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

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

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

#rockbox log for 2009-11-19

00:00:04Utchybannold_wheel_value never get assign -1 after long 'untouch'.
00:00:41Utchybann*first 'touch'*
00:03:58 Quit Zorda (Read error: 104 (Connection reset by peer))
00:07:27 Nick krazykit is now known as krazykit` (
00:07:49 Nick krazykit` is now known as krazykit (
00:08:40 Join kugel [0] (n=kugel@rockbox/developer/kugel)
00:09:12*kugel wonders why the mtp guys are using an "outside" git repo
00:09:47kugelIMO they should've been given a branch on the svn repo
00:15:50*dewdude got the bootloader...and rockbox...installed on his beast
00:17:41JdGordon|kugel: get a chance to look at 10709?
00:19:18 Quit perfectdrug ("CGI:IRC (Ping timeout)")
00:19:33kugelJdGordon|: not before tomorrow
00:20:09JdGordon|arg.. i wantted to comit it tonight.... :(
00:20:35JdGordon|I guess i was really asking if you had objections to any of it?
00:21:19kugelI haven't looked at the latest one, but the one before was ok I think
00:21:55 Join fdinel [0] (
00:22:00JdGordon|and the api change? or we argue about that later?
00:22:48kugellater rather
00:23:15kugelI'm not opposed to api changes, but it should be a separate commit IMO
00:23:53JdGordon|yep fine, ok
00:24:18 Quit kugel ("exit(0);")
00:28:48 Quit bmbl ("Bye!")
00:33:27 Quit ender (" The only reason some people get lost in thought is because it's unfamiliar territory. -- Paul Fix")
00:36:27 Join kugel [0] (n=kugel@rockbox/developer/kugel)
00:37:43 Quit bertrik (Read error: 113 (No route to host))
00:39:45bluebrotherdewdude: what OS are you using? Are you by chance using WXP with WMP10?
00:40:07dewdudeblue: as a matter of is.
00:40:19dewdudei booted in to ubuntu and it worked fine that way
00:40:44 Part marek_ ("Konversation terminated!")
00:41:34dewdudei've got rockbox already installed and running
00:41:47bluebrotherwell, there's a known issue with XP / WMP10 acting up. No idea what's going wrong there, using WMP11 works fine. Probably some incompatibility with the MTP drivers provided by WMP
00:42:16dewdudei figured there was something up...because it'd constantly fail
00:42:46dewdudeubuntu wasn't really that easy either...cuz it wanted to mount MTP which made it i'd have to unmount and make sure everything reset, or LIBMTP would panic
00:43:07dewdudebut that's the only time I really should have to deal with that situation no biggie
00:43:13 Join AaronM [0] (
00:43:48 Join efyx_ [0] (
00:43:48 Quit seani (Read error: 60 (Operation timed out))
00:44:44bluebrotherMTP is not a filesystem you can mount on linux ... maybe there is some fuse implementation handling this.
00:45:10bluebrotherbut good to head you got it working :)
00:45:53 Part froggyman
00:46:19dewdudebluebrother, well, that's what i meant
00:46:42dewdudefuse was doing popped up offering to load rhythmbox
00:47:07dewdudefirst two times i hit cancel...which lead to libmtp panic...wasn't till i hit that unmount button in the left corner that things worked...but either's working.
00:47:18dewdudethat is the reason i bought the player...i'm completely happy now.
00:47:45dewdudebut thanks. i'll go ahead and jump outta here.
00:47:47 Part dewdude ("Leaving")
00:50:48 Quit BlakeJohnson86 ("Leaving.")
00:51:30 Join BlakeJohnson86 [0] (
00:54:36 Join salty_horse [0] (
00:58:26 Join seani [0] (n=seani@
01:03:21 Join dmb [0] (n=Dmb@unaffiliated/dmb)
01:03:40 Join darkham [0] (
01:09:27 Quit DerPapst ("Leaving.")
01:12:16 Quit salty-horse (Read error: 110 (Connection timed out))
01:12:49 Quit n17ikh (Read error: 104 (Connection reset by peer))
01:13:31 Join n17ikh [0] (
01:16:09 Join salty-horse_ [0] (
01:16:18 Quit salty-horse_ (Read error: 104 (Connection reset by peer))
01:17:10***Saving seen data "./dancer.seen"
01:20:04 Quit darkham (Remote closed the connection)
01:23:53 Quit salty_horse (Read error: 145 (Connection timed out))
01:31:13 Quit Thundercloud (Remote closed the connection)
01:33:25 Quit efyx_ (Remote closed the connection)
01:34:27 Quit kugel (Read error: 60 (Operation timed out))
01:34:33 Quit stripwax ("")
01:44:17 Join perfectdrug [0] (
01:46:44 Join phanboy4 [0] (
01:48:30perfectdrugis it just me or is the recent simulator for the mini2440 broken? it won't compile
01:56:58hillshumperfectdrug: Something like
02:00:04perfectdrughillshum: yes exactly
02:00:16hillshumIt's not just you
02:02:03perfectdrugdo you have a physical 2440? I quess I finish the fidelity wps blindly then, not much left, you wanna test?
02:02:34hillshumNo, I just tried building the sim to see if it was just you
02:02:55perfectdrugah ok, then thanks:D
02:03:16hillshumyou could try consolidating the definitions of 'usb_inserted'...
02:03:47 Join Strife89 [0] (
02:09:02 Join StealthyXIIGer [0] (
02:11:05 Join Strife1989 [0] (
02:11:10perfectdrugI guess I'm not capable of doing this
02:11:38 Quit Strife89 (Nick collision from services.)
02:11:50 Nick Strife1989 is now known as Strife89 (
02:12:09 Join Strife1989 [0] (
02:18:27 Part toffe82
02:25:38 Quit grndslm ("Leaving")
02:27:14hillshumperfectdrug: You might submit a bug report if there isn't one already
02:27:41perfectdrugwill do
02:32:43hillshumperfectdrug: Actually, never mind. Report in the development thread on the forum, not in flyspray
02:36:24perfectdrugups too late :/
02:50:15hillshumperfectdrug: What's the number? Ask here for it to be deleted
02:51:49 Quit AEnima1577 ("Leaving.")
02:54:14 Join Blue_Dude [0] (n=chatzill@rockbox/developer/Blue-Dude)
02:56:12 Join grndslm [0] (
02:58:13Blue_DudeWhat's the command to build a voice file?
03:16:36Strife89The USB driver for the Nano 2G is proving ... rather buggy so far. :(
03:17:11***Saving seen data "./dancer.seen"
03:20:43Strife89I've gotten two divide by zero errors already in the middle of playing music. :/
03:21:13Strife89And I can't even read the error code because the screen is badly cracked.
03:33:28 Join CaptainKewl [0] (
03:36:14 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
03:36:33 Join MethoS- [0] (n=clemens@
03:38:12 Quit perfectdrug ("CGI:IRC")
03:47:47 Join froggyman [0] (
03:49:03 Join AEnima1577 [0] (
03:56:28 Join T44 [0] (
03:56:31 Join TopyMobile__ [0] (
04:00:51 Part froggyman
04:01:33 Quit panni__ (Client Quit)
04:03:28 Part AEnima1577
04:05:40 Quit TheSeven (Nick collision from services.)
04:05:49 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
04:05:59 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
04:13:28 Quit Rondom (Nick collision from services.)
04:13:39 Join Rondom [0] (
04:14:01 Quit TopyMobile_ (Read error: 110 (Connection timed out))
04:14:27Blue_DudeOK, I build a brand new voice file from a freshly cleaned build directory using the defaults for my e200. I moved the voice file to the sim and enabled voiced menus. Some of the menu items are OK, but some say something completely wrong. What is going on?
04:14:50 Quit Topy (Read error: 110 (Connection timed out))
04:16:40JdGordonyour english.lang and the build are out of sync
04:17:06Blue_DudeI just built one from the latest update from scratch. How can it be out of sync?
04:17:35Strife89Make sure the firmware build itself is up to date as well.
04:17:55Strife89It's not enough to simply update the voice file.
04:18:04Blue_DudeI'm building from the trunk.
04:18:13Blue_DudeWhat else is there to update?
04:18:50JdGordonclean tree?
04:18:56JdGordonmake clean
04:19:15JdGordonwasnt it your patches which did this? :D
04:19:25Blue_Dudesvn status shows only my personal changes to pcmbuf that I'm working on. No changes to english.lang
04:19:28Blue_DudeMan I hope not.
04:19:54JdGordonif you changed any #defines or includes you can cause the .lang to not build the same
04:20:01JdGordonor features.txt can get out of date
04:20:01Strife89Checkout a fresh copy of the trunck to a new folder, just in case.
04:20:11JdGordonna, just do a make clean and make
04:20:21Blue_DudeI did a make clean.
04:21:17Blue_DudeHere's what I did: tools/configure, e200, advanced, voice, all defaults; make clean, make voice
04:22:06Blue_DudeIt created a english.voice file and it's a little screwy.
04:22:19JdGordonthats interesting...
04:22:31*JdGordon has a try
04:23:03JdGordonoh! you're using the e200 voice with a e200-sim?
04:23:13Blue_DudeYeah. Is that bad?
04:23:23JdGordonit shouldnt be, but there could be differences
04:23:36*JdGordon prepares to annoy people
04:23:37Blue_DudeRun it again with e200 sim, voice, etc?
04:23:57*Blue_Dude prepares to be annoyed.
04:25:05*Blue_Dude is working on the pcmbuf software mixer, FWIW.
04:27:13JdGordonwith the late mixing so we can have voie while music is pasued?
04:27:17Blue_DudeI'm beginning to understand why nobody has done it yet.
04:27:41Blue_DudeYeah, that's the hard part.
04:27:42saratogai like this last clip patch
04:27:50saratogaif no one has problems i'm going to commit it as is
04:28:17Blue_Dudesaratoga: no problems yet with v1a, with a couple of hours of playback
04:28:59saratogatry 2a
04:29:08*Blue_Dude still thinks that playback.c change is a little weird
04:29:09saratogasame logic but does a memmove when possible
04:29:18saratogawell if you've got any idea how to improve it
04:29:28Blue_DudeNo, I'm not sure what it does. :)
04:30:14saratogawell me neither, but apparently it prevents the buffer from corrupting itself
04:30:41JdGordonjust what we want to hear :p
04:30:50saratogaimo its a sideways step in code cleaniness and a leap forward in stability
04:31:02saratogawhich isn't a bad thing
04:31:29JdGordoncan anyone think of a good backronym for sbs which isnt statusbar skin?
04:31:37JdGordonseen as its no longer a bar
04:32:59JdGordonok, backronym doesnt mean what I thought it meant
04:33:01Blue_Dudestrikingly beautiful status?
04:33:13JdGordonnow shorten that for a filename?
04:33:43Blue_Dudesuddenly behaving skin?
04:35:16Blue_Dudeseriously bothersome svn?
04:35:20*JdGordon goes with "short bowel syndrome"
04:36:02Blue_Dudestill behaving strangely?
04:37:47Blue_Dudestatus bar surrogate?
04:43:42Blue_Dudeskinnable background status?
04:43:51Blue_DudeI coming up empty here.
04:45:02krazykitstatus bar: skinnable
04:49:42*Strife89 will probably try again with making a theme.
04:50:56JdGordonwell... fuck... patch doesnt work on target at all... fine in all the sims though
04:52:35*Blue_Dude covers his virgin ears
04:56:49Blue_DudeSo what's it wrecking on?
05:09:52 Join FOAD_ [0] (
05:11:53Blue_DudeFor the pcmbuf software mixer, I'm probably going to have to expand the pcmbuffer some. A second or so. Otherwise the pcm driver may starve.
05:13:29 Quit liar (Read error: 113 (No route to host))
05:13:43Blue_DudeMight cause problems on the clip and other low memory devices...
05:14:48 Quit togetic (Connection timed out)
05:16:58 Quit Blue_Dude ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
05:17:13***Saving seen data "./dancer.seen"
05:24:10 Quit FOAD (Read error: 110 (Connection timed out))
05:24:10 Nick FOAD_ is now known as FOAD (
05:24:57 Join AEnima1577 [0] (
05:30:26 Quit JdGordon ("Leaving.")
05:34:20 Join AndyIL [0] (n=pasha_in@
05:38:53 Quit AaronM ("Emo Time In My Corner... //_-")
05:43:03*Strife89 does research, hoping to gather material to propose a DS port of Rockbox.
05:43:44 Join shai [0] (
05:43:47 Quit AEnima1577 ("Leaving.")
05:45:08 Quit AndyI (Read error: 110 (Connection timed out))
05:54:05 Quit bzed (Remote closed the connection)
05:54:37 Join bzed [0] (
05:57:41 Quit bzed (Remote closed the connection)
05:57:54 Join bzed [0] (
06:01:47kkurbjunJdGordon|: you had a patch for me to try?
06:03:50 Quit Horscht ("Verlassend")
06:04:11 Quit bzed (Remote closed the connection)
06:04:24 Join bzed [0] (
06:06:15 Join AEnima1577 [0] (
06:07:26 Quit bzed (Remote closed the connection)
06:07:34 Quit rvvs89 (Remote closed the connection)
06:09:02 Join bzed [0] (
06:11:44 Quit JdGordon| (Read error: 104 (Connection reset by peer))
06:11:55 Quit HBK (Read error: 104 (Connection reset by peer))
06:13:01 Join HBK [0] (
06:16:05 Quit MethoS- (Read error: 54 (Connection reset by peer))
06:21:37 Quit evilnick (Read error: 104 (Connection reset by peer))
06:23:55 Join rvvs89 [0] (
06:27:15 Join BHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
06:30:39 Quit bzed (Read error: 110 (Connection timed out))
06:37:11 Quit Strife89 ("Sleep.")
06:43:42 Join bzed [0] (
06:48:40 Quit bluebrother (Nick collision from services.)
06:48:43 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
06:55:51 Join JdGordon [0] (n=JdGordon@rockbox/developer/JdGordon)
06:56:15 Part JdGordon
06:56:19 Join JdGordon [0] (n=JdGordon@rockbox/developer/JdGordon)
06:57:43 Quit CaptainKewl ("( :: NoNameScript 4.22 :: )")
07:01:06 Quit jfc (Read error: 104 (Connection reset by peer))
07:01:29 Join jfc [0] (
07:03:37 Quit JdGordon (Remote closed the connection)
07:06:07 Join jdgordon [0] (n=JdGordon@rockbox/developer/JdGordon)
07:12:42 Quit Strife1989 (Read error: 110 (Connection timed out))
07:17:16***Saving seen data "./dancer.seen"
07:21:04 Quit StealthyXIIGer (Read error: 110 (Connection timed out))
07:25:18 Quit jdgordon (Read error: 110 (Connection timed out))
07:25:31 Join einhirn [0] (
07:25:36 Join tchan [0] (
07:51:05TheSevenUtchybann: ACK, will look into this later
08:06:55 Join DerPapst [0] (
08:10:01 Join ender` [0] (
08:11:45 Quit parafin ("So long and thanks for all the fish")
08:16:49 Join parafin [0] (
08:17:22 Join Bagder [0] (
08:23:13 Quit Prot (Read error: 54 (Connection reset by peer))
08:30:13 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
08:35:37 Join flydutch [0] (
08:39:47 Join martian67 [0] (i=lol3@about/linux/regular/martian67)
08:50:58 Quit TheSeven (Read error: 113 (No route to host))
08:53:55 Join maruk [0] (
08:56:30 Join Rob2223 [0] (
09:01:24 Join petur [50] (n=petur@rockbox/developer/petur)
09:04:12 Join bertrik [0] (
09:04:43 Join mcuelenaere [0] (n=mcuelena@rockbox/developer/mcuelenaere)
09:05:42 Join Hilikus [0] (n=hilikus@unaffiliated/hilikus)
09:05:58 Quit phanboy4 (Read error: 54 (Connection reset by peer))
09:07:14amiconnUgh. Booting rockbox got even uglier now
09:07:25Hilikusare replaygain tags still only supported in id3v2 and not APEv2?
09:08:05amiconnFirst, the statusbar appears at the top of the splash screen. Quite ugly, since the splash screen has a black background but the statusbar is black on rockbox blue (my standard background.
09:08:33amiconnWhen the browser appears, the statusbar disappears for some time, then reappears
09:08:49amiconnThe latter effect is new
09:08:59marukHilikus: if I'm not wrong APE tag are not supported.
09:09:22marukHilikus: you can check the wiki.
09:10:32amiconn(this is on X5)
09:13:44amiconnmaruk, Hilikus: No, APE tags are only supported for .ape files, not for .mp3. This is by design
09:13:57Hilikuscool, that's what i thought
09:14:27marukamiconn: ok.
09:14:35Hilikusyou guys know a replaygain for mp3s that writes tags in id3v2? mp3gain uses apev2
09:15:02 Quit Rob2222 (Read error: 110 (Connection timed out))
09:15:36*amiconn has no idea as he doesn't use replaygain
09:15:42amiconnThe wiki probably lists some
09:17:20***Saving seen data "./dancer.seen"
09:17:35marukHilikus: I use mp3gain than a little script inspired by some tool found on the wiki.
09:17:42 Join jdgordon [0] (n=JdGordon@rockbox/developer/JdGordon)
09:18:10Hilikusbut mp3gain puts them in APE. is the tool to convert them to id3??
09:19:39marukHilikus: yes, it uses eyeD3 to create tag for rockbox.
09:19:41 Join Thundercloud [0] (
09:21:33marukHilikus: look at
09:26:34 Quit Xerion (Read error: 54 (Connection reset by peer))
09:27:07 Join Xerion [0] (
09:34:45Hilikusthanks maraz
09:35:06 Quit jdgordon (Read error: 60 (Operation timed out))
09:36:35 Quit bubsy ()
09:38:35 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
09:39:12 Quit GodEater (Read error: 113 (No route to host))
09:39:47 Join bubsy [0] (n=bubsy@
09:40:00 Quit bertrik (Read error: 148 (No route to host))
09:45:03 Quit Thundercloud (Remote closed the connection)
09:57:14 Quit mt (Read error: 104 (Connection reset by peer))
10:05:06 Quit mcuelenaere ()
10:07:50gevaertsHilikus: I believe that recent versions of mp3gain can write id3
10:15:02 Quit gevaerts (Nick collision from services.)
10:15:12 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
10:18:20Hilikusgevaerts: really? i can't find anything about that
10:18:49Hilikusi can't even find the website of the version i have from the ubuntu repo
10:20:02gevaertsI seem to remember someone saying that, yes
10:20:25gevaertsand have a look at the copyright file in ubuntu, that usually has the download site
10:34:42 Quit DerPapst ("Leaving.")
10:37:11linuxstbHilikus: The manpage for my mp3gain (also Ubuntu) links to - it doesn't seem to support id3
10:37:47Hilikus:( i was checking if the newer version 1.5 did but the website lacks information
10:37:55Hilikusit doesn't even have the name of the developer
10:38:30Hilikusi thought therewere 2 apps called mp3gain cause in ubuntu my version is 1.4.6 and on the website the stable release is 1.2
10:38:34Hilikusbut there's also 1.5 beta
11:02:37 Quit BHSPitMonkey ("Ex-Chat")
11:13:54Hilikusgevaerts: you were right
11:14:12Hilikusi looked at the code and v1.5 can write to id3v2 :)
11:17:24***Saving seen data "./dancer.seen"
11:22:23 Quit Hilikus ("Suerrrrrrte")
11:26:32 Join GodEater [0] (
11:27:40 Join jdgordon [0] (n=JdGordon@rockbox/developer/JdGordon)
11:40:55 Join matsl [0] (
11:48:25 Join MethoS- [0] (n=clemens@
12:01:55 Join DerPapst [0] (
12:07:28 Quit MethoS- (Remote closed the connection)
12:26:53 Quit jdgordon (Read error: 110 (Connection timed out))
12:33:38 Join funman [0] (n=fun@rockbox/developer/funman)
12:40:41 Join mt [0] (n=mtee@
12:54:08 Join AaronM [0] (
13:00:01 Join liar [0] (n=liar@
13:07:22 Join ender [0] (
13:07:49 Quit ender` (Read error: 60 (Operation timed out))
13:13:51 Join ender` [0] (
13:16:20 Quit antil33t (Read error: 104 (Connection reset by peer))
13:17:25***Saving seen data "./dancer.seen"
13:22:30 Join ender1 [0] (
13:27:48 Join [1]ender [0] (
13:27:48***Alert Mode level 1
13:27:48DBUGSent KICK ender to server
13:27:48DBUGSent KICK ender` to server
13:27:48***Alert Mode level 2
13:27:48DBUGsent MODE #rockbox +b *!*i=krneki@*
13:27:48DBUGSent KICK ender1 to server
13:27:48DBUGEnqueued KICK [1]ender
13:27:48***Alert Mode level 3
13:27:49Kick(#rockbox ender :*bang* too many joined users) by logbot!
13:27:49***Alert Mode level 4
13:27:49Kick(#rockbox ender` :*bang* too many joined users) by logbot!
13:27:49***Alert Mode level 5
13:27:49Mode"#rockbox +b *!*i=krneki@* " by logbot (
13:27:49Kick(#rockbox ender1 :*bang* too many joined users) by logbot!
13:27:49***Alert Mode level 6
13:27:50DBUGQ-Sent KICK [1]ender to server
13:27:50Kick(#rockbox [1]ender :*bang* too many joined users) by logbot!
13:27:50***Alert Mode level 7
13:33:01 Join antil33t [0] (
13:35:06 Quit freqmod_qu (Remote closed the connection)
13:35:18 Quit dudacgf ("CGI:IRC")
13:37:51***Alert Mode OFF
13:43:11 Join froggyman [0] (
13:45:29 Part froggyman
13:52:16 Part funman
14:02:29 Join FOAD_ [0] (
14:03:51 Quit FOAD (Read error: 60 (Operation timed out))
14:03:51 Nick FOAD_ is now known as FOAD (
14:05:04 Join teru [0] (
14:05:54 Join Strife89 [0] (n=michael@
14:19:00 Join jdgordon [0] (n=JdGordon@rockbox/developer/JdGordon)
14:21:11 Part Bagder
14:23:30 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
14:25:11peturLinusN: if you saw pondlife's post in the ML you'll know the question: did you write down between what two versions the h300 bootloader started to act up on my h380?
14:27:19LinusNmy memory is weak... :-(
14:27:44LinusNmaybe it's in the irc logs from the devcon
14:28:38petureow :(
14:37:58 Quit jdgordon (Read error: 60 (Operation timed out))
14:43:20 Join bughunter2 [0] (n=bughunte@unaffiliated/bughunter2)
14:47:23 Join freqmod_qu [0] (
14:49:09 Join StealthyXIIGer [0] (
14:49:40 Quit freqmod_qu (Remote closed the connection)
14:55:22 Join freqmod_qu [0] (
14:57:17 Part LinusN
14:59:21 Join TheKiller [0] (n=theuser1@
15:10:38 Quit TheKiller (Read error: 148 (No route to host))
15:17:29***Saving seen data "./dancer.seen"
15:19:28 Quit StealthyXIIGer (Read error: 110 (Connection timed out))
15:23:46 Quit FlynDice (Remote closed the connection)
15:24:46 Join Strife1989 [0] (n=michael@
15:27:29 Join FlynDice [0] (
15:29:10 Join watto [0] (n=watto@
15:30:31 Join Strife89DS [0] (n=nds@
15:30:53 Quit Strife89 (Nick collision from services.)
15:31:23 Quit Strife1989 (Nick collision from services.)
15:32:04 Nick Strife89DS is now known as Strife89 (n=nds@
15:36:50 Join wark [0] (n=ryanw@
15:39:36 Join evilnick [0] (i=0c140464@rockbox/staff/evilnick)
15:45:50 Quit DerPapst ("Leaving.")
15:47:56 Quit wark ("Leaving")
15:54:10 Quit matsl (Read error: 110 (Connection timed out))
16:01:28 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
16:19:00saratogahas anyone tried the clip buffering/playback fix on a HD target? it works great on flash devices but i'd like to know its ok on everything
16:19:12 Quit teru ("Quit")
16:26:04 Join n1s [0] (n=n1s@rockbox/developer/n1s)
16:26:35 Quit tchan ("WeeChat 0.3.1-dev")
16:30:06 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
16:30:14 Join toffe82 [0] (n=chatzill@
16:31:26 Join jdgordon [0] (n=JdGordon@rockbox/developer/JdGordon)
16:38:24 Quit jdgordon (Read error: 60 (Operation timed out))
16:38:42 Join solexx_ [0] (
16:46:03 Quit solexx (Read error: 145 (Connection timed out))
16:46:36 Join Rob2222 [0] (
16:49:24 Quit Strife89 ("ClIRC - IRC client for Nintendo DS")
16:54:49 Join Strife89DS [0] (n=nds@
16:55:54 Join Strife1989 [0] (n=michael@
16:57:22 Quit Strife1989 (Read error: 104 (Connection reset by peer))
16:57:44 Join Strife89 [0] (n=michael@
17:04:41 Quit Rob2223 (Read error: 113 (No route to host))
17:17:33***Saving seen data "./dancer.seen"
17:20:09 Quit Zagor ("Don't panic")
17:23:46 Join Omlet [0] (
17:26:54n1ssaratoga: about those mp4 files that require too much memory to play, how do i know that is why a file isn't playing?
17:27:15saratogaIIRC it cuts in at about 50 minutes of length
17:27:18saratogayou can also check in the sim
17:29:38n1show do i check in the sim?
17:29:42 Join fyrestorm [0] (
17:30:52 Quit freqmod_qu (Read error: 110 (Connection timed out))
17:31:14saratogait should print something if the file is too big IIRC
17:33:48 Join Lss [0] (
17:35:17n1ssaratoga: do you know where the copying is done?
17:38:21CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:38:21*n1s is looking at FS #10790, with the patch the file plays in the sim but apparently doesn't on target, i wonder if it was that memory thing, is the buffer the same size on the sim as on target?
17:41:35saratogan1s: that file is tiny i doubt it runs out of memory
17:41:50saratogai forget where the copy happens, probably in apps/codecs/libm4a
17:43:12n1sthe file is truncated but the sim says it should be about 5 minutes so ou are probably right
17:45:31saratogacan you see where the device stops decoding it?
17:45:44n1sugh, I think i found a possible memory corruption, probably unrelated to this though
17:45:56 Join pamaury [0] (n=pamaury@
17:48:20n1sum, how would i do that?
17:49:35saratogalogf I guess
17:50:34 Join KBH [0] (
17:51:41 Join n00b81_ [0] (n=n00b81@unaffiliated/n00b81)
17:52:54 Join jdgordon [0] (n=JdGordon@rockbox/developer/JdGordon)
17:54:12n1suh, that will have to wait a bit ;)
17:54:55 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:00:55 Join liteone [0] (
18:01:14 Nick liteone is now known as liteone|away (
18:02:58 Join bertrik [0] (
18:03:22 Quit Strife89 ("See ya.")
18:03:35 Quit petur ("work->home")
18:03:45 Nick liteone|away is now known as liteone (
18:05:12liteoneAnyone know when the sansa fuzes will be completely supported?
18:05:34 Nick Strife89DS is now known as Strife89 (n=nds@
18:06:10Torneno, nobody knows.
18:06:14Tornewhen it's done.
18:06:15 Quit jfc (Read error: 110 (Connection timed out))
18:06:44liteoneThats what I guessed.
18:06:44 Nick Strife89 is now known as Strife89DS (n=nds@
18:07:58 Nick liteone is now known as liteone|away (
18:07:59 Quit HBK (Read error: 110 (Connection timed out))
18:08:04 Nick KBH is now known as HBK (
18:08:05 Quit Strife89DS ("Lunch.")
18:09:30 Nick liteone|away is now known as liteone (
18:09:33gevaertsliteone|away: please review the IRC guidelines linked to from the topic
18:10:39liteonegevaerts: sorry... didn't know asking about the fuzes were against topic. I'll review the guidelines again
18:11:34gevaertsliteone: the fuze is entirely on topic :)
18:11:53liteoneI'm confused sorry :|
18:12:05*gevaerts was referring to "do not change your nick when you go away"
18:12:35 Quit AndyIL ()
18:13:01CIA-80New commit by nls (r23666): Avoid possible memory corruption
18:13:23liteoneah ok thanks. I'll avoid it as much as possible
18:14:08n1syay a *666 rev
18:15:25 Quit AEnima1577 ("Leaving.")
18:15:31liteone18:09 < gevaerts> liteone|away: please review the IRC guidelines linked to from
18:16:09liteone sorry, im using shell for first time and have trouble with controls..
18:19:44 Quit liteone ("leaving")
18:27:02 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:31:09 Part n00b81_ ("Leaving")
18:35:09 Join JdGordon1 [0] (
18:39:43 Quit maruk ("Leaving.")
18:40:42 Quit jdgordon (Read error: 110 (Connection timed out))
18:41:39 Join TheSeven [0] (n=theseven@rockbox/developer/TheSeven)
18:42:39JdGordon1(I know I sholdnt be asking this...) has anyone else had experienc with a sbs r wps loading fine in the sim but not on target?
18:42:55shaiDoes a Catalog playlist automatically updates as I load them, with new/removed files from the place I had originally added/created them with?
18:43:16shaiSo how do I update it?
18:43:26shaiRecreate it everytime?
18:43:45JdGordon1yeah, or use the database instead
18:43:46 Join einhirn [0] (
18:44:00JdGordon1or the file browser actually..
18:44:02shaiJdGordon1, How ?
18:44:04JdGordon1anthing but m3u's
18:44:14JdGordon1how what?
18:44:24shaiManually means browser...
18:44:34shaiYou said to use the database instead..
18:44:53shaiI'm talking about a specific folder, for example, that has lots of specific mp3s
18:45:00shaiI update it regularly ..
18:45:12shaiAnd I want to be able to play that directory "on demand" ...
18:45:23shaiSo .. just browse to it and play it every time or use a catalog?
18:45:38 Quit T44 (Read error: 104 (Connection reset by peer))
18:45:46shaiAnd if I use the file browser... what's the point of a catalog?
18:46:54JdGordon1yeah, if you are only using the catalog for a folder, then there is no point
18:46:55shaiBy far... learning Rockbox is the hardest task I've had in a long long time ...
18:47:07JdGordon1just play the first track in a folder and it will load the whole lot
18:47:08shaiSo what is the point of a catalog?
18:47:15JdGordon1have you read the manual?
18:47:22shaiOh yea...
18:47:26shaiover and over again ...
18:47:38 Join Biont [0] (
18:48:03JdGordon1the point of the catalog is... to have quick access to your m3u playlists
18:48:16JdGordon1which not everyone is going to find useful
18:48:32 Join Topy44 [0] (
18:50:08BiontIs it "normal" that after exiting a plugin/viewer, the menu stays in fullscreen instead of going back into the UI viewport defined in the .sbs? Or is there anything I can do about it?
18:50:40shaiJdGordon1, So instead of 5 clicks to play a directory and its content... they go to the playlists folder and play the m3u instead? lazy IMO.
18:51:06JdGordon1Biont: its a known bug
18:51:10JdGordon1dont use plugins :)
18:51:20Lloreanshai: If you aren't aware of all the features of the catalog, maybe you shouldn't whine about it.
18:51:31shaiouch ...
18:51:35Biont:) alright. Just wanted to make sure. I know there's some trouble with plugins and sbs
18:52:26n1sthe point of the catalog afaik is to be able to modify non-current playlists, in a slightly convoluted way
18:52:55Lloreanshai: The playlist catalog allows options for creating and modifying playlists beyond the ability to simply add things to the current playlist, and also ensures that those playlists are saved.
18:53:16 Join Grahack [0] (
18:54:21 Join Hilikus [0] (n=hilikus@unaffiliated/hilikus)
18:54:24Hilikushey guys
18:54:26 Quit linuxstb (Read error: 60 (Operation timed out))
18:54:28Lloreanshai: It even explicitly says that's what it's for in the manual you've read "over and over again". Nowhere does it claim to have functionality like you asked if it had.
18:54:36Biontanother basic question: What is the difference between preloading an image and displaying it right away? The only difference I know of is the possibility to display sub-images.
18:55:14Hilikusi have a question about replaygain tags in mp3s. are they supposed to be in id3v2.3 or id3v2.4 for rockbox to read them?
18:55:22domonokyBiont: or if you want to display this image conditionally.
18:55:37JdGordon1if you always want the image on the screen then you shouldnt preload it
18:57:01Biontdomonoky: Ah, yes
18:57:31 Quit bubsy (Read error: 60 (Operation timed out))
19:01:09BiontOkay, I take it it's not a big issue anyways :)
19:01:18 Join kugel [0] (n=kugel@rockbox/developer/kugel)
19:02:29 Join kugel__ [0] (
19:02:43 Quit kugel (Nick collision from services.)
19:02:49 Nick kugel__ is now known as kugel (
19:05:40 Join AEnima1577 [0] (
19:07:52saratogai wonder if assuming > word alignement in buffering.c is a good idea
19:08:43saratogamatsch says everything is aligned to at least 8 bytes but i don't know if thats intentional or by accident
19:09:03 Join freqmod_qu [0] (
19:12:17saratogaend = (int32_t *)((int32_t)buffer + (int32_t)buffer_len - 4);
19:12:25saratoga warning: cast to pointer from integer of different size
19:12:52 Quit amiconn (Nick collision from services.)
19:12:53 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
19:12:54 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
19:12:54 Quit pixelma (Nick collision from services.)
19:13:13 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
19:13:13 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
19:13:40n1ssaratoga: if that's on 64 bit linux you cast a 32 bit int to a 64 bit pointer
19:14:06 Quit freqmod_qu (Remote closed the connection)
19:14:09saratogan1s: how?
19:14:22n1s(pointer *to* int32_t is 64 bit)
19:14:41 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-eapgubgsclfnihjt)
19:15:16n1smaybe cast buffer to intptr_t
19:17:31 Join freqmod_qu [0] (
19:17:34***Saving seen data "./dancer.seen"
19:17:47saratogahow is that different?
19:17:48 Quit freqmod_qu (Remote closed the connection)
19:18:01 Join freqmod_qu [0] (
19:19:14 Join Thundercloud [0] (
19:19:47n1si mena like: end = (int32_t *)((intptr_t)buffer + (intptr_t)buffer_len - 4);
19:20:22n1sintptr_t is an int with the same size as a pointer
19:21:24saratogawouldn't that mean its 64 bit?
19:22:05n1syes, but if buffer is a pointer on a 64 bit system, casting it to 32 bits is um, unsafe
19:22:59n1sit's just the *address* that's 64 bits, it's still accessed as a 32 bit int due to the cast to int32_t*
19:24:23saratogathat does get rid of the nasty warnings and still works
19:26:07amiconnJdGordon1: Any chance the fugly status bar behaviour will be fixed soonish?
19:26:20JdGordon|which one?
19:26:53 Quit TheSeven (Nick collision from services.)
19:27:13 Join The_Seven [0] (n=theseven@rockbox/developer/TheSeven)
19:27:24 Nick The_Seven is now known as TheSeven (n=theseven@rockbox/developer/TheSeven)
19:28:31JdGordon|well.. If i had more help testing 10709... maybe it would be done already
19:29:14*amiconn thinks we need a bot that quotes tracker entries on request (#quassel does that)
19:33:24 Quit JdGordon1 (Read error: 110 (Connection timed out))
19:34:44*kugel has a script for that, but that's useless if people just write the number without fs before
19:35:48mtsaratoga: Just a check, in the ffts, bitreversals are done imlplicitly, right ?
19:37:11amiconnfs # or just # should probably trigger it (# with only digits after it, obviously)
19:38:13saratogamt: let me check
19:38:21 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
19:38:39 Join einhirn_ [0] (
19:39:22saratogamt: hmm fft_permute doesn't seem to get called
19:39:49mtthat's in ffmpeg's ?
19:40:10UtchybannTheSeven: replacing my logf(USEC_TIMER) by logf(FIVE_USEC_TIMER) fixes the clickwheel erratic behavior!
19:40:30TheSevenreplacing the *logf*?
19:41:46saratogamt: in ffmpeg their test driver calls fft_permute but it looks like their mdct does not
19:41:55saratogaso maybe its folded into the mdct?
19:42:13UtchybannTheSeven: yes. I have place some logf to identify the clickwheel problem. And this morning, I just want to check if FIVE_USEC_TIMER works.
19:42:38saratogamt: yeah its built into the mdct I think
19:43:01TheSevenUtchybann: i observe erratic behavior without any logf...
19:43:06saratogathe prerotation step uses a bit reversed load order, so i suppose they do both at once
19:43:15saratogathats probably why our fft tests look odd
19:43:30TheSevento name it exactly: the wheel jumping one step back or forth as soon as i touch it
19:43:37mtsaratoga: yeah, I thought so yesterday night, but didn't have time to check
19:44:10saratogamt: I wonder if the bit reverse tables aren't the same for the new and old ffts
19:44:17saratogathat would explain why the audio comes out all wrong
19:44:40mthmm .. possible.
19:45:10UtchybannTheSeven: I also have this jumping on my 2 nano with the svn version. Other people seems to have the same issue.
19:45:27saratogamt: yeah the bit reverse tables looks quite different
19:45:43UtchybannTheSeven: so I try to catch to problem and I found that USEC_TIMER is constant.
19:45:43saratogamakes sense since the split radix decomposition works in quite a different order then the radix 2 version . . .
19:45:49TheSevenUtchybann: so i'd say that this wasn't just caused by some logf
19:46:02UtchybannTheSeven: true.
19:46:12mtsaratoga: I didn't know the old fft was a radix-2 one.
19:46:24saratogai'm not sure it is :)
19:46:28TheSevenso is it logf'ing the other thing that fixes it?
19:46:29Hilikusi have a question about replaygain tags in mp3s. are they supposed to be in id3v2.3 or id3v2.4 for rockbox to read them?
19:46:31saratogabut it steps by powers of two
19:46:37UtchybannTheSeven: what I said, is that I tried this patch this morning And the jumping is fixed!
19:46:39TheSevenFIVE_USEC_TIMER and USEC_TIMER are derived from the same reg
19:46:48saratogaHilikus: I think either shold work
19:47:21Hilikussaratoga: you know the name of the tag? i've seen two different tags for replaygain, i don't know which one is the valid one
19:47:41UtchybannTheSeven: maybe, reading FIVE_USEC_TIMER have some action on USEC_TIMER.
19:47:55saratogaHilikus: no you'd probably have to grep the source code to see which are allowed
19:48:14saratogai would just use whichever one the replaygain people say to use
19:48:53TheSevenhm, maybe we need to read the upper part of it for some reason?
19:48:58TheSevenUSEC_TIMER discards it
19:49:00 Join salty-horse [0] (
19:53:05 Quit einhirn (Read error: 110 (Connection timed out))
19:55:09 Quit hillshum (Connection timed out)
19:56:20 Join jfc [0] (
19:57:29mtsaratoga: fft_permute makes a nice difference : !
19:58:27saratogahey that looks like an FFT!
19:58:43mtyep :)
19:58:54saratogaguess we need to replace the permute code in mdct_init_global with the permute code from ffmpeg
19:59:28mtI didn't normalize the frequency .. but the impulse is in the correct place after normalization (1/64)
20:00:25 Quit Thundercloud (Remote closed the connection)
20:01:08saratogalooking at that permutation i can't even tell if it'd be the same on different window sizes
20:03:14Hilikusdoes anyone know of a linux program that will store replay gain tags in id3v2? mp3gain stores them in APEv2
20:03:29mtsaratoga: I don't understand, how is it related to the window size ?
20:07:41AlexPHilikus: I just use foobar under wine for replaygain scanning. There isn't a decent linux option IMO
20:09:21*n1s recomends vorbisgain ;P
20:11:10 Part Llorean
20:12:32gevaertsHilikus: I thought you found that mp3gain 1.5 can do it?
20:13:05Hilikusit does handle it, but it doesn't write the required tag
20:13:20Hilikusit writes and undo tag, so it knows how to deal with id3
20:13:31Hilikusbut there's no replaygain_track_gain tag
20:13:43Hilikusi think that feature is not complete yet
20:13:50 Join Strife89 [0] (
20:13:54Hilikuswrites an* undo tag
20:15:39 Join hebz0rl [0] (
20:15:40 Join Llorean [0] (
20:17:55 Quit flydutch ("/* empty */")
20:20:06wattoHi all, I'm trying to port the rockbox bootloader to the ipod nano 4G.
20:20:10kugelUnhelpful: ping
20:20:14wattoCan anyone give me some advice on getting it to run in TheSeven's Unibugger?
20:22:24kugelHilikus: mp3gain writes APE tags, unfortunately
20:22:45TheSevenwatto: general rule of thumb: for now, leave the interrupt controllers and IRAM alone
20:23:04TheSevenalso don't ever mask IRQs
20:23:05kugelI think mp3gain isn't being actively developed since id3v2
20:23:29TheSevenyour image will be loaded to 0x08000000 with it's entry point at base+0x20
20:24:30CtcpPing from gevaerts!n=fg@rockbox/developer/gevaerts
20:24:48wattoOK, I uploaded the image to 0x08000000 but nothing much happened when it ran
20:25:01TheSevenjust use run <whatever>
20:25:01wattoi could have just done something stupid of course
20:25:23TheSevenand what code did you actually try to run?
20:25:43 Join domonoky1 [0] (
20:25:53wattoI started by copying the nano2G bootloader and changing the lcd register
20:26:14wattohoping to get something on the screen, but as yet nothing
20:27:30wattowhen i examined the disassembly, it looked like there were hardcoded jumps to 0x2200xxxx addresses
20:28:20wattoi thought that seemed wrong since that's where unibugger is located
20:28:55Hilikuskugel: i see
20:29:31Hilikuskugel: do you know of any other program? i could use that script that's around to move them from ape to id3 but i rather not
20:31:11 Quit einhirn_ (Read error: 101 (Network is unreachable))
20:32:29TheSevenwatto: yes, the linker script is all-wrong for an ibugger build
20:32:36 Quit domonoky (Read error: 104 (Connection reset by peer))
20:32:39TheSevenand most of the drivers won't work either
20:32:46TheSeveni would start from scratch...
20:33:01 Join jdgordon [0] (n=JdGordon@rockbox/developer/JdGordon)
20:33:56kugelHilikus: no, I use windows for that :(/
20:34:01 Quit polobricolo (Read error: 110 (Connection timed out))
20:35:08wattoi see. I'm afraid i'm quite a novice at this, could you briefly explain IRAM to me, is it to do with interrupt handling?
20:35:46 Join einhirn [0] (
20:44:09BiontIs there any hope that the screen buffer gets increased anytime soon? at least for the D2. the current size is barely enough for my touchscreen theme, but it will be *very* hard to do more in future themes
20:44:51BiontI'm asking because the D2 is certainly able to handle more than ~200kb worth of images on the WPS
20:45:12JdGordon|its easy enough to change if you can do your own build
20:45:27JdGordon|you're filling the buffer though?
20:46:03saratogamt: sorry was away and am heading back out
20:46:15saratogabut the permutation size depends on how big the fft window size is
20:46:35saratogathe wma code exploits the fact that the old permute was the same at each size except scaled
20:46:58saratogai don't know if that will work with the new one, so some of the logic might need changing (specifically computing 5 different bit reverse tables)
20:47:01saratogaanyway, got to go
20:47:55BiontHm....even if I could change it myself and make my own build, it'd be annoying for all D2 users using my theme to have to use custom builds just because of the small screen buffer :( It's not a big problem now, since I can't add much more atm. But once there are more features to use in a WPS/theme, this is definitely going to be a major problem
20:48:31kugelBiont: you can code efficiently instead
20:48:52*kugel the current skin buffer ought to be enough for everyeone :p
20:48:52gevaertsBiont: first show that we actually do need more, and then we'll talk :)
20:50:19Biontgevaerts: one more image in my theme and bam!
20:50:33gevaertsWe're not going to make it larger because one known theme almost reaches the limit. We might do it if one known theme actually needs more
20:51:02 Join bubsy [0] (
20:51:02BiontI'm not sure how much room for improvement there is code-wise, but I think I already did everything to keep my images as small as possible filesize-wise
20:51:47LloreanBiont: There are plenty of themes for other 320x240 targets, so the buffer selected for that screen size isn't proving to be too large a problem for a lot of people.
20:53:10gevaertsLlorean: the sbs is new since then though
20:53:20BiontLlorean: I know. But a lot of people make very simple themes. I use some popup menus and a custom status bar
20:53:59Lloreangevaerts: This is true. The existing formula for maximum wps size should include a "+ maximum SBS size" equation then/
20:54:50LloreanBiont: Ah, so the problem isn't so much that it's not big enough over all. It's that you want it to be big enough to allow significantly more complexity?
20:55:33BiontLlorean: pretty much, yes.
20:55:45gevaertsBiont: by the way, does the AA still clear after the backlight goes off?
20:55:56gevaertsWhile paused or something like that I mean
20:56:08Biontgevaerts: yes it does. I don't think I can fix that
20:56:23LloreanBiont: Be aware that increasing it is something that will affect everyone, and can reduce the overall experience quality for people not using that extra complexity in their themes.
20:56:28BiontLlorean: you can have a look at the theme here if you haven't seen it before
20:56:47LloreanThe buffer size needs to be balanced so as to offer the majority of theme sizes compatibility, rather than all of them, cutting off before the most outliers.
20:57:12JdGordon|no.. the size should be user/theme dependant
20:57:14 Join mirak_ [0] (
20:57:16JdGordon|it shuoldnt be hardcoded
20:57:36LloreanJdGordon|: Yes, but since that's not an option at the moment since nobody's fixed that, a size does need to be picked still.
20:57:52*Llorean still thinks the size should be "the size of the theme" and theme selection should require reboots.
20:58:05BiontLlorean: most targets wouldn't need that much complexity anyway since there are very few touchscreen targets atm. and I believe most touchscreen targets should be able to handle the increased buffer size, too
20:58:12LloreanAt least, reboots if you're going up, and if you reboot after going down in size, it reduces memory footprint.
20:58:24LloreanBiont: You believe based on what?
20:58:25AlexPBiont: It isn't a question of "handling it"
20:58:44AlexPBiont: It is a question of all of the finite memory used for the WPS can't be used to buffer music
20:59:16pamaurygevaerts: I have a working implementation (at least with libmtp and my sansa e200) for os descriptor. See github commit here: godeater/rockbox/commit/38a48236e44ca484e785d1defae6bc6c0758b30a">
20:59:20AlexPI'm not saying it should/shouldn't be bigger, but e.g. processor speed is irrelevant
20:59:22LloreanAlexP: Which, to be fair, isn't as important on flash.
20:59:30LloreanI mean buffer for audio.
20:59:31AlexPLlorean: Indeed not
20:59:38LloreanBut it still seems to matter based on those tests on the Nano way back.
20:59:42gevaertsHow about we use a fixed "normal" size (possibly smaller than we have now), and if the selected theme doesn't fit, reboot?
20:59:47 Quit jdgordon (Read error: 110 (Connection timed out))
20:59:51AlexPLlorean: But handling it sounds like it is based on how fast the target is
20:59:58LloreanAlexP: True.
21:00:07AlexPAs being touchscreen makes no difference to the RAM size
21:00:55 Part watto
21:00:57BiontI think I'm lost now. I though the buffer size is about RAM. Why are we tralking about how fast touch targets are?
21:00:59Lloreangevaerts: In that case why not shrink on reboots too if a smaller theme is selected and a manual reboot happens?
21:01:11LloreanBiont: Because you were talking about "handling it" and being unclear what you meant.
21:01:22AlexPBiont: It is about RAM - why does being a touch screen make any difference to that?
21:01:32LloreanBiont: It's not about how much total RAM there is. It's about how much RAM we're willing to accept being permanently unavailable for other uses.
21:01:33kugelBiont: you waste a lot on the images just for the borders of each
21:01:45gevaertsLlorean: oh yes. I forgot to mention that. Basically store the size of the currently selected theme, and on boot allocate MAX(current theme, standard setting)
21:01:48Bionthm okay
21:01:53AlexPThe touch screen targets have inherently no more or less RAM than any other target
21:02:04Biontbe a bit patient with me :) I'm glad I figured out the wps code^^
21:02:25Lloreangevaerts: Why MAX(those) rather than allowing it to shrink too? We let people select maximum files in browser to free up a little RAM, why not let them select minimalistic themes too?
21:02:49gevaertsLlorean: convenience
21:03:23BiontI just though that most touchscreen devices usually come with pretty huge-ass interfaces which leads me to think that there's enough ram to store them while still being able to play audio
21:03:41LloreanBiont: "enough" and "it's the most efficient option" aren't the same thing
21:04:02LloreanBiont: There's enough RAM to store a far more complex interface on nearly any Rockbox player, but you can usually get more battery life (by far) by putting some restrictions in place.
21:04:20BiontLlorean: Okay I get that
21:04:28AlexPBiont: There is no more RAM on the touch screen ones than the others
21:04:48AlexP(in general terms, there are always exceptions)
21:04:54Lloreangevaerts: Maybe a setting of some sort, though I'm hesitant to suggest more settings.
21:05:10*Llorean doesn't remember how much RAM the D2 has at the moment.
21:06:30JdGordon|just remember than there is no way of knowing how much ram a theme needs without allocating the entire audio buffer and loading it into there first, and then resize the actual skin buffer for it
21:06:41JdGordon|so a max-usage setting is the only one that makes sense
21:07:41LloreanJdGordon|: Weren't you the one who was just saying a few minutes ago that the size should be theme-dependent?
21:07:56kugelbtw, in which format are the bitmaps saved in the buffer?
21:08:02TheSevenwatto: IRAM is internal ram (vs. external dram chips)
21:08:09TheSevenit's usually small but fast
21:08:19TheSevenwhile dram is relatively slow, but huge
21:08:23JdGordon|Llorean: yes
21:08:58kugelin the native format or in the format that they are on the disk?
21:09:05TheSevenso iram is where ibugger itself lives, dram is where the code to be debugged is running from
21:09:12JdGordon|kugel: native
21:09:26*bertrik is still not completely sure that we really need to maximise audio buffer RAM for flash based targets
21:10:23LloreanJdGordon|: So how does the theme determine the size work with the "a max-usage setting is the only one that makes sense"? I don't quite understand what you're saying.
21:10:24gevaertspamaury: I'm not entirely sure if the device-level get_ms_descriptor() should be outside of usb_core.c, but I really can't make up my mind about it, so just ignore that. For the rest I think that implementation looks good
21:11:14Lloreanbertrik: It's less important to, at least.
21:11:26kugelJdGordon|: so, all bitmaps of that fidelity theme are 174K. and he says the theme uses 175k
21:11:38JdGordon|Llorean: yes they dont work entirely well together
21:11:48Lloreanbertrik: But even if you don't need it for buffering audio, some plugins and other things will still need to claim it, so having as much as possible available is still good
21:11:56LloreanJdGordon|: Okay. That's fine then. I was just confused. :)
21:12:07JdGordon|I'm just saying the problems
21:12:19JdGordon|I tihnk upping the static usage arbitrarily is silly
21:12:21kugelis the rest of the data really 60k?
21:12:53JdGordon|60K sounds way too high, but plausable
21:12:57LloreanJdGordon|: I think upping the static usage in a formulaic way (to handle SBS) makes sense, but it should just be a variation of the normal WPS formula.
21:12:58pamaurygevaerts: what exactly do you mean ? You think usb_core could handle a little bit more of the request (like setting up the header) ?
21:13:01bertrikLlorean, agreed
21:13:32LloreanJdGordon|: Did the buffer size increase for SBS (or does it use its own buffer?)
21:13:37*Llorean does not know these things.
21:13:50JdGordon|it hasnrt been touched
21:14:00BiontIs there anything I can contribute to the discussion?
21:14:07JdGordon|and beer
21:14:20AlexPbeer would be good
21:14:20 Join Thundercloud [0] (
21:14:52gevaertspamaury: no. I mean look at wValue
21:15:08gevaertsto see if it's device or interface, and handle it in core if device
21:15:12BiontI'd love to give a beer to each of you for developing rockbox, that's for sure^^
21:15:17gevaertsagain, I'm not at all sure about this
21:17:38***Saving seen data "./dancer.seen"
21:18:34pamauryAh yes, perhaps it can be implement by each get_ms_descriptor: the current implementation continues to loop if get_ms_descriptor returns 0. But again, I'm not sure MTP was really thought for multi-interfaces devices and at least libmtp doesn't handle it well. Perhaps it's not the case with Windows. I'm not sure about this too.
21:18:57JdGordon|Llorean: actually they could work together... if the theme has a config value saying how much it uses, and its hiegher than the users setting we could ask if they actually want to load it
21:18:59 Join petur [50] (n=petur@rockbox/developer/petur)
21:19:32gevaertspamaury: oh, another thing. usb_ch9.h has a typo, onstant instead of constant :)
21:19:46pamauryah thanks :)
21:21:32*Biont is afk for a cigarette
21:22:16LloreanJdGordon|: Is how much a theme uses consistent between Sim and Target?
21:24:22 Join StealthyXIIGer [0] (
21:24:31 Quit hebz0rl ()
21:26:48*Llorean still thinks a lot can be improved by a _good_ checklist for each target for testing.
21:27:07 Quit bertrik (Remote closed the connection)
21:28:40 Join bertrik [0] (i=54a0d7b4@gateway/web/freenode/x-dqvegyjapgdwlzsx)
21:28:58 Quit bertrik (Client Quit)
21:29:04 Join bertrik [0] (
21:31:47*Biont is back
21:32:47JdGordon|Llorean: in theory yes... there may be very slight differences though
21:33:23bertrikLlorean, I like the checklist idea
21:34:37BiontJdGordon|: remember my thread about the custom actions/tags/tokens/whatever? I'd love to use them to turn my fidelity-theme into a monster, but that won't work with the current skin buffer size
21:34:50LloreanJdGordon|: I was going to ask if the "Save theme settings" thing could also include theme size, but if it might be different between Sim and Target, then a theme could report less size than it needs (also multi-target themes with different native bitmap formats?)
21:35:18JdGordon|Biont: yes, I remember... its on my todo list (or you could do it if you really wanted :) )
21:35:24evilnickLlorean: +1 for the checklist idea
21:35:49*gevaerts agrees
21:35:51Lloreanbertrik, evilnick: It's something I suggested before, but I think I didn't make it clear how it differed from our "release checklist"
21:35:55LloreanAnd of course someone has to do it.
21:36:12LloreanBut I think the big problem facing our releases is that when day 0 comes we really have little idea what has or hasn't even been looked at.
21:36:26JdGordon|I'm nto sure what checkwps does, but it could be used to get the exact value for the real target
21:36:32gevaertsIt's very similar with what I had in mind (not necessarily what I *said*) with the report templates
21:36:44BiontJdGordon|: believe me, I did look at the code and set up a dev environment to see if I have *any* clue. But I can't code and I don't think I'll learn it soon enough to be of any help
21:36:57evilnickLlorean: Yes, which gevaerts' idea addresses, so we can have more confidence in the releases
21:36:58LloreanJdGordon|: Still, themes that work on multiple targets are probably a problem. Maybe the maximum possible size used by a theme?
21:37:02JdGordon|thats the wrong attitude :)
21:37:23Lloreangevaerts: I thought it might be, which is why I wanted to get to work on an idea for the content of those report templates. :)
21:37:35 Quit Thundercloud (Remote closed the connection)
21:37:53 Join froggyman [0] (
21:38:12gevaertsLlorean: I was mostly inspired by the day I spent before the 3.0 release going through my targets and checking every single plugin keymap against the manual :)
21:39:25gevaertsWe now also have so doing codec checklists isn't that hard either
21:40:58kugelthe size of a theme can be sufficiently approximated before loading it
21:41:21BiontLlorean: What's the problem with (optionally) restricting specific themes to specific targets?
21:42:25AlexPWhy would we want that?
21:42:38bertrikwhat would be nice would be some kind of matrix of features vs players with color indicating how well some feature has been tested and by whom, to show coverage
21:42:59AlexPIf they work on multiple targets and the only issue is buffer size, it seems not right to limit them
21:43:11gevaertsbertrik: let's start by actually collecting the data :)
21:43:32bertrikit would be nice too if we know who tested what, so we can thank them in the release announcement
21:43:36AlexPbertrik: You know full well that if we collect some data gevaerts won't be able to resist graphing it :)
21:43:40JdGordon|kugel: how?
21:43:46JdGordon|everything is malloced remember?
21:43:49gevaertsAlexP: only if rasher doesn't get there first!
21:45:06BiontAlexP: If 2 targets have a 320*240 screen but only one of them uses a touchscreen, there's no point in restricting the touchscreen device (I tend to believe that you need more screen buffer size for a good touch theme)
21:45:18AlexPtouchscreen is different
21:45:27kugelfilesize of all images + ~10k for the rest should be fine
21:45:40AlexPbut we were talking generally about all the other targets, many of which share screen size
21:45:47JdGordon|are images always smaller in ram than on disk?
21:46:01kugelnot always
21:46:03BiontAlexP: Okay then maybe add another theme section for touchscreen targets and allow more screen buffer for them or whatever... :/
21:46:03JdGordon|smaller or the same even
21:46:34JdGordon|so checkwps could do it
21:46:50JdGordon|otherwise, you still have to open the skin before allocing the buffer
21:46:52BiontAlexP: I'm just making up stuff now. It's just that I think that touchscreen WPS should be able to use a bigger buffer
21:47:10JdGordon|I agree that touchscreens would need more buffer than non
21:47:11kugelcheckwps could even do it accurately (if run on a 32bit machine I assume)
21:49:23kugelbut you don't need to load the images before making the guess. Parsing it, and then get the filesize of the images in the image-linked list before loading them
21:49:53 Nick Strife89 is now known as Strife1989 (
21:50:18JdGordon|the images arnt loaded untill the end anyway
21:50:52JdGordon|but you cant resize the used buffer once its started parsing already
21:51:04JdGordon|unless we get rid of all the pointers
21:51:06kugelyou could even get the depth at parsing time, then you could determine it accurately before loading too
21:51:16 Join Strife89DS [0] (
21:53:41 Quit thegeek (Read error: 104 (Connection reset by peer))
21:53:51 Join thegeek [0] (
21:55:02 Join hd [0] (
21:55:32 Join ehntoo [0] (
21:55:47 Quit Strife1989 ("If you hold a Unix shell to your ear, you can hear the C.")
21:56:11 Quit Grahack ("Leaving.")
21:57:00 Join Zambezi_ [0] (i=Zulu@
21:57:10 Quit FOAD (
21:57:10 Quit antil33t (
21:57:10 Quit Rondom (
21:57:10 Quit HellDragon (
21:57:10 Quit Zambezi (
21:57:10 Quit blithe (
21:57:10 Quit AlexP (
21:57:10 Quit dionoea (
21:57:10 Quit shaggy-h (
21:57:10 Quit Utchybann (
21:57:10 Quit jasio (
21:57:10 Quit Hadaka (
21:57:10 Quit J-23 (
21:57:39NJoinAlexP [0] (n=alex@rockbox/staff/AlexP)
21:58:24 Join Strife89 [0] (
21:58:58NJoinFOAD [0] (
21:58:58NJoinantil33t [0] (
21:58:58NJoinRondom [0] (
21:58:58NJoinblithe [0] (
21:58:58NJoindionoea [0] (
21:58:58NJoinUtchybann [0] (
21:58:58NJoinHadaka [0] (
21:58:58NJoinJ-23 [0] (
21:58:58NJoinjasio [0] (
21:58:59 Quit Hadaka (Remote closed the connection)
21:59:09 Join Naked [0] (
21:59:19 Nick Naked is now known as Hadaka (
21:59:37 Quit blithe (Remote closed the connection)
21:59:47 Join blithe [0] (
22:00:33 Nick Zambezi_ is now known as Zambezi (i=Zulu@
22:10:28 Quit J-23 (Read error: 110 (Connection timed out))
22:16:33 Join J-23 [0] (
22:17:33froggymanBiont: I would imagine that freenodes webclient is having problems, or some other web client
22:17:38 Quit jasio (
22:17:38 Quit antil33t (
22:17:38 Quit Utchybann (
22:17:38 Quit FOAD (
22:17:38 Quit Rondom (
22:17:38 Quit dionoea (
22:17:59Biontfroggyman: It does look like that :/
22:18:07B4gderor just freenode itself I'd say
22:18:38froggymanboth are possible :)
22:18:57B4gderright, but the netsplits we see are the network, not a client
22:19:23NJoindionoea [0] (
22:19:40NJoinFOAD [0] (
22:19:48NJoinRondom [0] (
22:20:02Biontjust when you thought you're getting somewhere.....^^
22:20:18*froggyman gets a global notice saying that one of their sponsors is having routing problems and that ~3000 users are affected
22:20:28*gevaerts glares at froggyman
22:21:26NJoinantil33t [0] (
22:21:26NJoinUtchybann [0] (
22:21:26NJoinjasio [0] (
22:21:34 Join jasio_ [0] (
22:23:27CIA-80New commit by Domonoky (r23667): make checkwps able to check .rsbs files.
22:24:10 Quit einhirn (Read error: 104 (Connection reset by peer))
22:25:27*Biont is playing around with the /me command
22:25:28domonoky1rasher: could you take a look at FS #10794 (patch for theme site to also check sbs files). please ?
22:26:47saratogaBagder: did laurent get SVN access?
22:27:25B4gderyes he did
22:28:02B4gderroughly 24 hours ago
22:28:05Biontby the way: where can I get those nifty "rockbox artist" badges for the forums? just curious
22:28:18saratogasounds good
22:34:02 Quit jasio (Read error: 113 (No route to host))
22:34:25LloreanBiont: I used to hand them out. My plan was to revamp the entire badge setup, and so stopped. I'm not involved with that side of things any more, so I don't know what the current policy is either way.
22:35:44BiontLlorean: okay XD No problem, thank you
22:36:09 Join MethoS- [0] (n=clemens@
22:36:50AlexPWe are basically paused
22:37:01 Part AEnima1577
22:39:42gevaertsAlexP: is the hold switch turned up though? ;)
22:39:43 Quit Hilikus ("Suerrrrrrte")
22:39:53AlexPgevaerts: Eh?
22:40:02*AlexP no comprenno
22:40:14LloreanAlexP: There's a baffling forum post.
22:40:34LloreanAlexP: "is there any way to disable pause from turning the hold switch up (in sansa clip build)?"
22:40:55 Join jdgordon [0] (n=JdGordon@rockbox/developer/JdGordon)
22:41:06Lloreangevaerts: Maybe it's pausing on headphone unplug or something, even while hold is on, and he wants to disable that?
22:41:34AlexPAh, I see the post - I've asked him to rephrase
22:42:02LloreanWell, I mean I'm pretty sure it's "can I prevent pause by turning on hold" but turning on hold *should* prevent pause in most circumstances
22:42:52LloreanOooh, right, Clip
22:43:01LloreanIsn't the Clip glitch sometimes presenting itself as randomly pausing audio?
22:44:02saratogayeah i read that post like 5 times without being able to parse any sort of meaning or idea
22:45:48LloreanIf it said "is there any way to disable pause by turning the hold switch up" it'd be more or less fine. I think it's just a language/translation issue.
22:46:29 Join LambdaCalculus37 [0] (n=LambdaCa@rockbox/staff/LambdaCalculus37)
22:46:34saratogawell yeah thats grammatically correct but it still doesn't make any sense
22:49:14 Join antil33t1 [0] (
22:49:20 Quit antil33t (Nick collision from services.)
22:52:06bertrikI think what is happening is that he pushes the power-button/hold slider up too far when unlocking hold, hitting the power-up button that way, causing playback to stop (not pause)
22:52:40bertrikand he could be confusing stop and pause
22:54:08kugelI assume he's accidentally pushing the switch to far
22:54:13kugelwhen he hits power the playback stops
22:54:28kugelI think the clip doesn't have the mechanism to disable power for 1s after hold as been released
22:54:44AlexPIt seems to
22:55:04AlexPI just tested, and going deliberatly from hold to pushing up to off, it continues playing
22:55:39saratogamine stops if i do that
22:55:40kugelit doesn't here, it stops
22:55:46Unhelpfulkugel: pong?
22:55:47AlexPI can push it up once or twice more as well if I do it quickly before it finally works
22:55:58kugelUnhelpful: I got 4.4.2 to compile
22:56:05Unhelpfulkugel: oh? how?
22:56:16kugelsubscribing to the gcc-help ml turned out fruitful :)
22:56:20AlexPThat's rather odd, I have SVN from yesterday (plus v1 of saratoga's patch)
22:56:33AlexPAnd it definately doesn't stop
22:57:12kugelUnhelpful: well, I think it's still a bug anyway, but apparently you aren't really supposed to do just make when going for bare metal
22:57:12Unhelpfulkugel: interesting, i searched and found stuff on there but it was all along the lines of "you need to provide a system library, or a board library if you're building for bare metal"
22:57:54kugelthere are 2 make targets that compile *only* gcc and libgcc each
22:58:01AlexPHmmm, it does occasionally stop, but most times it doesn't
22:58:04kugelwithout all the lib blob
22:58:20Unhelpfulkugel: eesh... we should probably *always* do that, save some time ;)
22:59:10kugelmake all-gcc all-target-libgcc to compile, and make install-gcc install-target-libgcc to install
22:59:16 Part froggyman
22:59:29kugelobviously those are badly named and documented
22:59:55saratogaits matsch's patch
23:00:02saratogaand it wont' change how that works
23:00:02kugelI already tried, it still pulls in the unwind shit
23:00:05saratogaso its got to be something else
23:00:22AlexPsaratoga: Yeah, I know it won't change it, I just mentioned it for completeness
23:00:23saratogabut imo its probably worth putting something in the button driver to disable power for 1 second after hold
23:00:37kugelthat's what the fuze does
23:00:55saratogaalso, making select still ignore usb insert even if other buttons are accidentally pushed
23:01:09saratogaits rather hard to get into charge mode on the clip due to the small surface area to button ratio
23:01:17AlexPthat also is a good plan
23:01:44kugelthat button is not very fat finger friendly indeed (although I have no problem :p )
23:02:05kugelUnhelpful: I have a patch for configure to add eabi switches
23:02:27Unhelpfulkugel: what do you mean?
23:02:57kugeltools/configure −−eabi/−−no-eabi
23:03:08kugelI think that's going to be useful for the transition phase
23:05:14Unhelpfulwhat's wrong with changing the target to arm-elf-eabi? that also makes it easy to keep both compilers handy...
23:06:02 Join Thundercloud [0] (
23:06:39kugelbeing able to switch the compilers easily
23:07:09kugel(using eabi while it's not default or using the old one while eabi is still "experimental")
23:08:05 Join Tomis2 [0] (n=Tomis@
23:09:12 Join matsl [0] (
23:09:12kugeljust changing to -eabi doesn't make it easy to keep the old compiler handy IMO
23:10:14 Quit pamaury ("exit(*(int *)0 / 0);")
23:11:14Unhelpfulcouldn't we just make libgcc multilib for both ABIs and use -mabi?
23:12:51kugelwhy should we do that?
23:13:18kugelit's not like switching two 4.4.2 (non-eabi) is a safe thing on its own
23:14:23CIA-80New commit by mcuelenaere (r23668): CheckWPS: set screen to SCREEN_MAIN when not a r{sb,wp}s. ...
23:15:39 Quit Tomis2 (Read error: 60 (Operation timed out))
23:15:58 Quit Tomis (Success)
23:16:00kugelI don't think we would switch to 4.4.2 if we weren't after eabi
23:17:18kugelalso, we need 2 binutils versions. I think those cannot be build with multilib, so they would override each other if they were for the same target
23:17:42***Saving seen data "./dancer.seen"
23:18:32kugelUnhelpful: I plan to add eabi to, so that we can get some other people to compile eabi builds
23:19:15*Unhelpful really can't see how this is better than just switching to arm-elf-eabi
23:19:57kugelyou think we don't need a transition phase?
23:20:54domonoky1think about all those build-servers.. we need a transition phase :-)
23:21:41domonoky1the best would be a way to have both versions available, and be able to choose which to use via configure.
23:21:56kugelwhich is what I'm proposing :)
23:22:22Unhelpfulwell, pastebin your patches, i'll look at them tonight. :)
23:22:49saratogai assumed we'd just svn up the configure script and kill any build servers which didn't have the right version
23:25:07kugelUnhelpful: they seem trivial enough to commit as is
23:25:59kugelsaratoga: that's probably going to happen
23:26:41 Quit domonoky1 (Read error: 104 (Connection reset by peer))
23:29:07 Join DerPapst [0] (
23:31:22 Join fdinel [0] (
23:33:28 Quit matsl (Remote closed the connection)
23:34:07 Join Tomis [0] (n=Tomis@
23:42:12 Quit n1s ("Lämnar")
23:45:20 Quit Omlet ("( :: NoNameScript 4.22 :: )")
23:47:16CIA-80New commit by kugel (r23669): Add −−eabi to let configure prefer the arm-elf-eabi-* toolchain. Add −−no-eabi to achieve the current behavior (which is still default).
23:53:17 Quit DerPapst ("Leaving.")
23:56:50 Quit bertrik (Read error: 60 (Operation timed out))

Previous day | Next day