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 2009-12-08

00:00:39mrtok1i never did a plugin - do i just have to mod this file and add it to sthe sources ?
00:00:44stripwaxThe profiling code will only capture *instrumented* code in that thread. So you need three things - 1. make a Profiling build; 2. instrument the code (i.e. compile with -D$(PROFILE_OPTS) in the build settings for that lib/codec/plugin/etc) and 3. somewhere have a profile_thread that runs in the same thread as the code you want to profile, and runs once early one and a profstop that runs at the end (or some suitable location) later on in
00:01:38***Saving seen data "./dancer.seen"
00:02:21mrtok1to start with the plugin: it should be easier because i can start and stop my plugin from menu?
00:03:20stripwaxI just checked: codec_pcmbuf_insert_callback calls dsp_process. This executes in the same thread as the codec, when you call ci->pcmbuf_insert from codec_main
00:03:49 Quit archivator ("Leaving")
00:12:28 Quit petur ("Zzzzz")
00:14:07 Join Thundercloud [0] (
00:17:56 Quit Omlet05 ("( :: NoNameScript 4.22 :: )")
00:18:27mrtok1stripwax: thank you very much - i think i have to investigate this topic a bit ;-) br
00:19:25stripwaxmrtok1 - test_codec would probably work as-is for you - it offers the possibility of running with dsp effects turned on. so if you profile_thread / profstop in test_codec itself (e.g. profile_thread at start of the test_codec function called "codec_thread" and profstop at end before the 'codec_playing=false' bit) it might 'just work' −−
00:19:53stripwaxmrtok1 (so long as the code you are actually profiling - some dsp code? - is compiled with -D$(PROFILE_OPTS) .. )
00:22:23mrtok1how do i enable test_codec?
00:23:44stripwaxadd test_codec.c to apps/plugins/SOURCES
00:24:13 Quit domonoky1 (Read error: 104 (Connection reset by peer))
00:26:02 Quit Guest47578 (Client Quit)
00:27:44 Quit LambdaCalculus37 ("Leaving")
00:29:12mrtok1stripwax: okay - i give it a try - modded SOURCE, test_codec.c as you suggested - and additionally added -D$(PROFILE_OPTS) to plugin.make - where the CFLAGS are mentioned
00:30:41stripwaxI think you will need it (at a minimum) in the codecs.make file , rather than plugin.make (oh, unless codecs.make reuses the cflags from plugin.make - not sure)
00:31:22 Join FOAD_ [0] (
00:33:22stripwaxWhich dsp code in particular are you trying to profile?
00:34:10saratogaif you just want to profile a dsp function, commenting out the actual decoding test_codec and just feeding the dsp code some stored audio data might be easier
00:34:20saratogabut i don't know how much that will actually tell you
00:34:30saratogathe profiler is for finding slow functions
00:35:35mrtok1saratoga: if there would be another way to measure the code it would be great - but i found no way ...
00:35:57saratogawhat do you want to measure exactly?
00:35:58 Join BHSPitLappy [0] (
00:36:16mrtok1yes - as good as possible
00:36:47mrtok1i thought about highfrequency counter or so?
00:36:54mrtok1using logf
00:40:57mrtok1saratoga: sorry - only saw "exactly" ;-) I wanna measure time spent in my code or how much MIPS I use
00:41:37saratogai don't think the profiler can give you that
00:41:53saratogaif you want to time it, the microsecond timer for your device might be a better choice
00:42:06saratogaor just using test_codec with the DSP enabled and disabled
00:44:49stripwaxright - if you just want to measure the 'real time' impact of enabling the dsp, don't use profiling at all, just run test_codec with and without dsp, and see the resulting decode time and/or MHz reported by test_codec
00:45:10mrtok1saratoga: i thought there would be a cool tool which can disassemble and does the calculation on instruction basis - but I didnt found one
00:45:24saratogathats not possible on this kind of hardware
00:45:40saratogathe profiler is really just a profiler, it tells you what percentage of time is spent in each function
00:46:16mrtok1saratoga: why? - you talk about the cool tool I mentioned?
00:46:28saratogabecause theres no way to get that kind of information
00:47:29 Quit FOAD (Read error: 110 (Connection timed out))
00:47:30 Nick FOAD_ is now known as FOAD (
00:47:46mrtok1but i know how much clocks a specific instruction needs and i know each instruction - so why not?
00:52:25 Quit FlynDice (Remote closed the connection)
00:53:08 Join efyx_ [0] (
00:53:21saratogamrtok1: well you can certainly compute it
00:53:38kugelsaratoga: any interested in updating libfaad?
00:53:45saratogakugel: not really
00:54:00saratogai skimmed the logs and didn't see anything all that appealing done to it
00:54:17kugellibfaad is aac and the like, right?
00:54:23saratogayeah all the aac profiles
00:54:41kugelhow invasive are our changes?
00:54:45saratogaFree AAc Decoder i think
00:54:49saratogaours aren't all that invasive
00:55:10saratogamostly I just cut out the backend of the decoder, the part that handles the inverse transform, windowing, and passes it to rockbox
00:55:14mrtok1saratoga,stripwax: I wonder why there is no such tool ... hower thank you guys - buy!
00:55:26saratogathe front 70% is basically unchanged, and i haven't even looked at SBR
00:55:52saratogamrtok1: well you're welcome to write one, but i think you'll find emulating a complete arm core is not entirely simple
00:56:15saratogaand then all the hardware needed for things like display, etc
00:56:23saratogawe have a partially working one on the tracker somewhere
00:57:02mrtok1;-) good night
00:58:39 Quit mrtok1 ()
01:02:01 Quit kugel (Read error: 104 (Connection reset by peer))
01:02:31stripwaxmrtok1 (if you read the logs) - you could of course just read your disassembler dsp function and see if you can spot any inefficient code, manually ..
01:02:54*stripwax builds libfaad with profiling
01:04:57stripwaxsaratoga - I don't seem to be able to download your file (it wants to open automatically in windows media player). any chance you could put it in a zip archive?
01:05:21saratogastripwax: no wget?
01:05:25stripwaxah, nevermind. link works fine
01:05:42stripwaxlife's too short to wget everything. click, click, click ...
01:05:50 Join FlynDice [0] (
01:06:36stripwaxI have a hunch the mime type might be messed up though. it really wants to download it and rename it as .mp3 ..
01:07:09saratogause chrome it works great
01:08:01 Join phanboy4 [0] (
01:10:33JdGordon|anyone around that can fix the theme page images?
01:11:02 Join PaulJam_ [0] (
01:11:09 Quit PaulJam (Nick collision from services.)
01:11:34stripwaxhopefully libfaad doesn't have fast amounts of inline functions - that really makes profile output garbage. I've put a few __attribute__((no_instrument_function)) around the few functions that are defined as inline that I could fine - hopefully nothing hidden away..
01:14:07 Part captain_kewl
01:14:13saratogaits not well optimized so probably not
01:14:39 Quit flydutch ("/* empty */")
01:16:51stripwaxbizarre. pitch_detector doesn't compile with profiling enables. It barfs on the NO_PROF bit on line 1132. as if NO_PROF is defined empty if profiling disabled; and defined to be something that doesn't compile if profiling enabled
01:17:44stripwaxsaratoga - yeh. hopefully the profile results will make sense. (Remember that time when I was seeing 80% of the vorbis decode time showing up in render_line/render_point, and it turned out to be a red herring due to inlining?)
01:18:36saratogawell we'll see
01:20:06CIA-6New commit by mcuelenaere (r23894): Onda VX747(+)/VX777: update battery discharge curve
01:23:16 Quit Tomis (Read error: 60 (Operation timed out))
01:24:03 Part toffe82
01:32:48 Nick Ypsy is now known as YPSY (
01:33:45 Quit saratoga (Ping timeout: 180 seconds)
01:37:25stripwaxHm. profile build appears to crash out with an undefined instruction shortly after (presumably) hitting the profiling.
01:43:11 Quit Thundercloud (Remote closed the connection)
01:49:26 Quit n17ikh (Read error: 104 (Connection reset by peer))
01:51:19 Join n17ikh [0] (
01:51:26 Join fdinel [0] (
01:53:46stripwaxsaratoga (for the logs) - the illegal instruction address appears to be in iram. so sounds like something really rather messed up somewhere in profile builds.
01:53:51stripwax(so no profiling, for now)
01:55:45stripwaxoh, weird. the address matches the *second* instruction in __cyg_profile_func_enter.
01:57:23 Join Strife89 [0] (
01:59:26stripwax(which seems to disassemble fine according to objdump, so some memory-blatting going on in iram)
02:01:40***Saving seen data "./dancer.seen"
02:02:16 Join GeekShad__ [0] (
02:03:34 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
02:06:01 Quit GeekShado_ (Read error: 145 (Connection timed out))
02:06:26*Unhelpful saw very little marked inline or even static in common.c :/
02:06:53Unhelpfuleven though ones32 and floor_log2 are never called outside of it
02:08:58 Quit BHSPitLappy (Read error: 110 (Connection timed out))
02:16:02stripwaxbooh. so profiling seem borked, at least on ipod video. undefined instruction, as described above, reproducibly. rebuilt from same source as a non-profile build and works fine (though obviously sans profile). can anyone think of something i might have done / not done ?
02:16:08 Join JdGordon2 [0] (
02:21:17 Quit stripwax ("")
02:30:22 Join z35_ [0] (
02:32:44 Quit efyx_ (Remote closed the connection)
02:32:47 Join CaptainKewl [0] (
02:38:16 Quit Soap_ (Read error: 113 (No route to host))
02:53:55 Quit phanboy4 (Read error: 104 (Connection reset by peer))
02:54:09 Join phanboy4 [0] (
02:54:13 Quit JdGordon2 (Read error: 145 (Connection timed out))
02:54:19 Quit z35_ (Client Quit)
02:59:41 Quit PaulJam_ (Read error: 145 (Connection timed out))
03:19:16 Quit GeekShad__ (Read error: 113 (No route to host))
03:24:20kkurbjunJdGordon, with the latest SVN I am still getting the blue background in plugins
03:24:41JdGordontrying to figure out why some update code is crashing :(
03:25:23JdGordonsee anything obvious ?
03:30:04 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
03:31:56kkurbjunhmm, not from what I can see
03:32:18JdGordonNOTE: set_viewport out of bounds: x: -1295507920 y: 32561 width: 9775216 height:0
03:32:25JdGordonno idea where thats coming from :(
03:33:09kkurbjunwhere that message is coming from, or where that viewport is coming from?
03:33:31JdGordonthe latter
03:33:40JdGordoni tihnk my problem might be div0 though
03:33:57JdGordonwould width or height ==0 cause problems in the lcd drivers?
03:34:24kkurbjunthe drivers are mostly do while loops
03:34:35Unhelpfulsurely it would be of *some* value to profile on sim... i mean, we'd at least get an idea which codec funcions take most of the time, even if results are skewed a bit...
03:35:07kkurbjunbut they do bounds checking on everything but the viewport
03:35:51kkurbjunjdgordon, which target are you using teh sim on?
03:35:55Unhelpfulstripwax: you're not also using the eabi toolchain, are you? there are funny things that can do w/ iram load address in some cases
03:36:20JdGordonkkurbjun: mr500... putting in a safety check seems to work
03:36:49kkurbjunthe mr500 has the safety checks enabled in the driver by default
03:37:04kkurbjunit will clip the viewports
03:37:31kkurbjundo you see those messages if you turn off HAVE_VIEWPORT_CLIP?
03:37:41JdGordonwidth or height are getting set to 0 which apparently sticks it in a while(1) loop
03:37:55JdGordonso thats ok.. but now im still getting aretefacts
03:38:16JdGordonah because i havnt fixed it yet :p
03:39:08JdGordonno more splash artefacts, and no more list flicker :)
03:39:13JdGordonnow to fix the boot splash
03:39:36JdGordonkkurbjun: where is that define?
03:40:25kkurbjunit's in the mr500 config, I'm not sure where that is now though
03:41:28kkurbjunline 69
03:41:43JdGordonyep found it.. rebuilding
03:41:59JdGordonlots of error lines still
03:43:29kkurbjunok, I was worried that the clipping might be causing an adverse effect somehow.
03:44:32JdGordonWTF? they disaapear when I add some printf() lines!
03:46:26*JdGordon decides to add a dirty hack to get around the splash problem
03:48:20 Quit Horscht (Read error: 113 (No route to host))
03:50:17kkurbjunJdGordon, are those messages possibly coming from the remote?
03:50:33JdGordoncould be
03:51:05kkurbjunyou might check if they still show after setting the font to 8
03:51:10kkurbjuna size 8 that is
03:51:31kkurbjunand then a 16, and then go over 16
03:52:22JdGordonhow bad really would a second show_splash() be?
03:53:30JdGordon:( looks crap
03:53:55JdGordonthemes need to be the last thing loaded
03:54:28JdGordonscrew it... the fix for that is pulling the theme loading out of settings_apply()...
03:55:28JdGordonok, got a simple fix... care to test?
04:01:44***Saving seen data "./dancer.seen"
04:01:56 Quit DerPapst ("Leaving.")
04:03:11JdGordonkkurbjun: patch updated... there is a tiny flicker on boot, but not going to get any better so good enough
04:03:14*JdGordon wants to commit
04:04:29 Part froggyman
04:06:02kkurbjunI can do some testing a little later
04:06:10kkurbjunI'll give it a try though
04:15:57mc2739FlynDice: I tested your patch and it works without problems on my e260v2 and microsd card.
04:17:01 Join Rondom [0] (
04:17:40FlynDicemc2739: Thanks a bunch
04:18:03 Quit Rondom_ (Read error: 60 (Operation timed out))
04:18:12 Join kcynice [0] (n=kcynice@
04:18:17FlynDicenow if we can get a fuze and a class 2 uSD card I'll be comfortable...
04:19:30kcynicehi, all. this is the first time that i know rockbox, but when i run the latest installer, it says it can't run on my player( 6th).
04:22:13kcynicebut, i can find the download link for 60/80GB iPod in current builds list. what should i do?
04:23:00mc2739kcynice: what model is your player?
04:23:44kcynice80GB. but im not sure why it told me this is 6th version.
04:24:04kcyniceim trying to install as 5th
04:25:20krazykitkcynice, try to identify your model using this page:
04:26:21krazykitif it is a classic, you're out of luck, as rockbox doesn't work on the 6th gen devices.
04:26:54kcyniceyes. classic
04:28:43kcyniceno way, no choice?
04:30:04krazykitlike i said, rockbox doesn't run on that device.
04:32:00kcynicewhat a pity
04:39:36mc2739FlynDice: btw, my microsd card is a class 2. It works fine, and shows up in the debug screen as 50 MBits/s
04:43:42kkurbjunJdGordon, you are on an older svn version
04:43:55kkurbjunI'm getting alot of artifacts after leaving plugins
04:44:07kkurbjunand the plugin colors are still messed up in the menus
04:44:20kkurbjunI see that light blue background
04:44:40kkurbjunthe flickering is back inthe SBS when you transition top level menu entries
04:45:18 Join froggyman [0] (
04:45:40kkurbjunapplying the custom sbs in the sbs menu is acting worse than the last patch too..
04:47:04FlynDicemc2739: re microsd I find that interesting, I was starting to think that perhaps a class 2 wasn't a HS card, but yours shows 50.0 so that can't be it. Maybe we need a delay in there somewhere
04:47:56kkurbjunthe sbs still shows the album art after the playlist finishes playing too
04:48:05kkurbjunand the scrolling lines persist
04:48:27kkurbjunactually, now they don't go away even when transitioning menus
04:48:54kkurbjunwell, after a few menu transitions they go away
04:49:08kkurbjunthe splashes clear properly now
04:49:42kkurbjunafter hitting resume a couple of times I got a stack overflow though
04:50:02kkurbjun*PANIC* Stack overflow... viewportmanager
04:53:40kkurbjunI should clarify, that scrolling lines do not show in the plugins anymore though
04:56:05kkurbjunJdGordon: the sbs glitches do not show when music is not playing, but if I start playing music, the sbs starts to glitch when I transition the top-level menu items.
04:58:37JdGordonok, ill have a looky
05:02:54 Quit kcynice ("Leaving")
05:03:25 Join dukeman [0] (n=dukeman@
05:05:04dukemanSansa e200 with current build of rockbox installed via rbutil-qt. Host computer runs Linux x86_64. USB transfers yield I/O errors and the device node for the sansa disappears when anything is done with USB HID mode disabled
05:05:34dukemanTurning on USB HID mode in the player (in "Multimedia" mode, for example) and then plugging it in, transferring files, etc works
05:06:10dukemanWhat information do I submit to help locate the issue?
05:08:21froggymanis anyone aware of the device image problems on the theme page?
05:16:20 Quit Xerion (Read error: 104 (Connection reset by peer))
05:16:53 Join Xerion [0] (
05:17:32 Part froggyman
05:31:51 Quit goffa (Read error: 113 (No route to host))
05:42:08 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
05:44:03 Join Tomis [0] (n=Tomis@
05:45:09 Join goffa [0] (n=goffa@
05:49:46 Quit Horschti ("Verlassend")
05:51:12dukemanActually, disregard that. Closer inspection reveals the same problem appears regardless of HID setting and also in the original firmware. I'll chalk it up to hardware failure and replace the grumpy old brick.
05:52:05 Part dukeman
06:01:48***Saving seen data "./dancer.seen"
06:04:37 Quit panni_ ("( :: NoNameScript 3.81 :: )")
06:47:23 Quit Strife89 ("See ya.")
06:53:46 Quit ansuz ("A motion to ajourn is always in order.")
07:05:31 Quit bluebrother (Nick collision from services.)
07:05:34 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
07:12:50 Quit Tomis ()
07:16:42 Join Tomis [0] (n=Tomis@
07:22:04 Quit z35 (Read error: 110 (Connection timed out))
07:22:44 Quit JdGordon ("Leaving.")
07:40:22 Quit n17ikh (Connection timed out)
07:43:09 Quit CaptainKewl (Remote closed the connection)
07:48:43 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
07:50:11 Join PaulJam_ [0] (
07:54:53 Join einhirn [0] (
07:55:36 Quit Tomis ()
07:56:38 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
07:59:42 Join jae [0] (
08:00:15 Join Tomis [0] (n=Tomis@
08:01:52***Saving seen data "./dancer.seen"
08:10:03topiki'm compiling in your patch now, FlynDice
08:13:30 Join n17ikh [0] (
08:17:34 Quit Tomis ()
08:26:02 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:28:09topikFlynDice: so far your patch doesn't seem to have negative effects. playing flac smooth for a while now from the uSD card in my Fuze. silly card is still at 25Mbit/s though but that's probably the card's fault.
08:32:52 Join Rob2222 [0] (
08:36:35FlynDicetopik: Thanks
08:38:54topikany way i could test writing to check for those aborts?
08:39:02topikor crc failures even
08:40:01 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
08:40:11 Quit Rob2223 (Read error: 60 (Operation timed out))
08:44:29JdGordonkkurbjun: you round?
08:45:15 Join Tomis [0] (n=Tomis@
08:48:31 Join pondlife [50] (n=Steve@rockbox/developer/pondlife)
08:51:03 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
08:51:57FlynDicetopik: The easiest way is to copy a directory from internal to uSD & then back to internal. Using test_disk is better but you have to add it to a SOURCES file and then build & I've found the copy test just as reliable
08:54:20JdGordonstupid splashf() is causing wierd glitches :(
08:56:31*JdGordon is an idiot :p
08:58:26JdGordon is now really ready to go... can I get some testing?
08:59:03pondlifeJdGordon: I was just testing the previous one
08:59:18JdGordonthe previous one is fucked :)
09:00:52 Quit arohtar (Read error: 104 (Connection reset by peer))
09:01:14 Join arohtar [0] (n=faemir@
09:03:13topikcopying is not very fast
09:04:37 Join maruk [0] (
09:07:48 Quit gevaerts (Nick collision from services.)
09:08:00 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
09:08:46JdGordonI tihnk the last issue is changing from non sbs to inbuilt bar, which is not something that I particularly care about... but good to know
09:15:10FlynDiceGotta take some people back to Seattle now, I'll check back when I get in
09:15:44 Quit FlynDice (Remote closed the connection)
09:16:01 Quit Dhraakellian (Remote closed the connection)
09:16:09pixelmaJdGordon: I didn't follow the patch yet but remember some discussion - did you really remove the battery and volume display settings? If not - the could use the %St tag for displaying either depending on the user's choice but that would need preparation with conditional viewports. The viewports themselves are set up correctly (if you comment the lines out in the .sbs)
09:16:29JdGordonI havnt yet
09:16:36JdGordonthis patch doesnt remove the inbuilt bar yet
09:16:50JdGordonjust cleans stuff up
09:16:58JdGordonand fixes some redraw bugs
09:17:58pixelmathen my remark is something to keep in mind for the case when it is removed ;)
09:18:27JdGordonok cool
09:19:26JdGordonI'm going to bed now... if anyone has a few minutes please test the above patch... I'll probably commit it tomorow night my time unless I hear bad stuff about it...
09:19:43JdGordonthen have a mad race to skin fm and rec before release :p
09:22:51topikFlynDice: my test copying 165MB from uSD to internal, then back resulted in identical files. it's only a small test, but I think your patch might solve the corruption
09:25:02 Join Thundercloud [0] (
09:25:52 Join petur [0] (n=petur@
09:28:19 Quit Tomis (Read error: 110 (Connection timed out))
09:40:30linuxstbZagor: Good morning. I was wondering why you made this change, instead of renaming the bootloaders to bootloader-$modelname.sansa -
09:41:00linuxstbAlso, shouldn't bootloader-ipodnano.ipod be renamed bootloader-ipodnano1g.ipod on the download server (with a symlink)
09:41:17Zagorlinuxstb: simply because I didn't touch any bootloader names. I don't know which (if any) need to have specific names.
09:42:26 Join DerPapst [0] (
09:44:44 Quit Thundercloud (Read error: 104 (Connection reset by peer))
09:44:46linuxstbThe "bootloader-*.sansa" names can be changed to anything. It's only things like "firmware.mi4" or "pp5022.mi4" which need to stay like that.
09:45:07 Join funman [0] (n=fun@rockbox/developer/funman)
09:47:28 Join flydutch [0] (
09:52:55 Quit phanboy4 (Read error: 104 (Connection reset by peer))
09:53:26 Join phanboy4 [0] (
09:58:13 Quit n17ikh ()
10:01:53***Saving seen data "./dancer.seen"
10:04:44 Quit arohtar (Client Quit)
10:05:49 Quit JdGordon1 ("Leaving.")
10:14:50 Quit pondlife ("Leaving.")
10:18:53 Quit amiconn (Nick collision from services.)
10:18:55 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
10:19:15 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
10:20:35 Join Rob2223 [0] (
10:31:04 Join kugel [0] (i=kugel@rockbox/developer/kugel)
10:31:54kugelJdGordon: doesn't your splash.c change mean that non-blocking ones don't show (or only show for a fractional second)?
10:32:17kugelclearing directly in splash.c only works for blocking ones, I already experimented with that
10:34:50kugelalso you remove too much, add_event(LCD_EVENT_ACTIVATION, false, do_update_callback); is particularly unrelated
10:35:02kugelbut I'm not surprised that you blindly remove it
10:36:33 Quit Rob2222 (Read error: 113 (No route to host))
10:38:04kugel+ FOR_NB_SCREENS(i)
10:38:04kugel+ screens[i].stop_scroll();" is bad too IMO
10:38:28gevaertsindeed it is. That should be on a pastebin!
10:38:33*gevaerts hides
10:40:37kugelI though 2 lines are OK, but I didn't notice that it's going to be 3 with an empty one :(
10:46:47kugelI mean, the previous screen should be stop scrolling on its viewports, or the sbs on all of its viewports. just doing it for the whole screen isn't nice. although it's probably not that bad in this case
10:48:29kugelit might be better to lcd_update() the whole display in toggle_theme, instead of each deadpart vp separately, to reduce flicker
10:50:49Zagorkugel: how can I specify "no options" with −−ttsopts ?
10:50:58Zagoror rather, default options
10:51:18Zagor−−ttsopts='' still gives me an interactive question
10:51:57kugelno idea, it might need some code changes.
10:53:02kugeladvanced options via command line is not my work, I just tested it. I may not have tested this case though
10:54:10Zagoraha. the ttsopts stuff was last touched by you :-)
10:55:22kugelcan't you leave out −−ttsopts entirely?
10:55:35Zagorno, that gives me the interactive question
10:55:40kugelah right
10:56:19 Quit phanboy4 (Read error: 113 (No route to host))
10:56:37kugelwell, a special value for "take the damn defaults" isn't so usual so I assume you need to pass them. or add such a special value
11:07:01 Join Bagder [0] (
11:16:01Zagorthe manual build looks broken
11:17:29Bagderdid you check the log?
11:17:41ZagorI'm trying to build
11:18:31 Join shai [0] (
11:19:13 Join Omlet [0] (i=omlet05@
11:25:01CIA-6New commit by zagor (r23895): Do not prompt for encoder and tts options. They are now command line only.
11:31:09pixelmaZagor: can you translate this commit message for me a bit? How do I pass tts options now?
11:31:19Zagorpixelma: −−ttsopts=
11:32:07pixelmafor "make voice"?
11:32:16Zagorno, for ../tools/configure
11:37:00 Quit kugel (Read error: 110 (Connection timed out))
11:48:38 Join DerPapst1 [0] (
11:58:04 Quit Sajber^ (Read error: 131 (Connection reset by peer))
12:01:55***Saving seen data "./dancer.seen"
12:06:14 Quit DerPapst (Read error: 110 (Connection timed out))
12:11:17 Join Tomis [0] (n=Tomis@
12:11:58 Join kugel [0] (i=kugel@rockbox/developer/kugel)
12:13:00 Quit jae (Read error: 104 (Connection reset by peer))
12:15:43 Join watto [0] (n=watto@
12:17:46 Quit amiconn (
12:17:46 Quit einhirn (
12:17:46 Quit Hadaka (
12:17:46 Quit jon-kha (
12:17:46 Quit Slasheri (
12:17:46 Quit HellDragon (
12:17:46 Quit antil33t (
12:17:46 Quit liar (
12:17:46 Quit crwll (
12:17:46 Quit HBK (
12:17:46 Quit tchan (
12:17:46 Quit advcomp2019 (
12:17:46 Quit niekie (
12:17:46 Quit mc2739 (
12:17:46 Quit bzed (
12:17:46 Quit elcan (
12:17:46 Quit lostlogic (
12:17:46 Quit chaos (
12:17:46 Quit DerPapst1 (
12:17:46 Quit funman (
12:17:46 Quit gevaerts (
12:17:46 Quit BHSPitMonkey (
12:17:46 Quit goffa (
12:17:46 Quit jasio (
12:17:46 Quit LinusN (
12:17:46 Quit AlexP (
12:17:46 Quit seani (
12:17:46 Quit aevin (
12:17:46 Quit jfc (
12:17:46 Quit ehntoo_ (
12:17:46 Quit J-23 (
12:17:46 Quit PaulJam_ (
12:17:46 Quit Xerion (
12:17:46 Quit Topy44 (
12:17:46 Quit rasher (
12:17:46 Quit ThomasAH (
12:19:02NJoinDerPapst1 [0] (
12:19:02NJoinamiconn [0] (i=quassel@rockbox/developer/amiconn)
12:19:02NJoinfunman [0] (n=fun@rockbox/developer/funman)
12:19:02NJoingevaerts [0] (n=fg@rockbox/developer/gevaerts)
12:19:02NJoinLinusN [0] (n=linus@rockbox/developer/LinusN)
12:19:02NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
12:19:02NJoineinhirn [0] (
12:19:02NJoinPaulJam_ [0] (
12:19:02NJoingoffa [0] (n=goffa@
12:19:02NJoinXerion [0] (
12:19:02NJoinTopy44 [0] (
12:19:02 Join HellDragon [0] (n=jd@Wikipedia/HellDragon)
12:19:02NJoinantil33t [0] (
12:19:02NJoinliar [0] (n=liar@
12:19:02NJoinHadaka [0] (
12:19:02NJoincrwll [0] (
12:19:02NJoinHBK [0] (
12:19:02NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
12:19:02NJoinadvcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
12:19:02NJoinniekie [0] (i=quasselc@CAcert/Assurer/niekie)
12:19:02NJoinmc2739 [0] (n=mc2739@rockbox/developer/mc2739)
12:19:02NJoinbzed [0] (
12:19:02NJoinjon-kha [0] (
12:19:02NJoinelcan [0] (
12:19:02NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
12:19:02NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
12:19:02NJoinjasio [0] (
12:19:02NJoinchaos [0] (n=chaos@gentoo/user/ch4os)
12:19:02NJoinAlexP [0] (n=alex@rockbox/staff/AlexP)
12:19:02 Join rasher [50] (n=rasher@rockbox/developer/rasher)
12:19:02NJoinseani [0] (n=seani@
12:19:02NJoinaevin [0] (i=eivindsy@unaffiliated/aevin)
12:19:02NJoinjfc [0] (
12:19:02NJoinehntoo_ [0] (
12:19:02NJoinJ-23 [0] (
12:19:02NJoinThomasAH [0] (
12:20:10 Join jae [0] (n=jae@
12:22:44kugelZagor: you now dropped the interactive tts questions entirely?
12:22:44 Quit Topy44 (Remote closed the connection)
12:23:02Zagorkugel: only ttsopt and encoderopt
12:23:56kugelI don't use voice so it doesn't touch me really, but I wonder why that was necessary?
12:25:25 Quit ChanServ (
12:25:28Zagorbecause we need to be able to run configure non-interactively. the alternative is adding extra "nottsopt" and "noencoderopt" parameters
12:25:59NJoinChanServ [0] (ChanServ@services.)
12:25:59Mode"#rockbox +o ChanServ " by
12:26:20kugelor −−ttsopts=default, that sounds more appealing than dropping the interactivity
12:26:42 Join Topy44 [0] (
12:27:54ZagorI don't consider interactivity for encoder and tts _options_ a very important feature
12:47:51 Quit PaulJam_ (Read error: 110 (Connection timed out))
12:53:59 Quit shai (Read error: 110 (Connection timed out))
12:56:12 Quit jae (
12:56:12 Quit Slasheri (
12:56:12 Quit jon-kha (
12:56:12 Quit einhirn (
12:56:12 Quit Hadaka (
12:56:12 Quit amiconn (
12:56:12 Quit chaos (
12:56:12 Quit lostlogic (
12:56:12 Quit mc2739 (
12:56:12 Quit crwll (
12:56:12 Quit antil33t (
12:56:12 Quit elcan (
12:56:12 Quit HBK (
12:56:12 Quit bzed (
12:56:12 Quit tchan (
12:56:12 Quit advcomp2019 (
12:56:12 Quit HellDragon (
12:56:12 Quit liar (
12:56:12 Quit niekie (
12:56:12 Quit Topy44 (
12:56:12 Quit jasio (
12:56:12 Quit goffa (
12:56:12 Quit funman (
12:56:12 Quit DerPapst1 (
12:56:12 Quit gevaerts (
12:56:12 Quit BHSPitMonkey (
12:56:12 Quit seani (
12:56:12 Quit AlexP (
12:56:12 Quit J-23 (
12:56:12 Quit ehntoo_ (
12:56:12 Quit jfc (
12:56:12 Quit aevin (
12:56:12 Quit LinusN (
12:56:12 Quit rasher (
12:56:12 Quit ThomasAH (
12:56:12 Quit Xerion (
12:56:12 Join dfkt [0] (
12:59:50 Join Slasheri [0] (i=miipekk@rockbox/developer/Slasheri)
12:59:50 Join jon-kha [0] (
12:59:50 Join Hadaka [0] (
12:59:50 Join einhirn [0] (
12:59:50 Join amiconn [0] (i=quassel@rockbox/developer/amiconn)
12:59:50 Join jae [0] (n=jae@
12:59:50 Join jasio [0] (
12:59:50 Join goffa [0] (n=goffa@
12:59:50 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
12:59:50 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
12:59:50 Join funman [0] (n=fun@rockbox/developer/funman)
12:59:50 Join DerPapst1 [0] (
12:59:50 Join Topy44 [0] (
12:59:50 Join chaos [0] (n=chaos@gentoo/user/ch4os)
12:59:50 Join lostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
12:59:50 Join elcan [0] (
12:59:50 Join bzed [0] (
12:59:50 Join mc2739 [0] (n=mc2739@rockbox/developer/mc2739)
12:59:50 Join niekie [0] (i=quasselc@CAcert/Assurer/niekie)
12:59:50 Join advcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
12:59:50 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
12:59:50 Join HBK [0] (
12:59:50 Join crwll [0] (
12:59:50 Join liar [0] (n=liar@
12:59:50 Join antil33t [0] (
12:59:50 Join HellDragon [0] (n=jd@Wikipedia/HellDragon)
13:00:31 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
13:00:31 Join AlexP [0] (n=alex@rockbox/staff/AlexP)
13:00:31 Join seani [0] (n=seani@
13:00:31 Join aevin [0] (i=eivindsy@unaffiliated/aevin)
13:00:31 Join jfc [0] (
13:00:31 Join ehntoo_ [0] (
13:00:31 Join J-23 [0] (
13:00:40 Join Xerion [0] (
13:00:40 Join rasher [50] (n=rasher@rockbox/developer/rasher)
13:00:40 Join ThomasAH [0] (
13:01:31 Quit BHSPitMonkey (Remote closed the connection)
13:05:40 Quit J-23 (
13:05:40 Quit ehntoo_ (
13:05:40 Quit seani (
13:05:40 Quit AlexP (
13:05:40 Quit jfc (
13:05:40 Quit aevin (
13:05:40 Quit LinusN (
13:05:40 Quit ThomasAH (
13:05:40 Quit rasher (
13:05:40 Quit Xerion (
13:05:40 Quit antil33t (
13:05:40 Quit crwll (
13:05:40 Quit mc2739 (
13:05:40 Quit lostlogic (
13:05:40 Quit chaos (
13:05:40 Quit elcan (
13:05:40 Quit HBK (
13:05:40 Quit tchan (
13:05:40 Quit bzed (
13:05:40 Quit liar (
13:05:40 Quit niekie (
13:05:40 Quit advcomp2019 (
13:05:40 Quit HellDragon (
13:05:40 Quit amiconn (
13:05:40 Quit einhirn (
13:05:40 Quit jon-kha (
13:05:40 Quit Slasheri (
13:05:40 Quit jae (
13:05:40 Quit Hadaka (
13:05:40 Quit Topy44 (
13:05:40 Quit DerPapst1 (
13:05:40 Quit goffa (
13:05:40 Quit jasio (
13:05:40 Quit funman (
13:05:40 Quit gevaerts (
13:06:38NJoinThomasAH [0] (
13:06:38NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
13:06:38NJoinXerion [0] (
13:06:38NJoinJ-23 [0] (
13:06:38NJoinehntoo_ [0] (
13:06:38NJoinjfc [0] (
13:06:38NJoinaevin [0] (i=eivindsy@unaffiliated/aevin)
13:06:38NJoinseani [0] (n=seani@
13:06:38NJoinAlexP [0] (n=alex@rockbox/staff/AlexP)
13:06:38NJoinLinusN [0] (n=linus@rockbox/developer/LinusN)
13:06:38NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
13:06:38NJoinjon-kha [0] (
13:06:38NJoinHadaka [0] (
13:06:38NJoineinhirn [0] (
13:06:38NJoinamiconn [0] (i=quassel@rockbox/developer/amiconn)
13:06:38NJoinjae [0] (n=jae@
13:06:38NJoinjasio [0] (
13:06:38NJoingoffa [0] (n=goffa@
13:06:38NJoingevaerts [0] (n=fg@rockbox/developer/gevaerts)
13:06:38NJoinfunman [0] (n=fun@rockbox/developer/funman)
13:06:38NJoinDerPapst1 [0] (
13:06:38NJoinTopy44 [0] (
13:06:38NJoinchaos [0] (n=chaos@gentoo/user/ch4os)
13:06:38NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
13:06:38NJoinelcan [0] (
13:06:38NJoinbzed [0] (
13:06:38NJoinmc2739 [0] (n=mc2739@rockbox/developer/mc2739)
13:06:38NJoinniekie [0] (i=quasselc@CAcert/Assurer/niekie)
13:06:38NJoinadvcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
13:06:38NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
13:06:38NJoinHBK [0] (
13:06:38NJoincrwll [0] (
13:06:38NJoinliar [0] (n=liar@
13:06:38NJoinantil33t [0] (
13:06:38NJoinHellDragon [0] (n=jd@Wikipedia/HellDragon)
13:12:02 Quit LinusN (
13:12:02 Quit seani (
13:12:02 Quit ehntoo_ (
13:12:02 Quit aevin (
13:12:02 Quit AlexP (
13:12:02 Quit jfc (
13:12:02 Quit J-23 (
13:12:02 Quit rasher (
13:12:02 Quit ThomasAH (
13:12:02 Quit Xerion (
13:12:02 Quit antil33t (
13:12:02 Quit crwll (
13:12:02 Quit mc2739 (
13:12:02 Quit lostlogic (
13:12:02 Quit chaos (
13:12:02 Quit elcan (
13:12:02 Quit HBK (
13:12:02 Quit tchan (
13:12:02 Quit bzed (
13:12:02 Quit liar (
13:12:02 Quit niekie (
13:12:02 Quit advcomp2019 (
13:12:02 Quit HellDragon (
13:12:02 Quit amiconn (
13:12:02 Quit einhirn (
13:12:02 Quit jon-kha (
13:12:02 Quit Slasheri (
13:12:02 Quit jae (
13:12:02 Quit Hadaka (
13:12:02 Quit Topy44 (
13:12:02 Quit DerPapst1 (
13:12:02 Quit goffa (
13:12:02 Quit jasio (
13:12:02 Quit funman (
13:12:02 Quit gevaerts (
13:12:36NJoinThomasAH [0] (
13:12:36NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
13:12:36NJoinXerion [0] (
13:12:36NJoinJ-23 [0] (
13:12:36NJoinehntoo_ [0] (
13:12:36NJoinjfc [0] (
13:12:36NJoinaevin [0] (i=eivindsy@unaffiliated/aevin)
13:12:36NJoinseani [0] (n=seani@
13:12:36NJoinAlexP [0] (n=alex@rockbox/staff/AlexP)
13:12:36NJoinLinusN [0] (n=linus@rockbox/developer/LinusN)
13:12:36NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
13:12:36NJoinjon-kha [0] (
13:12:36NJoinHadaka [0] (
13:12:36NJoineinhirn [0] (
13:12:36NJoinamiconn [0] (i=quassel@rockbox/developer/amiconn)
13:12:36NJoinjae [0] (n=jae@
13:12:36NJoinjasio [0] (
13:12:36NJoingoffa [0] (n=goffa@
13:12:36NJoingevaerts [0] (n=fg@rockbox/developer/gevaerts)
13:12:36NJoinfunman [0] (n=fun@rockbox/developer/funman)
13:12:36NJoinDerPapst1 [0] (
13:12:36NJoinTopy44 [0] (
13:12:36NJoinchaos [0] (n=chaos@gentoo/user/ch4os)
13:12:36NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
13:12:36NJoinelcan [0] (
13:12:36NJoinbzed [0] (
13:12:36NJoinmc2739 [0] (n=mc2739@rockbox/developer/mc2739)
13:12:36NJoinniekie [0] (i=quasselc@CAcert/Assurer/niekie)
13:12:36NJoinadvcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
13:12:36NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
13:12:36NJoinHBK [0] (
13:12:36NJoincrwll [0] (
13:12:36NJoinliar [0] (n=liar@
13:12:36NJoinantil33t [0] (
13:12:36NJoinHellDragon [0] (n=jd@Wikipedia/HellDragon)
13:12:57 Quit tarbo (SendQ exceeded)
13:13:11 Quit Tomis (Remote closed the connection)
13:13:34 Join Tomis [0] (n=Tomis@
13:15:22 Quit LinusN (
13:15:22 Quit seani (
13:15:22 Quit ehntoo_ (
13:15:22 Quit aevin (
13:15:22 Quit AlexP (
13:15:22 Quit jfc (
13:15:22 Quit J-23 (
13:15:22 Quit rasher (
13:15:22 Quit ThomasAH (
13:15:22 Quit Xerion (
13:15:22 Quit antil33t (
13:15:22 Quit crwll (
13:15:22 Quit mc2739 (
13:15:22 Quit lostlogic (
13:15:22 Quit chaos (
13:15:22 Quit elcan (
13:15:22 Quit HBK (
13:15:22 Quit tchan (
13:15:22 Quit bzed (
13:15:22 Quit liar (
13:15:22 Quit niekie (
13:15:22 Quit advcomp2019 (
13:15:22 Quit HellDragon (
13:15:22 Quit amiconn (
13:15:22 Quit einhirn (
13:15:22 Quit jon-kha (
13:15:22 Quit Slasheri (
13:15:22 Quit jae (
13:15:22 Quit Hadaka (
13:15:22 Quit Topy44 (
13:15:22 Quit DerPapst1 (
13:15:22 Quit goffa (
13:15:22 Quit jasio (
13:15:22 Quit funman (
13:15:22 Quit gevaerts (
13:15:27 Join tarbo [0] (n=me@unaffiliated/tarbo)
13:16:34NJoinHellDragon [0] (n=jd@Wikipedia/HellDragon)
13:16:34NJoinantil33t [0] (
13:16:34NJoinliar [0] (n=liar@
13:16:34NJoincrwll [0] (
13:16:34NJoinHBK [0] (
13:16:34NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
13:16:34NJoinadvcomp2019 [0] (n=advcomp2@unaffiliated/advcomp2019)
13:16:34NJoinniekie [0] (i=quasselc@CAcert/Assurer/niekie)
13:16:34NJoinmc2739 [0] (n=mc2739@rockbox/developer/mc2739)
13:16:34NJoinbzed [0] (
13:16:34NJoinelcan [0] (
13:16:34NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
13:16:34NJoinchaos [0] (n=chaos@gentoo/user/ch4os)
13:19:37 Join archivator [0] (n=archivat@
13:19:49NJoinTopy44 [0] (
13:19:49NJoinDerPapst1 [0] (
13:19:49NJoinfunman [0] (n=fun@rockbox/developer/funman)
13:19:49NJoingevaerts [0] (n=fg@rockbox/developer/gevaerts)
13:19:49NJoingoffa [0] (n=goffa@
13:19:49NJoinjasio [0] (
13:20:33NJoinXerion [0] (
13:20:33NJoinrasher [50] (n=rasher@rockbox/developer/rasher)
13:20:33NJoinThomasAH [0] (
13:20:36 Quit preglow_ (
13:20:36 Quit topik (
13:20:36 Quit dionoea (
13:20:36 Quit SIGSEGV (
13:20:36 Quit Kohlrabi (
13:20:36 Quit YPSY (
13:20:36 Quit zu_ (
13:21:43NJoinjae [0] (n=jae@
13:21:43NJoinamiconn [0] (i=quassel@rockbox/developer/amiconn)
13:21:43NJoineinhirn [0] (
13:21:43NJoinHadaka [0] (
13:21:43NJoinjon-kha [0] (
13:21:43NJoinSlasheri [0] (i=miipekk@rockbox/developer/Slasheri)
13:21:45NJoinYPSY [0] (
13:21:45NJoinpreglow_ [0] (
13:21:45NJoindionoea [0] (
13:21:45NJoinSIGSEGV [0] (n=user@
13:21:45NJoinzu_ [0] (
13:21:45NJointopik [0] (
13:21:45NJoinKohlrabi [0] (
13:22:16NJoinLinusN [0] (n=linus@rockbox/developer/LinusN)
13:22:16NJoinAlexP [0] (n=alex@rockbox/staff/AlexP)
13:22:16NJoinseani [0] (n=seani@
13:22:16NJoinaevin [0] (i=eivindsy@unaffiliated/aevin)
13:22:16NJoinjfc [0] (
13:22:16NJoinehntoo_ [0] (
13:22:16NJoinJ-23 [0] (
13:22:18 Quit tarbo (SendQ exceeded)
13:23:00 Join MethoS- [0] (n=clemens@
13:29:27CIA-6New commit by mcuelenaere (r23896): Onda VX777: use POWER button to exit Plasma and Starfield
13:29:31 Join angelwolf71885 [0] (
13:31:24 Join z35 [0] (
13:32:19 Quit ps-auxw (
13:32:19 Quit markun (
13:32:19 Quit crashd (
13:32:19 Quit Unhelpful (
13:32:19 Quit fish_ (
13:32:19 Quit rjg (
13:32:19 Quit scorche (
13:32:53NJoinps-auxw [0] (
13:32:53NJoinscorche [50] (n=scorche@rockbox/administrator/scorche)
13:32:53NJoinUnhelpful [0] (n=quassel@rockbox/developer/Unhelpful)
13:32:53NJoincrashd [0] (
13:32:53NJoinrjg [0] (
13:32:53NJoinfish_ [0] (
13:32:53NJoinmarkun [50] (n=markun@rockbox/developer/markun)
13:34:56 Quit z35 (Read error: 113 (No route to host))
13:35:25 Join tarbo [0] (n=me@unaffiliated/tarbo)
13:36:46 Join z35 [0] (
13:55:05 Join teru [0] (
14:01:32CIA-6New commit by rmenes (r23897): GoGear SA9200 plugin keymap for snake2. Just get it under ...
14:01:58***Saving seen data "./dancer.seen"
14:04:02 Quit antil33t (Read error: 110 (Connection timed out))
14:04:53 Quit angelwolf71885 ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
14:05:28 Join angelwolf71885 [0] (
14:11:59 Join antil33t [0] (
14:15:19 Join PaulJam [0] (
14:30:54 Join Robotnik [0] (
14:32:51RobotnikSorry for disturbing you guys so early in the morning, but I was wondering what was the "OF"... because trying to find that word through the search engine gave me nothing.
14:33:25Robotnik(somekind of OF for file transferts)
14:33:27funmanOF = Original Firmware
14:33:40RobotnikOh thank you! : D
14:35:11 Join Jaykay [0] (
14:37:40Jaykaybluebroth3r: could you close FS #8025?
14:38:45 Quit Robotnik ("CGI:IRC")
14:40:21 Quit angelwolf71885 ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
14:40:48 Join angelwolf71885 [0] (
14:47:48kugelJaykay: I added it as related that should be enough for now
14:50:51Jaykaykugel: would merging the three tasks into one new task help in any way?
14:50:52CIA-6New commit by zagor (r23898): Renamed file
14:51:07kugelI don't think so
14:56:28 Join FlynDice [0] (
14:58:48 Join stoffel [0] (
15:03:19 Nick Omlet is now known as Omlet^away (i=omlet05@
15:08:02 Quit funman ("free(random());")
15:28:01Jaykaykugel: they are duplicates, why didn't you want to delete them?
15:32:30 Join jgarvey [0] (
15:34:11 Quit DerPapst1 ("Leaving.")
15:36:23kugelbecause I can't tell for sure if it's the same bug or just similar symptomes
15:39:04 Nick YPSY is now known as Ypsy (
15:40:21*Jaykay would vote for "same bug"
15:40:52Jaykayanother thing: database creation is extremely slow when the disk is full...
15:41:30Jaykaykugel: how about FS #8554? it says rockbox freezes, but it doesnt
15:42:14Jaykayerm, a question: the database is built in two steps (before and after rebooting), does rockbox write in both steps on disk? is?
15:43:22Tornewhat do yo umean, two steps?
15:43:47Jaykaytorne: as i said, before and after booting
15:44:12Jaykaybefore rebooting it says something like "searching.... (x found) while x is increasing
15:44:29Jaykayand after rebooting it says "committing.... (x/9)
15:44:53Tornereally? i've neve rhad to reboot to build the db
15:46:04Jaykaywtf... you are also speaking from creating a new database on a fresh rockbox right?
15:48:20Torneyou have to reboot to enable the db..
15:48:30Tornei may be remembering wrong..
15:48:33Tornebut i've done it quite a few times
15:48:49Tornehave been formatting my ipod a lot lately :) you also need to reboot, wheres the problem then?
15:48:57pixelmamaybe it is related to having dircache and "load database to RAM" enabled and/or using (auto)update. I usually don't have either (sometimes dircache though) and have to wait for the committing step on next reboot to have the new entries available
15:48:58 Quit gentgeen__ (Remote closed the connection)
15:49:13TorneJaykay: i mean, you reboot to enable the setting
15:49:18Tornenothing is *done* before
15:49:23Torneeverything is done after rebooting
15:49:45Jaykaytorne: so for you the "searching..."-step is "missing"?
15:50:01Torneno.. it searches and then it commits
15:50:02 Join FOAD_ [0] (
15:50:24Jaykaybut both searching and commiting are done after reboot?
15:50:51Tornethey're done whenever i tell it to initialise the databse
15:50:57Torneas long as the db is enabled
15:51:04Torneotherwise it enables the db and tells you to reboot
15:52:29Jaykaytorne: you can disable the database!?
15:52:58pixelmaTorne: what happens when you initialise again (or update)? I think you either remember wrong, or there's a misunderstanding, or something changes with your settings.
15:53:09Tornei've never updated the db
15:53:14Torneor initialised it again :)
15:53:23Tornewithout having blasted the entire contents of the disk inbetween
15:54:06Tornemy files haven't ever changed :)
15:54:12CIA-6New commit by zagor (r23899): Renamed file
15:54:46 Join DerPapst [0] (
15:54:46Jaykaytorne: how does your "enabling" of the database work?
15:57:29pixelmaI'm curious too
15:58:21 Quit angelwolf71885 ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
16:01:07Tornecan't see anything in tagcache_build that implies the need to reboot..
16:01:33Tornegoes through the disk and writes stuff to the temp file, then commits to the real db
16:02:00***Saving seen data "./dancer.seen"
16:02:07pixelmathe committing step is usually done after a reboot
16:02:12Torneif that's where you are rebooting between then yes, it's writing to disk in both
16:02:27Torneit writes to a temp place first, then into the real db files
16:02:49Zagorthe manual build is horribly noisy. Mr Someone should look into trimming that down so real warnings become visible.
16:03:47 Quit teru ("Quit")
16:04:06pixelmaI believe someone said it wasn't possible, but maybe ask bluebrother - he knows more about TeX
16:04:12Jaykaytorne: does the code also say what hapens if the disk gets full while writing?
16:05:06Torne"nothing", by the look o fit
16:05:49Jaykaytorne: hwat does that mean?
16:06:50 Quit FOAD (Read error: 110 (Connection timed out))
16:06:50 Nick FOAD_ is now known as FOAD (
16:07:17Torneall the calls to write() in tagcache.c don't check the return code
16:07:21Torneso.. nothing happens.
16:07:25Tornei assume.
16:07:29Tornelooks like it just carries merrily on.
16:08:43Jaykayit does.... but extremely slow...
16:08:48Jaykayshould i report a bug?
16:09:50Tornenot sure why it shouldbe slow, but that depends on our fat implementation i guess
16:11:01Jaykayi'm wondering how it can carry on at all... if the disk is full, it's full...
16:11:01gevaertson a full disk, you're likely to suffer from fragmentation
16:11:11Torneif it's *actually* full then it shouldn't have a problem though, no?
16:11:22Tornewhen you try and write you notice the next free cluster is no cluster at all
16:11:25Torneshould go faster! :)
16:11:34TorneJaykay: it carries on because it doesn't seem like it's checking for failures
16:11:55Jaykaytorne: i mean how it can write even if the disk is full
16:12:00TorneIt doesn't
16:12:01Torneit fails
16:12:03Torneevery time
16:12:55UnhelpfulJaykay: it can *call* write all it pleases
16:13:12Unhelpfulyou can call it with a number you make up instead of a real file handle if you wish ;)
16:13:18Jaykayso when it's getting slow at 24 and is now at 29, 24-29 are not written at all?
16:13:27TorneNo, how could they be if the disk is full
16:13:49Jaykayand why does failing need so long? :D
16:15:05Tornethat is possibly an interesting question, yes ;)
16:15:30gevaertsmaybe the fat driver goes through the bitmap every time?
16:15:55Torneif it does then the fat driver is dumb
16:15:59Tornethat's easy to fix :)
16:16:00 Nick Omlet^away is now known as Omlet (i=omlet05@
16:16:01*Unhelpful was just about to say, perhaps finding that there are no free clusters is not very fast...
16:16:19Tornegenerally i'd expect a FAT implementation to maintain a pointer to the first free cluster
16:16:27Tornewhich means you know instantly if the disk is full
16:19:13Torneah, yes, we do that, but if we're appending to an existing file it ignores it and looks starting from last-cluster-of-file + 1
16:20:03 Join Tomis2 [0] (n=Tomis@
16:22:21Torne(and we don't quite maintain it the way i expect anyway)
16:22:33Torneanyway, yah, this looks like one of the many reasons why tagcache can go horribly wrong
16:22:49Torneno evidence it even tries to handle running out of disk :)
16:27:24 Quit Tomis (Read error: 110 (Connection timed out))
16:27:24 Nick Tomis2 is now known as Tomis (n=Tomis@
16:32:26 Join Thunder_Drop [0] (n=chatzill@
16:34:36Thunder_DropWould it be possible to compile a custom build of rockbox for the cowon d2 that would only use the buttons... I've got a d2 that went through the wash and lost touch screen capabilities. Everything else works. I can compile myself and change the code myself but do you think it is possible?
16:35:15gevaertsI don't think you'll get very far with three buttons
16:35:42gevaertsHow long ago did this happen? I'd let it dry for a bit longer (at least a few weeks) before giving up
16:36:07Thunder_Dropi could probably get playback if I used three buttons and used them as three more buttons with the hold switch on. Also long pushes could be configured to different buttons right?
16:36:54 Join linuxguy4 [0] (
16:37:15CIA-6New commit by zagor (r23900): Corrected filename spelling mistake
16:37:16Thunder_Droplong time ago... It didn't work at all for a while but then I came back to it after a while and it runs fine except for the touch screen
16:38:16Thunder_DropI've also had it apart but it seems that there are corroded connections on the back of the screen itself
16:38:54Thunder_DropBut I'm just wondering if I could at least get playback functioning with the physical buttons
16:39:10Thunder_DropI could write configure files with my other rockbox player
16:42:51ByanThunder_Drop: sounds like more trouble than it's worth
16:43:04Byanand you could just write the config files by ahdn
16:43:17Byanbut, sure, it's possible, why not
16:44:01Thunder_Dropwell I have an S9 that I have been using but I have to send it in for repairs and it will by out of commision for a few weeks. I don't know how long I can live with my sansa :)
16:44:27Thunder_DropWell I'll have a look at it latter
16:44:28Byanbut in the end.. you'd have taken hours of your time changing the code, only to get a boneified shuffle
16:44:42Thunder_Dropbut good sound :)
16:45:00Thunder_Dropplus its a challenge... I like challenges
16:45:53Byanyeah.. but a challenge that would be useful in other cases would be better =p
16:46:01amiconnTorne: The reboot for db commit is only necessary if neither dircache nor "load database to ram" are enabled
16:46:38Thunder_Dropbyan: too true... I don't see how I can fix my problem and be helpful for others though
16:46:41amiconnIf either option is enabled, the database uses the respective buffer for the commit step
16:47:06ByanThunder_Drop: I guess you could make it really easy for people to remap buttons
16:47:18Torneamiconn: aha, that's why i've not seen it then :)
16:47:36Thunder_Dropbyan like a configure page that remapped buttons :)
16:47:51Thunder_Dropbyan: maybe I'll look into that
16:47:53Byanit'd never get put into Trunk though
16:48:04Thunder_Droptoo confusing?
16:48:05ByanRockbox devs aren't gonna like that
16:48:24Thunder_Dropmeh could be a patch that ppl would use
16:48:33Byantrue enough
16:48:55 Quit linuxguy3 (Read error: 110 (Connection timed out))
16:48:59Thunder_Dropwell class calls so I'll take a look latter :)
16:49:01gevaertsThunder_Drop: if you want to have a go at that, talk to JdGordon. He says he knows how to do it
16:49:23Thunder_DropI'll have to make contact in a bit
16:50:31Byanshould I sleep.. or study more..
16:53:13ByanI think am gonna go take a power nap..
16:53:15 Join toffe82 [0] (n=chatzill@
16:54:05 Join funman [0] (n=fun@rockbox/developer/funman)
16:57:00 Quit MethoS- (Remote closed the connection)
17:01:31 Quit DerPapst ("Leaving.")
17:02:52 Quit Zagor ("Don't panic")
17:04:52 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-hsdqszkhmkrijapp)
17:05:18saratogaawesome i love getting PMs from users I've never heard of asking me random installation questions
17:07:22Tomishai, i'm trying to instal this lunix into my yellow dog? i think he's a golden reteiver? he won't accept the instal disk, keep throwing it back up, what should i do?
17:08:17archivatorTomis: try iPodLinux, rockbox has nothing to do with linux :)
17:10:28 Join DerPapst [0] (
17:11:45gevaertsTomis: lunix only runs on commodore 64
17:16:04 Join Tomis2 [0] (n=Tomis@
17:17:30archivatorWill Rockbox be applying for GSoC 2010? I'll be eligible for gsoc this year and would like to try and work for rockbox :)
17:19:35 Quit Tomis (Read error: 104 (Connection reset by peer))
17:19:35 Nick Tomis2 is now known as Tomis (n=Tomis@
17:23:24 Join grndslm [0] (
17:27:40 Join Tomis2 [0] (n=Tomis@
17:28:18 Quit Tomis (Read error: 104 (Connection reset by peer))
17:28:18 Nick Tomis2 is now known as Tomis (n=Tomis@
17:29:31 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
17:32:55 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
17:39:34Jaykayi have 4 tasks which could be closed (maybe)... what should i do now? :)
17:39:58Jaykaythe problem is that in the tracker nobody answers... and here nobody does, too
17:41:35*jae just experienced his first WSOD (white screen of death) while trying doom on his Fuze...
17:42:20gevaertsjae: a less wasteful malloc() might help there :)
17:44:04jaeGood thing I don't "need" Doom on that (or anywhere else, for that matter)... (though "not need" goes for that whole player thing too, albeit a bit weakened)
17:44:45 Quit dfkt (Nick collision from services.)
17:44:48 Nick dfkt_ is now known as dfkt (i=dfkt@unaffiliated/dfkt)
17:48:22 Quit PaulJam (Nick collision from services.)
17:48:30 Join PaulJam_ [0] (
17:51:09 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
17:51:41TorneJaykay: which?
17:52:26 Join Tomis2 [0] (n=Tomis@
17:52:46JaykayTorne: FS #8766, FS #8769, FS #8982, FS #9130
17:53:21 Quit PaulJam_ (Nick collision from services.)
17:53:31 Join PaulJam__ [0] (
17:54:16 Join evilnick [0] (i=0c140464@rockbox/staff/evilnick)
17:54:41 Join kadoban [0] (
17:55:01 Join thegeek_ [0] (
17:55:25 Quit FOAD (
17:55:25 Quit thegeek (
17:55:25 Quit kadoban__ (
17:55:25 Quit yosafbridge (
17:55:25 Quit linuxstb (
17:55:25 Quit B4gder (
17:56:13 Quit Zambezi (Read error: 104 (Connection reset by peer))
17:56:56 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
17:58:01NJoinFOAD [0] (
17:58:02 Join killan_ [0] (
17:58:15TorneJaykay: closed 8766 and 8982
17:58:17 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
17:58:31Torne9130 I don't think your testing really helps; you are not using the same device as the submitter
17:58:31 Quit Tomis (Read error: 145 (Connection timed out))
17:58:31 Nick Tomis2 is now known as Tomis (n=Tomis@
17:59:40Tornenot sure about killing bug reports where the submitter never came back :)
17:59:56Jaykaytorne: does the h300 have a hard disk?
18:00:10 Quit killan (Read error: 104 (Connection reset by peer))
18:00:20Jaykayand maybe the tracker should have submitting and deleting closing rules :)
18:00:41*Torne shrugs
18:00:48TorneI don't think having open issues is a huge problem :)
18:00:52Torneif you don't care about them stop watching them
18:00:52Jaykaytorne: do you have a target with a hard-drive and could you test that? ;)
18:01:40 Join Rob2222 [0] (
18:01:40 Join PaulJam_ [0] (
18:01:40NJointhegeek [0] (
18:01:40 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
18:01:40NJoinkadoban__ [0] (
18:01:40NJoinB4gder [241] (n=daniel@rockbox/developer/bagder)
18:01:40NJoinyosafbridge [0] (
18:01:40Jaykaywho said i don't care about them? as you see i'm trying to reproducce some bugs to help you (the devs)
18:02:01***Saving seen data "./dancer.seen"
18:02:24Torneyes, but i mean, that just because you tested it on a different device doesn't mean it's ok
18:02:50Torneif you can reproduce the submitter's issue with an old build on the same device but can't with a new build on the same device then that's pretty good evidence
18:03:01Tornebut otherwise, it's very hard to say whether an abandoned bug is still valid or not.
18:03:02 Join Byan_ [0] (
18:03:07Tornei am erring on the side of leaving them open
18:03:33Jaykayok.... i need to get up a testing environment so i can test old builds too...
18:03:54 Quit DerPapst ("Leaving.")
18:03:55Torneif you can't reproduce the problem, but it's something pretty specific, then it may be that you *do* need to be using the same device..
18:04:02Torneor it may be that the submitter's settings impact it
18:04:11Torneor it may be fixed :)
18:04:21Torneeliminating variables is the only way to really prove it
18:04:58 Quit domonoky (
18:04:58 Quit dfkt (
18:04:58 Quit z35 (
18:04:58 Quit archivator (
18:04:58 Quit Rob2223 (
18:04:58 Quit flydutch (
18:04:58 Quit maruk (
18:04:58 Quit pixelma (
18:04:58 Quit Kitr88 (
18:04:58 Quit blithe (
18:04:58 Quit tha (
18:04:58 Quit Byan (
18:04:59 Quit TaZzZ (
18:04:59 Quit loyx (
18:04:59 Quit thegeek (
18:04:59 Quit Rob2222 (
18:04:59 Quit kadoban__ (
18:04:59 Quit yosafbridge (
18:04:59 Quit PaulJam_ (
18:04:59 Quit B4gder (
18:04:59 Quit linuxstb (
18:05:02Tornesomeone else might have a different opinion on what to do with old abandoned bugs, though..
18:05:37NJoindomonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:05:37NJoindfkt [0] (i=dfkt@unaffiliated/dfkt)
18:05:37NJoinz35 [0] (
18:05:37NJoinarchivator [0] (n=archivat@
18:05:37NJoinRob2223 [0] (
18:05:37NJoinflydutch [0] (
18:05:37NJoinmaruk [0] (
18:05:37NJoinKitr88 [0] (
18:05:37NJoinblithe [0] (
18:05:37NJointha [0] (
18:05:37NJoinTaZzZ [0] (
18:05:37NJoinloyx [0] (
18:06:16 Quit archivator ("Leaving")
18:06:26 Join archivator [0] (n=archivat@
18:06:58 Quit loyx (Killed by (Nick collision))
18:07:01 Join domonoky1 [0] (
18:07:27 Join loyx [0] (
18:07:27 Join pixelma [0] (
18:07:27NJoinRob2222 [0] (
18:07:27NJoinPaulJam_ [0] (
18:07:27NJointhegeek [0] (
18:07:27NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
18:07:27NJoinkadoban__ [0] (
18:07:27NJoinB4gder [241] (n=daniel@rockbox/developer/bagder)
18:07:27NJoinyosafbridge [0] (
18:08:14 Quit Rob2222 (Read error: 104 (Connection reset by peer))
18:08:33 Quit kadoban__ (Connection timed out)
18:08:46 Quit dfkt (Nick collision from services.)
18:08:54 Nick dfkt_ is now known as dfkt (i=dfkt@unaffiliated/dfkt)
18:10:41 Join Rob2222 [0] (n=Miranda@
18:13:34 Quit thegeek (
18:13:34 Quit yosafbridge (
18:13:34 Quit PaulJam_ (
18:13:34 Quit pixelma (
18:13:34 Quit B4gder (
18:13:34 Quit linuxstb (
18:13:34 Quit loyx (
18:14:45 Quit Rob2222 (
18:14:45 Quit maruk (
18:14:45 Quit Rob2223 (
18:14:45 Quit blithe (
18:14:45 Quit tha (
18:14:45 Quit Kitr88 (
18:14:45 Quit TaZzZ (
18:14:45 Quit flydutch (
18:14:45 Quit domonoky (
18:14:45 Quit z35 (
18:14:45 Join blithe_ [0] (
18:15:15NJoindomonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:15:15NJoinz35 [0] (
18:15:15NJoinRob2223 [0] (
18:15:15NJoinflydutch [0] (
18:15:15NJoinmaruk [0] (
18:15:15NJoinKitr88 [0] (
18:15:15NJointha [0] (
18:15:15NJoinTaZzZ [0] (
18:15:33 Join faemir [0] (n=faemir@
18:16:40 Join Robotnik [0] (
18:17:28RobotnikDoes anyone know a super secret hideout here I could find the 1.1.1 firmware for an iPod nano 2g?
18:17:51 Join pixelma [0] (i=quassel@
18:18:30 Join phanboy4 [0] (n=benji@
18:18:31 Quit pixelma (Killed by (Nick collision))
18:18:35 Quit domonoky (Success)
18:18:39NJoinRob2222 [0] (n=Miranda@
18:18:39NJoinloyx [0] (
18:18:39NJoinpixelma [0] (
18:18:39NJointhegeek [0] (
18:18:39NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
18:18:39NJoinB4gder [241] (n=daniel@rockbox/developer/bagder)
18:18:39NJoinyosafbridge [0] (
18:18:49funmanRobotnik: it's not related to rockbox, so offtopic here, join #rockbox-community for offtopic talk
18:18:54 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
18:18:54 Quit pixelma (Nick collision from services.)
18:19:11 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
18:19:13 Join Zambezi [0] (i=Zulu@
18:20:16RobotnikActually, I installed Rockbox on my iPod and it seam that I cannot up-load song on it because my firmware is too new. : /
18:20:29RobotnikBut I will go ask there
18:22:20 Join Kitar|st [0] (
18:22:20funmanrockbox is still in development for ipod nano 2g, so i understand you can't install the rockbox bootloader due to your OF being to recent ?
18:22:46 Quit thegeek (
18:22:46 Quit Rob2222 (
18:22:46 Quit yosafbridge (
18:22:46 Quit B4gder (
18:22:46 Quit linuxstb (
18:22:46 Quit loyx (
18:23:09funmanFlynDice, kugel: do you think fuze/e200v2 can go stable with 3.5 release ?
18:23:20NJoinyosafbridge [0] (
18:23:20 Nick Omlet is now known as Omlet^lunchtime (i=omlet05@
18:23:45RobotnikI was able to install it manually.
18:23:58topikit doesn't need usb for 'stable', funman ?
18:24:21Robotnikfunman: I was able to instal it manually
18:24:34funmantopik: no, e200v1/c200v1 have been stable without rockbox usb support; USB can be handled by the OF
18:24:36Robotnik*sorry, double posting*
18:25:00funmanRobotnik: well if rockbox runs and doesn't work properly then it's on topic here, provided you give details about what happens
18:25:18topikexcept for those tiny screen glitches i have had no problems with rockbox on my fuze that weren't quickly fixed
18:25:26 Join Rob2222 [0] (
18:25:31RobotnikAlright. : )
18:25:47RobotnikSo the problem is that I can't up-load song onto it.
18:26:12topiki fill my nano2g in the ipod's disk mode
18:26:14RobotnikI think I saw on the forum that is was due to the firmware of my iPod nano being to recent
18:26:48funmankugel: ^ 10 lines above
18:27:14RobotnikI tried that, but when I try to access the E: manualy, it appears complety empty and I can't see any files
18:27:16kugelfunman: I wouldn't mind. did we decide whether unstable can get releases?
18:27:29RobotnikNot even the Rockbox files
18:27:40funmankugel: well i meant "move the e200v2/fuze from unstable to stable"
18:28:29 Join loyx [0] (
18:28:45topikRobotnik: yet you have rockbox working on it fine?
18:29:32 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
18:29:32 Join phanboy_iv [0] (
18:29:37kugelfunman: well, I would like to fix the sporadic lcd corruption on the fuze, but we still have 2 weeks, so
18:29:51Robotnik(I've just uninstalled Rockbox so I could try to downgrade my iPod)
18:30:17 Quit Rob2223 (No route to host)
18:30:20Robotnik(But it's no good without the 1.1.1 file I guess)
18:30:29 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
18:31:42TorneRobotnik: where did you hear that you needed to downgrade the OF?
18:31:48funmankugel: do we need to push bertrik for that ?
18:34:38 Join panni_ [0] (
18:34:41kugelwe could but he doesn't seem very interested. I would commit it anyway (he didn't seem to be against going by his responses)
18:35:46 Quit Rob2222 (
18:35:46 Quit yosafbridge (
18:35:46 Quit pixelma (
18:35:46 Quit phanboy4 (
18:35:46 Quit maruk (
18:35:46 Quit tha (
18:35:46 Quit Kitr88 (
18:35:46 Quit TaZzZ (
18:35:46 Quit flydutch (
18:35:46 Quit z35 (
18:35:48 Join JdGordon| [0] (
18:35:55 Join Kitr88 [0] (
18:35:59 Quit petur ("*plop*")
18:36:31JdGordon|kugel: the stop scroll there is for the sbs scrolling lines... and yes 0tick splashes will get cut
18:36:35 Quit Kitr88 (Killed by (Nick collision))
18:36:37NJoinRob2222 [0] (
18:36:37NJoinyosafbridge [0] (
18:36:37NJoinpixelma [0] (i=quassel@rockbox/staff/pixelma)
18:36:37NJoinphanboy4 [0] (n=benji@
18:36:37NJoinz35 [0] (
18:36:37NJoinflydutch [0] (
18:36:37NJoinmaruk [0] (
18:36:37NJoinKitr88 [0] (
18:36:37NJointha [0] (
18:36:37NJoinTaZzZ [0] (
18:37:01 Quit Rob2222 (Read error: 104 (Connection reset by peer))
18:37:33 Join Rob2222 [0] (
18:37:51topikis there a known way to trigger those fuze glitches?
18:38:06kugelJdGordon|: and that's better?
18:38:16JdGordon|than what?
18:38:19kugelmost of the splashes are 0ticks ones actually
18:38:23 Quit phanboy_iv (Read error: 104 (Connection reset by peer))
18:38:35 Quit Kitr88 (SendQ exceeded)
18:38:42 Join Kitr88 [0] (
18:39:04JdGordon|there is a dodgey way to fix it by not disabling the theme on 0 tick splashes
18:39:04JdGordon|but that wont work any better
18:39:34kugelsvn has no problems with splashes, you could just keep the event
18:40:21JdGordon|tha wont actually help at all.... 0 tick splash shows, then the sbs can be updated over it anyway
18:40:36RobotnikTorne: I guess I've misread something about downgrading an iPod touch and though it was of nanos : /
18:40:37kugelif there's a sbs
18:40:53JdGordon|which is a perfectly valid asumption
18:40:57kugelthe dead parts problem is caused by ui vp mainly
18:41:03TorneRobotnik: the version of the original firmware shouldn't affect rockbox once it's running
18:41:08kugelit is not
18:41:11Torneif you can install and run rockbox you are done with the OF
18:41:16JdGordon|of course it is
18:41:38 Quit Kitar|st (Success)
18:41:38kugelyou can use custom ui vp without sbs..
18:41:55TorneRobotnik: if you mean you have problems accessing the disk in rockbox, with it showing up as empty and 0 bytes.. that's a problem several people have. not sure what you can do about it other than use the OF disk mode..
18:42:08kugeland custom ui vp has the dead parts already
18:42:30kugelI don't understand why you remove something that works at least (even if not entirely clean) and introduce something that makes things worse
18:42:36JdGordon|the sbs case will be far more liekly that just deadspace
18:43:56RobotnikTorne: Oh ok, I will go check in the faq for the disk mode
18:44:00JdGordon|... to get a rise...
18:44:07RobotnikTorne: Thanks a lot!
18:44:11JdGordon|no, actualy I prefer than that code is clean than minor glitches
18:44:33JdGordon|and 0 tick splashes arnt so important anyway
18:44:49JdGordon|they are only there to show that something is happening... which could be done better
18:45:12 Quit maruk ("Leaving.")
18:46:27topikyou just said you tried disk mode, Robotnik, and now you are going to read a faq about it?
18:47:48 Quit phanboy4 (Success)
18:49:09Robotniktopik: yeah because I must have missed something
18:49:34topikthere's not much to it besides holding select-play while it starts
18:49:41 Join KBH [0] (
18:50:16 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
18:50:31 Join stoffel_ [0] (
18:50:53 Join yawny [0] (
18:51:26 Join hd [0] (
18:51:50 Join bzed_ [0] (
18:51:59 Join niekie_ [0] (i=quasselc@CAcert/Assurer/niekie)
18:52:11 Join Rob2223 [0] (
18:53:07 Quit Rob2222 (
18:53:07 Quit loyx (
18:53:07 Quit bmbl (
18:53:07 Quit stoffel (
18:53:07 Quit tarbo (
18:53:07 Quit chaos (
18:53:07 Quit lostlogic (
18:53:07 Quit elcan (
18:53:07 Quit mc2739 (
18:53:07 Quit crwll (
18:53:07 Quit HBK (
18:53:07 Quit tchan (
18:53:07 Quit bzed (
18:53:07 Quit advcomp2019 (
18:53:07 Quit HellDragon (
18:53:07 Quit liar (
18:53:07 Quit niekie (
18:53:09 Nick bzed_ is now known as bzed (
18:53:12 Nick KBH is now known as HBK (
18:53:51NJointarbo [0] (n=me@unaffiliated/tarbo)
18:57:10NJoinchaos [0] (n=chaos@gentoo/user/ch4os)
18:57:50 Join Tomis2 [0] (n=Tomis@
18:58:04 Quit Tomis (Read error: 104 (Connection reset by peer))
18:58:07 Nick Tomis2 is now known as Tomis (n=Tomis@
18:59:01 Join Sajber^ [0] (
19:01:12 Quit kugel (Read error: 110 (Connection timed out))
19:04:07topikRobotnik: don't know if it matters, but my nano 2g has firmware 1.1.3 and works fine with rockbox
19:05:25 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
19:07:26NJoinbmbl [0] (n=Miranda@unaffiliated/bmbl)
19:11:23Robotniktopik: after re-installing Rockbox I've understood that I mixed up 'multimedia mode' and 'disk mode'
19:11:36RobotnikNow everything works fine, ahaha
19:11:38 Join n17ikh [0] (
19:13:42 Join Tomis2 [0] (n=Tomis@
19:15:41RobotnikWell, thanks a lot for helping me guys! : )
19:16:50 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-zvfpyzegdeklmcrf)
19:19:30 Quit Tomis (Read error: 110 (Connection timed out))
19:19:31 Nick Tomis2 is now known as Tomis (n=Tomis@
19:24:21 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:24:37 Quit bmbl ("Bye!")
19:25:34gevaertsUnhelpful: h300, flac -8, r23879 : 620.23%. r23879+patch 631.82%, crc matches
19:25:36 Quit amiconn (Nick collision from services.)
19:25:37 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
19:25:57 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
19:27:03 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
19:28:08 Join loyx [0] (
19:28:08 Join liar [0] (n=liar@
19:28:08 Join crwll [0] (
19:28:08 Join mc2739 [0] (n=mc2739@rockbox/developer/mc2739)
19:28:08 Join lostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
19:28:09 Join pixelma_ [0] (
19:28:09 Quit pixelma (Nick collision from services.)
19:28:57 Quit Robotnik ("CGI:IRC (EOF)")
19:33:33Tuplanollashould rockbox play xm files?
19:33:34 Join bertrik [0] (
19:37:28 Join Hillshum [0] (
19:37:36 Quit Sajber^ ("Leaving.")
19:38:43 Quit archivator (Read error: 110 (Connection timed out))
19:39:50 Join Jaykay_ [0] (
19:42:07domonoky1Tuplanolla: rockbox plays many different chiptunes files, but i think xm is still not supported (i think there is a patch for it on the tracker)
19:43:10Tuplanollaah, ok
19:43:23domonoky1but you can play sid,mod,nsf,spc,gbs and .sap chiptunes files :-)
19:43:33 Quit stoffel_ (Remote closed the connection)
19:43:35Tuplanollayep, and that's awesome :p
19:46:24 Join archivator [0] (n=archivat@
19:55:19 Join captain-kewl [0] (i=2669ecc2@gateway/web/freenode/x-kpwcamczhwgqaxmt)
19:57:44 Quit Omlet^lunchtime ("( :: NoNameScript 4.22 :: )")
19:59:16 Quit Jaykay (Read error: 110 (Connection timed out))
20:00:17 Quit Jaykay_ ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
20:02:02***Saving seen data "./dancer.seen"
20:10:36 Join DerPapst [0] (n=DerPapst@
20:13:39 Quit loyx (Read error: 110 (Connection timed out))
20:21:41 Join kugel [0] (n=kugel@rockbox/developer/kugel)
20:22:25 Quit archivator ("Leaving")
20:23:57 Join Casainho [0] (
20:27:34 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
20:30:46 Join archivator [0] (n=archivat@
20:31:45kugelJdGordon|: mister band aid prefers glitchy over not entirely clean now? And IMO breaking 0-tick splashes isn't very nice, they are not unimportant (I think the majority of splashes are 0-tick ones)
20:32:38JdGordon|adding band-aids when you change the API is a bloody stupid thing to to
20:32:52kugeli didn't change the api
20:33:14JdGordon|did I say you did?
20:33:42 Quit Casainho ("ChatZilla 0.9.85 [Firefox 3.5.5/20091109125225]")
20:34:40*kugel blames the English "you" ambiguity
20:37:03kugeldo you want to commit it for 3.5? I'm not sure if that's a good idea if it introduces other problems. the problems in SVN are relatively easy to fix
20:38:13JdGordon|yes I do
20:38:30JdGordon|I actually tihnk the best thing for 0 tick splashes is to get rid of them!
20:38:49JdGordon|there is no clean way to do them
20:39:08JdGordon|places like the wps "loading" splash for example are pointless
20:39:25JdGordon|it could just as easily be done with a "busy?" token and a sbs
20:40:11JdGordon|but anyway... if those splashes are the worst issue left then I'm completly good to commit it
20:40:34JdGordon|the code is much cleaner than svn, and the glitches are no worse than svn
20:43:00 Join solexx_ [0] (
20:43:05amiconnThe "loading" splash isn't pointless
20:43:07JdGordon|adding a token for busy indication could actually be pretty cool... the sbs could show a silly animation in the middle of the screen while its doing that
20:43:31amiconnIt gives valuable feedback
20:43:31JdGordon|amiconn: yes, its not pointless.. but it doesnt have to be displayed like it is
20:43:47gevaertsJdGordon|: not really. That would require it to draw outside of the viewport it promised to stay in :)
20:43:54amiconnAnd zero tick splashes are useful, and should be doable
20:44:22JdGordon|gevaerts: well.. it doesnt have to be in the ui area :)
20:44:46amiconnIf you make updates synchronous as I suggested yesterday (?), the main window knows when to do a full update
20:45:05JdGordon|kugel: anyway, I tinhk sbs overdrawing the splash is a far bigger problem than deadspace because its alot more liekly that the splash doesnt even hit deadspace (narrow margins)
20:45:09amiconnThen it just needs to do a full update the next time it updates after a zero tick splash was shown
20:45:51JdGordon|how do you do that wthout adding a layer between the main thread and the lcd driver
20:46:09kugelthe 0-tick splashes are supper annoying
20:46:20JdGordon|but they taste yummy :p
20:46:49amiconnThere is another reason why these splashes need to stay: They're voiced
20:46:57 Part watto
20:47:16amiconnJdGordon|: Well, the main thread knows when it draws a zero tick splash
20:47:20JdGordon|people using voice wont care about artefacts :)
20:47:48kugelI've had an idea how to deal a tad bit better; let the viewportmanager redraw them after all other stuff is drawn. that would require to change the splash api so that the non-blocking ones also have a max timeout parameter
20:47:51amiconnYou suggested an animation (which would be *really* silly imo, and most proably wouldn't work properly anyway)
20:48:19 Join Thundercloud [0] (
20:48:23JdGordon|amiconn: yes, but it would be painful to have to go and modify every single screen to handle the descision
20:48:36amiconnAnd you're wrong that people using voice don't care about artifacts
20:48:42JdGordon|and I suggested a skin token... how its used is up to themers
20:48:59 Part captain-kewl
20:49:04 Join matsl [0] (
20:49:24JdGordon|kugel: thats not very nice
20:49:25amiconnAn animation won't work because the main thread is usually blocked in cases where a zero tick splash is shown
20:49:53kugel(that was for JdGordon| )
20:49:56 Nick fxb__ is now known as fxb (
20:50:23JdGordon|the splash api should be simple... I really tihnk working around these issues is a waste of time
20:50:58JdGordon|the splashes already look crap for 0 tick ones that change size
20:51:24amiconnJdGordon|: You will need that for the startup case anyway, if you want to get rid of atrifacts in a simple way
20:51:33kugelthe calling code wouldn't notice that the splash is redrawn
20:51:58JdGordon|I fixed the startup splash problem by reordering the settings_apply() calls
20:52:24amiconnThat sounds like a fragile solution
20:52:56JdGordon|it wont be fragile once the the theme loading is explicit
20:53:07JdGordon|which is on the todo list
20:53:18kugelthe splash api wouldn't change really, just having the timeout always apply and a do_block parameter. the redraw would me handled internally
20:53:38kugeland I'm sure no splash would leave artifacts ever with that
20:53:56JdGordon|0tick splashes are used because the main thread is busy.. we dont want to add to that with extra draws
20:54:05FlynDicefunman: re fuze/e200v2, I've got to confess I'm not up to speed on the state of the manual or the installer, but I would say that from a user perspective, Rockbox certainly runs well on these players with just "minor" deviations from perfect. Until I break it again that is.. ;-)
20:55:10JdGordon|a fix which I dont want to do for splashes is tell _theme_undo() to not clear the deadspace
20:55:16JdGordon|which would fix everything
20:55:20funmanFlynDice: manual & installer are alright, so let's go for it
20:55:30topikFlynDice: that 15.5Mhz patch seems fine
20:55:34kugelmy idea would make that possible easily
20:55:53 Join Casainho [0] (
20:56:09 Part LinusN
20:56:10funmanif i count correctly, we should start freeze in 2 days (or at the end of the week ?)
20:56:20JdGordon|but that means that the main thread DOES have to remember to do a full screen clear/update.. which sucks
20:56:55amiconnImo it's the only reliable way to handle this
20:57:07JdGordon|IMO its not worth the effort
20:57:17FlynDicetopik: Thanks for testing, I'm going to commit shortly
20:57:17amiconnThe splash artifacts also apply to timed splashes if a theme uses space wasters (aka margins)
20:57:30JdGordon|how about this.. we wait and see how much negative feedback there is.. if there is too much we can decide how to fix it
20:57:46 Quit parafin ("reboot")
20:57:52kugelwell, we could simply declare that 0-tick splashes aren't guaranteed to show up for longer than X seconds and have the splash code remember the last viewport. after X seconds the viewport would be cleared via the GUI_EVENT_ACTIONUPDATE event
20:57:54 Join n1s [0] (n=n1s@rockbox/developer/n1s)
20:58:29JdGordon|kugel: that sounds really complicated... and error prone
20:58:41amiconn...and undesirable
20:59:14JdGordon|and dangerous actually
20:59:51JdGordon|the clearing is done in theme_undo()... which must always be called with theme_enable() before hand... what you are proposing could very esaisly stkov the theme stack
20:59:56 Quit solexx (Read error: 110 (Connection timed out))
21:00:41JdGordon|splash()... splash()... splash()... splash()... ...time elapses... stkov because _undo() hasnt been called
21:01:02JdGordon|or.. we force splash to use the ui area
21:01:16 Quit domonoky1 (Read error: 110 (Connection timed out))
21:02:08kugelwhat has your strange theme stack to do with it?
21:02:23JdGordon|splash disables the theme
21:02:35kugelthere's no viewport_set_fullscreen anymore?
21:02:52 Join parafin [0] (
21:02:52JdGordon|there is... but thats very undesirable
21:03:03kugelsplashes don't need that theme_enable/theme_undo thing imo
21:03:10kugelwhat for?
21:03:16JdGordon|although... I had totally forgotten about that.. and using it would solve all this
21:03:20JdGordon|except the deadspace
21:03:54 Join Tomis2 [0] (n=Tomis@
21:05:49kugelmy idea is super simple, not complicated and error prone
21:06:06gevaertskugel: the last bit is the problem :)
21:06:19gevaertsunless you want the "not" to also apply there
21:06:22*gevaerts hides
21:07:40JdGordon|kugel: it still sounds like a bad idea... but go ahead and implement it against 10824 and we'll have a look
21:08:00*JdGordon| still thinks the splash issues arent going to actually be a problem
21:09:18 Quit Lss (Read error: 60 (Operation timed out))
21:12:54kugelwell, you're in fact breaking 0-tick splashes for everybody who's not using a fullscreen ui vp
21:14:44JdGordon|its not broken
21:15:45JdGordon|if the deadspace is small the user wont notice.. and if they arnt usign a sbs the deadspace *will* be relativly small
21:16:23kugelthey're cut off!
21:16:46JdGordon|and thats a hugly big deal why?
21:16:51kugelthat means they are broken
21:17:02JdGordon|no, they look shit...
21:17:22JdGordon|and they look shit in svn, for years if the size of the splash changes
21:18:17kugelthey look bad yes, but if the text is cut off they are also broken
21:18:49 Quit Tomis (Read error: 110 (Connection timed out))
21:18:49 Nick Tomis2 is now known as Tomis (n=Tomis@
21:18:55JdGordon|for 0tick splashes.. the text isnt really that important
21:19:06JdGordon|they are an indication that rockbox hasnt crashed... nothing more
21:19:46JdGordon|and as long as you dont have crazy stupid deadspace it wont cut off all the text anyway
21:19:51 Join stripwax [0] (n=Miranda@
21:21:44 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
21:24:28 Quit stripwax (Client Quit)
21:24:45 Join stripwax [0] (
21:26:34CIA-6New commit by FlynDice (r23901): Sansa AMS: Run all SD cards within SD Specification frequencies. ...
21:27:04domonokyrasher: ping ?
21:32:16*amiconn really doesn't understand JdGordon|'s weird logic
21:32:59JdGordon|if the informatino in the splash was important, then it would be guarenteed to stay on the screen long enough to read
21:33:03JdGordon|its not, so its not
21:33:56JdGordon|+ ''''''
21:34:01amiconnOn one hand you want to improve looks by extending theming, on the other hand you're knowingly breaking the look of things which are working now
21:34:13domonokyJdGordon|: sure a "busy" sign would probably replace all those 0time splashes, but untill you replace them all, its bad to commit code wich brakes it.
21:34:28amiconndomonoky: Don't forget voice
21:34:50JdGordon|voice can be fixed there easily
21:34:58domonokytrue, so we need a voiceable busy sign :-)
21:35:36JdGordon|give me an example of a splash that is absolutly required and its information is important and I'll back down...
21:35:44JdGordon|that has the 0tick splash
21:36:14amiconnThe splash glitches do not just apply to 0-tick splashes
21:36:41*kugel is also under the impression that everything worked well after the initial sbs commit, and that every change since then made things worse
21:36:43JdGordon|no they dont, but the cutting off wont happen untill after the tmeout on them
21:37:10kugelespecially r23606
21:37:23amiconnI'm not referring to cutoff, but to parts remaining in dead areas
21:37:40amiconnkugel: Even initial sbs broke things iirc
21:37:48kugelamiconn: non-0-tick ones could be easily handled within splash itself; even in the same function
21:38:51JdGordon|amiconn: no, parts remaining on the screen is entrily fixed.. and you would see if you tested the patch
21:39:12JdGordon|r23606 was required.. but yes, maybe was too soon
21:39:18JdGordon|those functions were just bad
21:39:52JdGordon|which is besides the point
21:40:21 Join FOAD_ [0] (
21:42:28JdGordon|the only 0tick splash I can think of that has any usefull information is the updating database one.. which should be replaced with something that isnt the splash
21:43:43stripwaxwhat about the pictureflow 'failed to load bmp' one?
21:44:24JdGordon|thats 0 tick?
21:44:41kugelmost are
21:45:15JdGordon|they shouldn't be if the text is importany
21:45:29JdGordon|there are a few solutions for this.. all of which should be delayed
21:45:47JdGordon|a possible one is skinning the splash screen! at least enough to just give it a viewport
21:46:11kugelsure those solutions are delayed, but your patch needs to go in very fast?
21:46:30kugeldo we freeze tomorrow, btw?
21:46:44JdGordon|am I really the only one who thinks the splash's arnt that important?
21:46:50JdGordon|oh fuck... do we?
21:47:30 Join froggyman [0] (
21:48:16funmanbertrik: did you have a chance to test mkamsboot change on c200v2 ?
21:48:50 Join thanzero [0] (
21:56:55 Quit FOAD (Read error: 110 (Connection timed out))
21:56:55 Nick FOAD_ is now known as FOAD (
21:58:08 Join kkurbjunW [0] (n=karlk@
21:58:37kkurbjunWkugel: svn has the same problem with splashes that jdgordon's patch used to have
21:59:51CIA-6New commit by tomers (r23902): Manual: Hide USB HID actions which are not supported on target ...
22:00:51JdGordon|used to... but not now?
22:01:00 Quit archivator ("Leaving")
22:02:04***Saving seen data "./dancer.seen"
22:02:08kkurbjunWjdgordon, I haven't tried the latest patch, I am assuming they are working properly now.. an alternate way to fix the splashes would be to center and draw them in the ui viewport
22:02:30kkurbjunWrather than always drawing in the center of the screen
22:03:04JdGordon|thats another possible fix
22:03:29kkurbjunWI think that would be a really clean way to fix them
22:04:11JdGordon|I actually like splashes being outside the user area... it makes them feel more "pop-up-y"
22:04:50kkurbjunWalso, I had a concern about the settings apply changes thta you did, I havn't tested it on target, but are you sure those changes won't effect things like the database and dir cache?
22:06:02 Join Strife89 [0] (n=michael@
22:06:15kkurbjunWJdGordon, yeah - I mean that's a matter of oppinion, I don't care either way, but it would be nice if 0 second splashes worked properly
22:06:45JdGordon|that was my worry also (settings change), but for now at least that should be completly sa.... oh hmm... shit.. yeah maybe not
22:07:46JdGordon|what does dircaceh and databse do when they are re-initialised to be enabled if they were already?
22:08:38kkurbjunWI really don't know - I don't know if anyone that is here regularly knows what the consequence of that would be offhand.
22:08:47 Quit toffe82 (Read error: 131 (Connection reset by peer))
22:12:59 Join Llorphone [0] (n=pocketir@
22:14:00 Quit kkurbjunW (Remote closed the connection)
22:14:45 Quit funman ("free(random());")
22:14:58 Join kkurbjunW [0] (n=karlk@
22:17:44 Join Kitar|st [0] (
22:19:50 Join mrtok1 [0] (
22:21:10JdGordon|so does anyone want to pipe in about the release time frame?
22:21:16JdGordon|are we in fact freezing tomroow?
22:22:00mrtok1saratoga: sorry for bothering you again - could you assist in setting up test_codec.c for my testing? what ive done is, adding test_codec.c to SOURCES and checked the CATEGORIES files - but i cannot find the test_codec within my plugin menu - target is ipod nano 1g
22:22:10 Quit Kitr88 (Read error: 110 (Connection timed out))
22:23:03 Quit kkurbjunW (Remote closed the connection)
22:24:02mrtok1saratoga: + please ;-)
22:24:31mrtok1or anyone else?
22:24:50 Join polobricolo [0] (
22:24:53stripwaxmrtok1 - it's a viewer, so you long-select the file, and Open With
22:25:37stripwax^the file you want to decode
22:27:16 Join fyrestorm [0] (
22:27:16 Quit HBK (Read error: 104 (Connection reset by peer))
22:27:16mrtok1hi stripwax - i saw it is a viewer, but i did not know how to deal with that information... i try
22:27:16stripwaxmrtok1- can you confirm it did actually build test_codec.rock ? :)
22:27:16stripwaxyep, that's how you launch viewers (e.g. you can launch Sliding Puzzle by long-press on a .bmp file)
22:27:16mrtok1yes - and this is also installed under rocks
22:27:16gevaertsJdGordon|: that's the original plan, yes. How much time do you need?
22:27:16 Join HBK [0] (
22:27:38 Quit Bagder (Read error: 104 (Connection reset by peer))
22:27:45 Join Llorean1 [0] (
22:27:51 Join Bagder [0] (
22:28:18 Quit Bagder (Remote closed the connection)
22:28:26 Quit Llorphone (Remote closed the connection)
22:28:30 Nick B4gder is now known as Bagder (n=daniel@rockbox/developer/bagder)
22:28:35JdGordon|its not really "need".... I'd like to commit this tonight...
22:28:57gevaertsoh, so the rest would be bugfixes :)
22:29:11JdGordon|well, yeah
22:29:21 Join toffe82 [0] (n=chatzill@
22:29:43mrtok1stripwax: currently running the test .... :-) thank you !
22:29:54stripwaxno problem!
22:30:37stripwaxwhat device do you have, by the way? (I don't know if you read last night's irc logs but I'm also having profiling issues - but the fact that you were getting an empty profile.out yesterday actually means you're doing less badly than I am!)
22:31:09mrtok1ipod nano 1g
22:31:46mrtok1so you didnt even get an empty one?
22:32:28stripwaxmrtok1 - nope, got Illegal instruction exceptions right at the point where profiling would start. ipod 5g so hardly at all different to nano 1g. I'll try a full rebuild..
22:35:52kugelkkurbjun: what did you mean by "used to"?
22:36:12kugelor rather, which specific problem were you refering to
22:39:51kugelFlynDice: I'm not sure if adding a delay into rather speed critical loops to fix a non-existent problem is a good idea
22:41:31kugelthe place there seems weird to me, did you investigate that further? i.e. if the problem is in the dma code?
22:42:25 Quit Kitar|st (
22:42:25 Quit lostlogic (
22:42:25 Quit mc2739 (
22:42:25 Quit crwll (
22:42:25 Quit liar (
22:42:25 Join Stephen_ [0] (
22:42:42*amiconn thinks that a release is almost pointless right now, with all those broken things
22:42:42kugelthat's why don't release *now* :)
22:43:19 Quit Llorean (Read error: 110 (Connection timed out))
22:43:19amiconn3.5 that is... 3.4.1 is overdue
22:43:22NJoinKitar|st [0] (
22:43:22NJoinliar [0] (n=liar@
22:43:22NJoincrwll [0] (
22:43:22NJoinmc2739 [0] (n=mc2739@rockbox/developer/mc2739)
22:43:22NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
22:43:27JdGordon|release numbers are arbitrary
22:43:28 Quit fyre^OS (Read error: 110 (Connection timed out))
22:44:19*kugel hasn't seen a serious attempt to get a 3.4.1 out
22:44:21mrtok1stripwax: i´ve done 2 tests - one with algo under test enabled and one disabled - no i got 2 speed results - when disabled i need 60MHz less for realtime ;-)
22:44:55kugeland I also haven't seen a report of the problem which that 3.4.1 should fix, except amiconn's one ;)
22:45:10stripwaxwhat was the reported clockspeed required with algo disabled?
22:45:31stripwax(that sounds like some slow algo - audio decode likely around 30MHz)
22:45:42stripwaxwhich codec?
22:45:57stripwaxah - heh - uncompressed
22:46:24stripwaxwhat does your nano 1g report for say vorbis or mp3
22:46:37pixelmakugel: then you didn't look hard enough
22:47:09stripwaxbut anyway, yeah, that's a very slow algo - assuming nano 1g hardware basically equivalent to 5g, your dsp needs the same amount of juice as TWO vorbis decodes..
22:47:26pixelmaI saw reports about early USB problems, e.g. in the forums
22:47:35 Join polobric1lo [0] (
22:48:02mrtok1how to tranlate that into delay and mips?
22:48:19kugelyea, I don't look through the forums intensively
22:48:25mrtok1just running one with mp3 disabled
22:48:34kugelwhy are you so keen on mips?
22:49:33stripwaxor delay - not possible to tell from this what the latence will be thru the dsp algo, but basically user won't see it if it's realtime decoding at 60MHz
22:49:36mrtok1stripwax: testresult->25MHZ for realtime
22:49:37*JdGordon| yells and pulls his hair out over early usb
22:49:51stripwaxmrtok1 yeah. so 60MHz sounds BAD for any kind of dsp
22:49:54stripwaxwhat effect is this?!
22:50:30mrtok1some kind of psychoacoustics - yes not very simple
22:50:34 Join MethoS- [0] (n=clemens@
22:50:55stripwaxsome kind −− anything specific? (you're among friends here...)
22:51:51mrtok1doing amplitude and frequency dependent delaying signals
22:52:03mrtok1difficult to explain
22:52:23stripwaxgot any links to docs on the web? pretty smart people hang out on #rockbox ..
22:52:33stripwaxbut basically fft/ifft processing?
22:52:43 Quit Casainho (Read error: 113 (No route to host))
22:52:43stripwaxso guessing you need a decent fast fft for that
22:52:50mrtok1not of that kind -
22:53:01stripwaxit's slower than fft+ifft ....
22:53:35mrtok1it´s not a simple just having these both in place
22:54:31stripwaxno, well, presumably something happens in between
22:54:35JdGordon|can someone get a list of ui viewport/SBS type bugs in svn together so we can decide to go ahead with 10824 pre release or not?
22:54:36mrtok1it´s a bunch of biquads and some glue
22:54:42stripwaxbut probably easier for you to describe than for me to guess .. :)
22:54:55 Join Thundercloud_ [0] (
22:55:06 Quit n1s ("Lämnar")
22:55:21JdGordon|pixelma, kkurbjun, kugel, amiconn, others... ^^
22:55:43 Quit Thundercloud (Read error: 101 (Network is unreachable))
22:55:54stripwaxmrtok1 - hrm, I'm not sure I understand how/why that could be so slow
22:56:02stripwaxgot source/patch/etc ?
22:56:10saratogaimplementing them as a transform is unlikely to be done in less then 30-35MHz for just the transform stage, so a filterbank is probably better unless you've got a very large number of channels
22:56:49 Quit polobricolo (Read error: 110 (Connection timed out))
22:57:42saratogaprobably just have to think more carefully about the math and find a better way to code it
22:58:00 Quit toffe82 (Read error: 104 (Connection reset by peer))
22:58:55 Quit bertrik ("De groeten")
22:58:56mrtok1saratoga: there is a state invariant part too, which steers a feedback
22:59:09FlynDicekugel: re delay, What I found was there needs to be a delay between the time the card gets the SD_WRITE_MULTIPLE_BLOCK command and when the actual transfer starts when MCICLK is in the 15 MHz range. If we don't delay we get data crc failures.
22:59:34kugeli saw that
23:00:33FlynDicekugel: It fits in where I put it because there's no need to add another if(write) condition if we don't have to.
23:00:58 Quit dfkt (Read error: 110 (Connection timed out))
23:01:28kugeldid you do speed tests this time?
23:01:47FlynDicekugel: It's not really a non-existant problem either, the cards are overclocked
23:02:12kugelthat hasn't been a problem for 62MHz and not for 31MHz
23:03:42 Quit mrtok1 ()
23:04:22FlynDicekugel: No, it isn't a problem at the higher frequencies, but to get a non-hs card down into a spec frequency with PCLK at 62 MHZ your legal options are 15.5 MHz, 15.5 MHz, or even 15.5 MHz ... ;-)
23:04:55kugelI realized that overclocking means not within specs
23:05:44 Join polobricolo [0] (
23:05:47 Join Armand` [0] (
23:08:50 Quit thanzero (Read error: 110 (Connection timed out))
23:08:57Armand`q: i need to replace the hard disk in my ipod 5th gen, originally the 80GB model. If I'm only going to run rockbox on it, can I use any 1.8-inch hard disk with the ZIF connection or are there limitations?
23:09:53stripwaxArmand` - there is a page in the wiki
23:10:03stripwaxsearch for HardDrive
23:10:10gevaertsArmand`: the main limitation is that the official rockbox build doesn't support more than 128GB, although if you're comfortable with building your own it's easy to get a build that does work with those
23:15:37FlynDicekugel: For what it's worth, what I found out while figuring this out has helped convince me that IDE_CLK is used as MCLK for internal SD. We can run IDE_CLK at 50 MHz, div by 2 for 25.0 on the internals, and make the delay conditional for only standard speed uSD cards.
23:15:55 Quit polobric1lo (Read error: 110 (Connection timed out))
23:16:23FlynDicekugel: I wanted to break those 2 parts up though.
23:17:27FlynDicekugel: 25 MHz does nt require a write delay.
23:19:23 Join efyx_ [0] (
23:24:52 Quit Strife89 ("See ya.")
23:26:56 Quit bmbl ("Bye!")
23:28:10stripwaxsomething seriously unhappy with profiling codecs thru test_codec plugin. just got "Prefetch abort error at C0EDBABE." That's a new one. profiling within an arbitrary plugin (e.g. test_fps) seems to work just fine, on the other hand.
23:32:43 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
23:36:16kugelFlynDice: did you do speed tests?
23:38:14 Quit domonoky (Read error: 104 (Connection reset by peer))
23:39:18 Quit jgarvey ("Leaving")
23:39:34 Quit efyx_ (
23:39:34 Quit Thundercloud_ (
23:39:34 Quit lostlogic (
23:39:34 Quit mc2739 (
23:39:34 Quit crwll (
23:39:34 Quit Kitar|st (
23:39:34 Quit liar (
23:43:05NJoinefyx_ [0] (
23:43:05NJoinThundercloud_ [0] (
23:43:05NJoinKitar|st [0] (
23:43:05NJoinliar [0] (n=liar@
23:43:05NJoincrwll [0] (
23:43:05NJoinmc2739 [0] (n=mc2739@rockbox/developer/mc2739)
23:43:05NJoinlostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
23:43:20Unhelpfulgevaerts: thanks... i wasn't expecting any large gain, but i wanted to be sure it didn't hurt.
23:44:08Unhelpfulstripwax: "C0EDBABE"? nice. what about just playing one track with the codec? you'd still get a profile of where the codec spends time, right?
23:47:56stripwaxUnhelpful - no, because it aborts and I have to reset ..
23:48:10stripwaxI don't get any profile.out - it doesn't even get that far
23:48:37Unhelpfuleep. i tried on sim, just to get a rough guess at where costs are, but gprof apparently can't see through dlopen :/
23:49:01Unhelpfuli guess i could get info of that low quality with a profile build of faad too, though :)
23:51:46Bagdercallgrind should work on the sim
23:51:55Bagder"Callgrind uses runtime instrumentation via the Valgrind framework for its cache simulation and call-graph generation. This way, even shared libraries and dynamically opened plugins can be profiled"
23:53:06 Join GeekShado_ [0] (
23:54:05 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
23:55:45 Quit evilnick ("Page closed")
23:55:47 Join linuxguy3 [0] (
23:56:05stripwaxI made some hacks to build a rockbox Profile Sim build, which seems to work, but I'm not sure about the granularity of the timer used
23:57:34Bagdervalgrind/callgrind requires no special build, which is very attractive
23:58:28 Nick Ypsy is now known as YPSY (

Previous day | Next day