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 2010-06-02

00:02:05 Part oldtopman ("Leaving")
00:02:44 Quit pamaury (Quit: exit(rand());)
00:03:29kugelis there a way to di 'if defined(X) || defined(Y)' in Makefiles?
00:04:09 Quit avacore (Ping timeout: 240 seconds)
00:05:17 Join avacore [0] (
00:07:06 Quit DataGhost (Ping timeout: 272 seconds)
00:11:14 Join kugel_ [0] (
00:11:22 Quit kugel (Disconnected by services)
00:11:28 Nick kugel_ is now known as kugel (
00:11:32 Quit kugel (Changing host)
00:11:32 Join kugel [0] (~kugel@rockbox/developer/kugel)
00:15:39kugelhah, Ladies and Gentlemen, RaaA plays sound!
00:16:24 Quit CGL (Remote host closed the connection)
00:19:45kugelBagder: thanks!
00:20:49gevaertsDo we get a mail? :)
00:20:59gevaertsAlso congratulations!
00:23:57kugelI think it's not RaaA enough
00:24:55wincentI would like to add that pdbox is playing all test designs flawlessly on the H300.
00:25:14kugelwincent: nice
00:25:59wincentSo, the Pure Data core is ported. From now on, errors therein count as bugs :-)
00:26:56kugelcan I make my first bug report?
00:27:09kugelit doesn't run on the other targets :)
00:27:12wincentOf course!
00:27:54wincentIt will become rather complicated because of GUI controls.
00:29:34wincentThe GUI needs a shift button (meaning that the corresponding switch does not serve any other button) and as far as I can see it is different on every target. The GUI will require some polishing too.
00:30:16tom`kugel: I noticed you're working on RaaA for Android - any idea how hard it would be to port it to a "generic" arm (or x86) Linux?
00:30:39 Join kugel_ [0] (
00:30:56wincentApart from that, some parts of the PD core are obsolete. I will elaborate on it on the mailing list.
00:31:28 Quit guymann_ (Ping timeout: 240 seconds)
00:32:15 Quit kugel_ (Read error: Operation timed out)
00:32:40 Quit kugel (Ping timeout: 245 seconds)
00:36:32wincentkugel: pdbox runs also on H100 targets (as far as I know). Feel free to try it with contents of , which is provided on the (GSoC) Wiki page of pdbox.
00:36:50 Join TexasRockbox [0] (
00:38:10 Join kugel__ [0] (
00:38:28 Nick kugel__ is now known as kugel (
00:38:32 Quit kugel (Changing host)
00:38:32 Join kugel [0] (~kugel@rockbox/developer/kugel)
00:40:12 Join kugel_ [0] (
00:40:32 Quit kugel (Disconnected by services)
00:40:32 Quit TexasRockbox (Client Quit)
00:40:36 Nick kugel_ is now known as kugel (
00:40:40 Quit kugel (Changing host)
00:40:40 Join kugel [0] (~kugel@rockbox/developer/kugel)
00:41:25kugeltom`: with or without sdl?
00:41:42kugelwith sdl, not that hard. without there's a bit more involved
00:46:22 Join voRia [0] (
00:46:50 Quit kugel (Ping timeout: 245 seconds)
00:49:38 Quit ender` (Quit: On the contrary, if you never procreate, neither will your kids.)
00:51:12 Quit efyx (Remote host closed the connection)
00:51:33 Join efyx [0] (
00:57:19 Quit esperegu (Remote host closed the connection)
01:09:58 Join kramer3d [0] (~kramer@unaffiliated/kramer3d)
01:10:55 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
01:26:14 Quit MethoS- (Remote host closed the connection)
01:30:00 Join Kitr88 [0] (
01:31:04 Quit liar (Read error: Connection reset by peer)
01:33:49 Quit Kitar|st (Ping timeout: 240 seconds)
01:34:50 Quit Kitr88 (Ping timeout: 265 seconds)
01:39:34 Quit elcan (Read error: Connection reset by peer)
01:40:11 Join Kitar|st [0] (
01:43:25 Quit efyx (Remote host closed the connection)
01:45:32 Join puetzk [0] (
01:52:56 Join elcan [0] (
01:55:48 Part toffe82
01:57:10 Join Richard_N [0] (
01:57:55wincentIf somebody would like to try pdbox, do it with designs from (beware of Daft_Widget.pd and dj_scratch.pd , though −− they make pdbox hang!)
01:58:15***Saving seen data "./dancer.seen"
02:02:38Richard_NI'm trying to create a build enviroment in Ubuntu, and I just did the sudo ./, and it's showing checking for ..., and it looks like it's looping through all the same stuff, how long should I let this run for?
02:15:51 Join anewuser [0] (anewuser@unaffiliated/anewuser)
02:16:18gevaertsThat depends on your system, but on slow machines it can take an hour or even more
02:17:47Richard_Nok, cool
02:17:49 Quit Richard_N (Quit: CGI:IRC)
02:18:14puetzkIs rockbox capable of recording/playing at the same time (to record while listening to a click track or similar?)
02:18:41puetzkSansa Clip+ if it matters
02:21:38 Quit detaos (Remote host closed the connection)
02:28:47 Quit DerPapst (Quit: Leaving.)
02:31:23 Quit voRia (Quit: Leaving.)
02:32:57 Join CaptainKwel [0] (
02:33:08 Quit bieber (Ping timeout: 240 seconds)
02:33:25 Join detaos [0] (
02:33:25 Join bieber [0] (
02:41:51tom`kugel: with SDL - that's good to hear
02:42:05puetzk /topic
02:45:35JdGordonbieber: thanks :)
02:48:42 Quit puetzk (Quit: puetzk)
02:48:59 Join Zigtown [0] (
02:50:41 Join puetzk [0] (
02:51:21CIA-6New commit by jdgordon (r26458): revert an accidental commit in r26078
02:57:14saratogapuetzk: no
02:57:25puetzksaratoga: thanks for the answer
02:57:30puetzkcouldn't find a way in the menus or manual
02:57:49 Quit wincent (Ping timeout: 240 seconds)
02:58:13puetzkBut I wasn't sure I'd looked everywhere
02:59:05saratogain rockbox if something can't be done at the same time as playback, it should just stop playback
02:59:14 Join krabador [0] (
03:01:29puetzksaratoga: which is what it seems to do (other than recording from FM, which can keep playing)
03:03:17puetzkHmm. I used to just playback from audacity and record the next piece, but the Clip+ seems to have a clock that's inaccurate enough to make that work poorly (the beat drifts noticeably off when I import the recording). It's better at 48kHz than at 44.1, but still too much to piece things together against a click track. Maybe I'll grab the Vmware build setup and see if FS #10906 helps recordings.
03:07:58JdGordonbieber: the editor is looking good :) though it looks like there might be a bug with the viewport change
03:08:11puetzkI am pleasantly blown away by how good the mic is :-) I was expecting something I could practice with, and got... a lot more
03:09:05pirokoARGH. Just found out that my ipod is useless because apple encrypted the firmware on it -_-
03:17:58 Quit steve|m (Ping timeout: 276 seconds)
03:29:38 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
03:31:23 Join steve|m [0] (
03:43:31 Quit CaptainKwel (Quit: Ex-Chat)
03:48:32bieberJdGordon: What was the bug?
03:48:45JdGordonhey, I've found a few issues
03:49:02JdGordonit seems the gui stops working as soon as a valid viewport line goes in
03:49:09JdGordonI dont know if that is the gui of the parser
03:49:35JdGordonnext, the * doesnt allow the braces to be left off completly (1 line fix)
03:50:07bieberOh, that reminds me, I think you have the * wrong in the tag table
03:50:20bieberIt doesn't make anything optional, it just specifies that any optional arguments must be grouped together
03:50:36bieberSo if you want everything optional you still need to have a | in there
03:50:38JdGordonand escaped chars dont seem to be parsed correctly... %% is coming up with illegal tag
03:51:16JdGordonwas * done so we can do a all or none thing?
03:51:27JdGordon*fIIII was meant to mean either all those params are there, or none
03:51:33bieberYes, if you want all tags or none of them then just start out with *|
03:51:54bieberBut you also have the option of, say, *ii|ii, which would require that either all or none of the optional integers be inccluded
03:53:41bieberAnd okay, I do see your Viewport issue now. Oddly enough, it doesn't do that if you have other stuff on the same line as the viewport declaration, I'm sure it's in the parser somewhere
03:53:44JdGordonI thought | meant it would accept if any after the bar are missing? *ab|cd to me looks like what I would have hoped) is either nothing or at least a and b, c and d are optional but only accepted if a and b are there
03:54:21JdGordonadd text after the viewport and the output is empty
03:54:29bieberYep, I see it now
03:54:41bieberHeh, I need to take down a list of things to fix tonight
03:54:59bieberIt also occurs to me that attempting to specify a negative integer as a parameter will also probably fail miserably
03:56:01bieberAnd the | just makes everything after it optional, but if there's no * then you're free to stop at any point after the bar
03:56:20bieberSo ab|cd would be satisfied by specifying ab, abc, or abcd
03:56:30bieberBut if you put the * at the beginning, it has to be either ab or abcd
03:57:17JdGordonright, but I want to say either nothing OR ab OR abc OR abcd
03:58:01JdGordonadding "|| (star && *cursor != ARGLISTOPENSYM)" to skin_parser.c line 413 seems to fix that
03:58:16***Saving seen data "./dancer.seen"
03:59:09bieberSo you want * to always allow an empty argument list?
03:59:58bieberOkay, I can make it mean that if you're sure you'll never want a situation where you have some required arguments and some optional arguments that must always be specified together
04:00:16bieberIf you want both of them to be possible, then we'll have to introduce groupings to the tag format codes
04:01:49JdGordonyeah, I don't think that will be required
04:02:02bieberOkay then, could you just go ahead and commit that change to SVN?
04:02:12JdGordonI'm somewhat surprised *fI|II doesnt just work
04:02:35bieberAs I have it coded now, anything before a | MUST be there in order for it to parse correctly
04:02:43puetzkhmm, I get it all set up and then notice FS #10906 is for as3525v1, not v2. D'oh
04:03:23 Join Eosphere46 [0] (IceChat7@
04:04:47bieberOther than that, the only issues you've found are failed parsing after a viewport declaration and character escaping not working correctly?
04:05:22JdGordonyeah, so far :)
04:05:24bieberI just confirmed that negative numbers don't parse at all, so I'll be on in a couple hours to fix all this
04:05:52bieberIIRC, there _can_ be negative numbers as parameters to tags, right?
04:06:11JdGordonI want to fix the debug output to be more useful.. right now I've hacked itto show the incorrect tag, but char and line numbers owuld be sweet (leave that to after gsoc though)
04:06:51 Quit Eosphere46 (Quit: IceChat - Keeping PC's cool since 2000)
04:07:14bieberOkay, I'll get it all taken care of tonight. I've got to run for now, should be back in two or three hours. Feel free to commit anything you've hacked up while I'm gone, I'll make sure to svn update before I do anything later
04:07:20 Join Eosphere46 [0] (Eosphere46@
04:11:51JdGordonbieber: oh the last thing.. is it really necessary for the NEWLINE element to be in the parse tree?
04:13:24bieberI suppose it can be left out now, since we have the LINE groupings
04:13:30 Quit Eosphere46 (Read error: Connection reset by peer)
04:13:40 Quit kenguest (Ping timeout: 240 seconds)
04:13:42bieberI'll cull that out
04:17:11 Join Eosphere46 [0] (IceChat7@
04:17:26JdGordonthat should make the tree slightly easier to manage I think
04:17:37JdGordonat least if it will be user visible
04:21:10 Quit Eosphere46 (Read error: Connection reset by peer)
04:21:36 Join Eosphere46 [0] (IceChat7@
04:26:22 Quit Eosphere46 (Read error: Connection reset by peer)
04:29:06 Join Eosphere46 [0] (Eosphere46@
04:31:27 Join kenguest [0] (
04:32:32 Quit BHSPitMonkey (Remote host closed the connection)
04:33:43 Quit TheSeven (Ping timeout: 240 seconds)
04:36:44 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:36:44 Quit pixelma (Disconnected by services)
04:36:48 Quit amiconn (Disconnected by services)
04:36:49 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:36:58 Quit Eosphere46 (Read error: Connection reset by peer)
04:37:04 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:37:11 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:38:00 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:44:32 Join Barahir [0] (
04:48:14 Quit Barahir_ (Ping timeout: 276 seconds)
04:49:49 Quit anewuser (Quit: for SELL 2 by the price of 1 now!)
04:51:22 Quit kenguest (Ping timeout: 260 seconds)
04:51:51 Join CGL [0] (~CGL@
04:57:55 Join phanboy4 [0] (
05:05:28 Quit fdinel (Quit: Miranda IM! Smaller, Faster, Easier.
05:13:32 Join kenguest [0] (
05:14:02 Join Eosphere46 [0] (Eosphere46@
05:14:53 Quit Eosphere46 (Client Quit)
05:21:19 Join krazykit` [0] (~kkit@
05:21:19 Quit krazykit (Read error: Connection reset by peer)
05:28:08 Quit elinenbe (Ping timeout: 272 seconds)
05:33:36 Quit Horscht (Quit: Verlassend)
05:49:15S_a_i_n_t <piroko> ARGH. Just found out that my ipod is useless because apple encrypted the firmware on it
05:49:32S_a_i_n_tpiroko: You do know that the Nano2g uses encrypet FW right?
05:58:19***Saving seen data "./dancer.seen"
06:00:28 Join Dhraakellian [0] (
06:01:31 Quit kramer3d (Quit: Leaving)
06:01:50Dhraakellianfor per-track albumart, ./filename.{jpeg,jpg,bmp} doesn't include the original file extension, does it?
06:02:38*Dhraakellian is finally making progress on his rockbox moodbar idea
06:03:19DhraakellianI have a ruby program to render the .mood files to bitmaps of arbitrary size
06:03:59*S_a_i_n_t fails to parse "include the original file extension"
06:04:18Dhraakelliansoon I'll need to get around to making a WPS that can make use of them as per-track album art above/below the progress slider
06:04:48DhraakellianS_a_i_n_t: 01-title.bmp vs. 01-title.ogg.bmp
06:05:01S_a_i_n_tI also don;t really understand "per track" AA, it seems useless to me...per album AA is fine for me.
06:05:33DhraakellianS_a_i_n_t: well, since the moodbar is a representation of what's going on in the track at a given time, per-track is kinda needed
06:05:44Dhraakellianand, thankfully, supported, according to the manual
06:05:53Dhraakellianand, hey, deja vu all over again
06:06:12S_a_i_n_tthe way I see it, all the tracks in an album are going to have the same AA no?
06:10:51Dhraakellianand the manual (appendix C in the PDF copy I have open here) appears to indicate that per-track should work
06:10:58Dhraakellianso I should probably just go test it out
06:25:48 Quit krabador (Quit: Leaving)
06:29:12 Quit panni_ (Read error: Connection reset by peer)
06:53:49 Join esperegu [0] (~quassel@
07:01:05 Quit saratoga (Ping timeout: 252 seconds)
07:03:52 Join Strife89DS [0] (
07:11:25kisakthere's some dispute between my cousin and I. There is a vibration with his Fuze v1 which is not consistant with the information available ...
07:13:59 Quit Strife89DS (Disconnected by services)
07:14:22kisakhe thinks there is a 1.5" hdd in it
07:15:00 Join Strife89DS [0] (
07:15:05S_a_i_n_twell, there isn;t.
07:15:23kisakindeed ... so where's the vibration coming from?
07:15:39FlynDicehis mind...?
07:15:53S_a_i_n_tthere shouldn't be anything in there capable of vibrating.
07:15:59kisakI noted the vibration as well a couple weeks ago
07:16:25S_a_i_n_tdoes the FuzeV2 have that clicky-wheel thing?
07:16:34kisakit does
07:16:42S_a_i_n_tclicky-wheel-spinng thing.
07:16:53S_a_i_n_tand, is it when the wheel is being used?
07:17:12 Join esperegu_ [0] (~quassel@
07:17:15 Quit esperegu (Read error: Connection reset by peer)
07:17:19S_a_i_n_tIIRC there's just a little piece of metal that rubs against a tootleh wheel to make that "click"
07:18:18kisakwell, he's trying to re-confirm this oddity
07:19:06pixelmaBagder: there are currently no swcodec target manuals available on the daily manuals page. I can build e.g. an X5 manual just fine locally so don't know where the problem is. Could you have a look?
07:19:32 Quit Strife89DS (Read error: Connection reset by peer)
07:19:32S_a_i_n_tkisak: I'd say "if it anint tobacco, don't smoke it" ;)
07:19:52S_a_i_n_tthere *shouldn't* be anything in there capable of vibration
07:20:09JdGordonbieber: I dunno if you're gonna like this... I twisted the param scanning logic inside out so it loops on the tag_args variable instead of the num_args (which I think feels more natural) and I have it working again with *ab|cd where %aa, %aa(a,b), %aa(a,b,c) and %aa(a,b,c,d) all are accepted.
07:20:19JdGordonI'm not sure if %aa() should be allowed
07:21:08bieberI just changed a couple lines that should make it treat the * specifier the way you want it to
07:21:51bieberIf you think your change makes it read more intuitively, though, you can commit that. Does it still error out the same way and everything?
07:22:57JdGordonwell it is your code so it is up to you :) I think the error handling is the same
07:23:23kisakwell ... to end that vibration line of thought ... looks like it was gremlins
07:23:31bieberI'll commit my change now, just try it out and see if it works the way you want it to
07:24:00S_a_i_n_tkisak: Well, you can *definitely* tell him/her that it is FLASH based, no HDD, none at all.
07:25:33kisaknow how about a basic port of rb for the ipod classic springing out of a rabbit hole?
07:26:29JdGordonbieber: ignore the 2nd arg to skin_error()
07:27:03S_a_i_n_tkisak: You're free to do the work on the port.
07:27:25S_a_i_n_tAsking for it won;t help it to happen unfortunately.
07:27:41CIA-6New commit by bieber (r26459): Theme Editor: Altered the way the parser handles asterisks in the tag argument specifiers
07:29:15kisakS_a_i_n_t: thanks for the reminder
07:29:28S_a_i_n_tanytime ;)
07:29:52 Join Strife89DS [0] (
07:29:53pixelmaBagder: seems it's not "every swcodec target" but everything that comes after the D2 manual. As I remember seeing a similar effect semi-recently, I'm checking the M3 one now which is next in the "list"
07:29:55bieberOh wait, so now if the * is specified we don't require all the optional arguments to be there?
07:31:56JdGordonbieber: did we have a misunderstanding again? :(
07:32:19bieberI think I've got it straight now
07:32:21*JdGordon tried to remove all ambiuity this morning
07:32:59pixelmaBagder: yep, M3 manual broken and seems to stop "further" manuals. I remember asking Zagor about the build script and he said it doesn't stop building if one manual fails... could it be the upload script then?
07:33:05bieberSo basically, a | specifies that anything after it is optional, and a * specifies that you're free to leave off ALL of the arguments including the (), without any affect on whether the optionals have to be included?
07:33:07JdGordonfrom the begingin I was hoping * and | wouldnt actually change eachother... * was supposed to mean that either use the param list or no param
07:33:18JdGordonbingo :)
07:33:20bieberOkay, I think I have it right now, then
07:33:38JdGordonmy test didnt work :)
07:33:39bieberIt wasn't doing that in that last commit, but now it is, and I think I'm about to have the character escaping bug fixed
07:33:44JdGordonalso there is a minor bug
07:34:26JdGordonI tihnk its a bug... s and f will accept an empty string
07:35:19JdGordon*fII|II doesnt work as expected.
07:35:25JdGordonthe | is ignored...
07:35:58bieberThat should be fixed in the next commit
07:36:10bieberAnd do we want to reject empty strings?
07:36:17JdGordonI don't know
07:36:26JdGordonI guess not
07:36:35JdGordonyeah, leave that as a featuer :)
07:42:26bieberWhen it's scanning a string for a parameter, do you want to strictly enforce what characters can be escaped?
07:42:56bieberi.e., if a string parameter includes %E, should I just go ahead and parse that as an E?
07:43:09Speedy2Ok, very offtopic, anyone here use RDP?
07:43:29S_a_i_n_tif it's very offtopic...then, why?
07:43:30 Quit Strife89DS (Max SendQ exceeded)
07:43:46Speedy2Because there is life in this channel.
07:44:02S_a_i_n_tthat's not a reason to ask offtopic questions here
07:45:35CIA-6New commit by bieber (r26460): Theme Editor: Fixed a small bug with asterisk handling in tag parameter argument codes and fixed bugs with handling escaped characters
07:45:44 Quit mc2739 (Ping timeout: 240 seconds)
07:46:38JdGordonbieber: cheers! quick tests looks good
07:51:02CIA-6New commit by pixelma (r26461): Correct the iaudio-remote button names which were added in r26377 to make the M3 manual build again.
07:52:11pixelmaif the changes in r26377 are in the release branch too, this ^ patch should be ported there too but I don't have the time right now
07:52:56pixelmaargh.. unrelated changes included :\
07:53:53 Quit esperegu_ (Remote host closed the connection)
07:53:57 Join solrize [0] (~phr@unaffiliated/solrize)
07:54:11 Join esperegu [0] (~quassel@
07:54:18solrizewhee, i bought a refurb sansa fuze from woot but i think they sent me a NEW unit (v2, sigh)
07:54:49pixelmaI believe these are correct too, just not mentioned in the commit message. Guess they could stay?
07:55:43CIA-6New commit by bieber (r26462): Theme Editor: Fixed code generation with escaped characters
07:57:03pixelmaAlexP: if you have the time, please port at least the last hunk in r26461 to the release branch if necessary. Maybe also have a look at the other changes as I don't have the time right now. Going to have a look in the evening again
07:58:20***Saving seen data "./dancer.seen"
08:08:45 Quit puetzk (Quit: puetzk)
08:12:20CIA-6New commit by bieber (r26463): Theme Editor: Removed the NEWLINE parse tree element
08:14:16 Join stoffel [0] (
08:15:10JdGordonbieber: code gen bug.. %% add a blank line after it
08:15:48 Quit Zigtown (Ping timeout: 260 seconds)
08:15:53JdGordonhmm... maybe not
08:16:09bieberYou mean the lack of an empty line in the parse tree?
08:16:44 Join DerPapst [0] (
08:16:57JdGordon%%\nwe <- the text... change "we" on the left and hit update code
08:17:04JdGordonit turns to %%\n\nwe\n
08:17:34JdGordonremove the blank line and hit update code again, it comes back
08:18:03JdGordonclicking enter in the text editor with no text there segfaults ;)
08:18:20JdGordontell me to shut up and I will :)
08:19:02bieberNo, I'm just glad you're finding this stuff
08:19:49bieberOkay, so I've got a rogue newline popping up in code generation, and the segfault thing I think is just another facet of the viewport handling bug I'm working on
08:25:45bieberFixed the newline issue, now to tackle this viewport parsing nonsense
08:30:13 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
08:32:44 Join B4gder [0] (~daniel@rockbox/developer/bagder)
08:37:13bieberWow, that bug turned out a lot easier to parse than I expected
08:37:54 Quit stoffel (Remote host closed the connection)
08:40:35 Quit GeekShadow (Read error: Connection reset by peer)
08:40:54 Join ender` [0] (
08:41:43CIA-6New commit by bieber (r26464): Theme Editor: Fixed bugs in code generation and viewport parsing
08:44:39JdGordonbieber: haha its gone pretty funny now
08:45:28bieberWhat's gone funny?
08:46:46JdGordon%V(0,0,0,0,0) in the text, parses ok, but then click update code
08:47:25bieberOh, I see it
08:47:35bieberD'oh, forgot to insert the %
08:52:18CIA-6New commit by bieber (r26465): Fixed another code generation bug with viewports and enabled negative numbers in tag parameters
08:52:54bieberOkay, that _should_ take care of all the bugs that have come up so far, and finishes off my list of changes for tonight
08:53:53JdGordonbieber: sorry :) Looks like line counting isnt working anymore
08:53:54 Join petur [0] (~petur@rockbox/developer/petur)
08:54:09bieberOh, duh
08:54:30bieberI think I really need to get some sleep after this next commit :P
08:54:42*JdGordon will stop finding bugs once it successfully parses cabbiev2.wps
08:55:41JdGordonand comments are adding a blank line
08:55:49 Join Strife1989 [0] (
08:55:53JdGordonafter cliking update coe i mean
08:57:31 Quit Strife89 (Ping timeout: 276 seconds)
09:04:34 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
09:04:36CIA-6New commit by bieber (r26466): Theme Editor: Fixed some more code generation bugs
09:04:43bieberHooray for shotgun commits
09:07:57JdGordonnick off before I post another bug :)
09:09:15 Quit moparx (Ping timeout: 240 seconds)
09:11:16JdGordonbieber: yeah bad news :) a real parser bug.. but I'll see if I can fix it if you are done for the night
09:11:37bieberWhat is it?
09:12:16JdGordonnested conditionals doesnt parse
09:12:22JdGordon%?mh<hello|%?mh<|>> fails
09:12:24bieberOh, crap
09:12:32JdGordon%?mh<hello|%?mh<>> works
09:13:55JdGordonI assume it is just a quick line in skin_parse_conditional() to skip over any <>'s when counting children?
09:14:14solrizeis that stuff written in lua these days?
09:14:39solrizehmm ok. maybe some day
09:14:42 Join einhirn [0] (
09:16:10JdGordonsolrize: if you want to do it you should be able to get pretty far with the current lua plugin
09:16:52 Quit r0b- (Read error: Connection reset by peer)
09:17:07solrizeyeah, last time i hung around here there was no scripting in rockbox and concern over memory footprint... i was interested in adding a small lisp runtime, but now i hear rockbox has lua, which is probably more palatable to lots of folks ;)
09:17:31*kenguest does a double take
09:17:41kenguestthere's scripting in rockbox?! :D
09:17:48bieberHmm, code generation is a mess with conditionals too
09:17:54solrizekenguest, yeah, rockbox has lua
09:18:09solrizeit would be cool to scriptify most of the user interface stuff
09:18:22solrizethat was hard coded C last time i looked
09:18:27JdGordonstill is
09:18:46JdGordonlua will let you do almost anything though.. full drawing and button support
09:19:00solrizeoh cool, so it's already integrated with the UI
09:19:35JdGordonno, lua can only be run in a plugin
09:19:45solrizehmm ok
09:19:47JdGordonso you could redo most of the ui in a plugin if you really wanted to :p
09:20:24S_a_i_n_tbut...obviously there's little point in doing so
09:20:52 Join Zagor_ [0] (~bjst@
09:21:23 Quit Zagor_ (Changing host)
09:21:23 Join Zagor_ [0] (~bjst@rockbox/developer/Zagor)
09:22:37 Join Rob2222 [0] (
09:24:14solrizei wonder whether javascript would have been ok since so many people know it
09:24:43B4gderit would not
09:25:08B4gderlua was made to be stand-alone and small
09:25:28B4gderjavascript was not
09:25:29solrizei think small js implementation is possible, i just dunno if anyone has done it
09:25:45 Quit Rob2223 (Ping timeout: 240 seconds)
09:25:48 Quit steve|m (Ping timeout: 276 seconds)
09:26:19solrizei guess i'll read up on lua, i've only looked at it briefly
09:26:21B4gderEVERYTHING is possible given enough effort, that's not what I'm talking about though...
09:26:36solrize <−− this is what i thought of integrating
09:26:44 Join steve|m [0] (
09:27:45 Quit Kitar|st (Ping timeout: 276 seconds)
09:29:56 Join lpereira [0] (
09:30:43solrize LTU thread about lua
09:31:10B4gderno need to tell us you know, we already have lua support...
09:32:16 Join Kitar|st [0] (
09:32:46 Quit Galois (*.net *.split)
09:32:46 Quit tchan (*.net *.split)
09:32:46 Quit joeyg (*.net *.split)
09:32:47 Quit dionoea (*.net *.split)
09:32:47 Quit Torne (*.net *.split)
09:35:57kenguestsolrize: next, B4gder will be saying not to bother reading
09:36:12solrizethat pdf is pretty good, i feel like i understand lua a lot better now
09:36:53B4gderwell, dropping random URLs with no further explanation or reasoning is always a bad idea
09:37:20B4gderits just noise
09:37:24solrizeLTU is a blog about programming languages, i guess i shoulda said that
09:38:21bieberJdGordon: Okay, I got the nested parsing working, now I just have to get it to codegen right
09:38:31kenguestso how long has lua been in rockbox? is it there by default? I mean, I upgraded to rockbox 3.5.1 a month or three ago, so does that mean I can just drop in a lua file and run it on my player or is there more involved?
09:38:58 Join Galois [0] (
09:38:58 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
09:38:58 Join joeyg [0] (
09:38:58 Join Torne [0] (torne@rockbox/developer/Torne)
09:38:58 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
09:39:09JdGordonbieber: great :)
09:39:33Zagor_kenguest: it's been there for about a year
09:40:08JdGordonthat wiki page is exactly a year old
09:42:03S_a_i_n_tDoesn't the game "boomshine" run in LUA?
09:42:15*S_a_i_n_t thought so at least. is indeed.
09:43:30 Quit esperegu (Read error: Operation timed out)
09:47:34 Quit slck (Ping timeout: 276 seconds)
09:48:49CIA-6New commit by bieber (r26467): Theme Editor: Fixed parsing and code generation for nested conditionals
09:49:13 Quit lpereira (Ping timeout: 265 seconds)
09:50:08 Join slck [0] (Venci@Slackware.SlackPix.Com)
09:50:35 Join esperegu [0] (~quassel@
09:51:27JdGordonbieber: quick test looks like it works, real world copy doesnt though... %s%al%?ia<%ia|%?d(2)<%d(2)|(root)>>
09:52:06JdGordonoh hmm.. it doesnt like (root)
09:52:13JdGordonthat needs to be escaped?
09:52:40JdGordon(root%) works...
09:53:02JdGordon() should only need to be escaped inside a paramaterised tag
09:53:51JdGordonah I see why
09:54:20 Join efyx [0] (
09:54:33bieberOh, I think it bails out on either an ENUMLISTCLOSESYM or an ARGLISTCLOSESYM
09:54:38bieberWant me to drop the latter condition?
09:55:43JdGordonwont that break argument parsing?
09:55:59bieberOh yes, I think it would
09:56:11JdGordonit should onyl be checking for them if we are inside a param list
09:56:29 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
09:57:00JdGordonwe can of course make them always need to be escaped if that is better?
09:57:15JdGordonwhich your code gen does anywa
09:57:28pixelmaZagor_, B4gder: could you trigger a new manual build round - and check why everything following the broken M3 manual is missing too (the other should be fine, I know for sure the X5 manual is)? I remember Zagor_ checked the build script last time this happened, maybe it's the upload script (just a guess)
09:58:24***Saving seen data "./dancer.seen"
09:58:49JdGordonbieber: yeah, I think just force them to always be escaped is the easiest solution everywhere... I'll fix my update program
09:58:55 Join Rob2223 [0] (
09:59:55bieberIt's actually an easy fix, because parsing parameter strings and plaintext elements are done in separate functions
10:01:07bieberJust delete line 607 of skin_parser.c and the line you pasted will work fine
10:01:23bieberI'll commit it tomorrow wtih whatever changes I make, I'm gonna sleep now
10:01:46 Quit Rob2222 (Ping timeout: 245 seconds)
10:01:57JdGordonwoo! cabbiev2.wps parses correctly
10:02:15JdGordonI'm not sure that one line change is correct though
10:02:16bieberDoes the tree look correct? And does running a code gen break anything horribly?
10:02:52JdGordonquick glance both look correct
10:03:19JdGordonthat one line change means that , is still not allowed in free text inside conditionals
10:03:26bieberYeah, you should also get rid of the first clause in that conditional
10:04:29bieberThat'll be in the next commit as well
10:04:47JdGordonno, i dont like that "fix"... very easy to break it still
10:04:54JdGordon p, li { white-space: pre-wrap; } %s%al%?It<%It|%Fn,>
10:04:58JdGordonnote the last comma
10:05:18JdGordonand ignore that first garbage
10:05:59JdGordonescaping all those special chars makes the most sense
10:06:07bieberThat probably is the best idea
10:06:31JdGordongreat to see the full wps parsed though :)
10:06:43bieberThere should be a simple fix for that one too: you probably just have to look a level or two up in the parser, but escaping is definitely the easy way out
10:07:24bieberYeah, that too ;)
10:07:48S_a_i_n_tone of the few situations where *that* will be true ;)
10:08:16S_a_i_n_t(easy/correct being one in the same)
10:09:06JdGordonit does mean that the parser needs a line to make it more strict to force the escaping though... but not a big deal just yet
10:09:52JdGordonhaha I completly destroyed my updater though :p
10:09:54bieberTwo lines ;) Strings are scanned both in the scan_string function and in the parse_text function
10:10:11JdGordonit escapes those chars everywhere now... doh
10:10:55bieberHonestly, I wouldn't worry too much about forcing it. I figure if the syntax is unambiguous enough for the parser to figure it out, may as well let them use it
10:11:15bieberThe HTML3 approach to parsing, if you will :P
10:12:33*JdGordon now has a updated cabbiev2.wps parsing in svn themeditor
10:13:14bieberAwesome :)
10:13:22JdGordon$ diff out.a.wps out.wps
10:13:22JdGordon< %ax
10:13:22DBUGEnqueued KICK JdGordon
10:13:22JdGordon< %Vl(a,105,30,-,90,1)
10:13:22***Alert Mode level 1
10:13:22JdGordon> %ax%Vl(a,105,30,-,90,1)
10:13:41JdGordonthat is the whole diff between the input and regenned output
10:14:04JdGordonoh HAHA you're going to kill me :D
10:14:21JdGordonit is important that %ax attach to the next tag...
10:14:59bieberWait, what's %ax do?
10:15:19JdGordonit tells the next tag to align itself with the language direction...
10:15:45bieberOh, that shouldn't be a problem
10:16:06bieberDuring rendering, just set a flag when you encounter one and apply it the next time you render a tag
10:16:25JdGordonyeah I guess so
10:16:25bieberIf it's not a supported tag, you'll have to throw an error during rendering instead of parsing, if that's alright
10:16:42bieberOtherwise we can just add a hard-coded check in skin_parse_tag
10:17:03JdGordonno, I'll sort it out in the middle step (between parse and render)
10:17:09bieberOkay, sounds good
10:17:35bieberI really will sleep this time, let me know if you find any more bugs
10:17:50JdGordongood night
10:18:27 Join lpereira [0] (
10:18:49CIA-6New commit by jdgordon (r26468): we need to escape more chars now, so fix that. test skins with this update parse correctly with the themeditor :)
10:19:26Zagor_pixelma: this is what happened:
10:20:18Zagor_plus a bug in the build script that exited on some errors
10:21:07 Quit Unhelpful (Read error: Connection reset by peer)
10:21:22 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
10:21:43CIA-6New commit by zagor (r26469): Don't exit on build errors.
10:21:56 Quit petur (Quit: reboot)
10:23:22 Join petur [0] (~petur@rockbox/developer/petur)
10:23:23***Alert Mode OFF
10:25:01JdGordonpixelma: what do we want to do about the AA tag... %Cl(15,32,s75,s75), those 2 s's are alignment chars.. really they should be seperate params..
10:26:17JdGordon%Cl(x,y,width,height,optional x align, optional y align)?
10:27:45 Join kugel [0] (~kugel@rockbox/developer/kugel)
10:29:19 Join r0b- [0] (
10:31:11 Quit shai (Quit: Leaving)
10:34:13CIA-6New commit by jethead71 (r26470): FFT Plugin: Revamp the main code to rid it of 64-bit math. Use 32-bit kiss_fft_scalar because 16-bit integers are generally a poor choice for ...
10:37:59r0b-i cant figure out how to install rockbox via the VMWare Build Env
10:39:21JdGordonr0b-: install it on what?
10:39:29r0b-my Clip Plus
10:40:10r0b-i have the file built
10:40:23 Join shai [0] (
10:40:31 Quit phanboy4 (Ping timeout: 245 seconds)
10:41:57JdGordoncan the vm see the clip+?
10:42:02JdGordonyou need to share the usb port
10:42:26r0b-i enabled the passthrough
10:42:41r0b-the VMWare build environment for Rockbox is weird
10:43:01r0b-the Clip+ shows in lsusb
10:43:06r0b-but i dunno where its mounted
10:47:23JdGordonwhat should we do with the alignment modifiers in the %Cl tag which are ignored? just drop them in the update?
10:47:54JdGordond i s are legal for backwards compat reasons
10:47:56tmztr0b-: it might not be mounted automatically
10:48:15tmztif you need filesystem access, use dmesg to see what device it appears as
10:48:19tmztand mount it manually
10:49:07r0b-Directory sread (sector 0x3b10) failed
10:49:07r0b-hub.c: new USB device 00:07.2-1, assigned address 6
10:49:07r0b-WARNING: USB Mass Storage data in
10:49:26r0b-new USB device 00:07.2-1, assigned address 6
10:49:26r0b-WARNING: USB Mass Storage data integr
10:49:38tmztthis is dmesg? doesn't look familiar
10:49:39 Join blairb [0] (
10:49:58r0b-Directory sread (sector 0x3b10) failed
10:49:58r0b-hub.c: new USB device 00:07.2-1, assigned address 6
10:49:58r0b-WARNING: USB Mass Storage data integrity not assured
10:49:58DBUGEnqueued KICK r0b-
10:49:58r0b-USB Mass Storage device found at 6
10:50:03r0b-thats dmesg
10:50:20tmztwhat hci does vmware emulate?
10:50:33tmztlspci should tell you, but don't paste the full thing here
10:50:37tmztalso, pastebin lsusb -v -v
10:51:50r0b-lsusb shows the same stuff dmesg did
10:51:54 Join funman [0] (~fun@rockbox/developer/funman)
10:53:04tmztthat's the kernel console
10:53:24tmztdmesg -n 1
10:53:34 Quit lpereira (Quit: Leaving.)
10:53:39tmzthow are you copying text from vmware?
10:53:59r0b-im using vmware tools
10:54:04r0b-im about to KILL vmware
10:54:19r0b-all i was trying to do was fix a problem in rockbox
10:59:40CIA-6New commit by jethead71 (r26471): FFT Plugin: Should use scale_factor, not LCD_WIDTH/HEIGHT to break the spectrogram loop.
11:00:08tmztr0b-: vmware tools work in console?
11:00:28tmztwhy do you need to install from vmware, can't you install from windows?
11:02:04r0b-i gotta get the .zip rockbox's build shit made
11:02:08r0b-from the VM to windows
11:06:56tmztdownload sbd.exe and use nc in the rockbox live environment?
11:07:33 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:07:54CIA-6New commit by funman (r26472): backport r26444 ...
11:07:57CIA-6New commit by funman (r26473): backport r26438: Italian translation update
11:08:01 Quit JdGordon (Ping timeout: 245 seconds)
11:17:17CIA-6New commit by funman (r26474): Fix version to "3.6RC" instead of svn version + build date
11:17:22AlexPpixelma: I'll have a look when I get home this evening if you haven't got to it already
11:17:30funmando we need to do RC builds?
11:18:24kugelnot 1 day before the release...
11:18:45funmanwhy not?
11:18:47CtcpIgnored 2 channel CTCP requests in 36 minutes and 36 seconds at the last flood
11:18:47*kugel wonders if anybody has actually tried the release branch
11:18:57*funman just did
11:19:29kugelwe should have RC1 when freezing and RC2 when branching IMO
11:21:32funmanI can make bins, put them on my home page, and post on the forum use 'make -j' , which brings my system down to its knees :/
11:24:29 Join LinusN [0] (linus@rockbox/developer/LinusN)
11:25:11funmanBagder: did you see my msg about the e200v1 ?
11:35:27 Join JdGordon [0] (~6e17cc41@gateway/web/freenode/x-dsqtsqthippfcyrx)
11:35:49funmankugel: did you see FS #11267 ? fuzev1 freezes but playback continues, the button led can be activated but not deactivated, and backlight doesn't work
11:39:11 Join JdGordon1 [0] (
11:40:50funmanbacklight is handled in a thread but buttonlight turning on in the tick
11:42:30kugelfunman: that's the same freezes I experienced
11:43:36funmanhow did it happen? i can trigger them with recording
11:43:48 Quit JdGordon (Ping timeout: 252 seconds)
11:43:52kugelfunman: the only possiblity is that there's a dead lock in sd operations, i.e. wakeup_wait never returns
11:43:57kugelI think
11:44:01JdGordon1what header is toupper() in?
11:44:05funmanthat's what i'm thinking yes, we could make it timeout
11:44:25funman[fun@kru /media/bordel/rockbox/firmware/libc]% grep toupper .
11:44:25funman./include/ctype.h:int _EXFUN(toupper, (int __c));
11:44:28 Join elinenbe [0] (
11:44:33kugelwell, it works on all other targets without timeout so there's something else wrong
11:44:56funmankugel: all others = others AMS?
11:44:57kugelalso, how do you know if the transfer finished when the timeout wears off?
11:45:11kugelfunman: PP too, no?
11:45:13funmanwakeup_wait has a return value
11:45:20JdGordon1thanks funman
11:45:26 Join watto [0] (~watto@
11:45:47funmanwell wakeup_wait not returning just indicates a bug in the driver
11:47:39kugeland what would you do in case the timeout wears off?
11:48:26funmanretry = true;
11:48:28*S_a_i_n_t is looking at a thread on the forums...
11:48:29S_a_i_n_tI've handled a FuzeV2 a few times, doesn't the OF *always* say "Goodbye" when powering off the unit?
11:48:45S_a_i_n_tIs it possible that the OP has only just noticed this?
11:49:05funmanS_a_i_n_t: the message changed in an OF upgrade i think
11:49:05S_a_i_n_tI find it highly unlikely installing RB made the OF gain an additional feature.
11:49:44S_a_i_n_tHe needs to come to terms with the fact that something in his unit has gone boom.
11:49:56S_a_i_n_tIt's a shame that he'll blame RB for it though.
11:50:50funmangevaerts: Torne: isn't there some ipod feature that needs to be turned off for the release ?
11:51:07CIA-6New commit by jdgordon (r26475): change the %Cl tag to be cleaner. it it now %Cl(x, y, width, height [, xalign][, yalign]) ...
11:52:18gevaertsfunman: yes, usb
11:53:00funmani'm making testing builds now, could you make the change in export/config so I can see which targets I need to rebuild?
11:53:18gevaertsnot now, sorry. Look at the 3.5 branch
11:53:23 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
11:54:11funman4g, color, mini1g, mini2g, nano1g, video <- does that sound right ? (r24350)
11:55:40S_a_i_n_tWhat is the reason USB gets turned off again?
11:56:12funmanUsing our own USB code exposes some problems, such as the current charging issues (for many people, charging basically doesn't work)
11:56:29S_a_i_n_tthen, I'd leave it for the Nano1g
11:56:34S_a_i_n_tit charges fine.
11:56:38S_a_i_n_tneeds less power
11:56:51funmanfor you
11:56:58S_a_i_n_tI have *no* issues with it in SVN
11:57:04funmanIIUC it depends of the charger
11:57:04S_a_i_n_teven charging while using it.
11:57:04TorneIt doesn't charge fine..
11:57:16TorneThey *all* only draw 100mA
11:57:28TorneThat will eventually charge your nano1g..
11:57:31S_a_i_n_twell, define "fine"...mine charges in less than 40 mins.
11:57:38S_a_i_n_tfrom flat
11:57:41TorneI don't believe you :)
11:57:41CIA-6New commit by funman (r26476): Use OF USB on the ipods. ...
11:57:51Torneunless the nano1g's battery is *really* tiny
11:57:59S_a_i_n_tTorne: Seriously?
11:58:15Torneit's a 300mAh battery?
11:58:23TorneThat should take at least two hours to charge, then :)
11:58:26***Saving seen data "./dancer.seen"
11:58:30TorneYou are not charging it from flat to full, if it takes longer
11:58:32Torneer, less
11:58:46S_a_i_n_t40 mins, unless it reports wrong.
11:58:48Tornewith working charging you coul probably do it in well under half an hour
11:58:49JdGordon1S_a_i_n_t: obey physics...
11:59:31funmani dunno what exactly rockbox reports has the time left to full charge but it's definitely wrong on Sansa AMS
11:59:32Torne100mA, 5V charging 3.7V nominal, 300mAh, is 2.22 hours assuming linear charging..
11:59:41Tornewhich it isn't (it will be slower than that)
11:59:50Torne(but also, flat is not flat, and full is not full, generally)
12:00:13S_a_i_n_tthen it *must* be shutting off too soon and never fully charging.
12:00:19Torneanyway, the point is that "working" charging should generally be somewhere between 4 and 9 times faster
12:00:22S_a_i_n_tas it reports a full charge in 40 mins
12:00:23Tornedepending on rockbox's idle power draw
12:00:48 Quit esperegu (Remote host closed the connection)
12:00:59Torneso.. either your nano1g is in fact drawing 500mA in which case that's.. interesting, but you'd have to check it with a meter
12:01:00S_a_i_n_tand that's if RB reports it flat, and shuts down.
12:01:08TorneThat's not actually flat, though :)
12:01:21S_a_i_n_tI know, but y'know what I mean :P
12:01:29 Join esperegu [0] (~quassel@
12:01:45Torneanyway, meter it :)
12:02:03Torneunfortunately as far as i know only the ipodvideo has the 4066_ISTAT line available on an ADC has a 'doonly' variable to build only specified targets but it doesn't seem to be used
12:02:14Torneso all the others, you need a multimeter and a cut-up USB cable
12:02:24S_a_i_n_tHmmmm, I'll need to dig out the multimeter.
12:02:42S_a_i_n_terrr, probably buy a better one actually.
12:03:18funmanhm no i can't read
12:03:21Torneanyway, as well as the charging thing I believe there are still some models where it's noticably faster to use OF USB
12:03:24JdGordon1funman: it is used if you only want to build one target
12:03:38S_a_i_n_tWell, there is that.
12:03:53S_a_i_n_tTransfer on the Nano1g in particular is a REAL drag.
12:06:44TorneS_a_i_n_t: i'm hoping to sort out USB charging in general (and add support for at least some of the ipod models) at devcon
12:06:59 Join n1s [0] (~n1s@rockbox/developer/n1s)
12:07:21Tornei don't like the way we control it at the moment but i'm having a bit of a problem nailing down exactly what I *do* like, so i'll borrow some opinions from othe rpeople this weekend ;)
12:09:48funmanTorne: i'll tell you what you like, no problem
12:10:49 Quit petur (Quit: *plop*)
12:11:03*S_a_i_n_t doesn't think that'd be appropriate in this channel :P
12:11:08funmankugel: recording PCM@96kHz seems to be the fastest way to make it crash
12:12:09kugelfunman: for me, just pressing play to resume playback after boot, then quickly enter the playlist viewer and select another track (before the first track finished buffering I think) repro'd it 90% of the time
12:12:37 Join jitu [0] (~jitu@
12:12:50jituhi all, need your help in Ipod mini 2g (4gig)
12:12:57jitumy ipod hdd was gone and it was producing disgusting noises .. i can see a sad face on my ipod screen .. and i tried using the hdd with a CF reader .. but no luck .. so I finally decided to go for a CF card
12:13:03jituI got a Sandisk 4gig card and bt I'm not able to make it work .. :((
12:13:04kugelsometimes it would only crash playback, i.e. the UI still works, but if I then tried to stop playback it would freeze completely
12:13:48kugelprobably it depends on which thread causing the deadlock in the driver
12:14:03funmanplaying from µSD?
12:14:19kugelI'm not sure but IIRC it didn't matter
12:14:24S_a_i_n_tdon't you need a specific revision for a CF'd mini? S_a_i_n_t looks at JdGordon1
12:14:41JdGordon1CF doesnt need anything to work
12:14:50kugelbut now I remember that I seemed to be unable to repro it with dircache disabled
12:14:55JdGordon1*maybe* the bootloader, but i dont remember
12:15:24n1sjitu: here's a page with some info about cf modding
12:16:32jituhumm when I plug my ipod to my ubuntu .. its not able to load it .. throwing errors like "Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK" in the log .. I can use the same CF card with my card reader .. format it into 2 partition one with type empty and other with fat32 and put the fimware on the partition1 (downloaded from rockbox) .. still after that my ipod doesnt work ..
12:16:39 Quit bieber (Ping timeout: 265 seconds)
12:16:56 Join bieber [0] (
12:17:22n1syour computer doesn't show it as a msc device or the ipod doesn't boot?
12:18:12jitumy computer doesnt shows it as it is not able to read the device and but the ipod boots .. first it will come up with a folder with exclamation sign and then Do not diconnect sign
12:18:34jitubut the computer stills doesnt shows it and keep throwing "Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK" in log ..
12:19:07n1sthat sounds to me like the ipod is having trouble with that CF card, maybe someone else knows more
12:19:37jituI have Sandisk Ultra .. 4 gig
12:20:10tmztwhere is the iPod running firmware from when you try to connect to it?
12:20:41jitusorry tmzt, I didn't understand your question
12:21:15funmanwho is Dwayne Asbury ? (on the wiki)
12:21:28tmzthow are you booting the ipod when you tried mass storage
12:22:35jituhumm actually I am doing nothing .. my ipod circuit board is out .. i just plugged in the CF card (after doing the formatting using CF card reader and running dd if=/media/ipodjunk/mbr-mini2g-4gb.bin of=/dev/sdb1) .. as soon as I connect my data cable .. it boots up
12:23:01tmztso it works in the iPod disk connect firmware but not rockbox?
12:23:20tmztcan you follow those instructions in the wiki or get the identity information from your host os (won't work through USB though)
12:24:30Tornejitu: the MBR doesn't go on sdb1..
12:25:04Torneit really should work without preparing the card though
12:25:35funmanwe can't revert a change in FosWiki ? (except manually)
12:25:43jituoh is it ..
12:26:40funmankugel: same symptoms with
12:27:38 Quit GeekShadow (Ping timeout: 264 seconds)
12:28:07jituTorne and tmzt .. correct me if I am doing anything wrong .. I got this new CF card .. now I need to have the firmware on this before I use it with my ipod .. so here is what I do .. 1) if I plug this to my ipod .. and connect it with my pc using usb .. my ubuntu doesnt recognize it .. throwing sd 48:0:0:0: [sdb] READ CAPACITY failed
12:28:07jitusd 48:0:0:0: [sdb] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
12:28:08jitusd 48:0:0:0: [sdb] Sense not available.
12:28:08DBUGEnqueued KICK jitu
12:28:08jitusd 48:0:0:0: [sdb] Write Protect is off
12:28:08jitu\sd 48:0:0:0: [sdb] READ CAPACITY failed
12:28:10jitu\sd 48:0:0:0: [sdb] Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK
12:28:12jitusd 48:0:0:0: [sdb] Sense not available.
12:28:14jitusd 48:0:0:0: [sdb] Attached SCSI removable disk
12:28:16jitu 2) so I got a CF card reader and now my ubuntu is able to recognize this 4 gig device
12:28:26funmanjitu: please use pastebin
12:28:27kugelfunman: I wonder what this caused. I'm fairly sure it's a regression since I don't remember any such problems around when we promoted the fuzev1 stable
12:28:29TorneWhat is the ipod running, though?
12:28:38jituapologies funman
12:28:43TorneDid you hold select+play to go into emergency disk mode?
12:29:17jitunothing .. select+play doesnt do anytihng because the CF card is the new and empty one
12:29:43funmankugel: 13 commits in this file since 3.5
12:29:56S_a_i_n_tI'm not the only one confuzed here right?
12:30:27Torneselect+play disk mode is implemented in ROM and always works
12:30:34Torneit doesn't matter what's on the storage device.
12:30:45TorneIf you *don't* press that, then i can't see what it's booting (except the same disk mode, perhaps?)
12:30:47jituoops .. then it seems like my ipod is screwed .. let me try the same again :(
12:30:51Tornewhat is on the screen when it boots?
12:31:27S_a_i_n_tyeah, I'm confused by "booting", and "boots"
12:31:40S_a_i_n_tif the card is empty, I don't get it.
12:32:49Tornethe easiest way to do this, if you have access to a Windows computer, is to put the card in the ipod, boot disk mode with select+play, connect it to itunes and hit restore. messing about dd'ing bits of mbr and firmware partition onto the card is only worth attempting if the former is just not possible :)
12:35:11n1si thought it booted into the emergency disk mode automatically if it failed to boot the firmware
12:35:22CIA-6New commit by jdgordon (r26477): clip int scanning to 15 digits to remove an unecessary malloc(). ...
12:35:29jituseems like something is screwed big time .. freaking buttons are not working .. with and without the ipod case :(
12:35:31Tornen1s: it will for some errors, yes
12:35:43n1swhich reminds me that i should find a ribbon cable to finally fix my ipod mini
12:35:51JdGordon1jitu: did you snap the wheel cable?
12:36:01JdGordon1ribbon cable from the wheel
12:36:02Tornejitu: not working how? what are you running on the ipod at the moment?
12:36:03jituyeah I tried tht too :(
12:36:05CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
12:36:05*JdGordon1 did that fucking 3 times
12:36:07n1sthat cable is stupidly fragile
12:36:14 Nick JdGordon1 is now known as JdGordon (
12:36:33n1sJdGordon: where did you find new ones?
12:36:40S_a_i_n_ts/fragilt/brittle/suicidal/ i think
12:36:47S_a_i_n_tall of the above
12:37:00funmankugel: can you try r26163 ?
12:37:47funmanthis one could only be affected by the udelay() change
12:37:59jituhumm seems like the wheel cable is gone .. when I plug in the usb cable the ipod boots automatically .. but if I try to boot it using the wheel .. it doesnt do anything
12:38:42kugelfunman: the problem exists longer than the udelay
12:38:47JdGordonyeah, you broke the ribbon
12:38:58jitujust plugged in the USB cable .. it booted up and shows me the sad icon again :( damn
12:38:59funmankugel: hm you say 3.5 was fine?
12:39:15kugelyes, but udelay came way after 3.5
12:39:32kugelr26163 doesn't change anything for fuzev1?
12:39:33funmanthen the problem could be outside the .c file
12:39:34jitushit man .. i thought I'll try to repair this freaking ipod because its my first one and got it as a appreciation award 6 years back
12:39:45funmankugel: no but r26164 does ^^
12:40:25kugelno, it's definitely older than r26164
12:40:32JdGordonn1s: jitu: looks like they ont have jus the cable so you might need to buy a whole clickwheel
12:40:33JdGordonemail them
12:40:44kugelI seem to remember you tried to fix this problem with that commit?
12:42:38jituthanks JdGordon but I'm in India .. not sure how to get it ship here ..
12:42:46 Quit blairb (Ping timeout: 258 seconds)
12:42:48JdGordonthey might ship
12:42:57funmankugel: yes, the exact problem was a freeze when close()ing in test_codec
12:43:23kugelI thought it could be the same problem
12:43:28funmani have put a timeout in each infinite loop of sd-as3525.c, let's see
12:43:46funmankugel: try to make sd_enable() not disable the clocks?
12:43:55funmanit's what i did before adding this delay
12:44:57jituhumm .. Thanks a ton guys .. gotta go now .. I will be back on this forum soon and will update u guys how I fixed it ;-) (If by any chance I can :D)
12:45:37 Quit shai (Quit: Leaving)
12:45:54kisakfunman: there was a problem with pointing to 2.03.33 for the Fuzev2 on the wiki?
12:46:02 Join shai [0] (
12:46:30funmankisak: yep, mkamsboot 1.2 doesn't support it
12:47:07*kisak bangs his head on the desk
12:47:50kisakwhy does it matter? mkamsboot 1.2 doesn't support the fuzev2 anyway?
12:48:04funmanperhaps we should make another release with the fuzev2/clip+ USB detection fix
12:48:10funmanit does
12:48:37funmanthe only problem is that it always detect USB as connected on some models so it always boots to OF
12:48:38kugelrelease early, release often? :)
12:49:08funmankugel: i wanted to wait USB driver to work :/
12:50:22funman3.6RC builds (r26476) funman/rockbox-3.6RC/"> - should I announce them on the forum only ?
12:51:08kisakfunman: well, fair enough, sorry for the hassle
12:51:14 Join DataGhost [0] (~dataghost@unaffiliated/dataghost)
12:54:20CIA-6New commit by uchida (r26478): tta codec: optimize for arm target. pp502x: ~1% speed up.
12:55:11 Join ucchan [0] (
12:56:50funmanDunno in which forum i should post, I see no RC builds in Announcements
12:57:21 Join guymann [0] (~charlie@
12:57:32funmanand nothing rockbox/rockbox-dev archives
12:57:47JdGordonofficial test builds
12:57:59funman"When a new feature needs user testing in a custom build before going into SVN, a thread will be started here."
12:58:24Mode"#rockbox +o funman" by ChanServ (ChanServ@services.)
12:58:43Topic"3.6 now branched - RC builds on | Please read before speaking: | Please direct offtopic/social chat to #rockbox-community | This channel is logged at | DevCon2010 will be from June 4th to 6th in Ghent," by funman (~fun@rockbox/developer/funman)
12:58:49Mode"#rockbox -o funman" by funman (~fun@rockbox/developer/funman)
12:59:09*funman washes hands - his job's done here
13:02:44 Join Zigtown [0] (
13:02:53 Quit solrize (Ping timeout: 240 seconds)
13:03:20 Quit jgarvey (Quit: Leaving)
13:04:07 Join puetzk [0] (
13:05:26ucchanS_a_i_n_t: May I not commit my yesterday's patch because the change in r26458 ?
13:07:52S_a_i_n_tucchan: I guess it's already taken care of.
13:08:14S_a_i_n_tthe "-j$" was removed, it turns out it was an accidental inclusion
13:08:23S_a_i_n_t"-j4" rather
13:11:04 Join mk429 [0] (
13:14:46ucchan funman: I applied FS #11335 (thumb-friendly patch) on iPod video. and change GCCOPTS="$CCOPTS -mcpu=arm7tdmi -mthumb-interwork" in configure. Then building failure. (PLUGIN_RAM is full).
13:15:00ucchanIs tmy method of the build wrong?
13:16:44 Join solrize [0] (
13:17:13funmanucchan: yes, some files can not be built with -mthumb
13:17:57funmanget the from and change Makefile CC=/path/to/ccache /path/to/arm-elf-gcc into CC=/path/to/ccache /path/to/ /path/to/arm-elf-gcc
13:18:12funmanthen run make clean and make normally
13:18:45ucchanfunman thanks. I try build using our advices.
13:19:13funmanadding only -mthumb-interwork only will just make object files bigger because no thumb code is used, only function 'veneers' for calling to/from thumb
13:23:34Rob2223funman : could i have 1 minute of your time regarding the fuze firmware diff? you just have to take a look at this screenshot:
13:24:59funmani see a 'V02.3.33A', is it present in 2.3.31 ?
13:25:17Rob2223do you think 0x8 and 0x60 is code? cause these values are even for .31 and .33, AND they are even vor .28 and older
13:25:40Rob2223so this could theoretically match
13:25:42funmani don't think, i know
13:26:06funmanuse a disassembler for arm code and you will see it
13:26:10Rob2223but it doesnt look like a opcode
13:26:41funmanhow should an opcode look like?
13:27:14Rob2223i expect it to be longer and not so much 00's
13:27:17Rob2223but i dont know
13:27:56Rob2223but, as sayes, these bytes are even for all the 2 firmwares the fuse accepts and they are even for the 3 FWs that the fuze doesnt accept
13:28:05funmanit's 32 bits and little endian, the difference you see is in the relative address part (the op code starts at 0xB and is: 0xe59ff218)
13:28:31funmantry changing it then, if it bricks your fuzev2 you can still open it
13:28:55funmanyou need to adjust the whole file checksum and the header checksum though
13:29:21funmanFlynDice: ping
13:29:54Rob2223funman: thats what i wanted you to ask, do you think these bytes are code that gets executed?
13:30:04Rob2223i wanted to ask you*
13:30:21funman100% sure
13:30:42Rob2223well then i shouldnt change it blind
13:30:51Rob2223a pity
13:32:01Rob22230x60 is a normal opcode, too?
13:32:11funmanuse a disassembler, i don't know
13:32:17 Join Stummi [0] (
13:32:22jituguys I just uploaded the pic of my screwed ipod mini wheel here .. seems like its broken :-( can someone pls confirm
13:32:36 Join robin0800 [0] (
13:32:37funmanRob2223: there's one in /utils/disassembler/arm
13:32:57Rob2223ah thx for the hint.
13:33:03TorneRob2223: ARM instructions are all four bytes long, for reference
13:33:03Stummihello, will there be a rockbox-version for sansa fuze v2 in any time?
13:33:08TorneRob2223: They don't have seperate opcode/operands
13:33:18TorneRob2223: And the vast majority of them will start with 'e' :)
13:33:27funmanStummi: check
13:34:07Rob2223Torne: thx 4 the hint. is there a specific program entry point?
13:34:10Stummifunman, there I only found a version for fuze withe the note, that this is not compatible to fuze v2.
13:34:15TorneRob2223: the beginning?
13:34:33funmanStummi: read more carefully
13:34:38TorneI don't know anything specifically about the sansa firmwares
13:34:58Stummifunman, argh, my failure.. found it. thanks :D
13:35:30funmanRob2223: entry point is 0x0 after you strip the header : it's also mapped at address 0x0 in the player's ram
13:35:51funmanso the first 32 bytes / 8 instructions are the ARM vectors
13:38:11*Torne feels kinda nerdy for being able to read any of the hexdump.
13:38:49funmanTorne: if you can read 128kB of it, i'll have some questions for you at DevCon ;)
13:39:02Tornefunman: that would take more time ;)
13:44:11 Nick fxb__ is now known as fxb (
13:47:43 Join petur [0] (~petur@rockbox/developer/petur)
13:50:59funmani'm now testing
13:58:13 Quit JdGordon (Changing host)
13:58:13 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
13:58:27***Saving seen data "./dancer.seen"
13:59:57funman17mins recording fine
14:00:34ucchanfunman: The error occur while building thumb version source. error: see
14:00:34 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
14:01:17funmanucchan: you modified Makefile after running configure ?
14:03:31ucchanI try 1) thumb-v2.diff apply. 2) configure add -mthumb 3) executes configure 4) change Makefile. (inserts 5) make
14:04:08funmandon't add -mthumb in configure, will do it when possible
14:04:44funmanit tries to build with -mthumb -mthumb-interworking and if it fails it tries only -mthumb-interworking
14:05:05ucchanfunman thanks. I try again.
14:05:07funmanalso I need to use eabi gcc because my not-eabi libgcc doesn't support interworking
14:06:22 Quit TheSeven (Ping timeout: 265 seconds)
14:06:23funmanif the .c file uses ARM asm building with -mthumb will not work, that's why the script is needed
14:06:43funmanother solution is to make a list of files which can be built with -mthumb
14:07:11funman13:50 < funman> i'm now testing <- test_disk still locks
14:17:29funmanlinux mmci driver doesn't use dma
14:22:06funmanhm i spot a nasty (potential) bug
14:22:43funmanah no, wrong alarm :o
14:23:16funmanZagor_: ping
14:27:34funmancan you bring the e200v1 with JTAG to DevCon?
14:29:24B4gderZagor_: I could bring it if we can meet up somewhere tomorrow or so
14:29:49B4gderif I can find it at home ;-)
14:30:09Rob2223funman: have you solved your non-charging fuze? it can be that the charger wont charge it cause the voltage is below safety treshold
14:30:35funmanRob2223: no problem with my fuze
14:30:40Rob2223ok :)
14:39:16 Quit B4gder (Quit: It is time to say moo)
14:39:35 Quit jitu (Quit: Leaving)
14:42:12 Quit Zigtown (Ping timeout: 260 seconds)
14:45:02 Quit JdGordon (Ping timeout: 265 seconds)
14:45:39CIA-6New commit by jethead71 (r26479): Threading: Split processor support code into respective target files. C files from /target/xxx are included into thread.c because of essential ...
14:45:54 Join JdGordon [0] (~jonno@rockbox/developer/JdGordon)
14:48:36 Quit robin0800 (Remote host closed the connection)
14:51:18kugeljhMikeS: any ETA on the x86 implementation? :)
14:54:06ucchanfunman thumb version Rockbox can build !! very thanks. It succeeded without changing configure. (patch apply, change Makefile, and use arm-elf-eabi).
14:54:40ucchanI try use thumb's rockbox.
14:56:07CIA-6New commit by jethead71 (r26480): Fix PP5002 builds. Somehow an #else went missing along the line.
14:56:42jhMikeSkugel: hrm...7 days (like the TV show)
14:57:11*kugel doesn't get it
14:57:50*jhMikeS neither (other than it involves time travel somehow)
15:02:14ucchanoh!!! thumb's version Rockbox does not boot from iPod Loader2 (
15:03:32FlynDicefunman: pong
15:05:33Stummibtw: the attachment-links doesn't works
15:06:05Stummioh, now it works
15:06:46Stummia few minutes ago they didn't worked
15:09:23ucchanI use the version that doesn't build thumb enabled. sigh...
15:11:16ucchanfunman sorry, I quit soon. Please teach when I login next time if there is a solution.
15:11:26S_a_i_n_tucchan: Not that it is important, but the page you linked to for Loader2 doesn't even exist anymore...
15:11:31S_a_i_n_tit hasn't for some time IIRC
15:12:58ucchan exists. I visit now.
15:13:15ucchanperhaps ) includes.
15:13:39Rob2223"There is currently no text in this page. You can search for this page title in other pages, or search the related logs. "
15:13:43Rob2223i cant see it, too
15:14:02S_a_i_n_tHmm, yes it included the )
15:14:31Rob2223ah lol and for me it included the /
15:14:45S_a_i_n_twhy you'd want to use it is another story :P
15:17:14ucchanS_a_i_n_t: It is easy to start Rockbox and Original Firmware.
15:17:29S_a_i_n_tso's the RB bootloader.
15:17:44S_a_i_n_tbut...that's a flamewar waiting to happen
15:18:13ucchanMoreover, even if the start of Rockbox fails, OF can be started.
15:18:35 Join wincent [0] (
15:19:02n1sif the rb bootloader doesn't allow that, it could be fixed :)
15:19:25S_a_i_n_tdoes it not allow that?
15:19:35 Quit wincent (Changing host)
15:19:35 Join wincent [0] (~wincent@rockbox/developer/wincent)
15:19:37S_a_i_n_teven so "hard reset" anyone?
15:20:17S_a_i_n_tthe new bootloader drops quite nicely to dickmode if anything fails booting RB, which is kinda nice.
15:20:41S_a_i_n_twow, nice twice in one sentence...that was bad.
15:21:03n1salso "dickmode" :P
15:22:07S_a_i_n_tc & s are *kinda* close ;)
15:25:50ucchanIf the person who is using iPodLoader is a little, it might not be big problem.
15:26:17 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
15:26:25funmanFlynDice: i think a problem still lies in sd_enable(false), i'm trying some things. Do you remember if we measured the effect of this function on battery life?
15:26:27S_a_i_n_t"a little"?
15:26:44ucchana little means minority.
15:26:52funmanucchan: kugel and JdGordon have tried PP (arm7tdmi) builds with thumb and it didn't work either
15:27:19funmani have tried on Sansa AMS (arm9tdmi, both armv4t and armv5) and it works
15:27:27funmanI hope i can find the problem at DevCon
15:27:40funmani tried single core build : same problem
15:28:00funmani mean JdGordon tried it, i have no PP here
15:28:54ucchanuum i'm tired. then English sentences are wrong more sorry.
15:28:58n1sis there any gain from thumb on targets with lots of ram?
15:29:14funmann1s: see FS #6734
15:29:58 Join evilnick_B [0] (~0c140464@rockbox/staff/evilnick)
15:30:01funmanif you can test on other devices it would be nice
15:30:16FlynDicefunman: I don't remember anyone measuring that. I think you said that the OF doesn't do that but instead uses the pl180 power saving mode. I'm assuming you're talking about the as3525 v1's?
15:30:24funmani only checked battery life on clipv1 where any difference would be noticeable (it's 8 hours now)
15:30:45funmanFlynDice: yeah, v1 - i would like to get this bug fixed in 3.6
15:31:17FlynDicefunman: what bug are you looking at?
15:31:27funmanSD freeze
15:31:31ranmafunman: Is there anything special to be done to switch to the external SD on Clip+?
15:31:54funmanreproducible with test_disk write&verify freezing when closing the file (before verify)
15:32:03funmanreproducible with recording PCM@96kHz
15:32:10ranmaI've set CCU_IO and GPIOB_DIR and GPIOB bit 5, but I always seem to get a response from the internal SD...
15:32:29funmankugel has another way but it didn't work for me perhaps i'm not fast enough
15:32:38*ranma is trying to get rockbox to boot from external SD...
15:32:48funmanranma: well you are lucky because the SD expert is just here..
15:33:03ucchansorry I quit. i prepare to go to bed.
15:33:13*FlynDice looks around for said expert
15:33:19 Quit ucchan (Quit: Leaving...)
15:33:24*funman holds a mirror to FlynDice
15:33:36funmanranma: you set B5 to 0, right?
15:33:47ranmai.e. I added "CCU_IO |= (1<<2); GPIOB_DIR |= 1<<5; GPIOB_PIN(5) = 0x00;" to sd_init_card()
15:34:13ranma(This is the bootloader)
15:35:20ranmaMaybe the switching between internal and external SD is also broken on this Clip+?
15:35:58ranmaI could try modifying and loading the main rockbox binary instead, but I assume I'd have to enable the SDRAM first?
15:36:10funmanafaik it's alerady enabled
15:36:52funmanranma: what makes you think the response is from internal SD ?
15:37:12n1sfunman: ah, so basically only binsize savings, which i expect to be most interesting for the targets with very little ram
15:37:29ranmaBecause the card doesn't initialize.
15:37:36funmann1s: perhaps we can get different results if we fine-tune which files to build with thumb
15:37:42 Quit TheSeven (Ping timeout: 240 seconds)
15:37:50funmanranma: where's it stuck at?
15:37:52n1sbtw, eabi never needs long calls, the linker fixes up short calls that need to be long with a trampoline (IIRC that's what it's called)
15:38:05ranmawhile (!(card_info[drive].ocr & (1<<31)), it just times out...
15:38:07funmann1s: it was with gcc 4.4.3
15:38:11funman4.4.4 works fine
15:39:00n1smaybe i could do some testing on sansa c200v1, how does it fail currently
15:39:17funmane200v1 just shows black screen, and ipod video too it seems
15:39:57n1swith arm-elf-gcc (4.0.3) or eabi? (4.4.4)
15:40:11kugeln1s: s/trampoline/veneer/ :)
15:40:16ranmaI expect that send_cmd() should error out, but that doesn't happen, even if I remove the card.
15:40:19funmanmy 4.0.3 doesn't have interworking multilib so i only used eabi
15:40:31ranmaSo I'm assuming it still talks to the (broken) internal one.
15:40:34FlynDiceranma: In order to get the uSD you need to make sure it gets selected with a CMD7(SD_SELECT_CARD) and in send_cmd() you give it drive 1 as an arg. I'm not sure if the GPIOB5 is actually necessary.
15:41:12funmanranma: try to fake a drive0 init (set MCI_CLKENA bit for example)
15:41:28funmanFlynDice: but for init you don't have the rca
15:42:55*FlynDice grabs some caffeine & tries to wake up...
15:43:18FlynDiceIs the card not initing?
15:45:52 Quit puetzk (Quit: puetzk)
15:46:02funmanranma: GPIOA_PIN(2) says the µSD is present?
15:47:46ranmaYep: removed: 0xc80b03fc: 00000024; inserted: 0xc80b03fc: 00000020
15:47:54FlynDiceDuring init I think all you need to do to send the cmd to the uSD is use drive 1 as an arg. If you want to transfer data you need to make sure it gets selected at some point. I think that's more clear than what I said before.
15:48:37ranmaFlynDice: Keep in mind that it's the bootloader, so HAVE_MULTIDRIVE is not defined.
15:49:20funmandefine it in the .c
15:49:48ranmaHmm, can try that...
15:50:03FlynDiceah now I'm catching a clue! But I've got to go now. I should be back in a couple of hours.
15:51:57ranmaI'll try loading rockbox.bin directly first, I suppose the bootloader must have already enabled SDRAM at that point...
15:52:59Zagor_anyone wanna build his "own" dap?
15:53:50Zagor_nothing ground-breaking, but a nice portable package and fairly complete
15:54:19 Join panni_ [0] (
15:55:34ranmaI wonder though: In send_cmd() GPIOB_PIN(5) is only updated with the drive value if sd_present(SD_SLOT_AS3525). What happens if the last send_cmd() was to external and then you remove it, then nothing switches back to the internal one?
15:57:16funmanbrowsing µSD and ejecting it, the internal storage still works
15:58:30***Saving seen data "./dancer.seen"
15:59:22ranmaDepending on how the controller handles multiple cards internally, it could be that the internal card failing prevents me from accessing the external one I think.
16:02:08 Part LinusN
16:03:18 Nick CGL is now known as [CGL] (~CGL@
16:07:04ranmaMaybe bitbanging the XPD SD-MCI pins directly would work :)
16:10:15ZambeziAnyone here using Rockbox with a Sandisk Sansa Clip+? I want to buy one to have as a spare incase my iPod Mini descides to give up. Cause I love the slot for an extra memorycard.
16:12:22evilnick_BZambezi: Did you have a question about that target?
16:13:44JdGordonFUCK YEAH! funman: I got clip+ rec mode
16:14:55JdGordoni need a of file
16:15:46JdGordon7min to download from there :(
16:15:56 Join saratoga [0] (~9803c6dd@gateway/web/freenode/x-vdzhfjodojpgzoic)
16:16:16Rob2223i just dl it in 10 secs ?!
16:16:50JdGordonwhich link?
16:16:54Rob2223Zambezi: well then buy one
16:17:22JdGordonfunman: have you got the pic still? I'm pretty sure it was your first guess
16:17:53JdGordoninternational link is currently screwed to .au... so tat would be why so slow
16:18:03JdGordonyep, the red ones did it
16:18:10JdGordonpatience and a very steady hand is needed :)
16:18:16funmanyeah ;)
16:18:36funmancan you upload pic to SansaAMSUnbrick?
16:19:13JdGordonwe cant use that pic?
16:20:03funmani uploaded them in case they are removed from imageshack
16:20:35JdGordonok will do
16:23:52ranmaIs the unbrick bridge connected to some GPIO? *idly wondering*
16:25:09ranmaThe "update button" from the Datasheet presumably?
16:27:04 Join lpereira [0] (
16:28:03JdGordonwoooo!!!! unbricked
16:28:11JdGordonbut a crazy high pitch squeel!
16:28:21FlynDiceranma: Just back for a second, I had this idea I thought may help, define HAVE_MULTIDRIVE and define INTERNAL_AS3525 1 (vs 0) and that may trick it into using the uSD as internal?
16:28:45FlynDiceJdGordon: That's you voice dummy... ;-)
16:28:50ranmaAssuming that's gpioa0 and gpioa4 you could probably abuse those for a bitbanging serial console.
16:29:09pixelmaZagor_: sorry for late reply - my commit fixed the M3 manual. I just wondered why all "following" manuals were affected too
16:29:19JdGordontime to rebuild my clip+ :)
16:29:23Zagor_ok. that should be fixed now. again... :-)
16:30:15Zambezievilnick_B: Well, yeah. How it works with the extra memoryslot, swap card etc. If volume is higher then stock firmware and if radio works good.
16:30:38funmanFlynDice: i try to use select/deselect on v1, anything i miss ? (no µSD inserted so i just hardcoded INTERNAL)
16:30:55ranmaFlynDice: ATM I'm assuming the problem I'm seeing is that the internal SD not initializing somehow also prevents the controller from accessing the external SD completely
16:33:12 Nick YPSY is now known as Ypsy (
16:33:31 Quit S_a_i_n_t (Ping timeout: 276 seconds)
16:33:56Rob2223funman: do i understand AMSUnbrick right, that these players have a reserved area where the 1st half is that what we have in the *.bin firmware file (without the 0x400 header) and the 2nd half is some data, too?
16:34:23funmanRob2223: yeah, the whole file is there (with the header)
16:34:40funmanyou can try update to an older OF this way
16:34:54Rob2223but the 2nd half says in the player (as long he didnt modify it by itself)?
16:35:05Rob2223well, ive ordered a 2nd fuze
16:35:12Rob2223if thats a new one, too
16:35:15Rob2223i will open one
16:35:25 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
16:35:46Rob2223i mean the firmware.bin fle just takes 1/2 of the reserved space, or?
16:36:07funmanyeah, i don't know what's in the 2nd half but the OF uses it (see what i wrote for the fuzev1)
16:36:14Rob2223yeah i know
16:36:18Rob2223i read i mean
16:36:30Rob2223cause of that i think thats some kind of unit specific storage
16:37:47Rob2223funman: if you need to unbrick a V2, can you dump the whole reserved space, if you havent already don? maybe i need it then
16:37:57Rob2223from a working unit
16:38:08funmanif you need it i'll open the fuzev2
16:38:33Rob2223ok, i will decide depended on the 2nd unit i get
16:39:19evilnick_BZambezi: I don't own that player myself, sorry
16:39:44 Join S_a_i_n_t [0] (S_a_i_n_t@
16:41:11funmanFlynDice: got it! in v1 code i need to set MCI_ARG as flags
16:41:54funmani'll make it like v2 code
16:42:03 Join halmi [0] (~netbook@
16:46:34funmandamn deselecting the card still freezes :'(
16:49:55JdGordondoes rbutil support the clip+ yet?
16:54:22pixelmaZagor_: so most manuals should be available tomorrow again?
16:54:31Zagor_pixelma: yes
16:58:30 Quit Zagor_ (Quit: Leaving)
17:05:12 Quit tchan (Quit: WeeChat 0.3.3-dev)
17:06:08 Quit esperegu (Read error: Connection reset by peer)
17:07:22JdGordonthe clip fms is empty...
17:07:37 Quit bieber (Ping timeout: 245 seconds)
17:07:52 Join bieber [0] (
17:08:29funmanworks for me ?
17:10:32JdGordoncompltly empty for me :/
17:12:11JdGordonoh /me is an idiot
17:13:54 Join esperegu [0] (~quassel@
17:14:34 Quit DerPapst (Quit: Leaving.)
17:26:27 Quit bieber (Ping timeout: 264 seconds)
17:28:44Rob2223funman: is there a chance to get from a _working_ unit with a key combo into that recovery mode? the problem is i will loose 2 year warranty, cause of that i try to avoid openting if its not really nessessary
17:29:46 Join flydutch [0] (
17:30:12funmanif there was a way you think we would probably use it and not open our players
17:31:49Rob2223i thought: pressing a key on a working unit maybe sets a 0 or 1 to that pin and it goes in a recovery mode. is the unit doesnt boot up anymore maybe that didnt work and you have to force the voltage by connecting the pins. sorry for asking
17:32:44 Join hebz0rl [0] (
17:32:59 Quit GeekShadow (Quit: The cake is a lie !)
17:34:29funmanother players have a key combo to enter recovery mode but nobody found one yet on the sansa ams
17:34:40Rob2223ill try a few
17:37:05FlynDicefunman: There should be no need to deselect cards with as3525v1 because each card is paired with it's own pl180 controller.
17:37:33 Quit lpereira (Quit: Leaving.)
17:38:13funmanFlynDice: yeah just trying to see why sd_enable(false) cause problems. I was thinking deselecting the card would terminate all communication with the controller if that's the problem
17:39:09TorneRob2223: It's very unlikely; the pins you bridge are just a gap. It's more likely that on development units the gap contains a zero-value resistor so that development can go through rapid reflash cycles, which is then removed from production units
17:39:12funmanthe delay i added in 26164 seemed to help (sometimes), but i removed it in r26291
17:40:35funmanwhile(MCI_STATUS(INTERNAL_AS3525) & ( MCI_CMD_ACTIVE|MCI_TX_ACTIVE|MCI_RX_ACTIVE| MCI_TX_DATA_AVAIL|MCI_RX_DATA_AVAIL)); doesn't help (i expected it but i wouldn't know what else to wait for)
17:41:01Rob2223wif i just had rockbox like little text on the fuze2 but couldnt read it that fast
17:41:29funmanfilm it
17:41:41Rob2223Torne: well if its really only a gap then theres no chance
17:41:48Rob2223funman: must reproduce it first
17:41:50TorneRob2223: well, that's what it looks like from the pcb
17:42:07TorneRob2223: you would have to follow the traces to see where they go, to know for sure
17:42:14FlynDicefunman: If you want to stop all communication with the controller you can issue a CMD0 which resets it back to anidle state. I'm not sure if that's what you're interested in doing though...
17:42:25Torneit is certainly an SMT-resistor-sized gap with two SMT solder splashes though :)
17:42:27Rob2223Torne: couldnt it get in the PCB in a lower layer?
17:43:02TorneRob2223: dunno how many layers the board is :)
17:43:09TorneYou'd probably have to remove the SoC to actually trace it
17:43:20Tornewhich is not really easy to put back afterward ;)
17:43:38Rob2223lol yeah
17:43:59Torneit's worth doing if someone ever gets their hands on a broken one, maybe ;)
17:44:27funmani have a broken clipv1
17:45:22Tornewell if someone is sufficiently bored to yank the chips off and trace all the not-well-known lines ;)
17:45:25 Join bluebrother [0] (
17:45:25 Quit bluebrother (Changing host)
17:45:25 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
17:47:17ranmaAh, I found what I was missing, I have to set CMD_CARD_NO() correctly too
17:48:31FlynDicefunman: Have you tried using the MCI_CLOCK_POWERSAVE (1<<9) bit yet instead of sd_enable(true/false)? I tried that way back when but couldn't get it to work.
17:49:14 Quit bluebroth3r (Ping timeout: 276 seconds)
17:49:31 Quit guymann (Read error: Operation timed out)
17:49:33FlynDiceranma: That should be set with send_cmd argument
17:50:13ranmaYeah and I was missing that one, it was set wrongly and so it was still accessing the internal card.
17:50:24ranmaNow I'm getting somewhere :)
17:50:51funmanjust checked again that disabling sd_enable() fixes the freeze
17:51:04 Join DerPapst [0] (
17:53:08 Join phanboy4 [0] (
17:53:38 Join guymann [0] (~charlie@
17:55:03funmanlinux has a comment which says: "clk |= MCI_CLK_PWRSAVE; This hasn't proven to be worthwhile"
17:56:17funmanand setting this bit gives me Xfer panic
17:56:38funmanlinux suspend/resume code only sets interrupt mask to 0
17:58:34***Saving seen data "./dancer.seen"
18:01:46funmanusing only MCI_CLK_PWRSAVE has the same effect than doing nothing: test_disk works
18:02:35funmanranma: could you measure power consumption on c200v2 with a patch? just in case battery life is halved or so
18:04:41funmani can run battery_bench with this & with just returning but i would like to know before release ;)
18:05:04ranmaHmm, bootloader seems to work now, but rockbox complains "No partition found. Insert USB cable and fix it."
18:05:45funmanhow is the SD formatted?
18:05:56funmanranma: hm did you remove the AMS_OF_SIZE skip for drive 0 ?
18:06:08ranmaIt set AMS_OF_SIZE to 0
18:06:19ranmaSD is "mkdosfs /dev/sdc"
18:06:33funmanno -I ?
18:06:48ranmaIncluding -I :)
18:07:19 Quit hebz0rl (Ping timeout: 240 seconds)
18:07:19ranmaDoes it have to be FAT16 or something?
18:07:47funmanmkfs.vfat -F 32 -I works for me
18:09:27 Quit pamaury (Quit: exit(rand());)
18:11:48 Quit petur (Ping timeout: 245 seconds)
18:14:34funmanranma: i have another (better?) patch, let me test it first
18:21:39 Quit panni_ (Read error: Connection reset by peer)
18:23:25funmanranma: :just disable the clock instead of setting powersave bit (with unrelated diff)
18:24:41 Quit esperegu (Read error: Connection reset by peer)
18:25:58 Join esperegu [0] (~quassel@
18:29:53 Quit kugel (Ping timeout: 276 seconds)
18:36:27funmanjust removing the CGU_IDE bit seems to work too
18:37:47 Join Zigtown [0] (
18:38:22 Join toffe82 [0] (~chatzilla@
18:38:45 Nick Ypsy is now known as YPSY (
18:41:31 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
18:44:03 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
18:49:00kisakso most manuals are missing and this is known and expected for the moment
18:49:16kisakI must have missed the memo
18:49:53 Quit bluebrother (Disconnected by services)
18:49:55S_a_i_n_tthere wasn't really a "memo", I just caught a message from pixelma to Zagor earlier
18:49:56 Join bluebroth3r [0] (~dom@rockbox/developer/bluebrother)
18:50:35kisakdo we keep any archived versions? or do should I hop over to waybackmachine?
18:50:57 Quit Llorean (Read error: Connection reset by peer)
18:51:09S_a_i_n_tIs that the "royal we" ;)
18:51:12 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
18:51:29pixelmaI hoped that he could run an "unscheduled" build round to make them available today :)
18:51:50kisakno, it's the communal peon we
18:52:44kisakmy mistake, plebian
18:53:29 Quit guymann (Ping timeout: 265 seconds)
18:54:17kisakisn't that nice ... wayback machine is down
18:55:03 Join guymann [0] (
18:58:17 Join lpereira [0] (~lucien@
18:58:35funman <- works on fuzev1
18:59:27funmantest_disk is finishing to compare, i'll see if µSD is still fine, try recording for 1 hour .. and commit to branch
19:01:52funmanargl: Test passed. but it froze
19:03:38 Join Jerom [0] (~Jerom@
19:04:16pixelmakisak: unfortunately it had been broken for more than three days and unfortunately we only seem to keep the last three dailies of everything. Otherwise there should be old ones at least on In the past not all mirrors cleaned up as regularly and you could find even older versions there (if you are lucky there might still be one mirror out there that still does)
19:04:32Jeromfunman: Hi, you wrote that it took more than 24 hours to discharge your Fuze
19:04:46JeromHow was your battery level at that time ?
19:04:47funmanv2 yes
19:04:56funmanalmost full iirc
19:05:28JeromAlmost waited for 48 hours here
19:05:45funmanthat was about it
19:05:54gevaertskisak: has the manuals for the last release
19:06:54Jeromok thanks, because I'm tired of waiting ^^
19:08:56 Nick Strife1989 is now known as Strife89 (
19:09:20funmanJerom: i plugged headphones to see if there was some juice
19:09:54Jeromfunman: I still ear some "clics" when I plugged them in
19:10:16funmanthen its' on
19:10:36FlynDicefunman: I've got sd_enable disabled and I'm able to read from sd with the MCI_CLOCK_POWERSAVE bit set but writing is getting panics. I think it was the interaction with sd_enable that prevented using it befor.
19:10:42funmanthermodynamics law say it will discharge
19:11:30funmanFlynDice: if we can disable fully the clock while not using it, perhaps powersave is not worth looking at ? (that's what linux source seesm to say)
19:13:47 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
19:13:54FlynDiceYes , I just saw that your last try froze and was thinking if we can make this simpler with the MCI_CLOCK_POWERSAVE bit there may be less chance to freeze?
19:14:37funmanFlynDice: my last try while with mixing current code with MCI_CLOCK_ENABLE
19:14:49funmanafaict changing MCI_CLOCK_ENABLE bit alone worked
19:15:02FlynDice[10:02] <funman> argl: Test passed. but it froze
19:15:43funmanyeah that one
19:16:03funmanwith the pastie i pasted 2 lines before
19:19:08 Join bieber [0] (
19:19:24FlynDicefunman: So did I misunderstand then, it works fine?
19:19:55funman worked every time i tried it in test_disk(1 or 2 times?)
19:20:54FlynDiceok then, nevermind me then....
19:22:24ranmaHmm, somehow 4bit mode is broken on the microsd slot.
19:22:45 Quit TopyMobile (Ping timeout: 252 seconds)
19:22:57 Join stoffel [0] (
19:25:26 Quit einhirn (Quit: Miranda IM! Smaller, Faster, Easier.
19:26:41 Nick fxb is now known as fxb__ (
19:27:18 Quit DerPapst (Quit: Leaving.)
19:28:09funmankugel: (for the logs) I just noticed we use brightness*10 for the fuzev2 steps but the register setting is in steps of 8 (bits 2:0 unused)
19:28:50ranmaYay, it booted and played mp3 until rb decided to shutdown because the unconnected battery is empty (was powered over usb)
19:29:11funman(stupid rockbox)
19:32:15funmanif i comment out CGU_NAF_CLOCK_ENABLE & CGU_IDE bits in sd_enable() it still freezews
19:33:35 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
19:37:47*r0b- needs to test his change to the AMS FM Radio
19:37:57 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:41:27 Quit stoffel (Ping timeout: 264 seconds)
19:46:34funmanif i comment out everything else, the same
19:46:39 Quit kugel (Ping timeout: 265 seconds)
19:46:52 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:47:47 Join Zagor_ [0] (
19:47:57 Quit Zagor_ (Changing host)
19:47:57 Join Zagor_ [0] (~bjst@rockbox/developer/Zagor)
19:48:14r0b-funman could you impliment the change you suggested?
19:48:49r0b-about moving the ClipPlus to the Fuze v2's init?
19:48:54r0b-for the fmradio
19:48:59funmandoes it work?
19:49:08r0b-i havent tried
19:49:22*r0b- cant seem to get rockbox from the vm env to his Clip Plus
19:50:14r0b-but if you change the lines i will try it and if it does not work you can revert it
19:50:26 Join wincent_balin [0] (
19:51:04 Join kugel_ [0] (
19:51:23 Quit kugel (Disconnected by services)
19:51:26 Nick kugel_ is now known as kugel (
19:51:33 Quit kugel (Changing host)
19:51:33 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:51:52r0b-how can i get my custon rockbox from the vm to the target?
19:51:56 Join DerPapst [0] (
19:52:21funmansee the doc of your vm
19:52:26 Quit wincent (Ping timeout: 276 seconds)
19:53:01r0b-im using the rockbox build environment VM
19:53:16 Quit lpereira (Quit: Leaving.)
19:53:52amiconnWhat I am doing is exporting my home dir from within the VM to the host, then mapping a drive
19:54:02bertrikr0b-, maybe I could do make some test builds for you. What target do you have? Do you have a patch I could apply?
19:54:18funmanFlynDice: works
19:54:29r0b-bertrik its not a patch
19:54:34funmanranma: ^ can you check if this patch gives too much power consumption ?
19:55:14 Quit komputes (Read error: Connection reset by peer)
19:58:38***Saving seen data "./dancer.seen"
20:09:42S_a_i_n_tr0b-: Can't you just search the drive for ""?
20:09:54S_a_i_n_tSeems like the sensible (or easy) thing to do.
20:10:18S_a_i_n_tfind it, right
20:14:04 Nick esperegu is now known as mrHacksaw (~quassel@
20:16:21 Quit guymann (Read error: Operation timed out)
20:17:49bertrikfunman, do you have documentation for the CCU_IO register is the fuzev2 (and the clip+)? Bit 12 is being set in the fmradio i2c driver
20:18:09funmanbertrik: no, the as3543 only covers i2c registers
20:18:21funmanit's also used in buttons & lcd i think
20:18:51 Join kugel_ [0] (
20:19:16*S_a_i_n_t wonders why "make voice" isn't smart enough to do "nothing to be done for 'all'"
20:19:49 Join guymann [0] (
20:20:21 Join Llorean [0] (
20:20:23 Quit Llorean (Changing host)
20:20:23 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
20:22:05 Quit kugel (Ping timeout: 248 seconds)
20:22:50 Quit S_a_i_n_t ()
20:23:05 Join S_a_i_n_t [0] (S_a_i_n_t@
20:23:06 Join komputes [0] (~komputes@ubuntu/member/komputes)
20:23:48 Quit kugel_ (Ping timeout: 264 seconds)
20:23:58 Quit flydutch (Quit: /* empty */)
20:26:16 Quit Zarggg (Quit: Zarggg)
20:27:17r0b-0005 9000 F080 0104
20:27:23r0b-thats the regs for this fm chip
20:27:37 Join petur [0] (~petur@rockbox/developer/petur)
20:28:23funmandoes it work?
20:28:39funmanthen that's not the regs
20:28:47 Join Horscht [0] (~Horscht2@xbmc/user/horscht)
20:28:59funmanfirst there are more and then they should start with '1242'
20:29:31funmantry power off/on and see if that changes completely
20:30:32r0b-the first set of digits of each set 1st0005 2nd00F1 3rd5594 4th5803
20:30:46r0b-the first set of digits for the 4 different rows
20:30:55CIA-6New commit by funman (r26481): as3525v1: do not disable clocks in sd_enable() ...
20:31:35 Join hebz0rl [0] (
20:32:11 Join TexasRockbox [0] (
20:32:40funmanbattery bench takes 14hours or so, i can have results tomorrow
20:33:00bertrikfunman, I wonder if maybe the clip+ also needs the CCU_IO thing (bit12) that the fuzev2 does
20:33:17TexasRockboxHello, are the new changes to the theme code ready to use in the current build?
20:33:17funmanbertrik: i think that's what r0b- is trying
20:33:53bertrikhe's using binaries that I'm building for him, this one just enabled the fm debug values (and slowed the i2c clock down, just to be sure)
20:35:23*r0b- loves the microsd slot on his clip +
20:35:27AlexPS_a_i_n_t: He is using a VM
20:35:33 Join stoffel [0] (
20:35:34r0b-makes uploading firmware easy!
20:37:23TexasRockboxIs it possible to disable Rockbox USB with the current build (iPod 5.5g 64MB) or is a custom build required?
20:37:48 Join lis6502 [0] (~Owner@
20:37:52lis6502hi y'all
20:38:05lis6502i just like to say big THANK you for Rockbox.
20:38:37lis6502now my sansafuzev2 has new life, and it's much more usable than on sandisk's firmware :)
20:38:59 Join lpereira [0] (
20:39:22lis6502if there's something to test (like new plugins, codecs or something) i'll gladly help
20:39:26AlexPTexasRockbox: You need to alter source (slightly), then compile
20:39:56lis6502btw, question to developers: is module (*tracker) plugin based on mikmor
20:40:00lis6502d* ?
20:40:28TexasRockboxOK, thank you.
20:40:55TexasRockboxAre the theme code changes ready to use in the current build?
20:41:16ranmafunman: No VIC glitch on the Clip+ so far :/ (I was kind of hoping to get one :))
20:41:33funmanranma: which method did you use?
20:42:40TexasRockboxOK, thank you again! I've been seeing all the changes but wasn't sure if it was ready to use. The sticky in the forum seemed to make me think the changes were going to be available last weekend.
20:43:33 Join Luca_S [0] (
20:43:37ranmaI'm using r26412 with the patch reverted (so it's directly writing to VIC_EN_CLEAR), on the 8th Song or so now...
20:44:31 Quit TexasRockbox (Quit: CGI:IRC (EOF))
20:44:45funmani didn't try on clip+ but i can test it
20:45:16ranmaAlso tried toggling between pcm_lock and pcm_unlock in the debug menu between screen refreshes, but that also didn't trigger it.
20:47:13funmanat very high rate?
20:48:21ranmaOr more if I hold a key down (up to max key repeat rate, I guess HZ/30?)
20:48:47CIA-6New commit by alex (r26482): Backport r26461 to 3.6 branch.
20:49:04 Join kugel_ [0] (
20:49:17bertrikare we going to release soon?
20:50:11 Quit kugel_ (Read error: Connection reset by peer)
20:50:31 Join kugel [0] (
20:50:47 Quit kugel (Changing host)
20:50:47 Join kugel [0] (~kugel@rockbox/developer/kugel)
20:51:12funmanbertrik: tomorrow?
20:52:17 Join kugel_ [0] (
20:52:24bertrikI guess so, I think the idea was to release before devcon :)
20:52:38 Quit kugel (Disconnected by services)
20:52:39bertrikThere's no blockers left for a release, right?
20:52:42 Nick kugel_ is now known as kugel (
20:52:46 Quit kugel (Changing host)
20:52:46 Join kugel [0] (~kugel@rockbox/developer/kugel)
20:53:30 Quit mrHacksaw (Ping timeout: 252 seconds)
20:53:32funmani would like to sort out if fuzev1/e200v2 battery life takes a big hit or not
20:53:54funmani should know tomorrow morning
20:54:42 Join moparx [0] (~moparx@unaffiliated/moparx)
20:54:58 Quit stoffel (Remote host closed the connection)
20:57:57lis6502about rc builds
20:58:07funmanranma: perhaps jtag could help with r11297
20:58:16lis6502rockbox for Fuzev1 will not work with v2 ?
20:58:26funmanno it will not
20:58:47CIA-6New commit by alex (r26483): Update Spanish ...
20:58:53pixelmaAlexP: thanks
20:59:11AlexPpixelma: no worries :)
20:59:13pixelmaalthough... you broke it ;)
20:59:30AlexPFully willing to admit that :)
20:59:49AlexPAlthough I thought I copied what existed (clearly not well enough) :)
21:01:45pixelmaI haven't checked whether other remote target manuals are broken too though
21:03:12CIA-6New commit by alex (r26484): Update Spanish ...
21:03:27AlexPpixelma: I tried the other two (I think)
21:03:48AlexPI built quite a lot of them before I committed, but evidently not the iaudio ones
21:04:27AlexPyeah, they are fine
21:05:17pixelmaH100/300 and Gigabeats?
21:06:57pixelmaAlexP: only the M3 defines HAVEREMOTEKEYMAP, the other Iaudios don't currently but they could...
21:07:16ranmaMaybe, but I think I should fix the jtag connector first, the ground wire has come lose :/
21:07:21funmanbertrik: clip+ OF only sets bit 2 of CCU_IO
21:07:31AlexPpixelma: Probably why - I built x5 as a check, but that obviously didn't bring it up
21:08:20pixelmawhen it was added to the M3 the button mapping diffeered between the other too and a unification was pending
21:08:34AlexPah, OK
21:08:36pixelmawhich happened quite some time ago
21:08:42bertrikfunman, hm, that's in xpd_func_sel according to the as3525 datasheet, so probably the CCU_IO has a different mapping on the clip+
21:09:41AlexPpixelma: Do the others need adjusting to add the remote?
21:09:41 Quit Tuplis (Ping timeout: 260 seconds)
21:09:45bertrikfunman, when / where does the OF set bit 2 of CCU_IO?
21:10:06funmanbertrik: bit 2 is set in SD driver (dunno if it's related)
21:10:18ranmaBuilding C200v2 for testing of commit 26481 now...
21:10:19funmansearch for C810000C gives 6AFC
21:10:46pixelmaAlexP: that's what I ask myself too... I don't think so but am not a 100% sure
21:11:23pixelmatrying if it builds should be enough though
21:12:28 Join MethoS- [0] (~clemens@
21:12:46CIA-6New commit by alex (r26485): Update Chinese (Simple) translation ...
21:13:02 Join esperegu [0] (~quassel@
21:13:09 Quit lis6502 (Quit: Leaving)
21:13:17pixelmaAlexP: feel free if you want to try
21:13:28bertrikranma, you have no problems with fm radio on your clip+, right?
21:13:34CIA-6New commit by alex (r26486): Update Chinese (Simple) translation ...
21:13:38funmani have no problems either
21:14:59 Quit kugel (Ping timeout: 276 seconds)
21:15:01funmani see that cpu/pclk freq seems to be changed in FM code
21:15:11funmani had seen that on other models but ignored it
21:15:25 Part watto
21:16:14bertrikhuh, really, where? I don't see why FM code would do that?
21:16:23AlexPpixelma: I'll see if it is just a case of enabling it, and check if it builds. I can't check if it is correct though (tonight anyway)
21:16:37funmanbertrik: in clip+ 1.02.09
21:16:48funman2A1A8 calls 1C48
21:17:01ranmabertrik: Didn't try, can do that tomorrow
21:17:56ranmaThat was the first time I got the Clip+ to boot and play music today after all :)
21:18:53 Quit Luca_S (Quit: CGI:IRC)
21:22:18 Join liar [0] (
21:23:07 Join obo [0] (~obo@rockbox/developer/obo)
21:23:56funmani start benchmark on fuzev1
21:24:24CIA-6New commit by alex (r26487): Update Slovenian (Slovenscina) translation ...
21:24:43 Quit komputes (Ping timeout: 245 seconds)
21:25:28 Join evilnick_B_ [0] (~0c140464@rockbox/staff/evilnick)
21:25:34CIA-6New commit by alex (r26488): Update Slovenian (Slovenscina) translation ...
21:26:03 Join komputes [0] (~komputes@ubuntu/member/komputes)
21:26:50 Quit evilnick_B (Ping timeout: 252 seconds)
21:29:16 Join Tuplis [0] (
21:30:33 Quit komputes (Ping timeout: 245 seconds)
21:31:44 Quit wincent_balin (Changing host)
21:31:44 Join wincent_balin [0] (~wincent@rockbox/developer/wincent)
21:33:35wincent_balinI just made some noise on the PD-anywhere mailing list and it seems the maintainer is interested in integrating changes done by pdbox back into the PD mobile fork of Pure Data.
21:36:16 Join komputes [0] (~komputes@ubuntu/member/komputes)
21:37:55 Nick fxb__ is now known as fxb (
21:38:22 Quit obo (Remote host closed the connection)
21:38:33 Join obo [0] (~obo@rockbox/developer/obo)
21:42:59saratogawincent_balin: does PD work on all rockbox targets or just some?
21:43:54 Quit komputes (Excess Flood)
21:43:59wincent_balinsaratoga: For now it does on H300 (tested by myself) and presumably on H100.
21:44:11saratogawhats hardware specific about it?
21:44:28wincent_balinsaratoga: GUI controls.
21:44:55wincent_balinNext target I will test it on is Sandisk e200.
21:44:59funmani tried to add a keymap for fuze but i'm not sure it worked
21:45:26wincent_balinpdbox needs mre than 2 MB of RAM.
21:45:34wincent_balin... more ...
21:46:22wincent_balinBut I did not benchmark to which extent this capacity is used.
21:46:51wincent_balinfunman: Now that I corrected some silly bugs you may tiy again.
21:47:07wincent_balin... try ...
21:47:27funmani remember also that the screen was too small
21:47:50funmanbut then it's the same than h300
21:49:44wincent_balinfunman: As I do not have a target other than a H320 I would like to ask you, if you have time and will for it, to test pdbox on fuze and to tell me about the bugs and whether they occure on target only or in the simulator too.
21:51:09funmanyou have some .pd files?
21:55:05 Quit dfkt (Quit: -= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.)
21:57:13 Join stripwax [0] (
21:57:55 Nick evilnick_B_ is now known as evilnick_B (~0c140464@rockbox/staff/evilnick)
21:58:39***Saving seen data "./dancer.seen"
22:00:07 Join DomasoFan [0] (
22:00:27 Quit esperegu (Read error: Connection reset by peer)
22:03:28 Join stephen [0] (~stephen@
22:04:02 Nick stephen is now known as Guest86259 (~stephen@
22:04:56 Join esperegu [0] (~quassel@
22:07:34wincent_balinfunman: You might want to download the file from
22:09:03 Nick Guest86259 is now known as Stephenc____ (~stephen@
22:09:58DomasoFanhi guys. is there one who can make me a mkamsboot from svn for win32? i just have cygwin here and it won't compile here.
22:13:20funmanwincent_balin: seems to work
22:13:53funman2 bugs: bpm & pitch are uninitialized in l8p.pd, i hear cracks when switching frequency in test.pd
22:15:46CIA-6New commit by funman (r26489): pdbox: add fuze keymap
22:20:47funmanAlexP: Slovenian update caused yellow
22:21:07 Quit leavittx (Ping timeout: 240 seconds)
22:24:07AlexPhmmm, I wonder why he removed those translations
22:24:28bluebroth3rbieber: any thoughs on making the theme editor build put the intermediate files into a subdir?
22:24:59bieberYou mean the .o's and such?
22:25:14bluebroth3ryes. Basically the same as done for Rockbox Utility:
22:25:25AlexPah, I see
22:25:42bluebroth3r(the code is basically copied from :)
22:26:05funmanAlexP: some people have been removed from CRDITS also
22:26:13funmanperhaps he merged from an outdated tree
22:26:16AlexPdo what
22:26:25AlexPOK, give me a bit :)
22:26:48funmanranma: clip+ crashes in 5 minutes with r26489, r26316+r26444 reverted
22:27:39 Quit stripwax (Read error: Connection reset by peer)
22:27:59bieberOh, I didn't realize you could do that
22:28:15 Join kugel [0] (~kugel@rockbox/developer/kugel)
22:28:30funmanbluebroth3r: osx has no ccache ? :/
22:28:55bluebroth3rbieber: another nice thing is the option -config silent with qmake. Works fine at least on Linux, has been an issue in the past with MinGW
22:29:29bluebroth3rfunman: haven't checked, and haven't updated that. It's been that way since ages :)
22:30:00bluebroth3rthough checking for ccache on Mac shouldn't hurt, even if XCode doesn't ship it one can always install it via macports
22:30:09CIA-6New commit by alex (r26490): Restore people accidently removed from CREDITS in r26488
22:31:07bieberHow do you get mingw to static link its runtime, btw?
22:31:24CIA-6New commit by alex (r26491): Restore people accidently removed from credits in r26487
22:31:26bieberI finally got Qt compiled in a Windows VM to do static linking, but apparently the executable still needs a MingW DLL
22:31:46bluebroth3rfunman: just checked, Mac comes without ccache. It's in MacPorts, though.
22:31:55funmanbieber: you can cross compile qt apps from linux although it's a bit of a PITA
22:32:04bluebroth3rbieber: you need to use -no-exceptions when building Qt
22:32:25 Join Zarggg [0] (
22:32:42funmanvlc is built from linux with - although this build only has core & gui afaik
22:32:49bluebroth3rsee the RockboxUtilityDevelopment wiki page, it lists how to build Qt for static linking :)
22:33:27*bluebroth3r thinks its much easier to simply build in a VM running Windows
22:33:53bieberI got that far, it's just MingW that's my problem
22:33:57bieberI'll take a look at the Wiki page
22:34:02bluebroth3rbieber: which version of gcc did you use?
22:34:35bluebroth3ras far as I've understood from the MinGW ML there seem to be changes that could eliminate the dependency on mingwm10.dll with gcc 4.5. Haven't looked into that though.
22:34:46 Quit Zigtown (Read error: Connection reset by peer)
22:35:37bieberOh, I think I got 4.4
22:35:41 Quit guymann (Read error: Connection reset by peer)
22:36:31CIA-6New commit by bieber (r26492): Changed build subdirectory
22:36:35 Part DomasoFan ("I'm a happy Miranda IM user! Get it here:")
22:37:18bieberAlso, is there a way to temporarily redirect stderr from within a program?
22:37:53bluebroth3rfoo 2> file
22:38:06 Join anewuser [0] (anewuser@unaffiliated/anewuser)
22:38:28bluebroth3rbieber: afaics you last commit contained changes unrelated to the commit comment. Please try to avoid such commits, they are confusing :)
22:38:45bluebroth3r(at least it confused me for a second right now)
22:39:45bieberOh, sorry
22:39:49bieberI thought I'd already committed that
22:40:15bluebroth3rno problem. Just wanted to be sure this wasn't intentional ;-)
22:40:50bluebroth3r(I've seen too many sloppy commits at work, sometimes even causing major breakage)
22:42:22funmanI think git useful for that because I can check my commits after committing (before sending them to the server)
22:42:58 Quit bertrik (Read error: No route to host)
22:44:04bluebroth3rfunman: definitely
22:44:09 Join guymann [0] (~charlie@
22:44:24bieberOh, that reminds me: are there any serious issues with using Git instead of SVN to commit to the repository?
22:44:49funmanno, the only thing missing is setting svn properties, but personally i don't mind
22:45:23bieberOkay, I'll have to give that a try
22:45:32AlexPFor the lang yellow, he had included a few untranslated items in there - if I remove the whole item that should just "mark" the as untranslated again, right?
22:45:43bieberI worked a couple hours without Internet access the other day, and I was wishing I could do local commits :/
22:47:03*AlexP looks around for rasher
22:47:17bluebroth3rbieber: I'm using git-svn since quite a while now without any problems. I have a svn checkout to fix properties in case I need to. Properties can get some kind of an issue when dealing with binary files, but at least for the code base that's no issue.
22:49:55 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
22:50:34bieberDo we have a wiki page with info on the structure of an _entire_ theme, not just the WPS file?
22:50:47bieberFrom what I've gathered so far, it's just a cfg file, a wps file, and some images?
22:51:27bluebroth3rcheck ThemeGuidelines −− no idea if it's up to date though.
22:51:57bluebroth3rbasically, yes. You can include a rwps for targets that have a remote too
22:52:02bluebroth3rplus fonts, if needed.
22:52:25 Quit ender` (Quit: Where there's a will, there's an inheritance tax.)
22:52:31bluebroth3rno idea about that sbs stuff. Guess you can have that in the theme too.
22:53:19 Quit n1s (Quit: Lmnar)
22:53:46 Join ender` [0] (
22:55:09 Quit phanboy4 (Ping timeout: 245 seconds)
22:55:21bieberI think sbs is the status bar?
22:55:36bieberAnd what do you mean by a remote? Like a physical device with a screen?
22:55:58bluebroth3ryes, some devices have a physical remote. For example the h100 series.
22:57:01AlexPalso the fms
22:57:06AlexP(fm radio screen)
22:57:39AlexPso wps, rwps, fms, rfms, sbs, rsbs, cfg
22:57:46wincent_balinfunman: BPM/pitch - I am looking into it. Cracks when switching frequency seem to be a sad reality.
22:57:56AlexPAnd only certain settings are allowed in the cfg
22:58:46bieberWow, I had no idea such a thing existed
22:59:30AlexPThe iaudios have them too
22:59:39bluebroth3rthose are the three remotes available for the h100 / h300 series
22:59:49AlexPbieber: In fact, the M3 doesn't have a screen on the main unit, just the remote lcd
22:59:55bluebroth3rthere's even the m3 which doesn't have a screen :)
23:00:02*AlexP wins!
23:00:20*bluebroth3r feels bad
23:00:26AlexPsorry :)
23:01:26pixelmathe remote on the M3 is treated like the main screen there though
23:01:59pixelmabut the same remote can be used on the other Iaudios
23:03:36*AlexP wants to unyellow the build table, but would like to check a lang question first
23:03:50pixelmaweirdest combinations are IMO - the M5 with the remote (160x128x2 main, 128x96x2 remote) and the MRobe500 (640x480x16 main, 79x16x1 remote)
23:04:15bieberfms is for radio tuning?
23:04:31AlexPit is like the wps, but for when the radio is playing
23:04:47bieberOkay. But the tags and everything are all the same for every screen?
23:05:10bieberErr, I mean as far as the basic semantics go
23:05:21AlexPbieber: You might want to look at
23:05:23bieberViewports, sublines, conditionals and such. I'm assuming there will be different specific tags
23:05:24pixelmaAlexP: if the phrases didn't exist before, you could safely remove them again I think. They'll show up as missing in genlang again
23:05:31AlexPbieber: yes
23:05:36AlexPpixelma: That's what I thought
23:06:18AlexPpixelma: There are 3 that did exist before, but I think I'll remove them just in case - I assume they were removed by accident, but it might have been deliberate
23:06:33AlexPSo if I remove them, it'll get flagged for him to double check
23:07:02AlexPbieber: Sorry, - but yes, the format is the same
23:07:42wincent_balinfunman: Could send me a diff of the keymap for fuze?
23:07:48pixelmawhich reminds me.... petur: last time I looked, nederlands.lang contained "### note to the translator". Was that on purpose?
23:08:28pixelmait doesn't actually say "note to", I was just too lazy to look it up ;)
23:11:36CIA-6New commit by alex (r26493): Fix yellow, remove some untranslated strings from Slovenscina
23:12:31funmanwincent_balin: it's in svn ;)
23:13:06wincent_balinThe compilation of pdbox for fuze too?
23:13:07funmanhowever the keymap doesn't go fine with the wheel test .pd
23:13:40funmani made the keymap from the keys description but center/up/down aren't center/up/down on the scrollwheel like GUI
23:13:59CIA-6New commit by alex (r26494): Fix yellow, remove some untranslated strings from Slovenscina
23:14:08peturpixelma: remind me again on devcon
23:14:19 Quit petur (Quit: Zzzz)
23:15:03bieberProbably a stupid question: Do all the Rockbox targets only tune FM radio, or is fm just short for fm/am?
23:15:07pixelmapity if it can't get fixed till release but...
23:15:41AlexPbieber: fm only at this point
23:15:56pixelmabieber: FM and some D2s have DAB but I doubt it's supported
23:16:20pixelmathe Logik DAX has a DAB tuner too ;)
23:16:29wincent_balinfunman: I ponder whether I should change the standard layout of pdbox. Currently it is the one I got from PdPod, which was running on iPod 3G.
23:16:53funmanah this is why the fm.pod looks like a ipod wheel :)
23:17:56wincent_balinfunman: :-)
23:18:26funmani think it would be a good idea, especially if we add more targets
23:20:14wincent_balinfunman: I looked into pdbox.h and SUBDIRS, but pdbox is compiled for H100 and H300 only.
23:21:00funmani added CONFIG_KEYPAD == SANSA_FUZE_PAD condition in apps/plugins/SUBDIRS (r26489)
23:22:18wincent_balinfunman: And it is in the SVN?
23:23:24wincent_balinfunman: Ok, now I see. My fault −− forgot to svn update.
23:23:42funmansvn update a day, keeps the bugs away
23:24:58 Quit Stephenc____ (Quit: Leaving)
23:26:04wincent_balinfunman: Another question: Do you hear the cracks (pdpod_test.pd) in the sim or on target?
23:26:35funmanTorne: you remember where the default startup plugin can be activated? grep PLUGIN apps/ turns nothing
23:27:03kugelfunman: you mean at boot?
23:28:38kugellook at main.c
23:29:02wincent_balinfunman: Hmm... pdpod_test.pd is more or less an oscillator connected to DAC. Cracks mean there is something wrong with index of the cosine table.
23:29:27wincent_balinfunman: Fuze is an ARM target, isn't it?
23:29:48bluebroth3rfunman: look for AUTOROCK
23:30:11funmanwincent_balin: yes it's arm9tdmi, i tested on fuzev2 (armv5)
23:30:23funmanbluebroth3r: thanks ;)
23:30:27 Quit DataGhost (Ping timeout: 265 seconds)
23:31:02wincent_balinfunman: Does ARM have a signed/unsigned peculiarities? And do the cracks occure in the simulator too?
23:35:24 Join intrados [0] (
23:35:47funmanyes but much less
23:36:31funmani don't know ARM arithmetic but i don't know of any specific thing about sign
23:36:41 Quit GeekShadow (Quit: The cake is a lie !)
23:40:47wincent_balinfunman: How high is the HZ rate on Fuze?
23:41:02 Quit lpereira (Quit: Leaving.)
23:41:04funman100 - i think it's the same on evrery target
23:41:35funmanwincent_balin: look in firmware/export/kernel.h
23:42:40wincent_balinfunman: Indeed... Hmm...
23:43:30 Quit hebz0rl (Ping timeout: 265 seconds)
23:43:38wincent_balinfunman: On H320 I get the cracks too. Will look into it later.
23:44:08 Join Buschel [0] (
23:45:49Buschelas you are talking of pdbox and cos_table -> will anyone change the formatting of the cos_table[] declaration in a way to not declare this multi-10k array in one single line?
23:46:15*Buschel talks about cos_table.c and cos_table.h in the pdbox folder
23:48:55wincent_balinI hope so.
23:49:43wincent_balin.c file - certainly, .h may have to remain like this because of compatibility.
23:51:04bluebroth3rwhy because of compatibility?
23:52:07Buschelthe declaration in the .h file is not used when ROCKBOX is defined and could be removed...
23:52:46bluebroth3rregardless of that I don't see why whitespace changes would not be permitted because of compatibility reasons.
23:53:02 Quit [CGL] (Remote host closed the connection)
23:54:57bluebroth3rwow, that line is really extreme
23:55:02wincent_balinOk, then I'll change that.
23:55:25CIA-6New commit by bluebrother (r26495): Move constructing the themes info download link to rbutil.ini. ...
23:55:39*bluebroth3r never looked at that before
23:56:44*Buschel 's editor is always complaining about those 2 files
23:57:26Buschelbtw, the cos-tab has 4 times the size it would need. we could save a lot of memory (if needed)
23:58:09 Join komputes [0] (~komputes@ubuntu/member/komputes)
23:58:40***Saving seen data "./dancer.seen"

Previous day | Next day