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 2009-12-28

00:02:22 Quit JdGordon1 (Read error: 110 (Connection timed out))
00:07:52 Quit merbanan (Read error: 60 (Operation timed out))
00:11:27***Saving seen data "./dancer.seen"
00:11:30 Quit einhirn (Read error: 60 (Operation timed out))
00:22:35 Join einhirn [0] (
00:40:52 Join n1s [0] (n=n1s@rockbox/developer/n1s)
00:56:39 Quit StealthyXIIGer (Read error: 110 (Connection timed out))
01:04:56 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
01:13:50Tornewhen are we branching for release?
01:19:20n1sgood question
01:20:10Tornedunno if the discussion ever came to any conclusion :)
01:20:16Tornei guess too many people are still doing xmas stuff
01:21:33 Quit pamaury ("exit(*(int *)0 / 0);")
01:22:27n1sIIRC we usually freeze 2 weeks before release and usually release on the 23rd but we agreed to postpone the release 14 days this time so we should have frozen on the 23rd and branched one week after that, so wednesday, but we aren't frozen afaik
01:23:26 Join bonjour [0] (
01:24:31 Quit einhirn (Read error: 104 (Connection reset by peer))
01:26:02bonjourI have a question about using rockbox stable on a 30 gig ipod..... is it possible to jump back to the now playing screen while browsing the other menus? the only way I can find is to go through the main menu.
01:26:21LloreanIn almost all menus the play button takes you to the WPS.
01:26:37LloreanThis is in the manual. Or are you talking about it not working as described?
01:27:56bonjournope I just couldnt find anything about it in the manual. Thanks, it's nice that everything I keep finding is missing really isn't at all.
01:28:34Lloreanbonjour: I'd recommend reading the manual fairly thoroughly. At the very least look at all the keymap tables.
01:30:52bonjourI did indeed read the manual, just missed that key fact. I feel like I understand just about everything else. thanks for the help
01:30:55 Quit DerPapst ("Leaving.")
01:30:56 Quit bonjour ("CGI:IRC")
01:31:56 Quit Llorean (Read error: 54 (Connection reset by peer))
01:32:24 Join Llorean [0] (
01:36:03 Quit bmbl ("Bye!")
01:36:56 Quit Omlet^away ("( :: NoNameScript 4.22 :: )")
01:57:06 Quit bertrik (Read error: 113 (No route to host))
02:04:15 Join Sajber^ [0] (
02:07:54 Quit Sajber^1 (Read error: 60 (Operation timed out))
02:09:41 Join JdGordon1 [0] (
02:11:31***Saving seen data "./dancer.seen"
02:21:57 Quit Phyber0ptik (Read error: 60 (Operation timed out))
02:28:12 Quit bluebrother (Nick collision from services.)
02:28:15 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
02:28:45jaeWhat's the endianess of the ARM CPU in the Sansa Fuze?
02:29:15Tornelittle endian
02:29:29Tornebig endian ARM is all but dead
02:29:32*jae is just tring to (finally) get the DB read intelligently, and robustly (taking the module from the Wiki as template)
02:29:39 Join Space [0] (
02:32:15SpaceCan I use rockbox on it?
02:32:53jaeBut coldfire and SH1 are big-endian? Which players use them, and... would I see "coldfire" in .rockbox/rockbox-info.txt?
02:34:03TorneSpace: supported players are listed on the front page. if it's not there, then no.
02:34:06JdGordon1isnt the db format endian-agnostic?
02:34:28 Join Reilithion [0] (
02:34:30Spacebut I read that with somethings sansa players don't work
02:34:36n1sjae: yeah, coldfire and sh1 are BE
02:34:37Spacebut with another yes, and
02:35:31JdGordon1Space: yes, but that looks way to expensive
02:35:35n1sJdGordon: i think the db supports both BE and LE db files
02:39:01 Quit n1s ("Lämnar")
02:42:33Spaceexist a list of mp3 players support by rockbox and his components? micro and RAM?
02:42:51Spacerecording. fm, etc
02:43:00jaeYeah, maybe it would, but... if it creates one, since I haven't seen any (obvious) signs of endianess conversion...
02:43:43jaeAnd: why should it support both? Since the DB is (usually, excepting weirdos like myself) only read and written on and by the player
02:44:35*jae is (still, or only now real seriously) looking for a way to read *and write* data to the db
02:47:44JdGordon1i think i remember it got some endian handinling in the code so the db file is always big (or little?) endian
02:50:01*jae slaps himself with a pretty big trout
02:50:22jaeWhy didn't I just search for "endian" in tagcache.c to begin with? :P
02:53:36SpaceJdGordon1, e200 v1 are supported and stable, but the v2 is unstable
02:53:42Spacethat say e250
02:54:09JdGordon1it could be either
03:03:54Spacewhat target have you? JdGordon1
03:07:37Spacesimilar cost
03:09:04 Join bzed__ [0] (
03:11:47 Quit GeekShado_ (Read error: 54 (Connection reset by peer))
03:11:59 Quit bzed__ (Read error: 104 (Connection reset by peer))
03:12:04 Join bzed___ [0] (
03:16:53 Quit bzed (Read error: 111 (Connection refused))
03:16:53 Nick bzed___ is now known as bzed (
03:20:21 Nick fxb is now known as fxb__ (
03:29:57 Quit JdGordon1 (Read error: 60 (Operation timed out))
03:34:59beta2kAnyone around familliar with the Phillips GoGear line?
03:36:46 Join FOAD_ [0] (
03:39:23 Join krazykit` [0] (
03:41:46 Join togetic [0] (n=togetic@unaffiliated/ibuffy)
03:42:22 Quit krazykit (Read error: 60 (Operation timed out))
03:45:04 Join krazykit [0] (
03:45:04 Quit krazykit` (Read error: 54 (Connection reset by peer))
03:46:58 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
03:50:47 Join krazykit` [0] (
03:52:54 Quit FOAD (Read error: 110 (Connection timed out))
03:52:55 Nick FOAD_ is now known as FOAD (
03:53:10 Quit krazykit (Read error: 60 (Operation timed out))
04:05:11 Join fdinel [0] (
04:08:10 Nick Ypsy is now known as YPSY (
04:10:13DogBoyI have that sansa player Space
04:10:51 Join krazykit [0] (n=kkit@
04:11:35***Saving seen data "./dancer.seen"
04:14:33 Quit krazykit` (Read error: 60 (Operation timed out))
04:31:56 Part froggyman
04:36:56 Quit r0b- (Read error: 110 (Connection timed out))
04:38:04 Quit kugel (Read error: 113 (No route to host))
04:38:55 Quit mc2739 (Read error: 60 (Operation timed out))
04:43:42 Quit Sajber^ (Read error: 104 (Connection reset by peer))
04:52:09 Join dys`` [0] (
05:03:57 Quit BHSPitLappy (Read error: 60 (Operation timed out))
05:07:39 Quit dys` (No route to host)
05:08:04 Join negru_voda [0] (n=negru_vo@
05:08:20negru_vodahi there...
05:08:42negru_vodaanybody have an ideea how to install rockbox on a 120Gb ipod ?
05:10:31krazykityou can't
05:10:54krazykitthe ipod classic is not a supported target
05:11:30 Quit efyx (Remote closed the connection)
05:11:39negru_vodadamn it ...
05:11:56negru_vodaand there's any way how to boot rockbox?
05:13:25krazykitnot at this time.
05:16:37negru_vodakrazykit could you tell me please... what i can do to add some movies/books to my ipod? i'm kind of new in this domain
05:18:15krazykiti don't know, as i don't own an ipod. i'm sure google has plenty of guides that can help
05:18:34negru_vodathank you :)
05:18:37negru_vodahave a nice day
05:18:40 Quit negru_voda ()
05:35:33 Join darkham [0] (n=darkham@
05:51:38 Quit darkham (Client Quit)
06:11:39***Saving seen data "./dancer.seen"
06:13:07beta2kkrazykit: Could have been a 5.5g with a 120gb disk in it :)
06:13:38krazykitbeta2k, i considered that, but figured if he's savvy enough to swap disks he's savvy enough to read the front page :)
06:13:50beta2kHEhehe :)
06:14:14DogBoyhow do you swap disks like that
06:14:15beta2kYou wouldn't happen to have a GoGear would you?
06:14:50beta2kDogBoy: Pull out disk A, insert disk B? :) There's a guide for different models on the RB wiki
06:15:16beta2kTrying to find someone who might have an idea about why these stupid things fail so often
06:15:22krazykitDogBoy, the newer ipods and gigabeat s use ZIF drives which are easy to swap
06:15:41DogBoyand there is a howto on the wiki?
06:15:42krazykitthe gigabeat f, and some of the irivers and cowons can too with a cheap adapter
06:15:48beta2kI dunno, I think the 50pin disks are easier to swap
06:16:09krazykiti have a 120gb drive in my gigabeaft with the 50pin->zif adapter
06:16:26beta2kand if you're going to do a swap in a gigabeat F you'll want a really small adaptor or a F40/60 back for it
06:16:41beta2kMine unfortunately doesn't fit :(
06:17:03krazykita cf card might, though
06:17:12beta2kkrazykit: You wouldn't happen to have a extra gigabeat you're will to part with? :)
06:17:22DogBoyI want one too
06:17:32krazykiti have a broken gigabeat s with a good lcd :)
06:17:33beta2kCF cards use a different pinout then 50pin ATA don't they?
06:17:43krazykitthey do, but i think there's an adapter
06:17:56beta2kkrazykit: I'd be more likly to go for it if it was functional with a bad LCD :)
06:18:04 Quit Strife89 ("Bed.")
06:18:04DogBoywhere would that be on the wiki
06:18:12DogBoyI'm searching for swap or
06:18:41DogBoyoh I see it
06:18:47DogBoyhehe, just found that
06:18:59beta2kI love my gigabeat f :) Kicks the dorry pants off the PP based players
06:19:16krazykitsee also
06:19:43DogBoyjeepers, crashed my fearfox
06:20:28beta2kDo you think rockbox is friendly enough to give to a mother in law that can't set their VCR time?
06:21:32 Quit BlakeJohnson86 (Read error: 60 (Operation timed out))
06:22:18krazykitmaybe. i can't set my vcr time either
06:22:54krazykitbeta2k, the CFModGuide wiki page may be of interest to you :)
06:26:17 Quit Space (Read error: 110 (Connection timed out))
06:26:38 Join BlakeJohnson86 [0] (
06:28:04 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
06:44:57 Join dmb [0] (n=Dmb@unaffiliated/dmb)
06:53:34 Quit dmb (Client Quit)
07:06:33 Join n17ikh [0] (
07:18:02 Join StealthyXIIGer [0] (n=stealthy@
07:27:53 Join Horschti [0] (n=Horscht2@xbmc/user/horscht)
07:31:20 Quit FlynDice (Remote closed the connection)
07:35:10 Join FlynDice [0] (
07:46:12 Quit Horscht (Read error: 110 (Connection timed out))
07:46:50 Quit FlynDice (Remote closed the connection)
07:48:23 Join FlynDice [0] (
07:48:29 Quit StealthyXIIGer (Read error: 60 (Operation timed out))
08:11:43***Saving seen data "./dancer.seen"
08:31:24 Join fyrestorm [0] (
08:38:42 Join FOAD_ [0] (
08:40:03 Quit FOAD (Read error: 60 (Operation timed out))
08:40:03 Nick FOAD_ is now known as FOAD (
08:49:05 Quit kaniini ("BRB! IRL!")
09:00:39 Join flydutch [0] (
09:06:02 Join kaniini [0] (
09:20:32 Join florinp3 [0] (
09:23:52 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
09:46:24 Join Grahack [0] (
10:03:14 Join n1s [0] (n=n1s@rockbox/developer/n1s)
10:05:20 Join robin0800 [0] (
10:06:58 Quit robin0800 (Remote closed the connection)
10:07:13 Join robin0800 [0] (
10:07:41 Quit robin0800 (Remote closed the connection)
10:09:56 Quit tha (Read error: 104 (Connection reset by peer))
10:10:00 Join tha [0] (
10:10:19 Join joe_ [0] (
10:11:44***Saving seen data "./dancer.seen"
10:13:19 Quit tha (Read error: 60 (Operation timed out))
10:13:56 Quit joe_ (Client Quit)
10:14:33 Join nima [0] (
10:29:08 Join bertrik [0] (
10:36:56 Join tha [0] (
10:43:22 Quit Res1 (Read error: 60 (Operation timed out))
10:44:57 Join Res1 [0] (
10:49:21 Join pamaury [0] (
10:58:24 Join DerPapst [0] (
10:59:33 Quit Reilithion (Read error: 104 (Connection reset by peer))
11:02:05 Quit Grahack ("Tu m'as vu ?")
11:02:50 Join stoffel [0] (
11:33:47 Quit n1s (Read error: 60 (Operation timed out))
11:54:21 Join n1s [0] (n=n1s@rockbox/developer/n1s)
12:00:01 Quit stoffel (Read error: 60 (Operation timed out))
12:11:16 Quit pamaury (Read error: 60 (Operation timed out))
12:11:45***Saving seen data "./dancer.seen"
12:17:05 Join Sajber^ [0] (
12:32:34 Join einhirn [0] (
12:39:28 Join stooo [0] (
12:40:39 Part stooo
12:45:53 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
12:50:31 Quit einhirn (Read error: 54 (Connection reset by peer))
12:55:43 Join einhirn [0] (
12:59:30 Quit einhirn (Read error: 54 (Connection reset by peer))
12:59:44 Join einhirn [0] (
13:03:18 Join mc2739 [0] (n=mc2739@rockbox/developer/mc2739)
13:11:09 Join kugel [0] (n=kugel@rockbox/developer/kugel)
13:12:35 Quit einhirn (Read error: 54 (Connection reset by peer))
13:12:47 Join einhirn [0] (
13:14:58 Quit einhirn (Read error: 104 (Connection reset by peer))
13:15:15 Join einhirn [0] (
13:17:13 Quit einhirn (Read error: 54 (Connection reset by peer))
13:17:58 Join einhirn [0] (
13:18:53 Join stooo [0] (
13:19:36 Quit einhirn (Client Quit)
13:19:49 Join einhirn [0] (
13:29:59 Quit TopyMobile (Read error: 110 (Connection timed out))
13:30:33 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:36:06 Join Phyber0ptik [0] (
13:36:11 Join TopyMobile [0] (
13:42:02 Quit FlynDice (Remote closed the connection)
13:42:22 Join FlynDice [0] (
13:45:59 Join Omlet^away [0] (
13:52:45 Join PaulJam [0] (
13:52:46 Quit einhirn (Read error: 54 (Connection reset by peer))
13:55:14 Quit TopyMobile (Operation timed out)
14:00:16 Join Grahack [0] (
14:02:31 Join TopyMobile [0] (
14:04:30 Quit Omlet^away ("( :: NoNameScript 4.22 :: )")
14:07:04kugelthe samsung yh-j70 arrived
14:07:56 Join efyx [0] (
14:11:46***Saving seen data "./dancer.seen"
14:16:45 Join Omlet^away [0] (
14:16:54 Quit TopyMobile ("Ex-Chat")
14:23:34 Quit Grahack (Remote closed the connection)
14:23:53 Join einhirn [0] (
14:26:16kugelwhat happens if you put a bad bootloader/firmware onto an X5?
14:26:51kugelamiconn: ^
14:27:12amiconnDepends on what is bad about it
14:27:36amiconnIf it isn't recognised as a proper firmware file by the built-in loader, nothing will happen
14:27:48amiconnIf it is recognised, it will be flashed
14:28:08amiconnBut since the built-in loader is never overwritten, it is always recoverable
14:28:30kugelso there is some kind of bootloader usb?
14:29:18kugelthe j70 seems to have a similar mechanism, a .bin file is placed on the root and then flashed
14:29:48kugelthe yh925 does have a bootloader usb too so you can basically throw everything at it
14:30:28n1skugel: what's inside this j70 player?
14:30:49amiconnYes, the built-in loader handles charging and usb
14:31:21amiconnIf it finds a firmware file, it will flash it (iirc only if the charger is connected)
14:31:53kugelthe "system loading" screen is definitely way shorter if i turn it on via usb
14:33:07kugeln1s: well, I hope that page isn't lying
14:34:32 Join Grahack [0] (
14:34:37n1sso another cf target :)
14:37:08n1sis it the same chip as an x5? if so that page is lying :)
14:37:49 Join pamaury [0] (
14:39:05 Join fdinel [0] (
14:39:19kugelwell, pyro_maniac doesn't have a j70, I assume he gathered that from the interweb
14:39:43n1sthe chip in x5 is called mcf5250 (or mcf5250)
14:40:11kugelmaybe I find a way to open that thing so I can investigate myself
14:40:24n1svery similar to the 5249 used in irivers and m5
14:42:50kugelI'm not sure that cf5050 even exists :(
14:43:39 Join einhirn_ [0] (
14:44:19 Quit einhirn_ (Client Quit)
14:46:09kugelthere's no screws anywhere :(
14:46:36 Join markun_ [0] (i=54f18043@rockbox/developer/markun)
14:48:55 Join einhirn_ [0] (
14:49:44 Quit faemir ("Leaving")
14:51:22 Quit einhirn_ (Client Quit)
14:53:32 Join einhirn_ [0] (
14:53:49 Quit einhirn_ (Client Quit)
14:54:13 Join einhirn_ [0] (
15:00:20 Quit Phyber0ptik (Read error: 60 (Operation timed out))
15:01:21amiconnn1s: The M5 has a 5250 like the X5
15:01:27amiconnThe M3 has a 5249
15:02:37kugelI managed to open it
15:18:37 Join froggyman [0] (
15:25:32 Join JdGordon1 [0] (
15:27:16 Quit einhirn (Read error: 60 (Operation timed out))
15:34:04 Join blithe [0] (
15:34:53n1samiconn: ah
15:36:16 Quit n1s ("Lämnar")
15:39:31 Quit einhirn_ (Read error: 110 (Connection timed out))
15:39:54 Quit JdGordon1 (Read error: 60 (Operation timed out))
15:43:35 Join pyro_maniac [0] (
15:45:07pyro_maniacif i may interuppt, the infos ive put into the wiki are from samsung repair manual
15:45:49pyro_maniacit included also a diassemble howto
15:48:41kugelcan you link to it?
15:49:07kugelpyro_maniac: ^
15:49:41pyro_maniaci will try to PM it to you tomorrow
15:50:27pyro_maniacthere are also pcb images in but i didnt published them because of copy right
15:51:00pyro_maniacyou should revert some changes of the wiki
15:51:36pyro_maniacJ50 and J70 have a different first letter on the chip
15:52:38kugelfeel free to correct it again. I got a j70 20gb today and updated what I could find on the board
15:53:07kugelI can also upload some pictures later
15:53:39pyro_maniacthat should be more secure then the pics of the manual
15:54:15 Quit fyrestorm (Read error: 54 (Connection reset by peer))
15:57:04kugelis that manual confidential?
15:58:47 Join fyrestorm [0] (
15:59:25pyro_maniacits branded
15:59:49pyro_maniaci can not remember completely
16:01:03pyro_maniaci found the link again nichlas sent me:
16:01:52 Join faemir [0] (
16:03:13 Join webguest [0] (
16:04:15 Quit webguest (Client Quit)
16:11:48***Saving seen data "./dancer.seen"
16:14:43 Join b1uebrother [0] (n=bluebrot@
16:14:57kugelpyro_maniac: the j50 seems a bit different at some places
16:16:05pyro_maniacyes i mentioned that but i thougth they are similar enough for one wiki page
16:18:04 Quit killan_ ("( :: NoNameScript 4.22 :: )")
16:20:26 Join Space [0] (
16:23:58kugelheh, it has a hardware piezo
16:27:41Spacethe thumbwheel of sansa players is similar to the clickwheel of ipods?
16:28:14kugelwell, it's a physical wheel
16:30:39 Quit b1uebrother ("leaving")
16:30:52 Quit DerPapst ("Leaving.")
16:35:24 Join arohtar [0] (
16:35:29 Nick Omlet^away is now known as Omlet^RPM (
16:37:55 Quit faemir (Read error: 60 (Operation timed out))
16:38:03 Join klapaucjusz [0] (
16:39:37 Quit arohtar (Client Quit)
16:39:49 Join faemir [0] (
16:40:42Spacekugel, what is the difference?
16:41:10Tornesimilar in what sense?
16:41:13SpaceI see that somes models only have a thin circle, other more...
16:41:21Tornethe way it's controlled in software? the way it feels to use?
16:41:39Spacefeels to use
16:41:48Tornewell it's a physical wheel. it turns when you turn it
16:42:02Torneso it feels completely different to an ipod, except the very old ipods which also had an actual turning wheel
16:42:11 Nick Riku is now known as Lss (
16:42:53Spacerockbox never couldn't use the WiFi conection of a music player?
16:43:01Torneit could if someone wrote the code for it
16:43:08kugelisn't the ipod "wheel" just a ring of touch sensors?
16:43:38Tornewell, except the 1st gen i think?
16:44:11TorneSpace: they would have to not only write a driver for the player's wifi interface, but also port a network stack to rockbox (lwip or similar) and implement something at the application layer to actually use the connection for something
16:44:25TorneSpace: it would be a lot of work and few players have such hardware, so nobody has bothered.
16:45:03Spaceclickwheel was in ipods since 4th generation
16:45:06Spaceof ipods
16:45:16Spacehas been
16:45:33 Nick fxb__ is now known as fxb (
16:46:14SpaceTorne, with the new generations of players has problems because the firmware are encrypted?
16:46:37Spacein my country only can find two class of sansa players
16:47:09Spaceone is more expensive his cost, and the other (sansa connect) is fine, but I think that I can't use rockbox on it
16:47:14TorneSpace: depends which players you are talking about..
16:47:37SpaceSandisk Sansa Connect
16:47:44TorneI know nothing about that player
16:48:12Tornebut in general yes, there are a reasonable number of DAPs now which have encrypted firmwares which make it difficult/impossible to run rockbox
16:49:02SpaceI didn't see anything of that player on the rockbox's web
16:49:46klapaucjuszHow does one get a patch merged into Rockbox?
16:50:19klapaucjuszI've been trying to get FS #10832 committed for over a month now, with no success.
16:53:06Tornewhen you say trying, what have you done?
16:53:11*klapaucjusz didn't want to kill the conversation.
16:53:21Tornedo people have objections, or is just nobody paying attention? :)
16:53:41 Join killan [0] (
16:54:04klapaucjuszTorne: submitted patch, irced around, sent e-mail to rockbox-dev, had a lengthy discussion on this channel.
16:54:15Tornewhat was the result of the discussion? :)
16:54:27klapaucjuszSaratoga expressed interest, he said he agrees with the patch, but he wants the opinion of somebody who understands the mp4 code.
16:54:38Torneaha, i see
16:54:47klapaucjusz...and of course he was unable to point me at somebody suitable.
16:55:04klapaucjuszFolks −− I'm seriously thinking of giving up on contributing to rockbox.
16:55:22klapaucjuszIf it takes so much effort to get a simple fix committed, I cannot imagine how one actually gets a new feature in.
16:55:44 Part pyro_maniac (" - Chat comfortably. Anywhere.")
16:55:52Tornestuff goes in when someone with commit access cares sufficiently. this is not always prompt, no :)
16:56:13Tornewhile i'm looking at it, does this do anything for super-long AAC files?
16:56:18Torneones that don't play even if they are streamable?
16:56:38klapaucjuszNo. What's the issue?
16:56:50Tornethe codec expects some index or table to fit in a fixed size buffer
16:57:08Torneso 20 hour audio books or whatever don't load, even if their chunks are arranged in the best possible way :)
16:57:18Tornei was just curious, because that is a problem i do have ;)
16:57:30klapaucjuszNot difficult to fix, probably.
16:57:47klapaucjuszBut I'll be damned if I make another patch against rockbox before FS #10832 is committed.
16:58:08Tornewell, you have to be somewhat accepting that this is a volunteer effort..
16:58:12Tornei understand your frustration
16:58:29Tornei had quite a lot of stuff i wanted to see committed. but then they tricked me into becoming a committer, so it's mostly done ;)
16:58:54klapaucjuszI'm regularly contributing to a bunch of volunteer projects.
16:59:08Torneperhaps we are lazier than the others
16:59:10Tornewho can say :)
16:59:12klapaucjuszNowhere have I had this kind of problems, except GNU Emacs (where I had to go through some silly copyright assignments).
16:59:33Tornealso, the nature of projects is not all the same
16:59:36Tornerockbox is a very diverse project
16:59:54Tornei know absolutely nothing about codecs, very little about the application layer in general, and nothing about most of the targets either :)
17:00:05Tornei just fiddle with ipod stuff, and occasinoally plugins
17:00:31Tornebut yes, it looks like it would be good to get that committed
17:00:37Tornedoes it ned all three of the posted patches?
17:01:09Torneyou might find it easier if you repost it as a single patch, synced to current svn
17:01:32Tornealso we are in theory preparing for a release at the moment
17:01:46klapaucjuszPlease define "easier".
17:01:55Tornewell i mean other people might find it easier to review
17:02:02Torneso it might also be easier once the 3.5 release branch has been created and there's less concern about breaking an important feature like aac playback
17:02:44*Torne has a look to see if there's any likely people to pester about this.
17:02:55klapaucjuszThe patches apply cleanly to the current tree (nobody has touched the MP4 and AAC stuff in a long time).
17:03:08amiconnTorne: You're right - only the 1st Gen iPod has a physical wheel
17:03:19Tornesure, but the way they are numbered isn't entirely clear that you need all three
17:03:29Tornesince they are numbers 1, 2, and 1 again :)
17:03:35*Torne shrugs. it doesn't matter ;)
17:03:39klapaucjusz1 and 2 are the minimum.
17:03:56klapaucjuszThe last patch deals with a different kind of not-quite-well-formed file.
17:04:07 Quit markun_ ()
17:04:24amiconnAll later generation have touch-based wheels; the 2nd and 3rd Gen and mini 1st Gen wheels only output the same signals as a physical wheel does
17:04:56 Quit Grahack ("Tu m'as vu ?")
17:05:23Torneklapaucjusz: anyway, i don't really know codecs, or aac specifically. but if i have the time i will take a look later..
17:05:47Torneand if you really can't find anyone else then once the release branch is created i will consider committing it for you :)
17:05:59Torneif nobody has *objected*..
17:06:24Tornethere does come a point with some things where the easiest way to test it is to stick it in trunk and wait for people to complain ;)
17:12:09 Join dys``` [0] (
17:14:53 Quit froggyman (Read error: 104 (Connection reset by peer))
17:15:26 Join froggyman [0] (
17:15:44Torneis there a convenient source of unstreamable aac files?
17:15:50Tornei don't have any, i suspect
17:16:16Tornei've only ever not been able to play files that are many hours long, which is a different problem as I mentioned ;)
17:21:17Tornecd ~/big/rockbox/ipod
17:21:30Tornetoo many terminals
17:23:38klapaucjuszTorne: grab a "HQ" file on Youtube, and pass it through "mplayer -dumpaudio".
17:23:47Torneok, cool
17:24:01 Quit dys`` (No route to host)
17:24:21 Quit togetic (Read error: 110 (Connection timed out))
17:38:49 Quit FlynDice (Remote closed the connection)
17:39:02 Join Kitr88 [0] (
17:41:53 Join FlynDice [0] (
17:42:42 Quit FlynDice (Remote closed the connection)
17:43:28 Join FlynDice [0] (
17:44:33 Join togetic [0] (n=togetic@unaffiliated/ibuffy)
17:50:01 Join StealthyXIIGer [0] (n=stealthy@
17:50:41 Quit Sajber^ (Read error: 60 (Operation timed out))
17:52:00 Join Sajber^ [0] (
17:53:56 Join hebz0rl [0] (
17:57:33 Quit Kitar|st (Connection timed out)
17:57:33 Quit Kitr88 (Connection timed out)
17:57:49 Join Kitar|st [0] (
18:01:29 Join Phyber0ptik [0] (
18:01:47 Quit florinp3 ("CGI:IRC")
18:01:51 Quit togetic (Success)
18:11:51***Saving seen data "./dancer.seen"
18:14:06 Nick dys``` is now known as dys (
18:14:07 Join togetic [0] (n=togetic@unaffiliated/ibuffy)
18:16:13 Join kramer3d [0] (n=sai@unaffiliated/kramer3d)
18:16:38kramer3dwho are the people working on rockbox S series port?
18:19:36bertrikWhat is "S series port"? Can you specify the exact model number?
18:20:50kramer3dToshiba Gigabeat MES30VW
18:21:06kramer3dmy fav mp3 player :)
18:21:07froggymankramer3d: i think lambdacalculus has done work for it
18:21:12kramer3dthanks a lot
18:21:21beta2kNeah, F's are my fav :)
18:21:24kramer3ddoes he come here often?
18:22:11kramer3dlambdacalculus i mean
18:24:04froggymani dont know, if you needed to contact him, your best bet would be to send him a PM on the forums
18:24:13Tornewhat is the problem/question/etc?
18:29:40 Quit Phyber0ptik (Read error: 110 (Connection timed out))
18:29:42 Join instinct- [0] (i=5439ce63@gateway/web/freenode/x-dmhliomrtkcifedk)
18:29:53kramer3di just want to ask him about the current status
18:29:58kramer3dof the development
18:30:08Tornein what sense? the status is on the front page :)
18:30:21TorneAnd a lot more people than just him will know about it..
18:30:29TorneI've worked with the beast some
18:30:46instinct-hi there
18:31:19instinct-just wanted to say thanks to whomever contributed to the sansa e200v2 builds. Finally installed it today and am a very happy camper
18:39:35Unhelpfulamiconn: ldr on arm7tdmi is 3 cycles, right, regardless of whether the result is used immediately after? if it takes longer than that to fetch the data from memory, what happens? is the CPU stalled, or can non-dependent instructions be processed during that delay?
18:40:36Torneit stalls, instructions can't be run out of order at all
18:44:28 Join Hillshum [0] (
18:45:58 Nick YPSY is now known as Ypsy (
18:47:48UnhelpfulTorne: i'm not expecting it to reorder instructions... but it stalls even if the next instruction does not depend on the result?
18:48:27Tornesure, why wouldnt it?
18:51:45Unhelpfulthat's what i was afraid of. i'll need to rethink bs_generic for such cases, i'm pretty sure i'm seeing a stall loading the address of the LUT it uses (which is in iram)
18:52:18Torneretiring them out of order is still reordering
18:52:43Tornein fact that's harder for most cores.. lots of OoO cores still only retire in order
18:54:15 Join toffe82 [0] (
18:55:03 Join stoffel [0] (
18:55:33 Join funman [0] (n=fun@rockbox/developer/funman)
18:55:37 Quit pamaury ("exit(*(int *)0 / 0);")
18:57:02 Join dmb [0] (n=Dmb@unaffiliated/dmb)
18:59:07Unhelpfulshame there's not an adrll or something like it that would use 3 or 4 adds to construct an address. it would still be faster than what's happening with my problem code.
19:01:13 Quit hebz0rl ("Ex-Chat")
19:01:31Torneyou can just do it yourself, no?
19:01:37Torneadr isn't an instruction anyway
19:05:10UnhelpfulTorne: it's a pseudo-op that's converted to add <reg>, pc, #imm, i know. i believe adrl emits *two* adds. the issue is that i can't construct the address explicitly since the address isn't known until link time, and adrl won't work as it's in another section, and definitely too far away to construct with two adds, anyway. i had hoped that perhaps coding the function in asm myself and putting the table load at the beginning might help... it
19:05:10Unhelpful arm11 but arm7 is, as usual, less nice. :)
19:05:45Torneah right, yeah
19:05:53Tornethe linker doesn't have a suitable relocation type for that
19:06:26Unhelpfuland it doesn't have one presumably because no address-construction pseudo-op is specified for that. :)
19:06:46 Quit stoffel (Remote closed the connection)
19:07:13Unhelpfulhrm, i wonder if mov reg, #<symbol>&0xff would work? and then of course some adds to fill in the other bits...
19:07:24funman < do we plan a freeze ? the mail proposed to freeze on the 23rd
19:07:27Tornei am prettu sure it won't
19:07:36Torneit will whine that it doesn't know the symbol value
19:07:47Tornebut having said that there are over 110 relocation types for ARM
19:07:50Torneso who the hell knows
19:07:57Tornemaybe there's one for that :)
19:08:48Unhelpfulthat's insane :)
19:09:15Tornethat's one of the many exciting reasons why if we wanted properly-loaded plugins we wouldn't want to load them as ELF :)
19:09:57Unhelpfulhrm. asm("mov %0, #table&0xff" : "=r" (var)) appears to work, actually.
19:10:19TorneDo arbitrary masks work? :)
19:10:29Torneif so i'd be curious to see what the object file contains
19:11:48Unhelpfulnot sure how to check :)
19:13:38 Quit flydutch ("/* empty */")
19:13:58Tornearm-elf-objdump -d foo.o and look for the R_ARM_whatever for the location in question
19:17:42Unhelpful...embarassingly i was editing the wrong file. it doesn't compile. :/
19:18:47Unhelpfulthis does explain, though, why ffs-arm.S uses table-follows-code rather than putting the table in iram (which the coldfire version does)
19:19:29 Quit Zarggg (Read error: 104 (Connection reset by peer))
19:19:33 Join Zarggg_ [0] (
19:24:01kugelUnhelpful: I think all-non PP ARMs have iram that isn't much faster than normal ram anyway
19:24:41 Quit Zarggg_ ()
19:25:29 Join Strife89 [0] (
19:26:26Unhelpfulkugel: i'm testing on e200... which is where i first noticed that an inline function with about 33% fewer ops was actually slower than saving registers and calling __builtin_clz.
19:26:39amiconnThe difference is that in the PP chips (and coldfires) IRAM is true SRAM, whereas some of the newer SoCs only have internal DRAM that isn't much faster than ordinary DRAM
19:27:59Tornewell, you just assume the cache will take care of it, no?
19:28:16Tornethe internal ram is more so you have some way to boot conveniently when the sdram controller isn't initialised yet, rather than a performance thing
19:28:18Unhelpfulprobably the only way i could make it faster, if loading the table address is going to cause a stall, is to inline the table as well, with something like asm(<code> "bx 1f\n\t" "<table>>" "1:")
19:28:30Torneand some way for the crypto processor to work without trusting external ram, if it does that kinda thing
19:28:40*kugel wonders how just putting the X5 bootloader on the yh-j70 would turn out
19:28:53Unhelpfulkugel: do it! ;)
19:28:58kugelAFAIK someone did that on the yh925 and noticed it works
19:29:16kugelexcept he took the h10 binary
19:30:23 Join Zarggg [0] (
19:30:29kugelthat service manual says that you can fix a non-working firmware by upgrading to another version, which suggests that the flash content is recoverable by the bootloader
19:30:30amiconnTorne: The cache needs to load stuff from DRAM the first time it is used. That might be unacceptable
19:31:02amiconnAlso, controlling larger caches is certainly more complex than just having another RAM area that happens to be as fast as a cache
19:31:03Tornewell yah. but system designs tend not to care about that kind of thing these days :(
19:31:36amiconnThere must be a reason why those older SoCs do it that way - they have both SRAM as IRAM and a cache
19:32:09amiconn(coldfire only has an instruction cahce - not a data cache - which points towards the complexity theory)
19:32:25kugelmaybe because the processors were slower than nowadays
19:33:46amiconnEven the slowest processor in our targets (SH7034) has SRAM as IRAM, but no cache at all...
19:34:37 Quit funman ("free(random());")
19:34:51kugelI assume faster processors with built-in cache are cheaper than extra SRAM, and still faster
19:35:02amiconnIt can do a nice trick with this IRAM which I think wouldn't be possible with cache - it can access IRAM and external memory *in parallel*
19:35:17amiconnThe caches essentially *are* SRAM
19:35:34amiconnBut with significantly more complex control logic
19:36:41Tornei think if people really care about that for special cases they tend to lock cache down as TCM
19:36:51Tornerather than explicitly designing systems which have seperate SRAM to function as TCM
19:37:21Torneand i think it's a catch-22 to some degree as well: standard OSes don't make good use of TCM and only benefit from increased caches
19:37:21amiconnCache line locking needs even more logic
19:37:38Tornekugel: tightly coupled memory
19:37:47amiconnPP can't do that afaik
19:37:52TorneNo, it can't
19:37:57Tornewell, not as far as we know
19:38:00Tornebut standard ARM caches can
19:38:14amiconnNot that it would make sense - 8KB cache but 96KB resp. 128KB IRAM
19:38:54TorneI think it's really just because OSes don't tend to use it
19:39:04Torneso system designers use the area for something else
19:39:11Tornebigger caches are a win for almost everyone
19:39:27 Quit amiconn (Nick collision from services.)
19:39:30 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
19:39:46Torneproviding a small fast internal ram is not as generically applicable, even if it's cheaper
19:39:49 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
19:39:50kugeleven we need to optimize for iram on a target base if it's bigger than the usual 128K
19:39:54 Quit pixelma (Nick collision from services.)
19:39:56 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
19:39:56Unhelpfulunless your load is from an address that *still* manages to be forced out of cache repeatedly :/
19:40:08TorneUnhelpful: well yes, there are always bad cass :)
19:40:11 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
19:40:18kugelwhile caches only need a few calls here and there which are the same for all targets
19:41:27Unhelpfulshame there's no way that, in specifying our own CLZ function, we could, say, indicate that it doesn't need any register saves (aside from the ones used for argument/result passing)
19:43:06kugelaren't there some attributes for that? naked maybe?
19:43:20Unhelpfulkugel: no, naked just says "don't generate prolog/epilog"
19:44:11Unhelpfulthe only way i can think of is to provide an inline that branches to the function and explicitly states lr is clobbered, and input/output is r0.
19:44:12kugelisn't it inline'd? the caller saves regs then?
19:45:11Unhelpfulkugel: the whole problem here is that inlining clz is losing because of the cost of loading the table address. it's cheaper to *call* gcc's clz function because it can construct its table address.
19:45:44Unhelpfulit would be cheaper still if the call could be specified as clobbering only lr :)
19:46:01Unhelpfuli don't think branching out of inline asm is supposed to be valid, though
19:47:04kugelI think 4.5 allows to branch to C labels
19:48:01Unhelpful"As of GCC version 4.5, asm goto may be used to have the assembly jump to one or more C labels. In this form, a fifth section after the clobber list contains a list of all C labels to which the assembly may jump. Each label operand is implicitly self-named. The asm is also assumed to fall through to the next statement."
19:48:11TorneYeah, you are wanting to be flexible about the calling c onvention there
19:48:16Tornewhich is indeed a useful notion
19:48:16 Join tomers [0] (
19:48:32Tornebut it's very hard to do that in a sensible way in the toolchain in general
19:50:12Tornei tried to write an assembler which can globally allocate registers without requiring a calling convention, and the result was nto pretty
19:50:31Torneand it's much easier to do for asm than for C :)
19:51:22UnhelpfulTorne: the compiler is responsible for generating code. if an attribute is provided indicating a specific clobber or i/o register list, and is used to declare the function, that should suffice
19:52:22Tornewell, yes. but that's not a very generally applicable solution
19:52:35Torneit works for the specific case of calling quick asm functions
19:52:42Unhelpfulinvalid or not there's a chance that cheating it with asm might *work*... provided the compiler really does assume that execution will arrive after your asm block with results in the specified registers and clobbers as specified, etc.
19:52:43Tornei was trying for global optimisation :)
19:52:54Torneand yah, it's worth a go maybe
19:56:05 Quit tomers (Read error: 60 (Operation timed out))
19:57:17 Join tomers [0] (
19:58:10tomersI want to commit "FS #10879 - Cabbie: Fix usage of spaces to position strings" - it looks straight forward to me... Any objections
20:03:31kugelto me too
20:05:41tomerskugel: I will commit in about half an hour if no one says anything else...
20:11:54***Saving seen data "./dancer.seen"
20:14:00 Quit FlynDice (Remote closed the connection)
20:15:37 Join Bagder_ [0] (
20:16:26 Join FlynDice [0] (
20:17:14 Join captain-kewl [0] (i=2669ecc2@gateway/web/freenode/x-tpjdstthzunqsgcs)
20:21:42 Nick fxb is now known as fxb__ (
20:28:51 Join AaronM [0] (
20:49:46 Quit StealthyXIIGer (Read error: 110 (Connection timed out))
20:53:41 Join merbanan [0] (
20:55:03 Join einhirn [0] (
21:03:05 Nick Horschti is now known as Horscht (n=Horscht2@xbmc/user/horscht)
21:04:50 Quit kramer3d ("Leaving.")
21:07:01UnhelpfulTorne: the more i think of it the more i think that *would* work. i would think the restriction on jumping out of asm only applies when it changes flow of control in the surrounding C code - because the compiler wouldn't know which code is potentially live in which cases.
21:07:16Tornewell do it then :)
21:08:33Unhelpfulso essentially i need to write an __clz with the implementation and a static inline clz that does a bl and explicitly puts the input and output in r0... and the asm needs to clobber lr obviously.
21:10:29CIA-6New commit by tomers (r24120): FS #10879 - Cabbie: Fix usage of spaces to position strings
21:10:29 Join petur [50] (n=petur@rockbox/developer/petur)
21:12:50 Join stoffel [0] (
21:13:15 Quit killan (Read error: 54 (Connection reset by peer))
21:13:17 Join killan_ [0] (
21:18:12 Join Xqtftqx [0] (
21:20:06 Join Grahack [0] (
21:41:20 Quit PaulJam (Read error: 113 (No route to host))
21:45:22 Join Spamhost [0] (
21:46:25 Quit stoffel (Remote closed the connection)
21:46:52 Quit tomers (Read error: 113 (No route to host))
21:51:02 Quit Xqtftqx (Remote closed the connection)
21:56:55 Quit Space (Read error: 104 (Connection reset by peer))
21:59:40 Quit togetic (Read error: 110 (Connection timed out))
22:00:03 Join TopyMobile [0] (
22:06:44 Quit petur ("Zzzzzz")
22:07:44 Join slam [0] (
22:08:10 Nick slam is now known as Guest38781 (
22:08:57 Nick Guest38781 is now known as grndslm (
22:10:05 Join Zagor [242] (n=bjst@rockbox/developer/Zagor)
22:11:55***Saving seen data "./dancer.seen"
22:15:16evilnickinstinct-: What's the Show Files setting on?
22:16:10evilnickIf it's on Playlists, then RB will only show directories and .m3u fies
22:16:21evilnick(or .m3u8 files)
22:16:26instinct-checkin. and thanks for the heads up on the channel confusion
22:17:46instinct-works like a charm. thanks a lot. Thought that had worked before. is that a theme dependent setting ?
22:18:06instinct-just installed the e280v2 build before and am a happy camper playing around hehe
22:20:15evilnickIt's not theme dependent, but it's on the Quickscreen by default and can be changed without the user necessarily being aware of it
22:21:04instinct-ah, k. good to know. thanks again
22:21:22instinct-i should dive into theme dev i guess...but i prolly wont find the time as much as i'd love too *sigh*
22:25:56 Quit soap (Remote closed the connection)
22:33:45 Quit liar (Read error: 60 (Operation timed out))
22:40:33 Join r0b- [0] (
22:40:59r0b-why isthe Rockbox Utility not detecting my Sansa e200v1 properly?
22:41:21 Quit merbanan (Read error: 60 (Operation timed out))
22:41:52instinct-it isnt deteting the v2 neither for me (e280v2, xp pro 64)
22:42:34r0b-it says Target Mismatch
22:42:46 Quit Grahack ("Tu m'as vu ?")
22:43:13 Join soap [50] (n=soap@rockbox/staff/soap)
22:45:13 Join Reilithion [0] (
22:46:13r0b-someone probably fucked the build up
22:47:30 Join Omlet05 [0] (i=omlet05@
22:51:48 Quit Zagor ("Leaving")
22:52:38 Quit Topy44 (Read error: 60 (Operation timed out))
22:54:54 Join liar [0] (n=liar@
22:56:25 Quit FlynDice (Remote closed the connection)
23:01:59 Join DerPapst [0] (
23:02:00 Join Topy44 [0] (
23:03:13 Join JdGordon1 [0] (
23:03:53 Quit Omlet^RPM (Read error: 110 (Connection timed out))
23:09:10 Join FlynDice [0] (
23:19:07 Nick Spamhost is now known as Space (
23:24:16 Quit Bagder_ ("It is time to say moo")
23:32:40 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
23:43:55 Quit Hillshum (Read error: 60 (Operation timed out))

Previous day | Next day