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 2009-06-19

00:00:23 Join Rob2222 [0] (
00:00:31Bagderbluebrother: done!
00:00:47bluebrotherBagder: thanks
00:02:23Bagderdevcon day started!
00:04:31Bagderall files are uploaded to the download mirror master now
00:06:08 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
00:08:16CIA-70New commit by bagder (r21337): 1) use buildv- as prefix for the dirs this creates while building so that it ...
00:08:17 Quit Programan (Read error: 110 (Connection timed out))
00:08:24 Quit bmbl ("Bye!")
00:09:17Nico_Pfunman: are you still editing the release notes ?
00:09:39 Join stripwax [0] (
00:09:59funmanyes i only removed strike/red and fixed a "bugfixes since 3.3" into 3.2
00:10:10funmanyou can edit now
00:10:41Nico_PI am told you still have the write lock
00:10:45 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
00:10:51 Part safetydan ("Leaving.")
00:11:25 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
00:11:49funmani properly cancelled instead of hitting previous, sorry
00:12:12Nico_Pit works now :)
00:12:23 Nick fxb__ is now known as fxb (
00:12:26CIA-70New commit by bagder (r21338): s/echo -e/printf to make it run more portably
00:13:40 Nick fxb is now known as fxb__ (
00:16:11 Quit robin0800_ (Read error: 110 (Connection timed out))
00:17:20CIA-70New commit by bluebrother (r21339): Initial script to automate rbutil release builds.
00:18:13*Nico_P sees 77 closed bug tasks since 3.2 in the tracker
00:22:57 Part pyro_maniac ("Leaving.")
00:24:05 Quit Zagor ("Clint excited")
00:31:34 Quit bluebrother ("zzz")
00:33:05CIA-70New commit by funman (r21340): FS #10306 by Thomas Martitz : button light doesn't change on SD transfers on Sansa AMS (Fuze & e200v2)
00:33:25 Join swears [0] (n=swears1@
00:33:40swearshey, anybody here using Gigabeat S-10?
00:33:47swearsi'm looking for an updated build to load
00:35:07JdGordon|"Bagder started pondering on the design of the new build system" <- have it so the only interaction with configure is through command line args... no more of this echo | ../tools/configure...
00:35:29*JdGordon| doesnt know if thats in the ideas list or not yet
00:35:40saratogaJdGordon: theres a wiki page for it
00:35:51BagderI consider that specific thing a rather minor detail
00:35:54saratogaswears: you mean S30 or S60? I don't think theres an S10
00:35:56 Quit ender` (" I believe that God left certain drugs growing naturally upon our planet to help speed up and facilitate our evolution. Okay")
00:36:26swearsI had rockbox installed before, but, the charging wasn't working
00:36:31swearsbut now I see that has been fixed
00:36:36swearsso I'd love to put it back on!
00:37:06saratogahave you read the install directions in the wiki?
00:37:25swearsbut you need linux right?
00:37:40swearsto get the right build environment or something?
00:37:44saratogai don't think so
00:38:15 Quit Thundercloud (Remote closed the connection)
00:39:51 Quit loyx ()
00:40:50Mikachuyou can use cygwin or vmware or something
00:41:07saratogaBagder: was the updated beastpatcher for windows ever put on the download servers?
00:41:12Nico_Paren't there more fixed bugs since 3.2?
00:41:57swearsAny word on an actual release coming for the S60?
00:42:01Bagdersaratoga: no
00:42:12swearsI feel like I might mess up my player if I try this myself
00:42:17swearsi had someone send me the files before
00:42:38saratogaif you're not comfortable don't install it
00:42:56*JdGordon| glares at the chan topic
00:43:08*martian67 glares at JdGordon
00:43:24*swears glares
00:43:43***Saving seen data "./dancer.seen"
00:45:34CIA-70New commit by funman (r21341): Apply FS #10306 to Sansa c200v2 (fix build in the same time)
00:46:46 Quit Nico_P ("bed time")
00:47:06 Join n00b81 [0] (n=Taylor@unaffiliated/n00b81)
00:47:07 Join loyx [0] (
00:47:34Bagderfinally is one of the mirrors holding 3.3 files
00:47:49stripwaxIn pictr
00:48:00*stripwax triies again
00:48:12CIA-70New commit by bagder (r21342): 3.3 is released
00:48:31 Part n00b81 ("Leaving")
00:48:42JdGordon|I was about to say we should wait an hour or so for all mirros to get the 3.3 files
00:48:58Bagderthey're only two these days
00:49:04stripwaxIn pictureflow on ipod, is it 'by design' that you can't just keep rotating the touchwheel to dial past the cover art? What seems to work is swiping repeatedly in one direction, which is 'kinda' like swiping the ipod touch screen.. but I'm wondering if that is due to a bug or deliberate..
00:49:13*stripwax hoorays at 3.3
00:52:32loyxsweet 3.3
00:52:43loyxand they fixed charging on ipod 4g.. i think
00:53:49JdGordon|go they!
00:54:11Bagder <= some first design thoughts, more tomorrow in another country
00:54:31 Part swears
00:56:22*funman sends the announces
01:02:44robis the 3.2 -> 3.3 upgrade worth doing
01:03:08JdGordon|well.. at the very least.. its 3 months worth of fixes
01:03:16funmanrob: sure, that's the point of releasing
01:03:25*rob starts upgrading
01:03:44CIA-70New commit by funman (r21343): Sansa AMS: fix yellow for models with buttonlight & without multivolume
01:09:43 Quit PaulJam_ (Nick collision from services.)
01:09:46 Join PaulJam [0] (
01:16:05lilltigerstill loonging to the day i can try out Rockbox :P
01:19:33 Quit stripwax ("")
01:20:56 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
01:21:02 Quit mt ("ChatZilla 0.9.84 [Firefox 3.0.11/2009060215]")
01:22:32 Quit markun ("sleep")
01:28:58funmanbertrik: can you try this on c200v1 ?
01:29:17 Join bmbl [0] (
01:29:20funmanif it still doesn't improve i'll just revert and use the old code
01:29:31 Quit bmbl (Remote closed the connection)
01:30:01 Join thegeek [0] (
01:31:02 Join robin0800 [0] (
01:31:03 Quit hillshum (Remote closed the connection)
01:32:53 Quit robin0800 (Remote closed the connection)
01:33:09 Join robin0800 [0] (
01:34:05 Join hillshum [0] (n=quassel@unaffiliated/hillshum)
01:35:42loyxis there a way to build rockbox without building the plugins
01:35:50JdGordon|make bin
01:35:54loyxoh ok
01:35:58JdGordon|that will only do the main binary
01:37:18Mikachuyou can also just press ctrl-c after you see it creates rockbox.yourtarget
01:39:05Unhelpfulstripwax: that sounds very odd, pictureflow uses the scroll events on scrollwheel targets.
01:39:19Mikachuit worked more or less fine for me when i tried it on my nano
01:41:55Unhelpfulit works great on my e200, but those have a physical wheel.
01:41:57 Quit robin0800 (Remote closed the connection)
01:42:13 Join robin0800 [0] (
01:43:10 Quit bertrik (Read error: 113 (No route to host))
01:46:20 Join funman_ [0] (n=fun@
01:48:10 Quit funman ("Reconnecting")
01:55:56 Quit robin0800 (Remote closed the connection)
01:56:10loyxbattery drains while plugged in via usb
01:56:18loyxthats only while using the ipod g4 though
01:56:25Mikachuweren't ipod charging changes disabled for the release?
01:56:42loyxthey were?
01:56:57loyxit does say on my ipod that a charger is plugged in
01:57:06 Nick funman_ is now known as funman (n=fun@
01:57:08 Join robin0800 [0] (
01:57:54loyxhmm got a higher charge while in usb transfer mode
01:58:09loyxwell, to me usb charging isn't really a big deal
01:58:23loyxthe only thing that matters is that ac adapters can charge it
01:58:31 Quit robin0800 (Remote closed the connection)
01:59:11 Quit efyx (Remote closed the connection)
02:01:13 Join robin0800 [0] (
02:02:26 Quit robin0800 (Remote closed the connection)
02:02:47 Join robin0800 [0] (
02:03:02 Quit robin0800 (Remote closed the connection)
02:04:10loyxalso just curious, but how come when i plug in the cord the battery level goes down immedietely
02:24:40 Join mi__ [0] (
02:25:53mi__hi just a couple questions, i am guessing you cannot splice a game controller to your usb connection since it has no usb host on the device correct? I have a sansa e260
02:27:44mi__also since it plays doom are there any other ports of games like comething commander keen style jill of the jungle? gameboy emulation is slow at best (I wasn't expecting much, emulation is always taxing) but doom runs rather well
02:28:13JdGordon|yes and no
02:28:19funmannow you can use your player as a keybioard (USB HID)
02:30:57mi__funman, couldn't you always? maybe i am misunderstanding but the keys act as the hid. I was wondering if you could splice in a controller and have the device driver on a sansa.
02:32:29krazykitmi__, you could if you made the connector right, wrote the code for usb host, added the drivers for the gamepad, and made doom / rockboy use the controller.
02:32:31funmanah ok. you can't, and i don't think it's possible to do
02:32:50funmanif it's possible to use the usb chip as host which is not guaranteed
02:32:57krazykiti was under the impression that the software usb targets could do host
02:33:20mi__one of the toughest things with doom is a lack of strafe since there are no enough keys but an eternal controller.
02:33:22funmani understood that some needed pins could not have been soldered
02:36:54 Quit gregzx ("ChatZilla 0.9.84 [Firefox 3.0.11/2009060215]")
02:37:15 Part wincent ("Kopete 0.12.7 :")
02:39:38 Quit funman ("free(random());")
02:43:09 Quit tvelocity (Remote closed the connection)
02:43:47***Saving seen data "./dancer.seen"
02:43:54 Join Bunia [0] (
02:45:19 Quit jgarvey ("Leaving")
02:46:30 Part Bunia
02:47:29 Join tvelocity [0] (
02:49:04 Join Bunia [0] (
02:49:37Buniais anybody here?
02:49:51Genesishi Bunia
02:50:09Genesisthanks for help guys , my rockbox now run on the cowon pretty well
02:51:16Buniai'm for the first time in IRC xD
02:52:29krazykitBunia, in that case, i recommend you read the IrcGuidelines page linked in the channel topic
02:53:32Buniaok, thanks ;D
03:03:08 Join fdinel [0] (
03:06:03 Quit timc (Read error: 60 (Operation timed out))
03:07:31 Part Bunia
03:09:24 Quit krazykit (Read error: 60 (Operation timed out))
03:10:41 Join killan_ [0] (
03:16:17 Quit killan (Read error: 60 (Operation timed out))
03:20:01 Join krazykit [0] (
03:20:06 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
03:22:13 Join timc [0] (n=aoeu@
03:24:52 Join ArtificialGekko [0] (
03:25:04 Part ArtificialGekko
03:28:22 Join DarkSpectrum [0] (n=ZX@
03:29:30 Join Rob2223 [0] (
03:29:30 Quit Rob2222 (Read error: 104 (Connection reset by peer))
03:29:36 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
03:30:04 Quit killan_ (Read error: 110 (Connection timed out))
03:30:45 Join BlakeJohnson86 [0] (
03:31:04 Quit pixelma ("No Ping reply in 90 seconds.")
03:31:07 Join pixelma [50] (n=pixelma@rockbox/staff/pixelma)
03:43:17 Quit Genesis ("parti fabriquer des IP")
03:43:58 Quit pixelma (Nick collision from services.)
03:44:00 Join pixelma_ [50] (n=pixelma@rockbox/staff/pixelma)
03:44:12 Quit amiconn (Nick collision from services.)
03:44:14 Join amiconn_ [50] (n=jens@rockbox/developer/amiconn)
03:44:20 Nick pixelma_ is now known as pixelma (n=pixelma@rockbox/staff/pixelma)
03:44:35 Join Greek-Boy [0] (n=greek@
03:44:36 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
03:45:28 Quit Greek-B0y (Read error: 54 (Connection reset by peer))
03:46:18 Quit JdGordon| (" ajax IRC Client")
03:47:01 Quit Horscht (Read error: 110 (Connection timed out))
03:48:42 Part DarkSpectrum
03:49:09 Join DarkSpectrum [0] (n=ZX@
04:05:50Topic"3.3 has been released | Please read before speaking: | Please direct offtopic/social chat to #rockbox-community" by ChanServ (ChanServ@services.)
04:06:05Topic"Rockbox 3.3 has been released | Please read before speaking: | Please direct offtopic/social chat to #rockbox-community" by ChanServ (ChanServ@services.)
04:07:29 Quit hillshum (Remote closed the connection)
04:12:31LloreanWe really ought to update the released bootloader for some targets to those used in FS #9955 or similar
04:21:29 Join saratogahome [0] (n=41becb3b@gateway/web/cgi-irc/
04:21:39saratogahomeLlorean: did we just forget to update the bootloaders?
04:22:15Lloreansaratogahome: It's not something we normally do unless there's a specific reason to
04:22:28LloreanIn this case, we ought to for USB enabled targets to bypass the "boot OF for USB" in many of the bootloaders
04:23:01saratogahomeas i recall the plan was to update the bootloaders when USB was enabled
04:23:13saratogahomewas there any discussion as to why we shouldn't?
04:23:48LloreanI don't *think* so.
04:24:20saratogahomeupdating the bootloaders is as simple as copying them to the download server right?
04:25:39LloreanProbably want new sansapatcher binaries with them compiled in too, so the manual install steps are still accurate, but I think that's the only other thing besides uploading new ones.
04:26:15saratogahomewe should do that before people notice
04:29:08LloreanWe should probably have a news post about the need to update the bootloader. Does RBUtil have the ability to let users known their bootloader is out of date?
04:29:40 Join hillshum [0] (n=quassel@unaffiliated/hillshum)
04:29:42saratogahomei'm not sure
04:30:25saratogahomei need to sleep
04:30:55 Quit saratogahome ("CGI:IRC (EOF)")
04:35:49 Quit mi__ ("Leaving")
04:43:49***Saving seen data "./dancer.seen"
04:48:19CIA-70New commit by lowlight (r21344): Clean up the Synaptics touchpad driver.
04:56:03CIA-70New commit by unhelpful (r21345): Core JPEG decoder improvements: ...
05:05:04 Join bubsy_ [0] (i=Bubsy@unaffiliated/bubsy)
05:07:33 Quit bubsy (Read error: 60 (Operation timed out))
05:11:02 Join r0b- [0] (
05:11:26 Quit bubsy_ (Read error: 54 (Connection reset by peer))
05:11:26 Join cool_walking_ [0] (i=cb3b81c3@gateway/web/ajax/
05:12:03 Quit hillshum (Remote closed the connection)
05:12:31 Join hillshum [0] (n=quassel@unaffiliated/hillshum)
05:23:40CIA-70New commit by lowlight (r21346): Philips SA9200: Working touchpad and button lights! Also, improvements to the keymap (still needs work).
05:27:21 Quit krazykit (Read error: 54 (Connection reset by peer))
05:29:04 Quit rob (Read error: 110 (Connection timed out))
05:31:37 Join krazykit [0] (
05:32:44 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
05:34:33 Quit PaulJam (".")
05:39:10 Join low_light [0] (i=ad58bb86@gateway/web/ajax/
05:41:54CIA-70New commit by lowlight (r21347): Philips SA9200: 1) reset all devices in system_init, 2) revert an earlier change in the sd driver, 3) properly detect the charger, 4) add lcd register ...
05:43:16 Quit Horschti ("Verlassend")
05:50:32CIA-70New commit by lowlight (r21348): Philips SA9200: add the UI sim
06:19:42 Quit hillshum (Read error: 104 (Connection reset by peer))
06:24:09 Quit DarkSpectrum ()
06:26:30 Join DarkSpectrum [0] (n=ZX@
06:30:26low_lightBagder: Time to make the build table wider :)
06:30:37low_lightBagder: Please add the Philips SA9200 (normal, bootloader, & sim)
06:30:52low_lightBagder: You could add the 3 Samsung sims too
06:34:29 Quit perrikwp (" ajax IRC Client")
06:34:39 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
06:43:52***Saving seen data "./dancer.seen"
06:46:23 Quit JdGordon ("Leaving.")
06:48:28 Join JdGordon [0] (
06:59:11 Join Zarggg_ [0] (
06:59:22 Quit thegeek (Read error: 104 (Connection reset by peer))
06:59:33 Join thegeek [0] (
07:12:07 Join kugel [0] (i=kugel@rockbox/developer/kugel)
07:14:28 Quit low_light (" ajax IRC Client")
07:15:07 Quit AndyI (Read error: 110 (Connection timed out))
07:15:23 Quit jordoex_ (Remote closed the connection)
07:15:45 Quit Zarggg (Read error: 110 (Connection timed out))
07:20:40 Quit perrikwp (" ajax IRC Client")
07:20:47 Join perrikwp [0] (i=18ac0c41@gateway/web/ajax/
07:24:54 Join markun [50] (n=markun@rockbox/developer/markun)
07:25:47 Join kugel_ [0] (
07:38:32 Join funman [0] (n=fun@rockbox/developer/funman)
07:40:51 Quit daurn (Read error: 110 (Connection timed out))
07:42:03 Nick Zarggg_ is now known as Zarggg (
07:42:56 Quit kugel (Read error: 110 (Connection timed out))
07:49:03 Join daurn [0] (n=daurnima@unaffiliated/daurnimator)
08:06:35 Join funman_ [0] (
08:07:49 Quit funman ("Reconnecting")
08:09:20 Quit JdGordon (Read error: 104 (Connection reset by peer))
08:15:29 Quit safetydan ("Leaving.")
08:21:48 Join Rob2222 [0] (
08:32:56 Quit kugel_ (Read error: 110 (Connection timed out))
08:39:54 Quit Rob2223 (Read error: 110 (Connection timed out))
08:41:03 Quit markun (Read error: 60 (Operation timed out))
08:41:07 Quit funman_ ("free(random());")
08:43:56***Saving seen data "./dancer.seen"
08:50:13 Quit rasher (Read error: 104 (Connection reset by peer))
08:50:53 Join markun [50] (n=markun@rockbox/developer/markun)
08:53:13 Join rasher [0] (
09:03:32 Join robin0800 [0] (
09:05:06 Join ender` [0] (
09:05:26 Join flydutch [0] (
09:06:54 Join jds_ [0] (n=jds@
09:13:12 Nick tmzt_ is now known as tmzt (
09:14:50 Quit ender` (" The latest survey shows that 3 out of 4 people make up 75% of the world's population.")
09:19:01 Quit robin0800 (Remote closed the connection)
09:19:51 Join robin0800 [0] (
09:19:57 Quit robin0800 (Read error: 104 (Connection reset by peer))
09:22:06 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
09:23:53 Join petur [50] (n=petur@rockbox/developer/petur)
09:32:57 Join Thundercloud [0] (
09:33:14webmindno luck with the ipod :(
09:33:20 Join DarkDefender [0] (
09:36:38 Quit DarkSpectrum (Read error: 104 (Connection reset by peer))
09:37:01 Join robin0800 [0] (
09:37:15 Quit robin0800 (Remote closed the connection)
09:37:40 Join robin0800 [0] (
09:37:48 Quit robin0800 (Remote closed the connection)
09:39:58 Quit cool_walking_ (" ajax IRC Client")
09:40:21 Quit rvvs89 (Read error: 113 (No route to host))
09:43:23 Join DarkSpectrum [0] (n=ZX@
09:45:13 Join Grahack [0] (n=chri@
09:48:02 Quit Thundercloud (Remote closed the connection)
09:52:55 Quit MrDuck (Read error: 113 (No route to host))
10:00:08 Join bertrik [0] (
10:11:18 Quit r0b- (Read error: 104 (Connection reset by peer))
10:19:45 Nick zitune[afk] is now known as zitune (n=zitune@bearstech/zitune)
10:26:10CIA-70New commit by unhelpful (r21349): 2-point and 1-point JPEG IDCT ARM assembly, remove comment in jpeg_load.c about inline asm, change loop condition to be a bit safer in case of bad ...
10:26:18linuxstbwebmind: What's the problem?
10:26:47*GodEater sees we can leave the channel in linuxstb's capable hands whilst devcon happens
10:31:42 Join einhirn [0] (
10:34:46 Nick zitune is now known as zitune[afk] (n=zitune@bearstech/zitune)
10:43:50 Join efyx [0] (
10:43:58***Saving seen data "./dancer.seen"
10:47:48webmindlinuxstb, well, I changed the microdrive in my ipod mini
10:48:00webmindlinuxstb, I can copy the mini's mbr
10:48:12webmindbut that's partition table is awkward
10:48:22webmindcfdisk doesn't recouldnize the sda1
10:48:37webmindI can't resize sda2
10:49:02webmindI supose I could delete sda2 and replace it with a new one starting at the same position
10:49:16webmind(which wouldn't work in cfdisk I think)
10:50:12linuxstb"fdisk" works for me - but the original fdisk, not the newer one (from GNU I think)
10:50:14gevaertswebmind: replacing it works, or resizing it with parted. You can also just have a good look at it and make an entirely new one that has a first partition that's big enough
10:52:58webmindgevaerts, parted crashed on it
10:54:10webmindnot for resizing
10:54:27gevaertswebmind: I would clear it and make a new one. The only important thing is to have the first partition of type 0, and big enough to hold the Apple firmware
10:57:46webmindgevaerts, thanks
10:59:21 Join MrDuck [0] (
11:01:12 Join n1s [0] (n=n1s@rockbox/developer/n1s)
11:02:34webmindnow rockbox won't recouldnize a partition
11:02:41webmindmakes sense actually :)
11:03:10webmindwrong type
11:03:55webmindbut select+play won't put it in disk mode :(
11:08:24LloreanHave you done that before? Disk mode is practically impossible to disable.
11:11:28 Join Zagor-cgiirc [0] (n=5ae89de2@gateway/web/cgi-irc/
11:14:48 Quit Zagor-cgiirc (Client Quit)
11:20:06 Quit DarkSpectrum (Read error: 110 (Connection timed out))
11:22:15 Quit Lss (Read error: 104 (Connection reset by peer))
11:23:34 Quit intrados_ (Read error: 104 (Connection reset by peer))
11:27:10 Join matsl [0] (
11:28:08 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
11:30:06 Join intrados_ [0] (
11:46:47 Nick zitune[afk] is now known as zitune (n=zitune@bearstech/zitune)
11:49:37 Join robin0800 [0] (
12:03:49 Join fml [0] (n=4fd3ff25@gateway/web/cgi-irc/
12:04:51fmlWould anybody object to the following patch: ? It imroves code readability IMO. No RAM or binary changes (and of course no functional changes).
12:07:23fmlAlso, I'd like to change the speed display in the pitch screen to show the real playback speed and not the internally used factor. It's been discussed here or in the mailing list. But IIRC, no agreement has been achieved.
12:08:51gevaertsfml: I'd go for both the real speed *and* the factor
12:09:50gevaertsI seem to remember that someone was going to work on this though, so maybe you can coordinate
12:09:58fmlgevaerts: would you show three values? Pitch, Speed and the factor?
12:10:43gevaertsfml: I'd just add the "real speed" to all modes, yes
12:10:52fmlIMO the factor is just a technical detail, it's not something that is normally used to describe a sound property
12:13:26fmlgevaerts: what is "real speed"? The speed at which the file is played back? I.e. if the song would normally take 1 min to play but now it takes 2 min to play then the real speed is 50%, right?
12:13:53fmlNote that this does not depend on the pitch
12:14:29 Quit matsl (Read error: 110 (Connection timed out))
12:15:10gevaertshm, do we want the relative speed, or the relative time?
12:15:15*gevaerts isn't sure
12:15:15fmlIn the old pitch modes, we change both pitch and speed. In the new mode, they are changed separately.
12:15:18Llorean"speed" should be what the user sees. if it says 100%, the played time should be going down at 1 second per second, no matter what happens behind the scenes.
12:15:54LloreanSpeed should basically adjust the rate at which the progress bar moves, pitch should adjust what the user actually hears without changing that progress bar movement rate, and everything else should be "behind the scenes" in my mind
12:16:20fmlLlorean: this my view as well. But it's not the case in the current SVN
12:16:35LloreanI know
12:16:45*gevaerts probably agrees
12:16:58LloreanBasically, the values should represent what the user can observe.
12:17:09fmlLlorean: so a patch would probably be accepted (if implemented properly)?
12:17:17LloreanI'd be all for it.
12:17:36fmlOk, I just wouldn't like to waste time :-)
12:17:36LloreanHow much RAM does enabling timestretch cost?
12:18:31fmlLlorean: how can I see it? I only have the statically produced rb-info.txt. Everything else goes at the cost of the audio buffer IMO.
12:18:54gevaertsLlorean: 64K according to flyspray
12:19:12fmlSo the question is probably how much of the audio buffer is spent for the feature
12:19:25Lloreangevaerts: Maybe it should just be always on.
12:20:13LloreanIt might be possible to reclaim it later with improvements to how we allocate memory for buffering, etc, but right now having to enable it and then also switch modes makes it kinda confusing to use.
12:20:21Llorean*reclaim it when not in use
12:20:31gevaertsLlorean: being able to get bits of the audio buffer without stopping playback would be really useful for this sort of thing
12:20:46LloreanEspecially tiny bits
12:21:27gevaertswell, tiny... this is 1/32 of the total RAM on the clip :)
12:21:34Unhelpfulmalloc ;)
12:21:37LloreanOnly HD targets count. :-P
12:21:51gevaertsUnhelpful: malloc is a word, not an implementation :)
12:22:27fmlWhat about the patch above? ( Would I be sentenced for committing it?
12:24:47Unhelpfulactually, there are some buflib features i want to see in the core buf* functions, most notably the ability to compact the buffer after some items are bufclose'd, freeing contiguous space at the end, and to find the "last" buffer item for audio data (the one latest in the playlist of those buffered, i suppose), and either close or truncate it to make space.
12:25:14Lloreanfml: I don't see that as really improving readability too much. As well "midheight" being "height * 2" seems odd to me.
12:25:25LloreanJust, name-wise
12:25:26gevaertsfml: well, the main problem with those is when you want to find out where code comes from, as it makes svn blame much less useful
12:26:03Unhelpfulactually, buflib can only do the first of those things, but even that would be useful by itself, if we add a refcount field and only load cover images once, for example.
12:26:20LloreanUnhelpful: That at least has been long talked of
12:26:47fmlLlorean: yes, the name was suspicious to me too. It's the height of the middle viewport. Its value is double height. What's the best name for it?
12:27:33UnhelpfulLlorean: amiconn and i talked about this a few weeks ago. it's on my list, i'd like to use such things to, for example, carve the plugin buffer dynamically from the end of the audio buffer, and have it disappear entirely when no plugin is loaded.
12:27:39fmlgevaerts: so no code changes without any real (functional?) improvements?
12:28:28Unhelpfulhe had also suggest that being able to change the plugin buffer size at runtime would be a cleaner solution than overlay plugins :)
12:29:59LloreanUnhelpful: So the ability to de-buffer audio too? Compact to the front, then if there's not enough free, drop a little at the end?
12:30:43LloreanBeing able to change the plugin buffer at runtime would also be excellent for, for example, Rockboy which really needs about a 768KiB buffer to improve compatibility while listening to music
12:32:30Unhelpfulpictureflow and the other plugins that use overlay on hwcodec could also use such functionality to run during playback, by requesting a larger plugin buffer.
12:33:57Unhelpfuland yes, i already know that we'd need to implement it *twice* to have it on hwcodec as well :/
12:35:30LloreanHonestly, I'd consider it better than the current situation if it required playback to restart and rebuffer from disk to re-allocate.
12:35:44LloreanAs long as the restart and rebuffer were automatic, I mean music would only be gone for a second anyway
12:36:45gevaertsfml: if the style is really horrible, it's different of course, but personally I'd prefer not to change this particular bit
12:38:26fmlgevaerts: ok, then I'll leave it as it is now
12:43:06UnhelpfulLlorean: i don't see why we should *have* to restart and rebuffer, though. it seems like buffering.c already contains some support for moving individual buffer. it even helpfully updates the pointers in bitmap buffers for you. :)
12:43:52LloreanUnhelpful: My point was that I'd consider that still an improvement on the current situation, though, was all. :)
12:44:02***Saving seen data "./dancer.seen"
12:44:48Unhelpfuli can imagine that codecs, or anything else, that uses coprocessor threads and the buffer could possibly get burned by doing that sort of thing?
12:45:11LloreanFrom restarting playback entirely?
12:45:24Unhelpfulno, from moving buffers.
12:46:48 Quit bertrik (Remote closed the connection)
12:48:18Unhelpfulbuflib has the same problem if you want to use cop - there's really no way to know if a buffer has been moved since you fetched a pointer for its handle. it's fine on one core, because this can only happen when you yield, but a second core ruins that assumption.
12:50:10 Join nazir [0] (
12:50:10Unhelpfulthe best scheme i could come up with for that sort of thing was to have cop+buffer users register a semaphore that they hold *most* of the time, but release on some regular basis. the basic idea would be that to move a buffer associated with such a semaphore you would need to acquire it, and the process using that buffer would understand that the buffer would normally not move, but might move each time it releases the lock on it.
12:50:59 Join bubsy [0] (i=Bubsy@
12:51:45Unhelpful(i am assuming, of course, that we have semaphores which do not rely on non-preemption?)
12:52:46gevaertsYou could fill up the pcm buffer, restart playback without flushing pcm, and hope for the best :)
12:53:26Unhelpfuldo we *do* sample-accurate seeks on resume? :)
12:53:44gevaertsthat's a major part of the "hope for the best" bit
12:54:46Unhelpfulwell, at least on my e200 i can stall buffering for about 6s before things go wrong :)
13:01:09 Join AndyI [0] (i=AndyI@
13:02:52 Nick fxb__ is now known as fxb (
13:08:19 Quit tvelocity (Remote closed the connection)
13:10:59 Part Grahack
13:17:38 Join killan [0] (
13:21:11 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:23:07CIA-70New commit by alle (r21350): Slightly rearrange statements. No functional change but RAM and binary are reduced.
13:25:18fmlHrm... I'd also like to be able to change pitch in semitone steps without affecting the speed. Would it be a fourth mode?
13:29:09LloreanThe whole interface should probably be improved. The screen is unmanageable (or the implied controls incorrect) on several targets anyway
13:32:46GodEaterdefinitely wrong on the ipod
13:33:01fmlYes, it's not very intuitive. And I'd throw out the nudge functionality. I think hardly anyone uses it.
13:33:20Llorean"nudge functionality"?
13:33:57*GodEater is bewildered too
13:34:38GodEateralthough I see a #def refering to nudge in there =/
13:35:23fmlLlorean: yes. Temporarily changing pitch +/-2%
13:35:30 Join d3v14710n [0] (i=d3v14710@unaffiliated/d3v14710n)
13:35:34 Join p3tur [50] (n=petur@rockbox/developer/petur)
13:36:20LloreanThat's actually useful if you know the the BPM of something else, and you're trying to get the beats lined up after adjusting the speed or pitch of the current thing playing on your player
13:36:34GodEatertrain is bordering
13:36:45 Quit p3tur (Remote closed the connection)
13:36:45 Join wark [0] (
13:37:28 Quit _lifeless (Read error: 60 (Operation timed out))
13:37:46 Join _lifeless [0] (n=lifeless@
13:38:06fmlLlorean: in that case I'd just change the speed/pitch with "normal" controls, and then back. The BPM is not displayed in the screen anyway so you have to use your ears. And with normal controls you're more flexible.
13:38:48Lloreanfml: If your target speed is 120.3%, it's very hard to go back to it *exactly* from a faster or slower speed, especially instantly
13:39:05LloreanBeing able to drop from 122.3 to 120.3% the instant you hear the beats occur at the same time is handy.
13:39:37LloreanThe regular UI just doesn't allow this to be done easily at all.
13:40:04fmlLlorean: ok, maybe. This is from the "advanced DJ tools" area :-)
13:41:56LloreanI'd say instead of a "pitch screen" it should probably just use menus like normal settings.
13:42:23*Llorean doesn't really understand why it needs a custom screen.
13:42:47LloreanWith normal menus, you could set speed to 120.3, then after setting it, move it to 122.3 and as soon as you catch up, cancel out of the menu so it drops back to 120.3
13:45:44fmlSo we'd have two settings: one for the pitch ond the other for the speed? If timestretch is enabled the two can be adjusted independently. If it's not enabled changing one setting would also affect the other. Is it what you're thinking about?
13:46:23 Join Lss [0] (
13:46:38fmlOr, if timestretch is disabled, we could show only one setting but call it "pitch and speed"
13:47:20LloreanPersonally, I think we could be rid of the option to turn time stretch on and off, and just have the speed and pitch settings always independent.
13:47:42LloreanBut otherwise, yes, have changing one also change the other probably
13:47:49 Join __lifeless [0] (n=lifeless@
13:48:10 Quit _lifeless (Read error: 60 (Operation timed out))
13:49:47 Quit markun ("off to rockbox devcon")
13:50:23fmlBut what about the principle "allow the people to use as much of the sound buffer as possible"? By always enabling the feature, you'd always reduce the buffer. But I personally would be for it (always enable the feature, cut the setting)
13:53:41fmlGetting rid of the screen (replacing it with settings) and always enabling the feature would probably mean a huge green delta
13:55:59 Quit n1s (Read error: 110 (Connection timed out))
13:58:12 Quit jfc^2 (Read error: 104 (Connection reset by peer))
13:59:11LloreanThe problem with the feature as it is now is it's one that requires a reboot, but doesn't seem like it should to a user. As well, with the way the pitch screen is set up, even after the reboot it's not obvious how to use it.
14:00:30LloreanI'd say it's one of the least intuitive things to use (even once you know how Rockbox tends to work)
14:00:41LloreanThe EQ also suffers a lot from difficulty to use.
14:01:00 Join bertrik [0] (n=57d33175@gateway/web/cgi-irc/
14:01:59 Quit robin0800 (Read error: 104 (Connection reset by peer))
14:03:03fmlLlorean: for the EQ it's justified by the fact that you should see it all at the same time, and it's "two dimensional" by the nature
14:03:42 Join funman [0] (n=fun@rockbox/developer/funman)
14:03:57fmlBut with the menu, how would you implement the semitone mode?
14:04:55LloreanA setting for "step size" with options of .1%, 1%, and "Semitone" possibly?
14:05:07 Quit bertrik (Client Quit)
14:07:46fmlBut it's not very "compatible" with the setting menus as they are used otherwise. We should think about it (I mean the others as well).
14:08:32Mikachufwiw i think the pitch screen is very hard to use on ipods, it's almost completely random if it goes .1% or 1%
14:08:51Mikachuit used to repeat (ie 1%) as long as you touched the wheel, but now it seems to be timing dependant
14:10:11LloreanWould it make sense to restrict "Pitch" to semitones rather than using percentages?
14:10:34Mikachumaybe the original song is out of tune :)
14:11:11LloreanAll of the instruments are the same amount out of tune? :-P
14:11:34Mikachuyeah, maybe a halfbroken record player
14:11:43LloreanThen you'd fix that in the encoding.
14:12:34 Part nazir
14:14:39funmansaratoga: i found a simple way to reduce buffer size
14:16:11 Quit bmbl (Read error: 110 (Connection timed out))
14:16:23funman.pad (NOLOAD) : { . += 0xsomething; } > DRAM just before audiobuf in
14:19:06LloreanIs there a point in it being measured in semitones actually?
14:19:45Mikachuif you want to play along?
14:20:00 Join syn4pse [0] (
14:20:31 Join LambdaCalculus37 [0] (n=44a0430d@rockbox/staff/LambdaCalculus37)
14:22:16fmlLlorean: yes. Semitones are a term from the music and are often used when transposing or adjusting the music to the voice etc.
14:23:00Lloreanfml: I'm aware what a semitone is.
14:23:34LloreanMy point though was that Rockbox is primarily a playback device. Options should, generally, be things you might want to change "in the field." Wouldn't adjusting that be more something you'd prepare in advance if you're singing / playing along?
14:23:56LloreanI mean, the original pitch shift was there because we didn't have the ability to time stretch
14:26:31*LambdaCalculus37 has sound and touchpad on his GoGear SA9200 :)
14:26:49fmlOr the other way around: the speed change was there because we didn't have the ability to time stretch. I think both are useful. What if you use the same song but different people would like to sing along with it?
14:27:21LambdaCalculus37And no one ever sent a "Ladies and Gentlemen..." mail for that target in over a year. ;)
14:29:47Lloreanfml: Traditionally the song is played with specific notes independently of what the people are singing... But anyway, I was just asking to be sure semitones are the more useful.
14:32:58fmlLlorean: they are just useful, not "more useful". We'd still need fine tuning IMO. And that with specific notes is true −− but only for the people with absolute pitch. For the rest, it's more or less the same.
14:33:52LloreanWhy is fine tuning "needed"?
14:34:05LloreanSaying it's needed doesn't explain anything on its own.
14:35:05 Join n1s [0] (n=n1s@rockbox/developer/n1s)
14:35:18fmlLlorean: in order to be able to exactly match the pitch −− to e.g. how an instrument is tuned.
14:36:22LloreanIf the song is the right pitch, doesn't that mean the instrument is out of tune?
14:36:33fmlSay you want to play guitar along with different songs. Which are not exactly at an integer number of semitones from each other.
14:37:17LloreanIf you need fine tuning always, what's the point in semitones in the first place? It's just an approximation that you can get to by scrolling through the percentage list quickly.
14:38:30Mikachuit's easier to go to the right general area by semitones and then fine tune i would think
14:39:36LloreanYou can scroll through lists pretty darn fast.
14:39:50Mikachui don't mean mechanically fast, but figuring out where to go in the first place
14:40:00fmlFine tuning is more powerful, yes, but tuning in semitones is more handy. More generally, I'd say it's good to have larger steps than 0.1%, and semitones are just natural larger steps.
14:40:59fmlMikachu: that too, yes! Everybody who makes music can approximately tell how many semitones the difference is.
14:42:04LloreanIn a list, large steps is less important anyway since we have scroll acceleration.
14:42:17fmlLlorean: if in the fine list we'd have additional marks on the entries that correspond to semitones that would be a solution maybe
14:42:46LloreanThe normal settings list wouldn't allow that, I believe.
14:43:29LloreanIf you're playing an instrument along with a song, shouldn't it be tuned properly for the song?
14:43:29fmlLlorean: with a custom formatting function it should be possible
14:43:38LloreanI'm confused as to why you want to adjust the song to your instrument, rather than the other way around.
14:43:58Mikachuit's kind of hard to adjust a piano, for one
14:44:00LloreanAnd if the song is out of tune, why are you playing along to that recording instead of a good one?
14:44:05***Saving seen data "./dancer.seen"
14:44:28fmlLlorean: because it's harder to tune the instrument than to adjust the pitch. And if I have different songs it can became a nightmare to retune it every time.
14:44:38 Join robin0800 [0] (
14:45:20Lloreanfml: How often would you say you use the fine tuning to match it to an instrument?
14:45:26LloreanAs in, how often have you personally done it?
14:46:02 Quit robin0800 (Remote closed the connection)
14:50:04fmlLlorean: I can't tell exactly. Not very often. Most of the time I use semitone adjustment if I need it. That's why I'm also annoyed by the fact that that's not the default mode :-)
14:50:59LloreanI think semitone adjustment would be adequate. The primary function of Rockbox is playback, not "adjusting out of tune songs to play along with possibly slightly differently tuned instruments"
14:51:20Llorean And if your songs are in-tune, and your instrument is too, there shouldn't be a problem right? It's just for solving the *error* case.
14:51:31LloreanOr is there a case where everything is right and it's still needed?
14:51:55fmlLlorean: the song is probably OK, just the instrument is tuned differently.
14:52:03LloreanBut is there a case where everything is right and it's still needed?
14:52:36fmlIf everything is right it still might be needed if you like to play it in a different tune
14:52:45LloreanRockbox shouldn't be given the job of making up for your instrument. And there are a thousand things that can be wrong with files or songs that we already don't fix because it's outside the scope, this really should be too.
14:53:19Lloreanfml: Is that a *realistic* example there? Really?
14:53:22fml(because you can play in some tunes but not the others)
14:54:31 Quit timc (Read error: 110 (Connection timed out))
14:54:52fmlWhy not have the fine list with additional marks? It should be easy to implement as we can have a custom formatting function for the values in the list
14:55:30LloreanBecause we could save more binsize by not implementing custom formatting for all of one list, for a feature that is almost never used outside of correcting errors that shouldn't be Rockbox's problem?
14:55:49 Join LambdaCalculus [0] (i=44a0430d@gateway/web/freenode/x-e58b8794bf34af37)
14:56:22LloreanPeople might want to adjust their equalizer in .1db steps, or adjust the frequency by 1hz at a time. There's a thousand settings in Rockbox that *could* offer users more freedom, but instead are constrained to a "common sense" range. Semitones really seems like that range for pitch.
14:57:06 Quit LambdaCalculus37 ("CGI:IRC 0.5.9 (2006/06/06)")
14:57:20 Nick LambdaCalculus is now known as LambdaCalculus37 (i=44a0430d@gateway/web/freenode/x-e58b8794bf34af37)
14:58:07 Join timc [0] (n=aoeu@
14:59:50fmlI could live with just semitones quite well
15:00:13fmlBut wouldn't mind if it were possible to do fine tuning as well
15:00:35LloreanMaybe 10 cent increments instead of whole semitones?
15:00:45LloreanSo that you can fine tune, but you're still on a scale where semitones are obvious?
15:01:53fmlLlorean: what is cent? 10%?
15:02:26LloreanIf my understanding is correct, a semitone is 100 cents, so 10% of a semitone yes.
15:02:45fmlThe problem with semitones is that it's not a linear scale. It's logarithmic.
15:03:12LloreanSo is the cent
15:03:23LloreanOtherwise a semitone couldn't consistently be 10 cents
15:03:33fml...(or exponential): one semitone = frequency * (1/2)^(1/12)
15:06:17LambdaCalculus37lowlight: (for the logs) I have a new bootloader and your build for the GoGear SA9200 installed. It'w working great!
15:06:43fmlLlorean: (after loking to the wikipedia) Ah, ok, it's just more fine (too fine for common use IMO) unit. 1/2 semitone would be enought in most cases I think
15:07:13fmlLlorean: I.e. 50 (music) cents
15:07:20LloreanYou could display it as .1 semitone rather than 10 cents, for easier understanding. .5 works too. My point, really, was "keep the scale in semitones but use finer units, rather than custom formatting"
15:07:45LloreanI don't object to finer tuning, I just want to try to keep things "simple" with how the menu works.
15:08:33 Quit syn4pse ("ChatZilla 0.9.84 [Firefox 3.0.11/2009060215]")
15:08:58fmlLlorean: that would be a good way. A scale with a 1/2 semitone steps, with internal transformation to the real pitch
15:11:38 Quit arohtar (Client Quit)
15:15:07 Quit fml ("CGI:IRC 0.5.9 (2006/06/06)")
15:17:48funman : do not try to load album art indefinitely if the buffer is full (there is not enough room on c200v2 with logf enabled), is it ok (see r20149) ? I think the buffer will not empty itself while album art hasn't been loaded so it's an infinite loop
15:21:02SlasheriUnhelpful: hi, are you already in the CREDITS? I am just looking at the open_files issue you found from tagcache
15:22:40 Quit Llorean (Read error: 104 (Connection reset by peer))
15:22:42oboSlasheri: if he's not he can always add himself :)
15:23:07Slasheriobo: oh, great :)
15:23:09 Join Llorean [0] (
15:23:46LloreanSlasheri: Doesn't he have commit access?
15:23:55n1she does
15:24:59n1sand he's already in the list
15:25:03SlasheriLlorean: i didn't knew, i haven't keeped track :)
15:25:23 Join PaulJam [0] (
15:25:39 Join jfc [0] (
15:25:51 Quit jfc (Read error: 104 (Connection reset by peer))
15:26:11Slasherii am trying to finally find was is causing duplicates in the db some people are reporting
15:26:42 Join jfc [0] (
15:27:53LloreanIt seems pretty common, or at least, pretty often reported
15:28:05*LambdaCalculus37 has to get to working on some plugin keymaps for the GoGear SA9200
15:28:06LloreanA lot of people have stopped using auto-update because of it (people say manual updating seems to work better)
15:29:38Slasheriit's weird because i have never had them. But i have always used the db with load-to-ram enabled so that very likely hides the problem
15:34:20PaulJamhmm, is the IRC log viewer on the site broken? for me it stops at 15:03 for todays log...
15:34:59funmanyes it breaks when some urls are typed in the channel
15:35:21 Quit flydutch (Read error: 110 (Connection timed out))
15:35:30funmanthe watermarks in buffering.c need to be adjusted when the audiobuffer is really small
15:41:30 Quit linuxstb ("Leaving")
15:41:44 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
15:48:13funmanhmm chainging the watermarks and default filechunk produces a data abort in memset() ..
15:48:29funmanin the middle of the song, not at the beginning though
15:48:39Lloreanfunman: Didn't Zagor look into the watermarks a few months back?
15:48:45Mikachuthe 'current.txt' link works fine
15:48:47Mikachu(irc log)
15:49:06funmanLlorean: it still doesn't handle very low buffers though
15:49:21funmanit would need dynamic checking of the buffer size
15:50:01LloreanToo low buffers are going to be harmful, though, anyway right? If we didn't need the buffer, we wouldn't have it in the first place.
15:51:01n1sMikachu: it's just the reader that breaks
15:51:21Mikachuit was just a suggestion for the person who wanted to read the log
15:51:23funmanLlorean: well we have to live with whatever room we have
15:52:51funmanperhaps the pcm buffer could be made smaller as well
15:53:42linuxstbLlorean: They can probably be lower on flash targets anyway, as there is no spin-up time to worry about.
15:54:20Lloreanlinuxstb: Yes, that should be true.
15:54:21linuxstbfunman: I _think_ the pcm buffer needs to be that size for crossfade to work. I'm not 100% sure though...
15:58:24 Quit __lifeless (Remote closed the connection)
15:59:23PaulJamMikachu: thanks
16:06:29 Join robin0800 [0] (
16:19:56 Quit funman ("free(random());")
16:23:08Slasheriwho was Unhelpful? To include his name in the commit log :)
16:23:32LambdaCalculus37Slasheri: He's already in the commit log.
16:23:35LloreanAndrew Mahoney?
16:23:44*Llorean is trying to learn names and go by memory.
16:23:46Mikachuthere's no y on the ircnicks page
16:24:02SlasheriLambdaCalculus37: i meant the commit message i will do shortly :)
16:24:07SlasheriMikachu: thanks
16:28:08 Quit PaulJam (Nick collision from services.)
16:28:13 Join PaulJam_ [0] (
16:31:03CIA-70New commit by miipekk (r21351): Fixed searches from tagcache when there is a sudden need to open a tag file on the fly. Thanks to Andrew Mahoney for pointing out the issue.
16:31:19Mikachuno y
16:32:37lilltigerbtw when rockbox builds it's database, does it check Album Artist before it checks Artist when generating an album?
16:36:30LloreanThat's not an issue for database building, but for your tagnavi I believe
16:38:12lilltigerok, not went any deep into it i just tagged and fixed my whole music collection and noticed that amarok does it wrong :(
16:38:21lilltigerand so does my Sansa Fuze
16:38:48 Join bmbl [0] (
16:40:16lilltigernd thought that it might be a neat thing for rockbox to do correctly :)
16:40:52 Quit petur ("brb")
16:41:51 Quit PaulJam_ (Nick collision from services.)
16:41:56 Join PaulJam__ [0] (
16:41:57 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
16:42:34 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
16:42:38LloreanSlasheri: Would it be possible for Rockbox's database to have the concept of "fallback tag"? For example, as this person wants "Album artist, but fallback to artist if not present" and a common want is the fallback from "Sort Artist" to simply "Artist" etc.
16:42:54LloreanI suppose they'd need to be used in generation rather than sorting for performance
16:43:18 Quit robin0800 (Remote closed the connection)
16:43:46linuxstbB4gder: Before the enjoyment starts, did you put a copy of the 3.3 source on the download servers?
16:43:53B4gderah I did not
16:43:57B4gderme will fix that
16:44:09***Saving seen data "./dancer.seen"
16:45:39 Join Ubuntuxer [0] (
16:45:51 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-3dbf3bca8d237f31)
16:45:59 Quit PaulJam__ (Nick collision from services.)
16:46:03 Join PaulJam [0] (
16:46:05 Join Zagor [242] (n=bjst@rockbox/developer/Zagor)
16:47:17SlasheriLlorean: hmm, that should be definately possible :)
16:48:07B4gderpetur works on getting the webcam pc up
16:49:13LloreanSlasheri: It would kill a fairly large number of birds.
16:49:36lilltigerkilling birds would be neat ;p
16:49:53Mikachuwith one rock?
16:49:55*obo wonders if he has found lcd drawing routines....
16:50:44lilltiger"Rockbox, now we kill birds"
16:51:00 Quit wark (Read error: 104 (Connection reset by peer))
16:53:01 Join Domonoky [0] (n=domonoky@rockbox/developer/domonoky)
16:53:10Horschthave fun at devco, guyse
16:53:33*Domonoky arrived at Devcon :-)
16:53:57Horschtno you did not. they sent you to a fake location :p
16:55:10UbuntuxerHi, can someone commit my patch FS #10281, please ?
16:55:40 Join toffe82 [0] (n=chatzill@
16:57:30scorche|shhooray for DevCon!
16:59:11 Quit bmbl ("Bye!")
16:59:29 Join bmbl [0] (
17:04:22 Quit toffe82 (Remote closed the connection)
17:04:32webmindLlorean, not with my mini I think. doesn't seem to respond
17:05:20 Join toffe82 [0] (n=chatzill@
17:09:03 Join petur [50] (n=petur@rockbox/developer/petur)
17:10:07 Join captain_kewl [0] (i=2669ecc2@gateway/web/freenode/x-e636f717c3e77208)
17:10:52 Join BryanJacobs [0] (i=809743f3@gateway/web/freenode/x-3220e7906f83f8f2)
17:14:43 Join JdGordon [0] (
17:19:00 Join p3tur [50] (n=petur@rockbox/developer/petur)
17:23:39B4gder mentions rockbox
17:25:38 Part evilnick
17:26:26 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-3dbf3bca8d237f31)
17:26:27 Join Rob2223 [0] (
17:26:33 Quit perrikwp (" ajax IRC Client")
17:26:37LambdaCalculus37B4gder: Onto the Articles page it goes. :)
17:28:44saratogawe've got a lot of patches for plugins sitting around on the tracker
17:28:50saratogawe should give some of these people SVN access
17:29:04saratogai get the feeling a lot of them are never reviewed just because no one is familar with the code
17:29:19 Join ocean_ [0] (i=d59c23e8@gateway/web/freenode/x-960c544bdb4bb768)
17:30:28ocean_hi, someone knows how to backup firmware from a jz47**/chinachip device? (is that the right channel to ask in? :P)
17:31:29ocean_since i'm planning to play around with rockbox/firmware upgrading and don't want to break it
17:31:33B4gderocean_: I think you're better off asking in the suitable forum thread as not that many of us know things about the ondas
17:32:04JdGordonLambdaCalculus37: haha we got a tiny link.... thats hardly an artile about rockbox :) (although dd-wrt didnkt get linked so we wins there :D )
17:32:11ocean_B4gder: i opened a topic in Repairing and Upgrading forum but has been locked up
17:32:26ocean_maybe i should ask it in the injenic topic?
17:32:59B4gderthose are new targets, you should probably ask in the existing thread
17:33:20saratogai think that would be fine, provided you make it clear to mcuelenaere that you're interested in development
17:33:30saratogahe will probably want to know
17:33:58LambdaCalculus37And since it may prove a couple of new challenges, he may be interested in those.
17:34:26ocean_interested in development yes though i'm not that good but i know C++/x86 asm and how to use IDA if that could be of help
17:39:42 Part Ubuntuxer
17:42:03LambdaCalculus37ocean_: Grab the Rockbox source and get to reading. Sounds like you have taken a couple of first steps.
17:42:17LambdaCalculus37Rockbox is mostly C, with some assembly for speed-critical bits.
17:42:32LambdaCalculus37But the assembly is SH-1, m68k, MIPS, and ARM.
17:43:36ocean_yes i know C i have done some university exams with C :P
17:43:40 Quit Rob2222 (Read error: 110 (Connection timed out))
17:43:49ocean_don't know mips but i could learn
17:44:14Torneto work on an early stages port you'd almost certainly need to kjnow the asm for the processor in question
17:44:31LambdaCalculus37ocean_: The best place is to check out the existing Onda code in SVN.
17:44:44LambdaCalculus37But also Google for some books on MIPS assembly.
17:45:18B4gder"see mips run" being a popular book on mips
17:45:32*Torne finds that inappropriately hilarious
17:45:36LambdaCalculus37Torne: While that's most certainly true, in this case ocean_ can work on the existing code that's there.
17:45:50ocean_i'll give a look whenever i have time (i have some exams with university now) after having backed up fw :P
17:46:09ocean_thanks lambda
17:46:26saratogamips is extremely easy to learn
17:46:28 Join markun [50] (n=markun@rockbox/developer/markun)
17:46:32ocean_i think there's also something in my university library
17:47:06LambdaCalculus37ocean_: If not, Google is your best friend.
17:47:35ocean_i know :D google and ... power searching :P
17:47:43Tornehow does lcd_puts (not putsxy) measure horizontal cells if you are using a proportional font?
17:49:15 Join bertrik [0] (n=bertrik@
17:55:12 Nick fxb is now known as fxb__ (
17:57:03*B4gder and zagor discussed distributed build design on the flight here
17:57:24B4gderand now we await zagor to paste our notes in a wiki page... /me looks
18:01:18*Zagor tries to format it in a sane way
18:03:59 Join pyro_maniac [0] (
18:04:09saratogaoh wow theres a lot of replaygain related emails on the mailing list
18:07:29LambdaCalculus37saratoga: Besides plugin patches, there are also a lot of language patches in the tracker that have to be taken care of.
18:07:44Zagorrough notes:
18:08:32 Join ender` [0] (
18:08:37LambdaCalculus37I also see a few patches that add new DAP images for manuals, etc.
18:09:39Torne...hang on, does lcd_puts() really determine what 'x' means by multiplying it by the width of the string divided by the number of characters?
18:09:55*linuxstb hangs on tight
18:10:02Tornesurely that makes it totally useless on proportional fonts
18:10:11Torneno way to get anything to ever line up
18:10:20oboTorne: what proportional fonts?
18:10:22B4gderZagor: it struck me that the upload in the bg works fine with the protocol we worked on
18:10:23ZagorTorne: well, is it used?
18:10:40ZagorB4gder: ah, true! that makes bandwidth a near non-issue
18:10:42TorneZagor: hard to tell if anything calls it without setting the font to sysfixed first
18:10:49Tornebut i would guess not
18:11:40Tornesince my interpretation of the code is that lcd_puts(1, 0, "iiiiii") leaves a gap on the left the width of an 'i' whereas lcd_puts(1, 0, "wwwwww") leaves one the width of a 'w'
18:11:48Tornewhich makes it difficult to use sensibly :)
18:12:27Torneobo: most of the fonts other than sysfixed are proportional..
18:12:58Mikachuusually i think programs use the with of 'm' for proportional fonts
18:13:02 Quit MrDuck (Read error: 110 (Connection timed out))
18:13:12TorneMikachu: other programs might, yes, but rockbox doesn't appear to
18:13:27Tornei'm looking in lcd_puts_style_offset in lcd-16bit.c
18:15:12linuxstbTorne: I think that function is almost exclusively used with an "x" of 0 - so no-one cared. Use lcd_putsxy() instead.
18:17:43Torneit just seems overcomplicated to get a useless answer. :)
18:17:47linuxstbamiconn _may_ have an explanation for that logic when he's around though - he's done a lot of the work on the graphics API.
18:17:58Torneif it's only going to do something sensible for fixed width fonts it may as well just use the font width
18:18:07Tornerather than calculating the string width then dividing it by the length :)
18:18:28Tornei expect it's just the result of generalising from character cell display
18:20:05pyro_maniaclogbot seems to be down
18:20:20B4gderso cheer him up!
18:20:38 Join JdGordon| [0] (i=ad813e9c@gateway/web/freenode/x-84a7f26208e694f1)
18:20:44*Zagor rushes to give cpr
18:20:53 Quit antil33t ()
18:21:11 Join antil33t [0] (n=Mudkips@
18:22:00 Quit ocean_ ("Page closed")
18:22:43pyro_maniacB4gder: who adminstrates the logbot?
18:23:00B4gderpyro_maniac: zagor does
18:23:16saratogaanother devcon idea: making the web logs not freeze on wiki links!
18:23:28Zagorthe bot is fine actually
18:23:41saratogayeah its the perl script
18:23:52saratogathe regex fails on wikipedia links i think
18:23:58saratogathough i'm not sure why
18:26:22 Quit antil33t ()
18:26:36 Join antil33t [0] (n=Mudkips@
18:26:52 Join Hillshum [0] (n=chatzill@unaffiliated/hillshum)
18:27:07 Quit antil33t (Client Quit)
18:27:22 Join antil33t [0] (n=Mudkips@
18:28:18 Quit antil33t (Client Quit)
18:28:23 Join antil33t [0] (n=Mudkips@
18:31:35Zagorlog reader script fix. reload.
18:31:54saratogabah was just debugging it
18:32:14saratogawhile you're there any idea why it doesn't update in google chrome?
18:32:25 Quit antil33t (Client Quit)
18:32:36Zagorbasically only mozilla supports those chunked transfers
18:33:05Zagoryou are welcome to do a real fix, I really just moved it so it doesn't occur as often...
18:33:19saratogawhat did you change? i don't see anything in SVN?
18:33:33Zagorah no didn't commit yet
18:33:47 Join antil33t [0] (n=Mudkips@
18:34:14CIA-70New commit by zagor (r21352): Only nick-fix url if it actually matches
18:34:23saratogai assume its just one of the regex in parsechunk getting stuck
18:35:21 Quit JdGordon| (Ping timeout: 180 seconds)
18:40:07 Join ssoap [0] (n=42c07542@gateway/web/cgi-irc/
18:43:13 Join BHSPitLappy [0] (n=BHSPitLa@unaffiliated/bhspitmonkey)
18:43:21 Quit ssoap (Client Quit)
18:44:10***Saving seen data "./dancer.seen"
18:44:25 Join kugel [0] (n=kugel@rockbox/developer/kugel)
18:45:56 Join JdGordon| [0] (i=836b0046@gateway/web/freenode/x-3be4c527520a01d5)
18:47:11*obo has found some very e200 looking code - lcd_write_reg, lcd_send_msg & lcd_send_byte, but no luck getting anything to display yet...
18:48:30 Quit Hillshum ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
18:52:21 Nick p3tur is now known as petut (n=petur@rockbox/developer/petur)
18:52:44 Quit petur (Nick collision from services.)
18:52:49 Nick petut is now known as petur (n=petur@rockbox/developer/petur)
18:53:17 Join petur2 [50] (n=petur@rockbox/developer/petur)
18:53:50bertrikobo,have you been able to enable the backlight already?
18:54:19 Quit petur2 (Client Quit)
18:54:40obobertrik: I've found a GPIO which turns it on and off - but I'm running my code after the OF bootloader, not a cold boot...
18:57:30 Quit bertrik (Remote closed the connection)
19:08:29 Join p3tur [50] (n=petur@rockbox/developer/petur)
19:08:51 Quit petur (Nick collision from services.)
19:08:53 Nick p3tur is now known as petur (n=petur@rockbox/developer/petur)
19:09:28 Join kugel_ [0] (
19:13:04 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:13:37 Join bertrik [0] (n=bertrik@
19:13:40 Join froggyman [0] (
19:16:01 Quit kugel (Read error: 110 (Connection timed out))
19:17:33 Quit DarkDefender (Read error: 110 (Connection timed out))
19:20:25 Quit Zagor (Read error: 110 (Connection timed out))
19:20:40 Join Zagor [242] (n=bjst@rockbox/developer/Zagor)
19:20:44 Join ocean__ [0] (i=d59c23e8@gateway/web/freenode/x-477f77eac10a1cbf)
19:21:03ocean__someone could explain me what mculenaere means with FLT (flash layout)?
19:21:31linuxstbNot FTL?
19:22:06DomonokyFTL = Flash Translation Layer..
19:25:33bertrikfaster than light?
19:26:38ocean__ok thanks i'm gonna read some specs about it
19:26:56 Quit ender` (Read error: 60 (Operation timed out))
19:27:57 Join jgarvey [0] (
19:28:15 Join davidfg4 [0] (n=david@
19:30:15saratogaocean__: its the mapping between logical addresses in software, and physical cells of memory on the flash chip
19:30:51saratogathe software continually changes them around to ensure that writes to the same logical address don't all write to the same physical cell in memory, wearing it out over time
19:34:36ocean__yep read the intel paper and specs
19:34:48ocean__and why the chinachip should be different?
19:35:20ocean__also, as i wrote on the forum on linux 2.6.30 i cannot see the flash fat partitition on windows yes
19:35:30ocean__so maybe windows drivers support that kind of FTL
19:36:26ocean__if it is so i could try to get a hand on reversing the driver
19:37:30saratogaocean__: the FTL is a piece of software in the original firmware's flash memory driver
19:37:39saratogait happens below the level of the USB protocol and windows driver
19:38:43saratogaeach manufactorer likely comes up with their own implementation, or licenses someone elses, so there will be many differences between different devices
19:39:07ocean__mmm ok so maybe it's not possible to get to a raw read of the sectors behind fat partition as i hipotized in forum
19:39:37saratogayou can get raw sectors in rockbox, and get remapped sectors in the OF's USB mode
19:40:03ocean__mmmm mculenaere post hit the nail
19:40:28ocean__he says tha OF do the work of reading/writing
19:42:46 Join ender` [0] (
19:46:41 Join perrikwp [0] (i=18ac0c41@gateway/web/freenode/x-567526a1d1daf9e5)
19:48:45ocean__so one way to get informations on FTL should be reversing original onda firmwares right?
19:50:15 Quit kugel_ (Read error: 110 (Connection timed out))
19:53:57jds_hi guys, im playing around with and ftl, too. im wondering what vfl means, the acronym is around in a lot of papers about ftls out there but seems to be nowhere explained. do you know?
19:54:32 Join Thundercloud [0] (
19:58:33 Quit perrikwp ("Page closed")
19:59:55ocean__mmm no in the intel paper i'm reading there's no trace of vfl
20:02:46jds_you are right, i think i found it in some samsung docs. but mainly in the (mostyly uncommented) sourcecode of an reverse engineer reed only driver for samsungs whimory
20:03:58jds_found the paper again (spotlight rocks): PBFilter: Indexing Flash-Resident Data
20:03:58jds_ through Partitioned Summaries
20:05:51ocean__mmm in the document it tells it's a api of FTL
20:06:16ocean__"several FTL provide different API level"
20:06:30jds_thats not exactly true
20:06:55ocean__and cite FTL, VFL, FIL levels in Samsung S-SIM FTL
20:07:04jds_in the sourcecode the vfl seems to map logical to virtual 'ftl' addresses
20:07:21jds_sry, virtual vfl addresses
20:07:32jds_and the vfl determines the physical address
20:07:40ocean__could be some kind of virtual address translation
20:08:57jds_you're right
20:09:05jds_but i'm interested in the acronym
20:09:36ocean__could be virtual file layer with file intended as memory page
20:09:46ocean__or something like that
20:10:06jds_sounds goot
20:13:19 Join tessarakt [0] (
20:13:53B4gderchaos on the cam!
20:14:07B4gdernow we'll just see Jens' back there
20:16:06 Join kperri [0] (
20:18:02 Quit ender` (" Why shouldn't truth be stranger than fiction? Fiction, after all, has to make sense. -- Mark Twain")
20:18:15 Join ender` [0] (
20:19:35 Join Hillshum [0] (n=4ba5ebce@gateway/web/cgi-irc/
20:22:07 Quit Hillshum (Client Quit)
20:23:14 Join Ubuntuxer [0] (
20:25:43 Join mcuelenaere [0] (
20:26:17mcuelenaereocean__: still here?
20:26:52MikachuB4gder: maybe put the devcon in progress thing on the news thing?
20:27:45Tornejds_: FTLs hav eto deal with three similar things and they sometimes do it in seperate layers: bad block remapping, erase blocks being bigger than sectors, and wear levelling
20:27:55Tornejds_: so that's possibly what the different layers there are
20:28:07ocean__mcuelenaere: yes
20:29:07mcuelenaereocean__: yes, the NFTL can be reverse engineered from the OF
20:29:17mcuelenaereI already identified the functions, but haven't looked into reversing them
20:29:44*Torne would help with that but is a bit NDAed. :)
20:30:28Tornebut if people do have ftl questions i can try and answer..
20:31:34ocean__mcuelenaere: right now i don't know mips assembly only x86 but could help reverseing and learn while reversing
20:31:34 Join funman [0] (n=fun@rockbox/developer/funman)
20:31:34mcuelenaereTorne: not every FTL is the same
20:31:49Tornemcuelenaere: no, but there aren't that many basic schemes
20:32:03Tornemost of which samsung have hilarious patents on ;)
20:32:27mcuelenaereocean__: well the FTL isn't top priority (for me) atm, but if you want to I can point you to the location in the binary
20:32:49mcuelenaereTorne: do you have experience with reversing a FTL through looking at a NAND dump?
20:33:00ocean__well if there are other priority i can help with them :P
20:33:06Torneno, but i am familiar with how quite a few work
20:33:17Tornebut, NDAs and shit
20:33:18mcuelenaereocean__: did you already try running Rockbox? :)
20:33:42ocean__i'm waiting for my dxt vx747+ to get charged (it arrived in the morning)
20:33:44Tornedo you know whose ftl it is?
20:33:50ocean__and will try running it on :D
20:34:27ocean__it's for sure a jz/chinachip because the 747 model has hxf firmware binary
20:34:42mcuelenaereTorne: no idea, and I didn't find anything recognizable.. (no strings)
20:35:17Tornewhat kind of flash?
20:35:19mcuelenaereocean__: it is probably compatible, there are a lot of clones out there and most of the 'real ones' work
20:35:22ocean__samsung flash
20:35:26mcuelenaereTorne: NAND? or what do you mean?
20:35:35TorneMLC or SLC?
20:35:55saratogasurely MLC given the capacity
20:36:18mcuelenaereright, MLC
20:36:29ocean__mcuelenaere: i have looked at some videos on youtube the firmware looks like the same, it only changes the inital splash screen
20:36:37Torneyah, so the write restrictions there are a bit of a clue at leatst
20:36:39mcuelenaereTorne: contains a NAND dump (8GB) of my player, if you're interested ;) (no OOB included)
20:37:01mcuelenaereocean__: yeah, that sounds compatible
20:37:19Tornemcuelenaere: i'm not *that* motivated :)
20:37:34ocean__tonight after the first 8 hours charge i'll get rockbox on :P
20:37:44Tornedo you know if the ecc is using the whole spare region?
20:37:46ocean__and hope it works
20:37:56Torneor if part of that is the BBR/levelling data?
20:38:59mcuelenaereI don't know what BBR is, and I'm also not sure if the ECC is using the whole region (I don't think so); isn't that also part of the FTL specification?
20:39:25Tornebad block relocation
20:39:39Tornequite alot of mlc flashes use the entire region for ECC
20:39:49Torneand have to store the actual FTL structure data in data blocks
20:39:56Torneyou need a stronger ecc on mlc flash generally
20:40:14Torneand yes, the ecc will be implemente din the ftl as well
20:40:43 Quit Ubuntuxer ("Leaving.")
20:40:51mcuelenaerehmm but generally speaking, is the definition of where to store the ECC and how big it is defined by the NAND manufacturer or the FTL one?
20:41:10mcuelenaerewith 'where to store' meaning in the OOB
20:41:31ocean__mcuelenaere: i have a little idea to avoid bricking up the little box, if i dump the every single bit in the nand like you've done and restore it, i also restore the firmware, the i could also try upgrading to some new onda firmware, and if goes wrong i restore the entire dump
20:41:51 Quit timc (Read error: 110 (Connection timed out))
20:41:57Tornemcuelenaere: not entirely sure but i think the ecc is likely to be dictated by the chip
20:42:01 Join Zagor_ [0] (
20:42:13Tornemcuelenaere: the ECC is always going to be in the OOB
20:42:32mcuelenaereocean__: you could try that, but I can't guarantee that will work (you never know what weird things get stored/verified by the OF)
20:42:36ocean__howewer i'm lucky included in the package there's a little minisd so i could try rockbox without doing damage :P
20:42:37 Quit Zagor (Nick collision from services.)
20:42:41 Nick Zagor_ is now known as Zagor (n=bjst@rockbox/developer/Zagor)
20:42:45Torneit's a questin of whether it uses the whole OOB or not; if it doesn't then lots of FTLs use the space for their erase counts and sometimes mapping data
20:42:57Torne(at which point looking at an OOB-less dump is useless)
20:43:03mcuelenaereocean__: Rockbox currently *doesn't* require any 'damage' to your player, it fully runs in RAM and SD
20:43:57 Join Nico_P [50] (n=nicolas@rockbox/developer/NicoP)
20:44:11***Saving seen data "./dancer.seen"
20:45:29mcuelenaereTorne: also, the hardware offers a choice between doing Hamming or RS algorithms; I suppose the two algorithms yield different ECC data?
20:45:37ocean__yeah but i also wanted to play a little with the box, linux/moddded firmware :P
20:45:45Tornethey may also be different sizes
20:45:56funmanbertrik: if you're going to work on clip/m200v4 problem at devcon perhaps i can arrange to be on irc at the same time
20:46:05ocean__but i think until i don't have a way to get hold of the OF i'll stick to play with rockbox :)
20:46:20bertrikfunman, ok good idea, probably not tonight though
20:46:22mcuelenaereocean__: linux also doesn't require any NAND modifications ;)
20:46:36ocean__i read on the jz-kernel
20:46:44bertrikI don't really trust the current ams sansa memory layout
20:46:46funmanbertrik: i'm working from 12AM->3PM and 6PM->1AM
20:46:55ocean__they flashed it and pointed to a 747write link
20:47:02ocean__to restore the original OF
20:47:04funmanbertrik: it's fine, really ;) i think the problem is in buffering.c / playback.c
20:47:18mcuelenaereocean__: could be, but last time I tried it; it purely ran from SD
20:47:25bertrikwe now basically have the codecs in iram and the rest in dram, right
20:47:29Tornemcuelenaere: you got datasheets for the chip?
20:47:36mcuelenaereTorne: yes
20:47:38funmanbertrik: some rockbox code is in iram also
20:47:40*B4gder adds committer #75
20:47:41Tornewhich is it?
20:48:11mcuelenaereSamsung K9HBG08U1M on my DAP
20:48:27funmanB4gder: \o/
20:49:31*mcuelenaere should probably add a link to the leaked Ingenic datasheets too
20:52:28funmanobo: the sansa view screen is 320x240?
20:52:34 Nick zitune is now known as zitune[afk] (n=zitune@bearstech/zitune)
20:52:36obofunman: yup
20:53:05funmanyour code for backlight doesn't work on a cold boot ?
20:53:17 Join low_light [0] (
20:53:34funmanperhaps there is another pin for powering lcd controller
20:53:48obofunman: I've never tried a cold boot. I don't yet want to competely replace the bootloader...
20:54:30low_lightBagder: did you see my reqest?
20:54:34funmani think cold boot is better unless you really miss something mandatory
20:55:04funmanBagder: while you're at it, did you see my request for c200v2 ? :)
20:56:13Tornemcuelenaere: where are you getting the info on the ecc?
20:56:19Tornemcuelenaere: datasheet implies the hardware doesn't care
20:56:27mcuelenaereTorne: which info on the ECC?
20:56:34Mikachufunman: i think you want to use B4gder for hilighting him right now :)
20:56:49Tornefrom teh datasheet i would guess it was a dumb chip and its all software's problem
20:56:50mcuelenaereTorne: the hardware offers RS/Hamming decoding
20:57:09Torneoh, sorry, you mean the controller in the soc?
20:57:19Tornesorry. im being dumb
20:57:50mcuelenaerenah, you're not with the datasheets in front of you :)
20:57:58 Quit prona (Read error: 110 (Connection timed out))
20:58:08obofunman: well, I'm not sure about recovery details... recovery mode seems to be built into the bootloader, which I'd be replacing?
20:58:14Torneis the soc manual around?
20:58:25low_lightfunman: we don't replace the OF bootloader on portalplayer devices (assuming the view is a pp decendent)
20:58:40mcuelenaereTorne: NAND flash can be connected to static memory bank 4~ band 1. Both 8-bit and 16-bit NAND flashes are supported. Hardware ECC generator is implemented (including Hamming and RS codes correction). A mechanism for booting from NAND flash is also supported.
20:58:44obolow_light: PP6100
20:59:04Torneah right
20:59:07mcuelenaereTorne: (if you have an account)
20:59:50low_lightobo, funman: the rockbox bootloader always loads after the OF bootloader
21:00:15 Quit Nico_P (Remote closed the connection)
21:00:40 Join Bagder_ [0] (
21:01:42 Quit Bagder (Nick collision from services.)
21:01:49 Nick Bagder_ is now known as Bagder (
21:02:11 Join Bagder_ [0] (
21:02:48obolow_light: good to know. does it still load the OF from a file, or does it read from the second (hidden) partition?
21:02:56funmanoh you only replace the .mi4 (which is a _rockbox_ bootloader, right?)
21:03:48 Quit BryanJacobs ("Page closed")
21:04:15low_lightobo: if it's like the c200/e200 then it's loaded from the hidden partition
21:06:41obolow_light: yup, the view OF does that (the magic in that partition is a little different, but it's very similar to an e200) - I just wasn't clear about how the rb bootloader did it
21:07:11 Part captain_kewl
21:08:59 Quit B4gder (Connection timed out)
21:10:14low_lightobo: the rb bootloader replaces the firmware, and simply loads rockbox.mi4 from the disk
21:12:58obolow_light: Okay. I guess I'll worry about that if I actually get around to having worth installing :) Back to looking at LCD code
21:15:40low_lightobo: have you tried the e200 emulator (FS #6524)? It has been very useful for me in the porting PP devices.
21:16:50obolow_light: I had a look at it a month or two ago, but I didn't have much luck in getting anything to run - will have to have another try
21:21:03low_lightBagder: did you see my reqest :)
21:21:04 Join Zagor_ [0] (
21:21:24Bagderah nice
21:21:26 Quit Zagor (Nick collision from services.)
21:21:29Bagderwider table!
21:21:30 Nick Zagor_ is now known as Zagor (n=bjst@rockbox/developer/Zagor)
21:21:32ocean__mcuelenaere: couldn't you upload it somewhere else?
21:21:51ocean__i don't have an account on that website
21:22:06ocean__if could upload the document somewhere else thank :)
21:22:14funmanBagder: did you see my request to add c200v2 boot/sim/build ?
21:22:39funmanXXXXXL build table ;)
21:23:23LambdaCalculus37We're going to need a wiiiiiiiide screen now. :)
21:23:27Mikachumaybe time to put sims in a separate table? :)
21:24:10mcuelenaereocean__: did you get my PM?
21:25:00Domonokyor switch the buildtable orientation again ?
21:25:31BagderI tried to rotate it once, it didn't make much of a difference in usability
21:25:42ocean__mcuelenaere: yes
21:26:32Bagderok, I'll add a bunch of new builds a bit later
21:26:44BagderI'll poke on the new build server code first
21:26:45funmanperhaps the table can be circular ?
21:27:16Mikachumaybe some sort of mandelbrot?
21:27:23 Quit bubsy ("Panic.")
21:27:38 Join bubsy [0] (i=Bubsy@
21:28:45 Join kugel [0] (n=kugel@rockbox/developer/kugel)
21:28:55obolow_light: which version of the e200 emulator do you use? The last one posted on that task?
21:29:05funmani was thinking of something like svg (written by the buildserver)
21:31:03low_lightobo: I believe I started with the last one. You'll need to adjust the ram offset. The e200 loads to 0x10600000 whereas you say the view loads to 0x10f00000.
21:32:05 Join n00b81 [0] (n=taylor@unaffiliated/n00b81)
21:32:26 Part n00b81 ("Leaving")
21:32:32 Quit PaulJam (
21:32:32 Quit dfkt (
21:32:32 Quit CaptainKewl (
21:32:32 Quit J-23 (
21:32:32 Quit jordan` (
21:32:32 Quit scorche (
21:35:00NJoinPaulJam [0] (
21:35:00NJoindfkt [0] (i=dfkt@unaffiliated/dfkt)
21:35:00NJoinCaptainKewl [0] (
21:35:00NJoinJ-23 [0] (
21:35:00NJoinjordan` [0] (i=gromit@
21:35:00NJoinscorche [50] (n=scorche@rockbox/administrator/scorche)
21:35:26CIA-70New commit by Ubuntuxer (r21353): added myself to the committers
21:36:01 Quit Lss (Read error: 104 (Connection reset by peer))
21:36:03mcuelenaereUbuntuxer: \o/
21:38:15*LambdaCalculus37 congratulates our newest committer \o/
21:39:00*mcuelenaere awaits his/her first commit
21:40:15Mikachui would go with his
21:40:20LambdaCalculus37mcuelenaere: Besides adding to docs/COMMITTERS?
21:40:59Bagderhe is a he
21:41:07mcuelenaereLambdaCalculus37: wasn't Ubuntuxer mostly accepted because of the fact that more patches were going to get committed?
21:41:15Mikachuand that wasn't prejudice, i checked the commit and saw "johannes" :)
21:41:28LambdaCalculus37mcuelenaere: Yes, mostly plugin fixes that should really go in IMHO.
21:41:28saratogaUbuntuxer: congrats
21:41:38saratogaoh hes not online
21:43:03 Join matsl [0] (
21:46:43 Quit bmbl ("Bye!")
21:53:33 Nick ocean__ is now known as ocean[afk] (i=d59c23e8@gateway/web/freenode/x-477f77eac10a1cbf)
21:59:00 Join CaptainKwel [0] (i=2669ecc2@gateway/web/freenode/x-909b8d2d1aed7794)
21:59:19 Quit LambdaCalculus37 ()
22:03:17 Quit ender` (" It's bad luck to be superstititious. -- Law of Superstition")
22:05:00 Quit evilnick (Ping timeout: 180 seconds)
22:06:15 Quit mcuelenaere (Remote closed the connection)
22:07:14 Part CaptainKwel
22:09:21 Join ender` [0] (
22:11:26 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
22:11:47 Quit kperri (Read error: 110 (Connection timed out))
22:12:05 Quit low_light ("ChatZilla 0.9.84 [Firefox 3.0.7/2009021910]")
22:13:15 Join edhelas [0] (
22:13:22edhelashye :)
22:13:53 Nick zitune[afk] is now known as zitune (n=zitune@bearstech/zitune)
22:18:38 Join Hillshum [0] (n=chatzill@unaffiliated/hillshum)
22:21:54 Quit Zarggg ()
22:31:31 Quit funman ("free(random());")
22:34:27 Quit JdGordon| ("Page closed")
22:34:56 Join Zarggg [0] (
22:39:50webmindrockbox boots again
22:39:56webmindbut no keys are working
22:40:04webmindI think I broke the interface connector :(
22:44:16***Saving seen data "./dancer.seen"
22:44:18 Join Ac-town [0] (
22:46:34 Join timc [0] (n=aoeu@
22:46:50 Join JdGordon| [0] (i=ad809766@gateway/web/freenode/x-ad6f68dfe17a5465)
22:47:17 Quit n1s (Read error: 110 (Connection timed out))
22:48:21 Quit ocean[afk] ("Page closed")
22:54:25 Join _lifeless [0] (n=lifeless@
22:55:22pixelmaLlorean: I experienced the wrong resumed playlist with songs added by "insert shuffled" too on my Ondio, running r21323. If I'm not confusing things, the playlist position index in the WPS showed the correct position in the playlist but I need to try and reproduce...
22:56:57 Quit edhelas (Read error: 113 (No route to host))
22:58:30 Join edhelas [0] (
23:00:11 Quit JdGordon| (Ping timeout: 180 seconds)
23:01:25 Join dmb [0] (n=dmb@unaffiliated/dmb)
23:02:13CIA-70New commit by bagder (r21354): Initial commit of early work on the new build server system. See ...
23:03:01 Quit Thundercloud (Remote closed the connection)
23:04:48edhelashttp:// I'm working on a new version of the wesite interface of rockbox wesite, someone is interested ?
23:06:49 Join __lifeless [0] (n=lifeless@
23:06:55Zagoredhelas: we are working on a new version too:
23:07:14edhelasgood :D
23:07:48 Quit d3v14710n (Read error: 60 (Operation timed out))
23:09:25edhelasgood job ;)
23:11:29 Join B4gder [241] (n=daniel@rockbox/developer/bagder)
23:12:47 Quit Hillshum ("ChatZilla 0.9.83 [Firefox 3.0.3/2008092417]")
23:12:50 Quit jfc (Read error: 104 (Connection reset by peer))
23:18:18 Join matsl_ [0] (
23:18:33 Quit matsl (Read error: 104 (Connection reset by peer))
23:18:37 Quit _lifeless (Read error: 113 (No route to host))
23:20:32 Quit pyro_maniac ("Leaving.")
23:20:55*amiconn wants a better mpegplayer benchmark than just staring at fps values
23:21:30LambdaCalculus37amiconn: What about a test_mpeg plugin for video benchmarks?
23:21:47amiconnA plugin for a plugin?
23:21:51 Quit Bagder (Read error: 110 (Connection timed out))
23:22:25LambdaCalculus37Now that I think about it, it is kind of stupid.
23:26:52*amiconn is a bit disappointed
23:27:32amiconnI've got the armv5+ idct working, but it looks like it isn't faster than the plain one on arv6 at least
23:27:46amiconn(mpegplayer idct that is)
23:28:14amiconnThere's some hope for the clamp/add/store stage though
23:37:14matsl_B4gder: Hows devcon going?
23:37:45B4gderbeer, pizza, crazy devs
23:37:52B4gdercouldn't be better!
23:38:06matsl_No midsummer snaps?
23:38:17B4gderbelgium beer makes up for it
23:38:36ZagorI tought about bringing sill and snaps, but the silly airlines would have stolen it anyway
23:39:34matsl_Your all sitting up and even standing. Are you sure you had any beer?
23:39:59matsl_It is a live cast isn't it?
23:40:03 Join funman [0] (n=fun@rockbox/developer/funman)
23:40:12B4gderwe need more beer ineed
23:40:17*B4gder adds a d
23:41:44funmanbertrik: did you bring your c200v1 to devcon?
23:42:13 Join n1s [0] (n=n1s@rockbox/developer/n1s)
23:42:17GodEatermatsl_: it's a live feed yes
23:43:23matsl_GodEater: Well bagder talks about beer and I see only coffe cups ;-)
23:43:35ZagorB4gder: ineed, youneed, weneed?
23:44:07funmanmatsl_: irish coffee perhaps O:-)
23:44:28matsl_Ah... missed that
23:44:45*pixelma wonders whether she should finally commit the c200 keymap patch...
23:44:45bertrikfunman, yes
23:45:02*B4gder spots many beer bottles here, but perhaps the beer bottle filter is on the cam
23:45:09AlexPpixelma: yes
23:45:14*bertrik votes yes
23:45:29funmanbertrik: could you give a try to ? (i swear, last attempt)
23:45:32pixelmaI know.... still a bit hesitant as I'm not fully satisfied with it (but when will that ever be?)
23:45:39GodEaterWE'VE DRUNK ALL THE BEER!!!!
23:45:51AlexPpixelma: It is still much better than it is
23:46:10funmanGodEater: use get_beer_buffer()
23:46:23GodEaterbuffer underun
23:47:16 Join flydutch [0] (
23:47:19pixelmaeverything at once or in small chunks?
23:47:54funmanpixelma: i think everything at once is ok, keymaps are easy to review in the file
23:48:09AlexPI agree
23:50:19bertrikfunman, 41 fps at 30 MHz
23:50:29Mikachui think you mean beer underflow
23:50:43matsl_Must disconnect now. Happy devcon hacking!
23:50:50funmanbertrik: thanks
23:51:13bertrikI'll revert the change and test again just to be sure
23:51:28 Quit ender` (" Sometimes I lie awake at night, and I ask, "Where have I gone wrong?" Then a voice says to me, "This is going to take more")
23:51:48 Quit matsl_ ("Leaving")
23:52:52funmani would like to see how this affects c200v2 and perhaps use 2 different loops
23:53:26funmanperhaps make lcd_send_data() a macro for c200v1
23:55:21 Join MrDuck [0] (

Previous day | Next day