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 2007-08-18

00:00:01DefineByteithe iPod video definitely does (don't have a nano). there's a post on flyspray saying Slasheri figured out the registers for it. probably rubbish though.
00:00:06*amiconn is still not 100% sure what's the best way to implement that
00:00:16DefineBytelinuxstb said it. :D
00:00:37linuxstb_And it was rubbish - Slasheri later said he hadn't
00:00:48Nico_Pamiconn: didn't know about memoserv, I'll check it out
00:00:49DefineByteah well. ^^
00:00:49austriancoderamiconn: I am looking for a place to call usb_stack_shutdown(), which unloads current used driver, shutsdown controller...
00:01:01 Quit mexicankiller ("ChatZilla [Firefox]")
00:01:02amiconnNico_P: /msg memoserv help
00:01:08Nico_Pthanks :)
00:01:26***Saving seen data "./dancer.seen"
00:02:04amiconnaustriancoder: Wouldn't it be better to do that on disconnect rather than just on shutdown?
00:02:22amiconnWould save battery power, maybe quite a bit
00:02:46*amiconn remembers the power drain from the isp1362 on H300, until he figured out how to shut it down
00:04:37amiconnIdeally the chip will be run in a mostly-asleep configuration, just enough to detect a connection. If a connection is detected, it would then be powered up completely in order to do its work. And on disconnect it would be put to sleep again
00:04:47austriancoderamiconn: on a disconnect I do a usb_stack_stop, which stops the controller, but on shutdown I am doing this:
00:06:03austriancoderamiconn: so on shutdown I tell the current used usb device driver to kill him, free the linked lists and do a call to controller->shutdown (which gets removed, as its not needed, i think)
00:06:27peturis there any reason not to do the shutdown on disconnect?
00:07:07amiconnOn a regular shutdown I wouldn't bother doing that. It would be necessary for suspend though
00:07:41amiconnRight now I wouldn't bother at all, first the stack needs to work
00:07:49austriancoderpetur: on the disconnect I call usb_stack_stop(), which turns of the used controller.. on usb_stack_shutdown i clean up things
00:08:17austriancoderamiconn: stack works... arcotg driver is buggy -> can not transmit responsed to usb requests
00:08:57amiconnHow do you know that the stack works. Do you have another target+ driver to test with?
00:09:13 Join qweru [0] (
00:10:46DefineByteokay, now i've confused myslef again. what does HAVE_BACKLIGHT_BRIGHTNESS tell me?
00:11:51austriancoderamiconn: no I dont have an other target, but the stack does only connect controller with driver and notify controller about irq... the real work, does the arotg driver, he checks if he can handle the usb request, if not push the request to the used device driver. If no one handels the request a stall happens.... so the stack is not the big bad boy... its the arcotg dirver -> so I would say the stack works... but device drivers are
00:12:25preglowirc doesn't like lines that long
00:12:30preglowyou should split them up
00:12:51*amiconn only believes that something is working when he (or another dev) actually sees it working
00:13:13*petur is with amiconn
00:13:24amiconnEverything else is "should work in theory"
00:13:55 Quit Nico_P (Remote closed the connection)
00:14:05austriancoderamiconn: okay.. but I need help to get the $%"$%$% arcotg controller tx data... and as petur dont own an pp based target he can't help me that much
00:14:19preglowwhat target?
00:14:20DefineByteor 'heard it working' i guess. :)
00:14:22peturso, as my boss would say, you have 2 days and 2 nights to get it working :p
00:14:34austriancoderamiconn: but maybe you
00:15:11austriancoderpreglow: pp based.. code works on e200 and ipod video with 64 mb ram - other are not tested
00:15:20amiconnI have some PP502x targets with usb, yes
00:15:29preglowonly think i have that's pp is a nano
00:15:37preglowbut i doubt i have the time to do testing
00:16:00amiconnIt shouldn't matter which PP502x target, as long as it has usb (and all of them do afaik)
00:16:28austriancoderjhMikes tryed to help me last night... but he had problems with gpio interrups
00:16:45austriancoderamiconn: do you have time to help me?
00:17:02DefineBytedoes HAVE_BACKLIGHT_BRIGHTNESS equate to hardware pwm or somesuch?
00:17:09amiconnDefineByte: yes
00:17:30amiconnBut it's of course only defined for those targets where we know how to use it
00:19:31DefineByteyeah, the patch i'm looking at adds it for ipod video/nano for some reason. it's pretty dirty
00:20:00amiconnMore precisely, it tells the app layer that the firmware layer can adjust brightness
00:20:48amiconnIt doesn't define how this is done, but the only acceptable way is doing it in hardware
00:20:56austriancoderamiconn: you have put usb and fw detection/... into usb-fw-pp502x.c - is it okay, if i rename file back to usb-pp502x.c and come with a little fw api and put it into fw-pp502x.c ?
00:22:43amiconnI wanted to do that myself, in order to distinguish usb and firewire, but atm that's not done, and hence the function needs to report both connection types the same
00:23:17 Join datachild [0] (
00:24:36amiconnWhen doing that, most of the current USB_* macros will be replaced by the more generic STORAGE_*
00:25:14amiconn(likewise for functions, except of course for the lowest layer, which will come in 2 flavours)
00:25:42austriancoderamiconn: okay... so i will not touch it and will leave it... as usb_enable(true) is only called by bootloader the stack will work when rb is started... later if fw and usb is seperated we can use the stack in the bootloader too
00:26:17amiconnThe bootloader doesn't touch usb or firewire at all atm
00:29:18DefineByteso basically the patch detects whether the target is a video/nano and then creates software pwm routines for backlight brightness. acknowledgeing the fact that this isn't the best idea is their an official way to detect the target ()in the absence of CONFIG_BACKLIGHT?
00:29:33DefineBytei knid of feel dirty ofr asking it. :(
00:29:44austriancoderamiconn: ah.. yes I see...
00:30:59amiconnYou wouldn't need that if you put the code in the target tree. But imo such a patch won't get accepted into svn
00:31:55*amiconn wonders whether there are similarities between firewire masstorage and usb masstorage classes
00:32:17DefineBytei think it's just a fudge till proper hardware pwn is discovered.
00:32:52DefineBytemost people seem to find the iPod too bright so it was created I guess
00:32:56 Join mexicankiller [0] (n=chatzill@
00:34:01mexicankillerHi guys,any good news about the sansa port in general?,I'm a bit scared DEAN is getting closer and it may strike here in the next days :(
00:34:15 Part latchem1
00:34:16amiconnThen imho it would make more sense to try figuring out hardware pwm rather than doing a software pwm patch
00:34:30scorchemexicankiller: i have told you before where to look for progress...
00:35:05scorchethen again, i dont quite get what you are asking anyway
00:35:28DefineByteokay, if i was clueless on such matters but was willing to learn any good reading i could do?
00:37:39DefineBytethe fack that i'm asking that question probably means i'm not the best person for the job, but hey. :)
00:38:03*preglow waits for norwegiandriver to join
00:40:30 Join ddalton [0] (
00:40:42Derkwhat's wrong with the sansa port?
00:40:52 Quit ddalton (Client Quit)
00:41:58 Quit Thundercloud (Remote closed the connection)
00:42:29DogBoygood question
00:42:40DogBoywhat are you talking about mexicankiller
00:43:18mexicankillerrefering to ?
00:43:23mexicankillerthe sansa port?
00:43:39DogBoythat part was clear
00:43:41DogBoywhat about it
00:44:05 Quit Wofl (Remote closed the connection)
00:44:46mexicankilleris just that I wanted to now the a bit of the changes made to it
00:47:29DefineByteah well, thanks for asking my stupid questions. good night all
00:47:38 Part DefineByte
00:49:27*petur swears a bit more at the settings and menu code
00:50:18 Quit ender` (" Remember: A secretary isn't permanent until she's been screwed on the desk...")
00:53:30 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
01:05:47 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
01:10:54austriancoderthe usb mass storage driver will be of CBI type
01:16:32peturaustriancoder: is that august 17 patch the latest/greatest?
01:19:04peturit applies (only one error) but doesn't compile
01:19:48peturlinkedlist redefinition error
01:23:07*petur kicks austriancoder awake
01:25:16peturand regarding your comment in core.c line 68: are you planning to use overlays for driver loading? Maybe a bit overkill?
01:35:03 Join midkay_ [0] (
01:35:45 Quit petur ("Zzzzzzzzzzz")
01:39:45 Join hannesd_ [0] (
01:43:49 Quit pLy`Br3iZh (Read error: 110 (Connection timed out))
01:44:13 Join midgey [0] (
01:48:03 Join jhulst [0] (
01:48:24 Nick little is now known as r4 (n=agnth@
01:49:04 Quit barrywardell ()
01:53:12 Quit midkay (Read error: 110 (Connection timed out))
01:54:30 Quit hannesd (Read error: 110 (Connection timed out))
01:54:31 Nick hannesd_ is now known as hannesd (
02:01:28***Saving seen data "./dancer.seen"
02:07:21 Part n1s
02:10:39 Join Kyron [0] (i=458969f7@gateway/web/cgi-irc/
02:11:01 Quit mexicankiller (Read error: 110 (Connection timed out))
02:12:04 Quit jhulst (Remote closed the connection)
02:12:04 Quit Kyron (Client Quit)
02:13:03 Join robin0800 [0] (
02:15:18 Quit iamben (Remote closed the connection)
02:25:53 Quit midgey (Read error: 104 (Connection reset by peer))
02:26:28 Join sarixe [0] (
02:26:41 Quit Arathis ("Verlassend")
02:27:22 Join _Thomas_ [0] (
02:30:02 Quit robin0800 (Read error: 104 (Connection reset by peer))
02:30:10 Part _Thomas_
02:30:26 Join iamben [0] (
02:33:07 Join Gibbed [0] (
02:33:11 Quit Rick (Nick collision from services.)
02:33:25 Nick Gibbed is now known as Rick (
02:39:38 Join stateq2 [0] (
02:40:06 Part stateq2
02:42:26 Join x1jmp [0] (
02:46:39einhirnWell well, am I the last one not sleeping yet?
02:48:27 Quit bdgraue (Remote closed the connection)
02:49:29einhirnAnyway - Does anyone know "LANG_REPLAYGAIN_NOCLIP"? Somehow the compiler died on the Line in "settings_list.c" containing it. I already grepped the Source tree but didn't find it anywhere except in some language files...
02:52:56 Quit Rick (Read error: 110 (Connection timed out))
02:52:58x1jmpI have it in settings_list.c on line 897
02:56:02 Join webguest00 [0] (i=4208822a@gateway/web/cgi-irc/
02:56:43 Join ctaylorr [0] (
02:57:18 Quit webguest00 (Client Quit)
02:58:03pixelmaeinhirn: it got excluded from the hwcodec settings (in english.lang) - somehow that slipped in there although I tried to keep that out there because someone said that replaygain could be made possible on hwcodec
02:58:38einhirnI'll check... ;)
02:58:45pixelmamaybe it's excluded somewhere else to
02:58:51pixelmatoo even
02:59:28pixelmamost of the other replaygain settings should be ok
02:59:35einhirnYes they are...
03:00:59pixelmaIIRC, someone said that "prevent clipping" wouldn't be possible though as that means a "live check" all the time, or something
03:01:34pixelmabut I'm not too much in the technical details, so...
03:04:03einhirnThanks for the tip with the language files.
03:05:04einhirnWell Replay Gain Tags contain a nifty little feature "Peak value", so you wouldn't need to check the whole time, but could adjust the Gain so that the Peak Value doesn't clip.
03:06:02 Join stateq2 [0] (
03:09:28pixelmayou should discuss it with a more knowledgable person... noticing my very nifty phrasing, I should probably get some sleep
03:12:00DerPapstnight :)
03:12:13 Part pixelma
03:20:28 Quit x1jmp (Remote closed the connection)
03:23:12 Join Rick [0] (
03:24:36 Quit Shaid ("I will see you, in the end. And I will laugh at your pain...")
03:31:13 Join Llorean [0] (
03:49:53 Join Gibbed [0] (
03:49:57 Quit Rick (Nick collision from services.)
03:50:11 Nick Gibbed is now known as Rick (
03:50:38 Nick fxb is now known as fxb__ (
03:50:39 Join aliask [0] (
03:56:25 Join Gulopine [0] (
03:57:39Gulopinehello all
04:01:16einhirnHi Gulopine...
04:01:31***Saving seen data "./dancer.seen"
04:02:19 Quit Rick (Read error: 104 (Connection reset by peer))
04:02:31 Join Rick [0] (
04:02:51Gulopinejust wanted to drop in and thank you folks for all the great work
04:03:06Gulopinei finally decided to put more on my ipod than what apple sent, and i'm pleasantly surprised
04:06:08 Join Gibbed [0] (
04:06:12 Quit Rick (Nick collision from services.)
04:06:26 Nick Gibbed is now known as Rick (
04:06:56einhirnYup, I'm mostly user too - and on one of the first Hardwares supported ;) I just hack into the source sometimes, hopefully implementing something I (and perhaps others) can use ;)
04:07:29Gulopineyeah, i'm not much good with c, so i doubt i'd be any good with helping
04:07:41Gulopinei'll probably try my hand at something in the future, but it's not my priority for now
04:08:15Gulopinei was especially impressed to see support for game music
04:08:36Gulopinei'm very pleased to put loads of nes and snes music on there without the extra space mp3 requires
04:08:50einhirn*g* Yes, and you can (probably) play Doom on your Ipod now ;)
04:09:48Gulopineyeah, but why would i want to do that? :(
04:11:18Gulopinei was a little disappointed that it doesn't support rsn packages for snes music, but life works that way
04:12:07LloreanWhen you've got a lot of space, and minimal CPU power, archiving them isn't as great a solution.
04:12:29Gulopinewell, no
04:13:00Gulopinebut when you download an archive of them in rsn format, which filenames stripped down, relying on id666 ...
04:13:16Gulopineit becomes a bit of a chore to extract and rename them
04:13:21Gulopinei haven't gotten around to doing that yet
04:15:53LloreanAlso, multi-track RSN files are particularly bad for use on a portable.
04:16:01LloreanYou have to decompress all the songs before the one you want to play.
04:16:55LloreanIt's really a format that assumes you have a lot of processing power available.
04:17:08Gulopinei'm a big game music fan, so that's one area i'm interested in
04:17:35LloreanWell, it's really not likely to happen, and if it does I think most people (probably all those not using a Gigabeat) will simply find it unusably slow.
04:17:48LloreanYou might want to get working on extracting your SPC collection. ;)
04:17:59Gulopineheh, no, i didnt mean i'm interested in rsn :)
04:18:04Gulopinei mean i'm interested in game music
04:18:15Gulopinei'm find extracting the collection, that doesn't really phase me
04:18:21Gulopineit was a momentary disappointment, nothing more
04:19:21Gulopinethe only thing i might possibly think could be better is if nsfs could be resumed at a particular track
04:19:47Gulopinebut i'm certainly not one to look a gift horse in the mouth
04:22:10LloreanI think NSF could still do with a few improvements, yeah.
04:24:25 Quit Rick (Connection timed out)
04:26:27 Join mexicankiller [0] (n=chatzill@
04:46:27 Join Rick [0] (
04:54:07 Join advcomp2019- [0] (n=advcomp2@
04:59:54 Quit iamben (Read error: 104 (Connection reset by peer))
05:02:47einhirn... need ... sleep ... now ...
05:03:03*einhirn tugs himself in.
05:03:15 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
05:04:40 Join perrikwp [0] (n=chatzill@
05:09:47 Part Gulopine
05:10:38 Join iamben [0] (
05:12:27 Quit advcomp2019 (Read error: 110 (Connection timed out))
05:12:30 Quit DerPapst ("So Long And Thanks For All The Fish!")
05:14:44 Nick advcomp2019- is now known as advcomp2019 (n=advcomp2@
05:24:06 Quit rotator ()
05:27:16 Quit Rick (Read error: 110 (Connection timed out))
05:28:11 Join Rick [0] (
05:55:35 Quit ToHellWithGA (Read error: 110 (Connection timed out))
06:00:22 Quit malsyned (Read error: 110 (Connection timed out))
06:01:32***Saving seen data "./dancer.seen"
06:10:01 Join jhulst [0] (
06:15:43 Quit jhulst (
06:15:43 Quit perrikwp (
06:15:43 Quit mexicankiller (
06:15:43 Quit Llorean (
06:15:43 Quit stateq2 (
06:15:43 Quit sarixe (
06:15:43 Quit Derk (
06:15:43 Quit GodEater (
06:15:43 Quit ptw (
06:15:43 Quit Bagder (
06:15:43 Quit merbanan (
06:15:43 Quit austriancoder (
06:15:43 Quit RaRe` (
06:15:43 Quit GodEater_ (
06:15:43 Quit linuxstb (
06:15:43 Quit blithe (
06:15:43 Quit enyc (
06:15:43 Quit joshin (
06:15:43 Quit Kohlrabi (
06:15:43 Quit fxb__ (
06:15:43 Quit DiDjCodt (
06:15:43 Quit cmvjk (
06:15:43 Quit lostlogic (
06:15:43 Quit dionoea (
06:15:43 Quit jbit (
06:15:43 Quit mtoups (
06:15:43 Quit _jz (
06:15:43 Quit maddler (
06:15:43 Quit Rick (
06:15:43 Quit ctaylorr (
06:15:43 Quit chrisjs169 (
06:15:43 Quit darkless (
06:15:43 Quit BHSPitMonkey (
06:15:43 Quit courtc (
06:15:43 Quit bospaadje (
06:15:43 Quit krazykit (
06:15:43 Quit tonyyarusso (
06:15:43 Quit alberink_ (
06:15:43 Quit hannesd (
06:15:43 Quit midkay_ (
06:15:43 Quit gtkspert (
06:15:43 Quit Soap (
06:15:43 Quit ze (
06:15:43 Quit Bjoern-Erik (
06:15:43 Quit indro (
06:15:43 Quit thegeek (
06:15:43 Quit preglow (
06:15:43 Join Rick [0] (i=rick@unaffiliated/rick)
06:15:43NJoinctaylorr [0] (
06:15:43 Join chrisjs169 [0] (n=chrisjs@unaffiliated/chrisjs169)
06:15:43NJoindarkless [0] (
06:15:43NJoinBHSPitMonkey [0] (
06:15:43NJoincourtc [0] (n=court@unaffiliated/courtc)
06:15:43NJoinbospaadje [0] (
06:15:43NJoinkrazykit [0] (
06:15:43NJointonyyarusso [0] (n=anthony@ubuntu/member/tonyyarusso)
06:15:43NJoinalberink_ [0] (n=alberink@
06:15:45***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
06:15:48NJoincmvjk [0] (
06:15:48NJoinjbit [0] (
06:15:48NJoinlostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
06:15:48NJoinmaddler [0] (
06:15:48NJoin_jz [0] (
06:15:48NJoindionoea [0] (
06:15:48NJoinmtoups [0] (n=mtoups@HENSON.ISR.CS.CMU.EDU)
06:15:55NJoinjhulst [0] (
06:15:55NJoinperrikwp [0] (n=chatzill@
06:15:55NJoinmexicankiller [0] (n=chatzill@
06:15:55 Join Llorean [0] (n=llorean@rockbox/administrator/Llorean)
06:15:55NJoinstateq2 [0] (
06:15:55NJoinsarixe [0] (
06:15:55NJoinDerk [0] (
06:15:55NJoinGodEater [0] (n=bryan@rockbox/staff/GodEater)
06:15:55NJoinptw [0] (
06:15:55NJoinBagder [0] (n=daniel@rockbox/developer/bagder)
06:15:55NJoinmerbanan [0] (n=banan@
06:15:55NJoinaustriancoder [0] (n=austrian@rockbox/developer/austriancoder)
06:15:55NJoinRaRe` [0] (n=Laffin_B@
06:15:55NJoinfxb__ [0] (
06:15:55NJoinGodEater_ [0] (n=bryan@rockbox/staff/GodEater)
06:15:55NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
06:15:55NJoinblithe [0] (
06:15:55NJoinenyc [0] (
06:15:55 Join joshin [0] (n=joshin@unaffiliated/joshin)
06:15:55NJoinDiDjCodt [0] (
06:15:55NJoinKohlrabi [0] (
06:15:57***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
06:15:57***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
06:15:58 Quit Neovanglist (Connection reset by peer)
06:16:00 Join Neovanglist [0] (
06:16:33NJoinhannesd [0] (
06:16:33NJoinmidkay_ [0] (
06:16:33NJointhegeek [0] (
06:16:33NJoingtkspert [0] (
06:16:33NJoinSoap [0] (n=Soap@rockbox/staff/soap)
06:16:33NJoinze [0] (
06:16:33NJoinBjoern-Erik [0] (
06:16:33NJoinindro [0] (i=indro@
06:16:33NJoinpreglow [0] (
06:17:07 Part Llorean
06:17:16 Quit FOAD (Remote closed the connection)
06:17:27 Join FOAD [0] (
06:22:08 Join tonyy [0] (n=anthony@ubuntu/member/tonyyarusso)
06:26:24 Quit stateq2 (Success)
06:26:56 Quit tonyyarusso (Nick collision from services.)
06:27:00 Nick tonyy is now known as tonyyarusso (n=anthony@ubuntu/member/tonyyarusso)
07:00:08 Part Derk
07:15:28 Quit advcomp2019 (Read error: 110 (Connection timed out))
07:26:32ptwdoes anyone know if arcotg_udc.c implements device mode, or if it just implements usb otg?
07:28:26ptwdoes it implement an ums class?
07:30:57 Join advcomp2019 [0] (n=advcomp2@
07:32:58aliaskI'm getting an error when compiling the main build: "rockbologo.h" missing. Rockboxlogo is supposed to be generated from the bmp file, but I guess it isn't. Anybody got ideas why?
07:34:48ptwhave you gotten it to build before?
07:35:00aliaskNot that I remember
07:35:14amiconnaliask: Is that for gigabeat S?
07:35:33ptwmy first guess is maybe there is a typo in a file somewhere?
07:35:48amiconnWhat's the display resolution of the S?
07:36:55scorchethen it is 240x320
07:37:01amiconnHmm, so 240x320. But it should work as it's the same as for gigabeat F/X
07:37:17aliaskYeah, I checked the bitmaps/native/SOURCES and all seems to be in order
07:37:30ptwhave you tried make w/ debug options?
07:37:35aliaskBut the makefile never does the BMP2RB stage
07:37:43amiconnmake V=1 or sth
07:38:28ptwi had something similar happen previously, and it was a typo in one of the headers preventing the build....but it could be somethign else
07:39:12aliaskI know something similar happens with sysfont.h if you have errors in the SOURCES files, but I can't see any.
07:39:33ptwtry make w/ verbose enabled
07:39:53aliaskDoing that now
07:40:26*amiconn wonders whether it's possible to test-build the gigabeat S branch
07:40:40ptwit build bootloader for sure atm
07:40:42*amiconn never worked with svn branches so afr
07:40:58aliaskOh, as in add it to the build table?
07:41:10 Nick ptw is now known as ptw419 (
07:41:13amiconnI mean checking out from the branch and building myself, to see what happens
07:41:42ptw419so, to ask my question again. does anyone know if arcotg_udc is only a usb otg driver?
07:45:20ptw419udc=usb device controller...
07:58:30 Join kubiix [0] (
07:59:38 Join Llorean [0] (
08:01:37***Saving seen data "./dancer.seen"
08:06:30 Join davina [0] (
08:09:50aliaskIn a make file, does % act similarly to * (matches all strings)?
08:10:57amiconnHow do I check out from a branch?
08:11:35*amiconn should probably read himself
08:11:50aliaskJust as you would from trunk, but replace trunk with /branches/gigabeat-s (assuming you want the gigabeat-s branch)
08:13:04 Join datasleep [0] (
08:13:28 Quit datachild (Nick collision from services.)
08:16:40amiconnThe build already bails out in usb.c
08:17:03aliaskThat's an easy fix though, just need to add some key defines
08:17:39aliaskNote that this is the first time i've ever tried compiling the "normal" build for the gigabeat s, so it's no supprise that it isn't really working.
08:18:21amiconnYeah, but if firmware/ doesn't compile, how do you get as far as having a problem in apps/ ?
08:18:45aliaskusb.c doesn't get compiled in the bootloader builds, so we'd never noticed it before
08:19:50aliaskI just added the defines in my local build, it's just near the top of usb.c
08:20:13 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
08:20:21amiconnThere are several other errors on firmware. Even make -k doesn't get as far as compiling apps/
08:20:43amiconnMaybe there's a problem in the generated Makefile
08:23:41aliaskHrm, odd - make -k built the bitmaps, whereas regular make didn't.
08:23:46 Quit kubiix (Read error: 110 (Connection timed out))
08:24:21amiconnmake -k doesn't stop on errors
08:24:43amiconnSo tehre's an error before it comes to building bitmaps
08:29:47aliaskYep, I think I just found it, thanks for pointing me to make -k
08:30:15amiconnThere are lots of warnings in other stuff, and errors in pcm_playback
08:30:36*amiconn wonders how merging from branches works
08:31:02aliaskYes, I'm not sure why other targets don't build pcm_playback - there are errors in the trunk version
08:31:24amiconnI mean errors in the gbs branch
08:31:34amiconntrunk builds fine for all targets
08:31:53amiconnOf course not all targets use pcm_playback. Only swcodec targets do
08:32:11aliaskYes, I know. I can't work out why, say the gigabeat f build, doesn't build pcm_playback.c, and only builds pcm-meg-fx.c
08:32:58amiconnRegarding the merge - I noticed that the branch has quite some stuff that's already reworked/moved around/replaced in the trunk
08:34:59 Quit mexicankiller ("ChatZilla [Firefox]")
08:39:43ptw419we should merge it soon probably
08:39:59ptw419to keep up w/ the current build
08:40:24ptw419if we wait longer, it will probably break further when we try to a final merge
08:44:32 Join miepchen^schlaf [0] (
08:50:59linuxstb_ptw419: If I was you, I would merge as often as possible - especially when things change a lot in firmware/ (such as some of amiconn's recent commits)
08:52:50 Join Rob2222 [0] (
08:53:46XavierGrmedia coder is a fine gui front end for mencoder. Much easier to transcode many videos for my gigabeat
08:54:30 Join petur [0] (n=petur@rockbox/developer/petur)
08:55:24LloreanXavierGr: Tried WinFF yet?
08:56:15 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
08:56:23XavierGrLlorean: not yet, do you think it is better than mediacoder?
08:56:41LloreanIt's my personal favorite, though I'm sure tastes vary
08:56:59 Quit jhulst ("Konversation terminated!")
08:57:22XavierGrI will make sure to try it next time
09:06:39 Quit ptw419 ()
09:07:22aliaskIs IRAM the same as ICACHE?
09:09:28aliaskNever mind, I found what I was looking for.
09:10:34 Join advcomp2019- [0] (n=advcomp2@
09:11:02 Quit Rob222241 (Read error: 110 (Connection timed out))
09:11:14 Quit advcomp2019 (Nick collision from services.)
09:11:17 Nick advcomp2019- is now known as advcomp2019 (n=advcomp2@
09:14:48amiconnpetur: You peak release fix isn't entirely correct
09:15:22amiconnThere used to be a unit, "units per read".
09:16:34amiconnWe could add that unit to the list of units, but in fact I don't like that unit very much. The peakmeter should instead base its release behaviour on time, i.e. ticks, not a (target dependent and somewhat varying) readout rate
09:17:42 Quit sarixe ("Peace")
09:21:30 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
09:25:52 Quit aliask ("ChatZilla [Firefox]")
09:32:22XavierGrLlorean: WinFF had problems transcoding my current files. Lots of errors and no support for subtitles. (maybe I got the errors because the video has 23.98 fps?) Though I have to say that it looks simpler and faster
09:33:45LloreanI don't know about subtitle support, honestly.
09:33:57LloreanAs to your errors, it probably depends on how current your version of ffmpeg is
09:34:40XavierGrwell technically they should have behaved the same given that they use the same mencoder commandline?
09:37:09 Join darkraven [0] (
09:37:13LloreanWinFF uses the straight ffmpeg binary. Is that what mencoder does?
09:39:52XavierGrit seems that mencoder doesn't use ffmpeg binary (I am not sure though) But Media Coder has both
09:48:16 Join n1s [0] (
09:49:11 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
09:49:52pixelmamorning n1s
09:49:59peturamiconn: I see... I will have a look at that peakmeter code
09:50:15 Part Llorean
09:50:33peturaustriancoder: awake?
09:50:48pixelmadid you see my (kind of) question yesterday about the lang string of the "disk full" splash?
09:51:10*n1s reads the log
09:51:47pixelmaI wondered where the "h100,h120,h300" has gone
09:52:30n1shmm, I have no idea
09:53:01linuxstb_XavierGr: mencoder uses the libraries from ffmpeg, but it is a standalone app in its own right. IIUC, it adds more video/audio processing features (i.e. all those present in mplayer) on top of core ffmpeg, so mencoder will give you far more options than just ffmpeg. e.g. I don't think ffmpeg can render subtitles, but mencoder can.
09:53:26pixelman1s: before there were 2 different strings - one for the irivers (with the "stop" button) and one for the other recording devices (with "off" button)
09:53:53 Quit ctaylorr (Read error: 110 (Connection timed out))
09:53:59pixelmaI also found a few other thing in english.lang that I'd like to be changed
09:54:57n1spixelma: ok, either tell me or change them yourself :-)
09:55:55XavierGrlinuxstb: thanks for clarifying. At least know I know which utility to use.
09:56:15amiconnHmm, the 1st gen draws a bit more power than the 2nd gen
09:57:31pixelma1) I found 2 unused strings (VOICE_EXT_TXT, LANG_RECORD_CURRENT_DIR), 2) in LANG_BUILDING_DATABASE, you corrected the mentioned button for x5,m5 but it doesn't follow the convention of writing the button all uppercase
09:59:04pixelman1s: 3) throwing out as suggestion: I thought "Keys locked" or "Keys unlocked" would sound better than "Keylock is ON/OFF" - what do you think?
09:59:11n1spixelma: the lang v2 cleanup was what changed the LANG_DISK_FULL thing, will fix now, hmm wonder which button it is for other targets...
09:59:51pixelman1s: you can find that in not updated lang files (for example deutsch.lang in SVN)
10:01:00amiconnI also want to add some things regarding language: (1) We should decide whether we want to call it keys or buttons
10:01:06n1spixelma: found it in the svn diff, but there's only a specific button for irivers, no iaudios, ipods or anything
10:01:07pixelmathere is also a _SYSFONT_ version of the "disk full" splash
10:01:13amiconn(2) Hold is not a button or key, it's a switch
10:01:39***Saving seen data "./dancer.seen"
10:01:45n1sSeems like it's time for some minor cleanup :-)
10:02:09pixelman1s: for the rest there was a string that said "Press OFF to continue"
10:02:18linuxstb_"button" seems better, but it's longer...
10:02:32amiconnI'd also vote for 'button'
10:03:15pixelmaso that would make my suggestion "Buttons locked/unlocked"
10:03:27n1spixelma: yeah I found it, but I want to add correct buttons for all targets while I'm at it so I wondered if OFF is correct for all others than the irivers?
10:03:37*n1s votes for button too
10:04:08amiconnAh, the difference in power consumption is the wheel. Hence the enable bit...
10:04:33*amiconn made the hold switch disable the wheel in the G1/g2 button driver
10:04:50linuxstb_So turning on hold will save power?
10:04:50amiconnThat saves ~12mA (!)
10:05:02amiconnJust on 1st gen though
10:05:07pixelmait is OFF for Ondios... should be right for iaudios too (though haven't experienced it yet) but there only is the power off switch, no stop button
10:05:18 Join ouioui [0] (
10:05:42 Quit qweru ("KVIrc 3.2.0 'Realia'")
10:06:06linuxstb_amiconn: I wonder how the OF handles that (if at all)...
10:06:23pixelman1s: for Ipods that can record it should be STOP too, I guess? (Holding the play button)?
10:07:09n1spixelma: I found the action, just have to check the keymaps
10:08:29linuxstb_pixelma: Are you suggesting the message should be "Press STOP to continue" on ipods ? (or am I not following your conversation?)
10:09:54pixelmaI don't really suggest that - what do you have to press to get out of the "disk full" splash on Ipods?
10:10:33linuxstb_I've no idea...
10:12:30pixelman1s: whil you're at it... what amiconn was especially aiming at with the button and hold switch is the LANG_BACKLIGHT_ON_BUTTON_HOLD would probably sound better as "Backlight on Hold" only
10:12:54 Join desowin [0] (
10:13:30n1spixelma: sounds ok to me, Hold or HOLD?
10:13:41pixelmalinuxstb: on other devices it says "The disk is full. Press STOP to continue." (or OFF)
10:15:16pixelman1s: I tend to think all uppercase would be more consistent but I'm not sure
10:15:30amiconnI'd use lowercase
10:15:44amiconn(since hold isn't a button)
10:17:50pixelmalinuxstb: an opinion as native speaker on the keylock suggestion?
10:19:29 Quit darkraven (Read error: 113 (No route to host))
10:21:29pixelmaamiconn: but "buttons unlocked" wouldn't fit on the player's display but there's no possibility of shortening that...
10:21:57amiconnIt sure fits...
10:22:23amiconnSplashes wrap as necessary, and it's just 2 words with <= 11 letters each
10:22:45pixelmaah ok, they wrap
10:23:21 Join darkraven [0] (
10:23:56pixelmaso the different strings for player that are now are completely unnecessary
10:27:43 Quit perrikwp ("ChatZilla [Firefox]")
10:27:51amiconnlinuxstb: Hmm, I wonder how fast the wheel electronics are. Perhaps it's possible to only enable the wheel for a very short time per tick, and then only keep it powered (for a few ticks) if the value changed?
10:27:57 Join midkay [0] (n=midkay@rockbox/developer/midkay)
10:28:10amiconnThis would then also save power when hold is not enabled
10:29:51 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
10:32:10 Quit midkay (Client Quit)
10:32:52 Join midkay [0] (n=midkay@rockbox/developer/midkay)
10:33:18linuxstb_pixelma: What about just "Hold On" and "Hold Off" ?
10:33:46linuxstb_pixelma: I'm sorry, I'm not sure I understood your question...
10:35:45 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
10:35:59n1slinuxstb_: the keylock splash is for targets with no Hold switch
10:38:07 Join freqmod [0] (
10:38:34preglow12ma for the wheel???
10:39:38 Quit linuxstb_ ("CGI:IRC")
10:39:39 Join latchema [0] (
10:40:06amiconnpreglow: 1st gen...
10:40:29amiconnIt's a mechanical wheel. Probably it's the pull up (or pull down) resistors)
10:41:39 Quit midkay_ (Read error: 110 (Connection timed out))
10:45:08 Quit petur ("switching...")
10:45:17pixelmalinuxstb: if you record it could be that the harddisk fills up, if it is full, you get the mentioned splash (which stays there until you push the appropriate button to abort recording... I wondered which one it is on Ipods but if n1s said he found it in the code...
10:45:18 Join petur [0] (n=petur@rockbox/developer/petur)
10:48:57n1spixelma: it's quite funny actually because the sysfont and the regular version can't have the same buttons, because the regular is only used when recording from radio it uses the ACTION_FM_STOP while the sysfont version is in the recording screen and uses ACTION_STD_CANCEL and those are not the same for all targets (sansa, x5, m5) and the regular version will never be displayed for ipods (no fm)
10:49:04 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
10:49:35pixelmaoh, I see
10:57:03pixelmaamiconn: there?
10:59:28 Join kretender [0] (
11:00:23 Part kretender
11:06:57 Join Phill [0] (
11:09:57 Join ender` [0] (
11:13:22 Quit Nico_P (Remote closed the connection)
11:15:23 Join barrywardell [0] (n=barrywar@
11:17:26 Nick datasleep is now known as datachild (
11:20:45 Join einhirn [0] (
11:21:20 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
11:21:37 Quit r4 ("Leaving")
11:24:28einhirnDid the Developers take a look in the IRC Log of this night? At around 4am Gulopine dropped by to say thanks ;)
11:25:01pixelmahello einhirn, did you take out the restriction of LANG_REPLAYGAIN_NOCLIP being swcodec only in english.lang? Is that needed?
11:25:24einhirnYes, that was needed to get it to compile with that menu entry.
11:25:27einhirnAnd it worked.
11:25:51einhirnI also took that restriction out of the other .lang files that still contained it.
11:26:43einhirnNow I just have to find a way to replace "dsp_set_replaygain()" with something that puts the correct data into the MAS ;)
11:27:01pixelmasure... they'll get updated anyways (I'm working on the german lang file ATM, that's why I thought of it when you asked ;) ).
11:28:02einhirnAh, k ;)
11:29:26pixelmaI had a few remarks on english.lang with n1s and he said he'll work on that, so I thought if it's really needed that could be changed in SVN right now, too...
11:31:42peturpixelma: he did already a commit this morning
11:31:44 Part gtkspert
11:32:51pixelmathanks, didn't notice
11:33:50pixelmadamn, I wanted to suggest that earlier too... :/
11:34:13n1spixelma: gtg now, but if you include that string for hwcodec targets remember to mention in the commit message that it will break voice files for hwcodec targets...
11:36:12pixelmamaybe that has time until einhirn is a bit further with his patch, or at least until tomorrow evening when there is not such a big gap between commit and daily voice file builds
11:36:52*amiconn has dynamic wheel enable working on 1st gen
11:37:27amiconn50 s is enough to get a reading from the wheel, so it will only draw power for 0.5% of each tick, as long as it's not used
11:37:46n1samiconn: nice!
11:38:02amiconnIt then stays enabled for 1/4 sec after the last change
11:39:13 Quit Phill ()
11:39:58 Join kretender [0] (
11:45:19 Join spiorf [0] (
11:48:51pixelman1s: thanks for the quick reaction btw. :)
11:49:07 Part latchema
11:50:29*amiconn likes all those extra power savings in rockbox vs. ipl :P
11:51:37scorcheand the fact that it can boot twice in your case? ;)
11:52:10amiconnNever tried ipl on 1st/2nd gen
12:01:40***Saving seen data "./dancer.seen"
12:03:25*einhirn should have used a better editor for making his code changes.
12:04:05einhirnProton messed up the whitespace - again (I had the same problem the last time I worked on the code)...
12:04:34einhirnOh well, now I'll use Vim to start my changes from a clean SVN tree ;)
12:05:44barrywardellaustriancoder: I was looking at your mrh_try7 patch
12:05:45 Join Phill [0] (
12:06:27*petur wonders if austriancoder is already awake
12:06:46barrywardellI added back in the proper timeouts and jmp's and don't get timeouts in usb_send any more
12:07:02peturyou mean it is sending?
12:07:26barrywardellI'm not totally sure, but it's getting to the end of usb_send without errors
12:07:58peturgrrrr I made remarks in that direction already last month :(
12:08:14petur'no change' was his response
12:08:53*amiconn also verified that the wheel enable delay has enough safety margin as units might differ somewhat
12:09:06barrywardellthe timeouts definitely look wrong at least
12:09:28amiconn20 s were too short, 30 s already worked on my unit. So I think 50 s is a nice value
12:09:33peturyes, that was my thought too, which is why I asked him
12:15:36barrywardellpetur: here's my updated patch:">
12:15:58pixelmaI hope I didn't miss something
12:16:04barrywardellI just took mrh_try7 and added the lonjjmp and timer stuff as was suggested on irc a couple of days ago
12:16:47peturbarrywardell: thanks for putting time in this... I hope austriancoder comes around real soon...
12:19:05*barrywardell is keen to see USB support in PP targets
12:19:31*petur should have bought a pp target before gsoc :/
12:21:19 Join Shadows1990 [0] (i=54408fbb@gateway/web/cgi-irc/
12:21:30peturwould a H10 5GB do ok? I found one 2nd hand for 100 euro near my place... does that price sound ok?
12:21:56barrywardellthat sounds pricey
12:22:06Shadows1990Does anybody know if Rockbox has been complied successfully to the Elio P722
12:22:48barrywardellbut it would do just fine - it's a pp5020
12:23:50amiconnThe H10s seem to be quite pricey. I paid 87 EUR for mine
12:23:57amiconn(6GB, UMS)
12:24:04Shadows1990i looked on the current releases, and there was some mention about it on the forums, but no actual complete build
12:24:26peturI'm offering him 75, let's see if he responds
12:24:26Shadows1990does anyone know where i can find a working compile of it
12:25:23peturoh, I found another one for 50 :)
12:26:09 Join ctaylorr [0] (
12:26:31barrywardellpetur: doing some more debuging here. I recive s->bRequest 0x6 (get descriptor), then send, then receive s->bRequest 0x5 (set address)
12:26:55barrywardellset address doesn't seem to be implemented yet though
12:26:56Shadows1990Does anybody know about the Elio P722?
12:26:59*amiconn does actually deem usb device support on PP less important than usbotg support on H300 and X5
12:27:00peturdo you receive something on PC?
12:27:47peturamiconn: I'll resume usbotg work right after gsoc....
12:27:56peturunless you want the code ...
12:27:58barrywardellactually, it does set the address at the end of the function
12:28:21amiconnMissing host support is an actual lack of features, while rebooting into OF diskmode is just a minor inconvenience (slightly slower than if it would stay in rockbox)
12:28:55amiconnHowever, an ipod video user might see things different because of the diskmode slowness
12:29:06amiconnH10 OF diskmode is also quite slow btw
12:29:35amiconnBut then, our usb driver not only needs to work, it also needs to work fast to be a real advantage
12:30:12barrywardellpetur: I don't get very useful usb debug messages on my mac
12:30:23amiconnBoth mini G2 and 1st/2nd gen "emergency" diskmode are as fast as the OF's diskmode
12:30:24barrywardellpetur: AppleUSBEHCI[0x42b4000]::Found a transaction which hasn't moved in 5 seconds on bus 253, timing out!
12:30:44amiconn(in fact on 1st/2nd gen the "emergency" diskmode *is* the OF's diskmode)
12:31:27amiconn..i.e. the OF does the exact same thing as rockbox when detecting a firewire connection
12:31:31barrywardellamiconn: the thing that annoys me is that it is much slower to connect to usb because of the whole reboot proceesure
12:31:48barrywardellit's so much easier on e.g. the gigabeat
12:31:52amiconnbarrywardell: On H10? Rockbox doesn't detect USB there...
12:32:09barrywardellit does,but in a strange way
12:32:10pixelmaShadows1990: work only started (*pointing to linuxstb* ;) ). There is no working copy for the Elio afaik
12:32:24Shadows1990ahhh, thanks
12:32:30amiconnbarrywardell: When I plug in USB while in rockbox, nothing happens
12:32:40Shadows1990any idea how long itll be before a working copy is made?
12:32:42 Quit GodEater (Read error: 110 (Connection timed out))
12:32:43barrywardellamiconn: holding SELECT while connecting should reboot, continue holding to get "emergency disk" mode
12:33:01amiconnEh, why only when holding select?
12:33:11pixelmaShadows1990: no-one is actually working on it
12:33:14barrywardellI haven't found a way to automatically boot to bootloader disk mode
12:33:25barrywardellso you have to hold SELECT to get there
12:33:26amiconnI'd expect to hold a button when I do *not* want to go into diskmode, like on all other targets
12:33:34Shadows1990why not?
12:34:13barrywardellamiconn: on MTP H10's, the only way to get to disk mode is to hold SELECT while starting up with USB plugged in
12:35:02amiconnAha, hmm. Is it possible to distinguish MTP and UMS models from within rockbox?
12:35:50pixelmaShadows1990: (if I'm not mistaken). Well, ports are made by interested users/developers who own the player, so afaik only one dev has one and is also occupied by other things. I think the best info you could find is in the "new ports" forum thread, maybe - I don't know much more.
12:36:12barrywardellyou could possibly read the mi4 header from H10.mi4. 010301 = mtp, 010201 = ums
12:37:15Shadows1990ok, thanks for the info
12:37:56amiconnI also found descriptions how to convert an MTP H10 into UMS, even for H10 pure
12:38:33barrywardellbut not for 20GB?
12:38:34amiconn(for non-pure it's a choice in the OF's firmware updater afaik - never needed that because mine already was UMS)
12:38:57barrywardellthere never was a UMS version of the 20GB AFAIK
12:40:09amiconnHmm, didn't check for the big model
12:40:44amiconnConverting the H10 pure (which also comes as MTP verison only) involves tricking it into flashing the non-pure firmware
12:41:06amiconnThen you'll get a useless radio menu in the OF :)
12:41:42barrywardellpetur: here's a logfdump:
12:41:58barrywardellpetur: it looks like I'm still getting a timeout->port
12:42:16 Nick fxb__ is now known as fxb (
12:42:33barrywardellamiconn: unfortunately that won't work on my H10 because of different screen sizes, etc.
12:42:55amiconnYes, obviously.
12:44:32amiconnPlugging USB while holding Select works, btw, I don't need to continue holding it once rockbox shows the USB logo
12:46:46 Join Thundercloud [0] (
12:47:04barrywardellmaybe we should do something different for UMS models. the current setup makes the most sense for MTP only
12:47:33barrywardellunless austriancoder manages to get USB support into Rockbox soonish
12:47:33 Quit Shadows1990 ("CGI:IRC (EOF)")
12:55:13 Quit barrywardell ()
13:01:26 Quit Nico_P (Remote closed the connection)
13:12:41austriancoderhi all
13:14:51*petur goes to lunch - brb
13:17:27einhirnIs there some specific reason for the indirection when calling "sound_set_X" in sound.c?
13:20:53einhirnI find it obscure that there is a Function "sound_set", that looks up a "sound_set_X", which in turn just calls another function like set_channelconfig and the like...
13:21:26amiconnThere is a reason
13:21:30einhirnAh ;)
13:21:35einhirnThere had to be ;)
13:21:47amiconnsound.c is in firmware, and firmware isn't allowed to call app layer code, except via callbacks
13:22:11amiconnBut the software tone control for swcodec targets without hardware tone control does reside in apps/
13:23:00amiconnSo swcodec registers these functions on startup.
13:23:48 Quit pixelma (" bye.")
13:24:09einhirnOk, Can I use "global_configuration.X" in sound.c?
13:25:15einhirnHmm. Then I have to devise some other way of setting Replaygain _and_ Channel Config/Stereo Width at the same time.
13:26:06amiconnYes. Same thing as for volume/balance on almost all targets except MAS3587/MAS3539
13:28:27amiconnYou have one function per settable value, which only puts the changed value into a variable in sound.c, and then calls a common function that combines the values and applies them
13:29:49 Join davina_ [0] (
13:30:42einhirnYes. I've seen that.
13:31:29einhirnSo I need to add one setting for "Gain" - will I break something when I just define that?
13:34:37 Quit davina (Read error: 110 (Connection timed out))
13:36:30amiconnNo, but take care of the proper order. Array init will work even if you don't, but then it will have gaps, wasting (some) memory
13:37:08amiconnAnd I wouldn't just call it GAIN, in oder to avoid confusion with the recording gain values
13:37:30 Quit Rick ("I don't need to be here.")
13:37:30amiconnSOUND_PLAYBACK_GAIN sounds reasonable imho
13:40:34amiconnBtw, implementing replaygain using the digital volume matrix will also apply it to s/pdif out. While that might be a good thing to do, it also reduces the level there. And unlike for analog, there's no extra +6dB to enable
13:41:35 Join Rick [0] (
13:42:59 Join iwantanimac [0] (i=cb3b14ee@gateway/web/cgi-irc/
13:43:00peturaustriancoder: what's the meaning of arcotg_udc.c line 642-644 (patch of barrywardell applied)?
13:43:32austriancoderamiconn: the linkedlist code i use in the usb stack was taken from linux kernel. And they are using prefetch to speed up things. Should we also use prefetch, if taget supports it? on arm targets it seems that pld is not suppoprted :(
13:44:09*amiconn has no idea what austriancoder is talking about
13:44:22iwantanimachi all. Have there been any major changes to the WPS system in the past 6 or so months? I'm a little behind the times.
13:44:29einhirnamiconn: I already tried reducing the value in the digital volume control matrix and I can easily compensate the volume drop by cranking up the Amp. Actually I then use the same Volume range I use when Playing back something from my computer.
13:44:55amiconnThe amp won't help with s/pdif out
13:45:56*petur sees 6502 mentioned on the ML and gets some school flashbacks....
13:46:53 Join Buschel [0] (
13:47:28austriancoderpetur: looking if port reset finished
13:47:56austriancoderpetur: i tryed barrys patch.. but I am getting lines like
13:47:58austriancoderusb 2-6: new high speed USB device using ehci_hcd and address 38usb 2-6: device descriptor read/8, error -71usb 2-6: device descriptor read/8, error -71
13:48:09austriancoderon my linux machine
13:48:12peturyes, but why report timeout when the controller is no longer in reset?
13:48:34austriancoderpetur: its not my code.. its mrh's work
13:49:56iwantanimacAs well as my previous questions, did 'viewports' ever get implemented?
13:50:12scorcheno it hasnt
13:50:34iwantanimacwhat ended up holding it back?
13:50:44scorcheand the wiki page MajorChanges will be of use to you
13:51:11iwantanimacalright. thanks.
13:51:15 Quit iwantanimac ("CGI:IRC")
13:57:52einhirnamiconn: The Amp I was talking about was the one connected via S/PDIF
13:58:52amiconnAh, ok
13:59:55einhirnBut I can call sound_set(SOUND_PLAYBACK_GAIN,value) everywhere where dsp_set_replaygain() is called, right?
14:01:42***Saving seen data "./dancer.seen"
14:02:17einhirn(I'll assume that for now ;) )
14:04:08peturaustriancoder: according to we're beyond the TX problem (we are up to step 5), but it seems (from line 510 in the code) that we're not handling the set address request?
14:04:14amiconnI would think so, but if you do that, it probably won't work
14:04:40amiconnthe dsp_* functions are used on swcodec, not on hwcodec
14:04:53amiconnHwcodec has a completely different playback engine
14:05:14amiconnI also wonder how you will handle the decoder delay
14:05:49amiconnIt's not a constant delay, it varies with current bitrate
14:06:28austriancoderpetur: I will implement the setadress thing.. then we will see
14:06:35austriancoderJdGordon: ping
14:07:34peturdamn.. the h10 I was trying to get was sold this morning :(
14:07:36einhirnOh, so just setting a new gain when the track changes could result in applying the gain of the next track to the last n seconds of the previous track.
14:08:15amiconnNot exactly second_s_, unless you consider using very low bitrates for music
14:08:28austriancoderpetur: but we hang at step 4... host never gets device descriptor: usb 2-6: device descriptor read/8, error -71
14:08:41austriancoderneed to check what error code -71 means
14:08:57amiconnIt's just that the MAS tries to keep its buffer full at all times, causing a delay that's inveresely proportional to bitrate
14:09:56einhirnOk. I hope I can adress that problem later. First things first - I have to mangle global_configuration.replaygain* into one value thats passed to sound.c ;)
14:10:23einhirn(global_configuration and current track settings)
14:10:58peturaustriancoder: no, read the comment below the steps: after the host has the first 8 bytes, it issues a reset and sets the address, then it asks the descriptor again
14:11:23austriancoderpetur: okay.. I will try it
14:13:00*amiconn found a minor display bug on Player
14:14:29 Join bdgraue [0] (
14:14:47peturaustriancoder: I have to run off to some shops, back in 1-2 hours I hope...
14:15:01scorchethat describes what error -71 is
14:15:05austriancoderpetur: okay
14:15:26 Quit ouioui ("ChatZilla [Firefox]")
14:15:51scorchewell, it doesnt describe definitively, but it is more than a number =)
14:16:08austriancoderit could be anything *g*
14:16:15scorcheThere is a lot of advice if you google for the specific error (e.g., "device descriptor read/64, error -71") but many of these links will say to remove the high-speed USB module with rmmod, disable USB in BIOS, and other low-level, potentially serious fixes.
14:16:15scorcheI read in the Linux USB FAQ (see that high-speed usb devices are much more sensitive to signal quality, and therefore cable defects. I figured it wouldn't hurt to try swapping out the cable before doing anything crazy, and what do you know! It worked!
14:17:11JdGordonaustriancoder: pong
14:17:15peturscorche: I think my explanation above will be the cause...
14:17:36scorchepetur: likely, just providing more info...
14:18:23austriancoderJdGordon: to descripte what I want in the settings code, i have done a drawing ->
14:19:02JdGordonah yes... lemme finish this level in ghost rekon and ill get to work :p
14:19:10austriancoderJdGordon: I can give you a char* with all driver names "a,b,c" or I can give you the linked list
14:19:13scorchepriorities! =)
14:19:15austriancoderJdGordon: okay ;)
14:19:29JdGordona char* would be best.... but ill make it generic
14:27:01 Join Arathis [0] (
14:32:07 Quit bdgraue (Remote closed the connection)
14:33:36amiconn'make bin' doesn't rebuild rombox.ucl
14:33:54*amiconn stumbled on that, wondering why his fix didn't werk :(
14:34:42*petur fails to leave the house...
14:38:47 Join DerPapst [0] (
14:39:05DerPapstheh.. locked yourself? ;)
14:40:25einhirnSeems like there is no real way to avoid duplicating most of the replaygain handling code - otherwise I'd surey break the SWCODEC ;)
14:40:47peturDerPapst: no babysitting a 2 month old which seems to like me more than my wife and didn't like me leaving ;)
14:46:05amiconnBagder: around?
14:46:25amiconnThe build system didn't pick up my last commit...
14:48:23 Join aliask [0] (
14:51:34peturamiconn: Bagder implemented some AI to ignore commits from certain people :p
14:59:42DerPapstpetur: aha :)
15:01:13 Join My_Sic [0] (
15:07:26*petur finally runs off
15:15:31 Join sarixe [0] (
15:30:40 Quit amiconn (" HydraIRC -> <- *I* use it, so it must be good!")
15:31:56 Join smail [0] (i=52e0d0a5@gateway/web/cgi-irc/
15:32:22 Quit smail (Client Quit)
15:32:42 Join smail [0] (i=52e0d0a5@gateway/web/cgi-irc/
15:38:29parafintracker doesn't feel good in last time: Undefined index: 0 in /usr/share/flyspray/htdocs/includes/class.flyspray.php on line 344
15:39:11parafinfor example
15:42:16 Quit alienbiker99 ("( :: NoNameScript 4.02 :: )")
15:47:13 Quit smail ("CGI:IRC (Ping timeout)")
15:48:38 Join smail [0] (i=52e0d0a5@gateway/web/cgi-irc/
15:55:00 Quit smail ("CGI:IRC (EOF)")
15:56:11 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
15:56:24 Quit Buschel (Read error: 104 (Connection reset by peer))
15:57:11 Join smail [0] (
16:01:46***Saving seen data "./dancer.seen"
16:05:05peturparafin: it does that from time to time...
16:05:46 Join Robin0800 [0] (
16:06:12parafinwhat does charge during USB connection option do on ipod video?
16:09:35 Quit krazykit ("leaving")
16:09:44peturit's not in the manual?
16:10:04parafinno, as far as i can see
16:10:40petursorry, I'm not an ipod user...
16:10:53austriancoderpetur: sending does not work - as in barrys patch there is everything which is needed... setting device address handling all essential usb requests....
16:12:05peturaustriancoder: sending must be working because the host will not try to set the address unless it received the first eight bytes...
16:12:21austriancoderpetur: it will... when you do a stall
16:12:28n1sparafin: are you using an official build?
16:13:13austriancoderpetur: if sending works.. then barrys code should set device address and accecpt it.. but it doesnt
16:13:27parafinbut this option is in official build
16:14:05peturaustriancoder: best way to try is to comment out the code that tries to send. If we get the same packets from the host (first 6 then 5) then I'll believe it ;)
16:14:06n1sparafin: then it's unsupported, that option is not implemented for ipods in the official rockbox, I have no idea what it might do, it is however supposed to switch the usb charging to draw the maximum allowed current
16:15:03Phillpetur: i'm trying barrys patch on my e200, and logging under Windows shows that the first 8 bytes are _not_ received. logf shows packet 6 sent twice from the host, and no packet 5.
16:15:47peturstrange, barrys log showed both 6 and 5 being received
16:17:18Phillbarry is using a mac rather than windows though I think. That may explain the difference.
16:17:39*austriancoder is using linux
16:18:39*austriancoder is putting barrys code on his sansa and paste the log...
16:25:12parafini can't understand, what's the difference between artist and album artist entries in database browsing...
16:25:59 Join My_Sic [0] (
16:28:55 Join BobShield [0] (
16:31:35peturit has request '5' in there too...
16:32:44*austriancoder wishes petur had an ipod or sansa
16:33:18austriancoderpetur: with my borken sending code i get also two types of request... get descriptor and set address
16:33:18*petur is bidding on an h10 and keeps an eye on a sansa
16:34:03 Quit sarixe ("Peace")
16:34:35austriancoderpetur: I will continue to work on the storage driver... as i dont want more time one tx problem... I think mrh would be a great help
16:35:53 Quit My_Sic (Read error: 104 (Connection reset by peer))
16:36:09peturaustriancoder: I'd rather have working TX as some proof of concept than a whole stack that hasn't worked at all.
16:36:54austriancoderpetur: but I cant get it working by my own...
16:37:04austriancoderi need help
16:37:49 Quit idnar (Nick collision from services.)
16:37:51 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
16:37:58parafinahh, i guess i understood what is album artist =)
16:38:03 Join My_Sic [0] (
16:40:38parafinit seems that easytag doesn't support albumartist tag :/
16:42:01peturparafin: iirc, there is a subtle difference between the two, so a track on an album can have a different artist (like xyz feat. ggg)
16:44:20 Join rasher [0] (
16:45:18JdGordonaustriancoder: sorry, new comp forceing me to play games instead of code... ill do it tomorow
16:45:22*JdGordon gones
16:45:37 Quit JdGordon ("Konversation terminated!")
16:45:54*petur doesn't like gamers
16:47:20 Quit rasher_ (Read error: 145 (Connection timed out))
16:48:00n1sany opinions on always representing dB values as dB*10 internally and do the scaling when displaying/talking or implement some kind of formatting system for spoken values too (as is implemented for displayed values in settings screens)?
16:48:55peturthat would be cB then ;)
16:49:02n1sthe problem that I'm trying to solve is that some dB settings are spoken as dB*10 so 2.5dB becomes 25 etc
16:49:25n1spetur: yeah but I don't think people would be happy with that :-)
16:50:20peturif there are some spots that use dB and others that use dB*10, I'm all for making that uniform
16:50:33n1sfor eq, replaygain preamp etc it is represented in cB internally but for other things like volume, bass etc it's regular dB
16:53:24 Quit smail (Read error: 113 (No route to host))
16:57:26 Quit miepchen^schlaf ("Verlassend")
16:57:29 Quit aliask ("ChatZilla [Firefox]")
16:57:46 Join miepchen^schlaf [0] (
16:57:56dionoeaAnyone familiar with a "iPod's LCD stays blank even if you plug the USB cable in" ?
16:58:23 Quit rasher (Read error: 145 (Connection timed out))
17:01:45parafintry resetting it
17:02:09parafini had such problem
17:02:24dionoeahum, hardware reset fixed it
17:05:08 Join rasher [0] (
17:06:59*n1s curses the settings code... meh maybe JdGordon will be here tomorrow...
17:15:29ender`dionoea: which ipod?
17:16:03 Join bluey- [0] (
17:16:09dionoeavideo 30 GB (it's fixed now)
17:17:35 Nick Arathis is now known as Arathis|afk (
17:20:27 Join toffe82 [0] (
17:21:33 Join latchema [0] (
17:22:34einhirnn1s: there are even internal representations of dB * 100 - I just stumbled across them: get_replaygain_int(...) in firmware/replaygain.c expects that as a value...
17:23:35 Join rds [0] (n=rogelio@
17:23:39einhirnStill trying to figure out, what happens with the track/album gain values - they seem to be dB in 7.24 fixed point format.
17:28:06 Part latchema
17:32:38*austriancoder makes a break
17:32:42 Quit austriancoder ("Kopete 0.12.4 :")
17:34:22*petur takes it
17:38:51 Quit bluey- (Read error: 113 (No route to host))
17:39:38 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
17:40:55 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
17:55:07 Part kretender
17:57:12 Join perrikwp [0] (n=chatzill@
18:01:50***Saving seen data "./dancer.seen"
18:02:29n1shmm, I was mostly thinking of the values of various settings, but interesting still
18:06:28 Join bluey- [0] (
18:06:41 Join The-Compiler [0] (
18:06:47 Quit bluey- (Remote closed the connection)
18:06:52einhirnMaybe it's just me not being very adept with fixed point math, but according to my understanding it can't be right to simply multipy one 7.24 fixed point number with one 19.12 fixed point number...
18:07:06n1sbut as always when dealing with the settings/menu code It made my head hurt and I coun't do it in a clean way so I gave up...
18:07:39einhirnyes, my head hurts too at the moment...
18:07:53n1seinhirn: no that doesn't sound right...
18:09:45 Join datasleep [0] (
18:09:47 Quit datachild (Nick collision from services.)
18:11:05 Nick datasleep is now known as datachild (
18:16:42einhirnn1s: I'll google some and then hopefully wrap my head around it soon. But first:
18:16:52*einhirn has to take a break
18:17:02 Join lee-qid [0] (
18:17:08 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:19:23 Quit BigMac (Read error: 110 (Connection timed out))
18:24:34 Quit The-Compiler (Connection timed out)
18:25:15 Join Arathis_ [0] (
18:28:13 Join BigMac [0] (
18:29:33 Quit freqmod (Read error: 104 (Connection reset by peer))
18:30:12 Join smail [0] (
18:35:48 Quit midkay ("Leaving")
18:36:21 Join chrisjs169_ [0] (
18:40:40 Join chrisjs169__ [0] (
18:44:11 Quit Arathis|afk (Read error: 110 (Connection timed out))
18:44:32 Join thegeek_ [0] (
18:45:38peturreal life calling...
18:45:46 Quit petur ("plop")
18:51:06 Join Buschel [0] (
18:53:26 Quit chrisjs169_ (Read error: 110 (Connection timed out))
18:53:27 Quit smail (Read error: 113 (No route to host))
18:53:38 Quit chrisjs169 (Nick collision from services.)
18:53:43 Nick chrisjs169__ is now known as chrisjs169 (
18:56:47 Quit thegeek (Read error: 110 (Connection timed out))
18:59:26 Join kubiixaka [0] (
19:03:55 Join thundergirl [0] (n=francima@
19:04:14thundergirlhi people
19:04:20 Quit datachild (Remote closed the connection)
19:04:32thundergirlanybody speak spanish?
19:04:45 Join datachild [0] (
19:07:53 Part thundergirl
19:09:28 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
19:12:01 Join kubiixaka [0] (
19:13:21 Nick fxb is now known as fxb__ (
19:17:03 Quit iamben (Read error: 104 (Connection reset by peer))
19:33:14 Quit ctaylorr (Read error: 110 (Connection timed out))
19:35:29 Join bluey- [0] (
19:42:33 Part toffe82
19:45:41 Join GRaTT [0] (
19:46:16GRaTTping jdGordon
19:48:11GRaTTping chrisjs169
19:50:08 Quit Buschel ()
19:50:33GRaTTI have updated FS7538 custom splash screen. There is NO DELAY for non custom splash screen users.
19:53:22 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
19:53:54 Join toffe82 [0] (
20:01:51***Saving seen data "./dancer.seen"
20:04:19chrisjs169ping GRaTT
20:07:26 Join datasleep [0] (
20:07:27 Quit datachild (Nick collision from services.)
20:10:49 Join BigBambi [0] (n=Alex@
20:11:35 Quit rds (Client Quit)
20:12:36 Join My_Sic [0] (
20:15:02GRaTThi chrisjs169 just wanted to let you know about the update for fs7538
20:15:33GRaTTreally wanted to get some attention for it and get it committed
20:22:02 Part bluey- ("Leaving")
20:22:07 Nick datasleep is now known as datachild (
20:25:57chrisjs169I'm looking at digital cameras right now (out of nowhere I decided I need a new one) but I'll add it shortly
20:26:23DerPapstget a dslr :P
20:27:26DerPapstanything else is crap imo</off-topic><on-topic>
20:28:28chrisjs169<offtopic>I won't mind having one, but I dunno if I could spend 800 bucks on one</offtopic>
20:31:17 Quit Arathis_ ("Verlassend")
20:31:32DerPapstit's really worth it.
20:32:05 Quit My_Sic (Read error: 104 (Connection reset by peer))
20:33:52 Quit datachild ("- nbs-irc 2.35 - -")
20:34:20chrisjs169anything would be better than the one i have now
20:44:05parafinwhat's up with this bug: ? it's old but wasn't rejected or applied
20:48:40 Join The-Compiler [0] (
20:52:05 Quit Thundercloud (Read error: 110 (Connection timed out))
21:01:05 Join datachild [0] (
21:17:11 Join My_Sic [0] (
21:18:15 Quit ompaul ("later")
21:21:36chrisjs169GRaTT: updated 7538 in my build
21:22:17 Join Llorean [0] (
21:23:51 Quit BHSPitMonkey (Connection timed out)
21:28:31GRaTTcool. Have you had any feedback on it?
21:29:17GRaTTI mean on the custom splash screen in general.
21:38:19chrisjs169I haven't put it on my Sansa yet, I'm testing a basic build at the moment
21:42:53linuxstbGRaTT: One bad thing I can see with your splash patch is that it's declaring a static buffer the size of the screen - on the 320x240 colour targets, this is 150KB (not insignifant).
21:51:31 Quit Anakin (Read error: 110 (Connection timed out))
21:56:17 Join seablue [0] (
22:01:52***Saving seen data "./dancer.seen"
22:20:45 Join petur [0] (n=petur@rockbox/developer/petur)
22:21:47*petur will be the owner of an H10-5GB in the near future
22:22:12Bagderoh, welcome to the PP target world!
22:22:27peturit was about time, I fear ;)
22:23:15Bagderthat's your second target, right?
22:23:53Bagderunspoiled youth... :*)
22:25:03 Quit The-Compiler (Remote closed the connection)
22:27:23peturthe remaining work (apart from the common usb and power issues) are FM and recording, right?
22:27:35Bagderyes I think so
22:28:01 Quit desowin ("use linux")
22:28:06peturtoo bad it doesn't have a normal line-in connector
22:28:53 Join Thundercloud [0] (
22:31:10 Join bdgraue [0] (
22:35:04 Join TTSbox [0] (i=OzgurOzt@DNab42237f.Stanford.EDU)
22:35:23peturerrr... somebody has been adding java to rockbox?
22:35:34peturand dare to call it innovation...
22:35:37BagderI don't think so, merely talking about it
22:37:59peturhah, right now they're only talking at the 'zune' level :)
22:38:10peturlots of bla bla
22:38:54Bagderhaha, yeah
22:39:02Bagdermy guess is that they'll remain there
22:41:00peturzune-level it is....
22:50:03 Quit cmvjk (Read error: 110 (Connection timed out))
22:51:44 Join chadoh [0] (i=42f95557@gateway/web/cgi-irc/
22:53:11chadohdo you have rockbox on an ipod video?
22:54:21 Quit chadoh (Client Quit)
22:54:46 Join chadoh [0] (i=42f95557@gateway/web/cgi-irc/
22:56:32 Join bnakifnig [0] (i=0@
22:56:37peturchadoh: I suggest you look around on and look at the manual
22:57:20 Quit chadoh (Client Quit)
22:57:59 Join chadoh [0] (i=42f95557@gateway/web/cgi-irc/
22:59:03 Join crop [0] (i=3e865097@gateway/web/cgi-irc/
23:00:02 Join agm3nt [0] (
23:01:07 Join sarixe [0] (
23:04:03 Quit crop ("CGI:IRC (EOF)")
23:05:27 Join advcomp2019- [0] (n=advcomp2@
23:06:13 Quit advcomp2019 (Nick collision from services.)
23:06:17 Nick advcomp2019- is now known as advcomp2019 (n=advcomp2@
23:08:43 Quit chadoh ("CGI:IRC (EOF)")
23:10:49 Quit sarixe ("Peace")
23:11:28 Join Sivart0 [0] (
23:17:49 Join shiftplusone [0] (
23:18:50shiftplusoneHey, is it possible to send/receive stuff through the USB connector (on the ipod) yet?
23:20:38 Quit at0m|c ("clear channels event")
23:21:29 Quit bdgraue (Remote closed the connection)
23:21:33 Join jurrie_ [0] (
23:21:44peturshiftplusone: we're working on it
23:22:07shiftplusonepetur, any luck?
23:22:41peturpartial, sending seems to fail... austriancoder is the one to nag ;)
23:22:53shiftplusoneI am not talking about files.... just 0s and 1s....
23:23:10shiftplusoneso... at this stage we can receive data, just not send it?
23:24:25shiftplusoneI know this isn't the place to ask... but suppose it won't hurt... does ipodlinux have that support yet?
23:25:18shiftplusoneok, thanks.
23:27:23 Nick fxb__ is now known as fxb (
23:29:08 Quit seablue ("life, death, life, death")
23:29:41peturwe share lots of knowledge so if they had it working, we would have too
23:30:04peturit's an open source thing ;)
23:30:38shiftplusonedidn't they have working ipod to ipod file transfer and such?
23:31:01DerPapstyes we have.
23:31:07DerPapstonly firewire though
23:31:35 Quit jurrie (Read error: 110 (Connection timed out))
23:33:25shiftplusoneso... if I change my initial question..... would it be possible to to send/receive data via firewire through rockbox or is that support pointless or still needs to be coded?
23:35:18DerPapstafaik rockbox has no fiwi supprt yet. the pp5002 support is very young
23:36:55shiftplusoneI take the yet as it's on someone's to-do list.... so is that for the purpose of file transfer form within rockbox without having to reset?
23:38:04shiftplusone*from >.>
23:41:31 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
23:43:11 Quit miepchen^schlaf ("Verlassend")

Previous day | Next day