up in there :-/
20:07:35 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
20:07:48 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
20:09:17 Nick fxb__ is now known as fxb (n=felixbru@
20:14:45tomerspixelma: ping
20:15:27 Quit blitzwing1 ("CGI:IRC (Ping timeout)")
20:17:40pixelmatomers: pong
20:18:02 Join panni_ [0] (
20:18:50tomersDid you experience "FS #10666 - Rockbox software USB doesn't connect with OS X 10.4" ?
20:20:16pixelmayes, even tried to investigate (with the hekp of barrywardell), it looks like it is a MacOS bug with composite devices
20:20:28 Nick Ypsy is now known as YPSY (n=ypsy@
20:20:58tomerspixelma: Can you please test whether r23157 fix this issue?
20:21:01pixelmahelp too
20:21:36tomersFrom the log you've posted (if it was indeed you), there was some bug, which I think I managed to fix
20:21:57pixelmaI can try tomorrow (as I only have access to those Macs at work)
20:22:25tomerspixelma: Just try tomorrow with current build. I hope it helps. If so, please close r23157. Thanks
20:23:21pixelmathe posted error messages are from someone else, he was here too
20:23:51tomerspixelma: Anyway, if it is fixed, then it doesn't matter :-)
20:24:08pixelmaoh, the ones you found in the IRC logs... yes one was mine
20:24:57***Saving seen data "./dancer.seen"
20:25:15 Quit linuxstb_ (Read error: 145 (Connection timed out))
20:26:35Strife89GodEater: (in response to a #community statement) Rockbox has Bookmarks and the Skip Length function. What's so bad about those?
20:26:41Strife89With that said, perhaps if someone had a feature to auto-bookmark any file that's longer than, say, 20 minutes?
20:26:48Strife89And auto-resume-from-bookmark?
20:27:00pixelmatomers: midgey said he had problems with his target (can't remember which) on MacOS 10.6, the newest. Maybe he could test too
20:27:36evilnickpixelma: I'll ask lambda to test with Snow Leopard (no idea about the version numbering, sorry)
20:27:54pixelmathat's 10.6
20:28:08midgeypixelma, tomers: busy at the moment. I can test in about 24 hours
20:28:15JdGordon|woops... hey... my mac has the 1.4 bits now
20:28:50enthdegreeIt'd be close to impossible to port libavcodec to rockbox, though, right? ;D
20:29:02evilnickCool, which targets in particular should he test?
20:33:18tomersmidgey: Good. Waiting for your results :-)
20:33:26pixelmaI guess one of the RockboxUSB ones (I experienced problems with my c200 on 10.4, midgey on 10.6 with his ...)
20:34:13midgeyerr, oops. I have a e200, ipod mini2g, and ipod4g
20:34:15tomersevilnick: I think it's doesn't matter what target, provided it has USB HID feature used
20:34:37tomersmidgey: All of these use the RB's USB stack, so it's OK.
20:34:56tomersPlease report in FS #10666
20:35:04midgeytomers: i know :) i have about five others with HW USB
20:35:18 Join funman [0] (n=fun@rockbox/developer/funman)
20:35:23evilnickWill see if he's got chance to report back tonight then.
20:37:20 Join bertrik_ [0] (
20:37:38 Nick fxb is now known as fxb__ (n=felixbru@
20:37:59 Quit bertrik (Read error: 113 (No route to host))
20:39:19 Join antil33t [0] (n=Mudkips@
20:39:21 Join bubsy_ [0] (n=Bubsy@
20:39:37 Quit bubsy_ (Remote closed the connection)
20:40:23pixelmamidgey: what/how is plugin language support doing?
20:40:53midgeybringing conversations in from community ;)
20:41:21amiconntomers: I wondered what r23157 is supposed to do. That change looks to me like it does essentially nothing
20:41:23linuxstbenthdegree: You could port it, but it wouldn't be very useful... (i.e. fast).
20:42:16midgeywell, i've committed some of the work so far, but mostly genlang changes
20:42:21amiconn.wTotalLength is already an uint16, so implicit and explicit casting have exactly the same result
20:42:22linuxstbenthdegree: On the video ipod, the broadcom chip is basically a second CPU/SoC - there is a separate firmware that runs on it, and IIUC, all the main Apple firmware does when playing back video to it is to send it the compressed data, and it receives back uncompressed audio.
20:42:51linuxstbenthdegree: The broadcom chip is linked to the LCD, and the main CPU linked to the DAC.
20:43:01midgeyi probably will try and commit sorting of users sometime soon
20:43:13*midgey vanishes
20:43:17 Quit midgey ()
20:44:16tomersamiconn: Instead of 0x0039 word which (correctly) represents size, the log shows 0x3900. I thought that maybe the Apple system has different endianess which caused it.
20:44:47amiconnThat won't change with this cast for sure
20:45:01amiconnSounds like a bug in the apple usb stack - nothing we can do about
20:45:52amiconnBut if this is due to endianess - why would 10.6 be affected?
20:46:05tomersamiconn: Maybe if size < 0xff write size as byte, and 0x00 to higher byte
20:46:26enthdegreehaha, you'd think. Even if were optimised for specific devices?
20:46:47amiconn10.6 is intel only, and intel is little endian
20:46:51tomersamiconn: What you say is that the DAP sends 0x0039 but the host stack translates somehow to 0x3900 ?
20:46:57amiconnAfaik the problem was reported for 10.4 on ppc
20:47:04amiconntomers: yes
20:47:23 Join KBH [0] (
20:47:28pixelmaamiconn: from his description it sounded like it was a little bit different in 10.6 (e.g. HID correctly working) but still no auto-mount - maybe it's still a different OS bug
20:47:36tomersIn my response for the FS item I postulated it could be a bug in the host...
20:47:37enthdegreeIs it even possible to do device-specific optimisation? (My knowledge of code starts and ends at very basic c#, c++ :P )
20:47:47amiconnIf our usb stack would do it wrong, it probably wouldn't work on *any* host I think
20:48:26amiconntomers: Did you analyze this on macos or some other system?
20:48:42tomersI just thought that storing integer value to unsigned work value without proper casting would cause it on a big endian system...
20:48:46amiconngevaerts: Is there a way to do this kind of analysis on linux or windows?
20:49:10tomersamiconn: I have neither Mac OS nor USB sniffer
20:49:14amiconnsigned/unsigned has nothing to do with endianess. Not at all.
20:49:17pixelmaalthough I saw HID working correctly on 10.4 too but only with that USB prober tool
20:49:49gevaertslsusb -v should show that
20:50:04 Quit DerPapst ("Leaving.")
20:51:05amiconnIs USB little or big endian?
20:54:17pixelmatomers: by the way, the other log was posted a few days before by barrywardell. It was made with an e200 on MacOS 10.5
20:54:54tomersThanks for this info.
20:54:59amiconngevaerts: I noticed that the usb stack seems to ignore endianess, that's why I'm asking
20:55:17amiconnCurrently all swusb targets are little endian, so that shouldn't cause problems
20:55:34amiconnBut it might cause problems in the future
20:57:16amiconntomers: I get wTotalLength: 57 (== 0x39) in linux using lsusb -v, and that build is from before r23157
20:57:43amiconnpixelma: Do you know whether it was PPC or x86?
20:58:03 Quit funman ("leaving")
20:58:25tomersamiconn: I thought that since PowerPC has different endianess (is it?) is could cause that bug. I didn't investigate further, and couldn't test it, that's why I asked for testing
20:59:03amiconntomers: *If* the endianess causes this bug (and in fact I tend to agree with that), it can only be caused (and only be fixed) on the host
20:59:14pixelmaI think it was PPC (the older processor)
20:59:14amiconnRockbox can't do anything about it
21:00:37 Quit HBK (Read error: 110 (Connection timed out))
21:00:40FlynDiceAnyone who may have an interest in AMS Sansa voltage scaling should take a look here FS #10669 - AMS Sansa Reimplement Voltage Scaling and yell if they have concerns ;)
21:03:32tomersamiconn: Bad news for us, indeed...
21:03:35*tomers away
21:12:36 Join ferros [0] (
21:14:16 Join kugel [0] (n=kugel@rockbox/developer/kugel)
21:15:01 Join mwtb [0] (
21:15:37 Part mwtb
21:16:44 Quit MethoS- (Remote closed the connection)
21:17:06kugelFlynDice: you know my opinion on that FS
21:17:43 Join Eragon55 [0] (i=chatzill@
21:17:46kugelFlynDice: that your first patch causes problems around microsd_insert is probably that you set it after doing the queue_broadcast()
21:19:21kugelqueue_broadcast() sends a message to *all* queues. as a result, anything using that queue and responding might simply access the SD before the variable is actually set (most importantly the SD thread which inits the cards)
21:19:30FlynDicekugel: Yes but is the implementation ok with you... yes, while I was working with other SD stuff I made it to the bottom of the file and found card_detect_target and saw that's what I really wanted...
21:20:57kugelI think checking if it's inserted is a good idea, can you create a patch which sets it before the broadcast?
21:21:49FlynDicethat boosts before the broadcast?
21:21:51kugelah, I didn't look at your very last patch, card_detect_target probably has the same effect
21:22:57kugelFlynDice: no, but setting it after is likely to result in the variable not being set before a thread causes a sd_enable call
21:23:56FlynDiceyes, card_detect_target does what I was trying to do, bad habit of mine trying to poorly reinvent something that is already done well somewhere else ;)
21:24:43 Join w1ll14m| [0] (
21:25:58pixelmalinuxstb: am I right in thinking that the ipodnano2g sim uses the same button mapping as the first gen one?
21:26:48kugelpixelma: probably not only the sim, but real target too
21:27:01pixelmaand are the labels on the latter's sim background correct? I would have to compile a sim first
21:27:16pixelmakugel: just making sure ;)
21:27:26kugelI remember one of the ipods being wrong
21:29:00kugelat least the wheels seem correct
21:29:11pixelmathey are all the same except the 3rd gen of course
21:29:24linuxstbpixelma: Yes, physically they're the same (from the outside).
21:29:53pixelmathe sim/ your keyboard/ your mapping? ;)
21:31:04linuxstbI've got a feeling I fixed the button labels on some UI sim backgrounds at some point...
21:32:05 Join Rob2222 [0] (n=Miranda@
21:32:45linuxstbpixelma: But yes, the nano1g UI sim button labels are correct.
21:32:49 Quit Eragon55 ("ChatZilla 0.9.85 [Firefox 3.5.4/20091007001339]")
21:32:56kugellinuxstb: I maybe remembered the time before then, I'm fairly sure the wheel was labelled incorrectly on some ipod
21:32:58linuxstb(for the nano2g too)
21:33:09linuxstbkugel: Yes, that's what I recall fixing.
21:34:21linuxstbLooks like they were all wrong...
21:36:37Lloreankugel: You never really responded to my question on the forums about the plugin actions rework build you posted.
21:36:56LloreanIt's more or less impossible for anyone to know if it's working as intended if you just say "some keymaps may change" and don't let people know which ones are expected to.
21:36:57kugelI didn't see you posted something
21:37:48kugelLlorean: I wrote that all targets are affected
21:38:31kugelI also mentioned the plugins
21:39:09LloreanYou mentioned that button mapping "may have changed slightly"
21:39:09Llorean*may* indicates that it also may not. So is it changed for all of those plugins on all targets?
21:39:12Unhelpfulkugel: i had a suggestion, see the log. otherwise i think it looks fine. :)
21:39:15LloreanOr some plugins on all targets? Or all of the plugins listed but only on some targets? Or some on some?
21:39:41kugelI picked may for a reason
21:39:55Unhelpfulbasically if you have an opaque user argument to bufopen it might make more sense to pass a struct with the size than to pass a callback to retrieve the size.
21:39:58LloreanSo my questions are valid - the user doesn't know which plugins are expected to have keymap changes.
21:40:02kugelit's nearly impossible to tell which targets have changed keymap in which plugins exactly
21:40:13 Join froggyman [0] (
21:40:33LloreanSo should users report keymap changes as a bug, then, since you don't know which ones are intentional?
21:40:34kugelall the ones I wrote down, for all targets. didn't I mention that clear enough?
21:41:42LloreanYou didn't write down which ones are expected to have keymap changes, and what sort. How is a user supposed to tell the difference between the keymap changes you "knew" would happen (but only in the vaguest sense) and keymap changes that result from a bug in the patch?
21:41:43kugelthere's not a single intentional keymap change, but I'm 100% a change is unevitable for some targets on some plugins
21:41:48 Quit ferros ("CGI:IRC")
21:41:50kugel100% sure*
21:41:54amiconnThe 1st/2nd Gen wheel also has a slightly different layout
21:41:56LloreanAgain: Should users report keymap changes as bugs in that build then?
21:42:35kugelI'll make it more clear what I want to get posted
21:43:56 Quit ruj1 (Read error: 60 (Operation timed out))
21:45:10*Llorean also thinks people posting in the "Official Test Builds" forum should subscribe to their own posts to be notified of replies.
21:45:29 Join ruj1 [0] (n=ruj1@
21:45:32LloreanThey don't do much good if you're not watching them.
21:45:32Unhelpfulyou could also later replace it with a pointer to a zero-terminated list of struct dim, if the loader eventually supports loading once with multiple output sizes.
21:45:48pixelmaI don't think they are "inevitable", they would "jiust" need to be taken care off in a different way. If nothing's changed I know one off the top of my head from testing some previous version of the patch - in bubbles, "Up" doesn't shoot anymore on the Ondio
21:46:14Unhelpful(see bmp.h for struct dim, it contains only a width and a height field)
21:47:22 Join barrywardell [0] (
21:48:23pixelmait would be some kind of exception to use Up there because the game layout and the fact that the button pad is below the screen on the Ondio makes it intuitive - in contrast to other games where "action" is more suited for the Mode button
21:48:39kugelLlorean: I thought I had email notify on
21:48:48kugelwasn't the case
21:49:10linuxstbamiconn: Speaking of the 1st/2nd gen wheel, do you have any ideas about ? It would be nice to close that one...
21:49:46kugelUnhelpful: the problem is that the callback also returns whether the album art is used (the function I pass for the wps is "wps_uses_albumart"). albumart.c checks for this return value as well when looking for the files
21:50:07kugelI can't do that with just the dim struct (although I didn't know it exists)
21:50:36Unhelpfulif there's no album art, surely you're never calling bufopen to load it in the first place?
21:50:56 Join compguy248 [0] (n=chatzill@
21:51:08 Quit compguy248 (Client Quit)
21:51:23kugelI haven't looked very deeply, but the function seems to have a way to return false even if the %Cl was passed successfully
21:51:29amiconnlinuxstb: That's still not fixed in svn. The simple fix might be potentially dangerous, that's why I don't like it, and I still didn't get around to analyzing how the OF fills the sysinfo struct if the rom doesn't have it
21:52:24kugelJdGordon|: what exactly is WPS_ALBUMART_NONE (skin_albumart.state can have that value)?
21:52:37Unhelpfulalso i like being able to have a list of sizes later. i know roughly how a scaler might handle multiple output sizes... i can even think of a few ways that would lump some of the work together.
21:53:10kugelmy patch is about having multiple sizes of the same AA
21:53:10JdGordon|kugel: I think it was when no aa was found for the track
21:53:25kugelbut it doesn't do it at the bufopen level (and I think it shouldn't be done there)
21:53:56kugelJdGordon|: that wouldn't make any sense
21:54:17*JdGordon| never suggested it did!
21:54:49kugelwell, in that case find_albumart() would already fail, and wps_uses_albumart() doesn't need to check that
21:55:07 Quit ruj1 ()
21:55:12JdGordon|it could very well be redundant
21:55:41kugelJdGordon|: find_albumart() calls wps_uses_albumart() which returns false if WPS_ALBUMART_NONE is set (i.e. if it's set before something has actually looked for albumart)
21:56:56kugeland the get_token_value determines whether AA was found by looking at the handle id of the buffered albumart (which is -1 if not found/not buffered)
22:06:48JdGordon|well yeah... I did a stright conversion to the skin stuff... I didnt change any logic
22:08:12 Join barrywardell_ [0] (
22:10:49kugelUnhelpful: should be safe to use such a dim struct then
22:11:35kugelthanks for having a look, by the way
22:12:15*TheSeven is stuck with an usb controller that's sending undocumented interrupts and refusing to send any data to the host once the first transaction has finished
22:17:52 Join Omlet [0] (i=omlet05@
22:18:00 Join Omlet05 [0] (i=omlet05@
22:19:26 Quit Omlet (Client Quit)
22:19:26 Quit Omlet05 (Client Quit)
22:21:19 Quit bmbl ("Bye!")
22:21:19CIA-85New commit by bluebrother (r23173): Don't trace every device in the machine when scanning for USB devices on Windows.
22:21:25CIA-85New commit by bluebrother (r23174): Add cache index file to HttpGet class to maintain a list of hash - file origin mappings.
22:24:43 Quit froggyman (
22:24:44 Quit w1ll14m| (
22:24:44 Quit dfkt_ (
22:24:44 Quit tomers (
22:24:44 Quit faemir (
22:24:44 Quit robin0800 (
22:24:44 Quit thegeek (
22:24:44 Quit gitster (
22:24:44 Quit Llorean (
22:24:44 Quit AndyIL (
22:24:44 Quit linuxstb (
22:24:44 Quit z35 (
22:24:44 Quit gevaerts (
22:24:44 Quit togetic (
22:24:44 Quit goffa (
22:24:44 Quit Bob_C (
22:25:00***Saving seen data "./dancer.seen"
22:26:50*pixelma looks around for linuxstb... hmm
22:27:36pixelmaoh, just netsplitted
22:28:04TheSevenlindbohm is failing often these days...
22:28:19 Quit barrywardell (Read error: 110 (Connection timed out))
22:28:19 Nick barrywardell_ is now known as barrywardell (n=barrywar@rockbox/developer/barrywardell)
22:28:29domonokyZagor, Bagder: why are there no voice links on the daily-build page anymore ?
22:28:45 Join hd [0] (
22:30:16 Join Traveler [0] (
22:30:43amiconnbarrywardell: ping
22:30:47 Nick Traveler is now known as Guest36291 (
22:30:54 Join intrados1 [0] (n=intrados@
22:31:59 Quit intrados1 (SendQ exceeded)
22:32:21barrywardellamiconn: pong
22:32:45*TheSeven just caught a double-bug in his USB driver
22:32:48TheSevenone hiding the other...
22:33:08amiconnbarrywardell: Regarding your usb log from OSX - was that PPC or x86?
22:33:23barrywardellamiconn: x86
22:33:53amiconnAh, that explains the byteswap then
22:33:53amiconnObviously OSX does endian correction before printing the "raw" data
22:34:25amiconnYours is the correct order, because x86 and USB are both little endian
22:34:39barrywardellwhy was it only those two bytes that are swapped?
22:34:46NJoinBob_C [0] (
22:34:46NJoingoffa [0] (n=goffa@
22:34:46NJointogetic [0] (n=togetic@unaffiliated/ibuffy)
22:34:46NJoingevaerts [0] (n=fg@rockbox/developer/gevaerts)
22:34:46NJoinz35 [0] (
22:34:46 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
22:34:46NJoinAndyIL [0] (n=pasha_in@
22:34:46 Join Llorean [0] (n=DarkkOne@rockbox/user/Llorean)
22:34:46NJoingitster [0] (
22:34:46NJointhegeek [0] (
22:34:46NJoinrobin0800 [0] (
22:34:46NJoinfaemir [0] (n=faemir@
22:34:46NJointomers [0] (
22:34:46NJoindfkt_ [0] (i=dfkt@unaffiliated/dfkt)
22:34:46NJoinw1ll14m| [0] (
22:34:46NJoinfroggyman [0] (
22:34:57amiconnPPC pre-swapped, but interpreted it correctly (see above)
22:35:07amiconnIt's the only 16 bit value in the config descriptor
22:35:43 Join intrados1 [0] (
22:36:10pixelmalinuxstb: welcome back
22:36:29barrywardellah, makes sense
22:36:33 Quit intrados1 (SendQ exceeded)
22:37:15 Quit Guest36291 ("Java user signed off")
22:37:53 Join intrados1 [0] (
22:39:59 Quit enthdegree ("Leaving")
22:40:42 Quit intrados1 (SendQ exceeded)
22:41:54 Join intrados1 [0] (
22:42:44 Quit intrados1 (SendQ exceeded)
22:43:51CIA-85New commit by bluebrother (r23175): Fix a compiler warning showing up with newer gcc versions.
22:44:54*TheSeven wonders where a -71 ("EPROTO") error comes from in his wireshark trace
22:46:15 Quit HellDragon (Connection timed out)
22:48:34 Join intrados1 [0] (
22:51:18 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
22:52:36 Join bluebrother [0] (
22:53:03 Quit bluebroth3r (Nick collision from services.)
22:53:41 Join compguy248 [0] (
22:54:16 Quit intrados1 (Connection reset by peer)
22:55:15 Join intrados1 [0] (
22:57:34 Quit compguy248 (Client Quit)
22:58:07 Join compguy248 [0] (n=compguy2@
23:01:22 Nick bertrik_ is now known as bertrik (
23:02:16 Quit compguy248 (Client Quit)
23:05:25 Quit esperegu_ (Read error: 113 (No route to host))
23:05:34 Part froggyman
23:06:09 Quit intrados1 (Read error: 60 (Operation timed out))
23:09:39 Quit GeekShadow ("The cake is a lie !")
23:11:46 Join Creposucre [0] (
23:15:21CreposucreAnyone who may want to use their non supported ipod accessory might want to test FS10623
23:16:05 Quit kkurbjunW (Remote closed the connection)
23:17:56 Join kkurbjunW [0] (n=karlk@
23:25:08 Join Thundercloud [0] (
23:25:44 Quit Rob2222 ()
23:26:25linuxstbCreposucre: I don't think many devs have ipod accessories, so you may not find many testers here...
23:27:39domonokyCreposucre: if you want testers, try forums/mailing lists.
23:29:20Creposucreok, thanks for the advice ;)
23:31:36*JdGordon| is reminded he has 4 different hardware stuff to fiddle with :(
23:31:58JdGordon|also, i tihnk my ipod video is going to die... damn hard disk is playing up
23:32:14 Quit Strife89 (Read error: 60 (Operation timed out))
23:32:26 Quit TheSeven (Read error: 60 (Operation timed out))
23:33:33CreposucreJdGordon|: time to try the CF mod?
23:34:28JdGordon|na... CF in my mini... i want to keep at least one target over 30GB :p
23:34:55 Join TheSeven [0] (n=theseven@rockbox/developer/TheSeven)
23:36:28amiconnPfft, even CF is >30GB these days
23:36:42amiconn*available at
23:36:49JdGordon|crazy expensive though
23:37:02JdGordon|Creposucre: no promises.. but ill try that patch tonight
23:37:43kugelonly with a fancy adaptor if there is one, I guess
23:38:23JdGordon|its damn hard to find a ssd with the writght connector (my netbook uses the same one)... and they are the old slow type
23:38:36JdGordon|which is fine for ipods, but useless for a compuuter...
23:39:04 Join gywst [0] (
23:39:32amiconnkugel: There are 50pin SSDs as well as ZIF SSDs
23:40:24*TheSeven wonders when there will be ssd-based daps that use sata as the storage interface to piggyback on the laptop market
23:40:30CIA-85New commit by dave (r23176): Working Nano2G UI sim. UI-ipodnano2g.bmp courtesy of Marianne Arnold, based on her SVG for the manual.
23:40:52 Quit rasher (
23:40:52 Quit AB3JU (
23:40:55NJoinAB3JU [0] (
23:41:11NJoinrasher [0] (
23:42:03 Quit dfkt_ ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
23:42:44TheSeven[42085.884046] usb 2-1: new full speed USB device using uhci_hcd and address 126
23:42:45TheSeven[42091.025210] usb 2-1: unable to read config index 0 descriptor/start: -110
23:42:47TheSeven[42091.025233] usb 2-1: chopping to 0 config(s)
23:42:48TheSeven[42101.025522] usb 2-1: string descriptor 0 read error: -110
23:42:50TheSeven[42101.025700] usb 2-1: no configuration chosen from 0 choices
23:42:52TheSeven[42142.560124] usb 2-1: USB disconnect, address 126
23:43:14gywstibought a sansa/exit
23:43:33linuxstbTheSeven: That's good I assume?
23:43:33 Quit gywst (Client Quit)
23:43:35 Join gywst [0] (
23:43:43TheSevenwell, it means we're one set further
23:43:46gywstback... sorry
23:44:15 Quit domonoky (Read error: 131 (Connection reset by peer))
23:44:44gywstI bought a Sandisk Sansa Fuze... fount it is a v2 I'm a n00b... where can I start hacking
23:45:03linuxstbgevaerts: Can you add the Nano2G sim to the builds? I think that's all now...
23:46:13*gevaerts testcompiles
23:48:51*TheSeven spots bogus bus reset interrupts
23:48:52 Quit evilnick ("Page closed")
23:49:31bertrikTheSeven, bogus?
23:49:55bertrikwindows does a bus reset after reading only the first part of the device descriptor IIRC
23:50:13TheSevenwell, they don't correspond to what I see in wireshark
23:50:32 Quit Thundercloud (Remote closed the connection)
23:50:33CIA-85New commit by gevaerts (r23177): Add nano2g sim
23:50:36TheSeventhere seems to be a fake one after each transaction, making the get configuration descriptor reuest fail
23:51:09linuxstbgevaerts: Thanks
23:51:39 Quit Creposucre ("CGI:IRC (EOF)")
23:52:25 Quit mcuelenaere ()
23:54:32 Quit jgarvey ("Leaving")
23:54:51 Join bertrik_ [0] (
23:55:13 Join Sajber^ [0] (
23:55:35n1shmm, this new port categorization makes me think we should downgrade ipod 1g/2g to "unstable" because of the FS #8778 bug and the whole reboot to OF to sleep thing, i think that qualifies as "less usable or has problems that limit it to advanced users"
23:56:33 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)

Previous day | Next day