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

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

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

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2009-12-14

00:01:21dedchkdsk on Windows didn't find anything. I just tried a different player and it shows the same behaviour. (ipod 5.5 this one running 3.4)
00:01:26 Quit thegeek_ (Read error: 104 (Connection reset by peer))
00:01:37 Join thegeek [0] (n=nnscript@s168c.studby.ntnu.no)
00:04:43***Saving seen data "./dancer.seen"
00:06:04 Quit efyx_ (Read error: 113 (No route to host))
00:06:17 Quit n1s ("Lämnar")
00:06:40kugel_ded: with 3.4 you can only control multimedia. a current build gives you more possiblities
00:08:35 Join efyx_ [0] (n=efyx@lap34-1-82-225-185-146.fbx.proxad.net)
00:11:34 Quit Casainho ("ChatZilla 0.9.85 [Firefox 3.5.3/20091020122601]")
00:11:36dedSo on both of my machines resume playback resumes the penultimate playlist, whereas I'm expecting it to play the last playlist played. But this seems much too obvious to be a bug? I don't understand. Is there something wrong with my experiment?
00:12:53ded(experiment described above)
00:13:54 Quit phanboy4 (Read error: 104 (Connection reset by peer))
00:14:15 Join phanboy4 [0] (n=benji@c-24-98-43-198.hsd1.ga.comcast.net)
00:17:26mc2739ded: that sounds like a bug that should be added to the Flyspray bug tracker.
00:27:25dedOK. Does resume playback work by reading /.rockbox/most-recent.bmark?
00:29:49mc2739no, resume does not use .rockbox/most-recent.bmark
00:41:27kugel_resume doesn't involve bookmarks
00:47:26JdGordonkugel_: is the sync_data struct in the gui_Wps really needed?
00:48:00 Quit bmbl ("Woah!")
00:48:58kugel_not really but you need that member
00:49:04kugel_so it doesn't matter in the end
00:49:11JdGordon?
00:50:10kugel_for managing full updates
00:53:04JdGordonok
00:53:18JdGordonarg... is there a better prefix for the radio screen tokens than 't'?
00:53:31JdGordont is already used for subline timeouts though
00:53:42kugel_f?
00:54:13JdGordonnot so nice :(
00:54:20JdGordonwhy the heck is the database using r?
00:56:28 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
01:00
01:00:23 Quit petur ("Zzzzz")
01:02:40dedmc2739: I am concerned that my "bug" is much too severe for it to be likely to be a bug in 3.4. Does anyone have time to try to replicate it? (set idle poweroff to 1min ; play track from album A ; play track from album B ; pause ; wait for idlepoweroff ; reboot ; resume playback ; what is playing?)
01:03:58JdGordonany ideas how to make the presets list avilable to the skin? I'm thinking it would be very nice to be able to show the freq/name of any of the loaded presets with a single token... something like %pn (preset name) would show the current one, %pn|1 would show the next ones name, %pm|-1 would show the previous, both wrapping so the token would always be legal
01:07:56JdGordonor try to copy the %d token? (i.e just get rid of the | and use 0 to specify "this preset"... eg %pn0 %pn1 %pn-1
01:11:52Lloreanded: Does it replicate if you manually power off, or does it only happen with idle power off?
01:12:37mc2739ded: I cannot reproduce your resume problem. This is on an e260v2, not version 3.4, but a recent build (r23933)
01:13:59JdGordonhow does this look for the initial set of radio tokens? http://pastebin.com/m43f104ef
01:14:48JdGordon%pb (progressbar) will be fixed to work as the frequency bar in the fm screen
01:15:29kugel_maybe %i* should just act as fm info in the fm screen
01:16:11JdGordonthere isnt really a logical mapping for them
01:16:34JdGordonand I could imagine that would be very confusing anyway
01:24:48dedLlorean: good question, I'll try to properly pin down the conditions under which I'm seeing this behaviour.
01:27:49Lloreanded: If it's only with idle poweroff, could you add one more step into the test? After you pause, change one setting (set the volume higher, set repeat mode, do something, but don't browse or cause the disk to spin) and then let it idle poweroff. On next boot is the setting at the value it was changed to, or the original value? Basically I'm curious if somehow idle poweroff isn't writing everything necessary to disk. *do* be testing with the lates
01:30:59 Join perfectdrug [0] (n=5b0ef41e@giant.haxx.se)
01:32:42perfectdrugSVG for the Onda VX747 is ready FS #10852 :) maybe pixelma can finish the necessary steps to include it as thumbnail to the themesite
01:39:05JdGordonis there any point adding the frequency step size to the skin tokens?
01:39:10JdGordonfor the radio
01:40:29LloreanI don't think it's very important.
01:49:35kugel_isnt that basically the same as the region which is available via St?
01:49:39kugel_or should be, that is
01:50:25JdGordonthe region number is.. the the freq's arnt
01:50:26 Quit GeekShadow ("The cake is a lie !")
01:50:40JdGordonand even if they were... adding doubles for certain tokens is not such a bad thing...
01:51:06JdGordonis this math correct to make sure preset is always between 0 and radio_preset_count()-1? "preset = (preset + radio_preset_count()) % radio_current_preset();"
01:51:54JdGordonas long as someone doesnt do something crazy like -10 thats should be safe?
01:55:00 Quit DerPapst ("Leaving.")
01:55:42 Quit perfectdrug ("CGI:IRC")
01:56:54*JdGordon is just about ready to start hacking at the screen to use the skin! tokens are all done
01:57:17JdGordonapart from the parser for the preset name/freq :/
01:58:43saratogahttp://forums.rockbox.org/index.php?topic=23394.0
01:58:57saratogaisn't that what the "running time" screen gives?
01:59:35JdGordonrunning time gets reset when it charges
01:59:58JdGordoncharges in rockbox that is
02:00
02:00:24 Join klapaucjusz [0] (n=jch@88-121-17-157.rev.libertysurf.net)
02:00:43saratogamines at like 200 hours and I always charge in rockbox
02:01:11Lloreansaratoga: I think it never got updated for USB-charging players
02:01:19klapaucjuszJust passing in the hope to attract some committer's attention to FS #10832
02:01:44Lloreansaratoga: Or maybe it's broken/changed now. It used to be on the H100 that it would reset when you inserted the wall charger while in Rockbox
02:02:25 Quit MethoS- (Remote closed the connection)
02:02:48saratogaklapaucjusz: I'd like to see a fix for that but i'm not sure enough about the code you're changing to commit it
02:02:53saratogahave you asked lear ?
02:02:55JdGordonklapaucjusz: email the dev list... that needs someone who actually knows the mp4/aac stuff
02:03:56klapaucjuszJdGordon: I looked at it, but it appears there's no way I can send a message without subscribing myself.
02:04:08klapaucjuszI'd be grateful if you could do it for me.
02:04:11JdGordonso subscrive?
02:04:16saratogai guess FS #10833 - Protect against division by zero in AAC (mp4) codec looks fine
02:04:45***Saving seen data "./dancer.seen"
02:04:58klapaucjuszsaratoga: yep. Zero-risk, and only triggered by incorrect files.
02:05:55kugel_JdGordon: you can't set the scan step specifically I think, only the region
02:06:09JdGordonyes
02:06:19kugel_so why an extra tag for it?
02:06:23JdGordonthe step size changes with region
02:06:36JdGordonits probably a pointless token, which is why IO was asking :)
02:07:09kugel_if we wanted to let the user know the step, we wouldn't have the region setting
02:07:48klapaucjuszJdGordon: I prefer following it over Gmane.
02:08:03JdGordonklapaucjusz: you can subscribe without getting replies
02:08:21Lloreanklapaucjusz: If you want a patch you're involved with discussed there, you pretty much need to subscribe because it's incredibly unlikely a single message from you will be enough
02:08:46JdGordonkugel_: not necessarily... but yeah, I wont bother with it
02:09:03CIA-6New commit by 03saratoga (r23983): Commit both patches in FS #10833 - Protect against division by zero in AAC (mp4) codec by Juliusz Chroboczek. Adds some return value checking so that ...
02:09:10LloreanJdGordon: I disagree with the "if we wanted the user to know..." bit, but I do agree that a tag to show the step is basically useless.
02:09:36klapaucjuszYay!
02:10:50saratogaif you really want to get me motivated to commit your patches start optimizing libfaad
02:11:08klapaucjuszHeh.
02:12:14JdGordonanyone in here care enough about skin tokens to help chose which letters go to which info?
02:12:38saratogamake it a character thats hard to type on european keyboards
02:12:42JdGordonor even... anyone bored enough to help?
02:13:07JdGordonascii only makes things easier :p
02:13:14JdGordonlatin only even more so
02:16:26JdGordonhttp://pastebin.com/m43f104ef is the list... probably going to use 't' for the grouping prefix
02:16:31JdGordonsuggestions very welcome...
02:34:42 Quit Rondom (Nick collision from services.)
02:34:45 Join Rondom [0] (n=quassel@dslb-084-057-181-026.pools.arcor-ip.net)
02:37:17 Nick fxb is now known as fxb__ (n=felixbru@h1252615.stratoserver.net)
02:38:56kugel_JdGordon: you could have artist etc info via RDS?
02:39:11kugel_or only station info
02:41:23JdGordonrds maybe
02:41:31JdGordonI wouldnt put the presets there though
02:44:17JdGordonradio AA is going to be fun :p
02:45:20 Quit robin0800 (Read error: 110 (Connection timed out))
02:51:45klapaucjuszJdGordon: done.
02:51:47klapaucjuszMail sent.
03:00
03:02:37 Join fdinel [0] (n=Miranda@modemcable235.127-131-66.mc.videotron.ca)
03:03:17 Quit ded (Read error: 104 (Connection reset by peer))
03:03:28 Join ded [0] (n=user@pool-96-235-185-249.cmdnnj.fios.verizon.net)
03:11:42 Quit ded (Remote closed the connection)
03:22:47 Quit BlakeJohnson86 ("Leaving.")
03:26:16 Join BlakeJohnson86 [0] (n=bjohnson@c-24-118-162-123.hsd1.mn.comcast.net)
03:27:18 Join garblesnarky [0] (n=alan@cpe-24-28-74-115.austin.res.rr.com)
03:27:42 Quit kugel_ ("exit(0);")
03:29:00garblesnarkycan i shuffle by album in rockbox
03:29:14 Join googo [0] (n=admin@pool-72-66-29-116.washdc.fios.verizon.net)
03:29:21googolol rockbox and your pretty things, STUPID
03:29:32 Join n00b81_ [0] (n=n00b81@unaffiliated/n00b81)
03:29:44garblesnarkygoogo was that to me
03:29:59googogarblesnarky: say what now?
03:30:08garblesnarkynevermind im retarded
03:30:20googoindeed, in fact, all of you are :P
03:30:39garblesnarkywell thats just uncalled for
03:30:52googogarblesnarky: oh please. what is there to be in here for anyway?
03:31:13garblesnarkyIM trying to find out if i can shuffle by album in rockbox
03:31:14garblesnarkyi dont know why anyone else is in here
03:31:21googoshuffle by album in rockbox
03:31:50googohey RB developers, who the hell said you can stall up my boot and "initialize" WTF
03:32:29Lloreangoogo: You did, or you wouldn't have turned on the option.
03:32:48googoLlorean: heh, nice one. all i did was open it and close it
03:33:09LloreanBy "it" do you mean Rockbox?
03:33:25Unhelpfulit sounds like he probably means the DB
03:33:45googowait.. nvm... i thought this was
03:33:55googorockapp channel
03:34:05googoAPOLOGIES, i am trolling the wrong people
03:34:13Unhelpfulwtf is rockapp?
03:34:25googoUnhelpful: a stupid app for iphone
03:34:38googon00b81_: said to troll here, i mixed up the name
03:35:00googoyou are all very nice people, and may not deserve trolling.
03:35:11n00b81_?
03:35:46googon00b81_: yeah see you told me: "n00b81: troll #rockbox :P"
03:35:55n00b81_googo: are you 'veeloc' from #linux4nano-dev?
03:36:05n00b81_lol, nvm your host says it..
03:36:16krazykitplease try to stay on-topic in here
03:36:42n00b81_krazykit: sorry, he's a regular spammer on #linux4nano, and I banned him once. Probably just getting back at me.
03:37:37 Join ArmandiuxGS [0] (n=c84ff9d7@giant.haxx.se)
03:37:41googon00b81_: lol man ur silly
03:38:07googohttp://grab.by/1aYu
03:38:25n00b81_Not going to continue to argue. Please stay on topic or leave.
03:38:31googokrazykit: righteo, sorry for the confusion. rockapp fans deserve it.
03:38:33 Quit Tomis (Read error: 54 (Connection reset by peer))
03:38:43googon00b81_: XD dude, look at the pic. im out now.
03:39:25advcomp2019__googo, you are getting offtopic plus you could have photoshopped that
03:39:27 Quit ArmandiuxGS (Client Quit)
03:39:44 Nick advcomp2019__ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
03:39:58n00b81_advcomp2019__: Not sure about photoshop, but he obviously used something :P
03:40:00googoadvcomp2019__: ...now... why would i do that. n00b81_ and i are actually pretty good friends were both just messing with eachother
03:40:32googoso wait.. what is rockbox anyway?
03:42:43 Part googo
03:43:20 Join Tomis [0] (n=Tomis@70.134.86.126)
03:44:50 Join ArmandiuxGS [0] (n=quassel@200.79.249.215)
03:45:44 Quit ArmandiuxGS (Client Quit)
03:49:46kkurbjunJdGordon: you around?
03:50:01 Part n00b81_ ("Leaving")
03:50:20kkurbjunthe scrolling breaks with the playlist menu and the system menu
03:50:59JdGordoni r
03:51:02kkurbjunto reproduce just open the mr500 sim with cleangreen, then start playing a song that requires scrolling in the SBS, and then go into one of those menus
03:53:51JdGordoninteresting...
03:54:58 Quit efyx_ (Remote closed the connection)
03:55:06 Join n00b81_ [0] (n=n00b81@unaffiliated/n00b81)
03:58:15garblesnarkyalbum shuffle can be done really easily with "random folder advance configuration"
03:58:18 Join lennyk [0] (n=lenny@dynamic2-254-138.usc.edu)
03:58:31garblesnarkyin case anyone finds the log of this on google
03:58:42JdGordonthats sort of what its for :p
03:58:50krazykitgarblesnarky, it's also in the manual :)
03:59:12garblesnarkyyea, thats where i found it, but it wasnt exactly straightforward to find
03:59:50garblesnarkyim not complaining
03:59:53lennykhas anybody had any success building a toolchain on mac os x with rockboxdev.sh?
03:59:59garblesnarkyi just hate it when you find a question without an answer on google
04:00
04:00:05garblesnarkytrying to prevent that
04:00:36lennykI seem to try it about once every 6 months, and every time I do it's broken again
04:04:38saratogawhat fails?
04:04:49***Saving seen data "./dancer.seen"
04:05:01lennykhttp://forums.rockbox.org/index.php?topic=23398.0
04:05:07saratogaoh saw the thread, but you should really consider putting some info in there
04:05:22lennykwhat additional information should I post?
04:05:31saratogathe console log
04:05:39saratoganot just the error message
04:05:47lennykOk, I can do that
04:06:11saratogagcc version probably isn't a bad idea either
04:07:18 Quit n00b81_ ("Leaving")
04:07:58Lloreanlennyk: Can you compile that gcc cross-compiler manually, or does it only fail when rockboxdev.sh is involved?
04:08:46lennykLlorean: I haven't tried to compile it manually. It would be more of a learning curve than I have time for this evening.
04:09:02lennyksaratoga: the text file of the log was too large to post on the forum
04:09:10lennykwhat should I do with it?
04:09:17saratogalink it?
04:09:22lennyk... ok
04:13:31garblesnarkythanks for help people... another satisfied customer
04:13:35 Quit garblesnarky ("Leaving")
04:13:51lennyksaratoga: is this what you wanted? http://www-scf.usc.edu/~koepsell/temp_files/rockboxdev_log_2009.12.13.txt
04:15:22saratogalennyk: yeah but theres not much to go on from the log
04:15:40saratogaugh editing the stupid user registration page broke it
04:15:43saratogai hate the new wiki
04:16:32lennykwhat prompted the switch?
04:16:58saratogathe old one was really out of date apparently
04:17:13saratogai think its a new build of the same software
04:18:19lennykon the wiki main page it says that rockbox switched from twiki to foswiki
04:18:25lennykno that it's important
04:18:29lennyk*not
04:19:07saratogaoh then i'm wrong
04:19:23Lloreansaratoga: Well, foswiki is a fork of twiki
04:19:31LloreanSo you weren't really wrong, but it wasn't the only reason.
04:19:49lennykwhat should my next step be, time permitting, to get a working crosscompiler toolchain?
04:20:19LloreanThe twiki project itself did something "bad" (I can't remember what) so foswiki forked and is generally considered preferable to use.
04:20:28 Quit klapaucjusz ("leaving")
04:20:35saratogayou could try following the manual directions but i think they're pretty similar so it might not help
04:20:50saratogamaybe try google and see if you can figure out what apple broke with your version of the os
04:21:07saratogayou're probably not the first person to notice gcc not compiling
04:21:15lennykfwiw, rockboxdev.sh wasn't working for me recently on os x 10.5 either
04:21:22lennyki would hope so!
04:21:37lennykthanks
04:22:12saratogai guess also check back here later in case someone sees the logs and knows the answer
04:22:32lennykok. i posted it on the forum in hopes that might help.
04:23:17lennykmy OS has gcc 4.2.1. could that be a problem?
04:27:21 Quit panni_ (Broken pipe)
04:29:38 Quit froggyman (Read error: 104 (Connection reset by peer))
04:30:55martian67lennyk: rockbox uses a custom patched version of 3.5.x iirc
04:31:09martian67for good reason too :)
04:31:34lennykmartian67: I was wondering if using that version of gcc wouldn't work to compile 3.5.x
04:31:49lennyki'm trying to setup a new toolchain
04:32:22martian67i know some gcc versions have issues compiling themselves/other gcc versions
04:32:38lennykI found out that my OS has gcc 4.2.x and 4.0.x installed. I'm trying to get rockboxdev.sh to use the 4.0.x version right now
04:33:37Lloreanmartian67: Why do you think Rockbox uses 3.5?
04:34:23lennykit looks like rockboxdev.sh downloads gcc-4.0.3
04:34:28LloreanFor ARM we've been using 4.0 from the start.
04:35:13Lloreanmartian67: Did you read 3.5.X is used somewhere, that can be corrected maybe?
04:35:15lennykit looks like rockbox still uses 3.4.6 for coldfire
04:35:34saratogathats probably what he meant to say
04:36:22martian67Llorean: sorry, i got that impression from chat a long time ago :x
04:36:30martian67well dont listen to me :D
04:36:45martian67i thought 4.x had rather serious bugs for many targets
04:37:01 Quit fdinel ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
04:37:12Lloreanmartian67: >4.0 (4.1, 4.2) have showed some difficulties.
04:39:08lennykI'm having a hard time trying to get rockboxdev.sh use something other than "gcc" for the compiler. Any ideas?
04:39:26lennyk... I'm sitll working through the script trying to understand it.
04:39:48Lloreanlennyk: Have you looked at the manual instructions yet?
04:40:01LloreanThey're probably much simpler than trying to deconstruct rockboxdev.sh
04:40:16lennykthanks Llorean I'll try that
04:43:14 Quit TheSeven (Nick collision from services.)
04:43:31 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
04:43:43 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
04:44:57saratogamartian67: newer gcc builds work about as well as 4.0.3 for most rockbox targets
04:45:19saratogaso theres not much incentive to change, except for a few newer targets that are better supported in newer versions
04:49:27lennyksaratoga: so are there still any compelling reasons to setup a new toolchain?
04:50:12saratogayeah a few, its been discussed recently
04:50:16saratogacheck the logs
04:50:36lennykoh ok. thanks
04:53:26lennykso compiling manually failed on building binutils
04:53:27lennykchecking whether the C compiler (gcc -no-cpp-precomp -g -O2 ) works... no
04:53:27lennykconfigure: error: installation or configuration problem: C compiler cannot create executables.
04:53:41 Quit Sajber^1 (Read error: 54 (Connection reset by peer))
05:00
05:05:09 Quit ansuz ("A motion to ajourn is always in order.")
05:14:04 Join Strife89 [0] (n=michael@adsl-80-164-135.mcn.bellsouth.net)
05:14:32lennykwell I'll have to try this again later. thanks for your help everyone.
05:23:15 Quit ps-auxw (Remote closed the connection)
05:23:33 Join ps-auxw [0] (n=arneb@dyn37.ps-auxw.de)
05:39:58 Quit Horscht ("Verlassend")
05:48:01CIA-6New commit by 03kkurbjun (r23984): M:Robe 500 Add some additional pin initialization for the SPI devices.
06:00
06:04:52***Saving seen data "./dancer.seen"
06:08:48 Quit bzed (Remote closed the connection)
06:12:08 Join bzed [0] (n=bzed@devel.recluse.de)
06:16:09JdGordondoes anyone have any ideas how do insert a line in a skin based on a conditional without using viewports?
06:16:22JdGordonI dont want a blank line at the top if the conditional fails
06:18:58JdGordonseems svn has the same blank line.. so I'll keep it
06:20:38 Quit Strife89 ("Bed.")
06:22:26 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
06:28:20FlynDicemc2739: ping
07:00
07:02:22 Quit lennyk ()
07:05:20pixelmaJdGordon: the blank line in the radio screen that is there additionally on small screens?
07:05:38JdGordon?
07:06:23pixelma??
07:06:38 Quit Zaba_ (Read error: 110 (Connection timed out))
07:06:40JdGordonI'm slightly confused
07:07:48pixelmaI thought you were working on the radio screen and meant the blank line at the top of that - which is only there on small screens (I think domonoky said that the check was the wrong way around there)
07:08:15JdGordonthats the one... its only on small screens? thats odd...
07:09:47pixelmaI think it was meant the other way around
07:10:38JdGordonprobably... see any obvious mistakes in http://pastebin.com/m3d21a7c7 ?
07:12:00JdGordoncheckwps isnt being very helpful :(
07:12:37pixelmathe hwcodec radio screen also has a peakmeter because you can record directly from the radio screen (not via recording screen and source "FM radio")
07:13:03JdGordonyeah, i'll worry about that later... right now I just want to get a skin working
07:13:04 Quit bluebrother (Nick collision from services.)
07:13:06 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
07:13:16JdGordonstupid checkwps just says that file cant load
07:13:42 Quit liar (Read error: 113 (No route to host))
07:15:49pixelmaare the colons really part of the translatable string for "Mode:" and does "Station:" even exist?
07:16:23JdGordonyes and yes
07:16:31JdGordonits not the Sx's.. removed them and still failed
07:17:11JdGordon%T is the error
07:18:44pixelmawhy is it uppercase? %T also already exist on touchscreen devices
07:18:53*JdGordon slaps self
07:19:16JdGordon%T is for the presets tokens, %t is for the radio ones...
07:19:22JdGordonPLEASE suggest better ones :p
07:19:51JdGordonhttp://pastebin.com/m38b650d3 is the list as it is
07:20:20 Join kood777 [0] (n=79b45a2c@giant.haxx.se)
07:21:33JdGordonyay! its working
07:28:08 Quit tha (Read error: 104 (Connection reset by peer))
07:28:32 Join tha [0] (i=1038@ccc2.rbg.informatik.tu-darmstadt.de)
07:29:16CIA-6New commit by 03kkurbjun (r23985): DM320: Add generic pin-setup function.
07:32:31CIA-6New commit by 03kkurbjun (r23986): M:Robe 500: Start adding some composite output stuff based on OF. Initialize LCD pins with pin-setup function.
07:39:18 Quit saratoga ("Page closed")
07:52:37CIA-6New commit by 03kkurbjun (r23987): M:Robe 500: Initialize SPI pins with pin setup function
08:00
08:00:39CIA-6New commit by 03kkurbjun (r23988): M:Robe 500: Initialize UART1 pins with pin setup function
08:04:29JdGordonhas anyone tried preset and frequency "testing" in the sim?
08:04:35JdGordonis it toally broken?
08:04:53***Saving seen data "./dancer.seen"
08:06:07pixelmathe radio (and recording screen) are just dummy things in the sim, I wouldn't wonder if there isn't much functionality behind it besides button handling etc.
08:06:40JdGordonI was hoping it does some faking...
08:06:53JdGordonotherwise debugging these new tokens is going to suck!
08:08:18JdGordonah, it does do some incredibly simple faking
08:14:51 Join stoffel [0] (n=quassel@p57B4DAEB.dip.t-dialin.net)
08:18:56CIA-6New commit by 03kkurbjun (r23989): M:Robe 500: Initialize TSC2100 pins with pin setup function
08:21:45 Quit TheSeven ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
08:32:57*JdGordon fails to get the current frequency on the skin :(
08:33:36JdGordondoes anyone know the tuner drivers?
08:34:02 Join lennyk [0] (n=lenny@dynamic2-254-138.usc.edu)
08:36:49JdGordonarg... apparently they dont keep track of the current frequency :(
08:41:22 Quit kood777 ("CGI:IRC (EOF)")
08:50:15 Quit faemir ("Leaving")
08:51:24 Join flydutch [0] (n=flydutch@host140-41-dynamic.116-80-r.retail.telecomitalia.it)
08:54:45 Quit Tomis (Read error: 145 (Connection timed out))
08:56:54 Join petur [50] (n=petur@rockbox/developer/petur)
09:00
09:01:40 Quit mt (Read error: 54 (Connection reset by peer))
09:02:11 Join mt [0] (n=mtee@rockbox/developer/mt)
09:04:14*JdGordon giggles like a school girl
09:04:21JdGordongot the progressbar working in the fm skin
09:05:01 Quit gevaerts (Nick collision from services.)
09:05:13 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
09:31:08 Join Bagder [0] (n=dast@giant.haxx.se)
09:42:48 Join einhirn [0] (n=Miranda@bsod.rz.tu-clausthal.de)
10:00
10:04:55***Saving seen data "./dancer.seen"
10:19:18 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
10:28:28 Join maruk [0] (n=papier@titanium.sdv.fr)
10:28:33 Join archivator [0] (n=archivat@77.70.28.57)
10:31:41 Join Kitar|st [0] (n=Kitarist@89.142.73.118)
10:31:44 Join shai [0] (n=Shai@l192-117-110-233.cable.actcom.net.il)
10:35:54 Quit phanboy4 (Read error: 110 (Connection timed out))
10:37:26archivatorif anyone has a few minutes to test out a new plugin, I'd like to point your attention to FS #10065 (a spectrum analyzer). This was a public service announcement, thank you for listening! :)
10:42:50 Join Sajber^ [0] (n=Sajber@h-143-7.A213.priv.bahnhof.se)
10:49:31 Quit Kitar|st (Read error: 110 (Connection timed out))
10:49:39 Join Kitar|st [0] (i=Kitarist@BSN-77-22-115.dial-up.dsl.siol.net)
10:49:51 Quit Kitr88 (Read error: 110 (Connection timed out))
11:00
11:12:57 Quit einhirn (Read error: 104 (Connection reset by peer))
11:13:13 Join einhirn [0] (n=Miranda@bsod.rz.tu-clausthal.de)
11:14:25 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
11:15:48topikwill just applying the patch make it build and be included ?
11:15:52topik(to archivator)
11:16:10 Quit maruk ("Leaving.")
11:16:24archivatortopik: you need to apply the patch and then compile a build for your target
11:16:56archivatorThe plugin will be under demos/fft
11:18:44 Join maruk [0] (n=papier@titanium.sdv.fr)
11:18:53topikit's not building. no rule to make target `..../fft_colors.16.o', needed by `..../libpluginbitmaps.a'.
11:19:17archivatortopik: which target?
11:19:20topikfuze
11:21:23topiksame for nano2g
11:21:54archivatortopik: it compiles fine on my dev tree, will try it on a fresh checkout
11:22:38gevaertstopik: you applied using patch I guess?
11:23:09gevaertsIf so, you probably don't have apps/plugins/bitmaps/native/fft_colors.16.bmp
11:24:10topikyes gevaerts
11:24:13topikx2
11:24:38 Quit lennyk ()
11:25:17topiki'll grab fft.v5.zip
11:25:54archivatortopik: that file *is* in the zip but under a different name
11:26:08archivatorfft_colors.320x240x16.bmp I think
11:26:14archivatorYou'll have to rename it.
11:26:15gevaertsYou can use "git apply" even if your tree isn't a git checkout. That will give you all files. Alternatively, you can wait for archivator to attach that file to the task separately
11:27:12 Quit archivator (""restarting X :("")
11:29:14 Join archivator [0] (n=archivat@77.70.28.57)
11:29:26topikno such thing as git on this build vm
11:30:57 Join DerPapst [0] (n=DerPapst@p4FE8F2ED.dip.t-dialin.net)
11:31:26 Join Omlet [0] (i=omlet05@130.7-65-87.adsl-dyn.isp.belgacom.be)
11:31:32topikcopying/renaming the bmp makes it work
11:33:32archivatorI also attached the file to the task. It *is* the same file from the archive, just renamed.
11:34:22gevaertsarchivator: please split the lines in const.h
11:34:29archivatorIs the fuze faster or slower than the older PP targets (ipod video and the like)?
11:34:40archivatorgevaerts: out of curiosity, why?
11:35:11gevaertsbecause 50000 characters on a single line is too much
11:35:16archivatorThose are not meant to be readable or edited. It's just a dump of data.
11:35:30gevaertsI know, but even then
11:36:06archivatorSplit at 80 chars or can we make an exception and have a bit more?
11:37:38gevaertsI'd just split at 80. Since as you said it's not meant to be edited, any number will work equally well, so I don't see a good reason to pick another number
11:39:04 Join pondlife [0] (n=Steve@cpc1-rdng11-0-0-cust362.winn.cable.ntl.com)
11:39:41topiki get a black screen with your plugin, archivator
11:39:50topikand no idea how to get out of it :)
11:41:23archivatortopik: sounds bad.
11:41:45archivatorThe sim worked, I think.
11:41:53topik'button_power'
11:42:05topiki'm guessing i'm missing bitmaps
11:42:10archivator?
11:42:37archivatorYou're not. There's just the one. It's a bug of some sorts.
11:42:44topik'button_power' is assigned to fft_quit
11:42:54topiki'll reassign that to button_home
11:42:57topikand try again
11:44:44archivatortopik: ah, the keymaps *are* probably very inconvenient. Sorry, I threw those together when I started, stealing ideas from the oscilloscope...
11:44:52 Join funman [0] (n=fun@rockbox/developer/funman)
11:47:48 Join evilnick [0] (i=4d63ba66@rockbox/staff/evilnick)
11:49:26topiksame lack of anything on my nano2g
11:49:56archivatortopik: I'm looking into it, give me a few minutes
11:51:09archivatortopik: you *do* have something playing, right? :) The plugin should exit if you don't but who knows..
11:51:31topikhehe, yea
11:52:22archivatortopik: and if I tell you that the fuze sim is working, would you believe me? :-/ i hate this type of bugs.. gr..
11:52:45topikmust be the wrong bmp?
11:52:49topiki coped fft_colors.320x240x16.bmp to fft_colors.16.bmp
11:52:52topikor copied
11:52:53archivatortopik: it's not.
11:53:02topikwhat else could be different?
11:53:13archivatortopik: the bmp is only used in spectrogram mode, you should be seeing lines mode at least..
11:56:07funmanarchivator: i believe CPU is the 1st core, while COP is the 2nd core
11:57:30archivatorfunman: Where does that leave my thread with IF_COP(,CPU), still on the first core on dual-core targets? More importantly, where will it go on single-core targets?
11:59:44funmanIF_COP(, x) evaluates to nothing on single-core targets, so the thread will go on the only core: the CPU
12:00
12:00:01funmanon dual-core it will go on the specified core: here it is CPU as well
12:01:25archivatorfunman: right, so it shouldn't break anything on single core targets but wouldn't give an improvement on dual-core targets either.
12:01:31archivatorBack to square one..
12:01:42 Join MethoS- [0] (n=clemens@134.102.106.250)
12:04:23archivatortopik: on the nano, when you press left and right, do you at least see the popups saying "spectrogram" or "lines" or something else? Also, does playback continue when you enter the plugin?
12:04:58***Saving seen data "./dancer.seen"
12:10:49 Quit BHSPitMonkey ("Ex-Chat")
12:10:54topikplayback does continue yes
12:13:11 Quit MethoS- (Remote closed the connection)
12:13:45archivatortopik: I have an idea what might be happening and why the sims work.. will you be here to test a new patch?
12:14:03archivatorI'll need about 10 minutes..
12:14:47topikyeah i'll be here
12:14:48archivatorScratch that. can't be what I thought it was..
12:14:51archivatorSorry.
12:19:07 Join perfectdrug [0] (n=5b0ece1c@giant.haxx.se)
12:19:18archivatortopik: and I take it you don't see *anything* when you randomly press buttons?
12:20:08perfectdrugSVG for the Onda Vx777 is also ready ( FS #10854 ), who ownes this device?
12:21:09 Quit rvvs89 (Remote closed the connection)
12:23:28topiknothing whatsoever
12:24:09topikyou are saying the left and right button should give some text?
12:24:32archivatorThey should.
12:24:41archivatorAnd they do, in sim builds.
12:25:38archivatorSomething locks up on the actual target. Could be threading issue, although I don't see why it would only show on single core targets (I have an ipod video and it works fine on target)..
12:26:37topikcould it be the screen size ?
12:26:50topikthere's some lcd_width >= 480 stuff in your patch
12:26:56topikneither the nano2g or fuze are that wide
12:26:57archivatorShouldn't be. The sims for both targets work just fine..
12:27:06archivatorlcd_width > 480?
12:27:09archivatorWhere?
12:27:21topikline 66 of patch.v6.patch
12:27:29 Join rvvs89 [0] (n=ivo@bright-snat.ucc.asn.au)
12:28:11archivatorThat's context for the patch to apply properly, it's not part of the actual patch (only lines starting with + or - are).
12:31:56topikdoes the rockbox site list display resolution for the devices somewhere? before you could see the resolution through the links on the themes pages
12:33:34archivatorBuyersGuide has some of the players..
12:35:32rvvs89DeviceChart
12:35:40archivatortopik: out of curiosity, you *did* restart the players after copying the new build, right? ROLO would suffice but you'll need the new rockbox image...
12:37:47topiki restarted yes
12:37:59topikfuze doesn't have usb, nano2g usb i don't quite trust
12:38:36archivatorI
12:38:42archivator'm out of ideas then, sorry
12:39:11topikthanks rvvs89. somewhat outdated though
12:39:21archivatorI'll need more feedback but I tried a few things on my player, trying to force it to crash in the way yours are but it still worked :(
12:41:14topikit doesn't really crash i think
12:41:21topikmusic still plays
12:42:33topiki must be doing something wrong. hopefully someone else can test it too
12:42:59archivatorI have a theory that's easily testable.
12:43:10archivatorCould you please change FFT_SIZE to be 1024 in fft.c ?
12:43:32archivatorline 179
12:47:44topikbuilding
12:51:48 Join Lynx_ [0] (n=Lynx@xdsl-78-34-254-130.netcologne.de)
12:54:41topiksomeone awesome should patch out the fuze's OF's 'refreshing your media' nonsense
12:55:22topikthat did it archivator
12:55:42topikpressing left gave me a 'data abort at 3077E268' though
12:55:58topikbut the line thingy worked before that
12:56:05archivatorthat could easily be the bitmap now.
12:56:35 Join watto [0] (n=watto@193.203.81.165)
12:56:58topiksorry, mixed up left and right. right gave the data abort. left gives colorful stuff
12:56:59archivatorAt least I know the reason now - I need a better way to access the PCM data :( Which means I have no idea which target works now and which doesn't..
12:57:56archivatorAh, the bars thing is probably going to be removed some time soon. It's full of incorrigible bugs..
12:59:10topikall bars are max length as well
12:59:12topikwhile they last
13:00
13:00:20archivatorYeah, those are pretty useless, anyway
13:02:02archivatortopik: you could try FFT_SIZE 2048 as well. Might work, might fail. Depends on implementation details in rockbox. I have no idea why the buffer size is not the same on all targets but there probably is a valid reason..
13:02:21 Join Pingu1084 [0] (n=54df3b31@giant.haxx.se)
13:03:36topikif you want i can try that
13:03:44Pingu1084p
13:03:49Pingu1084fu
13:04:11archivatortopik: If you don't mind, I'd like to know where those 2 targets stand. Also, you'll get more accurate graphs that way
13:04:19 Quit perfectdrug ("CGI:IRC")
13:06:29topiki haven't tried 1024 on my nano2g
13:06:49 Quit antil33t (Read error: 104 (Connection reset by peer))
13:06:54 Join antil33t [0] (n=Mudkips@203-184-54-232.callplus.net.nz)
13:07:28 Quit Pingu1084 ("CGI:IRC")
13:07:41archivatortopik: that's okay, it most likely works..
13:08:03 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:09:57topikbuilding again
13:14:29 Join philip__ [0] (n=chatzill@cpe-173-171-133-36.tampabay.res.rr.com)
13:14:52 Nick philip__ is now known as angelwolf71885 (n=chatzill@cpe-173-171-133-36.tampabay.res.rr.com)
13:17:39 Quit angelwolf71885 (Client Quit)
13:19:51archivatortopik: any luck?
13:20:33 Join perfectdrug [0] (n=5b0ece1c@giant.haxx.se)
13:20:40 Join Pingu1084 [0] (n=54df3b31@giant.haxx.se)
13:21:21Pingu1084How can I get the permissions for creating new topics in the wiki?
13:21:27 Join webguest07 [0] (n=18185071@giant.haxx.se)
13:22:10 Quit webguest07 (Client Quit)
13:24:22Pingu1084a
13:25:40Pingu1084a
13:26:48Pingu1084ò
13:27:30archivatorPingu1084: refrain from spamming, please. You'll need someone with admin privs to enable your account after you've registered.
13:27:39 Quit Pingu1084 ("CGI:IRC")
13:30:08topiksorry archivator, work took my attention. unzipping now
13:30:17topikunzipping rockbox.zip, don't get any weird ideas
13:30:58archivatortopik: wouldn't think of it until you pointed it out :)
13:32:15 Join Pingu1084 [0] (n=chatzill@host-84-223-59-49.cust-adsl.tiscali.it)
13:33:07Pingu1084a
13:33:23 Quit Pingu1084 (Client Quit)
13:41:14topikseems it works pretty well now, archivator
13:41:24 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
13:41:40 Quit Sajber^ (Read error: 104 (Connection reset by peer))
13:41:52 Join BlakeJohnson86 [0] (n=bjohnson@c-24-118-162-123.hsd1.mn.comcast.net)
13:42:08topikall 3 modes, and unless i'm wrong i now get (for the first time) the text in the top-right corner too
13:46:19 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
13:46:33 Join BlakeJohnson86 [0] (n=bjohnson@c-24-118-162-123.hsd1.mn.comcast.net)
13:47:51 Join pondlife1 [0] (n=Steve@cpc1-rdng11-0-0-cust362.winn.cable.ntl.com)
13:48:32 Part pondlife1
13:50:04 Nick YPSY is now known as Ypsy (n=ypsy@geekpadawan.de)
13:51:57 Join teru [0] (n=teru@KD059133115245.ppp.dion.ne.jp)
13:52:33 Quit perfectdrug ("CGI:IRC")
13:53:29 Join froggyman [0] (n=sopgenor@pool-72-69-220-194.chi01.dsl-w.verizon.net)
13:56:22 Join Pingu1084 [0] (n=chatzill@host-84-223-59-49.cust-adsl.tiscali.it)
13:58:02gevaertsPingu1084: what's your wiki username?
13:59:05 Join perfectdrug [0] (n=5b0ece1c@giant.haxx.se)
13:59:28Pingu1084It's PingU1084
13:59:41archivatortopik: yeah, that's what I meant when I said "popup" .. Well, at least now I know some targets can't give me enough data.. thanks for testing! I need to figure out what to do with that information, though :(
13:59:51archivatorPingu1084: It needs to be your real name
14:00
14:00:21 Quit Pingu1084 (Client Quit)
14:00:47gevaertsarchivator: actually, any registered wiki user can add others
14:01:26archivatorgevaerts: I didn't know that. Always assumed the regular devs had admin access or something. How do you add others?
14:01:47gevaertsarchivator: by editing WikiUsersGroup
14:01:51 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
14:02:12archivatorgevaerts: right, I'll have that in mind the next time someone asks! :)
14:02:29gevaertsarchivator: you can do this one if he comes back :)
14:03:28 Quit pondlife (Read error: 110 (Connection timed out))
14:03:58 Part froggyman
14:04:00 Join chykle897 [0] (i=riaa@c-98-197-54-196.hsd1.tx.comcast.net)
14:05:03***Saving seen data "./dancer.seen"
14:05:28archivatorCan anyone suggest a way of accessing low-latency PCM data other than the peak buffer?
14:12:58chykle897i have a sansa fuze v2 do i have to downgrade it to v1 to get rockox to work?
14:13:10topikyou can't downgrade a hardware revision
14:13:21chykle897aww
14:13:37rvvs89archivator: I would also be interested in this.
14:14:01chykle897so i wonr be able to rockbox my sansa/
14:14:16topiknope, not until rockbox supports the v2 fuze
14:14:30chykle897aww...that sucks
14:14:35topikand there hasn't been any movement on that lately afaik
14:14:59chykle897ohh well i guess i'll wait till it's supported
14:15:41chykle897cya
14:15:58 Quit chykle897 ("( www.nnscript.com :: NoNameScript 4.22 :: www.esnation.com )")
14:17:26archivatorre: PCM - I was thinking of having the playback thread call a user-defined callback (or just memcpy to a user-defined buffer) - but you'd need to ensure that you're not changing data midway and to do that in the playback thread, you'd need a mutex_lock that doesn't block - if the mutex is locked, you just skip this part entirely, otherwise, you copy the PCM data to the buffer. Make sense?
14:20:10archivatorOn the other hand, I assume there's a good reason why there isn't a mutex_locked() function..
14:21:00 Join fml [0] (n=5dd2dc5d@giant.haxx.se)
14:21:28topikis it because mutexes want to be free?
14:23:40fmlAre we in the freeze phase now? Or may patches still be committed? I'm talking specifically about FS #10786
14:23:50gevaertswe're not in a freeze
14:27:18fmlOK, I'll then try to commit it. Will test compile for some platforms first.
14:29:16gevaertsmaybe ask teru first. It's his patch, maybe there's a good reason why he hasn't committed it yet
14:29:25teruCould anyone test/review patch on FS #7772(Sudoku: "Solve" Crash)?
14:29:59fmlAlso, if there's no objections against FS #10843 I'd like to commit it as well. I never liked that thin font on the front page.
14:30:54terugevaerts: because i don't have non-RTC targets and don't know if it's desired.
14:47:05 Quit parafin (Read error: 60 (Operation timed out))
14:49:33 Join parafin [0] (i=parafin@paraf.in)
15:00
15:02:45topikteru: i tried your patch. is there any way to reproduce the situation it crashed in before?
15:02:54topik(suduko patch that is)
15:03:16fmlHrm... I can't commit to svn. Either I forgot my password or I have some forewall in the way. How can I check what's wrong? E.g. how can I verify my svn password?
15:03:26topikor perhaps sudoku ... fs#7772
15:06:04 Join Mark__T [0] (n=Mark__T@foresight/developer/mark)
15:06:05terutopik: select "new" in menu and save it as is (i.e. no numbers are set). then select solve in menu.
15:12:03fmlStill no success in committing. For everyone and anybody who can commit: I think FS #10786 and FS #10843 can be committed and the FS tasks closed.
15:13:17 Quit shai (Read error: 104 (Connection reset by peer))
15:13:43 Quit fml ("CGI:IRC")
15:13:53 Join shai [0] (n=Shai@192.117.110.233)
15:20:44topikseems to work fine, teru
15:21:29 Join Omlet05 [0] (i=omlet05@130.7-65-87.adsl-dyn.isp.belgacom.be)
15:22:18 Quit funman (niven.freenode.net irc.freenode.net)
15:22:18NSplitniven.freenode.net irc.freenode.net
15:22:18 Quit Omlet (niven.freenode.net irc.freenode.net)
15:22:18 Quit bluebroth3r (niven.freenode.net irc.freenode.net)
15:22:18 Quit Rondom (niven.freenode.net irc.freenode.net)
15:22:18 Quit xavieran (niven.freenode.net irc.freenode.net)
15:22:18 Quit GeekShadow (Read error: 104 (Connection reset by peer))
15:22:58CIA-6New commit by 03mc2739 (r23990): Make title on the front page sans serif ...
15:23:45 Quit thegeek (niven.freenode.net irc.freenode.net)
15:23:45 Quit Slasheri (niven.freenode.net irc.freenode.net)
15:23:45 Quit fyrestorm (niven.freenode.net irc.freenode.net)
15:23:45 Quit jon-kha (niven.freenode.net irc.freenode.net)
15:24:03NHealniven.freenode.net irc.freenode.net
15:24:03NJoinfyrestorm [0] (n=nnscript@cpe-69-203-150-85.si.res.rr.com)
15:24:13 Quit antil33t (Read error: 104 (Connection reset by peer))
15:24:15 Join jds2001 [0] (n=jds2001@fedora/jds2001)
15:24:31 Join bluebrother [0] (n=dom@f053154001.adsl.alicedsl.de)
15:24:43 Join funman [0] (n=fun@2001:0:53aa:64c:ee:1b26:b23e:7ca8)
15:25:12 Nick funman is now known as Guest28841 (n=fun@2001:0:53aa:64c:ee:1b26:b23e:7ca8)
15:25:15 Join GeekShadow [0] (n=Antoine@LPuteaux-156-15-25-14.w82-127.abo.wanadoo.fr)
15:25:32 Join antil33t [0] (n=Mudkips@203-184-54-232.callplus.net.nz)
15:25:33NJoinRondom [0] (n=quassel@dslb-084-057-181-026.pools.arcor-ip.net)
15:26:30NJoinOmlet [0] (i=omlet05@130.7-65-87.adsl-dyn.isp.belgacom.be)
15:26:30NJoinbluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
15:26:30NJoinxavieran [0] (n=xavieran@ppp118-209-112-86.lns20.mel4.internode.on.net)
15:26:33terutopik: ok, thanks.
15:26:35 Join thegeek [0] (n=nnscript@129.241.123.168)
15:27:34 Quit perfectdrug ("CGI:IRC (EOF)")
15:27:48 Quit bluebroth3r (Read error: 104 (Connection reset by peer))
15:27:58 Join shai_ [0] (n=Shai@192.117.110.233)
15:29:00 Quit Omlet (Success)
15:29:57 Nick shai_ is now known as Guest62080 (n=Shai@192.117.110.233)
15:31:39 Quit shai (Read error: 60 (Operation timed out))
15:31:41 Quit Guest62080 (Connection reset by peer)
15:32:30 Join hebz0rl [0] (n=hebz0rl@dslb-088-067-207-158.pools.arcor-ip.net)
15:37:19Lynx_Are there touch-enabled WPSs available for the D2 yet?
15:42:16 Quit DerPapst ("Leaving.")
15:45:13topikteru: noticed another odd thing. when i save the empty sudoku puzzle and get back to the menu, i have to click a button before scrolling in the menu works
15:51:39 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
15:52:02 Quit thegeek (Read error: 110 (Connection timed out))
15:52:51 Part LinusN
15:53:57terutopik: which target are you using?
15:56:02topiksansa fuze
15:59:21teruit sounds like FS #10097.
16:00
16:00:58 Join GeekShado_ [0] (n=Antoine@LPuteaux-156-15-25-14.w82-127.abo.wanadoo.fr)
16:01:00topikit's still within the plugin though
16:01:15topikit's after the 'save' thing when you create a new empty sudoku
16:04:20 Quit einhirn (Read error: 104 (Connection reset by peer))
16:05:00Lynx_How can I press on/off/hold and +, - and Menu in the D2 simulator?
16:05:07***Saving seen data "./dancer.seen"
16:05:25Lynx_Hmm, + and - is just that on the keyboad, but menu is not 'm' it seems.
16:09:45 Quit dfkt (Read error: 110 (Connection timed out))
16:10:07terutopik: I haven't ever seen issue like that using gigabeat, and I have no idea about that.
16:14:58 Quit GeekShadow (Read error: 110 (Connection timed out))
16:19:10 Join Kitr88 [0] (n=Kitarist@89.142.55.251)
16:21:53 Join hanselda [0] (n=hanselda@fkplap05.physik.uni-ulm.de)
16:22:14 Quit teru ("Quit")
16:22:47hanseldaHello all. I have just installed the rockbox on my sansa clip. It seems that the battery life is much shorter than the original firmware. Is there anything I can tune to increase the battery life?
16:30:58Lynx_In WPS files, is x 0 y 0 the top left or bottom left?
16:33:39 Quit Omlet05 (Client Quit)
16:34:06pixelmatop left (I think the CustomWPS page in the wiki also says so)
16:36:51Lynx_pixelma: ah, couldn't find it on there
16:37:07 Quit Kitar|st (Connection timed out)
16:37:32Lynx_wait, and with images, the bottom left corner is the one defined by the coordinates? like in %X?
16:38:20 Quit Kitr88 (Success)
16:41:21Lynx_pixelma: Somehow I don't get it, so the cabbiev3.wps contains: %xl|M|vol0.bmp|134|0| So x is 134, somewhere in the middle, and 0 for y should be at the top. But the volume icon is at the bottom in that wps. I'm confused.
16:44:46 Join ilam [0] (n=18185071@giant.haxx.se)
16:44:54ilamhello
16:45:25ilamanyone there im a noob and have a problem with my rockbox iriver h120
16:45:58 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-zstkkcnhetrtsvgd)
16:46:41ilamhello can give me some help with rockbox, thanks
16:47:13saratogaask question
16:48:46ilamok so i have a h120 with a cf mod and it has been flashed, recently i have tried the auto install of 3.4 from 3.3 now the iriver wont play anything and keeps skipping tracks and flashing "codec failure"
16:48:55ilamwhat should i do?
16:51:32saratogacodec failure usually means you have a corrupted install and it can't load codec files
16:52:32ilami am willing to use 3.3 iinstead of 3.4 if it is to diffucult. But I cannot find the files for the 3.3 on the rockbox site. when i check the info on the player it still says its version 3.3, i fiqure it has something to do with flashing and a incompatabilty with the new 3.4
16:53:49 Quit Mark__T ("Leaving.")
16:55:30 Join DerPapst [0] (n=DerPapst@p4FE8F2ED.dip.t-dialin.net)
16:59:08 Join toffe82 [0] (n=chatzill@12.169.218.14)
17:00
17:00:52ilamdoes anyone know where i can get a copy of the last stable version of 3.3?
17:01:47Zagorilam: http://download.rockbox.org/release/3.3/
17:03:00 Quit Zagor ("Don't panic")
17:04:19 Nick fxb__ is now known as fxb (n=felixbru@h1252615.stratoserver.net)
17:07:33 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
17:09:15pixelmaLynx_: is the bitmap (the correspndent %xdM) inside a viewport? See the "notes" and examples here: http://www.rockbox.org/wiki/CustomWPS#Images
17:11:24 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
17:12:10 Quit ilam ("CGI:IRC (EOF)")
17:13:16 Nick marga_ is now known as marga (n=marga@186.18.92.38)
17:19:46 Quit hanselda (Remote closed the connection)
17:22:15 Join ivanrocks321 [0] (n=18185071@giant.haxx.se)
17:22:52 Join jon-kha [0] (i=jon-kha@kahvi.eu.org)
17:22:52 Join thegeek [0] (n=nnscript@s168c.studby.ntnu.no)
17:22:52 Join Slasheri [0] (i=miipekk@rockbox/developer/Slasheri)
17:23:39 Join liar [0] (n=liar@83.175.83.185)
17:24:29 Join einhirn [0] (n=Miranda@bsod.rz.tu-clausthal.de)
17:25:11Lynx_pixelma: Ah, I didn't think of that, maybe that explains it. Thanks!
17:25:47 Quit dfkt_ (Read error: 110 (Connection timed out))
17:30:32ivanrocks321I have a question with installing 3.4. I have a cf modded h120 with 7pre4 bootloader installed and flashed replaceing the original iriver firmware, before it had rockbox 3.3 installed. I have tried to install the new 3.4 but I cannot play any music, skips tracks, and cannot display any of the album art like it usually does. when i check the player it says its still version 3.3. So i tryed to reflash but says incompatable version. Im am completely lo
17:30:43 Quit GeekShado_ (Connection timed out)
17:31:01 Quit Guest28841 ("free(random());")
17:32:54topikare you trying to install manually?
17:33:22topiki don't think you need to reflash to upgrade, just extract the rockbox zip again to the device
17:33:25topikor use rbutil
17:35:29ivanrocks321I have tried both ways auto and by myself. and i get the same problem with 3.4. it works sometimes but when i turn the player off and on the problem comes back.
17:36:13ivanrocks321I have also tried reformatting the drive as well, then reinstalling
17:36:28pixelmaivanrocks321: if it is somewhat like flashing an Archos (I don't have an H100 so be careful with my advice) you need to make sure that the plugins and the running rockbox version are the same so that the flash plugin works. Unzip the new version fully - load rockbox.iriver (I guess) and then the flash plugin
17:37:31pixelmamaybe wait for someone else who can tell for sure, be a bit patient
17:37:35ivanrocks321how to i make sure about the plugin, sorry im a noob and i flashed this thing a while ago
17:37:42 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
17:38:18ivanrocks321yeah, your right i can wait but ill still check the plugin issue.
17:39:17pixelmaas I said if you unzip the new version and start by chosing rockbox.iriver (then codecs and other plugins will work too)
17:41:04pixelmabut if you don't flash the new version then the bootloader will start the old version if you turn the player off then on again and you'll get the codec failures because the main firmware version (the flashed one) doesn't match the newer codecs or plugins which are started from the disk
17:41:42 Quit dmb (Read error: 54 (Connection reset by peer))
17:42:50 Join dmb [0] (n=Dmb@unaffiliated/dmb)
17:44:33ivanrocks321ok, i think i got the theory, so now ill just have to work out each step. The first is to unzip 3.4 to the root of the drive, Should i at this point reformat as well?
17:45:12pixelmashouldn't be needed
17:46:53ivanrocks321ok then whats the next step.
17:47:32 Quit GeekShadow (Read error: 104 (Connection reset by peer))
17:48:47 Part Bagder
17:52:32 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
17:55:14 Join Tomis [0] (n=Tomis@70.134.70.250)
17:57:26 Join SuperUser [0] (n=5950bebb@giant.haxx.se)
17:57:43SuperUserHaloa Rockbox
17:57:54markunhi SuperUser
17:58:22SuperUserThose times, rockbox boot had begun to take times :(
17:59:02SuperUseris that a known issue, or something on my end? For me, since this viewport/skin things reworks, boot is slow like hell
18:00
18:00:41SuperUsersince those status bar reworks in fact, things began to beiing slow :( (a very good Rockbox advantage that you made guys)
18:01:44markunSuperUser: I don't know if it's normal. Which player do you have?
18:03:34 Quit petur ("work->home")
18:03:35SuperUsermarkun: a gigabeat. That definitively since status bar reworks i.e since 23904
18:04:00markunSuperUser: my gigabeat broke so I can't check..
18:04:26markunhopefully someone else can help you. Just stay connected to the channel for a while.
18:04:30gevaertsso 23903 is fine?
18:04:46 Join Jaykay [0] (n=chatzill@p5DDC5EB6.dip.t-dialin.net)
18:05:09***Saving seen data "./dancer.seen"
18:05:18toffe82markun: you don't want to fix it ;)
18:05:29markuntoffe82: I
18:05:41markunI'm thinking about just buying a flash based player
18:07:13SuperUseryeah
18:07:20SuperUser@gevaerts
18:09:14SuperUsergevaets: and even, it slowed a bit before, still because status bar thing
18:09:18markunSuperUser: do you still have that build?
18:09:33markunr23903 I mean
18:09:57SuperUserguys, that pitty to lost the super fast rockbox boot for status bar? (/me use the inbuilt one since long)
18:10:48markunSuperUser: well, it sounds like a bug, not like a trade-off for having a skinnable status bar.
18:14:58 Quit SuperUser ("CGI:IRC (Ping timeout)")
18:19:28markunhm..
18:19:46markunI was prepared to make a build for him, but if he disconnects..
18:21:21 Quit evilnick (Ping timeout: 180 seconds)
18:21:46 Join evilnick [0] (i=4d63ba66@rockbox/staff/evilnick)
18:22:49 Join faemir [0] (n=faemir@78.33.109.163)
18:24:42ivanrocks321does anyone have the zip file for rockbox 3.3 for iriver h120?
18:25:52topikivanrocks321: http://download.rockbox.org/release/3.3/
18:27:00 Join dfkt_ [0] (i=dfkt@chello062178002170.1.11.univie.teleweb.at)
18:28:21ivanrocks321i think it was topik that provided the link before, when i extract it in the .rockbox folder there is only on folder. How do i get all the other folders, etc?
18:29:05krazykityou don't extract it in the .rockbox directory, you extract it to the root directory
18:29:42krazykitwhich will create something like /target/.rockbox. the former (if i understand you correctly) would give you /target/.rockbox/.rockbox
18:29:56Jaykayjust extract the zip to your player, and everything is fine
18:30:26 Quit maruk ("Leaving.")
18:31:44ivanrocks321i did that, but im still getting the problem with it not being able to play anything.
18:36:06 Join bertrik [0] (n=bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
18:38:27ivanrocks321my iriver is flashed and is running 7pre4 Bootloader and had 3.3, but started to mess up after trying to use the auto installer for 3.4
18:39:15 Join Highlander [0] (n=Highland@mek33-4-82-236-45-205.fbx.proxad.net)
18:44:53ivanrocks321Would it be easier to reinstall the original iriver firmware by using a converted hex file to bin? then redo everything with the new 3.4
18:44:58 Quit dfkt (Read error: 110 (Connection timed out))
18:45:36 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
18:48:22domonokyivanrocks321: if you flashed the iriver, its probably running the version in flash, and not on the disk, and the the codec/plugins wont work if they are newer.
18:50:53domonokyivanrocks321: for flashed irivers, the update procedure would be: unzip 3.4, rolo into new binary. flash new binary to flash, reboot and all should be fine.
18:51:55ivanrocks321how do i rolo into new binary?
18:52:18domonokyjust "play" the rockbox.iriver file in .rockbox
18:53:08domonokyinstead of unzipping you could ofcourse also use rbutil (which does the same for rockbox installation)
18:53:30ivanrocks321ok, i have rolo'ed what is the next step
18:53:52domonokynow the codecs should work.. try with a mp3 file.
18:53:53ivanrocks321thanks for the help btw
18:54:03 Quit dfkt_ (Read error: 145 (Connection timed out))
18:54:14ivanrocks321would flac be good too?
18:54:35domonokyto put this new binary into flash, you need to flash it (i think there is a plugin for it? , take a look at the wiki)
18:54:45domonokyyes, any codec should now work
18:56:12ivanrocks321is it rombox.iriver?
18:56:30domonokyhttp://www.rockbox.org/wiki/IriverFlashing under flashing Procedure
18:57:59ivanrocks321i there is no "bootloader.iriver" option under the .rockbox folder
18:58:17domonokyno need to change the bootloader.
18:58:35 Quit Lynx_ (" HydraIRC -> http://www.hydrairc.com <- The professional IRC Client :D")
18:58:52domonokylook a bit later on this page.
18:58:54ivanrocks321i do not have iriver_flash as option
18:59:14domonokyoh
18:59:38ivanrocks321maybe i need a reboot after the rolo step?
18:59:49 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
19:00
19:00:26krazykitno, rolo is a kind of rebooting
19:00:47domonokyyes, rolo is a "reboot" into the new code.
19:01:08 Join Grahack [0] (n=Grahack@ip-222.net-82-216-222.rev.numericable.fr)
19:01:08ivanrocks321humm should iriver_flash be under .rockbox folder
19:01:10domonokyhm.. iriver_flash.rock is included in 3.4
19:01:28domonokyyes, under .rockbox/rocks/viewers
19:01:29ivanrocks321maybe restart the whole thing?
19:02:21ivanrocks321wait, is it iriver_flash.rock?
19:02:26domonokythe wiki also talks about a way to change the boot order, so setting that to boot from disk, also would be a good first workaround.
19:03:11domonokyyes
19:03:27ivanrocks321i cant do that gives me an error, think its because im using a CF card not HDD
19:03:30domonokyand you should get it in the context menu of a .iriver file.
19:03:48domonokyah.
19:04:27ivanrocks321ok, what do i do with the flash_iriver? do i open something else with it?
19:04:41Unhelpfulivanrocks321: it *sounds* like you need to use it as a viewer... so go to the .iriver file's context menu, then to "open with, then flash_iriver
19:05:07Unhelpfulthough i should make it quite clear i'm only guessing and do not have an iriver :)
19:05:10domonokyivanrocks321: just follow the steps given the wikipage under Flashing Procedure (RAM):
19:05:27domonokyor (ROM) if you want to update the rom-image instead of ram-image.
19:08:14ivanrocks321ok it says sucess
19:08:59domonokythen reboot, and try if now all is fine :-)
19:09:45 Join Pingu1084 [0] (n=chatzill@host-84-223-58-135.cust-adsl.tiscali.it)
19:10:23ivanrocks321the problem is still present
19:11:14Pingu1084When I try to create a wiki page it says "Access Denied", can anyone help me?
19:11:34JaykayPingu1084: you need to register i guess
19:12:10Pingu1084Jaykay: I'm registered
19:14:05archivatorPingu1084: what's your username
19:15:20domonokyivanrocks321: and which one did yo flash ? (ram or rom version)
19:15:23 Quit ivanrocks321 ("CGI:IRC (EOF)")
19:15:31domonokyand to which is the boot order set ?
19:18:16 Quit dfkt (Read error: 110 (Connection timed out))
19:18:37 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
19:27:22 Quit dfkt_ (Read error: 110 (Connection timed out))
19:30:18 Quit amiconn (Nick collision from services.)
19:30:20 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
19:30:42 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
19:32:07 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
19:32:07 Quit pixelma (Nick collision from services.)
19:32:27 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
19:37:55 Join phanboy4 [0] (n=benji@c-24-98-43-198.hsd1.ga.comcast.net)
19:42:30 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:47:42Jaykayhi, FS #10787 and FS #10097 seem to be the same as FS #7332 and FS #8816 :)#
19:48:15 Join Hillshum [0] (n=hillshum@75-165-227-188.slkc.qwest.net)
19:48:58 Join efyx_ [0] (n=efyx@lap34-1-82-225-185-146.fbx.proxad.net)
19:49:10archivatorIf anyone finds some free time, please test FS #10065 (spectrum analyzer) :)
19:51:00Jaykayarchivator: give me a build and i will test it ;)
19:51:22archivatorJaykay: what target?
19:51:29Jaykaye200
19:51:32Jaykayv1
19:51:40 Quit mt (Read error: 113 (No route to host))
19:52:30archivatorno. 50 in the configure script, right?
19:52:35 Join kugel [0] (n=kugel@rockbox/developer/kugel)
19:52:45Jaykayi think so, yes
19:53:06Jaykayah and what should i test?
19:53:30Jaykaymaybe i cant because i don't have something you need
19:53:39 Join Emanon [0] (n=Emanon@c-67-183-42-162.hsd1.wa.comcast.net)
19:53:45 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-mdmzyzfuokcpsnbu)
19:54:08Emanonrockbox support the Creative Zen Vision M yet?
19:54:22HillshumEmanon: No
19:54:23AlexPno
19:54:30Emanonbummer thanks
19:54:32AlexPEmanon: look at www.rockbox.org
19:54:45 Quit Pingu1084 ("ChatZilla 0.9.86 [Firefox 3.5.5/20091102152451]")
19:54:47AlexPAll the players that Rockbox works on are listed on the very front page
19:54:57Emanonthanks
19:55:40 Part Emanon
19:56:33archivatorJaykay: what's that you don't have?
19:57:06Jaykayi don't know, maybe i need some special files for testing your whatever
19:57:30Jaykaye.g. it works only with monkeys audio or whatever
19:57:31 Join stripwax [0] (n=Miranda@87-194-34-169.bethere.co.uk)
19:58:13archivatorNope. You just need to play something, anything. Right now I'm at the "does it crash?" stage in testing.
19:58:25Jaykayok :)
19:59:29 Join mt [0] (n=mtee@rockbox/developer/mt)
20:00
20:01:23archivatorUnless I figure out a better way to grab PCM data, I'm considering converting the spectrum analyzer into a viewer plugin, where I'd be able to do proper analysis..
20:04:14stripwaxsaaratoga/mt - yo, do you have performance stats for the new fft/mdct?
20:04:17*stripwax is back
20:05:11***Saving seen data "./dancer.seen"
20:05:26stripwaxhm, typo on autocomplete. I meant saratoga, of course.
20:07:45*stripwax pulls the svn branch..
20:17:35 Quit stoffel (Remote closed the connection)
20:19:40 Quit FlynDice (Read error: 54 (Connection reset by peer))
20:23:08JdGordon|archivator: you really should be assuming that the devs use svn and not git if you actually want your patches looked at...
20:23:47archivatorJdGordon|: the patch applies using the standard patch tool, apparently
20:24:37archivatorI thought it didn't but gevaerts proved me wrong. So, I unintentionally didn't assume anything :)
20:24:41UnhelpfulJdGordon: patches generated with git diff apply with patch or with git apply, which according to gevaerts can apply them to non-git working trees
20:25:07JdGordon|the binary blob part also?
20:25:18archivatorNo but I uploaded that separately
20:26:06*JdGordon| shuts up :)
20:27:38gevaertsarchivator: admit it, only after I complained ;)
20:27:46 Join FlynDice [0] (n=FlynDice@c-24-19-225-90.hsd1.wa.comcast.net)
20:29:02archivatorgevaerts: I can't fight the logs! Guilty as charged!
20:34:13 Quit einhirn ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
20:34:24 Quit Highlander ("Quitte")
20:39:37 Quit evilnick ("Page closed")
20:48:13 Join stripwax__ [0] (n=Miranda@87-194-34-169.bethere.co.uk)
20:51:48 Join solexx_ [0] (n=jrschulz@e182089108.adsl.alicedsl.de)
20:57:33 Quit phanboy4 ("Leaving")
20:59:29 Part watto
21:00
21:00:51 Quit flydutch ("/* empty */")
21:03:18 Quit solexx (Read error: 110 (Connection timed out))
21:04:10 Join n1s [0] (n=n1s@rockbox/developer/n1s)
21:06:17 Quit stripwax (Read error: 110 (Connection timed out))
21:12:25 Join tomers [0] (n=chatzill@bzq-84-109-85-100.red.bezeqint.net)
21:12:31 Quit archivator ("Leaving")
21:14:16stripwax__saratoga/mt - in that branch, nothing yet uses the new mdct - will you be committing modified codecs that do use the new mdct? I just tried 'trivially' changing wma to use the new mdct, but despite vestiges of ffmpeg still existing (commented out) in wma, making libwma use the ffmpeg mdct seems completely nontrivial in fact!
21:14:54stripwax__You mentioned on the wiki page that you've tested with audio - would be great to have that committed to the branch so people can look at optimising for targets etc
21:18:16JdGordon|domonoky: do you know if the checkwps output for themes is saved anywhere?
21:18:40CIA-6New commit by 03tomers (r23991): FS #10848 - Cowon D2 manual
21:18:46 Join stoffel [0] (n=quassel@p57B4DAEB.dip.t-dialin.net)
21:18:46domonokyJdGordon|: yes, its saved in the checkwps table.
21:19:51 Quit Horscht ("Verlassend")
21:19:56 Join archivator [0] (n=archivat@77.70.28.57)
21:19:57JdGordon|awesome... would it be possible to dump any "FIXME:" lines to a website or txt file or something along with the target it was run on?
21:20:09JdGordon|only unqiue lines of course
21:20:25JdGordon|those lines means there is no debug info for that token which needs to be fixed
21:21:09domonokyshould be possible to extract that from the db.
21:21:53domonokyi just worked a bit on sorting. http://imagebin.org/75478 (it already works fine) any better ideas to allow the user the sorting ?
21:22:44JdGordon|that seems good enough :)
21:23:22 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
21:26:04domonokyabout the FIXME lines, do you just want this onetime, or everytime checkwps is run ?
21:26:53 Quit tomers (Read error: 60 (Operation timed out))
21:27:12JdGordon|no need for every time unless its cheap...
21:27:55domonokycheap would be to just run a sql query via console, and dump to text :-)
21:27:59 Join tomers [0] (n=chatzill@bzq-84-109-85-100.red.bezeqint.net)
21:28:40JdGordon|can you hook up a very simple php to do that?
21:28:58JdGordon|means I wont be able to accidently break the site
21:29:04JdGordon|or need to ask someone to do it
21:31:48archivatorQuestion - can mutexes be safely used to synchronize between 2 threads running on different cores or can they cause deadlocks?
21:32:10domonokyJdGordon|: i just see, its not stored in the db. :-/
21:32:40JdGordon|archivator: yes and yes
21:33:00archivatorJdGordon|: do you mind clarifying on the second part?
21:33:11domonokyarchivator: yes and no :-)
21:33:28JdGordon|mutexes are the standard way of doing thread/core syncronisation.. but done wrongly and they can cause deadlocks
21:33:48domonokymutexes are used to synchronise threads on cores, but if you use them wrong, they can deadlock ofcourse
21:34:03JdGordon|stop coyping me!
21:35:17archivatorHmm, I was thinking more rockbox-specific but I guess there aren't any specifics here.. My logic is simply wrong, it seems :(
21:35:20 Quit bertrik (Read error: 110 (Connection timed out))
21:35:47 Join bertrik [0] (n=bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
21:39:23CIA-6New commit by 03Domonoky (r23992): allow sorting of themes.
21:39:28 Quit tomers ("ChatZilla 0.9.85 [Firefox 3.5.5/20091109125225]")
21:40:48 Quit bmbl ("Bye!")
21:41:44CIA-6New commit by 03Domonoky (r23993): fix default for theme sorting.
21:42:47JdGordon|domonoky: sort by name?
21:43:11domonokygood idea. new sorting options are now easy.
21:43:20domonokythemename and author ?
21:43:44JdGordon|themename, but yeah, author also maybe
21:44:28JdGordon|also, I'd maybe get rid of the ascending/descending options.. always use descending and then have a seperate link to flip it?
21:44:32 Join Utchybann [0] (n=lolo@ede67-1-81-56-102-26.fbx.proxad.net)
21:44:35JdGordon|so the drop list isnt too big
21:45:13gevaertsarchivator: rb->thread_exit(); with the mutex held doesn't seem like a good idea
21:46:03archivatorgevaerts: that's not the problem I'm tackling right now, and no, it's not a particularly good idea
21:46:20*gevaerts would also not do if(sonething){code;continue;}else{something else;}
21:46:25gevaertsWhy the else block?
21:46:39archivatorYou almost instantly hit a deadlock on dual core targets (with IF_COP(,COP))..
21:46:48archivatorgevaerts: remnants of an older version
21:47:32archivatorCode style is not what's bothering me. My inability to solve a simple problem, on the other hand..
21:48:01gevaertsit's not a problem. I noticed because it tricked me into believing you had a double unlock
21:48:08n1sgood code style can be helpful for finding problems
21:48:54gevaertsI would suspect that part of the problems might be due to cache issues
21:49:11archivatorgevaerts: and in fact, the comment in that if() is entirely wrong. You'll never get discontinuities anymore because fft_idx is reset on every new buffer. If you insert those discontinuities, you'll get horrible graphs.
21:50:47gevaertsIf input_thread_has_data (or similar variables) happens to be put in a register, you're stuck. It needs to be volatile at least, but it may also require a cache flush (ask PP experts)
21:51:23archivatorThat's something I hadn't considered.
21:52:10gevaertsthere might be other similar issues as well. I only know the basics :)
21:53:56n1sisn't there some kind of NOCACHE attr thingy?
21:55:07gevaertsyes, but I'm not sure if/how that works with unaligned addresses and/or other variables that happen to be nearby.
21:56:33n1si have no idea either, iirc sratoga made the mp3 codec multithreaded so maybe he knows, or if you cna track down JhMikeS
21:56:40CIA-6New commit by 03Domonoky (r23994): improve the new sorting feature.
21:57:49JdGordon|domonoky: awesome :)
22:00
22:04:06archivatorwell, the mp3 codec uses semaphores which are not exposed to plugins :-/
22:05:14***Saving seen data "./dancer.seen"
22:09:40gevaertsIf you really need them, they can be exported
22:10:26JdGordon|hang on.. how can the mp3 codec use them if they arnt being exported? or are they in the codec api and not the plugin api?
22:10:59n1suh, mpegplayer is multithreaded too i think and that's a plugin
22:11:46archivatormy bad, they are exported..
22:11:52n1sbut yeah, if you need something from core, just add it to the api
22:12:09gevaertsah yes, mpegplayer could be a useful example
22:12:36archivatorJdGordon|: that's what it looked like but I somehow missed them in plugin.h (or rather, grep missed them)
22:13:14n1smaybe Mr Someone should create a simple example plugin that demonstrates the multithread stuf
22:13:15n1sf
22:13:41archivatorI've looked into mpegplayer but only as far as thread creation is concerned (there's an undocumented flags argument to thread_create). I'll go look at the sync stuff no..
22:14:02archivator*now
22:15:27 Join froggyman [0] (n=sopgenor@pool-72-69-220-194.chi01.dsl-w.verizon.net)
22:15:57archivator.... and here kids you can see the third type of synchronization mechanism available in rockbox - the queue! Now, in this next room you'll see the remains of someone who tried to figure all this out ...
22:16:25JdGordon|n1s: multithread and simple in the same sentence?
22:17:12n1syes! (i did mention i have no idea about it :) )
22:18:02n1salthough mpegplayer is quite a beast so something simpler than that must be possible
22:19:39archivatordoesn't batterybench use a thread to stay alive after exiting it?
22:19:42 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
22:19:51archivatorI always assumed that's how it does it..
22:19:58saratogawhat are you trying to multithread?
22:20:19archivatormy spectrum analyzer. And I'm failing miserably in the dual core case.
22:20:40domonokyarchivator: yes, but thats not multithreading.. there is no communication between tasks in battery bench
22:20:45gevaertsthreads in rockbox are easy. It's the COP that makes things hard
22:20:47archivator(but that was to be expected - I'm trying the *proper* solution now)
22:20:53 Join klapaucjusz [0] (n=jch@134.157.168.130)
22:21:45 Quit Grahack ("Tu m'as vu ?")
22:21:46archivatorgevaerts: eeeexactly.
22:22:40saratogais using the cop really worthwhile here?
22:22:46 Join PaulJam [0] (i=Paule@vpn-3077.gwdg.de)
22:22:49saratogafor mp3 the cop and main CPU will have about the same load
22:23:09saratogaso i doubt you'll see a speed up, except maybe on slower formats that aren't multithreaded like AAC or APE
22:25:32saratogahow many FFTs are you doing a second anyway?
22:25:33archivatorsaratoga: Granted, the load will be very uneven (as I'm putting just the thread that *steals* PCM data on the COP and the FFT is still on the CPU) but I'm hoping to win a few cycles by having the actual drawing parallel to the data input.
22:25:47archivatorI haven't benchmarked it. Few.
22:25:48saratogathey should be quite fast
22:26:04archivatorThe problem is that I don't have enough data to do large FFTs
22:26:32archivatorI can only do 2048 samples as some targets don't have enough data in their peak buffers for a 4096-point transform :(
22:26:38saratogarockbox manages something like 5 1024pt FFTs/1MHz on PP for most of the not very fast implementations I've tried
22:26:47saratogaif you're using more then a couple mhz here you're probably not doing it right
22:27:52gevaertshm, maybe a stupid question, but is actually getting the PCM data from the COP a good idea?
22:28:02archivatorsaratoga: I think the FFT part is pretty straightforward - it's the waiting for data that really slows down the whole thing
22:28:20saratogathen do one FFT while waiting for next PCM batch
22:28:22 Quit Ypsy ("Oh noes! My ZNC obviously just shut down :(")
22:28:26saratogawaiting is the easiest thing to parallelize
22:28:35archivatorgevaerts: I didn't see any problems in my simple tests (read: no sync, just that thread running)
22:28:55gevaertsIt's put there by the main CPU, so I'd imagine that calling pcm_get_peak_buffer() from the COP might be a bit tricky
22:29:03saratogapassing data to the cop won't hurt the core but theres a decent chance you'll get corrupted data occasionally i think
22:29:20archivatorsaratoga: kiss fft does not have an inplace transform, meaning I would need yet another buffer to do what you're suggesting
22:29:21saratogaoh, well yeah calling core functions from the cop could be dangerous
22:29:45saratogause a better transform
22:29:59gevaertsI'd put the FFT on the COP. It probably has a better isolation
22:30:10gevaertsi.e. less interaction with core functions
22:30:24saratogaif the problem is just waiting for data to be ready, putting both threads on the main CPU seems like the best option
22:30:38saratogaits not like a sleeping thread benefits from running on COP
22:31:08saratogawhere did you find a non-inplace fft? all the rockbox ones are in place
22:31:25archivatorsaratoga: KISS FFT, as I said - not a rockbox one
22:31:42archivatorlibspeex still uses it, I think
22:32:59 Join Ypsy [0] (n=ypsy@geekpadawan.de)
22:34:49archivatorYou know what, you two are probably right. I should really just scrap what I'm doing and think about it for a second. So, the two threads remain on the same core. I need to hunt down a better FFT. Which would then require that I do my own post-transform transformations, since I'm using real data (and I don't really understand those). And then I can parallelize that..
22:35:03archivatorsounds like a plan (albeit a more difficult one :( )
22:36:53saratogathe real data optimization is a good idea, but its not necessary i think
22:37:09saratogaif you're only doing 5-10 FFTs a second the computation time should be negligible
22:37:10archivatorpad with 0s?
22:37:17saratogai would do that first
22:37:21Unhelpfularchivator: are you polling for fresh peak buffer data?
22:37:28archivatorUnhelpful: yes
22:37:50Unhelpfulthat's probably a big part of why you're finding "wait for data" to be such a large part of your trouble, no? ;)
22:37:57 Join petur [0] (n=peter@d54C6F9B2.access.telenet.be)
22:38:12archivatorAbsolutely. What else am I supposed to do?
22:38:17 Part froggyman
22:38:23saratogafor reference the mdct-fft algorithm does something like 80 FFTs a second plus rotations in something like 18 MHz with not very fast transforms
22:39:03saratogayou may want to look at the new mdct branch we created this weekend, it has a fast fft that we're optimizing
22:39:06Unhelpfularchivator: create a register-for-callback system for things that need to get at audio data in a reliable fashion? or at least something that can be polled very quickly?
22:39:30 Join fml [0] (n=5dd2dc5d@giant.haxx.se)
22:39:37Unhelpfulsaratoga: congrats, i didn't know any of that was in svn yet :D
22:39:51saratogasee FasterMDCT on the wiki
22:39:54archivatorsaratoga: I suppose I will. Thanks.
22:40:28archivatorUnhelpful: I was considering that (in fact, today - should be in the logs) but my kernel-fu is not up to the task.
22:40:34saratogabut really it sounds like the most immediate problem is the lack of a good way to get data, not the fft
22:40:36archivatorI'll inevitably mess something up
22:40:50archivatorsaratoga: it is and always has been.
22:41:07 Quit Jaykay ("ChatZilla 0.9.86 [Firefox 3.5.5/20091102152451]")
22:41:07saratogaif you only want 10 fps, couldn't you just poll for data, sleep 50 ms, then poll again?
22:41:48archivatorsaratoga: that's what I'm essentially doing. I get the buffer, check if it's valid and if it's not, yield(). If it is, I use it.
22:41:57saratogaand it doesn't work?
22:42:02Unhelpfularchivator: if you want to assume that only one thing will ever do this at a time, why not just create a global event queue pointer for it, initialized to null, and a function to set it? then "wait for data" becomes "wait for event" and leverages existing code to do it quickly.
22:42:22archivatorIt does but it's slow. Very slow. You *might* get ~4-5 fps.
22:42:26fmlHello. I need some assistance in fixing my svn access. I run Ubuntu 8.04 as a guest OS in VirtualBox. The host is WinXP. Rockbox development is done in Ubuntu. I can check out svn, but can't commit. How can I fix this?
22:42:44JdGordon|why cant you commit?
22:42:50JdGordon|is there an error message?
22:43:12Unhelpfuland the cost in core should be tiny, 4B for a pointer, probably on the order of a few tens of bytes for a setting function and probably a call to it on plugin exit to ensure that it is always cleaned up...
22:43:13saratogai don't see why that would be slow unless the way you're getting data has issues
22:43:34fmlJdGordon: yes. The password is not accepted. I'm not sure whether I just forgot it or whether it's a firewall problem or such.
22:43:52 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
22:44:17Unhelpfulactually, i think we have some simpler wait-for-something mechanism than an event queue...
22:44:18JdGordon|are you using the right username?
22:44:22n1sfml: ask B4gder to reset it maybe?
22:44:25archivatorUnhelpful: So what data should I be pushing to the queue? The peak buffer again? How often is that updated?
22:45:08archivatorsaratoga: note the *might* bit. It is usually far slower.
22:45:21 Quit PaulJam (Nick collision from services.)
22:45:25fmlJdGordon: is the correct user name what's in docs/COMMITTERS?
22:45:27 Join PaulJam_ [0] (i=Paule@vpn-3077.gwdg.de)
22:45:35Unhelpfularchivator: all that should go *in* the queue is "data is ready"
22:45:47JdGordon|fml: yeah
22:46:04JdGordon|but remember the svn login assumes your local username first which you might have different
22:46:05 Quit n1s ("Lämnar")
22:47:15fmlJdGordon: yes, I specified −−username on the command line
22:47:15archivatorUnhelpful: I see. But you still didn't answer my question. As long as I'm using the peak buffer, I'll be hitting a bound - the speed it's updated at. And I have no idea what that is.
22:47:37fmlB4gder: Hello. Could you please reset my svn access password?
22:47:58 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
22:48:15Unhelpfularchivator: surely you'll find out if you create a notifier mechanism for it. and if that's not good enough, you'll need to find or create some other way...
22:49:11archivatorUnhelpful: Thanks for the tips! I'll see what I can do in the next few days/week/weeks :)
22:49:48Unhelpfulwhat exactly uses the peak buffer? oscilloscope and peak-meter plugins?
22:49:53 Quit Horscht (Read error: 113 (No route to host))
22:50:49Unhelpfulperhaps not only a "data is ready" notifier but also a way to set update frequency is desirable - it could even *not* be updated when not in use... if that's not what happens already ;)
22:51:03archivatorUnhelpful: Yep. If by peak-meter plugin you mean WPS volume meter thing.
22:51:18Unhelpfularchivator: i don't ;)
22:51:29archivatorUnhelpful: well then, add that to the list
22:51:44CIA-6New commit by 03alle (r23995): Enable calendar plugin for non-rtc targets (FS #10786 by Teruaki Kawashima)
22:52:05Unhelpfulplugins->demos->vu_meter
22:52:29archivatorAh, yes, I'd forgotten about that.
22:53:52 Quit stoffel (Remote closed the connection)
22:59:33archivatorsaratoga, Unhelpful: I'll be off but thank you for all the help! I really appreciate it! (don't think I won't ask for more - this fixed point math stuff is killing me!)
23:00
23:02:00 Join Omlet [0] (i=omlet05@130.7-65-87.adsl-dyn.isp.belgacom.be)
23:02:05Unhelpfulfixed-point math is hard, let's go shopping? ;)
23:03:59archivatorHonestly? Calculus is easier than that stuff. Especially when you want to convert 64-bit math to 32-bit, maintaining linear dependence. :(
23:04:47stripwax__saratoga - hi. so I checked out the branch and built my own libwma that uses the ffmpeg mdct. Is that something that you/mt are already going to check into svn or would it make more sense for me to commit it?
23:05:36stripwax__i wasn't super-happy about exposing the MDCTContext stuff from codeclib, rather than keeping it internal, but it was just a quick hack to get wma to compile that way ...
23:06:05 Quit archivator (""No message, just hatred at BA. Lots and lots of hatred!"")
23:06:50*Unhelpful lacks the book-learning on the subject and usually just wings it with as many fractional bits as he can afford, and then checks for overflows in the integer part...
23:12:12 Nick Ypsy is now known as YPSY (n=ypsy@geekpadawan.de)
23:14:34stripwax__saratoga - but there's also some code in libwma relating to rescaling from one version of fixedpoint (used internally by ffmpeg?) to a different version (used/expected by the tremor mdct?). Since you mentioned on the wiki that you've already tested with audio −− is that patch available elsewhere?
23:17:42stripwax__also - does the new mdct work inplace or does it require that input and output buffers are physically distinct? (i'm just trying to drop it into libtremor to see what will happen..)
23:20:35 Join froggyman [0] (n=sopgenor@pool-72-69-220-194.chi01.dsl-w.verizon.net)
23:30:23kugelfml: did you wipe the old system? svn saves passwords in plain text..
23:32:04B4gderkugel: he already did his commits...
23:32:17 Join LambdaCalculus37 [0] (n=LambdaCa@rockbox/staff/LambdaCalculus37)
23:32:23kugelyea, I saw. but still
23:36:49bluebrotherkugel: svn does not necessarily save passwords in clear text. Newer versions support keyrings like gnome-keyring or windows-cryptoapi
23:37:08kugeloh, nice to know
23:37:11 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
23:37:15bluebrotherB4gder: you're aware that some manuals are missing?
23:37:23B4gderno...
23:37:25B4gderI am now!
23:37:31 Quit PaulJam_ (Read error: 145 (Connection timed out))
23:37:36bluebrotheripod mini and fmrecorder afaics
23:39:07 Join panni_ [0] (i=hannes@ip-95-222-21-143.unitymediagroup.de)
23:41:37B4gderI'll leave that for zagor, I don't see why they don't build but I can see that they're not mentioned in the build log
23:42:19bluebrotherthey don't build? My guess was that they simply aren't linked due to the changed names. But that's only a guess of course ;-)
23:45:40B4gderI just checked the fm recorder one now and it doesn't get built at all
23:47:44mc2739I think that they are not building due to targets that do not have proper manuals in the build process - see FS #10842
23:47:50 Quit domonoky (Read error: 104 (Connection reset by peer))
23:53:28stripwax__saratoga - ok, well, to answer one of my own questions - libtremor does seem to work ok with some small changes, but the new fft requires that the imdct does not transform inplace in the pcm buffer, so libtremor needs additional (wasteful?) buffer space. but as I say, it does seem to work.
23:53:56stripwax__would be great if you could add some more details and/or let me know if there is a forum thread/some other way to contact you on details
23:54:17 Quit bertrik ("De groeten")
23:54:24stripwax__I'll add my findings to the wiki anyway (even though I assume you are already very many steps ahead of me and I'm just doing something you already know about)
23:54:41 Quit Horschti ("Verlassend")
23:55:25 Quit Tomis ()

Previous day | Next day