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

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

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

#rockbox log for 2011-10-28

00:04:32 Quit tchan (Read error: Connection reset by peer)
00:05:17 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
00:05:49***Saving seen data "./dancer.seen"
00:07:29 Join [Saint] [0] (~Saint]@
00:07:29 Quit [Saint] (Changing host)
00:07:29 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
00:09:59 Quit z180 (Ping timeout: 256 seconds)
00:10:08 Quit fdm (Remote host closed the connection)
00:21:10 Quit Unhelpful (Ping timeout: 240 seconds)
00:22:13 Quit wodz (Quit: Leaving)
00:29:49 Join mystica555 [0] (
00:29:54 Quit liar (Remote host closed the connection)
00:33:57 Quit ageis (Read error: Connection reset by peer)
00:35:08 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
00:38:19 Join Scromple [0] (
00:39:34 Quit Unhelpful (Ping timeout: 240 seconds)
00:51:21 Quit dhrasmus (Quit: Leaving)
01:02:35 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
01:22:06 Join [Saint_] [0] (~Saint]@unaffiliated/saint/x-8516940)
01:25:10 Quit [Saint] (Ping timeout: 240 seconds)
01:28:01 Quit MethoS- (Remote host closed the connection)
01:40:46 Quit [Saint_] (Ping timeout: 240 seconds)
01:44:56 Join [Saint] [0] (~Saint]@
01:44:56 Quit [Saint] (Changing host)
01:44:56 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
01:56:22 Quit hilbert (Quit: Textual IRC Client:
01:59:50 Join JdGord [0] (
01:59:58 Join ChickeNES [0] (~ChickeNES@
02:05:53***Saving seen data "./dancer.seen"
02:07:49 Join Scr0mple [0] (
02:09:34 Quit JdGord (Ping timeout: 240 seconds)
02:09:34 Quit Scromple (Ping timeout: 240 seconds)
02:11:33 Join JdGord [0] (
02:14:36 Quit zchs (Ping timeout: 265 seconds)
02:24:39 Join zchs [0] (
02:31:39 Quit KiwiCam (Quit: Leaving)
02:51:50 Join jdgord_ [0] (
02:51:51 Quit JdGord (Ping timeout: 258 seconds)
02:53:23 Quit acorn (Ping timeout: 265 seconds)
03:01:19 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
03:23:04 Join ageis [0] (
03:35:41 Nick Scr0mple is now known as Scromple (
03:49:07 Join JdGord [0] (~AndChat@
03:49:20 Quit JdGord (Client Quit)
03:49:33 Quit jdgord_ (Read error: Connection reset by peer)
03:49:34 Quit Scromple (Ping timeout: 240 seconds)
03:51:14 Join Scromple [0] (
04:02:10semitonesso I guess I have to crack my clip+ open :)
04:02:14semitonesor :/
04:05:54***Saving seen data "./dancer.seen"
04:12:45 Quit GeekShad1w (Remote host closed the connection)
04:15:07semitonesI still want to try the "hacking" section of
04:19:34semitonesbut there are parts I don't understand about it: "The size reserved by the OF is exactly twice the size of the firmware file when the reserved space is 0xF000 sectors (models with 8MB of ram)"
04:19:56 Quit TheSeven (Disconnected by services)
04:20:14 Join [7] [0] (~TheSeven@rockbox/developer/TheSeven)
04:20:22semitonesthe OF file for the clip+ is 16mb, and the drive I can see in fdisk is 32 mb. Does that mean I'm at the right place?
04:22:05semitonesbut I don't understand this sentence: "Copying an OF image over this drive might not be enough to recover operation, you might need the full part reserved by OF."
04:46:06 Quit amiconn (Disconnected by services)
04:46:07 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:46:07 Quit pixelma (Disconnected by services)
04:46:09 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:46:11 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:46:28 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:59:55 Quit zchs (Ping timeout: 265 seconds)
05:24:02 Join zchs [0] (
05:35:51 Join Rob2222 [0] (
05:37:13 Join tjb0607 [0] (
05:39:32 Quit Rob2223 (Ping timeout: 245 seconds)
05:46:13 Quit Horschti (Quit: Verlassend)
05:50:37 Quit nosa-j (Read error: Connection reset by peer)
06:00:41 Join nosa-j [0] (
06:05:57***Saving seen data "./dancer.seen"
06:34:40 Quit ChickeNES (Quit: Computer has gone to sleep.)
06:34:56 Quit robin0800 (Quit: Leaving)
06:53:55[Saint]Posting here as its logged and someone might be interested in this:
06:54:06 Quit AlexP (Remote host closed the connection)
06:54:14[Saint] <−−- ALAC has been open sourced with an Apache license
07:18:41 Quit Scromple (Ping timeout: 255 seconds)
07:20:13 Join robin0800 [0] (
07:28:13 Quit robin0800 (Quit: Leaving)
07:55:42 Join nosa [0] (
07:57:05 Quit nosa-j (Ping timeout: 259 seconds)
07:57:05 Nick nosa is now known as nosa-j (
08:02:57 Join Buschel [0] (
08:05:59***Saving seen data "./dancer.seen"
08:08:11 Quit [Saint] (Ping timeout: 240 seconds)
08:15:06 Join markun [0] (
08:21:02 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:25:31JdGordonanyuone know rough batt runtimes for ipod video with svn?
08:28:14 Join GodEater_ [0] (93722cc9@rockbox/staff/GodEater)
08:28:16BuschelI hope the numbers on the battery runtime page are still valid
08:29:42kugelmc2739: \o/
08:30:01kugelI have a fix for FS #12325
08:30:02fs-bluebot e200v1 screen corruption after USB connection since r30475 (bugs, new)
08:30:41JdGordonCF-modded so i shold get up around 20 hours :)
08:31:48BuschelJdGordon: you should switch off line-out if not needed, this consumes quite some power, and of course apply the 24Mhz default clock + your gui boost :)
08:32:47JdGordondone and done :)
08:32:59JdGordonjust doing a top up charge before starting a bench
08:33:07kugelJdGordon: fwiw, the wheel driver don't debounce AFAIK
08:33:28kugelthey only set BUTTON_REL after BUTTON_REPEAT, not for individual scrolls
08:34:41JdGordonthats fine, worst case is it stays boosted for longer than it needs to (very unlikely though)
08:34:44kugelyou can't really do BUTTON_REL, except perhaps with guessing (same issue as with the trackball on android).
08:35:16kugelI didn't look at the patch, but if it only unboosts on BUTTON_REL individual scrolls will keep it boosted forever, no?
08:35:49JdGordonerr, hmm.. ok, maybe its easier to do a single scroll than i tohught
08:36:23JdGordonyeah, you're right
08:36:55JdGordonpressing any button will unboost it though
08:36:58Buschelok, gotta go to work. see you later
08:37:01 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 7.0.1/20110928134238])
08:37:35kugelI think the 1s-timeout approach is nicer anyway because it avoids fast boost-unboost-boost-unboost scenarios which can affect user experience
08:38:36JdGordonunboosting on release wont be noticable by the user *by definition* because they stop interacting with it when it happens
08:38:36kugelbtw, I too think doing it in the action subsystem is a better approach
08:39:08JdGordonbut I was under the impression scrollwheels sent the REL button
08:39:42kugelthey don't necessarily stop interaction
08:39:46JdGordonactually I thught they would happen because the button driver adds that button event in implictly on the button_read() after a button is stopped being presed
08:40:17kugelstuff happens still on the display
08:40:41kugelalso I can imagine scrolling slowly possibly suffers
08:41:16kugelJdGordon: scroll wheels don't go through that
08:41:50kugelthey do queue_post() directly instead of button_read_device()
08:42:09JdGordonok, didnt know that
08:42:12kugel10ms polling isn't enough to properly do acceleration
08:42:41JdGordonand no, single scrolling is not noticable at all from the user, though at that speed there is no need to boost anyway
08:42:49JdGordonso perhaps boosting only once we see repeat would work
08:42:57 Join wodz [0] (
08:43:08kugelwhat's wrong with the 1s timeout?
08:43:12 Join ender` [0] (
08:45:36JdGordonit is unecessarily complicated
08:46:12kugelthe patch looks very simple
08:46:42JdGordonthats the way I'd like to keep it
08:46:44kugelyours a bit simpler indeed
08:46:53JdGordona simple bool to control it is perfect
08:47:01wodzRepeat mode behaves erratically. I am observing two bugs but can't find reliable way to reproduce. 1) selecting file from file browser -> change repeat to 'one' -> skip doesn't change the file (that is expected) but next track displayed in WPS advances
08:47:11JdGordonI don't have time now to test how it feels only boosting after repeat, ill do it sat night
08:47:40 Quit sinthetek (Read error: Connection reset by peer)
08:47:45wodz2) select file from file browser -> change repeat mode to 'one' -> change repeat mode to 'off' -> skip doesn't work
08:47:55JdGordonwodz: that's always been there, it shuld be easy to fix thugh (1)
08:48:14wodzJdGordon: should I fill FS bug?
08:49:03JdGordonif you want it fixed :)
08:49:17 Join sinthetek [0] (
08:49:18 Quit sinthetek (Changing host)
08:49:18 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
08:50:00wodzJdGordon: If thats easy to fix I may also ask you very kindly to look at this :-)
08:50:36JdGordonso here is the conundrum... where should the correct place for the fix go?
08:50:49JdGordonshould the wps be smart and check repeat mode? or should playback?
08:51:00 Join nick-p [0] (
08:51:22JdGordonperhaps audio_next_track() ?
08:51:49wodzI think playback is the place
08:52:04 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
08:52:06JdGordonI don't know if that knows what repeat mode it is in though?
08:52:17JdGordonI suspcect it doesnt
08:52:29kugelI would think that playback knows the next track but the wps just isn't updated properly
08:53:15wodzFS #12356
08:53:16fs-bluebot next track advances when skip in repeat one mode (bugs, new)
08:53:59wodzthe second bug I mentioned is harder to reproduce
08:54:37JdGordona quick look tells me that playback.c is scary and to ask jhMikeS how it works :)
08:54:50kugelwodz: by "select a file" you mean select one file in a directory with multiple files, i.e. more than one file is added to the playlist?
08:55:05wodzkugel: yes
08:55:54kugelwodz: are you playing the last file in the second case by chance?
08:56:17wodzand another bug I hit is famous scroll wheel doesn't work sometimes after deasserting hold
08:56:21kugelwithout repeat and if playing the last file in the playlist, then skip is expected to do nothing as there's no next track
08:56:39 Quit bertrik (Read error: Operation timed out)
08:56:58wodzkugel: but it shouldn't restart current track but end
08:57:22kugelyou just said "skip doesn't work"
08:57:48kugelthat meant to me "skip does nothing"
08:58:00wodzkugel: yeah, I mean doesn't work as doesn't account for changed repeat mode
09:00:10kugelhuh. repeat one is even more broken apparently
09:00:37wodzhmm, I described hold bug wrong - this is this: FS #5230
09:00:37fs-bluebot Keypad HOLD doesn't let go until scrolling is used (bugs, requires)
09:01:14wodzon MPIO however it is reversed - touchstrip doesn't work after hold while regular keys do.
09:01:23kugelwodz: for me in bug 1, skip also changes the file
09:02:30kugelthen the wps shows the file after (file 3)
09:02:34kugelas next track
09:02:45wodzsomething is really fishy there
09:02:56kugelbut the next skip goes back to file 1
09:08:12wodzReading about scrollwheel bug it looks like driver problem. The question is why it needs some time in hold state to be visible on MPIO.
09:09:23wodzhmm maybe I need to unconditionally ack scrollstrip interrupts after unmasking
09:11:04 Join petur [0] (~petur@rockbox/developer/petur)
09:33:09kugelJdGordon: does your font patch fix a bug, or is it only about the api?
09:33:21kugelthe one from the ml
09:43:18 Join [Saint] [0] (~Saint]@
09:43:19 Quit [Saint] (Changing host)
09:43:19 Join [Saint] [0] (~Saint]@unaffiliated/saint/x-8516940)
09:43:37JdGordonkugel: just api i believe
09:43:44JdGordonjust dont forget it after the unfreeze :p
09:48:50 Join dfkt [0] (~dfkt@unaffiliated/dfkt)
10:06:03***Saving seen data "./dancer.seen"
10:20:45 Join liar [0] (
10:21:36 Join dfkt_ [0] (~dfkt@unaffiliated/dfkt)
10:25:01 Quit dfkt (Read error: Connection reset by peer)
10:25:01wodzscrollstrip after hold doesn't look like a driver bug. The way to reproduce is: 1) select file for playback from filebrowser (this may be the last file in dir as well) 2) assert hold 3) wait for playback finish (this comes back to filebrowser) 3) deassert hold -> scrollstrip doesn't work. However backlight reacts so button events are passed by the driver
10:26:22kugelwodz: that's another issue, though
10:27:19wodzkugel: from ipod? I think it may have the same roots
10:27:27kugelso are you seeing FS #5230 or not?
10:27:28fs-bluebot Keypad HOLD doesn't let go until scrolling is used (bugs, requires)
10:29:13wodzkugel: hmm, can't reproduce on nano2g. Maybe I am mixing this with behaviour of my mini1g which definitely exhibits this bug - can't test though as Torne has my mini
10:29:51kugelwhat scrollstrip are you talking about?
10:30:13*kugel is confused
10:30:26kugelyou said you see FS #5230 with the scrollstrip, didn't you?
10:30:27fs-bluebot Keypad HOLD doesn't let go until scrolling is used (bugs, requires)
10:30:39kugelnow you reported a different buggy behavior
10:30:51wodzyes I confused myself
10:31:36wodzthis two are separate although striking similar
10:34:19kugelI think I know the issue you described now
10:34:50wodzgreat - do you know what is going on?
10:35:14kugeland both are quite the opposite. in the first only the scroll wheel/strip keeps working, in the second only the wheel/strips stops working
10:35:22kugelI know there are several FS# about it
10:35:55wodzyes exactly the case
10:37:11kugelFS #8816 is one of them
10:37:12fs-bluebot Extra button press required to continue browsing menus (bugs, unconfirmed)
10:37:25kugelbut there were more I cant find
10:37:45kugelah. see its related tasks
10:39:21kugelit has to do with blocking button presses on context switch (screen transition) to not make accidental presses
10:39:31wodzthis is damn irritating
10:42:11[7]jhMikeS: what's the semantics of pcm_play_dma_started_callback?
10:47:25 Quit Riviera (Remote host closed the connection)
10:49:38 Join mortalis [0] (~mortalis@
10:54:12wodzmortalis: the sdk you have found is extremely weird but very useful. It has tons of references to rk28 but code and data structures seems to match rk27xx
10:59:42mortalisseems rk28 an rk27 are very similar
11:01:45wodzmortalis: not at all - I have datasheets for both (although incomplete)
11:23:21 Join MethoS- [0] (~clemens@
11:39:27 Quit mortalis (Quit: KVIrc 4.1.1 Equilibrium
11:41:36 Join mortalis [0] (
11:44:35 Quit ukleinek (Ping timeout: 244 seconds)
11:49:20 Quit bluebrother (Disconnected by services)
11:49:22 Join bluebrother^ [0] (~dom@rockbox/developer/bluebrother)
11:50:20 Join ukleinek [0] (~ukl@2001:6f8:1178:2:21e:67ff:fe11:9c5c)
11:51:10 Quit fs-bluebot (Ping timeout: 240 seconds)
11:52:51 Join fs-bluebot [0] (
11:59:17 Join hilbert [0] (
12:06:04***Saving seen data "./dancer.seen"
12:20:31 Quit wodz (Ping timeout: 276 seconds)
12:35:28 Join lorenzo92 [0] (
12:36:08lorenzo92hi all...I was wondering how volumes are considered in an as3543 codec. I mean: headphones volume vs DAC volumes
12:36:34lorenzo92for balancing are they DAC? Otherwise they are always 100%?
12:54:08rasherHow much work has been done on our ALAC decoder?
12:54:35 Quit [Saint] (Ping timeout: 240 seconds)
13:00:50 Join TheLemonMan [0] (
13:01:27 Part Zagor
13:01:31 Quit nick-p (Quit: Leaving)
13:27:14 Join wodz [0] (
13:27:34 Join nls_web [0] (
13:27:38 Quit nls_web (Client Quit)
13:27:45 Join nls_web [0] (
13:28:41 Quit nls_web (Client Quit)
13:31:32 Quit lorenzo92 (Quit: ChatZilla 0.9.87 [Firefox 7.0.1/20111008085652])
13:49:31 Quit hilbert (Quit: Textual IRC Client:
13:50:17 Quit ReimuHak_ (Quit: Leaving...)
13:52:51 Join The_Seeker [0] (
13:57:40 Quit The_Seeker (Client Quit)
14:00:32 Join einhirn [0] (
14:00:57jhMikeS[7]: right after a callback to get more data, start DMA for the new buffer, then call the function
14:06:06***Saving seen data "./dancer.seen"
14:06:32jhMikeSit's there so the mixer can generate its next frame after the driver has started playing the last one that was prepared
14:08:37 Quit mortalis (Quit: CGI:IRC)
14:09:30 Join GeekShadow [0] (
14:09:48jhMikeSoh, and if get_more returns "no more" then you don't call it
14:11:14jhMikeSa simpler example of using it would be pcm-meg-fx.c for one
14:16:44[7]jhMikeS: so at which point in time should it be called?
14:17:19[7]it's basically releasing the previously played chunk?
14:17:47 Quit Bagder (Ping timeout: 260 seconds)
14:19:39[7]in that case it could be called right after pcm_play_get_more_callback, as the audio being played during that time is double buffered
14:27:26 Join bertrik [0] (
14:27:26 Quit bertrik (Changing host)
14:27:26 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
14:35:30jhMikeS[7]: right, it's based on the user's buffers, not how the driver splits it
14:43:35[7]but calling it a couple of microseconds later shouldn't cause that weird usb crash
14:45:29jhMikeSit also better not take _microseconds_ to reach it :)
14:46:15jhMikeSof course the cache flushing could slow things up alot
14:46:54 Join The_Seeker [0] (
14:47:37wodzjhMikeS: have you seen FS #12356 ?
14:47:38fs-bluebot next track advances when skip in repeat one mode (bugs, new)
14:49:40wodzalso anyone knew where to look for bug that some buttons events are ignored after screen change in hold mode?
14:49:49jhMikeSwodz: no. it wasn't doing that last time I did major playback work.
14:50:58jhMikeSthe playlist code. playback should only see the same track over and over again in repeat 1
14:51:23wodzwho is in charge of playlist code?
14:52:21jhMikeSI guess look at "maintainers" ??
14:54:27wodzjhMikeS: unfortunately I don't know maintainer called ''
14:54:50jhMikeSMr. Someone (our "stig")
14:56:35 Quit einhirn (Ping timeout: 240 seconds)
14:57:47wodzjhMikeS: Is it true that playback is always 44.1kHz samplerate and anything else is resampled first?
14:58:20bertrikwodz, this is true as far as i know
14:58:27jhMikeSwodz: for core playback, that is true
14:58:41wodzwhy it is so?
14:58:41bertrikour dsp stuff assumes a hardcoded 44100 sample rate
14:59:16jhMikeSwodz: lot of reasons, namely matching many things that need a common samplerate and certain effects are computer for that samplerate
15:00:08wodzso what for is samplerate setting in pcm drivers?
15:00:21jhMikeSif it were ever changeable I'd recommend a global preference, not switching on the fly since that is rarely a smooth operation and pretty much impossible to time properly anyway
15:00:46jhMikeSwodz: one's that use pcm directly can set it, like doom or pacbox, etc.
15:01:06jhMikeSalso for recording
15:01:23bertrikcrossfade will be difficult with different sample rates
15:04:35 Quit The_Seeker (Quit: Stay hungry, stay foolish)
15:27:23 Quit GodEater_ (Ping timeout: 265 seconds)
15:39:50 Join ReimuHak_ [0] (~reimu@
15:40:31 Quit ReimuHak_ (Client Quit)
15:40:56 Join ReimuHak_ [0] (~reimu@
15:41:39 Quit zchs (Ping timeout: 252 seconds)
15:47:24 Quit petur (Quit: *plop*)
15:49:55 Quit ReimuHak_ (Quit: Leaving...)
15:54:25 Nick Guinness` is now known as Guinness (
15:54:38 Join zchs [0] (
15:58:08 Join WalkGood [0] (~4@unaffiliated/walkgood)
16:05:52 Join Thra11 [0] (~thrall@
16:06:09***Saving seen data "./dancer.seen"
16:08:48 Join mortalis [0] (
16:22:14 Join Bagder [0] (
16:22:15 Quit Bagder (Changing host)
16:22:15 Join Bagder [241] (~daniel@rockbox/developer/bagder)
16:23:33 Join ChickeNES [0] (~ChickeNES@
16:24:22 Quit ChickeNES (Read error: Connection reset by peer)
16:24:32 Join ChickeNES [0] (~ChickeNES@
16:27:18 Join Horscht [0] (
16:27:18 Quit Horscht (Changing host)
16:27:18 Join Horscht [0] (~Horscht@xbmc/user/horscht)
16:28:44 Join y4n [0] (y4n@unaffiliated/y4ndexx)
16:31:19 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
16:31:59 Quit chkktri (Ping timeout: 256 seconds)
16:39:51 Quit mortalis (Quit: CGI:IRC (Ping timeout))
16:45:54 Join hilbert [0] (
16:49:51 Join funman [0] (~fun@rockbox/developer/funman)
17:01:17 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:02:32*bertrik is hacking the sansa clip zip
17:02:58dfkt_go bertrik \o/
17:03:03bertrikso far, I succeeded in running the OF with our dualboot loader code (which goes to the OF unconditionally now)
17:04:41bertrikfunman, can you do a quick review of the code to introduce a delay to prove the button reading code?
17:05:36bertrikI think the loop should take about 3 or 4 cycles, 24000000 (24 MHz) is the expected CPU frequency, so about 3 or 4 seconds delay
17:06:51funmanbertrik: ok to me
17:15:47 Quit ukleinek (Ping timeout: 240 seconds)
17:16:44 Join ukleinek [0] (~ukl@2001:6f8:1178:2:21e:67ff:fe11:9c5c)
17:17:56bertrikbutton readout works too, GPIO A6 is volume down :)
17:20:52 Quit ChickeNES (Quit: Computer has gone to sleep.)
17:22:13bertrikdualboot works now, but the rockbox bootloader still shows a black screen
17:22:59funmanno backlight?
17:24:01dfkt_it's oled
17:26:57 Join keyb_gr [0] (
17:28:34bertriknow, this is a bit tricky to find out what is preventing rockbox from displaying a logo
17:30:24funmanbertrik: does the screen has special commands to display all-white or something?
17:31:43bertrikI don't know, I saw only thing like auto-scroll and fade-in/out
17:34:17bertrikI added an lcd_enable at the end of the init sequence now
17:43:13 Join ChickeNES [0] (~ChickeNES@
17:43:26 Join mortalis [0] (
17:51:28funmanbertrik: can you downgrade zip OF ?
17:51:50 Quit hilbert (Quit: Computer has gone to sleep.)
17:52:07bertrikfunman, I haven't tried, but dfkt has and the answer is yes
17:59:34 Quit funman (Ping timeout: 240 seconds)
17:59:50 Join funman [0] (~fun@rockbox/developer/funman)
18:06:12***Saving seen data "./dancer.seen"
18:08:10 Quit pamaury (Read error: Operation timed out)
18:12:06bertrikhm, odd, we enable irqs in the as3525 bootloader, but after we've already sent some stuff over i2c (which uses irq)
18:12:35jhMikeSI think that has the provision that if IRQs are disabled, then it busy waits on completion
18:12:59bertrikI'd like to be sure
18:13:12bertrikthe ascodec driver is really complicated
18:13:13*jhMikeS used to be sure, and is no longer
18:13:57jhMikeSit does alot of stuff that will cause very long interrupt latencies
18:14:32bertrikit didn't really like it when it was changed from the much simpler waiting model
18:14:35jhMikeSmaybe not _alot_ but yeah, it can do that :)
18:15:17funmanbertrik: what was the incentive to do that?
18:15:27funmani think it was done by ranmachan
18:15:29jhMikeSlike the beast, I think it needs some asnyc ability
18:15:38bertrikto detect USB in the bootloader, or something like that
18:16:01jhMikeSeven so, it could be simpler even if it allows async transfers
18:16:51jhMikeSit's very similar to the beasts serial driver, but less straighforward
18:21:04bertrikthere is no simple thing on the clip zip that can be observed from outside, for the display to work, other things need to work first: general cpu stuff, i2c, spi
18:21:07 Join hilbert [0] (
18:28:06kugelI would like to commit a patch which doesn't strictly fix a bug, but makes reproducing FS #12325 on the sim possible (targets shouldn't be affected)
18:28:07fs-bluebot e200v1 screen corruption after USB connection since r30475 (bugs, new)
18:28:07funmanno button led, like the clip+ ?
18:28:38kugelperhaps other bugs too, so is it OK if I check it in during freeze?
18:29:20bertrikclip+ doesn't have a button led
18:30:31 Join Stummi [0] (
18:30:34 Quit Stummi (Changing host)
18:30:34 Join Stummi [0] (~Stummi@rockbox/developer/Stummi)
18:33:51-->"ods 100" received from cgthayer (~chatzilla@2620:0:1cfe:61:ec79:1c7f:2e12:1256)
18:34:24 Quit ChickeNES (Read error: Connection reset by peer)
18:34:44 Join ChickeNES [0] (~ChickeNES@
18:36:56 Quit ChickeNES (Read error: Connection reset by peer)
18:37:10 Join ChickeNES [0] (~ChickeNES@
18:38:54CIA-14New commit by kugel (r30845): Fix FS #12325 - screen corruption on early usb. ...
18:39:11kugelsvn is sloooooooooooooow these days
18:39:14kugelat least on commit
18:40:03 Quit Horscht (Quit: Verlassend)
18:41:02CIA-14r30845 build result: All green
18:41:44kugelit really waits until the build is over doesnt it?
18:41:53kugelcan't be coincidence
18:45:06 Quit ChickeNES (Read error: Connection reset by peer)
18:45:12funmangit still not on the horizon line?
18:45:24 Join ChickeNES [0] (~ChickeNES@
18:46:16 Quit ChickeNES (Read error: Connection reset by peer)
18:46:33 Join ChickeNES [0] (~ChickeNES@
18:47:36 Quit TheLemonMan (Quit: WeeChat 0.3.6)
18:54:10 Quit ChickeNES (Read error: Connection reset by peer)
18:54:29 Join ChickeNES [0] (~ChickeNES@
18:54:56semitoneshi, I'm trying to follow the SansaAMSUnbrick page for my clip+, however I'm somewhat confused about what to try next
18:55:48semitonesunder the "hacking" section it describes a drive that's twice as large as the OF file. I seem to have this drive (~32 mb (changes), where the OF is 16mb)
18:55:53semitonesam I on the right track?
18:57:53kugelI'm tracking down FS #12357. I think I have a fix but I need to verify
18:57:53fs-bluebot New backdrop not displayed after set backdrop from context menu (bugs, new)
18:58:18kugelanyway skin_backdrops.c is (IMO) very strange code, it deserves a re-write after the freeze
19:01:31 Quit ChickeNES (Read error: Connection reset by peer)
19:01:53 Join ChickeNES [0] (~ChickeNES@
19:08:02kugelalright, should be good to go
19:09:39CIA-14New commit by kugel (r30846): Fix FS #12357 - New backdrop not displayed after set backdrop from context menu ...
19:09:52kugelno, commit returned immediately
19:10:39 Join AlexP [0] (~alex@rockbox/staff/AlexP)
19:11:25CIA-14r30846 build result: All green
19:15:33 Quit mortalis (Quit: CGI:IRC)
19:15:41funmansemitones: yes it looks good
19:16:01funmani mean, 32 = 2 * 16
19:16:06 Quit ChickeNES (Quit: Computer has gone to sleep.)
19:17:39semitonesfunman: one problem I run into is how do I copy the OF to the drive? I tried dd, but after sending data to the device, the drive read the same
19:18:35 Join Buschel [0] (
19:21:48funmanif dd succeeds just reboot the player
19:22:51semitonesi think dd did not succeed
19:23:41semitoneshere's what I did, maybe you can notice something I can't:
19:23:41CIA-14New commit by buschel (r30847): Fix FS #12351: The poweroff timer was not reset on scroll wheel activity for e200 and simulator. Thanks to Nick Peskett.
19:24:25semitonesI looked at clppa.bin and it did not look like the beginning of firmwarefile1
19:25:29CIA-14r30847 build result: All green
19:25:36funmanno idea what 'firmwarefile1' is
19:26:25funmanis it the backup of what was on the drive?
19:27:14funmanafter running dd, just run 'echo $?'
19:27:32funmanif it shows 0 -> success, not 0 -> failure
19:28:36 Nick kugel is now known as kugelp (~kugel@rockbox/developer/kugel)
19:32:47 Join lorenzo92 [0] (
19:33:05semitonesfunman: alright, I'll try echo $ when I get back home. firmwarefile1 is what was on the drive
19:33:09 Quit funman (Ping timeout: 260 seconds)
19:33:26 Quit Buschel (Quit: ChatZilla 0.9.87 [Firefox 7.0.1/20110928134238])
19:35:48 Join Jerom [0] (~jerome@
19:39:54 Join saratoga [0] (18ad9e5e@gateway/web/freenode/ip.
19:40:28saratogaour ALAC decoder is pretty well optimized thanks to ffmpeg, but it could use some ARM asm functions
19:40:50saratogawe already have similar functions in flac and tta, so it might be possible to adaptthem
19:41:06saratogabut really its alac and it decode in under 40MHz as it is so i never bothered
19:41:13 Quit Thra11 (Ping timeout: 276 seconds)
19:45:06 Quit factor (Read error: Connection reset by peer)
19:45:19 Join Topy [0] (
19:45:40 Quit T44 (Read error: Connection reset by peer)
19:47:06 Join factor [0] (~factor@
19:52:44gevaertssemitones: it's "echo $?", not "echo $"
19:53:54 Join Thra11 [0] (~thrall@
19:53:56semitonesoh, thank you! also, will "cat /dev/sdb" output what's on the drive
19:54:02semitonesso I can see if it looks like the OF
19:54:32gevaertsIt will, but you probably don't want to do that
20:03:58 Quit WalkGood (Quit: ♪ ♫ ♪ ♫ ♪ ♫ ♪ ♫)
20:06:14***Saving seen data "./dancer.seen"
20:07:40 Quit Llorean (Read error: Connection reset by peer)
20:16:29 Join ChickeNES [0] (~ChickeNES@
20:17:06 Quit liar (Remote host closed the connection)
20:20:04 Join robin0800 [0] (~robin0800@
20:24:18 Quit ChickeNES (Quit: Computer has gone to sleep.)
20:45:58 Quit Bagder (Ping timeout: 240 seconds)
20:52:51 Join Riviera [0] (~Riviera@gateway/tor-sasl/riviera)
20:59:30 Quit Jerom (Quit: Leaving.)
21:03:17 Join funman [0] (~fun@rockbox/developer/funman)
21:03:34 Join KiwiCam [0] (
21:07:28 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
21:19:34 Quit robin0800 (Ping timeout: 240 seconds)
21:20:50semitonesgevaerts: how would you suggest I see if the OF was being written to the drive's storage?
21:21:43gevaertssemitones: copy back an equal amount of data with dd and compare checksums
21:22:07 Join robin0800 [0] (~robin0800@
21:22:44semitonesalright, I can try that out when I get home.
21:23:35bertrikhey, I see something on the screen
21:26:54wodzsomething as garbage or something as something useful? :-)
21:27:16funmanwodz: at this point, even garbage is useful ;)
21:27:48 Join Bagder [0] (
21:27:48 Quit Bagder (Changing host)
21:27:48 Join Bagder [241] (~daniel@rockbox/developer/bagder)
21:28:40bertrikATA error: -2
21:28:52bertrikand some ugly rendering of a rockbox logo
21:29:04funmanugly = wrong colors?
21:29:19bertrikit's all shades of blue/cyan
21:29:43bertrikyay, very nice to have at least text on the display
21:29:51funmanno yellow? what a shame, it's a clip after all
21:30:04bertrikunfortunately I can't remember what I did to make it work so far
21:30:29funmanbertrik: any insight on what kind of storage the zip uses?
21:30:48bertrikI think (hope) sd, like the clip+
21:31:10 Join TheLemonMan [0] (
21:35:14 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
21:37:46kugelpbertrik: i remember the cyan logo from the fuzev2 port. that was fixed by going for RGB565SWAPPED pixel format
21:39:02bertrikwhite text is cyan, the edge of the rb logo is blue, the yellow parts of the logo are black
21:39:15bertrikso, I guess I miss the yellow component somehow
21:39:44gevaertsThat will make Saint happy
21:49:47 Quit Riviera (Remote host closed the connection)
21:52:16 Quit lorenzo92 (Quit: ChatZilla 0.9.87 [Firefox 7.0.1/20111008085652])
21:56:38 Join freddyb [0] (
21:59:28freddybkugelp: did you look at the buffer move in dircache? I borked the parenthesis in the patch but you get the idea, right?
21:59:59kugelpyes I had a look
22:00:45kugelpI wasn't aware of even more pointers. I thought I made fd_bindings entirely static
22:03:20freddybThey're static but don't they point into the buflib allocation?
22:05:17 Join n1s [0] (~quassel@rockbox/developer/n1s)
22:05:42kugelpwell your patch shows they point into buflib doesn't it?
22:06:15***Saving seen data "./dancer.seen"
22:06:32freddybThat's why I thought those pointers should get updates, too.
22:09:55 Quit Stummi (Quit: Bye!)
22:12:47n1srasher, saratoga: yeah the RE'd alac decoder looks fairly good and as saratoga said could probably gain a bit of speed from some asm but it contains a couple of comments about things that were not completely understood so maybe those can be fixed up now
22:13:10rasherOh, hurray
22:13:14n1sit would certainly be interesting to compare them
22:13:51n1sthe fir stuff is very well suited to the coldfire emac btw :)
22:15:59n1sideally the ffmpeg guys fix up their codec and we can just cherry pick it ;)
22:18:20 Quit freddyb (Quit: Leaving)
22:18:27bertriknice, the bootloader appears to bootload already
22:18:55bertrikGetting a Prefetch abort at EA000034
22:20:46wodzkugelp: you mentioned that scrollwheel/touchstrip bug may come from some event suppression on screen change - could you give me some hints where to look?
22:21:43CIA-14New commit by bluebrother (r30848): Fix manual URL in Rockbox Utility (FS #12346) ...
22:23:11CIA-14r30848 build result: All green
22:23:24funmanbertrik: SD works?
22:23:46funmanit seems to be the same SD code than clipv2
22:24:13bertrikfunman, yes apparently internal sd works, I see text on the boot screen indicating has calculated a checksum, etc. but it's a bit too fast to read
22:24:18n1shmm, the alac decoder is doing something weird, it seems to be producing interlaced samples, and then deinterlacing them
22:24:40n1swhich the dsp will then interlace again
22:24:47bertrik"Loading rockbox", "Sum: <something>", "Executing", then the prefetch abort
22:27:29funmanis the address useful in prefetch abort?
22:27:57Torneif it's the right address, sure
22:27:57funmancould you show CPU ID ?
22:28:02bertrikLooks like a branch instruction
22:28:17Tornei would hope our handler prints r15 rather than FAR for prefetch :)
22:28:56bertrikhah, me browses a rockbox menu! :D
22:29:13bertrikexcept all the buttons are wrong ... :)
22:29:17bluebrother^a prefetch abort menu?
22:29:44bertrikI disabled some stuff which I'm now re-enabling, like set_cpu_frequency
22:30:20gevaertsbertrik: don't worry. People will keep complaining about buttons being wrong forever!
22:31:13bertrikoh, and radio works :)
22:31:40jhMikeSwell, if they're wrong forever, I don't expect them to cease their objections
22:32:29n1sthe apple code claims to support 20 and 32bit alac which the ffmpeg codec doesn't
22:33:42funmanbertrik: !!
22:33:56n1salso bits are in cpp it seems
22:48:53 Quit n1s (Remote host closed the connection)
22:51:28 Quit hilbert (Quit: Textual IRC Client:
22:52:57 Join Riviera [0] (~Riviera@gateway/tor-sasl/riviera)
22:55:19 Quit funman (Ping timeout: 260 seconds)
22:56:03 Quit y4n (Quit: To those frustrated with their marginal existence, we transmit: your time is now.)
23:01:16 Quit dfkt_ (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
23:02:50bertrikladies and gentlemen, boys and girls, clip zip is playing audio from the microsd card! :D
23:03:13 Join funman [0] (~fun@rockbox/developer/funman)
23:04:37 Quit saratoga (Ping timeout: 265 seconds)
23:05:19wodzthe fastest port ever :-)
23:06:09Bagderbertrik: awesome work!
23:06:44bertrikthanks :)
23:07:39 Quit Thra11 (Quit: kthxbai)
23:16:29kugelpbertrik: Wow
23:16:54kugelpthat's really impressive
23:17:23kugelpconsidering it as mostly a blond port
23:18:12 Quit fs-bluebot (Quit: So long, and thanks for all the fish.)
23:18:24bertrikit's very similar to a clip+ actually, except for the button map and display
23:18:32 Join fs-bluebot [0] (
23:19:57 Quit thomasjfox (Remote host closed the connection)
23:23:51 Join Mouser_X [0] (
23:24:30 Quit funman (Read error: Operation timed out)
23:26:46 Quit Mouser_X (Client Quit)
23:31:24 Join funman [0] (~fun@rockbox/developer/funman)
23:35:12 Join The_Seeker [0] (
23:35:24 Quit wodz (Quit: Leaving)
23:36:59 Quit funman (Ping timeout: 256 seconds)
23:38:34 Quit The_Seeker (Client Quit)
23:40:47 Quit benedikt93 (Quit: Bye ;))
23:40:48 Join PokeYouInDaEye [0] (
23:41:47PokeYouInDaEyeworked for a long time), the folders have folders within, of the same name as the folder I just entered, which I cannot get into.
23:41:57PokeYouInDaEyeso, start again:
23:42:14PokeYouInDaEyeseeking help with my Classic 1st Gen.
23:42:23PokeYouInDaEyemy file-tree navigation seems corrupt
23:42:31PokeYouInDaEyeas of only yesterday
23:42:35PokeYouInDaEyeit was working fine before that
23:45:43 Join funman [0] (~fun@rockbox/developer/funman)
23:47:50*PokeYouInDaEye pets the channel cat
23:49:45rasherbertrik: Nicely done
23:58:50 Join ReimuHak_ [0] (

Previous day | Next day