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 2006-09-03

00:00:04Bagderyeah, that's not the proper fix to this problem
00:01:13bluebrotherBagder, red builds :(
00:01:19BagderI know
00:01:43 Quit Siimors (Read error: 104 (Connection reset by peer))
00:01:46bluebrotherjust the moment I updated my source tree. Too bad :(
00:02:37 Quit rr3tzbach (Read error: 104 (Connection reset by peer))
00:03:07Bagderworking on a fix
00:03:58infamisit's a quick hack...button.h wasn't even included at all before 19:11...
00:04:05dionoeathe cvs builds table is getting larger and larger every day ... doesn't fit on my screen now
00:05:33Bagdermore More MORE!
00:05:59TucknDaronly idling in here, but I take it from your conversation that the now's not the best time to update to the latest bleeding edge build ;)
00:06:06linuxstb_We have a long list of potential PP5020 targets now...
00:06:12 Join TrueJournals [0] (
00:06:25TrueJournalsAlright... who broke rockbox? :-p
00:07:57BHSPitMonkeyI did! Got a problem with it!
00:08:14BHSPitMonkeyand I'll do it again, and again!
00:08:16TrueJournalsI want solitaire fixed but I can't have it because someone broke rockbox :-p
00:08:41BHSPitMonkeywell, I don't have CVS privs, do I doubt I could have really done it ^^
00:08:51BHSPitMonkeywhy don't you go look to the recent CVS activity for your culprit
00:09:35TrueJournalsCVS says...
00:09:36TrueJournalsDaniel Stenberg
00:10:39dionoeaBagder: i get a "Server Error" message when trying to access the wiki's main page
00:10:46dionoeaerror code is 500
00:11:10Bagderwhat URL?
00:11:38BagderI can view that...
00:11:40TrueJournalsthat's odd... works fine for me
00:11:59*dionoea presses Ctrl+Shift+r and it works ... weird ... ctrl-r didn't
00:18:01 Quit TeaSea (Connection reset by peer)
00:19:24 Join TeaSea [0] (
00:20:13infamis500 over here
00:20:22scorcheinfamis: =)
00:20:26infamis200 now
00:20:41infamisscorche, yeah I gave in
00:22:52infamisit sucks that I have to use pscp to transfer files between virtual machines though
00:23:03infamissince filesharing doesn't work
00:23:09scorchewhy not?
00:23:20scorchesamba is enabled on the vmware image
00:24:24infamisyeah, tried; kept saying no network provider accepts the address or whatever
00:25:28infamis"no network provider accepted the given network path"
00:25:44infamisI have filesharing and all that disabled but tried enabling services but still no go
00:26:05infamisno need for filesharing since I only have 1 computer
00:26:26scorchehave you tried disabling the vmware network adapter and enabling it again?
00:26:54infamishmm...never noticed these 2 new adapters
00:26:59infamiswhy do Ihave two?
00:27:21scorcheforgot the exact reason, but i do as well
00:27:39TrueJournalsWho will fix teh Rockbox???
00:27:41 Join maara [0] (
00:28:21infamislemme see...
00:28:22maaraonly one question - is it possible to play video on roxbox on ipod video??
00:28:31 Quit secleinteer (Read error: 104 (Connection reset by peer))
00:28:41scorchemaara: yes
00:29:39infamisscorche, yeah that worked!!!
00:30:08 Join secleinteer [0] (
00:30:20maararealy? without any problem? :)
00:30:35maaraand is it able to decode divx??
00:30:40scorcheit was probably just on a different subnet/vlan/whatever to start with
00:31:25scorchemaara: it is a work in progress....atm, it is only able to decode a specific format...see the mpegplayer wiki page for more information
00:31:48 Quit lodesi ("leaving")
00:32:26 Quit secleinteer (Read error: 104 (Connection reset by peer))
00:32:33infamisyeah they all are different subnets for some reason: vmnet1 on on (which has vmware bridge enabled) on
00:32:48infamissubnet mask 24bits
00:33:39Bagderok, there's still a silly dep problem in the codec makefile but I'll leave that for now
00:33:41maarascorche: ok thx... - i am gonna to have ipod video and i want to use rockbox on it from the begin so i have so lame questions...;)
00:33:52BagderI'll just fix the borked build now
00:36:00TrueJournalsthank you
00:36:15 Join secleinteer [0] (
00:37:06 Quit linuxstb_ ("CGI:IRC (EOF)")
00:45:22bluebrotherI've just created a patch which addresses FS #5929 −− does somebody wants to look into it?
00:47:03 Join damaki [0] (
00:50:05 Quit ender` (" I spilled Spot Remover on my dog... Now he's gone.")
00:57:07 Part TrueJournals
00:58:52 Quit Kohlrabi ("Quit")
00:59:12 Quit damaki_ (Read error: 110 (Connection timed out))
01:00:41infamisbluebrother: where's the definition of fm_emphasis?
01:01:44*dionoea tells infamis about "grep"
01:02:23bluebrotherinfamis, in settings.h, structure user_settings
01:03:38bluebrotheroops, that is missing in the diff.
01:04:22infamisgrep was the first thing I tried ;)
01:05:06bluebrotherseems the clean tree I tried wasn't really clean :(
01:05:24***Saving seen data "./dancer.seen"
01:09:37bluebrotherinfamis, fixed version attached.
01:09:55bluebrothernow I tried a _really_ clean tree for building
01:11:47Davide-NYCbluebrother: whatdoes this patch do/fix?
01:12:08Davide-NYCI just did a google search and came up with data that was way beyond my comprehention
01:12:54bluebrotherradio uses this thing calles "emphasis". You just need to set it correctly, and it is different depending on your location.
01:13:13bluebrotherlike the task description says, europe uses 50us, US 75us.
01:14:08bluebrotherhaven't looked into the technical details, just how to program the tuner chip correctly ;-)
01:14:58Davide-NYCgotcha, what was it set to before? 75us or 50us?
01:15:16bluebrotherafaics 50us
01:15:45Davide-NYChmm so I should expect better FM!
01:15:53Davide-NYC(I'm in NYC)
01:16:05bluebrotherpossibly. It's not too much better but I can at least hear a difference.
01:16:25bluebrotherlike someone changing your treble value a bit.
01:21:26bluebrotheranyway, gonna get some sleep. It's gotten late here.
01:21:52 Quit bluebrother ("cu later today")
01:27:43 Part telliott_
01:33:27 Quit maara (" HydraIRC -> <- IRC has never been so good")
01:34:18 Quit infamis (Read error: 110 (Connection timed out))
01:39:37[HO]vo|tin general, would there be any interest in my recorder v2 parts? it's the main board including lcd screen, but has poor work done on the headphone/line out jacks rendering those useless
01:40:18scorcheif they were v1 parts...
01:40:38[HO]vo|tbasically i'm asking if i should save them or smash them
01:40:48[HO]vo|tsmash it or throw it away
01:40:59Soapit's akin to the RIAA curve on records, pre-emphasis on original CDs, and Doby-B noise reduction. You pre-boost the part of the signal most succeptable to degredation, and then de-boost at the playing end.
01:43:45[HO]vo|tscorche: what do you need of v1?
01:44:13scorchepretty much the whole board considering that the usb 2.0 components are not pin compatible
01:44:31[HO]vo|ti see
01:45:05[HO]vo|tdid your direct deposit go through yet?
01:45:37Soapdon't say it has. I haven't even gotten the application conformation postal-mail yet.
01:46:04scorcheit either went through yesterday, or will go through monday
01:46:21Soapnot monday, tuesday.
01:46:32scorcheoh yeah..labor day
01:46:35[HO]vo|tcool - i went to the keybank myself and opened the checking account, also applied for the credit card, and i'm putting in my direct deposit request on tuesday
01:46:55scorcheSoap: what "application confirmation"?
01:47:50Soapi applied online, you did in the branch, they still need a signature, so they are /supposed/ to send me a form to sign.
01:48:09scorchewhy didnt you just ahve them fax it?....faster
01:48:17Soapit wasn't an option.
01:48:36SoapThough that is how I set up my IRA account, so obviously a fax of a signature is good enough legally.
01:49:06[HO]vo|tmy only hope is that they don't run out of their limited quantity of ipod nanos
01:49:35scorchei know it is alright legally...that is why i suggested it =)
01:49:43Soapmy hope is that they do and start shipping 2 giggers instead, for I have a feeling the 1gigger is "going down" on the 12th.
01:50:01scorcheif possible, when you get the confirmation, call then and see if they can accept a fax of it to speed it up
01:50:11[HO]vo|tSoap: "going down" ?
01:50:26scorcheas in going down for a new model?
01:50:31SoapI'm hopeful for a nano size increase on the 12th.
01:50:37Soapyea, 2,44, and 8
01:50:41scorchei doubt they are going to sub out the 1 gig for the 2 gig
01:50:51scorchethey would most likely apologize and give you a shuffle
01:51:08Soapif the part is out of stock and the 2 arrives at the same price point....a boy can dream?
01:51:22[HO]vo|tyeah :)
01:53:44*Joey_D wants a 3g mini
01:53:52*Joey_D is dreaming
01:56:42 Quit TucknDar ()
01:56:43 Join jhMikeS [0] (
01:58:00*BHSPitMonkey kicks Joey_D out of the dreamworld
02:00:51 Join infamis [0] (
02:14:22 Quit nicox (Read error: 110 (Connection timed out))
02:16:01 Join TCK [0] (
02:32:59 Join TrueJournals [0] (
02:38:04 Quit infamis (" HydraIRC -> <- IRC with a difference")
02:44:56 Join Rudy4Pez [0] (
02:51:24 Part TrueJournals
02:56:13 Quit jhMikeS (Read error: 110 (Connection timed out))
02:58:10 Join aliask [0] (
02:59:58 Quit TeaSea (Remote closed the connection)
03:05:27***Saving seen data "./dancer.seen"
03:06:44Davide-NYCis the new recording statusbar only for SWCODEC targets?
03:14:54 Join jhMikeS [0] (
03:18:20 Join spiorf [0] (
03:19:07 Quit spiorf (Remote closed the connection)
03:19:26 Join spiorf [0] (
03:20:40Davide-NYCquestion about the ico format, what do each of the hex values represent onscreen?
03:22:15dionoeacolumns i think
03:22:41dionoeabits are read as 0 for transparent / 1 for foreground color
03:22:52dionoeaico is 7*6 pixels i think
03:23:18 Quit jaebird ("Ex-Chat")
03:35:57 Quit xNibbler (Remote closed the connection)
03:39:34 Join Nibbier [0] (
03:39:40Davide-NYCinteresting. any way to generate bitmaps from the code?
03:41:16 Join Jazz00006 [0] (
03:46:13 Join actionshrimp [0] (
03:50:50 Join nudelyn [0] (
03:54:44 Quit nudelyn (Client Quit)
03:56:03 Join JdGordon [0] (
04:03:26 Join mantono [0] (
04:11:32 Quit mantono ("Don't worry, be happy")
04:18:05 Quit spiorf ("Read error: 104 (Connection reset by peer)")
04:26:37 Quit Jazz00006 ()
04:45:05 Quit EV|Toshiba ("#evilworld")
05:05:32***Saving seen data "./dancer.seen"
05:19:18 Quit actionshrimp ("( :: NoNameScript 4.01 :: )")
05:21:29 Quit BHSPitLappy (Read error: 104 (Connection reset by peer))
05:22:54 Quit TCK (Read error: 104 (Connection reset by peer))
05:29:13 Join TCK [0] (
05:30:45BigMacanyone know how to rate a song in jclix?
05:32:57scorcheBigMac: you should know better than to ask a question about an unsupported build in here =P
05:33:09BigMaci know i know
05:33:16BigMacwhen will i learn:-)
05:33:27scorchei had hoped you did already...
05:33:55BigMacwell I did post on the forums about it so i have learned a little
05:34:03scorchei saw
05:34:08BigMacjust thought i'd ask to see if anyone would know
05:34:11scorcheand in the right thread...i was so proud!
05:34:16BigMacme to
05:34:38BigMaci was like should this go in the button mapping thread??
05:34:44BigMacthen i thought wwjd
05:34:49BigMacand found the right thread
05:34:51scorchejesus being paul
05:35:45BigMacits amazing how the wpses have progressed
05:35:52BigMacin complexity
05:36:25BigMacfrom little text ones to jblackglass
05:36:29BigMacits amazing
05:41:35 Join Drkepilogue [0] (i=478df32d@gateway/web/cgi-irc/
05:46:05 Quit Drkepilogue (Client Quit)
05:48:20 Join vo|t [0] (
05:52:41JdGordondoes anyone know if you can get the english version of a translatable string if another language is being used?
06:06:30 Quit [HO]vo|t (Read error: 110 (Connection timed out))
06:11:52 Join Drkepilogue [0] (i=478df32d@gateway/web/cgi-irc/
06:26:42Davide-NYCanybody have a minute to help a drowning icon maker?
06:31:26 Quit vo|t ("( :: NoNameScript 4.02 :: )")
07:05:36***Saving seen data "./dancer.seen"
07:16:13 Join |AhIoRoS| [0] (n=ahioros@
07:25:23 Quit |AhIoRoS| ("Abandonando, see you")
07:29:26aliaskAnyone know how to get the sim to be forwarded over SSH? (X forwarding)
07:30:34JdGordonssh server -X ?
07:31:33aliaskTo a windows server. I think I've got it set up... I have xming running and I use putty with X forwarding enabled, and I've enabled x forwarding on sshd, just wondering if I'm missing anything
07:31:57JdGordonsounds right
07:32:39 Quit Drkepilogue ("CGI:IRC (EOF)")
07:33:24aliaskOh wait, I dont have SDL :)
07:33:46JdGordonon which comp?
07:34:08aliaskServer, I can't compile the sim.
07:34:16JdGordonhaha.. dill :D
07:34:33 Quit webguest93 ("CGI:IRC (EOF)")
07:34:51aliaskWell... my name *IS* Will...
07:35:06aliaskYou could make a rhyme if you were feeling creative.
07:35:08 Join midgey34 [0] (
07:35:29JdGordonwas it you who was tokenising the wps code?
07:36:10aliaskNo... was it Mmmm?
07:36:25JdGordoni cant rmemeber :p
07:36:47aliaskI've been working hard on my crappy ID3 tag editor.
07:36:55JdGordonhows it oging?
07:37:07aliaskIt can read/write id3v1.1 tags fine.
07:37:48aliaskBut I have no idea how to prepend data to a file, so V2 tags aren't happening yet.
07:38:30 Quit Criamos ("( :: NoNameScript 4.02 :: )")
07:38:31JdGordoncant u open with with O_APPEND or something for that?
07:39:01aliaskWouldn't that stick it on the end?
07:39:10 Join [HO]vo|t [0] (
07:39:13JdGordonoh.. prepend :p i misread
07:39:54aliaskAnyway, the plugin itself needs more work to make it prettier, and I should be studying for exams anyway.
07:40:50 Join |AhIoRoS| [0] (n=ahioros@
07:45:50 Join RoC_MM [0] (
07:50:11 Quit RoC_MM (Client Quit)
07:54:59 Join webguest62 [0] (i=da26adb4@gateway/web/cgi-irc/
07:55:07 Quit |AhIoRoS| ("Abandonando, see you")
07:55:39webguest62excuse me
07:55:49webguest62i have a question with rockbox..
07:56:11webguest62my bootloader doesnt get my ipod
07:56:21webguest62but my computer does..
07:56:47scorchemind rephrasing?
07:57:07webguest62doesnt get my ipod
07:57:20scorchei heard what you said
07:57:30scorchewhat do you mean by that though?
07:57:47webguest62you know bootloeader right?
07:58:02scorcheyes...i know what the bootloader is...
07:58:21 Quit secleinteer (Read error: 104 (Connection reset by peer))
07:58:24webguest62when i try to do it
07:58:33scorchedo what?
07:58:57webguest62to excess to it..
07:59:05webguest62i mean. to install the rockbox
07:59:12scorcheinstall the bootloader?
07:59:50webguest62it come out like
08:00:07webguest62error unfortunatley yout idpod drive name could not be found
08:00:23webguest62im stuck with this for one week.
08:01:27webguest62does anyone know how to solve it..?
08:02:00 Quit Davide-NYC ("Chatzilla 0.9.75 [Firefox]")
08:02:09scorchewell, it would help if you could explain clear enough to be able to understand you =)
08:02:59scorcheis english your primary language?
08:03:09webguest62im korean..^^
08:03:26webguest62iPod Application Manager.bat
08:03:30webguest62this thing
08:03:38webguest62when i open it.
08:03:53webguest62and click "next"
08:04:03webguest62i has to recognize my ipod
08:04:10webguest62but it doesn't
08:04:22webguest62even though my itune and computer does
08:05:30webguest62do you understand now??;
08:09:27midkaywhat on earth is that?
08:09:34midkayiPod Application Manager.bat?
08:09:48webguest62do you know?
08:09:59 Join perplexity [0] (
08:10:32webguest62isn't it the program to download the rockbox?
08:12:48midkaydownload the rockbox?
08:12:49midkaywhere did you get it?
08:13:27webguest62from korean cafe..
08:13:35webguest62like an rockbox cafe..
08:13:47webguest62 iPod Application Manager.bat?
08:13:49webguest62 iPod Application Manager.bat?
08:13:49DBUGEnqueued KICK webguest62
08:13:49webguest62 iPod Application Manager.bat?
08:13:51scorchethere are rockbox cafes?!?!?
08:13:59webguest62internet cafe.
08:14:22webguest62doesnt anybody know?
08:14:40webguest62it don't can anyone tell me how to install rockbox to my ipod video?
08:14:45 Join rebo123_ [0] (
08:16:03scorchewebguest62: we do not support unauthorized methods of installation...if you want help with your program, contact the creator of the program
08:16:16scorcheregulat installation methods are available on our site
08:16:30webguest62can you tell me where i can contact with the creator?
08:16:45scorcheno...i have not heard of that program
08:17:09scorchewhat about it?
08:17:23scorchei am referring to your "iPod Application Manager.bat"
08:17:26webguest62you don;t know rockbox? is that what you said?;;
08:17:49webguest62then.. how did you install rockbox.?
08:18:00scorchethe way dictated on the site
08:18:59webguest62thank you..
08:19:17webguest62\ how can i type this.. "\"
08:20:52 Quit linuxstb (Read error: 110 (Connection timed out))
08:20:53scorchei do not know the format of korean keyboards...
08:21:09 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
08:22:03 Quit midgey34 (Read error: 110 (Connection timed out))
08:23:45 Quit webguest62 ("CGI:IRC (EOF)")
08:25:37 Quit Stas_ (Read error: 110 (Connection timed out))
08:33:08dongslol @ rockbox people pagea.
08:33:32scorchethats nice..
08:37:42dongsdoes everyone who load it get a free core2duo which is needed to render that google map overlay of 1300 people?
08:38:27scorcheare you done trolling yet?
08:38:34dongsnope, i havent started.
08:38:39dongshave you opened that page yourself sir?
08:38:46scorchei have
08:39:02midkayyou're complaining because there are people on it? and if there weren't you'd complain that it was empty.. just shut it.
08:39:02dongsand its fast and snappy?
08:39:45midkayit's not fast and snappy, it's certainly usable..
08:40:32scorcheonly takes about 12-13 seconds for it to load for me...
08:40:55scorchei have been to other sites that are much worse
08:41:42dongsha ha ha
08:42:20scorcheyou are extremely funny...
08:42:22scorchedone yet?
08:43:05 Join webguest02 [0] (i=47ec866a@gateway/web/cgi-irc/
08:43:16webguest02can someone help me out?
08:43:26webguest02i am having wicked probs with rockbox
08:43:34dongsyoure not the only one
08:43:38scorcheit would help to know what they are
08:43:43scorchedongs: shut it
08:44:13webguest02installed the newest update..all songs are poofed...still shows a full drive..but cant find any songs
08:44:25webguest02used tagcache and it updated..still wont show them
08:44:35webguest02rebooted ....still no files
08:45:01webguest02showing 25 gigs free on a 60 gig ipod
08:45:39midkayare they in the hidden iPod_Control folder? and are you in ID3 Database browse mode?
08:46:37webguest02whoa...killer..slow do i show id3 browse mode? id3 now..but still no tunes
08:47:56webguest02oh lied
08:48:03webguest02i apparently have tunes now )
08:48:11midkaythat's how you "use" tagcache.
08:48:43webguest02cool maybe i can get a decent theme on here
08:49:08midkaygood luck..
08:49:09webguest02any other good websites for themes other than the rockbox-theme site?
08:49:16midkaythe wiki and forums.. i have what i want then..thanks ) been red in the face for an hour
08:50:15 Join itsmeguys|away [0] (
08:50:56 Quit webguest02 ("CGI:IRC (EOF)")
08:52:17itsmeguys|awayi need some help
08:52:28itsmeguys|awaywhenever i create a wps, it doesnt show the background i want it to
08:53:54itsmeguys|awayit says on the screen the screen
08:54:10itsmeguys|awayand then the song
08:54:16itsmeguys|awayit is a very simple wps, i am just starting out
09:03:03 Quit gtkspert (Read error: 110 (Connection timed out))
09:04:30 Join MarcoPolo [0] (
09:05:04 Quit itsmeguys|away ("( :: NoNameScript 4.02 :: )")
09:05:38***Saving seen data "./dancer.seen"
09:07:03 Join dishnecto [0] (
09:07:15dishnectohey can someone help me
09:07:33dishnectohow can i watch videos on the rock box
09:11:23lexconvert them to the right format and watch them
09:11:39lex that's for nanos
09:11:44dishnectowitch one
09:11:53dishnectoipod format or rvf
09:12:20dishnectoi did i put them in my ipod video
09:12:43lexthen just watch them
09:12:48dishnectoit say's loading then nothing comes
09:15:12dishnectoooh its mpeg-4
09:15:17lexyep yep
09:27:32 Quit _Veseliq_ (Remote closed the connection)
09:27:56billytwowillyis there any work on supporting mpeg-4/divx?
09:28:10dishnectowait to play the video does it matter what theme u chose
09:30:14lexbillytwowilly: it's not possible to decode it without the ipod video's chip
09:30:22lexbillytwowilly: so maybe in future - only for video
09:30:32lexdishnecto: nope
09:31:45billytwowillydo you know if the ipod's chip just supports the avc part of the mpeg-4 spec, or does it do the other part too?
09:32:29lexnope :<
09:32:30billytwowillyhmm. looks like it should if it's muxed into mp4 properly..
09:32:50billytwowillyoh well.
09:34:47billytwowillyaccording to the forum there is a port of the xvid codec out for rockbox, but it's huge.
09:36:32billytwowillynear the bottom of the first page.
09:39:42dishnectoi put a mp4- video file on root it does not work
09:39:46dishnectoneed some help
09:40:06dishnectodo i need codec or somthin elzz to make it work
09:40:10billytwowillyit won't play..
09:40:14billytwowillyit's mp4.
09:40:20billytwowillyneeds to be mpeg2 stream
09:40:34billytwowillyie no audio included, just mpeg 2 video stream...
09:40:35dishnectoyeaa thats what u use to play on a ipod
09:40:36billytwowillyI think.
09:40:40billytwowillyI've been wrong before...
09:40:42dishnectomp4 format
09:40:51billytwowillyjust ask scorche;)
09:40:59CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
09:40:59*scorche ahems
09:41:16*billytwowilly thought it was scorche anyway
09:41:23*billytwowilly has a bad memory for names
09:41:32 Join _Veseliq_ [0] (n=veseliq@
09:41:37scorchei was one of the 2 that one occasion, yes
09:41:56dishnectook what do i use to format to mv2
09:42:04lexdishnecto: apple's os plays mpeg-4s, rockbox plays mpeg-2s
09:42:07billytwowillyI don't know.
09:42:25billytwowillyno sound...
09:42:42scorchedishnecto: have you looked at the mpegplayer wikipa......yeah...
09:42:50dishnectono sound then whatz the point of that
09:42:56lexchange that scale=176:128 to scale=320:240
09:43:00scorcheit is a work in progress
09:43:06dishnectoooh kk
09:43:07lexdishnecto: it started a month ago
09:43:15dishnectoi c
09:43:22linuxstbWas it that long?
09:43:47linuxstbDamn, I've been neglecting it...
09:43:58scorchei remember that it was while i was at defcon that it happened
09:46:53amiconnlinuxstb: The initial mpegplayer commit is dated 7 Aug
09:46:54billytwowillyyou went to defcon?
09:47:20billytwowillydid you unplug the batteries to all your electronics and lock them in a safe at a bank in your home time before you went?;)
09:47:54scorchehah..not quite
09:48:31billytwowillyforcibly remove all the input devices from your laptop, pda, etc?
09:49:08scorcheit isnt that just have to know how to secure your things...which if you dont, you dont deserve to be there anyway =P
09:49:23 Join nicox [0] (
09:50:33billytwowillyyou can spot the feds because the are talking into cell phones encased in condoms...
09:50:59 Quit dishnecto ()
09:51:06billytwowillyholy crap. renderman was scheduled to talk.
09:51:12billytwowillyI know that dude.
09:51:19scorcheas do i
09:51:46billytwowillyhe's from my home town. Edmonton
09:52:28billytwowillyI haven't seen him in a couple years. Not since I stopped going to elug meetings. He used to idle on #edmonton-lug on here though.
10:00:19 Join bluebrother [0] (
10:01:00 Quit matsl (Remote closed the connection)
10:12:47 Nick dan_a_away is now known as dan_a (n=dan-mirc@
10:13:18dan_alinuxstb: I see you had some success with the coprocessor yesterday :)
10:14:27 Quit nicox (Read error: 110 (Connection timed out))
10:16:36 Join lodesi [0] (
10:16:41linuxstbdan_a: Sort of... It looks like a bootloader issue.
10:17:04Bagderdan_a: cvs [update aborted]: cannot write CVS/Entries.Backup: No space left on device
10:17:20Bagderon your build server
10:23:34 Join webguest67 [0] (i=c3ec4b1b@gateway/web/cgi-irc/
10:24:43linuxstbAnyone know the maximum number of PCM samples in an MPEG audio frame?
10:25:08Bagderno clue here
10:25:24billytwowillyout of curiousity.. if I wanted to speed up the sansa e200 port who would be the best person to send a sansa e200 to?
10:26:09Bagderif I knew that, I'd send one too
10:26:12dan_aBagder: Yes, I saw that. Can you do a "du -hs" in your ~rbclient directory?
10:28:09BagderI think a problem is that the build script doesn't do make clean after a build
10:28:19dan_alinuxstb: I'll have a look at the IPL sources and see what they do differently
10:28:21Bagderso it'll gather up quite some amount of files
10:28:26Bagderdu -sh
10:28:32linuxstbdan_a: My ~rbclient directory is 956MB
10:28:50linuxstbI guess ccache doesn't help either.
10:28:57dan_aI'd allocated a 200Mb tmpfs partition
10:29:08Bagderno ccache takes some 800MB by default I think
10:29:19BagderI mean up to 800
10:29:35linuxstbdan_a: What spec is your PC?
10:29:44linuxstb(CPU & RAM)
10:30:32dan_aAthlon XP 2400+, 512Mb RAM
10:30:36dongslol, amd
10:30:38JdGordonmorning Bagder, am i right that the sim packages are not sent back to the main server?
10:30:56linuxstbdan_a: But the good news about the COP is that it seems to work fine in Rockbox itself - I've got mpegplayer decoding video on the COP without problems.
10:31:08BagderJdGordon: so if you can build sim packages fast, you're welcome! ;-)
10:31:21dan_aBagder: I'll try adding a "make veryclean" to the end of the build script, if that's alright
10:31:30Bagderdongs: amd rocks the builds
10:31:30dan_alinuxstb: Excellent!
10:31:31 Join Siku [0] (
10:31:44Bagderdan_a: you can't do that easily...
10:31:50Bagdersince the master gets the zip
10:31:53linuxstbdan_a: The problem is that I need to run Rockbox directly - i.e. instead of a bootloader.
10:31:56JdGordonBagder: do u want to try my comp again and not have any of the cross compilers in the path and see if it helps?
10:32:07JdGordon.. so it will only compile the sim
10:32:16BagderJdGordon: I can certainly add it and we can see how it runs
10:32:32BagderJdGordon: still, right?
10:35:07amiconnMy /home/rbclient dir is 1.5GB
10:35:17dan_alinuxstb: I assume that since IPL use the COP, they must know how to set it up in the bootloader, even on PP5020
10:35:35dongsoh wow
10:35:40dongssansa looks fucking hot
10:36:17linuxstbdan_a: Yes, I've looked at their code, but can't see anything different. I'm planning on trying ipodloader2 (after I change it to load Rockbox the new way) and see if that works.
10:38:41dan_aI'm off out now, but I'll have a look at the COP stuff and at the build script when I get back.
10:38:46 Join rretzbach [0] (
10:38:48 Nick dan_a is now known as dan_a_away (n=dan-mirc@
10:40:51JdGordonBagder: ok, im just makgin sure the rbclient can build the sim, and im testing how long it takes
10:41:20Bagderits just a few keystrokes for me to re-add you in the loop
10:43:19Bagderdongs: its a neat thing but hard to figure out ;-)
10:43:28dongswhat is
10:43:31Bagderthe sansa
10:43:32dongsthe player?
10:43:37dongswhats wrong with it with default firmware?
10:43:42dongsit seems to support drag n drop mp3s
10:43:49scorcheit isnt rockbox =)
10:43:55Bagderoriginal firmware is plain wrong ;-)
10:44:06dongsi bet it skips to next song in repeat1 mode hwen you click 'next track' though.
10:44:15BagderI wouldn't know
10:44:16dongsso it cant be all that wrong. are still stuck on that?
10:44:48Bagderwell its lots of gui and things
10:44:48dongswell the bug wasnt fixed so
10:44:58billytwowillydongs: the default firmware doesn't have bookmarks
10:45:04BagderI bet if you like big pictures with many colors in the interface, you'll love the sansa
10:45:19*dongs has never heard/used such feature
10:45:31*billytwowilly listens to a lot of audio books.
10:45:35scorcheBagder: is anyone working on making it viable in the sim yet?
10:45:36dongsah those.
10:45:38billytwowillybasically, a bookmark lets you save your place.
10:46:01Bagderscorche: dan_a said he had it running locally so I'm expecting him to commit those fixes
10:46:19billytwowillythen restart later when you want to.. pretty handy when an mp3 is 5 hours long.
10:46:28scorchedrat...i was thinking about doing that =)
10:46:33dongsdefinitely not anything I'd ever need.
10:47:01Bagderscorche: you could get the LCD going instead then ;-P
10:47:05dongsits a portalplayer chipset so i dont see why lunixfgts not support it
10:47:12dongsand therefor rockbox as well
10:47:27JdGordonBagder: i just did h300-sim in about the same time as ~300s
10:47:32scorcheBagder: what is involved with that?
10:48:23Bagderscorche: disassembling bootloader, reading asm, trying to poke around, rebuilding test firmwares to see...
10:48:39Bagderagain and again
10:48:53scorcheheh...that would require actually having the device i assume
10:49:04Bagderdongs: you don't know much about these embedded things, do you?
10:49:51Bagderbtw, does IPL work on anything non-ipod yet?
10:50:05JdGordonBagder: i mean 200s
10:50:09scorchenot that i know of
10:50:34amiconnBagder: I guess they won't render their name inappropriate ;)
10:50:47amiconns/won'T/don't want/
10:51:17Bagderthey should've gone "pplinux" ;-)
10:51:29scorcheBagder: happen to know of things to do if one does not have the device?
10:52:05Bagderscorche: nah, then there's basically fixing the simulator and then making sure the keymapping all over is sensible
10:52:17JdGordonamiconn: did u see my email a few days ago?
10:53:13scorchei guess i could work on keymapping..
10:53:23BagderJdGordon: so what's your host name, let's give it another test
10:53:40 Join TucknDar [0] (, but ive reformatted since then so i need the pubkey and acbuildpl script again
10:55:04dongsbillytwowilly: sounds like sandisk contracted some rockbox developers to do a port, whcih will be closed-sores, the developers were/will be paid a fair amount to keep their mouths shut, and to avoid possible stink about GPL violations. more news as it becomes available. cat loves me...another mouse for me to throw out
10:55:38 Join rr3tzbach [0] (
10:55:40billytwowillydong is hilarious
10:55:43Bagderdongs: you're always full of crap
10:55:47scorchehe was
10:55:55scorchebut recently, he has just been annoying
10:56:01dongsthats the impression i get after reading the sansa thread on rockbox forums.
10:56:09dongsjust putting two and two together.
10:56:17Bagderyeah, using your twisted mind
10:56:24scorchetogether to get 3...
10:57:36Bagderdongs: I am the guy who's in contact with SanDisk
10:57:42BagderI think I know what's going on
10:58:02Bagderor possibly they're all just voices in my head
10:58:06scorchewait a tic...he used the term "closed-sores" uncharacteristic of him
10:58:21*billytwowilly bets on insanity
10:58:26scorchei thought he loved closed source..
10:58:44billytwowillyI thought you meant he usually has open sores
10:58:49Bagderhe's just one of our channel trolls
10:59:24 Join ender` [0] (i=null@
10:59:31amiconnJdGordon: I read it now. The a-b repeat actions missing for recorder is clearly a mistake introduced with actions. A-B repeat was possible on recorders before
10:59:49amiconnI didn't notice because I never actually use a-b repeat
11:00:44amiconnThe lastbutton variables should of cause be removed where they are no longer needed
11:01:14JdGordonso then i really dont know where the dead code you suspect is...
11:05:42***Saving seen data "./dancer.seen"
11:06:33dongsBagder: so how much are they paying you to keep from making a working port? :D
11:06:37dongsor is that under NDA too
11:06:57Bagderactually, I'm thinking of banning you
11:07:01dongsi see
11:07:11dongswell, im off until something new happens with rockbox on ipod
11:07:16Bagderto stop your stupidness from showing here
11:07:38BagderI can't be much more open about the SanDisk connection
11:07:47Bagderand you know that
11:08:11JdGordonyay! ban him :D
11:09:57 Join _jhMikeS_ [0] (
11:09:57 Join mantono [0] (
11:12:35 Quit rretzbach (Read error: 110 (Connection timed out))
11:14:57 Quit Bagder ("time to say moo!")
11:21:26Slasherihehe, now it's possible to see sleeping and running tasks marked in the debug menu.. :)
11:21:32Slasherinew scheduler almost ready
11:25:03 Quit jhMikeS (Read error: 110 (Connection timed out))
11:35:55 Join bbroke [0] (
11:36:30 Join nicox [0] (
11:39:48 Join n0n0sse [0] (
11:48:56 Join spiorf [0] (
12:04:37 Join daurn|laptop [0] (n=quae@unaffiliated/daurnimator)
12:06:09 Quit spiorf (Remote closed the connection)
12:07:15 Join spiorf [0] (
12:10:04JdGordondoes anyone here have the non-lcd remote for the irivers?
12:18:09*bluebrother spotted a broken black h120 on ebay
12:19:29Slasherihmm, this new debug menu is nice. It shows initially that dircache is active, then tagcache and finally all threads except codec thread are in sleep
12:21:05 Quit spiorf (Remote closed the connection)
12:21:24preglowSlasheri: new scheduler?
12:21:54Slasheripreglow: basically, or it uses now two separate two-way linked lists for running and sleeping processes
12:22:11preglowi've been thinking about that myself
12:22:37Slasherii think will commit that as soon as i have verified it wont break everything
12:22:55Slasheriat least iriver seems to be working very fine
12:23:29amiconnSlasheri: Your thread.c change yesterday was mostly for the better. It reduced binary size. The only slightly worse change it did was to increase iram usage (now the whole struct is in iram).
12:23:30preglowhow will it affect performance/code size?
12:23:34JdGordonisnt the picth screen relativly new?? why the hec wasnt it done with the multi-screen api!!?
12:23:35 Quit daurn|laptop ("daurn|laptop was using Daurnimator's Legendary Script")
12:23:51preglowJdGordon: timid didn't get that far before vanishing
12:24:00linuxstbJdGordon: The pitch screen existed on the Archos devices
12:24:02amiconnNothing breaks in mainline builds because of this, but it might break debug builds on archos (debug builds need slightly more iram)
12:24:10Slasheriamiconn: hmm, sounds good :) that's true that iram usage was slightly increased..
12:24:15JdGordonoh ok.. i guess i know what ill fix tonight then :p
12:24:36amiconnIt's only 128 bytes, but then the whole iram on SH7034 is only 4KB...
12:24:44Slasheripreglow: performance should be little better with the new scheduler as unnecessary context switches have been eliminated. but i haven't verified that yet
12:24:53Slasheriamiconn: hmm, that's bad
12:25:00amiconnI still wonder what you're trying to achieve with a new scheduler
12:25:06Slasheriat least the new scheduler would increase iram usage even more
12:25:21 Join bluey- [0] (
12:25:28amiconnThe current scheduler is small, simple, stable and predicatble
12:25:32JdGordonamiconn: have u thought about doing a slimmed down version of rockbox for bootbox? i.e without tagcache/ab/pitch screens to get it under size again?
12:25:50amiconnJdGordon: ??
12:25:56Slasheriamiconn: priorities could be nice to add it.. that would eliminate audio skipping if background processes does not yield often enough
12:26:17JdGordonbootbox for the FM so the bin is small enough again
12:26:26Jasonis somebody working on that iriver h20 lcd bug?
12:26:34amiconnBootbox already is a slimmed-down verison
12:26:47linuxstbJason: I believe there's a patch in the patch tracker to fix it.
12:26:51amiconnIt's essentially what the bootloader is for iriver etc builds
12:27:30amiconnHmm, you perhaps mean rombox
12:27:58JdGordonyes, thats probably what i mean
12:28:04amiconnIt wouldn't make sense to trim that
12:28:05Jasonlinuxstb: where can i find that patch tracker?
12:28:49amiconnRombox was introduced to save RAM when running rockbox. If one has to sacrifice features for that, he probably wouldn't want to run it that way
12:28:51bluebrotheror have a look at the menu on the homepage −− "patches"
12:28:53Jasonthank you
12:29:56 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
12:30:05 Quit amiconn (Nick collision from services.)
12:30:05 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
12:30:15amiconnSlasheri: Other threads not yielding often enough is a design problem in those threads, not in the scheduler
12:30:45amiconnMaking the scheduler more complicated to work around that is a step in the wrong direction imo
12:30:48Jasonnothing like that there
12:30:53 Quit perplexity (Read error: 113 (No route to host))
12:31:14 Join daurn|laptop [0] (n=quae@unaffiliated/daurnimator)
12:31:53Slasheriamiconn: not excactly. For example if tagcache yields quite often and with high bitrate audio files are skipping, we would need to significantly sacrifice tagcache performance and yield even more often for audio not to skip
12:32:18Slasheriand it's not possible to know how often yielding is enough
12:32:22Slasheribecause that will vary
12:32:46JdGordonso the new schedular can kick off naughty threads?
12:33:17SlasheriJdGordon: it just wont give cpu time to threads until higher priority threads are in sleep
12:33:21amiconnPriorities will introduce a whole bunch of problems while very probably it won't solve the skipping problems
12:33:34Slasheriso yield() wouldn't call necessary all threads
12:33:36JdGordonSlasheri: ah, ok
12:33:42amiconnYou can't give the UI thread a low priority since if you do, the UI will feel sluggish
12:34:15amiconnBut the UI thread is most often the cause of skipping because of not yielding often enough
12:35:35Slasheriamiconn: hmm, i will do some testing to see if it would (not) work
12:35:40amiconnIf you introduce absolute priorities the way it sounds, you can easily starve threads, which must not happen
12:35:45 Join spiorf [0] (
12:36:34 Join tucoz [0] (n=martin@rockbox/developer/tucoz)
12:36:55preglowit's worth a try
12:37:09JdGordoncant you up the ui thread's priority if it does start getting starved?
12:38:28amiconnThe starving is more problematic for background threads. The ui thread has to be set to a high priority anyway because of the ui feeling sluggish otherwise
12:38:37Slasherihmm, in fact we need to choice between UI becoming sluggish when buffer is running out or audio skipping
12:39:02Slasherii don't think it's possible to satisfy both
12:39:10JdGordonsluggish ui would proablby be better...
12:42:05 Join perplexity [0] (
12:42:09Slasherihmm, one way to do it could be to set codec threads priority at runtime above the ui thread if buffer is getting really slow
12:42:18Slasheriand set it back when buffer is safe again
12:42:59perplexityJdGordon: still after someone with an iriver non-lcd remote ?
12:43:20JdGordonperplexity: ye, does the quick and pitch screens both work correclty?
12:43:35perplexitydunno.. let me compile cvs and test..
12:46:56amiconnSlasheri: One possible problem with priorites: Let there be 2 threads, A and B. A has a high priority, B a low priority. Thread A starts an action (e.g. via message) that needs to be executed by thread B and then yields and waits for thread B to return the result, but A doesn't call sleep() because it wants to continue immediately when the result is ready.
12:47:31amiconnWhat will happen? The scheduler won't run thread B because thread A has a higher priority and didn't go to sleep, so A will wait forever
12:47:40Slasheriamiconn: hmm, that's true indeed..
12:48:00amiconnOnly one example from the priority can of worms...
12:49:53amiconnA very similar problem arises with mutexes
12:50:03Slasheribut mutexes do sleep
12:50:12preglowabsolute priorities won't work
12:50:16preglowthat's for certain
12:50:54amiconnOk, if mutexes do sleep, they shouldn't cause that problem
12:52:08Slasheriamiconn: in fact all kernel calls sleep (if those can block)
12:53:15amiconnYes, queue_wait() does as well
12:53:36amiconnBut not all communication is done via kernel calls iirc
12:53:37JdGordonscreen->width is in chars or pixels (for non char cell screens) ?
12:55:41JdGordondamn, it looks like its in pixels..
12:55:42bluey- <- how serios is that thing :P ?
12:56:03amiconnJdGordon: How could that be in chars (for bitmap displays)?
12:56:36JdGordonfor some reason i thought you could get the width in chars to make things easier :p
12:58:38perplexityJdGordon: help :) I can get into the pitch screen with the remote play button held down, I can do what I think is the correct behaviour altering the pitch.. I can't get out again with the remote
12:59:20fahdoes iboy work on rockbox on the nano yet?
12:59:57linuxstbfah: iboy is an ipodlinux application.
13:00:46fahoh.. rockboy i meant then. sorry.
13:01:12amiconnJdGordon: You simply can't give a bitmap display's width in chars. Font sizes can vary, and then there are proportional fonts...
13:01:23JdGordonperplexity: stop doesnt do it?
13:01:35perplexityThere is no stop button on the remote..
13:01:42JdGordonamiconn: ye, i guess that makes sense...
13:01:45linuxstbfah: Then no, no Nano user has made it work yet.
13:01:54JdGordoni must have thought it was initialized again after a font change
13:02:31JdGordonperplexity: is there a mode button on the remote?
13:02:49perplexitynope.. play. ff, rew, vol+, vol-, hold
13:05:00 Join damaki_ [0] (
13:05:22 Join lee-qid [0] (
13:05:45***Saving seen data "./dancer.seen"
13:07:12JdGordonperplexity: is there only one non-lcd remote for iviers?
13:07:34 Join Lear [0] (
13:08:15perplexityJdGordon: Not sure, I only know the one I have.. came with my 3x0. I'd like to get my hands on an LCD remote, but too lazy to search e-bay :)
13:08:33bluebrotherJdGordon, afaik there is only one non-lcd remote.
13:10:04*amiconn just had that odd playback start bug happening again on h140
13:10:23JdGordonperplexity: does the quick screen work correclty?
13:10:31JdGordonapart from not being able to exit..
13:10:52perplexityJdGordon: how do I get into it ? (I've got TFM in front of me but can't find it)
13:11:08amiconnSometimes trying to start playback loads the wps, starts buffering etc but actual playback doesn't start and the wps doesn't advance, even though it's not paused
13:11:20JdGordonperplexity: hold a-b on the main unit..
13:11:33amiconnSometimes stopping and restarting helps, but this time I had to reboot
13:11:36bluebrotheramiconn, I have this behaviour after a usb connection
13:12:04bluebrotherthe audio thread shows pcm empty while the codec is filled.
13:12:06perplexityJdGordon: yes, it appears to work correctly.. I figured if I could edit it on the remote I must have been able to get into it on the remote..
13:12:23bluebrotherin my case only rebooting helps
13:12:30JdGordonno, its not setup to.. i guess not enough buttons
13:12:57perplexityJdGordon: it's pretty bereft of function actually.. but it does enough to poke it in yer pocket..
13:13:09JdGordonperplexity: all 3 funcitons on the quickscreen work?
13:13:17perplexityJdGordon: yep
13:13:20JdGordonthe code suggests that repeat mode doesnt
13:13:23JdGordonbut ok..
13:14:52JdGordonperplexity: can i send u a new keymap fie to try?
13:15:18perplexityJdGordon: sure.. if you are lucky, my dcc might even work :)
13:15:33 Quit uski (Read error: 110 (Connection timed out))
13:15:47JdGordoni dont tinhk mine does tho, email in pm?
13:18:39 Quit damaki (Read error: 110 (Connection timed out))
13:19:12LearAnnoying... The encoders break the profile builds.
13:19:44 Join _belze [0] (
13:20:06LearAdding a call to codec_init would be needed. Anyone who knows if that could cause problems? (Don't see how myself, but...)
13:20:57 Join lightyear [0] (
13:24:30JdGordonhey, sound works in the sim!
13:28:12 Join JazzBone [0] (
13:29:22daurn|laptophey all
13:30:24perplexityJdGordon: I can now exit the quick menu with a single press on play/pause. The +/- 2% temp pitch bend no longer functions and it takes a long press of play/pause to exit the pitch screen
13:31:26perplexityJdGordon: the vol+/- buttons used to pitch bend and the |< >| buttons used to +/- 2%, now the vol do nothing and the |< >| pitch bend..
13:31:49JdGordonok, fixed
13:32:19 Quit daurn|laptop (Read error: 104 (Connection reset by peer))
13:32:42 Quit belze (Read error: 113 (No route to host))
13:32:56 Join daurn|laptop [0] (n=quae@
13:34:36 Join Rondom [0] (
13:36:13 Quit MarcoPolo ("Bye !")
13:37:55 Quit _belze (Read error: 60 (Operation timed out))
13:40:31 Join kexx [0] (
13:41:28 Nick kexx is now known as kex (
13:42:49*preglow kicks flac
13:43:02aliaskTake that!
13:51:12amiconnpreglow: ?
13:51:32amiconnI thought you like how well flac is working in rockbox?
13:51:57preglowamiconn: i did, until i tried 24 bit flacs
13:52:10preglowamiconn: it struggles at 125 mhz at the worst parts in the file.
13:52:58preglowi'll try making a 64 bit wide lpc routine when i'm home again
13:53:24amiconnHmm. What's the point in playing 24bit flac anyway?
13:53:34amiconnThe audio hw only outputs 16 bits
13:53:49preglowbecause the file is 24 bit
13:53:56preglowi don't transcode my music unless i absolutely hve to
13:53:59preglowhave to
13:54:14 Join Arathis [0] (
13:54:16preglowthis isn't my own rip, of course
13:54:16*amiconn thinks storing music in 24bit is a waste
13:54:25amiconnFor production it's another thing.
13:54:38preglowit's a high quality rip of a rare vinyl, so the ripper probably wanted to make it as high quality as he could
13:54:47preglowi understand the notion
13:55:07 Join groug [0] (i=3ecd5b41@gateway/web/cgi-irc/
13:55:39amiconnpreglow: I bet it's 96kHz as well?
13:56:01preglowamiconn: nope, actually it's 44.1khz
13:56:04preglowwhich is a bit weird...
13:56:09daurn|laptopi wish all music was in 24bit - 96kHz
13:56:09 Part kex
13:56:16daurn|laptopand i had a faster net connection
13:56:21preglowi'm just fine with 16 bit 44.1khz
13:56:48grougI don't know if it's the good place to ask questions about compiling rockbox
13:56:56*daurn|laptop wonders why the warez groups still (mostly) use 128kbps 44khz
13:57:40preglowthey don't
13:57:58daurn|laptopthey don't?
13:57:58Galois128k vorbis is actually pretty good
13:58:14amiconnHmm, that reminds me
13:58:33amiconnI should do a comparison test vorbis vs. mp3 in rockbox
13:58:45preglowamiconn: you mean for quality?
13:58:49amiconnI have one album as vorbis. It sounds muddy
13:59:01preglowwhat bitrate/encoder?
13:59:15amiconn...but it has a higher bitrate than average lame −−preset standard encoding
13:59:30preglowall vorbis files should be tagged with the encoder used
13:59:33amiconnI don't know the encoder, got it as vorbis. Bitrate is ~230kbps
13:59:44 Join banan [0] (
13:59:53Slasherihmm, now running rockbox with relative priority based scheduler and testing :)
13:59:56amiconnpreglow: Hmm, how can I check that?
14:00:01 Join MarcoPolo [0] (
14:00:01Galoislame is the best of the mp3 encoders. it's hardly fair to compare the best mp3 with an unknown vorbis encoder.
14:00:06preglowamiconn: all players i know of displays it
14:00:18preglowamiconn: usually in the metadata screen
14:01:15preglowman, flac sucks for 24 bit audio
14:01:24preglowit compresses badly as well
14:01:53bluebrotheryou can find the encoder in the first bytes of the file.
14:01:54bananGalois: what vorbis encoder are you talking about ?
14:02:08 Nick banan is now known as merbanan (
14:02:10preglowGalois: there's only one vorbis encoder
14:02:16amiconnpreglow: deliplayer doesn't display that kind of info
14:02:23preglowGalois: but there is plenty of material around encoded with old versions
14:02:27Galoisaotuv beta 4 is the one that all the tests use
14:02:51GaloisI use aotuv beta 4.51
14:02:52bluey-is it true that the latest aotuv beats musepack ?
14:02:53bluebrothere.g. I have here "libVorbis I 20040629" starting at 0x7b in the file
14:03:15preglowamiconn: just hexedit the file
14:03:19preglowamiconn: it's in the tag info in the start
14:03:27Galoisthe official, vorbis encoder is aotuv beta 2
14:03:40bluey-another reason to switch in the feature from musepack to vorbis
14:03:42amiconnlibVorbis I 20030909
14:04:06preglowamiconn: that should be decent
14:04:06 Quit groug ("CGI:IRC (EOF)")
14:04:12Galoisit isn't decent
14:04:33preglowwhy not?
14:04:47Lear's Vorbis 1.0.1...
14:04:48GaloisI mean, it's probably better than any 2003-era lame version, but things have progressed a lot
14:04:55preglownot in high bitrates
14:05:05Galoisaotuv is all about high bitrates
14:05:07preglowall the aotuv stuff is low bitrate improvements
14:05:08Galoisand low bitrates
14:05:10Galoisno nonononono
14:05:30Galoissee guruboolez tests, those are 130 and 160 kbit
14:05:31preglowthe 1.0.1 encoder is what i've been using for yonks
14:05:35preglowand it sounds completely decent
14:05:54Galoisat the high bitrate margins with super picky listeners 1.0.1 loses a lot
14:06:19amiconnWell, the files I have here don't sound decent, and the bitrates are 220..230kbps
14:06:31Galoisbeta4 breaks through the roof at 130kbit, beats out even aac and musepack
14:06:32amiconnMaybe it's the original source though
14:06:39preglowamiconn: are you able to compare with the cd
14:07:01amiconnBut that's what I should do - encode a known CD and compare
14:07:03preglowif you're doing a comparison, you'll do that anyway
14:07:04Slasherihmm, seems to work well. Audio buffer stays full while background processes are fighting for cpu time
14:07:19preglowSlasheri: how's the gui feel?
14:07:39Galoisaotuv is two separate sets of tuning improvements, one for low bitrates, one for high bitrates
14:07:53Galoisit may not even be fair to call them tuning improvements anymore, since a lot of the core code has changed too
14:07:58Slasheripreglow: currently it has the highest priority, so it's fast. But i will need to try out next what happens when the ui is the cause of skipping
14:08:15*preglow hands slasheri an ipod
14:08:22amiconnSlasheri: What about usb monitoring, power management etc?
14:09:58Slasheriamiconn: those have all the same priority, and because of the relative priority scheduling (priority == highest_priority || (current_tick - cores[CURRENT_CORE].running->last_run > priority * 10)), all threads gets some cpu time
14:13:15 Quit mantono ("will compile linux kernel")
14:17:10 Quit nicox (Read error: 110 (Connection timed out))
14:18:24preglowwavpack's lossy mode is actually quite nice
14:19:11preglowbut i'm out, later
14:19:31Galoiswoah, aotuv is out of beta now
14:19:34GaloisI must have missed it
14:26:04 Join Kohlrabi [0] (
14:34:22 Quit bluey- (Remote closed the connection)
14:34:53 Quit aliask ("Chatzilla 0.9.75 [Firefox]")
14:36:12Slasheripreglow: uh.. now i would seriously want an ipod because i can't make the UI so busy that audio would skip :P
14:36:31Slasherieven vorbis Q8 now doesn't skip no matter what i do
14:36:44Slasheribefore that skipped very easily
14:37:42daurn|laptopSlasheri: buy an archos gmini 402
14:37:55daurn|laptopmaybe even a 404
14:38:07daurn|laptop - but you'd need to find an exploit for that
14:40:39Slasheripreglow: playback buffer stays full, UI does not lag, q8 vorbis plays on background, tagcache is loading database from disk to ram and playlist cache control thread reads filenames for 5000 files while i try to put some load to the UI
14:41:24 Join nls [0] (
14:41:51 Join actionshrimp [0] (
14:42:48 Join deadhead [0] (
14:43:19deadheadhi dudes, is there any devel here? :D
14:43:56daurn|laptopbut they are idle
14:45:19deadheadwell couple of days ago I talked with Bagder about a bug I had on my ipod mini 2g with last cvs versions of rockbox
14:45:27deadheadI thought it was a bug
14:46:04deadheadinstead looks like they changed something in the way preferences are stored in config files
14:46:22deadheadit has been enought to recreate the config file
14:46:36deadheadand now everything work flawlessly
14:46:43Slasheripreglow: not to mention that disk stayd spinned up for almost eight minutes but at least playback didn't skip (it would have skipped all the time normally)
14:47:58dionoeaSlasheri: i can try on an ipod while playing xobox if you want
14:48:13deadheadin the specific my problem was that I cannot hear nothing from the player even if the display was ok . It has been enough to recreate the conf file
14:48:17Slasheridionoea: hmm, interesting. i can create a patch soon
14:48:37Slasheribut first really need to simulate the ui lag somehow..
14:48:46Slasheri+i argh
14:48:48 Join infamis [0] (
14:49:23dionoeaSlasheri: add a for( i=0; i<1000000; i++); loop somewhere in the ui code :)
14:50:01 Join tvelocity [0] (n=tony@
14:50:07Slasheridionoea :D i will probably do that :P
14:50:45 Quit MarcoPolo (Remote closed the connection)
14:54:40 Quit deadhead ("Sto andando via")
15:01:28 Join MarcoPolo [0] (
15:05:46***Saving seen data "./dancer.seen"
15:06:55 Join powr-toc [0] (
15:08:41 Join barrywardell [0] (n=barrywar@
15:09:18 Join TeaSea [0] (
15:10:49 Quit barrywardell (Client Quit)
15:15:37 Quit MarcoPolo (Remote closed the connection)
15:17:51 Join Criamos [0] (
15:18:45Slasherihmm, i will create a patch about the priority scheduling to the tracker.. it needs to be tested on all targets before committing
15:20:24 Join MarcoPolo [0] (
15:21:00Slasheriin fact priority scheduling is enabled only for software codec players
15:27:58perplexitynah, just commit it and see if it breaks ;)
15:32:58 Join Febs [0] (
15:34:40dionoeaPaprica: ?
15:39:03 Join gtkspert [0] (
15:49:21Papricayes :)
15:52:08dionoeaPaprica: did you see my comments on you chopper game patch ?
15:52:26Papricamm no
16:00:28 Quit TucknDar ()
16:01:00 Quit infamis (Read error: 54 (Connection reset by peer))
16:01:55 Join infamis [0] (
16:06:00 Quit infamis (Read error: 54 (Connection reset by peer))
16:06:23 Join infamis [0] (
16:12:28 Part n0n0sse ("Leaving")
16:15:52JdGordonSlasheri: have u tested on h300?
16:16:14SlasheriJdGordon: nope, i have only h140
16:16:21Slasheribut i am now simulating ui lag
16:16:28Slasheriit seems to work well then also
16:16:39JdGordonok, if u want me to test h300 i can
16:17:00Slasheriui becomes slightly sluggish when pcm buffering overrides the ui priority when it finds that buffer is critically low
16:17:14Slasherigood, i will create a patch soon
16:17:25 Join t0mas [0] (n=Tomas@rockbox/developer/t0mas)
16:18:43dionoeaso rockbox is now going to have mutexes and thread priority ... hehe
16:19:53 Join ep0ch_ [0] (
16:21:24perplexityI guess it's not that far away from dual core support then..
16:22:56JdGordon:'( someone found open source jumpman, but its bloody nasm so i cant port it :'( is there any nasm->c convertors?
16:23:24perplexityit'd likely be quicker and easier to re-write it..
16:24:06JdGordonif i had that and captain comic i would be in heaven :D
16:25:13pillcaptain comic !!
16:28:54 Join carini [0] (n=chatzill@pdpc/supporter/active/carini)
16:28:56ep0ch_any kind devs fancy commiting the startup behaviour patch? The patch maintainer is getting fed up maintaining it:
16:29:22mirakanyone did asm on coldfire ? I am wondering how I can determine if doing MAC a,b,acc0 MAC a,b;acc0 MAC a,b,acc1 MAC a,b;acc1 can be optimised by interlacing the register, for exemple doing instead MAC a,b,acc0 MAC a,b;acc1 MAC a,b,acc0 MAC a,b;acc1
16:30:00mirakI am wondering if any parallelisation can be done
16:31:09cariniJdGordon: You around?
16:31:16JdGordoni r
16:31:35Learmirak: don't think so... you should avoid doing a move from acc to reg immediately after a mac if possible, though.
16:32:30mirakI am ogliged to do one at some point
16:32:50Learamiconn: didn't you work on fast replacement mul and/or div for gcc? did you write any such for coldfire?
16:33:22mirakbut I do a mac on acc0 acc1 acc2 acc3 then move to acc0 acc1 acc2 acc3. is that ok ?
16:33:28JdGordoncarini: i am, but not for long... wassup?
16:33:58cariniI saw your h120 remote keymap fix for the quickscreen and it looks to have undid the change that was recently made.
16:34:05cariniLooks like lcd and nonlcd are swapped?
16:34:25carinileft is rew but right is bitrate?
16:35:10Learmirak: you mean move from (as in e.g., moveclr %acc0,%d0)? Such moves stall if done immediately after a mac.
16:35:17mirakLear: what as he done ? something like replace mul d0,d1 with mac d0,d1,acc0 move acc0,d1 ?
16:35:53mirakmoveclr acc0,d1
16:35:56JdGordoncarini: i only changed it for the non-lcd remotes... i dont tihnk i effected the other remotes
16:36:06Learthat doesn't improve things... mul use the same multiplier hardware as the mac...
16:36:27cariniIt looks like the lcd and non-lcd definitions for the h1xx are swapped.
16:37:34cariniPlus, cancel was changed from MODE to ON on the lcd remote
16:37:57*JdGordon checking what ive done
16:38:44cariniI just checked it in the sim and it's definitely changed.
16:38:54JdGordonok, fixing
16:39:09carini(I used to test it in the sim.)
16:39:51mirakLear: the mac uses one cycle
16:40:08mirakLear: on words at least, and 3 on long words
16:40:35JdGordoncarini: the iriver keymap file is too bloody big which is why the mistake :p
16:40:36mirakLear: mul uses 4
16:40:56cariniJdGordon: I feel ya
16:41:15mirakLear: unless I misunderstand you can save 2 cycles by going through a mac and a move
16:41:25mirakfor words at least
16:41:32LearI'm quite sure I've read that at least on emac CPUs, the same hardware is used for mul and mac...
16:41:33cariniI only care because I almost drove off the rode the other day trying to figure out how to leave the quickscreen from the remote :-)
16:42:00mirakthere are instructions times in MCF5249UM.pdf
16:43:50mirakin a mac.l takes also one cycle. the msac.l takes 3
16:43:59LearI just looked at the user manual for the MCF5249. It doesn't include instruction timings, but state that the emac instruction set includes mulu and muls.
16:44:00 Quit damaki_ (Remote closed the connection)
16:45:07mirakyou mean as functionality or internaly ?
16:45:25JdGordoncarini: do u want to test my fix? make ssure i got it right? the h100 lcd remote works fine here
16:46:25LearWait a minute, the manual do contain instruction timings...
16:46:54mirakLear: don't know, I have read everywhere on coldfire asm tutorials how MAC was wonderfull because it did things in just 1 cycle
16:48:28LearMul is a little slower, but I have a feeling that mul is internally treated very much like mac followed by a move...
16:50:40mirakdon't know, I though it was wired diferently inside the chip
16:51:56mirakI remember having done chip wireing and pipelining in univ, but that's far.
16:52:23mirakLear: do you know what the indirect access with 3 parmaters can be used for ?
16:53:00Learnot sure what you mean there...
16:53:07mirak2.2.7 indirect access mode with scale factor
16:54:58LearE.g., index an array in a struct, where you have a pointer to the struct?
16:55:18LearNot something one uses that often perhaps...
16:55:19 Quit solarflare (Read error: 104 (Connection reset by peer))
16:55:34Slasheriplease try out :)
16:55:35 Join stripwax [0] (
16:55:45mirakLear: in fact I maybe need it but I am not sure
16:55:46Slasheriand somebody with an ipod could check out that patch also
16:55:50 Join solarflare [0] (
16:56:09JdGordonSlasheri: getting
16:56:19SlasheriJdGordon: nice :)
16:56:33JdGordonill put it on and use it tomorow and let you know how it goes
16:56:33Slasheriand please report back ;)
16:56:44Slasherigreat :)
16:56:45JdGordon1am here almost falling asleep :p
16:56:48dionoeaSlasheri: what should i check ?
16:57:01dionoeaplayback ? plugins ?
16:57:12mirakLear: I have that 8x8 word block. I need to get 4 words in the column i from index 0 to 3 and store them in register d0-d1
16:57:17Slasheridionoea: you could try if it solves audio skipping when browsing a lot with ui or something like that
16:57:35*dionoea never had audio skipping when browsing (ipod)
16:57:43dionoeamaybe i'm not using the right file formats
16:57:44mirakLear: I am wondering if I can do that in just one movem
16:57:49Slasherioh.. but then you could try if it breaks anything
16:57:54dionoeabut i have audio skipping in some plugins
16:57:58dionoeamaybe it'll fix it
16:58:01Slasheripreglow: could you try that also? =)
16:58:11Slasheriyes, it might do that
16:58:18Learmirak: sounds like it.
16:59:19mirakmovem.w (0, a0, 8,4),d0-d1 ???
16:59:34*dionoea needs to checkout a clean trunk
16:59:56mirakhum something like movem.w (0, a0, 4,8),d0-d1
17:00:33mirakit seems powerfull
17:00:48 Join secleinteer [0] (
17:01:29mirakmovem.w (0, a0, 4,16),d0-d1
17:01:33BigMacMashed Potatoes
17:02:00LearHm.. Not quite, more like "movem.w (0, a0, d2.l*4), d0-d1", but it was long ago I used it... :)
17:02:07BigMacdo i need to have tag cache running to use the rating feature in the runtime database?
17:02:31SlasheriBigMac: yep, you do
17:03:00LearSorry, "*2" if you want word indexing, and assuming d2 contains element number.
17:03:29BigMacI have my music in root
17:03:32BigMacin a folder
17:03:38BigMacwhich is way easy
17:03:52BigMacand i already have all 800 songs tagged and album arted
17:04:17 Quit carini (Remote closed the connection)
17:05:41mirakLear: that's not clear if we can use registers in it.
17:05:48***Saving seen data "./dancer.seen"
17:06:05Learmirak: it is used in the coldfire asm in the flac decoder.
17:06:37Learmirak: the mode has displacement, address register and offset register, plus optional scaling of the offset register.
17:06:42Learthat much I'm sure of.
17:07:08LearAnd you don't have to specify the displacement if it is 0.
17:10:29stripwaxBigMac - why "boo"? tagcache enabled is not the same as id3 viewing mode. and what's that got to do with your files being tagged & album arted?
17:11:16mirakLear: in wich file is it ?
17:11:47 Quit secleinteer (Read error: 104 (Connection reset by peer))
17:11:55ep0ch_Slasheri: your patch modifies apps/plugins/iriver_flash.c which doesn't exist?
17:12:09stripwaxBigMac - you can still use your folders, and album art.
17:12:13BigMacwait so If i enable cache I don't have to view in Tag Cache Id3 stripwax(I remember I had to last time I did this)
17:12:22Slasheriep0ch_: ups.. it shouldn't touch that.. but it doesn't exist?
17:12:29Slasheriit should exist in the fresh cvs
17:12:35BigMacoh so I just have to have a database on the player
17:12:40BigMacnot use it though?
17:12:46stripwaxBigMac - I think so
17:13:05stripwaxi could well be wrong -
17:13:33stripwaxtry it?
17:13:45 Quit MarcoPolo (Remote closed the connection)
17:13:58 Join ootput [0] (n=ootput@tor/regular/five0)
17:14:29ep0ch_just done cvs up -dPC and there is no iriver_flash.c in plugins....
17:14:49Slasherihmm, lets see
17:15:07 Quit spiorf (Remote closed the connection)
17:15:28Slasheriep0ch_: Repository revision: 1.4 /cvsroot/rockbox/apps/plugins/iriver_flash.c,v
17:15:31Slasherithere is
17:15:43 Join spiorf [0] (
17:15:52Slasheritry to fetch a clean cvs :)
17:16:06ep0ch_yes its been a while :)
17:16:12 Join MarcoPolo [0] (
17:16:47dionoeaSlasheri: haha, i still have audio skipping when playing xobox (a lot less than before though) on ipod video
17:17:21dionoeawell it skipped only once
17:17:24dionoeacan't reproduce
17:18:36dionoealaunching the fireplugin completely forze the audio thread it seems
17:18:50 Quit JdGordon ("Konversation terminated!")
17:18:59stripwaxdionoea - I think fire just doesn't yield as much as it should
17:20:41 Part JazzBone
17:23:40Slasheridionoea: it seems fire doesn't yield at all
17:23:49Slasheriso scheduling cannot fix that
17:23:54dionoeait yields in the button loop
17:25:01dionoeabut overall the patch does reduce skipping
17:25:06dionoeawhich is nice
17:25:23 Join bluey- [0] (
17:25:42dionoeawoops, my ipod froze in mandelbrot
17:27:22dionoeait unfroze for like 1/2 sec ... still waiting. Does the sleep/wake code make sure that all threads get waken up at sometime ?
17:28:17Slasheridionoea: yes
17:29:07Slasheribut it seems there are still some problems with the patch
17:32:13 Join CriamosAndy [0] (
17:34:21 Part ootput
17:35:03 Join secleinteer [0] (
17:35:53 Join sharpe [0] (
17:43:42 Join TeaSeaLancs [0] (
17:46:26ep0ch_Slasheri: is it just me or is track buffering realllllllyyyly slow while audio is playing with the scheduling patch?
17:48:09 Quit bluey- (Read error: 60 (Operation timed out))
17:49:15ep0ch_and my player doesnt want to shut down hehe
17:49:21 Quit spiorf (Remote closed the connection)
17:50:13 Quit Criamos (Read error: 110 (Connection timed out))
17:50:26 Quit perplexity (Read error: 113 (No route to host))
17:52:47 Join spiorf [0] (
17:53:51ep0ch_but boost ratio seemed to be a good 3% lower
17:54:21 Join webguest83 [0] (i=3e21675b@gateway/web/cgi-irc/
17:55:40 Quit webguest83 (Client Quit)
17:57:20 Quit TeaSea (Read error: 113 (No route to host))
17:59:48 Quit apo (Read error: 110 (Connection timed out))
18:01:16 Join apo [0] (
18:03:19 Quit bbroke ("Chatzilla 0.9.75 [Firefox]")
18:13:08 Quit apo (Remote closed the connection)
18:13:20 Join apo [0] (
18:13:41 Join toed [0] (
18:13:55infamisdionoea, fire always freezes playback on mine
18:14:21toedis there a patch or something that lets me seek like with the original ipod firmware?
18:14:24infamisoops, a little late
18:14:30toedi.e. center then scroll
18:16:43nlsSlasheri, Tagcache question: on archos the auto update will never detect deleted files as there is no Dircache?
18:17:02 Quit Nibbier (Remote closed the connection)
18:17:07Soaptoed - no patch, but there is an older feature request for this.
18:18:12toedah, well I'm sure it'll appear with time, thanks
18:18:41Slasherinls: true, that would be too slow
18:19:53nlsok thanks, trying to make the tagcache chapter in the manual a little bit more correct.
18:20:30nlsIs the increasing of max files in browser still needed to see the end of a very long list.
18:21:21Slasherinls: that depends, for example showing complex entries containing sorting/entry count limiting that is necessary
18:21:32Slasheriwithout it either an error message is displayed or items are not sorted
18:22:50nlsok I will leave that note in.
18:24:06 Join Nibbier [0] (
18:27:13 Join Lynx [0] (
18:27:24 Join belze [0] (
18:28:15 Quit Lynx_ (Nick collision from services.)
18:28:41 Nick Lynx is now known as Lynx_ (
18:31:46 Join jorian [0] (
18:33:50 Quit Nibbier (Remote closed the connection)
18:35:32amiconndionoea: around?
18:36:08 Quit ep0ch_ (Read error: 110 (Connection timed out))
18:37:09dionoeaamiconn: ?
18:37:29dionoeanew solitaire bugs ? :)
18:37:48amiconnNot new bugs, no
18:38:09Slasheridionoea: ah! fire no longer halts playback :)
18:38:12Slasheriit just slows down :D
18:38:18 Quit thegeek ("( :: NoNameScript 4.03 :: )")
18:38:23dionoeaSlasheri: did you make it yield more often ?
18:38:30Slasherino, i fixed the scheduling =)
18:38:36amiconnThe invisible card is fixed, so playability is restored :)
18:38:42Slasherihmm, i will upload a new patch soon..
18:38:47amiconnThere are some graphical glitches though
18:38:54Slasheriamiconn: could you try that patch to see if it breaks archos?
18:39:01Slasheri(will upload it shortly)
18:39:02dionoeaSlasheri: i won't have time to test the new patch today. Tomorrow maybe
18:39:06Slasheriok :)
18:39:07dionoeaamiconn: where ?
18:39:27dionoeaon greyscale targets ?
18:39:40amiconnTarget independent
18:39:58amiconn(1) The bug I already described yesterday: If the draw stack becomes completely empty, it reappears. Slightly irritating
18:40:20dionoeaow ... indeed. I'll fix that ASAP
18:40:53Slasheriand mandelbrot works nice also, it just draws the screen slower when buffer is critical (UI priority overridden by the codec)
18:41:59dionoeaamiconn: what's (2) ?
18:42:56amiconn(2) When the cursor is placed on the remains' stack when there is no card, and then a card is drawn, the cursor disappears
18:44:00 Join Nibbier [0] (
18:44:30dionoeahum ... that shouldn't happen
18:45:35amiconn(3) (just discovered) Sometimes it's impossible to pick a card from the column. Couldn't reproduce it yet...
18:46:34amiconn(4) (easy one) The help texts refer to drawing 3 new cards, but solitaire now has selectable draw one / draw three
18:46:49dionoeai'm not fixing (4). drawing 1 card is cheating :)
18:47:02dionoeatoo easy
18:47:11*amiconn always plays solitaire in draw one mode
18:47:38*dionoea never would even think about using draw one mode
18:48:31Slasheriamiconn: please try out :) priority based scheduling is enabled only for software codec targets
18:48:41 Join perplexity [0] (
18:49:02Slasheriamiconn: and the improved debug menu now displays running tasks marked with * and priorities
18:49:09amiconnSlasheri: Another plugin causing high UI load is cube.rock when enabling "turbo mode"
18:49:31Slasheriamiconn: great, i will try that. How can i enabled turbo mode?
18:49:49amiconnJust press Navi, and you'll see...
18:50:18Slasherinice, lets see.. i will start playing a q8 vorbis on background :)
18:50:19amiconnIt just yields once between 2 frames instead of sleeping (in normal mode)
18:50:46amiconnAll these high-load plugins don't influence playback on archos
18:51:22dionoea(1) should be fixed (untested)
18:51:36dionoea(2) should be fixed too (hoping that i didn't trigger other bugs)
18:51:37amiconnThat's because the mpeg thread has very little work to do. Just rebuffer every now and then, and process user requests like skip, pause etc
18:51:54dionoea(3) can't reproduce -> wont fix
18:52:01dionoea(4) ... ok, i'll fix it
18:52:05amiconnFeeding the MAS is done via DMA, and an ISR sets up the next DMA block on request
18:52:17 Join bbroke [0] (
18:52:26amiconnYeah, fixing (4) is just replacing "3 new cards" with "new card(s)"
18:52:32 Join |AhIoRoS| [0] (n=ahioros@
18:52:53Slasheriamiconn: works great, no skipping at all :) but cube skips a lot :D
18:53:00dionoeaand next i need to code a flipper for rockbox
18:53:21Slasheriamiconn: at first it spins very fast, then it slows down and only spins a few times in second
18:54:05Slasheriand even that varies on how hard the current position of the vorbis is to decode
18:54:29Slasheriwhen there is just silence, it spins faster ;)
18:54:53amiconndionoea: Btw, I discovered that the final stacks are reachable with the cursor (up/down while on the remains' stack, on X5). I guess it's on purpose, but it isn't obvious from the help texts
18:54:57Slasheriand without turbo spinning, it spins smoothly
18:55:17amiconnIn normal mode, cube does a sleep(4) between 2 frames iirc
18:55:33dionoeaamiconn: that's on purpose... and the key mapping was better with the old archos layout
18:56:01amiconnYes, with the draw stack, remains stack and final stacks on the right
18:56:51amiconnRemapping that to left/right would be more intuitive, but less convenient when mvoing around for normal actions
18:57:11dionoeathat's exactly what i though
18:57:13amiconnIt's not that one needs to reach the final stacks often
18:57:25dionoeaonce in every 50 games i'd say
18:58:12amiconnPerhaps it could be documented in the wiki / manual.
18:58:18dionoeait could :)
18:58:22amiconnHmm, in fact it should
18:58:40amiconnI wonder how the old layout would look on the larger LCDs
18:58:58dionoeathe new layout is better
18:59:11dionoeawith the old layout you couldn't see the 3 cards that were drawn
18:59:15amiconnThe old layout had more space for long columns
18:59:16dionoeaonly the upper one showed
18:59:26amiconnAh, yes
18:59:50perplexitySlasheri: have you a link to the current sched patch please ?
19:00:01amiconnThe margins could be tweaked a bit more on the larger LCDs
19:01:33dionoeathey could. I'll let someone else do that
19:01:59amiconnCurrently the columns block is left adjusted. There is space between each 2 columns, but there's no left margin, only right margin
19:02:24amiconnLikewise, the draw stack does have a left margin, but the final stacks stick to the right
19:02:36dionoeadidn't i change that earlier today ?
19:02:51amiconnHmm, perhaps I have a version from before that running here...
19:03:59infamisperplexity: ?
19:04:34amiconnAnother thing that can be improved are the large suits shown on the empty final stacks. Archos doesn't use larger versions at all, and the larger versions on colour targets are looking strange
19:04:45amiconnI know whom to ask about this one :)
19:05:02 Quit Nibbier (Remote closed the connection)
19:05:18 Join Nibbier [0] (
19:05:36dionoeaamiconn: strange ? they look fine to me :) (seriously)
19:05:48 Quit Nibbier (Read error: 104 (Connection reset by peer))
19:05:51***Saving seen data "./dancer.seen"
19:06:00amiconndionoea: On which target?
19:06:12amiconnThey are looking strange on H340 and X5 to me
19:06:15dionoeai tested on recorder / h1x0 and h3x0
19:07:24dionoeadid you forget to run make clean a long time ago ? (like 3 weeks ago or something)
19:07:44amiconnI did a full rebuild including reconfiguring today
19:08:33amiconnHeart and Spades are looking fine, but the diamond looks squeezed horizontally, and Club looks completely strange
19:08:50dionoeadoes this look strange ?
19:09:05dionoeahmm ... ok
19:09:41infamisahh...we got a 'make install' for the sim now...
19:10:00dionoeado we have 1000x1000 LCDs to play solitaire ? :)
19:10:37dionoeawell they have smaller versions in the page. I'll let you change those if you want.
19:12:18amiconnOf course we don't have such large displays, but then the small versions shown on the cards are looking better than the large ones imho
19:12:44ghode|afkdionoea: you mentioned before that you don't get any skipping with the ipod 5g, is this even with using the scroll wheel a lot?
19:13:07dionoeaghode|afk: i don't have very long lists ... which might explain why i don't get skipping
19:13:12 Join Rob2222 [0] (
19:13:49dionoeai don't know if the fact that i use vorbis and not mp3 most of the time also changes anything
19:14:11ghode|afkah ok
19:14:51ghode|afki'm using the itunes folder structure, so in tagcache the lists are huge... about 4-5 mins to scroll from one end to the other
19:14:54amiconnSlasheri: If you change the plugin api in a backwards incompatible way, you are obliged to do two things:
19:15:16ghode|afkthis pretty much empties the audiobuffer and i get loads of skipping/tracking skipping
19:15:28amiconn(1) Bump the MIN_API_VERSION to be the same as the API_VERSION.
19:15:37 Join thegeek [0] (
19:15:42 Join lorijho [0] (i=536334d5@gateway/web/cgi-irc/
19:16:06dionoea(2) is far worse :)
19:16:22amiconn(2) Since the api isn't backwards compatible anyway, sort the api to restore proper grouping.
19:17:01amiconnI.e. put the functions that were added at the end since the last incompatible change into the group where they belong (lcd to lcd, sound to sound etc)
19:19:15amiconnSome more remarks:
19:19:46Slasheriah, yes. i will do that :)
19:20:07amiconn(1) Why did you change the sleep_thread() mechanism? With cvs, sleep_thread() can take less than one tick, since there are other interrupts than just the timer
19:21:13Slasheriamiconn: i did that because i wanted to move sleep tick counting to the scheduling and prevent unnecessary task switching
19:21:14amiconn(2) Are you sure that taking out that bunch of wake_up_thread() won't break the threading? It can easily leave the scheduler in sleeping stae
19:21:24Slasherido you mean that sleep_thread(0) should work too?
19:21:31amiconnSlasheri: Sleep isn't always in ticks, that's on purpose
19:21:53amiconnSometimes it's enough to just sleep until the next interrupt
19:22:01Slasheriamiconn: no, the scheduler works now in different way and it no longer requires wake_up_thread
19:22:19Slasheriamiconn: ah.. then sleep_thread(0) should do the old behaviour
19:22:50amiconn(3) You're carrying around the priority argument on archos even if it's unused
19:23:01 Quit Rob2222 (Read error: 104 (Connection reset by peer))
19:23:09 Quit perplexity (Read error: 110 (Connection timed out))
19:23:15amiconnMaybe there's a macro somewhere; didn't read the whole patch yet
19:23:51Slasheritrue.. i just thought the code might be more readable without defining a bunch of #ifdefs. But now as you mentioned it, defining a macro sounds like a better solution
19:23:54 Quit daurnimator (Read error: 110 (Connection timed out))
19:23:54amiconn(4) In battery_bench.c you're using a literal priority. Is that on purpose?
19:24:07amiconncreate_thread() could be a macro
19:24:14Slasheriups, no. i will fix that also
19:24:20 Quit daurn|laptop (Connection timed out)
19:26:44Slasheriamiconn: btw, did you already tested it? :)
19:26:52amiconnHmm, making the whole function call a macro is a nice solution in a totally different place as well :)
19:27:04Slasherihmm :)
19:27:41 Join daurn|laptop [0] (n=quae@
19:27:56 Join daurnimator [0] (n=daurnima@unaffiliated/daurnimator)
19:29:00 Join San [0] (n=San@
19:29:28SanI got a 2200mAh battery for the H300 today
19:29:40San:) I'm gonna install it tomorrow
19:30:44 Quit Rob2222_ (Read error: 110 (Connection timed out))
19:31:38infamisis it possible to send logf output over usb in real-time, say to a host computer?
19:31:42amiconnSlasheri: Your last commit causes YELLOW
19:32:17amiconninfamis: (1) Only on targets with USBOTG (2) if you write the necessary USB stack first
19:33:09infamisso it is possible, ok, I'll do some research then
19:33:35infamisI just need to see logf output in realtime
19:33:42amiconnIt's impossible atm, and depending on your target might be never possible
19:34:00Slasheriamiconn: ah, fixing
19:34:02amiconnX5 will be HARD
19:34:11amiconnThe USBOTG chip is undocumented
19:34:22infamislol...I give up before I even start then
19:34:41infamisoh well, back to just dumping logf to disk every 20 calls
19:34:48amiconnThe first target that is likely to see USBOTG support will be the H300, since it uses a documented USBOTG chip
19:35:15amiconnIt's still required to write the USB stack, which is a major task
19:35:35infamisany RE'ing being done on the Cowon firmware yet tha tyou know of?
19:35:47infamisregarding USBOTG
19:36:17amiconnWhen that is working, it _might_ be possible to implement USBOTG on X5 and ipod by reverse engineering
19:38:07 Join rotator [0] (n=e@rockbox/developer/rotator)
19:38:55 Quit MarcoPolo (Remote closed the connection)
19:39:41infamisis there any other approach I can use to get logf output?
19:39:48infamisw/o an lcd
19:39:55infamisI mean remote
19:40:09*|AhIoRoS| is away:
19:45:02 Quit Arathis (Read error: 110 (Connection timed out))
19:45:50 Join Arathis [0] (
19:45:59amiconninfamis: Dumping to disk or viewing in the debug menu
19:46:15Slasheriamiconn: hehe, that plugin api hasn't been sorted for some time it seems :D but i am almost finished with that now ;)
19:46:19amiconnYOu don't need to dump every 20 calls. The logf buffer is 1000 lines iirc
19:46:59amiconnSlasheri: It wasn't sorted since the last MIN_API_VERSION bump
19:48:09amiconnThat's how it works. During normal extension we are keeping backwards compatibility for 3rd party plugins by adding entries at the end and only bumping API_VERSION
19:48:46amiconnBut if the api is changed in an incompatible way, we have to bump MIN_API_VERSION, and that gives us the chance to sort the api
19:49:04amiconnThe create_thread() parameter change is backwards incompatible
19:49:11Slasheriyep, true
19:50:11dionoeaexcept on targets which don't need the change :)
19:50:30amiconnYeah, but do you really want to split versions?
19:50:59 Join matsl [0] (
19:51:00amiconndionoea: The new margins are better
19:51:43infamisamiconn, but dumping every X calls is not if there's an infinite loop I'll miss output
19:52:04amiconndionoea: Hmm, the help function is wrong on H300
19:52:06dionoeacan't you dump in the logf( ) call ?
19:52:38infamisyeah...that's what I'm doing now, but is bad IMO & not effective as I need it to be unless I dump every single logf call
19:52:41Slasheriamiconn: btw, do you think it should be ok to commit that new scheduler when problems have been fixed?
19:52:58dionoeaamiconn: the help thing needs to be changed anyways ... asking the user to find combos himself isn't very nice
19:53:09amiconnSlasheri: I still need to test
19:53:27Slasheriyes, testing first of course
19:55:07amiconnMy main concerns are: (1) It should work stable even in rare conditions. (2) The CPU sleep should work properly. It saves _a lot_ on archos, and is the reason why only rockbox reaches the battery runtime archos promises for the Ondio. (3) It shouldn't increase binary size on archos
19:56:04Slasheriamiconn: i think cpu sleep should work now better with the new scheduler and it should save power even more
19:56:29Slasheriyes, at least in theory.. but i haven't done any real testing with that
19:57:17Slasheribut binary size will probably increase a bit
19:57:45infamiseven dumping logf lines still have to view the file either off-device or in the text-editor
19:58:08infamisnot effective when trying to debug something that crashes the player
19:59:29*|AhIoRoS| is away:
19:59:43*|AhIoRoS| is away:
20:00:01*|AhIoRoS| is away:
20:00:02amiconn|AhIoRoS|: Please stop that
20:01:13amiconndionoea: There is no main help text at all on both X5 and H300, and on H300 the A-B buttons says REC: in its help text
20:01:16|AhIoRoS|amiconn: so sorry, i remove away option :-)
20:05:57 Join Bagder [0] (
20:10:23 Join Paul_The_Nerd [0] (
20:11:56 Quit Bagder (Read error: 104 (Connection reset by peer))
20:13:44 Join totoetlititi [0] (i=81664504@gateway/web/cgi-irc/
20:13:55totoetlititihi all!
20:14:35totoetlititiI have a big problem to cross compile in my mac using UIsimulator...
20:14:36lodesioops.. sry
20:15:00totoetlititiI always have the fatal message:"Unsupported system: Darwin, fix configure and retry"
20:15:52infamisyou don't crosscompile for the sim...
20:16:14totoetlititiI can't compile with the UIsimulator...
20:16:22totoetlititibut I can cross compile for my ipod !
20:16:41infamisthe sim runs on your computer's native architecture
20:17:08 Quit |AhIoRoS| ("Retarting X")
20:17:55 Quit secleinteer (Read error: 104 (Connection reset by peer))
20:18:27totoetlititiin the tools/configure I had a "case $uname in CYGWIN), Linux),FreeBSD), and darwin)
20:18:32Sanwould it be ok to charge a new battery for the first time through USB?
20:19:12totoetlititibut now in the new tools/configure I have juste for Cygwin, linux and Freebsd... nothing for Darwin, why ?
20:20:57 Join |AhIoRoS| [0] (n=ahioros@
20:20:58amiconnHmm, seems Bagder overlooked it when changing configure
20:21:04infamismaybe it's unsupported now?
20:21:17infamissomebody else with the same prob:
20:21:22 Join secleinteer [0] (
20:21:26Slasheriamiconn: hmm, i tried to declare create_thread as a macro but that will create a problem with plugins
20:22:24 Quit totoetlititi ("CGI:IRC (EOF)")
20:22:46Slasheriunless we use something like PRIORITY(PRIORITY_SYSTEM) as the last parameter everywhere (without the last ,)
20:22:57Slasheribut i am not sure if that's pretty
20:23:56 Join totoetlititi [0] (i=81664504@gateway/web/cgi-irc/
20:24:05amiconnThe macro doesn't get evaluated when used like rb->CREATE_THREAD(blah, blubb, priority);
20:24:21Slasherihmm, just a moment
20:24:59Slasheriplugin.c:267: error: `create_thread' undeclared here (not in a function)
20:25:18Slasheribecause macro can't be a function pointer
20:25:46 Quit totoetlititi (Client Quit)
20:25:46 Join totoetli1iti [0] (
20:25:57amiconnHmm, it's not chained...
20:26:17 Join barrywardell [0] (n=barrywar@
20:26:18amiconnIt would mean that we'd need another CREATE_THREAD macro for plugins
20:26:55amiconn...that would be used like that: CREATE_THREAD(rb, blah, blubb, priority)
20:26:57Slasheriah, yes.. now i got an idea
20:27:58amiconnWe could also use an IF_MV2() like mechanism
20:28:09amiconn...but that's looking rather ugly
20:28:18Slasheriyep, i was thinking about that also :)
20:28:52amiconnThe PRIORITY macro can take the comma inside its parameter list
20:29:27Slasheriah, true. i will change that
20:29:35amiconnThat's why the multivolume macro is called IF_MV2(). It takes 2 arguments, the second being an emty argument created by using the comma
20:29:52 Quit lee-qid ("Trillian (")
20:30:11amiconnIF_PRIO(, PRIORITY_SYSTEM) would be the equivalent
20:31:10Slasherihmm, that sounds nice
20:31:20Slasheriat least i could use that in the function declarations
20:31:34Slasheriand the macro when calling functions
20:31:49amiconnOr also just the IF_PRIO macro
20:39:30 Quit lightyear (Read error: 110 (Connection timed out))
20:39:41 Join carini [0] (n=chatzill@pdpc/supporter/active/carini)
20:39:55 Quit carini (Client Quit)
20:39:55 Join lightyear [0] (
20:42:25 Join Rob2222 [0] (
20:43:00 Quit jorian ("Leaving")
20:48:48 Quit lorijho ("CGI:IRC (EOF)")
20:59:57 Quit spiorf (Read error: 54 (Connection reset by peer))
21:01:34totoetli1itihi, when I compile rockbox for UIsimulator, I have an error with "LD rockboxui":
21:01:50totoetli1iti"/usr/bin/ld: ~/rockbox-devel/buil-dir/apps/lang.o bad magic number (not a Mach-O file)"
21:01:57totoetli1itiWhy ?
21:02:02totoetli1itiWhat does it mean ?
21:02:39 Quit merbanan (Remote closed the connection)
21:04:08 Join scott666 [0] (
21:04:14 Part scott666
21:05:48 Join TeaSea [0] (
21:05:52***Saving seen data "./dancer.seen"
21:06:24 Join carini [0] (n=chatzill@pdpc/supporter/active/carini)
21:09:49 Quit lodesi ("leaving")
21:10:23 Quit TeaSeaLancs (Read error: 60 (Operation timed out))
21:13:14 Part Paul_The_Nerd
21:13:49 Quit infamis (Read error: 104 (Connection reset by peer))
21:14:19 Join infamis [0] (
21:14:27 Quit powr-toc (Remote closed the connection)
21:18:30 Quit bbroke ("Chatzilla 0.9.75 [Firefox]")
21:18:50 Quit TeaSea (Remote closed the connection)
21:19:05 Quit totoetli1iti ()
21:19:56 Join TeaSea [0] (
21:25:38linuxstbdan_a_away: I've just tried a sim build with your COP patch, and create_thread_on_core() doesn't seem to be implemented.
21:26:08mirakah !
21:26:17mirakI finally managed to do the idct_row
21:26:21mirakin asm
21:26:36mirakamiconn: I have some doubts about the registers beeing cleared
21:26:45miraktha ACCX
21:31:17 Quit _Veseliq_ (Remote closed the connection)
21:32:30 Join _Veseliq_ [0] (n=veseliq@
21:34:37*linuxstb doesn't like it when things work in the sim but not on the target
21:35:05sharpelinuxstb: problems, or just meaningful thoughts?
21:36:03linuxstbI'm trying to get audio playback in mpegplayer, but libmad seems to be freezing on me and never returning from the mad_decode_frame() function. But it's working in the sim...
21:36:17infamiskeymap-x5.c's button_context_yesno has action_yesno_accept as button_select, but the lang strings specify "play to confirm...any other to cancel"
21:37:01infamisshouldn't it be button_play instead?
21:37:31sharpelinuxstb: well, if anything you can fork development off of rockbox and use the sim/mpegplayer as a new media player...
21:37:53linuxstbsharpe: I'm sure the mplayer and vlc people will be scared...
21:38:39linuxstbAlthough it seems that plugins can't play audio in the sim anyway... I'm just decoding the audio, not playing it.
21:38:59*linuxstb should try and fix that
21:39:12sharpewell, chances are that they're not as close to being run on a dap than the mpegplayer fork. :)
21:40:27amiconninfamis: The lang string needs fixing
21:40:41*amiconn would like to have Select as accept on ipod as well
21:40:45infamismost other keymaps have button_play though
21:40:56amiconnIriver has Navi
21:41:56amiconnThe fact that it is Play on most keymaps is because Archos had it on Play, as Play is the centre button of the cursor "cross" on archos recorders
21:42:38amiconn...and it's convenient to have accept on the same button as the select function
21:42:55sharpewell. i'm off, good day everyone.
21:42:59 Quit sharpe ("Leaving")
21:43:05infamisalright never mind...just wondering why delete_file never worked for me
21:43:09 Join TrueJournals [0] (
21:43:10amiconnThen langv2 was introduced, but not may devs cared enough to change the strings per target
21:43:39TrueJournalsHey, is anyone currently trying to fix the broken H10 build?
21:43:44infamis...unless I hit the wrong button on purpose
21:44:14linuxstbTrueJournals: How is it broken? Looks fine here:
21:44:30mirakamiconn: are you sure that all accumumators are cleared ?
21:44:35amiconnmirak: Yes
21:44:45amiconnIf not, there must be a bug somewhere
21:44:46TrueJournalslinuxstb: It seems to be fine... but if you try to boot into it, it either freezes at the end of the bootloader, at the logo screen, or when you change menus
21:45:12mirakamiconn: seems there is, because if I don't clear idct_row is garbeled
21:45:56amiconnEvery mac streak is obliged to use movclr for fetching the result, this will clear the accumulator in parallel
21:46:01TrueJournalsI have a feeling the broken-ness has to do with the CVS commit at 3 Sep 16:20 ("Fixed cpu incorrectly unboosted at boot." not sure... this is just a hunch...)
21:46:11mirakamiconn: and I only changed on function, I am 100% sure of clearing
21:46:47amiconnThen we need to search for the place that forgets to clear accumulators
21:46:55linuxstbTrueJournals: Maybe. It seems the PP5020 (not PP5021/2) targets still have occasional problems with cpu boosting. barrywardell is the only developer with a H10 though afaik.
21:47:00mirakamiconn: I use a move to acc also but at the end of the function I do also a movclr
21:47:24mirakamiconn: yes. that must be a in the core os
21:47:30TrueJournalslinuxstb: so we just have to wait till barry figures how what's wrong and he fixes it?
21:47:32 Join spiorf [0] (
21:47:40linuxstbTrueJournals: Or help him...
21:47:46mirakamiconn: I can't tell wich threads are running
21:47:54TrueJournalsIf I could do that... I would :-p
21:48:04 Join _FireFly_ [0] (
21:48:05linuxstbTrueJournals: One thing you can do, even if you can't program is to confirm exactly which CVS commit broke it.
21:48:20TrueJournalsHow can I figure that out?
21:48:42linuxstbCheck out the source code from CVS at different points in time, compile and test them.
21:49:12linuxstbOr checkout the latest source, and undo the change to the source that you think broke it.
21:49:17TrueJournalslol.. I'll have to figure out how to compile
21:49:20infamisuse: cvs update -D mm/dd/yyyy
21:50:00TrueJournalsJust so you know... I know nothing about computer programming... only web programming :-p
21:50:38amiconnTrueJournals: At least you could try the dailyy builds and find out which day it broke
21:51:14TrueJournalsThe problem with that is that the last daily build for the H10 is from 8/30
21:51:20amiconnHmm, something is wrong with the H10 dailies (!)
21:51:26TrueJournalsyup :-p
21:51:32*amiconn summons Bagder, the master of build system
21:52:22*linuxstb notices an absence of Swedes in the channel
21:52:24TrueJournalsAlso... I had another idea... since we now have a listings of all the PortPinAssignments for the H10 ( would it be possible for someone to make a plugin, or commit a CVS change to allow you to monitor all of the ports?
21:52:32TrueJournalsThat way, we could figure out scroll up from scroll down
21:52:48*Lear wonders what linuxstb is thinking about
21:52:49amiconnDebug->View I/O Ports
21:53:08TrueJournalsthat doesn't really seem to list all of them like the PortPinAssignments page does though...
21:53:15linuxstbLear: In general?
21:53:17infamis...and see what numbers change
21:53:22infamiswhen you press the button
21:53:31infamisprobably a single bit is being toggled
21:53:36Learlinuxstb: no, just about that comment about Swedes...
21:54:01 Join [|DaN|] [0] (
21:54:16TrueJournalsbut wouldn't there be a way to just monitor say GPIOA port pin 0?
21:54:18linuxstbLear: Sorry, I obviously forgot where you live... I'm talking about the Swedes that can do things with the servers.
21:54:19amiconnTrueJournals: It should show GPIO_A thru GPIO_L
21:54:41amiconnThis way you have everything that's listed on th eport pin page
21:55:11 Quit barrywardell ()
21:55:20TrueJournalsbut then how do you figure out what's port pin 1 and what's port pin 5? Sorry, I'm just trying to figure out exactly how this works :-p
21:55:39amiconnEach port pin is one bit in the byte read from the port
21:56:06linuxstbSeems that pcm_play_data works perfectly in the sim, but it's just #ifdef'ed out...
21:56:16linuxstb(in the plugin API)
21:56:18BigMachow come the rating feature isn't working
21:56:32BigMaci've listened to dani california three times
21:56:36LearHe, rockbox.elf on iriver doesn't contain a single divs or divu instruction... :)
21:56:37TrueJournalsamiconn: So, I believe the default value for all of them is FF (nothing), if it changes to say FC... that would be port pin 3?
21:56:41BigMacand no stars show up
21:56:56BigMacI'm no in id3 viewing mode though
21:57:05infamistruejornals, that's 2 bits...
21:57:46TrueJournalsso FC would be port pin 2?
21:57:50infamis0Ch=1100b; 0Fh=1111b
21:58:22infamisif it went from FF to FE (1110), that's pin 0; if it went from FF to FD (1101), that's pin 1, etc...
21:58:40TrueJournalswas forgetting about pin 0 :-p
21:59:49infamis0 is the most important number out there! without it, the world would be a crazy place
22:00:03infamisthen again 'without it' wouldn't exist...
22:00:22 Join barrywardell [0] (n=barrywar@
22:00:30SoapSong rating (%rr in WPS) is not the same as song playcount (%rp in WPS)
22:01:08SoapYou said you /played/ "Dani California" three times, but IIRC you are using %rr in your WPS.
22:03:56barrywardellTrueJournals: I've been away from my computer since yesterday
22:04:11TrueJournalsglad to see you back
22:04:11barrywardelljust read about the broken build now
22:04:23dan_a_awaylinuxstb: create_thread_on_core doesn't exist in sim builds? Ah - the sim uses its own threading code from a different file.
22:04:26barrywardelli'm gonna test it on my h10 and see what happens
22:05:43Soapand jClix uses %rr, not %rp
22:06:20linuxstbdan_a_away: Yes, so when I exported it via the plugin api, it wasn't found in the sim build.
22:06:46dan_a_awaylinuxstb: I'll get that fixed
22:07:00 Nick dan_a_away is now known as dan_a (n=dan-mirc@
22:08:16[HO]vo|tanyone need parts for a v2 recorder?
22:10:27infamisI'll take a hard drive
22:10:42TrueJournalsI just noticed something weird... On the wiki, it says that GPIOC and GPIOD are supposed to change together when the scroll pad is used... it also says GPIOC is supposed to be default F7
22:11:22TrueJournalsWhen I go into my View I/O ports, GPIOC is FF and does not change with the scrollpad, and GPIOD is DD and changes to FD with the scrollpad
22:13:26LearAnyone with m68-elf-gcc 3.4.x installed? If so, could you run it with the argument −−target-help and see if the option "-m5206e" is included?
22:14:15TrueJournalsHowever, I did look in the rockbox code a bit... and it seems to look for both GPIOC and GPIOD to change... so is there a bug in the View I/O Ports page?
22:14:29barrywardellTrueJournals: That's strange. I'll look at it again
22:15:12infamisLear, yes, on 3.4.6 (linux)
22:15:51linuxstbLear: Yes on 3.4.4 (Linux)
22:16:45 Quit mirak (Connection timed out)
22:18:16 Join mirak [0] (
22:19:41 Quit [|DaN|] ("|| Join BFE:CoD ~ A Call of Duty 2 Tournament || || #battleforeurope.cod2 || <3 ||")
22:21:41barrywardellTrueJournals: I'm trying latest cvs here with no new problems
22:22:22 Quit infamis (" jazzfest...chicago")
22:22:57tucozanyone using emacs around? sometimes i press ctrl-pgdown or something like that, and the left border gets filled with arrows pointing left. Do you know what this is, and how to 'get back' to the code?
22:22:58Learlinuxstb: Interesting... For our targets, using that instead of "-m5200" should generate noticably better code. Happens with gcc 4.0.3 at least.
22:23:27LearHaven't actually tried running it yet though (still building...).
22:24:21tucozthat is, i can get back to the code buffer by selecting it in the menu, but i wonder what kind of mode that is
22:24:35dan_alinuxstb: I've committed what I think should fix the sim problem - can you confirm it's OK
22:24:39TrueJournalsbarrywardell: really? That's odd... let me try again...
22:24:48tucoznevermind. ctrl+pgup got me back !??!
22:25:08barrywardellTrueJournals: delete the .rockbox directory and rockboxl.h10 before extracting the
22:25:14Leartucoz: jumped a "page" to the right (not an emacs user, so I'm guessing...)?
22:25:19 Join pixelma [0] (i=pixelma@
22:25:34tucozmaybe. Now i know what triggers it at least
22:25:36TrueJournalslol... seems to be building a new version right now...
22:26:49TrueJournalsok... downloading...
22:27:17tucozpixelma, nice work on the images
22:27:31tucozi committed them yesterday (in case you didn't know)
22:27:33barrywardellTrueJournals: the changes in the new build won't affect the h10
22:27:57TrueJournalsdoesn't work
22:27:57pixelmathanks :)
22:28:09TrueJournalsjust froze on rockbox logo screen
22:28:18TrueJournalsVersion: 060903-2223
22:28:47TrueJournalstrying again...
22:29:04TrueJournalsfroze on rockbox boot loader (all the text that's supposed to show up showed up)
22:29:32 Quit Pyromancer (Read error: 104 (Connection reset by peer))
22:29:32 Join Nico_P [0] (
22:29:56 Quit TeaSea (Remote closed the connection)
22:30:00TrueJournalsjust reset with same result
22:30:45 Nick belze is now known as belZe (
22:31:34 Join TeaSea [0] (
22:32:39barrywardelli'll build a version before the frequency scaling change you mentioned
22:32:54barrywardellTrueJournals: what model do you have?
22:33:03linuxstbdan_a: I get a warning: thread-sdl.c:64: warning: implicit declaration of function 'create_thread'
22:33:20TrueJournalsI have a 20GB
22:33:30linuxstbdan_a: Seems I'm not alone...
22:33:56linuxstbdan_a: And you also need to define remove_thread_on_core()...
22:35:09 Quit secleinteer (Read error: 104 (Connection reset by peer))
22:35:09 Join Angryman_h10 [0] (i=5438a665@gateway/web/cgi-irc/
22:35:41TrueJournalshey, Angryman... you're having the same boot problems with the CVS of rockbox, right?
22:35:57TrueJournalsok, I'll try it
22:36:01dan_alinuxstb: Take 2 coming up...
22:36:05barrywardellthat's a version without the cpu frequency changes
22:36:47Angryman_h10I setup cygwin again (better folders) and tried to compile with verion 1.36 bt i now got an missing arm-elf-gcc error
22:36:48 Quit nls (" HydraIRC -> <- Try something fresh")
22:37:28barrywardellAngryman_h10: what model do you have?
22:37:35Angryman_h105GB 6GB
22:37:50TrueJournalsbarrywardell: that version you sent me seems to be fine
22:38:03barrywardellso it's a cpu boosting issue
22:38:16TrueJournalsAppears so..
22:38:29TrueJournalsIf I look in the CPU Frequency under the debug menu, it says:
22:38:33barrywardellit's weird because I have no problems with my 20gb and cvs version
22:38:48TrueJournalsFrequency: 30000000
22:38:48TrueJournalsboost_counter: 0
22:38:49Angryman_h10Barry could you give me a tip with the arm-elf error??? I need to set a path in cygwin for that and cant remember how?
22:39:12Angryman_h10when time ^^
22:39:20linuxstbAngryman_h10: You can edit /etc/profile (or c:\cygwin\etc\profile from Windows)
22:39:48Angryman_h10so i dont have to set the path each time
22:40:08barrywardelllinuxstb: you mentioned problems with cpu boosting. does that still affect the ipods too?
22:40:13linuxstbNot if you edit /etc/profile (or a local .bash_profile in your user's home directory).
22:41:12linuxstbbarrywardell: There are unexplained freezes which seem to be attributable to cpu boosting. I've never been able to recreate it reliably enough to try and investigate it though. It seems to only affect PP5020 ipods (not PP5021/PP5022)
22:41:23 Quit TeaSea (Remote closed the connection)
22:41:51dan_alinuxstb: The sim doesn't seem to have a remove_thread(). Is that right?
22:41:54linuxstbbarrywardell: You'll make lots of 4g ipod owners happy if you can track it down though.
22:42:04linuxstbdan_a: It doesn't? I've no idea...
22:42:48linuxstbdan_a: It looks like in the plugin API (without an #ifndef SIMULATOR)
22:43:24TrueJournalslinuxstb: so it seems that Angryman and I are experiencing something no one can explain?
22:43:45linuxstbdan_a: Ah, it's a dummy function in uisimulator/common/stubs.c
22:44:10Angryman_h10export PATH=/home/linus/sh1/bin:$PATH <−− Is that what ive got to write in the profile ??
22:44:20 Join TeaSea [0] (
22:44:43linuxstbAngryman_h10: No - replace "/home/linus/sh1/bin" with wherever your arm-elf-gcc is installed.
22:44:53Angryman_h10ah ok
22:44:56amiconntucoz: There are 2 errors related to the new images in the manual:
22:45:02dan_alinuxstb: Thanks. I've not looked inside the sim code before - the threading things at least are scattered all through it!
22:45:21linuxstbdan_a: "grep -r remove_thread uisimulator" is your friend.
22:45:28amiconn(1) In the recorder v1 manual, the Left button isn't marked at all
22:45:31 Quit matsl (Remote closed the connection)
22:45:54 Join mantono [0] (
22:46:08amiconn(2) In the FM Recorder manual, the play button has a line pointing to it, but the text is missing
22:46:15*linuxstb curses libmad and goes to watch some TV
22:46:19tucozamiconn, thanks. I'll fix that
22:46:54TrueJournalsI gotta go
22:47:02TrueJournalshope somone can figure out this problem X-|
22:47:08barrywardelllinuxstb: what about kernel.c:397
22:47:08 Part TrueJournals
22:47:19barrywardellshould that be |= instead of =?
22:47:46pixelmatucoz: while your at it there's another tiny tiny glitch in the recorder drawing, I'm fixing it right now
22:47:55tucozoh. ok
22:48:04barrywardellbah. just missed him
22:49:57Angryman_h10Thx linuxstb that worked perfect
22:50:16 Quit mantono (Remote closed the connection)
22:50:46dan_abarrywardell: The TIMER1_CFG line?
22:51:28barrywardellin system.c:1262 it mentions TIMER1_CFG too
22:52:29dan_aI think the TIMER1_MASK line doesn't look right
22:52:53barrywardellsorry,that's what I meant
22:53:06barrywardella broken timer could definitely be causing a lock up
22:53:48dan_aI'm surprised it doesn't affect all iPods, though
22:54:17pixelmatucoz: updated Recorder_logo.svg - the problem was just in the pdf (and eps) again - but almost not noticeable
22:54:26barrywardelldoes it only affect PP5020 models?
22:55:05amiconnI wonder whether the Ondio's On/Off button should be renamed to Power
22:55:39LearNeat. With gcc 4.0.3, -m5206e cuts rockbox.iriver with 2.5 kB and reduces boost on an ogg test track by about 2 percentage points.
22:55:50amiconnOn/Off just sounds cumbersome, and on the X5 we also call that combined button Power
22:56:02LearOh well, enough testing of that by me for today...
22:56:22dan_abarrywardell: I haven't noticed any random lockups on the 3G. Or at least, the ones I have seen I thought I'd caused.
22:56:28barrywardelldan_a: yellow build table
22:56:35pixelmaamiconn: but it is not only used as "Power"...
22:56:43dan_abarrywardell: Working on it :)
22:57:16barrywardellcool, just wanted to make sure you noticed it
22:57:30amiconnNo, but it's just how we call the button. On X5 Power is also not only used for powering on/off
22:58:15cariniAnyone use the "tags" makefile target?
22:59:53 Quit lightyear (Remote closed the connection)
23:00:54 Join TrueJournals [0] (
23:01:12TrueJournalsbarry, you still here?
23:02:24TrueJournalsCan you make another "custom" rockbox build for the 20 GB for me without the CPU boosting CVS Commit mentioned earlier, and instead of having the scroll bar always go down, have it always go up?
23:02:44tucozpixelma, thanks. I'll commit them now.
23:05:49barrywardellTrueJournals: download that file again. it's updated
23:05:55***Saving seen data "./dancer.seen"
23:06:06TrueJournalswhat was the link?
23:06:35Angryman_h10Barry can i help you in any way in my freetime like (testing a row of numbers throug mybe theres a golden hit or just stopping spamming you ;-)">
23:06:58Jasonbarrywardell: is somebody workin on that lcd bug on H20?
23:07:20barrywardellJason: yes, I am. I think I understand the problem
23:07:29barrywardellnow I just have to write up a fix
23:07:39Jasonbarrywardell: oh thank you :D
23:08:45TrueJournalssweet :-D
23:08:52TrueJournals(on the LCD fix)
23:09:09barrywardellAngryman_h10: just keep testing the new cvs builds and let me know of any new problems that arise
23:09:34Angryman_h10is there a update for the 5GB model
23:09:42Angryman_h10so i will check it now
23:10:38TrueJournalsOne time, I went into the debug menu and accidentally changed boost_counter to 0
23:10:48TrueJournalsusing your up patch to change it to one makes the CVS work!
23:10:51 Part TrueJournals
23:11:57barrywardellAngryman_H10: did you by any chance do the same thing as TrueJournals in the debug menu?
23:12:16Angryman_h10No i sut downloaded the latest cvs
23:12:39Angryman_h10It hung up immediatly after the new install
23:13:00Angryman_h10removed now the lines by myself and it works
23:13:01barrywardellhad you ever been in the cpu frequency screen in the debug menu?
23:13:21Angryman_h10Those things i dont touch ^^
23:13:43preglowSlasheri: did you post a patch for the scheduler?
23:14:28barrywardelli mean">
23:14:39barrywardelltry that (rename it to rockbox.h10 though)
23:15:08barrywardellit's the same as i gave TrueJournals but compiled for 5gb instead
23:16:22Angryman_h10seems to work
23:16:39Angryman_h10but goes up now ^^ Doesnt matter ^^
23:16:55Angryman_h10up down who cares ^^
23:17:24 Quit tucoz ("Leaving")
23:19:31barrywardellI think I'll disable frequency adjusting for now then
23:19:57Jasonbarrywardell: and when do you think that lcd bugfix will be avalaible?
23:20:22barrywardellJason: hopefully in the next few days
23:20:34 Join Anna [0] (
23:20:41 Join webguest01 [0] (i=5024e36c@gateway/web/cgi-irc/
23:20:46 Quit webguest01 (Client Quit)
23:21:00barrywardellthe 5gb lcd bug is solved, so I'll commit that fix tonight
23:21:12Annai have some question
23:21:32Annawhen i boot to original ipod 5g firmware all settings reset
23:21:37Annait's normal?
23:22:30Angryman_h10Barry could you point me rto the jpegviewer plugin??? I am searcing it and cant find it in CVS
23:22:43dionoeain apps/plugins
23:22:56 Quit t0mas ("Doei")
23:23:18Angryman_h10got it^^
23:23:35pixelmadionoea: thanks for fixing solitaire :) but I have one last suggestion...
23:23:39BigMachow do I use the rating feature
23:23:44dionoeapixelma: go ahead :)
23:23:52BigMacthe manual only gives it as numeric constants
23:24:02barrywardellAngryman_h10: can you download that file again and test. I've updated it with frequency adjusting disabled
23:24:02BigMacbut i do not see it anywere on my wps
23:24:04 Join Anna_ [0] (
23:24:30Anna_sorry i lost my connection
23:24:44Angryman_h10download from CVS or your link ???
23:24:47pixelma...regarding the "shortcuts" (for example long press right on Ondio). They were chosen for your old layout and aren't intuitive anymore...
23:25:04 Join TrueJournals [0] (
23:25:07 Quit Nico_P ()
23:25:12TrueJournalsI'm back
23:25:15barrywardellmy link
23:25:48Anna_when i boot in apple os all settings reset, it's normal?
23:26:14TrueJournalsAngryman: did you try the fix that worked for me?
23:27:06dionoeapixelma: what keys would you change ?
23:27:21Anna_hi, before u said that solve a lcd issuer in 5g. which?
23:27:29Angryman_h10@Barry: Works fine !!!
23:27:49Angryman_h10@ Anna 5G iriver_H10
23:28:28 Join Bagder [0] (
23:28:41Anna_hehehe ok
23:28:53TrueJournalsAngryman_h10: what did you have to do to get yours to work?
23:29:08barrywardellTrueJournals: same thing
23:29:19TrueJournalssweet :-D
23:29:26pixelmadionoea: just like you can see it on the board, for example it is now "Hold Left - Put the card from the top of the remain's stack on top of the cursor" - should be "Hold Down" IMO because you put the card below the remains stack
23:29:36Anna_no idea for what all my settings in ipod 5G reset all time?
23:29:38TrueJournalsI wonder if that would solve the problems with the iPod 4g talked about in here before...
23:29:46Anna_i try to search in forums, but not luck
23:29:54Jasonand is there any progress with the touchpad driver for h10 series?
23:30:29dionoeapixelma: can you provide a patch ? :)
23:30:32Angryman_h10Jason come on... you already can see that barry works like hell... it will come on day^^
23:30:42 Quit scorche (Read error: 110 (Connection timed out))
23:30:55JasonAngryman_h10: im not saying that he should work harder
23:31:02barrywardellthe problem is it's not really a proper solution. disabling cpu frequency adjusting is far from ideal
23:31:04TrueJournalsJason: barry didn't say anything about a touchpad driver... but he did mention that he knew what was causing the LCD problems ;-)
23:31:05Angryman_h10i didnt wanted to say that^^
23:31:16Jasonim happy that rockbox is booting on h10
23:31:21Jasonim just asking
23:31:51barrywardelli'm going to try a few ideas to get the touchpad working today
23:31:53 Join scorche [0] (
23:31:55JasonTrueJournals: oh i know the solution for lcd driver?
23:32:15 Quit Anna_ (Read error: 104 (Connection reset by peer))
23:32:17Jasonshite... reading stupid things
23:32:21TrueJournalsbarrywardell: But could it be that the people who have the crashes have cpu frequency adjusting disabled? (I'm sure you don't know about this... and I may be wrong)
23:32:53TrueJournalsJason: Barry said he knew what was causing the LCD driver problems, and just needs to write code to fix this...
23:33:32barrywardelldan_a: what do you think about changing that = to |=?
23:33:42JasonTrueJournals: i know... and also i was asking about that touchpad... if somebody made a progress with it
23:33:44barrywardellshould I commit a fix?
23:34:22dan_abarrywardell: I don't know enough to say either way.
23:34:25TrueJournalsJason: As far as we know, no progress has been made... but I did stumble across something weird in the I/O ports..
23:34:53barrywardellJason: there are only two of us working on the H10 port. It's all just trial and error with the touchpad so there's no way to tell when we will happen across the solution
23:34:54JasonTrueJournals: thats good :)
23:35:05 Join senab [0] (
23:35:19senabevening people, just a quick question
23:35:20TrueJournalsthe wiki says that GPIOD and GPIOC should both change when touching the touchpad, but only GPIOD seems to change...
23:35:31 Quit Kohlrabi (Nick collision from services.)
23:35:33TrueJournalsso, that may be a step closer to solving the touchpad mystery
23:35:41 Join Kohlriba [0] (
23:35:41barrywardelldan_a: I'll leave it for now and ask linuxstb when he gets back
23:35:43TrueJournalsor it may just be nothing we need to worry about :-\
23:35:48senabyou know you've got wget to download files, is there a similar util to upload to ftp?
23:36:17JasonTrueJournals: i would use rockbox even if only down would work :)
23:36:18TrueJournalssenab: no offense... but what does that have to do with rockbox???
23:36:49TrueJournalsI agree with you Jason... there's nothing really major I've needed up for (except for changing volume :-\)
23:36:56senabbecause it's to do with my build environment, and i want my script to automatically upload the builds to webspace
23:37:33pixelmadionoea: no (I know it's a pity) - just followed your instructions to tell you about this in IRC ;)
23:38:22JasonTrueJournals: in the worst case you could use the fastforward and rewind button for navigation thru menus :)
23:38:59dionoeapixelma: pfff
23:39:53Jasonand the touchpad will be some sort of enter or advanced menu
23:40:45TrueJournalsheh... well, interesting idea if we never figure this thing out :-p
23:40:50 Quit Anna (Read error: 110 (Connection timed out))
23:40:54TrueJournalsI think we'll get it soon though :-)
23:41:48Jasoni will pray for you two :)
23:42:04TrueJournalslol, I don't develop rockbox :-p
23:42:12TrueJournalsI just try to help as much as I can :-\
23:42:21 Join SanAlmighty [0] (
23:42:23TrueJournalsI believe is Barry and Badger working on the actual code
23:42:45*Bagder just pretends to do things
23:42:58 Quit carini ("Chatzilla 0.9.75 [Firefox]")
23:43:15barrywardellonly myself and lorijho actually have H10s
23:43:25Jasoni will pray for whoever works on h10 rockbox :) <−−- FIXD
23:43:49markunBagder: to help people choose a font I'm working on this page:
23:43:49Angryman_h10@ TrueJournals do you have any problems displaying folders??? I always have to switch to display all in the filters
23:43:56markunFrom where should I link to it?
23:44:08Bagderthe h10 rockbox is of course 95% or so generic rockbox
23:44:30Bagdermarkun: very good question, I don't know ;-)
23:44:38markunMaybe you can also put a link in /fonts
23:44:40TrueJournalsAngryman_h10: From what I know, the default setting will show everything except hidden folders. If you want to see hidden folders, you have to switch to display all
23:44:47barrywardelland probably another 4% generic portalplayer
23:45:15dionoeathe touchpad is only 1% ?!?
23:45:18Angryman_h10I unhided the folders several times but they get hidden each time again
23:45:20Bagdermarkun: /fonts take you to the font package download page these days
23:45:25Bagderbut sure I can
23:45:43markunOr maybe I should rather link to /fonts
23:45:46barrywardellBagder: have you noticed that the H10 daily builds don't get updated?
23:45:52markunwill think a bit more about it
23:46:00Bagderbarrywardell: I think I've fixed that
23:46:05Bagdera permission thing on the server
23:46:07barrywardellcool, thanks
23:46:10TrueJournalsAngryman_h10: The folders go back to hidden, or your setting goes back to default?
23:46:35barrywardelldionoea: lots of small changes. the touchpad and lcd are the only big things
23:46:38 Quit senab (" HydraIRC -> <-")
23:46:42TrueJournalsalso barrywardell: did you ever look into what I found in the View I/O Ports?
23:46:58barrywardellyou mean with gpioc?
23:47:03 Quit _jhMikeS_ (Read error: 104 (Connection reset by peer))
23:47:11barrywardellit's on my list of scrollpad tests to do
23:48:10Angryman_h10@ Barry... i get a Data abort message now while displaying jpgs didnt happen before
23:49:33*barrywardell has never viewed a jpeg on his h10
23:50:08 Join DAK [0] (
23:50:14Angryman_h10? ^^ Really it worked before ^^
23:50:19Bagderwhat's a jpeg? ;-P
23:50:23TrueJournalsjpeg viewer works fine here..
23:50:27TrueJournalsDid you do anything weird?
23:50:35 Quit DAK (Client Quit)
23:50:36TrueJournalsCan you send me the jpeg you tried to open?
23:50:44Angryman_h10Its with evryone
23:51:19barrywardellAngryman_h10: post a bug report. I'll get to it once this scrollpad is working :)
23:51:30TrueJournalssend me one of them... like I said, works fine for me...
23:52:15TrueJournals(send it to
23:52:38dionoeaomg, you pasted your email address on an IRC chan
23:52:51TrueJournalsI know
23:52:54TrueJournalsI don't really care
23:53:00TrueJournalsmy e-mail is all over the internet
23:53:39TrueJournalswould you rather I said aimjournals[at]gmail[dot]com?
23:53:47*billytwowilly wonders just how big TrueJournals' penis is by now
23:54:09Angryman_h10True Journals... it doesnt depen on the files... i tested now several different ones... it always says Data abort at 00014248
23:54:24TrueJournalslol... that's why I like gmail... it filters out spam :-p
23:57:13Angryman_h10TrueJournals whioch model do you have ???
23:57:53TrueJournals20 GB
23:57:59TrueJournalsDidn't think about that
23:58:06TrueJournalsyou have a 5 or 6, right?
23:58:13 Quit San (Read error: 110 (Connection timed out))
23:58:44TrueJournalsthat's probably why :-p
23:58:55markunlinuxstb: this is probably the Elio chip with AZU on it (same as iriver h1xx)

Previous day | Next day