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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2007-08-28

00:00:57EnterUserNamelazka: do you have problems with charging?
00:04:03 Join dirkson [0] (
00:04:44dirksonHeyo all. I've just installed rockbox on an ipod mini.... Is it supposed to come with an mp3 codec? I get errors on every. single. mp3.
00:05:08dirksonMostly "No codec for this file" errors, with the occasional generic "codec error" mixed in the pot.
00:05:46lazkaEnterUserName: no, not really.
00:05:46saratogadirkson: that generally means the install didn't work
00:06:14EnterUserNamelooks like i gotta return it. I tried it on my macbook/pc/work computer all not charging fully
00:06:16dirksonsaratoga, Ah! Well, good to know... Umm... "install"? You mean, "Copy this directory over and configure your bootloader"?
00:06:20EnterUserNamethans lazka :)
00:06:55saratogadirkson: yes, i assume you're using today's SVN build the bootloader from the manual?
00:07:01***Saving seen data "./dancer.seen"
00:07:23 Quit jgarvey ("Leaving")
00:07:33dirksonsaratoga, I've been using loader2 from ipodlinux . (Another firmware that I've been having trouble making play mp3s >.< )
00:07:54saratogadirkson: well it sounds like you don't have something setup right
00:08:03saratogai'd recommend trying the rockbox bootloader and seeing if it works
00:08:15pixelmaand those are not files you bought from itunes or the like?
00:08:45dirksonsaratoga, I attempted to rename the rockbox folder to cover up an earlier failed install. Let me tinker with that- perhaps some sage-like programmer has coded in a hard value.
00:08:50linuxstbdirkson: Do you have a ".rockbox" (note the dot) folder in the root of your ipod's FAT32 partition? i.e. the same place as "Notes", "Calander", "iPod_Control" etc
00:09:14linuxstbInside .rockbox should be a directory called "codecs", and in there, an "mpa.codec"
00:09:15dirksonsaratoga, linuxstb This confirms my suspicion : ) I think I know how to fix this now. *tinker*
00:11:02dirksonP.S. Is there any way to stop rockbox from switching into usbdisk mode every time I plug it in?
00:12:16dirksonHahaha! For the first time since I got it, my mp3 player is successfully playing an mp3 ^.^ (Yes, yes I did switch to open source before even /testing/ the proprietary software ;) )
00:12:29dirksonThanks guys : )
00:13:09linuxstbdirkson: Hold MENU as you insert the cable (and maybe for a couple of seconds afterwards)
00:15:16dirksonlinuxstb, You're awesome...... Except that it just did it anyway
00:21:45 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
00:24:22BigBambidirkson: You may have to hold it for up to 30 seconds after inserting usb. There is a slight bug atm
00:25:39dirkson Ok
00:25:56dirksonAlso: Does rockbox support any other filesystem than fat32?
00:26:33dirksonsaratoga, That sucks. What're we running on top of, some variant of uclinux, or something else?
00:27:04dirksonrasher, Proprietary open source. Now I've seen it all. Thanks! Got to run.
00:27:08 Quit dirkson (Remote closed the connection)
00:27:18rasherThat did not make any sense at all.
00:29:01BigBambiI can't see anyway of interpreting that as proprietry
00:30:32JdGordonhe obviously thinks linux is the only non proprietry OS out there...
00:31:06 Part toffe82
00:32:56BigBambiespecially given propritery open source is surely a contradiction in terms
00:33:18*BigBambi is dismayed by his spelling
00:42:41 Quit Febs (Read error: 110 (Connection timed out))
00:44:32 Quit seablue ("life, death, life, death")
00:49:39 Quit JdGordon ("Konversation terminated!")
00:54:18 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
00:56:45Soapand his apparent sarcastic assertion that "a sage-like programmer has coded in a hard value." ie. the naming of the .rockbox directory.
00:58:32rasherNot sure what he was expectig
00:58:34 Quit qwm (Remote closed the connection)
00:58:37 Join qwm [0] (
00:59:39Soapexpecting a dope-slap. (well, should have been) ;)
01:00:45 Quit ender` (" (Coder's Ditty) 100 little bugs in the code, 100 bugs in the code, fix one bug, compile it again, 102 little bugs in the cod)
01:06:44 Quit jhulst (Remote closed the connection)
01:08:31 Join My_Sic [0] (
01:16:05nerochiarosaratoga: hi, are you around ?
01:21:18 Quit obo ("bye")
01:23:06saratoganerochiaro: am now
01:23:08saratogawhats up
01:23:41saratogaalso, been meaning to ask you, is the libasf in that zip i linked you the current libasf?
01:23:56nerochiarosaratoga: i was just wondering if you had any chance of trying the wma samples that crashed the rockbox simulator for me ?
01:24:09nerochiaroas for the libasf in the zip, i have not checked if it was the most recent one
01:24:12 Join HellDragon_ [0] (n=Nocebo@unaffiliated/helldragon)
01:24:35 Quit HellDragon (Nick collision from services.)
01:25:15saratoganerochiaro: yes one of them causes a nasty overflow error in the LSP decoding stuff
01:25:20saratogai didn't figure out a fix for it
01:26:00saratogai've been working on a problem with the changes made that enable us to decode one frame at a time
01:26:27saratogamost of the problem samples i have left seem to have that problem, or else something wrong with LSP if they've very low bitrate files
01:26:38saratogaonce i get it figured out, it might solve a few of your problems
01:27:05saratogaunfortunately, i'm back in school so my rockbox time is more limited
01:27:31saratogaand anyway, most of the remaining issues seem quite rare. so i'm not too concerned
01:27:53saratogai mostly want to fix the superframe bug i mentioned, and then get proper seeking in rockbox (though that shouldn't impact you)
01:28:42nerochiarosaratoga: isn't the rockbox decoder alrady decoding one frame at a time ?
01:29:02saratogayes, but on very rare files, the changes made to allow that seem to corrupt frames
01:29:41saratogathe problem does not impact my older version of the codec, so i can only assume thats the fault
01:29:57saratogaunfortunately, the problem has proven extremely difficult to track down
01:29:58nerochiarosaratoga: corrupt as in causing artifacts in the playback ? i have a couple of files that have bursts of static in just a couple places in the whole file
01:30:09saratogayes thats typically it
01:30:22saratogaVLC decoding will return nonsense on those files, or sometimes crash entirely
01:30:39nerochiaroi didn't have crashes so far due to that
01:30:45saratogaif the bitrate is over 32 kbps, this is the only remaining bug i am aware of
01:30:55saratogathough there may be others
01:30:58nerochiarohowever i'm still puzzling as to why i have that vlc overflowing problem that i mentioned earlier
01:31:05nerochiaroon the vlc buffers
01:31:07saratogai thought you resolved that?
01:31:49nerochiaroby increasing the buffers unfortunately
01:32:01saratogathat actually worked?
01:32:09nerochiaroit works, but i recognize it's ugly and probably wrong
01:32:30saratogawhat did you increase them to?
01:33:36nerochiarovlcbuf1 and 2, increased to 5100
01:35:01 Part pixelma
01:35:35nerochiaroVLC_TYPE vlcbuf1[5100][2];
01:36:38saratogathats very strange
01:37:00saratogais that the minimium size you could use, or just a larger one picked at random?
01:37:52 Join phalax [0] (
01:38:34nerochiaroit's actually the minimum, as far as i remember testing it. if you look at the original, it's VLC_TYPE vlcbuf1[2550][2]; i think, thus i doubled the buffer size
01:38:49 Quit phalax (Client Quit)
01:39:13saratogahave you tried checking sizeof(VLC_TYPE) just to be sure?
01:39:46saratogacould be one of the struct members is acidently typed differently on your system
01:39:54nerochiaroactually no, it occourred to me just as a few minutes ago when i told you it was double the original size
01:40:10nerochiaroi'll check that
01:40:26nerochiarowhat should it be the proper size ?
01:41:25saratoga2 bytes
01:41:28saratogalooks like its not a struct
01:41:42saratogai'm probably thinking of VLC
01:41:46 Join kk [0] (n=kkit@
01:41:50saratogastill, make sure that didn't end up an int or something
01:42:02nerochiarochecking now
01:43:20nerochiarosizeof(VLC_TYPE) is 2 on my system too
01:44:06 Join phalax [0] (
01:44:47saratogado you need static memory allocation on your system?
01:44:58nerochiarosaratoga: no, i can malloc
01:45:20saratogayou could replace the rockbox bistream.* files with the current ffmpeg ones
01:45:25saratogathey're only a few lines different
01:45:44saratogashould let you know if its a problem with my static method or not
01:46:12saratogabut really, if its working, maybe i'd ignore it for now
01:46:19 Quit qwm (
01:46:20 Quit preglow (
01:46:20 Quit FOAD (
01:46:20 Quit krazykit (
01:46:20 Quit EnterUserName (
01:46:20 Quit DogBoy (
01:46:20 Quit Bjoern-Erik (
01:46:30saratogaeventually it might be nice to save that memory for performance reasons, but i guess its not that pressing
01:47:01NJoinpreglow [0] (
01:47:01NJoinqwm [0] (
01:47:01NJoinFOAD [0] (
01:47:01NJoinkrazykit [0] (n=kkit@
01:47:01NJoinEnterUserName [0] (
01:47:01 Join DogBoy [0] (n=john@unaffiliated/dogboy)
01:47:01NJoinBjoern-Erik [0] (
01:47:01***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
01:47:17nerochiarowell, yes, i'm kind of ignoring it too. i moved it to the back burner, but it's a bit scary to have that hack there and not know why it's there. just that
01:47:33nerochiaroi'll try with ffmpeg bitstream files, though
01:47:49 Quit preglow (
01:47:49 Quit qwm (
01:48:20NJoinqwm [0] (
01:48:20NJoinpreglow [0] (
01:49:09saratogawhat was the status of your port anyway?
01:49:19 Quit Genre9mp3 ()
01:50:24nerochiaroit works right now, it decodes most of the samples i throw at it. i still need to fix the rare case when there are asf packets with > 1 payload. and implement seeking
01:50:37 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
01:51:03nerochiarobut some samples have these bursts of static and one is completely garbled
01:51:15nerochiarothe ones with static are probably for the reason that you mentioned
01:51:27nerochiarobut the garbled one is still unclear to me
01:52:46saratogawas it a really low bitrate file?
01:53:16 Quit krazykit (Connection timed out)
01:53:40nerochiarosaratoga: how do i check that, again ?
01:53:52 Join RudMan [0] (
01:54:47nerochiarosaratoga: ah, vlc player says 16Kb/s
01:57:00nerochiarosaratoga: is that very low bitrate ?
01:58:58nerochiarosaratoga: it's this one, by the way:
02:01:41saratoganerochiaro: yeah the wma lowbitrate features enable at 32 kbps (noise coding) and 16 kbps (Line spectral Pairs)
02:02:01saratogaboth are extremely difficult to implement in fixed point precision, so some problems likely remain
02:03:13nerochiaroi see. the situation is more clear now. let me check the bitrate of the files that have the bursts of static, i guess they will be low bitrate too
02:04:59 Quit HellDragon_ (" ")
02:05:41nerochiarosaratoga: two of the ones with static are low bitrate (<= 32). but one is 96Kbps and have the static too
02:06:01saratogathe 96k one is more interesting
02:06:11saratogacould you link or email it?
02:06:18nerochiarosaratoga: sure, one minute
02:07:03***Saving seen data "./dancer.seen"
02:07:04 Quit hannesd_ (Read error: 145 (Connection timed out))
02:07:14 Quit rasher (Read error: 60 (Operation timed out))
02:08:26nerochiarosaratoga: one thing that worries me a bit is that low bitrate wma files are pretty common. wma have been marketed as a codec suitable for low bitrate, so download sites seem to have low bitrate files often
02:08:50 Join Captain_A [0] (
02:09:40 Quit Captain_A (Remote closed the connection)
02:09:57 Join BigMac [0] (
02:10:00saratogamost people who use rockbox seem to have 128k or 192k files
02:10:27saratogai don't think i've had anyone submit a < 32k wma file with an error that they actually wanted to listen to
02:10:41saratogayou're actually the only person to ever send me music at less then 128k
02:11:04saratogawhich is why i suspect theres still a number of problems with low bitrate files
02:11:21saratogawell, that and the fact that the noise coding + lsp code paths are almost as long as the rest of the codec combined
02:11:28nerochiaroi'm not actually listening any of that. it's coming from the gals in the neuros testing team. they say they picked them from "popular download sites" (i didn't ask more)
02:11:31saratogaand much more FP intensive
02:12:24saratogaif you're finding problem files, you could probably fix them on your device pretty easily just by adding int64 variables to where ever the problem is
02:12:25 Quit BobShield (Read error: 104 (Connection reset by peer))
02:12:26 Join rasher [0] (
02:12:27nerochiarofunny, i would have expected that low bitrate codecs to be less complex for decoding than the higher bitrate ones. shows how little i know of audio coding tech, i guess
02:12:33saratogayou've certainly got the CPU power to afford that
02:13:00nerochiarocan you please explain better what you mean by adding the int64 vars ?
02:13:10saratogaright now the codec is all 32 bit
02:13:29saratogathe low bitrate problems are all just overflows in the LSP and Noise coding blocks
02:13:49saratogaswitching to 64 bit math neatly side steps the problem, but i don't do that because its quite slow
02:13:56saratogaand i'm working with ARM7
02:14:18saratogaso my fixes are generally more clever algebra to work around overflows
02:14:41 Join Febs [0] (
02:14:41saratogai basically just rescale variables, and rearrange multiplies and additions so that the values never get too big or too small to fit into a 32 bit int
02:15:14saratogaunfortunately, since i do not understand the math in the LSP or Noise coding sections, figuring out the needed range of each variable is quite difficult
02:15:21saratogatheres a lot of guess work
02:15:48nerochiarothese low bitrate algorithms aren't documented anywhere i guess ?
02:16:47saratogatheres various papers on LSP, but i don't really get the theory, so i didn't spend too much time on it
02:17:08saratoga"noise coding" is just a general idea, so who knows how it really works or what to look up
02:17:30saratogasadly ffmpeg does not provide citations with their comments
02:17:53nerochiaroaren't the authors reachable somehow ?
02:19:30nerochiaroand in original ffmpeg the LSP stuff is implemented in 64bit int math or float ? i was thinking to pick these parts up from ffmpeg is they don't use floats
02:19:42saratogaoriginally ffmpeg was all float
02:20:06saratogaexcept for things like huffman coding which don't need it
02:21:59 Quit Nico_P (Remote closed the connection)
02:22:01nerochiaroso all the int math in wma was added by you here in the rockbox code ?
02:23:43saratogamost of it
02:24:24saratogasome was done by Paul Jones last year
02:24:38 Join perrikwp [0] (i=98215e76@gateway/web/cgi-irc/
02:24:39saratogathough i've since redone most of his stuff
02:24:58saratogaall the low bitrate stuff is mine
02:25:19 Quit bdgraue (Remote closed the connection)
02:25:22nerochiarokudos for that, really
02:26:03saratogai guess if you can tolerate huge performance decreases, you could probably just do the LSP stuff in fp
02:26:13saratoganot sure i'd recommend that though
02:26:31saratogaint64 would probably be a factor of 5 or 10 faster, and really not any harder to implement
02:26:56 Nick iamben_ is now known as iamben (
02:27:10saratogawma_lsp_to_curve is probably where your lowbitrate problems are at
02:27:24nerochiarowell, fp is really too much for that processor. as for the int64 change, i'm not really sure i'm into the math enough to implement it, unfortunately. i might have to resort to just disable low bitrate support, at least in the beginning
02:27:49 Quit Vortex375 (Remote closed the connection)
02:28:09nerochiarounless it's just a matter of making the args of that func int64s ;)
02:29:08nerochiaro(btw, this is the 96kbps sample that has static bursts: )
02:31:25saratogaheres a fun idea
02:31:44saratogatry changing the fixed32 declaration in that function to fixed 64 and see if it works any better
02:32:21nerochiaroall of them ?
02:32:37saratogait'll get slower, but you're probably already at 500% realtime
02:32:39saratogamaybe faster
02:32:41nerochiaroallright, giving it a shot
02:33:14saratogathough for best results, you'd probably want to tweak the shifts in that function too so that smaller numbers are allowed (32->64 will allow larger numbers but not change the limit on smaller ones)
02:33:18nerochiaronot sure about the actual speed right now, but i guess it's at least 5x, yes
02:36:13saratogathat 96k file you sent me is quite strange
02:36:16 Join webguest70 [0] (i=4a7b3e0c@gateway/web/cgi-irc/
02:36:21saratogadefinately not encoded by wmp
02:36:46webguest70I was wondering about setting up the compiling environment on a mac
02:37:00nerochiarosaratoga: i can try to ask if they know with what it was encoded
02:37:15saratogaactually i take that back
02:37:22saratogaits just a very odd bitstream
02:37:32webguest70I've got it sorted, I think, but I don't know how to "run" from the terminal program
02:37:38saratogatheres maybe 5/5000 blocks that aren't the largest size
02:37:49saratogafile must be almost a constant tone
02:38:12webguest70can anyone help me out?
02:38:55 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
02:39:12iambenwebguest70: ./
02:39:22nerochiarosaratoga: mplayer plays it fine, it should give you an idea how it sounds like
02:39:32webguest70Thanks. I'll give it a try right now...
02:39:35saratoganerochiaro: i hear that static
02:39:45saratogavery strange, definately not a problem i've seen before
02:40:17nerochiarosaratoga: you only hear static, or you hear music and static ?
02:40:55saratogai get pure static for 1 block or so
02:41:02saratogai'm guessing the block didn't even decode
02:41:51nerochiarosaratoga: well, hope that problem sample helps
02:42:11 Nick fxb is now known as fxb__ (
02:42:15nerochiaroi tried the int64 quick fix, but it didn't improve things
02:42:25nerochiaroat least not appearently
02:42:36webguest70It works, but it now states that I must change the script because I don't have write permission.
02:42:51webguest70How do I get write permission?
02:43:11saratoganerochiaro: i take back what i said before about that being a normal file
02:43:22saratogaall the frames that aren't 2048 samples are just static
02:43:49saratogaand each one of those static burst is also 2048 samples long
02:44:03saratogai suspect that it doesn't detect the block size right on those samples
02:44:33nerochiarosaratoga: so it doesn't look hard to fix
02:44:45nerochiaroif it's indeed that the problem
02:45:21webguest70I guess I need to change the directory it will write to. Is that correct?
02:46:03nerochiarosaratoga: well, it's 3AM here, so i really need to go to sleep now. i needed to ask you some questions about seeking too, but we can maybe talk about that again the next time i catch you online
02:46:17saratoganerochiaro: sure
02:46:22saratogai have semi working seeking code already
02:47:05iambenwebguest70: yes you should edit the script and change install path and other options
02:47:22saratogaif macos allows sudo, just use that
02:47:36saratogaanyway, i'm going home
02:47:48nerochiarosaratoga: glad to hear that :) and thanks for all the help so far
02:47:54nerochiarosaratoga: that = the seeking thing
02:48:19nerochiaroi'm going away too. bye
02:49:09webguest70Do I change dlwhere, prefix, or builddir?
02:49:48iambenisnt this documented somewhere?
02:50:15iambenthe first is where the tgz's will be downloaded, the second is where it will be installed, the 3rd is where the build will take place BEFORE the install
02:50:46 Join Bagder_ [0] (
02:51:03webguest70I can't find the documenation anywhere, actually. All I found was "run from the shell"
02:52:30webguest70Do I change $HOME?
02:53:49webguest70I've been compiing on a PC using cygwin, but now I've changed to mac, and am all disoriented...
02:54:59iambennot sure how much i'll be able to help, ive never built on anything but linux
02:55:17webguest70I think it is fairly similar.
02:57:52 Part xaque
02:59:31 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
02:59:52 Join miepchen^schlaf [0] (
03:01:23webguest70I think I have to change prefix="/usr/local"
03:01:45webguest70Do I need to keep "usr" or "local"?
03:02:34iambenwell i kept my prefix like that, because i wanted mine installed at /usr/local/rbdev
03:02:44iambenbut you can put it wherever you like
03:03:15iambenif you put prefix="/home/user" then it will install at /home/user/rbdev
03:03:31webguest70How about "$HOME/rockbox/build"
03:04:50iambenthis $prefix is for where you want the build environment (ie gcc and friends) installed, not where you want the built rockbox to go
03:07:26 Quit Bagder (Read error: 110 (Connection timed out))
03:07:56webguest70Got it working! Thanks!!
03:09:30 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
03:09:51 Join miepchen^schlaf [0] (
03:14:02webguest70How much room is needed on the hard drive?
03:17:42 Quit Thundercloud (Remote closed the connection)
03:17:56 Join Nick_Brackley [0] (
03:18:01 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
03:18:22 Join miepchen^schlaf [0] (
03:20:16webguest70But now it says "Make your PATH include :/Users/federicosanchez/rockbox/devstuff/arm-elf/bin"
03:20:22webguest70ow do I do that?
03:22:07 Join kyuubiseal [0] (
03:22:14 Nick kyuubiseal is now known as cool (
03:22:33 Part cool
03:23:49webguest70In which file do I have to include this PATH?
03:25:28iambenyou can put it in your bashrc
03:26:01iambenor just export PATH="${PATH}:/Users/federicosanchez/rockbox/devstuff/arm-elf/bin" every time you are gonna build rockbox
03:26:36webguest70where is the bashrc?
03:27:01 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
03:27:22 Join miepchen^schlaf [0] (
03:28:19webguest70Do I put it in etc/profile?
03:28:21 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
03:30:18iambenyeah profile will work too
03:33:35 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
03:33:56 Join miepchen^schlaf [0] (
03:34:02webguest70I can't find profile or bashrc now
03:34:18 Quit BigMac (Read error: 110 (Connection timed out))
03:34:40 Quit Febs (Read error: 104 (Connection reset by peer))
03:36:01 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
03:36:22 Join miepchen^schlaf [0] (
03:40:13 Join annulus_ [0] (
03:41:19webguest70How do I add to bashrc?
03:43:04iambendo you have ~/.bashrc or ~/.bash_profile?
03:43:30webguest70I don't seem to have either
03:43:58webguest70I feel like I am so close to the end now, yet so far...
03:44:20iambeni don't know how you are supposed to set env vars on mac
03:47:02 Join elinenbe [0] (
03:47:43webguest70Maybe I'll just do as you said before
03:47:57webguest70and send "PATH ="${PATH}:/Users/federicosanchez/rockbox/devstuff/arm-elf/bin" when I build
03:48:28webguest70I suppose I just do that in the terminal window before sending ../tools/confgure
03:48:34webguest70Is that righ?
03:48:45webguest70Is that right?
03:48:55 Join thegeek [0] (
03:49:19Nick_Brackleyyeah thats right
03:51:11webguest70didn' work
03:51:36 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
03:51:57 Join miepchen^schlaf [0] (
03:52:44webguest70Sorry, I had an extra space. it's working now.
03:53:07webguest70Thank you very much!!!
03:53:18webguest70I am now off and running
03:53:31 Join Febs [0] (
03:54:56webguest70Bye for now.
03:55:11 Part webguest70
03:56:43 Quit Nick_Brackley ("ChatZilla [Firefox]")
04:07:08***Saving seen data "./dancer.seen"
04:15:18 Join xigxag [0] (i=cfcbfe6e@gateway/web/cgi-irc/
04:15:37xigxagyo rockboxers
04:16:05xigxagdoes anyone have a link for Senab build?
04:16:39xigxagthe links on are 404
04:16:59xigxagthat is...
04:17:09 Join donutman25 [0] (n=chatzill@
04:20:36DogBoysite down
04:23:05xigxagif anyone can setup a mirror or anything let me know
04:26:03 Quit jhMikeS (Nick collision from services.)
04:26:09 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
04:29:21 Join markun_ [0] (n=markun@rockbox/developer/markun)
04:29:43 Join Shaid [0] (
04:33:16Soapwhereas his main site is up
04:34:21DogBoythe document root?
04:34:35 Quit markun (Read error: 60 (Operation timed out))
04:37:21xigxagdocument root for the site, not for the browsable directory
04:38:14DogBoysay what
04:40:40kki'm sure another unofficial build has the features needed for those themes.
04:40:41 Quit perrikwp ("CGI:IRC (EOF)")
04:42:45xigxagare their other advantages to the custom builds other than patches included?
04:44:34DogBoythere are links in the forums for various custom builds
04:44:38DogBoyat least there used to be
04:44:59kkthat IS the advantage of a custom build
04:45:04kkis the patches.
04:45:43kkthe disadvantage is the lag time between builds. you may be missing some bugfix or feature for a few days.
04:46:11DogBoyyou could make your own
04:46:15Soapand the other disadvantage is you don't get support.
04:46:42DogBoyhe didn't have any of that before
04:48:27xigxagI'm gonna go ahead and dive into compiling to get my feet wet
04:48:41 Quit donutman25 ("ChatZilla [Firefox]")
04:52:00 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:52:05 Join miepchen^schlaf [0] (
04:56:30 Quit midgey ()
05:14:49 Join jhulst [0] (
05:22:55 Join homielowe [0] (
05:31:11 Quit xigxag ("CGI:IRC (EOF)")
05:32:35 Join Llorean [0] (
05:40:52 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
05:41:13 Join miepchen^schlaf [0] (
05:49:44 Quit Rondom (Read error: 110 (Connection timed out))
05:53:22 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
05:53:43 Join miepchen^schlaf [0] (
05:59:11 Join Rondom [0] (
06:04:49 Join desowin [0] (
06:04:56 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
06:05:17 Join miepchen^schlaf [0] (
06:07:11***Saving seen data "./dancer.seen"
06:16:05 Quit lazka (Remote closed the connection)
06:18:05 Quit jhulst ("Konversation terminated!")
06:25:26 Quit RudMan (Remote closed the connection)
06:30:25 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
06:30:44 Quit Hadaka (Read error: 60 (Operation timed out))
06:30:46 Join miepchen^schlaf [0] (
06:36:26 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
06:36:47 Join miepchen^schlaf [0] (
06:48:00 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
06:48:21 Join miepchen^schlaf [0] (
06:56:21 Join JerryLange [0] (
06:57:39JerryLangehello. just wondering what is the best way to go from a custom built version of rockbox to the default version of rockbox?
07:00:07desowinunpack .zip?
07:00:34desowin(if you don't have very old bootloader of course)
07:01:53JerryLangewell will any stuff be left behind from the custom build?
07:02:39desowinif you really feel like it, just remove .rockbox before unpacking new .zip
07:04:24JerryLangecool. thats what i think i am going to do
07:05:17 Quit JerryLange ("Leaving")
07:21:42 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
07:22:04 Join miepchen^schlaf [0] (
07:27:04 Join hannesd_ [0] (
07:31:05 Quit miepchen^schlaf (Connection reset by peer)
07:31:26 Join miepchen^schlaf [0] (
07:31:33 Join alienbiker99 [0] (
07:35:59 Join Siltaar [0] (
07:48:27 Join midgey [0] (
07:51:37 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
07:51:58 Join miepchen^schlaf [0] (
07:54:16 Quit miepchen^schlaf (Client Quit)
07:54:32 Join bsdfox__ [0] (
07:54:49 Join xigxag [0] (n=xenion13@
08:01:10 Quit desowin ("use linux")
08:04:39 Quit midgey ()
08:06:00 Join premhb [0] (i=530d0d73@gateway/web/cgi-irc/
08:06:09 Quit premhb (Client Quit)
08:06:26xigxagwhy do I need to use cygwin to compile?
08:07:05scorcheyou dont nee /cygwin/ persay
08:07:12***Saving seen data "./dancer.seen"
08:07:38 Join premhb [0] (i=530d0d73@gateway/web/cgi-irc/
08:07:49xigxagi suppose I mean why can't you use a win compiler, like studio
08:08:11scorchebecause it wont compile rockbox
08:08:14xigxagsorry, totally new to coding/compiling, obviously
08:08:46 Quit bsdfox_ (Read error: 110 (Connection timed out))
08:09:10 Quit tumu (Read error: 110 (Connection timed out))
08:09:28xigxagcould you provide a brief explanation as to why?
08:09:42scorchetry wont work =)
08:09:44xigxagsince it isn't a *nix app
08:10:03xigxaghah, no I believe you...would just like to wrap my head around why
08:10:14xigxagatleast vaguely..
08:10:41 Join Buschel [0] (
08:12:39 Quit premhb ("CGI:IRC (Ping timeout)")
08:13:25alienbiker99hey quick question can i change controls of a plugin without recompiling?
08:14:22GodEateralienbiker99: I can't think of any no
08:15:41 Quit Buschel ()
08:19:02xigxagholy crap, cygwin is incredible\
08:20:46midkayyou mean incredibly slow.
08:21:01GodEaterhe does
08:21:06GodEaterhe just doesn't know it yet :)
08:21:10midkayyeah, haha.
08:21:16xigxaghmmm, i dunno yet just lookin at the packages on the installer heh
08:21:43GodEateryou've reached an "incredible" verdict just looking at the installer ? Wow.
08:21:46xigxagI had no idea this was possible without a VM
08:22:26xigxagyeah, maybe premature but I just had an 'aha' moment
08:22:49GodEaterI wish more rockbox users were that easily impressed :)
08:24:11 Join miepchen^schlaf [0] (
08:26:02 Join CaptainSquid [0] (
08:27:06scorchehrm....anyone else want to laugh at midkay's sayings from 4 years ago? =>
08:27:18 Part midkay ("Leaving")
08:28:09xigxagmust be pretty bad
08:28:52scorchei dont even think visual studio even has a C compiler...C++, but not strictly C...
08:29:24xigxagthat sounds right
08:30:08xigxagit just seemed like C code was C code.
08:30:42xigxagIt didn't make sense why it would compile differently on the different OS
08:30:54 Join ender` [0] (
08:31:20scorcheit isnt about different is about different compilers
08:31:49scorchethere isnt One True Way to compile
08:32:13GodEateris there even a "reference" C compiler.
08:32:16*GodEater doesn't think there is
08:32:29xigxagso open-source projects are compiler exclusive?
08:32:40 Quit Llorean (Remote closed the connection)
08:32:55scorchethey typically are gcc, g++, etc
08:32:56xigxaglike if everyone is workin on the code, they must design it for a particular compiler
08:33:05 Join Llorean [0] (
08:33:14GodEaterxigxag: *this* project is like that
08:33:20 Quit Llorean (Remote closed the connection)
08:33:28GodEaterthere are of course others which use different compilers
08:33:36 Join Llorean [0] (
08:33:39GodEaterRockbox's requirements are rather exact though
08:33:44GodEaterLlorean: make your mind up!
08:33:58scorchewell, this is part of embedded, vs regular software
08:34:13LloreanThe newest version of Pidgin doesn't like being clicked on. At all, it seems. At least not in places with txt.
08:34:16 Join ddalton [0] (
08:34:20LloreanSorry for all the bouncing
08:34:29scorcheeasy...dont use pidgin =)
08:34:39LloreanWell, it's not just the IRC part of it.
08:34:39xigxagok, that makes some more sense
08:34:45 Join JdGordon [0] (
08:36:11TiMiD[FD]xigxag: visual studio cannot cross compile for mp3 player's cpu, so it's almost impossible to use it in the first place
08:36:12GodEaterLlorean: pidgin seems to not mind here
08:36:43LloreanGodEater: I'm using 2.1.1 on Ubuntu 7.10, and I'm assuming it's that combination that's somewhat unreliable.
08:37:11GodEaterhow odd
08:37:24xigxagyou use pidgin for an IRC client?
08:37:32LloreanThe odd bit is that it seems to shut down, not just crash, because I'm not leaving ghosts around.
08:37:48xigxagI see
08:38:31*GodEater prefers Xchat for IRC
08:39:29GodEaterI don't know why I bother trying to read the forums at this time of day
08:39:31DogBoyxchat is just imitation mirc though
08:40:11DogBoybut then I'm using it myself
08:40:23GodEaterexcept it's open source...
08:40:28 Join davina [0] (
08:40:45DogBoyI didn't know that
08:45:09 Join B4gder [0] (
08:50:46 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
08:52:22 Join prb_linker [0] (i=c27f0812@gateway/web/cgi-irc/
08:53:08prb_linkerCould somebody please help me sort out linking problems with
08:53:33prb_linkerThere are three source files, two .rock should be generated.
08:54:19GodEateryou've changed nick ?
08:54:42prb_linkerThe linker spits out something like [...]multiple definition of `___header'
08:54:52GodEaterone sec
08:54:52prb_linkerGodEater: tsss!
08:55:11 Join stripwax [0] (
08:55:33 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
08:55:37 Join ddalton [0] (
08:55:52prb_linkerI think this is a problem in the makefile. I don't see two separate LD calls there
08:55:57ddaltonsince I left rockbox users has the argument continued or is it over?
08:56:07pixelmaif you need more than one C file to compile your plugin, don't you need to add it to SUBDIRS (not SOURCES)? I don't know much about but...
08:56:16ddaltonI have seen two more messages but that is it.
08:56:34GodEaterI doubt it's a problem in the make file
08:56:34prb_linkerOne .rock should be produced from files _common and _view, the other .rock from _common and _append
08:56:58ddaltonso has anyone found out what players will run p7660?
08:57:21prb_linkerI think they are compiled and linked all three −− hence the probs with linking
08:57:28pixelmaprb_linker: same as plugins like sudoku and so one do?
08:57:50prb_linkerSince two of them include MEM_WRAPPERS
08:58:14prb_linkerpixelma: not quite like that since I want to produce two .rocks
08:58:31prb_linkerpixelma: whereas sodoku and co only produce one
08:59:00 Join B4gder_ [0] (
09:00:29 Quit B4gder (Connection reset by peer)
09:00:40 Nick B4gder_ is now known as B4gder (
09:00:50prb_linkerB4gder: now that the guru has arrived (and hopefully will read the last ten posts... ;-)
09:01:45B4gderI'm a bit busy now but I can have a look later
09:02:54prb_linkerB4gder: ok, no hurry please!
09:03:13prb_linkerB4gder: actually, I'm also a bit busy now
09:04:11 Join TiMiD[FD1 [0] (n=TiMiD[
09:05:31 Join petur [0] (n=petur@rockbox/developer/petur)
09:06:20GodEatergod, what a mess
09:07:54xigxagerrr, I can't seem to locate the daily source. It is included in the compiled zip or something?
09:08:11GodEaterxigxag: no
09:08:19GodEaterxigxag: try current builds
09:08:25GodEaterxigxag: not the daily ones
09:08:29Lloreanxigxag: It's the "Source Archive" on the current build page...
09:08:40B4gderxigxag: there's a current source and there's a daily source, a ~9MB zip
09:09:04B4gderand there's svn...
09:09:07xigxagthe link from source archive is the bleeding tar
09:09:23 Quit Jon-Kha (Read error: 145 (Connection timed out))
09:09:23xigxagyeah I definitely don't want to svn, heh
09:09:34scorchewhy not?
09:09:36B4gderwhy not? it is a lot simpler to update
09:09:41Lloreanxigxag: Why don't you want SVN or even the bleeding edge?
09:09:43B4gderand makes less things to download over time
09:09:48LloreanThe daily is just the bleeding edge made once a day...
09:09:50xigxagI thought that meant it was checked out?
09:10:07peturprb_linker: I think you can only have one source file for a .rock, if you need more move it to a subdir or move the common part in lib
09:10:24B4gderxigxag: checking out the code from svn gets you the latest and most up-to-date
09:10:27xigxagbecause I don't know what I'm doing
09:10:38B4gderyou'll learn
09:10:53scorche"checking out" is the same thing as downloading the source
09:10:56Lloreanxigxag: Checking out just means "downloading", it doesn't mean you can do anything to our copy
09:10:58xigxagyeah, but I didn't want to have code checked out if it would inhibit what someone else is doing
09:10:59ddaltonadn you can use svn diff and svn up
09:11:14 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
09:11:32LloreanIt's not analogous to say, checking out a book, where it prevents someone else access to the book.
09:11:37stripwaxxigxag - svn doesn't lock when you check out, it's fine!
09:11:55ddaltonisn't that check in or something?
09:12:29xigxagheh, good deal. I just knew it was some kind of versioning, so I thought it would have a way to prevent different people from altering the same parts of the code at the same time or something
09:12:51prb_linkerpetur: I moved all sources (three .c files) to a subdir (shortcuts). From those three files, two .rocks must be produced. Should be possible, I just don't know how
09:13:06stripwaxconflicts are handled at checkin only - nothing really prevents people altering the same code, so don't worry :)
09:13:20stripwaxunlike Visual SourceSafe, say
09:13:32 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
09:14:19 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
09:14:40linuxstb_prb_linker: Are you fm2?
09:15:14GodEaterprb_linker: you also appear to have multiple copies of the source within one file
09:15:20GodEaterlinuxstb: yes he is :)
09:15:43prb_linkerno, I am not!!!
09:16:08 Quit TiMiD[FD] (Read error: 110 (Connection timed out))
09:17:05B4gderspeaking of USB hubs I just read this:
09:17:13B4gder"hubs will not work on OSD, DM320 has a limitation of supporting only one device at a time."
09:17:24linuxstb_prb_linker: The problem (as others have said) is in the Makefile - it's only generating one .rock. You need to split the $(OUTPUT): rule into separate rules for each rock
09:17:45LloreanB4gder: Speaking of hubs?
09:17:59*linuxstb_ re-reads the logs...
09:18:03 Join bdgraue [0] (
09:18:05B4gderwe talked about usb, hubs and multiple usb devices yday
09:18:21LloreanI was wondering if this was an extension of that, or something I missed another time
09:18:25LloreanI was in favour of "no hubs"
09:18:33prb_linkerlinuxstb: that I have assumed. But I'm not very good at writing makefiles so if somebody could do this properly... I'd be grateful
09:18:36*linuxstb_ just want a disk mode
09:18:46*Llorean will be perfectly happy with just a disk mode
09:20:32*GodEater adds his vote for "just disk mode"
09:20:59linuxstb_And of course a serial port driver could have nice debugging uses...
09:21:13peturwas just going to say that too
09:21:49B4gderit would be amusing to get better usb support than the osd ;-)
09:22:16markun_maybe some code from the usb_serial driver we already have could be reused
09:22:20GodEaterI'm sure that patch file is broken too
09:22:28GodEaterit patches each file twice from what I can see =/
09:23:17ddaltonjust got p7660 to work on the sim (gigabeat)
09:23:41linuxstb_Is that espeak?
09:23:44markun_ddalton: great!
09:23:44*ddalton Thinks it will be great if we can some how get it to run on other targets
09:23:49markun_ddalton: it's real-time?
09:23:53ddaltonand sort the licence out.
09:24:11linuxstb_I assume it's on hold until we hear back from the espeak author...
09:24:21ddaltonyep it read a text file to me exactly as expected.
09:24:25scorchemarkun_: he said sim, so i would assume it is real-time ;)
09:24:31markun_ah :)
09:24:36ddaltonwhat is real-time?
09:25:00linuxstb_If it takes less than 1 minute to generate 1 minute of audio.
09:25:09markun_scorche: I think my brain just wanted to hear the gigabeat part :)
09:25:15prb_linkerGodEater: hrm... you are right. I don't know how it happened. I produced the patch with TortoiseSVN.
09:25:38ddaltonso what is different on the gigabeat to the h300?
09:25:46GodEaterddalton: the cpu
09:25:48ddaltonor is the sim different to the player?
09:25:53Lloreanddalton: The sim is very different from the player
09:25:54GodEateramongst other things :)
09:26:11linuxstb_Didn't toni say there were endian issues on big-endian?
09:26:13markun_ddalton: the sim is not an emulator
09:26:13ddaltonso if it runs on the sim it might not run on the actual gigabeat?
09:26:19LloreanI'd bet the H300 sim will work.
09:26:29Lloreanlinuxstb_: Yeah, it's endianness he said
09:26:39ddaltonlet me try just it might take a while to compile
09:26:43markun_ddalton: which player do you have?
09:27:52 Join Jon-Kha [0] (
09:27:53 Nick markun_ is now known as markun (n=markun@rockbox/developer/markun)
09:32:15ddaltonit says don't wrong.
09:32:44markunLooks like espeak also has support for chinese. Maybe the rockbox voice strings should be written in pinyin for that.
09:33:09B4gderlet's save that for the massive amount of chinese contributors...
09:34:09ddaltonIt voices in full punctuation. So it will not voice words like wouldn't ,shouldn't and don't. And so on.
09:34:24ddaltonit says them though
09:34:31GodEaterbetter than nothing
09:34:34markunB4gder: reminds me of the time I spent on arabic support vs the amount of users we have
09:34:36linuxstb_How does it say them?
09:34:42B4gdermarkun: :-)
09:34:46ddaltonwell that's right. A good start.
09:34:51GodEaterwouldn apostrophe t
09:34:54GodEaterI would guess
09:35:22ddaltonyep but it shouldn't be to hard to create some sort of dictionary file once everything is working.
09:35:24markunddalton: in the normal application it's an option to voice them I think, maybe we need to set some variable correctly
09:35:55markunthe −−punct option
09:36:17ddaltonwell as I might have said before I don't think I have the skills to port it yet so I will read my c book. But when I do I will deffenetly help out.
09:37:33ddaltoncan you start variables with -?
09:37:40markunlinuxstb_: I wonder what the espeak port changes for my GSoC student..
09:37:52markunddalton: no
09:38:04ddaltonso is the espeak project under the gpl v 3 and rockbox is under v2?
09:38:27ddaltonand what is the differences?
09:38:45ddaltonbetween v2 and 3?
09:38:48GodEaterlets just say they're not compatible
09:38:58*GodEater doesn't want to see this whole debate all over again
09:39:05ddaltonok then which one gives the developer more fredom?
09:39:10linuxstbmarkun: I don't think it changes anything - apart from helping him integrate his port with Rockbox by giving him some code he can steal.
09:39:19GodEaterddalton: that's the point people like to argue over
09:39:27markunlinuxstb: good point
09:39:32markunI'll mail him about it
09:39:38linuxstbmarkun: I think it will be very nice to have at least two TTS plugins, so we can make a genuine comparison of them
09:39:40 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
09:39:43ddaltonok then I am good at starting arguments I think. So won't start one now!
09:39:58ddaltonwhat's the other one?
09:40:26GodEaterddalton: flite is the other one being ported
09:40:29GodEaterI think
09:40:44 Join Rob222241 [0] (
09:42:15*linuxstb wonders how the SoC evaluations are going - have any mentors submitted them yet?
09:42:26ddaltonso which one is making better progress at the minute?
09:42:30*petur is still waiting...
09:42:31scorchewerent they supposed to be in yesterday?
09:42:38peturno 31/8
09:42:40markunlinuxstb: I have submitted a negative one for now
09:42:53linuxstbmarkun: I thought that an extension had been granted?
09:43:04markunyes, but this was the deal
09:43:05*petur still isn't sure if he should pass or fail
09:43:08*ddalton Wonders which one sounds better. (I am sure users will find that important) Doesn't worry me.
09:43:39markunI will change it to a positive one if he produces something during the extension
09:43:51GodEaterddalton: that's a matter of preference too. LinusN says he hates espeak, but I prefer it over the other two
09:43:59ddaltonHow does it read the language selection?
09:44:10LloreanI definitely prefer espeak over the other two, as well
09:44:15LloreanSounds less grating to me.
09:44:19GodEaterthink markun does too
09:44:23ddaltonwell espeak sounds fine to me. But I don't mind any.
09:44:31GodEaterLinusN really doesn't like it though
09:44:38ddaltonI just want them to read the screen to me so I can understand it.
09:44:42markunGodEater: nobody really liked espeak when I demo-ed it a bit during devcon
09:44:52GodEatermarkun: that's a shame
09:45:08markunI personally don't care, as long as I'm happy with it :)
09:45:49ddaltonso how does it read the language selection when you enter the plugin?
09:45:53markunI've listened to quite some books while making midnight walks using espeak ;)
09:45:56LloreanI don't like espeak in comparison to say... some of the windows voices.
09:46:10LloreanBut compared to flite and festival, that's a different story.
09:46:19 Quit bdgraue (Remote closed the connection)
09:47:26*ddalton I only got up to 217 kb per sec when downloading the h300 voice!
09:47:34GodEaterI've not heard any voices on windows that I like
09:47:50LloreanGodEater: heard the old AT&T ones?
09:47:58ddaltonwell I hate the microsoft ones but the IBM sounds ok.
09:48:31 Quit prb_linker ("CGI:IRC")
09:48:45ddaltonI don't mind the linux voices. I think I will be forced to use them when switching to linux.
09:49:09GodEaterLlorean: no - they were pulled before I ever tried voice
09:49:29markunI hope a espeak plugin for rockbox will make people submit patches to espeak for improvements and other languages
09:49:30LloreanGodEater: They sounded very natural and nice
09:49:50GodEaterLlorean: I'll take your word for it
09:49:53Lloreanmarkun: It'd be hard for us to incorporate those patches though, unless he goes back to v2 or later.
09:50:19markunLlorean: let's first wait what jonathan replies to aliask
09:51:23LloreanI'm really hoping we'll get a positive response there, or at least some form of limited dual licensing that will work for us.
09:51:31linuxstbLlorean: Or we go v3...
09:52:50ddaltonyep it works on my sim build for the h300.
09:53:04Lloreanlinuxstb: We could, but I'm uncertain how easily we could do that, re various contributions.
09:54:35 Quit Llorean (Remote closed the connection)
09:54:50 Join Llorean [0] (
09:56:00GodEaterstop clicking on things Llorean!
09:56:13 Quit xigxag (Read error: 110 (Connection timed out))
09:56:17GodEateror stop using pidgin
09:56:24ddaltonespeak reads the language selection but not the text file.
09:56:55LloreanGodEater: I'm tempted to stop using it for IRC at least, for the moment.
09:57:22*ddalton Better read this c book!
09:57:35 Quit Rob2222 (Read error: 101 (Network is unreachable))
09:57:36GodEaterddalton: testing on the h300 sim isn't the same as testing on an actuall h300
09:57:49GodEateryour PC's cpu has a different endian-ness to the h300's cpu
09:58:04ddaltonwell that's probably what happen it is different. I think my last on the player I did something wrong.
09:58:07peturand is a bit faster
09:58:12GodEaterjust a bit :)
10:05:37linuxstbAre there any US residents around who would be willing to buy a Gigabeat S for me and ship it to the UK?
10:06:13*JdGordon votes scorche
10:06:34scorcheheh...would be the 3rd S i have shipped abroad
10:06:44*scorche is still waiting on GodEater's
10:06:52linuxstbscorche: Then you know what to do... ;)
10:07:01scorchefrom the refurb depot place?
10:07:13linuxstbUnless you've found somewhere cheaper - I haven't.
10:07:14***Saving seen data "./dancer.seen"
10:07:26scorcheyou have paypal?
10:07:34 Join spiorf [0] (
10:07:45*Llorean is probably going to end up purchasing his from the refurb depot.
10:07:46B4gderhow much are they from there?
10:07:57linuxstb$169.95 for the S30 (single-platter 30GB)
10:08:10linuxstbscorche: I think so...
10:08:23scorchei put the figure at 200 USD to ship abroad
10:08:37markunSomeone at mygigabeat was trying to tell me that the rockbox port to the S was not making progress. I thought I had some authority over there by now..
10:08:45linuxstbscorche: That sounds reasonable to me.
10:08:53B4gdermarkun: the kids never learn you know...
10:09:03scorche(i still have yet to see the exact price, but if anything, i would be eating $4 or so)
10:09:26markunB4gder: mostly kids over there indeed
10:09:33linuxstbkids are everywhere...
10:10:02scorchemarkun: do you know if aliask unbricked his?
10:10:13markunno, didn't talk to him
10:10:22B4gderthe screen of the gigabeat S looks awesome
10:10:36GodEaterB4gder: you can't wait to play doom on it can you? Admit it.
10:10:43markunB4gder: the Gigabeat X has the same screen (and runs rockbox)
10:10:54scorcheB4gder: in the OF, it shows just 4 shades of grey was horrid
10:10:56GodEatermarkun: and is just as hard to find!
10:10:58B4gderthe S should run rockbox too any day now ;-)
10:11:25*B4gder sends some pressure to the porting guys
10:12:14*GodEater was sure he'd seen a coffee mug somewhere on the net the other day with a "strength guide" printed on the inside
10:12:57GodEaterand now I can't find it
10:13:18GodEatername somewhere other than thinkgeek (which I've checked) which might carry something like it!
10:20:11 Join obo [0] (n=obo@rockbox/developer/obo)
10:21:04GodEater <−− in case anyone is interested.
10:22:45webmindGodEater, milk in coffee ?
10:23:40 Join przemhb [0] (i=5035dcaa@gateway/web/cgi-irc/
10:23:52 Join Shaid` [0] (
10:24:49przemhbHi, my name is Przemyslaw Holubowski
10:26:06przemhbI would like to add a new codepage to rockbox - CP1250 (Central European). I have already edited codepage_tables.c and .h
10:26:37przemhband as I am completely new to rockbox community I've stuck at the point
10:27:01przemhbcould someone help me little bit saying how to add the codepage?
10:27:38markunprzemhb: yes, one moment
10:27:45amiconnBagder_: osd?
10:27:54B4gderamiconn: neuros "open source device"
10:28:08*amiconn googles
10:29:45markunprzemhb: you also need to change iso_decode in tools/codepages.c
10:30:08markunprzemhb: add a case for 0x08
10:30:50przemhbalready done
10:30:58przemhbwhat more?
10:31:43markunprzemhb: add it to apps/settings_list.c
10:32:08linuxstbB4gder: I thought it was a "mostly open source device" ?
10:32:18B4gderyes it is
10:32:32B4gderthough they call it OSD which supposedly stands for that
10:32:32markunprzemhb: just grep for "codepage", you will probably find a few places where to change the code
10:33:03B4gderI would even say "partly open source" since there's quite a lot of hidden stuff
10:33:12linuxstbPOS device?
10:34:34amiconnmarkun: Btw, is there a reason why unicode.c is in firmware/common, but arabjoin.[ch] and hangul.c are not?
10:35:00markunamiconn: no, not really
10:35:29markunnow that we have subversion it could be moved
10:38:00przemhbmarkun: Thank you.
10:38:29 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
10:38:37markunamiconn: maybe the arabjoin funcion from bidi.c should be moved to arabjoin.c
10:38:54 Quit Shaid (Read error: 110 (Connection timed out))
10:38:56 Nick Shaid` is now known as Shaid (
10:38:58przemhbI have one more question - is it a bug that rockbox does not play m4a files encoded with NeroAAC and -br or -cbr switches?
10:39:56linuxstbI think the main known bug with the m4a parser is that it won't play "non-streaming" files - i.e. files with the AAC metadata at the end of the file, rather than the more usual start.
10:40:01 Join bdgraue [0] (
10:40:04linuxstbPlus possibly very long files.
10:40:28linuxstbLear is the man to talk to.
10:42:16 Quit miepchen^schlaf ("Verlassend")
10:43:46ddaltonHas anyone got espeak to run on any player?
10:43:53ddaltonso the actual unit?
10:44:42markunamiconn: we now use short in the bidi code for the the unicode chars, should it be changed to long?
10:45:23markunamiconn: same for utf8decode in unicode.c
10:45:44markunddalton: I thought you were trying to do that
10:45:57ddaltonjust did it on the sim
10:47:13 Part Llorean
10:47:33markunddalton: so, which player do you have?
10:48:28 Nick fxb__ is now known as fxb (
10:48:36 Quit spiorf ("Read error: 110 (Connection timed out)")
10:48:36 Quit bdgraue (Remote closed the connection)
10:48:51ddaltonI have a h300. I will give it another go later.
10:49:03*Zagor reopened fs#2720
10:49:06ddaltonjust trying to sort out this linux box I am looking to buy.
10:49:55Zagordid any sort of rules get written down during the tracker cleanup effort?
10:51:44 Join Llorean [0] (
10:52:13Zagorfor example how to treat old patches with active users
10:52:29linuxstbmarkun: Can you remember how long the gigabeat F port took? i.e. from the time you first tried to run your own code on it, to audio playback?
10:53:05amiconnmarkun: I can't see why. Did you ever encounter unicode chars from outside the bmp?
10:54:14amiconnMany function use stack buffers, and doubling the amount of memory for those buffers would be bad
10:54:58markunlinuxstb: from the time I ordered my player to playback it was about 13 months
10:55:05 Join pondlife [0] (
10:55:22markunlinuxstb: but not sure at which time we had code running
10:56:33linuxstbmarkun: I meant from the time you started work on the port, so 13 months answered my question. Bad news about aliask's S...
10:57:47*amiconn reminds linuxstb about 2 inactive ports...
10:58:43*GodEater wonders why aliask said he had to use a windows PC to load code onto his S
10:59:11markunGodEater: because the modified Gigabeat V updater is a windows executable
10:59:54GodEaterso the merge thing is just to create the firmware, not to upload it
11:00:40markunyes, I believe so
11:00:51GodEaterthat's going to be a pain too
11:01:29scorchedoes it work in wine?
11:02:23markunnot sure. But if when have a bootloader with USB support you only need a windows PC 1 time
11:02:35markuns/if when/when we/
11:02:47LloreanI can't imagine we can't replicate what's sent to the device at some point, anyway.
11:03:22LloreanBut WINE is an interesting question.
11:09:11linuxstbamiconn: I'm back working on the AV300. I don't intend to work on the Gigabeat S port - but I know that once the port is working I will want to buy one, and they seem to be increasingly hard to find...
11:10:08*Nico_P got an S 2 two days ago :)
11:10:19Nico_Pmaybe I'll manage to be of some use to the port
11:10:29 Quit gtkspert (Remote closed the connection)
11:10:45 Join gtkspert [0] (
11:11:00Nico_Plinuxstb: "Bad news about aliask's S" => anything in particular happened ?
11:11:16linuxstbIt appear's that it's in a non-working state for some reason.
11:11:53 Join midkay [0] (n=midkay@rockbox/developer/midkay)
11:13:09 Join My_Sic [0] (
11:13:09Nico_Pso he doesn't know why ?
11:14:47GodEaterhe didn't say so at the time
11:14:58amiconnlinuxstb: What about the elio?
11:15:33B4gderthe elio port has served as a good starting point for other mi4 work
11:16:03B4gderlike the mrobe
11:16:18linuxstbA couple of people have been around recently showing interest in the Elio, but I'm not convinced they have the motivation to go through the learning curve of porting Rockbox
11:17:04amiconnThe elio itself is a somewhat interesting target imo, because it is hdd based, and has a card slot + usb host
11:17:19amiconnThe rest is standard pp
11:17:19peturdoes it have line-in?
11:17:29linuxstbI haven't completely abandoned it - it just needs the LCD breakthrough...
11:18:37linuxstbpetur: It has a standard USB socket, a dock connector similar to the ipods (but not the same), and a card slot. I don't know if there's a line-in in the dock connector.
11:18:46Nico_Pspeaking of the mrobe, what's its status ?
11:19:05B4gderNico_P: you need to ask low_light about that
11:19:16B4gderhe showed a screen shot in the forum
11:19:37linuxstbIt also has a very decent number of buttons - a four-direction pad and four buttons on the front, plus two buttons and a hold switch on the side. But it quite bulky
11:20:33Nico_PI remember the screenshot but wondered whether there had been more progres since
11:21:08 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
11:21:08linuxstb... plus another two buttons (volume) on the other side...
11:21:48B4gderNico_P: I've not seen any updates on that front recently, possibly he's been busy on the c200 front...
11:22:32B4gderhe's our new start-new-ports-without-having-the-previous-in-a-reliable-state guy ;-)
11:22:42Nico_Phehe :)
11:22:55preglowanother one? :P
11:24:54*linuxstb goes to the corner
11:25:47*scorche goes outside to have another look at the moon
11:26:09Nico_Pthat makes me wonder... any reason why the archopen code was never used to port rockbox to the tragets they have ?
11:26:31B4gderNico_P: because nobody ever bothered I guess
11:27:06linuxstbNico_P: That's what I'm working on at the moment - the av300 port. But the archopen-supported devices are more media players than DAPs, so Rockbox isn't ideally suited to them.
11:27:55scorchethe av300 is ideally suited to me =) (the screen isnt needed, but otherwise...)
11:28:08*scorche needs to start clearing off his projects list to he can move on to that...
11:29:00Nico_Pand how is archopen suited for these targets ? better than the OF or not ? One of my friends bought a gmini400 and I was wondering whether to point him to archopen or not
11:29:39 Join Thundercloud [0] (
11:30:56LloreanMy understanding is that Archopen can be seen more as a supplement to the OF than a replacement, right?
11:31:17 Join DerPapst [0] (
11:31:22DerPapstmorning :)
11:32:09linuxstbLlorean: I think so - the main focus seems to be similar to IPL - games.
11:32:23scorchethat is what i gathered as well
11:32:26linuxstbAlthough recently (in the last few months), they've added an audio player. I haven't tested that yet though.
11:32:29amiconnSpeaking about reliable state...
11:32:39*DerPapst hates all those stupid iPL games
11:32:40*amiconn really needs to get his head around powermgmt
11:32:44linuxstbBut afaik, there's no attempt at video playback, which is the main use of those devices.
11:33:19scorchehrm...last time i checked, i dont think they had audio playback
11:33:29*scorche goes and checks for a changelog
11:33:34Lloreanlinuxstb: Well *we've* got Video Playback.
11:33:41DerPapstsomeone made yet another emu for iPL which runs with incedibly 2-4 fps :-/
11:33:48Lloreanlinuxstb: They could just port Rockbox as one of their "plugins" or whatever. :-P
11:33:50linuxstbLlorean: Yes, another unfinished project of mine... ;(
11:33:55B4gderscorche: they have audio at least
11:33:56Nico_Pso basically rockbox would be more suited than archopen for those devices :)
11:34:20Lloreanlinuxstb: I don't consider video particularly unfinished. The only real thing lacking is FF/RW, right? (Something I've not actually felt the lack of, oddly enough)
11:34:35LloreanNico_P: *I* think it would be. ;)
11:34:55linuxstbLlorean: I suppose so. But it still has optimisation potential.
11:35:07linuxstb(especially on Coldfire)
11:35:10LloreanBut we've got a lot of codecs with optimization potential too.
11:35:18Nico_Pwasn't someone working on FF/RW ?
11:35:23Nico_PjhMikeS ?
11:35:41LloreanNico_P: There's a couple patches with two different implementations of nearly-seeking, and yeah, jhMikeS was looking into it too
11:35:46linuxstbLlorean: Yes, but at least they're working realtime. mpegplayer is only realtime (if we consider 25fps+ fullscreen video realtime) on a few targets.
11:35:47amiconnLlorean: And stability / sorting out the voice collision
11:35:52*Nico_P also remembers a patch for SRT support :)
11:35:56Lloreanamiconn: Ah, that's an important one.
11:36:32*linuxstb is expecting the av300 port to force him into working on hw/swcodec integration
11:36:32Lloreanlinuxstb: Considering most of those targets weren't designed for video, and the fact that the few that play video natively are beaten by Rockbox already, I'd consider it alright though
11:36:50LloreanI mean we match the Sansa's OF (and beat it in 16:9) and we beat or match the H300 don't we?
11:36:59LloreanOr are we behind on the H300, I seem to recall we were at 10-12 for a while
11:37:01amiconnlinuxstb: This morning I was thinking once more about shorten on archos....
11:37:21*preglow whispers "PP5020!!" to amiconn
11:37:48amiconnWell, if you fix the nano issue meanwhile... ;)
11:38:09linuxstbamiconn: My current thinking is that the av300 will be a hwcodec target. So I'll need to add support for the PCM codec and software decoders. The difference will simply be that the av300 can decode more formats.
11:38:14preglowdon't even know if my nano is affected, heh
11:38:51Lloreanpreglow: Update to svn-current, and play a mixed playlist of MP3, Ogg and FLAC for about 30 minutes, and you'll know. :-P
11:39:01amiconnEven if it's not it might be helpful for figuring out what's going on
11:40:40preglowwhat needs to be done?
11:41:29amiconnWell, I can't tell exactly what's needed, that's why I can't work on it without an actual unit
11:42:16amiconnI suspect it's ata timing, so if I'd have to start with an unaffected unit, I'd try increasing CPUFREQ_MAX to 100MHz and see whether this will make it an affected target
11:43:25amiconnI'd also start playing with the ata timing registers and see whether it's possible to fix the ata sleep hangup by using proper timing on wakeup
11:44:02amiconnThe toshiba hdds obviously aren't susceptible to timing errors
11:44:03preglowi'll see what i have time for
11:50:40 Part przemhb
11:51:56 Quit Rondom (Read error: 110 (Connection timed out))
11:52:09*pondlife thanks rasher for the perl voice building.
11:54:36 Join Rondom [0] (
12:02:22amiconnpondlife: It's still sub-optimal
12:03:26pondlifeIndeed, but much faster
12:04:01pondlifeMy old PC now builds a full set of voice files in about 20 mins, rather than 2 hours.
12:04:37amiconnI need 10 minutes for *one* voice now
12:04:43pondlifei.e. Cygwin + Windows file system was the problem
12:04:56rasheramiconn: Using the clip-pool, subsequent voices are quite fast
12:05:02amiconnThe actual problem in cygwin isn't filesystem, as I had to learn
12:05:03pondlifeI'm not using the pool
12:05:08amiconnIt's process creation
12:05:26pondlifeEither way, it's working in a reasonable time now.
12:05:30rasherpondlife: wow, then it should be even faster if you do that
12:05:45 Join spiorf [0] (
12:05:48rasheramiconn: did you try moving lame and wavtrim invocation to the vbs script?
12:05:49amiconnThat's why I hope to speed it up further
12:06:05pondlifeHmm, maybe I am using pool... that never worked for me before (just produced empty voice files).
12:06:30amiconnrasher: No, not yet. I'll also enforce clean coding in the vbscript (Option Explicit)
12:06:32 Quit spiorf (Remote closed the connection)
12:06:35linuxstbamiconn: I don't know if it's comparable (or if you even want to work on it), but you could try a CF card in your mini. IIUC, CF cards have the same sleeping problem that the Nano does.
12:07:06amiconnNot all CF cards will work, they need to support trueata (or however that's called)
12:07:09rasheramiconn: haha, yeah I'm just satisfied that it even runs.
12:07:18***Saving seen data "./dancer.seen"
12:07:33pondliferasher: Looks like I am using pool now...
12:07:46amiconnOption Explicit enforces to declare all used variables (with Dim). It might even speed up things, never measured...
12:08:31pondlifeAre we now starting a SAPI server? Or is it a new invocation each clip still?
12:08:58rasherpondlife: It's not a server as such, but we only run the vbscript once and keep it running throughout the script
12:09:07pondlifeOK, so a single invocation
12:09:20pondlifeWell, whatever you did works well here
12:09:32rasherIt's mainly the perl, probably
12:09:52rasherUsing shell means starting a million processes all the time
12:10:00*pondlife pings B4gder about uploading daily voice files.
12:11:08rasherpondlife: Which voice are you using, and does its license permit redistribution of the result?
12:11:26pondlifeMicrosoft Sam, should be ok, no?
12:11:39pondlifeI guess I need to learn about rsync?
12:11:45linuxstbAre you planning to replace the current voice files with windows-generated versions?
12:12:11pondlifeMaybe. They certainly sound better - despite the ploo-jins.
12:12:31rasherYou can adjust that in
12:12:38rasherif you're handy with regexps
12:12:52linuxstbWill that be various languages as well?
12:12:53pondlifeAh, did you do that work too?
12:13:03pondlifelinuxstb: No, I only have English :(
12:13:08rasher$string =~ s/plugins/pluck-ins/; # might work?
12:13:31pondlifeMaybe we should have a daily voice build farm?
12:13:54pondlifeSo we can have a selection of languages and voices.
12:14:17rasherI'd say the festival/flite/espeak voices could all be done by the build-server
12:14:32preglowwhy use anything but festival?
12:14:44rasherBecause festival isn't terribly hot
12:14:45preglowflite is festival with fewer features and espeak sounds like crap :)
12:14:57pondlifepreglow: Festival sounds awful.
12:15:10rasherFlite might actually be a better choice for the english voice, since it's faster
12:15:13preglowsounds ok enough to me
12:15:15rasherFaster to generate, that is
12:15:19linuxstbpondlife: That wouldn't be that hard - you can run sshd with cygwin so the existing build scripts could probably do the job with little tweaking.
12:15:51*pondlife knows little about Linux or shells (yet).
12:16:16pondlifeBut if there's a way I can help, tell me where to point my Windows box.
12:16:44GodEaterpondlife: call yourself a hacker? pfggh!
12:16:54pondlifeNah, I'm no hacker.
12:17:12pondlifeWell, I did do 8086 ASM for 4 years for money...
12:17:27pondlife... but that's about as far as my credentials go.
12:17:31rasherNow if B4gder could fix the "voice builds bitmaps", cygwin voice creation would be even faster
12:19:15 Join spiorf [0] (
12:21:19 Quit ddalton ("#emacs")
12:21:21 Quit Siltaar (Remote closed the connection)
12:21:25 Join ddalton [0] (
12:21:46*JdGordon is annoyed with his lack of coding recently...
12:22:10*pondlife is annoyed with JdGordon's lack of coding recently... ;)
12:23:25 Join bluebroth3r [0] (i=810dba01@gateway/web/cgi-irc/
12:23:35B4gderrasher: I still haven't thought out a fix for the problem so it won't happen very quickly
12:23:41JdGordonis there a way to automatically remove all non version controlled files?
12:23:51linuxstbB4gder: What about creating a completely separate Makefile for voice?
12:24:09B4gdersure that could work
12:24:29B4gderor possibly adding some condition in the existing for voice
12:24:44linuxstbDo we have the same problem for the manuals? Or are they not using features.txt ?
12:25:01 Join tvelocity [0] (
12:25:07B4gderI don't think it uses that, it has a separate makefile iirc
12:26:04bluebroth3rthe manuals don
12:26:10bluebroth3r't use features.txt
12:26:26pixelmabut if configured for manual, a simple "make" complains about the missing (or similar IIRC)
12:26:30bluebroth3rbut it might be feasible to do that ...
12:26:55bluebroth3rwell, that's a bit different −− the makefile system is tricked a bit
12:26:56pixelmabut compiles ok
12:27:16bluebroth3rwhich calls the stuff at the end
12:27:32*JdGordon found a semi-interseting bug to work on :)
12:28:26linuxstbJdGordon: You could script it with something like "svn status | grep ^? | (while read a b ; do rm $b ; done)" (NOTE: untested!)
12:28:50pixelmayes, I thought it was not exactly the same but was reminded of it ;)
12:28:54JdGordonlinuxstb: when i asked that i didnt realise I was on the last conflict :p
12:28:56JdGordonso all good now
12:28:58JdGordonbut thanks
12:30:06 Quit Llorean (Remote closed the connection)
12:30:19bluebroth3rusing features.txt to create the opt definitions could help to make the manual match better to the code
12:30:23 Join Llorean [0] (
12:30:37bluebroth3rbut I'
12:30:49bluebroth3rm not sure how hard it would be to add this. Might look into it later
12:31:18linuxstbbluebroth3r: Maybe a manual-features.txt would be more suited though - i.e. a file similar to features.txt that creates the \opt definitions based on the #defines in config.h
12:31:56linuxstbOr is that how it works now?
12:32:06bluebroth3rhmm. Parsing features.txt would require some gcc, right?
12:32:22B4gderat least a cpp
12:32:34bluebroth3rcurrently there are those definition files in manual/platform in which the definitions are set.
12:32:49bluebroth3rand plugins are opted separately
12:33:19bluebroth3rcould we add a simple perl preprocessor to tools/ and use that for this task?
12:34:02bluebroth3rnice. That could make some parts much cleaner ...
12:34:23linuxstbWouldn't using cpp be easier though? Does anyone not have at least the host C compiler installed?
12:34:54B4gderI think that's a safe assumption
12:34:54bluebroth3rwell, that's the question: can we safely assume someone who wants to work on the manual to have at least a host cc installed?
12:35:18linuxstbGiven all the other dependencies (Make, Perl, latex), one more isn't an issue IMO
12:35:19bluebroth3rif it's a safe assumption then we can of course use that. Even easier ;-)
12:35:43linuxstbEspecially as it's just ticking a box in setup.exe
12:35:44bluebroth3rlinuxstb: I agree
12:36:08*linuxstb was thinking cygwin for some reason...
12:36:25 Quit ddalton (Remote closed the connection)
12:36:27linuxstbBut vmware and Linux users should almost definitely have a C compiler.
12:36:35 Join ddalton [0] (
12:37:02bluebroth3rI could imagine some users trying to build the manual with activestate perl and miktex installed on windows.
12:37:14bluebroth3rNever tried, so I have no idea if this would work at all.
12:37:23linuxstbWouldn't they still need make and a posix shell?
12:38:30bluebroth3ryes, but there are native windows ports. At least for make I'm sure about this
12:39:03bluebroth3rbut I think we can simply ignore this case. Quite specific and I haven't heard of someone trying to use it.
12:41:04preglowrunnin through four different codecs again and again in a playlist here, doesn't seem to bug out
12:41:12preglowamiconn: what's the crash look like?
12:41:22*pixelma is reminded once again of tex4ht and cygwin :\
12:41:56rasherpreglow: Try heating it up
12:42:15Lloreanpreglow: Data aborts and illegal instructions and random addresses.
12:42:25Lloreanpreglow: As well, it's not uncommon to have distorted audio.
12:44:11 Join Vortex375 [0] (
12:44:14bluebroth3rlinuxstb: did you succeed building Qt4 on mac?
12:44:17preglowperhaps i need to use slower codecs
12:45:21 Quit ddalton (Remote closed the connection)
12:45:32Lloreanpreglow: People who've experienced it so far have reported it with FLAC, Vorbis, and mp3.
12:45:34 Join ddalton [0] (
12:45:57preglowthose are the very same ones i'm using, plus adx
12:46:09linuxstbbluebroth3r: No.
12:46:23 Join BigMac [0] (
12:47:07amiconnpondlife: I'd prefer mary and mike over sam
12:47:12linuxstbbluebroth3r: Qt hates me - my Debian box is also refusing to install it.
12:48:35pondlifeamiconn: I'll probably go with Mike
12:48:37amiconnpondlife: Oh, and with some further adjustment and the pool, it should be possible to build voice files for all targets in ~5 minutes on windows (one language)
12:48:39bluebroth3rI just found some entries in their installation faq −− have you noticed that?
12:49:14amiconnrasher: even plug-ins is sufficient
12:49:32amiconnOne entry in my old vbscript's tables does that
12:49:34linuxstbbluebroth3r: Ah, thanks. This would appear to be my issue -
12:49:53rasheramiconn: okay, should certainly be added to the script - possibly even for all engines (I seem to recall festival pronouncing it plooog-ins)
12:50:25amiconnNot for all engines. at&t speaks "plugins" fine without adjustment
12:51:11markunrasher: does festival sound better than espeak + mbrola?
12:51:23rashermarkun: no idea - what's mbrola?
12:51:36rasherPersonally I'm not a fan of espeak's somewhat metallic sound
12:51:53 Part ddalton
12:52:17 Quit My_Sic (Read error: 104 (Connection reset by peer))
12:52:19 Join ddalton [0] (
12:52:26rasherPerhaps the voice setup should include something like echo -e "test\ntest.wav\n\n\n" | cscript ../tools/sapi5_voice_new.vbs to check if the script can produce clips
12:52:40preglowif i have a small number of files in a playlist and set repeat on, the playback engine will just buffer all files again and again instead of reusing the data in the buffer, yes?
12:52:45Nico_PJdGordon: if you have nothing to do, I suggest you look here:;h=mob :)
12:53:10Lloreanpreglow: Yes.
12:53:23LloreanAt least there's a long, long unclosed task in the tracker relating to that
12:53:24markunrasher: mbrola is a diphone speech synthsizer with many voices
12:53:35JdGordonNico_P: maybe.... maybe...
12:53:43*JdGordon loves fixing simple bugs :p
12:54:06markunrasher: so with it espeak will just be used for the phoneme, pitch and duration info and should sound very different
12:54:45markunrasher: not sure how difficult it is to setup and how good it sounds
12:55:02amiconnmarkun: I created an espeak voice for my recorder for comparison. It sounds less distorted than festival, but very robotic instead...
12:55:33markunamiconn: using mbrola?
12:55:57amiconnNo. I didn't even know that it can use that
12:56:10rasherI found something really strange in Debian: epos - Text-to-speech system. Currently supported languages are Czech, Slovak, Latin and Vogon.
12:56:22rasherI scared to even try it
12:56:30markunshould make it sound less robotic since the mbrola databases are created with real voice recordings
12:56:44scorchewe have to try vogon now...
12:56:58markunas long as it's not poetry :)
12:57:07scorcheof course ;)
12:57:50bluebroth3rsomeone write a vogon lang file ;-)
12:58:13JdGordonyou mean bork bork... dont you?
12:58:16*linuxstb is surprised there are no "geek" translations in Rockbox yet - we're obviously attracting the wrong (or right?) kind of user
12:58:19rasherI believe someone once asked about Klingon
12:58:21amiconnSpeaking about lang files - I wonder whether we'll ever see japanese on the player...
12:58:52amiconnThat'd be possible now, using player: <katakana-string>
12:59:08linuxstbrasher: I was tempted to run english.lang through this site...
12:59:28markunlinuxstb: at least we have esperanto
13:00:13markunpreglow: I wonder if the 2 people who understand lojban even have rockbox capable daps..
13:00:16scorcheamiconn: well, there is always romaji as well
13:00:23bluebroth3rlinuxstb: do it! :D
13:01:04rasherLooks like the esperanto translation was a "shoot and run" deal
13:01:04preglowwell, it doesn't look like my nano is susceptible
13:01:07preglowi'll try raising the freq, then
13:01:32JdGordonmarkun: i know one of those 2 ppl and he has an h10.... but doesnt like rockbox!
13:01:34amiconntlhIngan_Hol.lang ;)
13:01:39markunrasher: yes, the guy who made it no longer uses rockbox "because he has an iaudio X5 which is much better than his ipod with rockbox"
13:01:41JdGordonhe is the biggest geek i know and doesnt use rb!
13:02:17B4gderisn't that a contradition if anything?
13:02:28B4gder(minus the spelling errors)
13:02:33Lloreanrasher: So with mbrola we get the better pronunciation (in my opinion) of espeak, but with better sounding voices?
13:03:13rasherLlorean: surely you meant to ask markun about that?
13:03:26Lloreanrasher: Indeed I did.
13:03:28markunLlorean: that would be the idea yes
13:03:50markunbut only works for english so far I think
13:03:57LloreanAs I agree, the espeak voice doesn't sound great, but I do think it has the better pronunciation (which is why I lean toward it, that's what grates on my ear more)
13:03:59rasheramiconn: I think we'd have to start with tlhIngan_Hol-ascii.lang
13:04:22*bluebroth3r leaves for food
13:04:25 Quit bluebroth3r ("bbl")
13:04:40scorche/kick Nico_P UGH
13:05:07Nico_Pscorche: notice the quotation marks :)
13:05:12preglowamiconn: don't need to do anything else but change CPUFREQ_MAX?
13:05:29JdGordonamiconn: oh, hows the cpu sleeping stuff going?
13:06:01linuxstbamiconn: hows the 1g/2g bootloader - are you near a release?
13:07:38 Quit BigMac (Remote closed the connection)
13:08:08*Llorean goes to try out the mbrola voices
13:08:24JdGordonstupid red delta!
13:10:01markunNico_P: not as bad as your quote, but people writing like this also piss me off:
13:11:16Nico_Pmarkun: yeah, that is pretty aweful to read
13:11:47scorchesounds like a good reason why i dont patrol the unsupported builds beat
13:12:21Nico_PI don't either and I think I'm better for it
13:12:46markunscorche: I was searching for posts with 'words' like "ne1" and "wut" ;)
13:13:09*Nico_P sometimes regrets having written the album art patch :)
13:13:24Nico_Pwhat does "ne1" mean ?
13:13:42B4gderan e one
13:14:44preglowi'm running at 100mhz, with full eq, crossfeed and dithering now, so it should be hot
13:15:36scorchei am surprised none of the unsupported builds bump up the clock just to get teh "overclocking coolness factor"
13:15:45rasherpreglow: Apply a hairdryer
13:15:59scorcherasher: i did that last night ;)
13:16:16rasheramiconn: might cscript /U be interesting?
13:16:36rasherOr rather, it's probably a necessity?
13:17:28 Quit khalid ("Ex-Chat")
13:18:31 Quit YouCeyE (Read error: 113 (No route to host))
13:18:32preglowrasher: don't own one
13:18:46 Quit ddalton (Remote closed the connection)
13:18:47amiconnpreglow: You need to change the pll values for CPUFREQ_MAX as well
13:18:52 Join ddalton [0] (
13:18:59*JdGordon wonders why the nano doesnt think its charging? its plugged into the ac powerpack
13:19:00preglowrasher: and if i need to use a hairdrier to debug, i'm not doing this :D
13:19:38amiconnJdGordon: cpu sleeping in itself works, but I didn't get any further with shutdown/usb/suspend thread synchronisation
13:19:45 Quit ddalton (Remote closed the connection)
13:19:50 Join ddalton [0] (
13:20:03amiconnlinuxstb: I could release a bootloader today, but I want to get suspend into place first
13:20:30JdGordonamiconn: is there a patch somewhere?
13:20:32 Quit ddalton (Remote closed the connection)
13:20:37 Join ddalton [0] (
13:20:41amiconnJdGordon: What patch?
13:20:48JdGordonfor cpu sleeping?
13:20:58 Quit spiorf (Read error: 110 (Connection timed out))
13:21:04preglowamiconn: eh, seems the control status register format isn't described in system-pp502x.c anymore, do you have a value i could use?
13:21:43amiconnThe register format was never described there. Holdd on...
13:21:52*amiconn should really wikify all the collected info
13:22:27preglowmight have been so in the form of code
13:22:47preglowat least there was something more descriptive than just constants
13:22:53preglowi'm quite sure since i put that code there
13:23:23*Llorean can't get the mbrola voices working. :(
13:24:37 Quit phalax ("Ex-Chat")
13:25:13amiconnpreglow: system-pp502x.c, change line 139 to: PLL_CONTROL = 0x8a021906; /* 25/6 * 24MHz */
13:25:41 Quit ddalton ("leaving")
13:26:55amiconnpreglow: To verify what clock you're running the PP at, go to 'View HW info'. There's an estimation based on loop counting
13:26:55 Join datasleep [0] (
13:27:27 Quit datachild (Nick collision from services.)
13:28:54preglowreally, now
13:29:04preglowi will test right now
13:30:25preglowline 141, you mean? :)
13:32:02amiconn'svn up' refuses to work...
13:32:37amiconn"svn: Can't connect to host '': Connection refused"
13:33:09B4gderzagor's upgrading the server atm
13:33:12B4gderI bet that's why
13:33:48Zagorcorrect, sorry I forgot to warn
13:34:04preglowesimation is still 81804 :/
13:35:11preglowi'm an idiot, brb
13:36:20preglownow estimation is 81799, hrmf
13:37:37preglowargh, what the hell is wrong with me
13:38:00Zagorupgrade complete
13:38:32preglowamiconn: ata error: -11 on startup
13:39:05preglowthis does king of strengthen your hypothesis, i guess
13:44:01Zagor(now upgrading
13:44:03 Quit saratoga ("CGI:IRC (Ping timeout)")
13:45:01pondlifeI like the ???!!??
13:45:27pondlifeSorry - just having a Monday moment on a Tuesday
13:45:48B4gdernonono, just go the corner!
13:46:03*B4gder adds a "to" in a random place in the above line
13:46:17*pondlife goes the to corner
13:47:15scorcheZagor: so what is getting upgraded this go-around?
13:47:53Zagorjust a standard debian dist-upgrade. "349 packages being upgraded"
13:48:14B4gder"14 petabytes being downloaded"
13:48:48Zagoractually I run the apt-get download nightly to avoid a long wait when actually performing it
13:49:20scorchewhat do the servers run?...stable?
13:49:32Zagorno, testing
13:49:56JdGordontesting on a production server! :O
13:50:08scorcheJdGordon: debian testing ;)
13:50:23Zagortesting is very stable. it's not the development branch. (which is called unstable)
13:50:40*amiconn needed to go unstable on amd64
13:50:50B4gderdebian testing has a nice trade-off for stability and recent software
13:50:53Zagoryeah I run unstable on all my desktops
13:51:00amiconntesting had too many unresoled dependencies
13:51:09B4gderfor anything I hack on daily I use unstable
13:51:19Zagoramiconn: unresolved??
13:51:23Zagorthat shouldn't happen
13:51:31amiconnShouldn't != doesn't
13:51:39B4gderunstable frequently goes dependency craziness
13:51:48Zagoryes, but not testing
13:51:53amiconnYes, but only for some packages, and a few days
13:51:57 Quit pondlife (Read error: 104 (Connection reset by peer))
13:52:00B4gdersince that's kind of the point of unstable => testing
13:52:02amiconntesting had dependency problems for months
13:52:07Lloreanmarkun: Have you tried the mbrola voices?
13:52:12rasherYou get odd periods of strangeness with testing. Sometimes if something breaks, it can take a while for the fix to trickle down to testing
13:52:21Zagoramiconn: that's news to me. what packages was that?
13:52:22JdGordonhow often is the unstable branch updated?
13:52:23preglowamiconn: any ideas about the ata error?
13:52:30B4gderJdGordon: all the time
13:52:41amiconnZagor: I don't remember, since I went unstable to resolve the problem
13:52:44ZagorJdGordon: every day
13:52:44B4gderJdGordon: the unstable changes trickle down to testing after a grace period
13:52:53Zagoramiconn: ok
13:52:55amiconnunstable sometimes also has dependency problems, but only for a few days
13:53:02Lloreanrasher: Disliking espeak as you do, interested in hearing the mbrola voices?
13:53:05markunLlorean: I've used them with festival in the past
13:53:15rasherLlorean: certainly - got a voicefile I can try?
13:53:19scorcheJdGordon: maybe this should clear things up:
13:53:34Lloreanrasher:"> <−− Just a couple sample wavs. The standard espeak, and then a male (uk) and female (us)
13:53:37JdGordonI'm doing a reinstall _soon_ and was probably going to go gentoo... but might just do debian for the less painfullness :p
13:54:40markunLlorean: I can't listen to them here at work. Do you think they sound ok?
13:54:46LloreanThe version of espeak that ubuntu installs didn't include the necessary bits for mbrola support to work (the folders were there but didn't contain the voice stubs) so I've been working on it.
13:54:53Lloreanmarkun: They sound much, much better than standard espeak.
13:54:54ZagorI'm a major debian fan :)
13:55:34preglowamiconn: also, what's the point with the udelay on pp5022 before waiting for relock? can't you just go straight to wait for relock?
13:55:36B4gderI'm a debian fan, at least when they don't torture my sonames... :-/
13:55:43markunLlorean: don't know about the mbrola license if we want to use it for our generated voices
13:55:50amiconnpreglow: Nope, that makes some PP5022s crash
13:55:53scorcheLlorean: wow...espeak is /really/ "metallic"
13:55:56preglowamiconn: wow...
13:55:57rasherB4gder: Haha, libcurl has been seing quite some turbulence lately
13:55:57Lloreanmarkun: They're free to use, but not free to distribute the output of?
13:56:01amiconnLooks like the locked bit doesn't get disabled immediately
13:56:12B4gderrasher: that's all debian's fault, _really_
13:56:20preglowamiconn: does retailos wait too?
13:56:20 Quit iamben (Connection reset by peer)
13:56:24Lloreanscorche: That's espeak without any tweaking of pitch etc, but yes, it's quite metallic but has much better pronunciation than flite and festival (in my limited experience)
13:56:31Lloreanscorche: But the mtest ones sound much better, no?
13:56:33ZagorB4gder: hehe, yeah i'll admit there are some flaws in the system :)
13:56:34amiconnpreglow: yup
13:56:44scorcheLlorean: much
13:56:44rasherLlorean: The pronounciation seems much better - far better than festival or flite, but I still don't like the metallic sound
13:56:51preglowamiconn: but yeah, any pointers on how to fix this ata error on boot?
13:57:12rasherLlorean: silly me, I was listening to the etest one...
13:57:19Lloreanrasher: etest is the bad one. :)
13:57:44rasherYeah, the mtest one is quite good
13:57:51amiconnprashantv: The ata timing registers need to be figured out, and then ata.c needs code to adjust timing properly
13:57:56amiconnErm preglow
13:58:01preglowlinuxstb: nice...
13:58:21Lloreanmarkun: They have no posted license for the output that I can see. :(
13:58:27LloreanOnly licenses for the databases and the program.
13:58:33Zagorscorche: that debian cycle image was cruel :)
13:58:38linuxstbpreglow: ;) I should get back to work now...
13:58:45scorcheZagor: very ;)
13:58:45Lloreanmarkun: Shall I email them?
13:59:02scorcheZagor: gives one a headache...
13:59:24markunLlorean: I'll first search a bit more
13:59:56 Part Zagor ("Client exiting")
14:00:06 Join Zagor [0] (n=bjorn@rockbox/developer/Zagor)
14:00:10Zagoroops, wrong window
14:00:35scorcheand you are the one updating? =P
14:00:51Zagoroops, wrong partition! ;)
14:01:58Lloreanmarkun: I was just thinking of something simple like this:
14:02:40markunLlorean: looks good
14:02:52markunI also can't find anything on using the generated sound clips
14:03:08LloreanYeah, I imagine it'll be "feel free" since the program is distributed freely so long as it's not sold.
14:04:56 Quit qwm (Remote closed the connection)
14:04:59 Join qwm [0] (
14:05:44peturB4gder: did you see :)
14:06:38B4gdernow coding truly is an art...
14:06:52peturthat is what I call graphics coding :)
14:07:11B4gdernow do that with the 'cat' source code!
14:07:11Lloreanmarkun: Or, apparently, I could try to send it and have it bounce. =/
14:07:19***Saving seen data "./dancer.seen"
14:07:54 Join Naked [0] (
14:08:10 Nick Naked is now known as Hadaka (
14:08:33markunLlorean: maybe you can try to email the person who created the database(s)
14:08:41markunehh, wrong link
14:10:35LloreanReading the site further, they say you're free to produce programs like a talking clock on top of mbrola, so long as it's non-commercial
14:10:44LloreanI'm almost positive we're covered by that.
14:11:45LloreanAs well, their terms make the database copyright the mbrola team anyway, independent of who created it.
14:11:59LloreanOnly the mbrola-format version of it, but still that's what we'd be using
14:18:00markunLlorean: perhaps you could generate a rockbox voice file using espeak+mbrola for people to try
14:19:15LloreanI'll have a look at it. It uses a different command line, and I'm not that clever with scripting so I may get it wrong a few times. :)
14:19:31 Quit Febs (Read error: 110 (Connection timed out))
14:19:39LloreanWow, espeak is fast
14:19:56 Quit davina (Remote closed the connection)
14:20:36LloreanIt took 11 seconds to read Alice in Wonderland with mbrola voices.
14:20:54markundoes it sound ok?
14:21:19LloreanOther than issues with line length, I think, yes.
14:21:50markunif treats the end of the line as the end of a sentence?
14:22:15LloreanI'd set a command line switch I didn't need, was all
14:22:38LloreanRemoving it fixes the problem. It's still imperfect, obviously, but it works well enough
14:23:14markunimperfections can be fixed. What mostly anoys me is "read" vs "read"
14:23:31JdGordon.... which doesnt work in irc :p
14:24:50markunwhere to submit a bugreport for bugs in the english language..
14:25:07B4gdermarkun: you can send them to me
14:25:15Lloreanmarkun: espeak+mbrola gets it right.
14:25:24Llorean"Have you read this book? Do you want to read it with me." works properly
14:25:49markunhe must have fixed that recently
14:27:02LloreanThe only problem with the mbrola voices is that they require a separate binary on linux that you have to pipe espeak's output through, and since it's binary-only it wouldn't work for an espeak plugin.
14:28:37 Join davina [0] (
14:28:58 Quit CaptainSquid ("Miranda IM!")
14:29:20markunLlorean: but not a show stopper
14:33:52LloreanShould still be usable with pre-generated voices, which is what we've got now anyway.
14:34:11linuxstbAnyone (B4gder) know what CPU the Sansa Express or the new Sansa Clip are using?
14:34:39B4gderI don't
14:35:09markunLlorean: yes, it's no option to use mbrola with espeak as a rockbox plugin
14:36:44markunor were you not talking about that?
14:36:46Lloreanmarkun: Do you know where the actual command line invoking espeak is?
14:36:53LloreanIn the voice generation? it seems.
14:39:17markunyes, line 280 I think
14:39:42LloreanI thought 281
14:39:46LloreanI'm not sure what 280 is actually doing
14:40:02markunprobably just displaying what is being executed
14:41:06LloreanAlright, attempting to make a voice with the mbrola voices.
14:43:39 Join jackcatess [0] (
14:44:20jackcatesshaving a problem when i plug in the ipod to the radio transmitter in the car
14:44:38jackcatessit goes into usb mode
14:46:15markunjackcatess: sorry, don't know anything about that
14:46:20linuxstbjackcatess: Are you using a recent and official version of Rockbox?
14:46:56jackcatessjust upgraded from a few months ago to current
14:47:04 Join bluebrother [0] (i=jFX7HNIo@rockbox/staff/bluebrother)
14:47:46linuxstbWhat kind of connection is it to the radio transmitter? And what radio transmitter is it?
14:47:57jackcatesscupholder transmitter
14:48:05jackcatessconnects to bottom
14:48:14jackcatesswhatever that's called
14:49:07linuxstbAnd it worked fine with the older version of Rockbox? Was that also an official build?
14:49:09markunI only found this:
14:49:24LloreanAlright, I apparently have no clue how to properly modify
14:49:30*Llorean knows nothing about shell scripting, at all.
14:49:35markunLlorean: what happened?
14:49:39linuxstbjackcatess: Also, do you have the hold switch on?
14:49:42LloreanI got a 370k voice file?
14:50:05markunLlorean: and what does line 281 look like?
14:50:51LloreanWait one second, I think I see the problem
14:50:58 Quit Rondom ("Ex-Chat")
14:52:01jackcatessnot sure what you mean, hold the switch on
14:52:22linuxstbThe "hold switch" - the switch on the top of your ipod.
14:53:07rasherLlorean: voicecommon is deprecated..
14:53:08bluebrotherthe hold switch, not hold some switch
14:53:14bluebrotherthe manual has a nice image describing how each button / switch is named.
14:53:25jackcatessipod always acted this way with rockbox. doesn't go into usb mode when i put it on the transmitter in apple ipod mode only when using rockbox
14:54:01jackcatesshold switch boots apple mode
14:54:08markunrasher: what's being used now?
14:54:09Lloreanrasher: Ah, well that's the first problem then
14:54:32Lloreanrasher: I'm looking for where espeak is actually invoked.
14:54:40 Quit qweru ("moo")
14:55:00rashermarkun, Llorean:
14:55:08rasherEverything happens in there
14:55:20*bluebrother tries the aol lang file and laughs :)
14:55:38rasherbluebrother: now do an aol voicefile!
14:55:43rasherrecorded from XBOX live
14:55:58bluebrotherwhy, is there an aol tts on the xbox?
14:56:04*linuxstb left the voice: strings alone...
14:56:11*Llorean attempts again
14:57:07LloreanAh, this appears to be working
14:57:22linuxstbjackcatess: I'm not sure what to suggest. There was a bug in Rockbox in the past which incorrectly detected a "usb charger" connection as a "usb connection to a computer", but that was fixed about 6 months ago. As far as I know, no-one has reported problems with that feature since then, apart from you...
14:57:50linuxstbjackcatess: Which ipod do you have?
14:58:07 Join Febs [0] (n=chatzill@
14:58:44linuxstbjackcatess: If you go into the "System -> Rockbox Info" menu, what does it report as the version of Rockbox you are running?
14:58:48jackcatessi have 30 gig vidoe ipod
14:59:11DerPapstyay for AOL kiddies :D
14:59:56Lloreanrasher: Alright, with a minor modification of it appears I can make voices properly. Testing the file shortly
15:00:15*DerPapst wonders when aoler.lang makes it into svn :P
15:00:37linuxstbDerPapst: If I commit it, I would have to maintain it...
15:00:50linuxstbBut then again, it's 100% automated...
15:01:18bluebrotherlinuxstb: script-generated?
15:01:34obolinuxstb: svn post commit hook to regen it every time english.lang is updated ;)
15:01:38linuxstbjackcatess: That's the current version, so I'm not sure why it's not working. Unless something is odd about that transmitter that makes Rockbox think it's a computer...
15:02:22linuxstbbluebrother: Yes. That AOLer translator is written in Javascript. so I ran it locally in a script via Rhino.
15:03:15jackcatessthanks guys
15:03:36DerPapstlinuxstb: how did you automate that?
15:03:53DerPapstnvm :P
15:04:02*DerPapst too slow
15:04:31 Quit Thundercloud (Read error: 104 (Connection reset by peer))
15:04:33markunLlorean: working now?
15:04:41*bluebrother looks up Rhino
15:05:03*DerPapst guesses
15:05:11*GodEater thinksl linuxstb should commit it
15:05:33DerPapstlet's vote for it ^^
15:05:43Lloreanmarkun: Indeed, everything seems to work.
15:06:59 Quit jackcatess ()
15:07:48rasherLlorean: Feel free to update
15:08:07 Join Nibbl [0] (
15:08:22 Quit Nibbl (Read error: 104 (Connection reset by peer))
15:08:31Lloreanrasher: I'll have to figure out the right way to do it.
15:08:50LloreanIt'd be quite easy to add espeak+mbrola as its own TTS engine, I suppose.
15:08:59LloreanBut I imagine there's a more elegant way
15:09:23 Join Nibbl [0] (
15:09:33 Join jgarvey [0] (
15:09:48 Quit Nibbl (Client Quit)
15:10:15markunLlorean: I'm really curious about the quality
15:10:45Lloreanmarkun: I'm trying to make voices for multiple targets now, if I can figure out how to wrangle this thingy
15:10:58morrijrhave you guys seen this ->
15:11:21*GodEater watches dapreview collapse under the rockbox effect
15:11:34GodEaterooh yummy
15:11:38GodEatergigabeaty goodness
15:11:45bluebrotherit has been rockboxed? ;-)
15:12:28Lloreanrasher: You... might want to have a look at the you posted to the patch tracker. It's still got some lines in it that are kinda useless (to other people), like an ftp upload.
15:13:35Nico_PI like the button configuration
15:14:18 Join aliask [0] (
15:16:19 Nick hannesd_ is now known as hannesd (
15:17:39 Join Thundercloud [0] (
15:19:02 Join ie [0] (i=c27f0814@gateway/web/cgi-irc/
15:20:25ieRB files have the following in the header: "All files in this archive are subject to the GNU General Public License". Shouldn't the license version be explicitly specified?
15:22:02Lloreanmarkun: Alright, I can't get the script that makes voices for all targets to work properly. =/
15:22:18LloreanAnyone have a specific request for a few targets for me to make mbrola voices for?
15:22:33 Quit aliask ("ChatZilla [Firefox]")
15:22:40linuxstbie: I think so. But as we haven't done it so far, it's too late. The version of COPYING included is v2.
15:23:14ieShould we do it for new files?
15:23:49ieOr better consistently wrong? Consistence is a good thing
15:24:53 Join miepchen^schlaf [0] (
15:25:19markunLlorean: I would like to try it on my gigabeat
15:25:43markunamiconn: any target you would like to try it on?
15:25:56*DerPapst votes for the target that gets the most downloads
15:26:14markunDerPapst: ipod video?
15:26:24*linuxstb wonders if that's the same target DerPapst is using...
15:26:44markunhe's such an apple fanboy ;P
15:27:40LloreanGimma about 5 minutes to work a few things out
15:27:49LloreanThen I'll put up an -f40 and a -ipvideo .voice file
15:30:19 Join seablue [0] (
15:30:26 Part ie
15:31:38 Nick kk is now known as krazykit (n=kkit@
15:32:36Lloreanmarkun: Okay, at there should be an f40 english file
15:33:24LloreanIt's worng
15:33:35Lloreanthat's using normal espeak, I accidentally broke something. :(
15:35:33 Quit DerPapst (Read error: 104 (Connection reset by peer))
15:35:49LloreanI had conflicts when SVN updating since I still had the USB patch on, and I was a little overvealous in reverting. Ah.
15:36:01markunbut it's easy to fix, right?
15:36:12LloreanOh yes
15:36:21LloreanIt's a very small change to now that I know where to make it
15:36:28 Join DerPapst [0] (
15:37:53Lloreanmarkun: *now* the one posted is using mbrola
15:38:45markunthanks, I'll try it when I get home from work
15:38:57 Join elinenbe_ [0] (n=elinenbe@
15:39:51markunLlorean: will you commit with espeak+mbrola added?
15:40:18Lloreanmarkun: Right now it's a quick hack. I'm not sure how to properly add it.
15:40:28amiconnmarkun: archos recorder
15:40:30LloreanThe best would be a change to configure.
15:40:45LloreanNot something I'd be able to do well.
15:40:58Lloreanamiconn: v1?
15:41:20Lloreanmarkun: It would need to prompt for the location of the mbrola binary, as well as the location of the actual voice file.
15:42:24markunthat was fast
15:42:38LloreanI'm using the pool
15:42:50 Quit JdGordon ("Konversation terminated!")
15:43:25B4gderthe pool rocks
15:43:33LloreanIt speeds things up drastically, yes.
15:44:03LloreanDoes the pool handle if a string is changed? For example if a clip is generated with a typo, the typo is fixed?
15:44:08amiconnfirefox tries to display it
15:44:59markunamiconn: something for you?
15:45:05amiconnAh, directory listing is allowed :)
15:45:26GodEaterI can use mbrola with festival
15:45:29GodEaterbut not espeak
15:46:18markunwhy not?
15:46:23LloreanGodEater: Which version of espeak are you using, and more importantly, have you checked the espeak-data folder to make sure the mbrola folder isn't empty?
15:46:43LloreanMine was empty for some unknown reason, so I ended up getting the binary version of espeak from the website to fix it.
15:48:00*elinenbe_ yells at elinenbe
15:48:01LloreanGodEater: Oh, wait, /usr/share/espeak-data/voices/mb
15:48:08 Nick elinenbe_ is now known as elinenbe_work (n=elinenbe@
15:49:06GodEaterah - I take it back
15:49:13GodEaterespeak is built with mbrola support by default
15:49:18GodEaterfestival has it as an option
15:54:55 Join BobShield [0] (
15:55:21LloreanB4gder: What TTS do we use for official voices right now?
15:56:03B4gdermy plan is to provide voices build with all the ttses we have on linux
15:56:35LloreanYou should skip normal espeak for espeak+mbrola, assuming you agree that there's no license problem.
15:56:45LloreanSame pronunciation, much much less metallic voices.
15:57:15markunLlorean: do we want a female? :)
15:57:24B4gderI won't get around to do that for a while anyway
16:01:21elinenbe_workwhy is GPL3 not compatible with rockbox?
16:01:35LloreanBecause GPLv3 isn't compatible with GPLv2
16:02:16bluebrotherhaving multiple versions of the voice files will require rbutil to let the user choose ...
16:03:15bluebrotherbut that voice sounds much nicer than espeak
16:03:16elinenbe_workwell, what's wrong with moving rockbox to GPLv3?
16:03:33bluebrotherall contributiors would need to agree?
16:04:20markunbluebrother: which voice?
16:04:36bluebrotherthe one Llorean provided. I tried the video version in the sim
16:04:47*bluebrother was really unprecise :/
16:05:05elinenbe_workI would think it would be impossible for any project to move to GPLv3 then...
16:05:09 Join midgey [0] (
16:05:21Lloreanelinenbe_work: It's very easy if your project is under a "2 or later" license.
16:05:31LloreanBecause then everybody has already agreed to allow you to move to 3 if you want to.
16:05:42markunor if people give away the copyright to the project
16:05:44 Quit Thundercloud (Remote closed the connection)
16:06:08elinenbe_workI say you create a developer poll.
16:06:10 Quit Zagor ("Client exiting")
16:06:12 Quit B4gder ("It is time to say MOOO")
16:06:15markunor can't "the rockbox foundation" own the copyright? (and sell it to an evil corporation)
16:06:29Lloreanelinenbe_work: Why do you think Rockbox needs to be GPLv3?
16:06:42elinenbe_workto satisfy the espeak license :-)
16:06:46Lloreanelinenbe_work: And how do you intend to get 100% of past developers to respond?
16:07:16elinenbe_workLlorean: well, that's why I was saying it would be impossible to move any project to v3
16:07:23***Saving seen data "./dancer.seen"
16:07:24markuncould we then still use GPLv2 stuff such as libmad?
16:07:32linuxstblibmad is v2 or later, so yes.
16:07:35elinenbe_worklet's say someone dies?
16:07:38Lloreanelinenbe_work: Unless you got them to agree in advance, as I said.
16:07:46LloreanThat's the whole point of "v2 or later" style licenses.
16:07:50linuxstbelinenbe_work: Which is not a hypothetical situation sadly.
16:07:57Lloreanlinuxstb: Is our "COPYING" stricly 2, or 2 or later?
16:08:02elinenbe_workNo, I know...
16:08:05linuxstbI think the COPYING is v2.
16:08:10GodEatermbrola is v. impressive!
16:09:06markunGodEater: best of the 4 so far I guess? (festival, flite, espeak, espeak+mbrola)
16:09:19Lloreanlinuxstb: Yes, the only mention in copying of later versions is where it says "If the program itself says 2 or later"
16:09:52GodEaterLlorean: which mbrola voice did you use ?
16:10:11 Join Cillian [0] (
16:10:26GodEaterI don't appear to have that one
16:10:29amiconnImo our headers aren't precise enough. They just say "gpl" and then refer to COPYING
16:10:35CillianAnybody else have the very curious problem that menu+select to reset their iPod doesn't work when the ipod has a case on?
16:10:38GodEaterunless they're called different things here
16:10:39Lloreanamiconn: And then copying refers back to the program.
16:11:03LloreanGodEater: mb-en1 for espeak, but the filename you point mbrola to is the en1 download from their site.
16:11:11LloreanWell the "en1" file in the en1 download
16:11:48GodEaterI have three voices to choose from - us1, us2 and us3
16:11:53GodEateralthough I can't get us3 to make any noice
16:12:10 Join Thundercloud [0] (
16:12:16linuxstbLlorean: I don't think that it would be impossible to switch to v3. If the majority of currently active developers are happy to go v3, and there is a practical need for it, then I would hope all inactive devs would agree. Although we would probably end up with a small amount of code where the copyright holder isn't contactable, which would need to be either removed or rewritten.
16:12:41Lloreanlinuxstb: I'm sure if the majority wanted to move, the minority would accept it for the greater good, yeah.
16:13:35LloreanGodEater: This is the voice I used:
16:14:27elinenbe_workwell, you can move sokoban to v3 :-)
16:14:29GodEaterLlorean: yeah - I just downloaded that too - us1-3 seem to come with my copy of mbrola already
16:14:44elinenbe_workscrew them TiVo people!
16:15:02GodEaterI think us1 > en1
16:15:18GodEateren1 is still slightly metallic
16:15:53LloreanI honestly didn't like us1. Is your version female?
16:16:00LloreanIt sounded kinda weird to me.
16:16:36GodEateryep - female
16:16:45GodEaterwell en1 certainly doesn't sound British
16:17:07LloreanIt would to many Americans.
16:17:22GodEaterin the same way Dick van Dyke does no doubt ;)
16:17:29LloreanThere's enough in the pronunciation to hint at it, I think
16:17:48LloreanDid you do mb-en1 or mb-us1 when you pointed mbrolo at the en1 file?
16:18:18 Join `sam` [0] (
16:18:21LloreanI wonder if you can mix pronunciations like that.
16:18:28GodEateronly one way to find out...
16:18:39 Part Cillian
16:18:44GodEaterlots of errors
16:18:46LloreanI know that it offers mb-de1-en for "English with a german accent"
16:18:46GodEaterand broken speech
16:19:29*GodEater HAS to try the afrikaans accent now
16:20:51GodEaterthat's hysterically bad
16:21:29GodEaterneat - the swedes get too attempts at english too
16:21:52LloreanClearly we need a swedish chef voice file.
16:22:03GodEaterit's almost mandatory
16:23:34DerPapsteh? English with a german accent? if i hear myself talking english i have enough of that already. I don't want my ipods to do the same ^^
16:23:51Nimdaeis there a reason to update the bootloader on my ipod if i haven't updated it in a few months?
16:24:10bluebrotheronly if there was a change
16:24:19bluebrothercheck the MajorChanges page
16:24:19GodEaterNimdae: only if it's not version 1.1 already
16:24:34bluebrother(and if you experience strange crashed update to bootloader also)
16:24:59Nimdaehow would i find the version? just curious because ipodpatcher no longer detects my ipod
16:25:34LloreanNimdae: Is this the official ipodpatcher binary, or one you've built yourself?
16:25:45Nimdaeofficial ipodpatcher
16:25:53bluebrotherhold right while booting
16:26:03LloreanNimdae: Is your fat32 partition 0x0b or 0x0c type?
16:27:33Nimdaei didn't partition it myself so whatever apple sets it to
16:27:49Nimdaebluebrother: unfortunately it goes by too fast to catch it (i must be getting old :( )
16:28:39LloreanNimdae: iTunes doesn't seem to be wholly consistent in this behaviour (for example if you've reformatted it yourself in the past, iTunes doesn't seem to necessarily change it back to what it originally was)
16:29:24Nimdaewell, i haven't touched partitioning, but i've used mkfs in linux to format it, i wouldn't figure that to change the partition flag
16:29:46linuxstbNimdae: Does the Apple firmware (or Rockbox) boot on your ipod?
16:29:47Nimdaei haven't have changed anything other than music files in the past year or so
16:29:54 Join kk [0] (n=kkit@
16:29:55Nimdaeyeah, rockbox boots fine
16:30:29Nimdaei was going to update it, i have a firmware from july 7 on there now
16:30:32linuxstbDoes the bootloader display any text, or just it just go directly from the Apple logo to the Rockbox logo?
16:30:38linuxstb^does it just
16:31:30Nimdaeit shows the version at the bottom with the rockbox logo, then into rockbox
16:32:17Nimdaelemme try resetting my docking station for my does weird things with usb sometimes and resetting it tends to work
16:33:59`sam`should the simulator play the song when it's selected?
16:34:12 Join obo_ [0] (n=obo@rockbox/developer/obo)
16:37:47Nimdaeyeah, that didn't work
16:37:56Nimdaei probably need to completely reset the ipod
16:39:02Nimdaei'll just update rockbox for now
16:40:11Nimdaeso i saw in the updates the 1g-3g battery life in rockbox exceeds the orgiinal firmware? any effect on 5g?
16:40:47preglowdamn, it's pretty wild we exceed retailos battery life...
16:40:50 Quit krazykit (Read error: 110 (Connection timed out))
16:41:10 Join toffe82 [0] (
16:44:14preglowbah, flyspray reposts a message if you refresh after having posted one
16:44:49Nimdaeah, well, yall will get there on the 4g and 5g
16:45:25Nimdaewhat about the usb stuff? i see it's experimental, but does it add usb functionality, for example, to the ipod?
16:45:37Nimdaei mean more than simple detection
16:45:53preglownot yet
16:45:56preglowbut it's coming around
16:46:06preglowi think we can expect something in not too long
16:47:05amiconnpreglow: I'd say exceeding OF battery runtime is the norm for rockbox, and PP502x is the exception
16:47:17amiconnE.g. on Ondio, we exceed OF battery life by ~30% ...
16:47:26 Join shagnar [0] (i=54a0d40a@gateway/web/cgi-irc/
16:48:41 Join lee-qid [0] (
16:48:47shagnarhi folks
16:48:51markunhi shagnar
16:48:55preglowamiconn: yeah, the norm, but a surprising one on 1g-3g
16:49:05markunamiconn: on the Gigabeat it's about the same
16:49:23preglowbeing on of the targets we know the least about
16:49:29preglowand not even utilising both cores
16:49:37markunshagnar: what's up?
16:49:48amiconnPerhaps it's exactly because we're not using both cores...
16:50:04preglowamiconn: yeah, but we'll probably want to anyway, for a more responsive gui
16:50:27*amiconn dares to say that we know more about PP5002 than PP502x
16:50:49amiconn...simply because the chip is less complex, and we know about the same in absolute terms
16:50:57shagnarhas anybody tried yet to connect the battery-loader of the h140 with the usb-port to load the batteries?
16:50:59 Quit obo (Read error: 110 (Connection timed out))
16:51:14preglowamiconn: i agree with that, but what we know is still not what i'd call much
16:51:31shagnargot some big delay, just wondered if i was disconnected.
16:53:47markunamiconn: do you know how fast vorbis playback is on the 1-3gen ipods?
16:53:52linuxstbmarkun: Rockbox gets about 30% longer battery life on the Gigabeat?
16:54:27markunlinuxstb: there was a guy who reported 19 hours with the OF
16:54:32amiconnmarkun: I don't know the percentage, but 170..190kbps vorbis is realtime with enough margin to apply sw treble+bass (after my little optimisation)
16:54:51amiconnEven works with a wps showing peakmeters
16:56:00amiconnAnd we still have the option to raise CPUFREQ_MAX by another 10MHz
16:56:09linuxstbpreglow: Are you intending to buy a gigabeat S? Sounds like your itching to write floating point DSP functions...
16:57:15 Join Shagnar^ [0] (
16:58:07markunShagnar^: you mean to charge over USB?
16:58:31Shagnar^yes. to add to cables to use the +5V from the usb-port
16:58:45preglowlinuxstb: nah, i'm just making sure our lives aren't going to be harder then they have to be when the time comes, heh
16:58:53preglowno one better to make sure a codec works than the codec author
16:58:58 Quit shagnar ("CGI:IRC")
16:59:04preglowlinuxstb: i'm not even sure using the fpu will be very beneficial
16:59:30linuxstbThat's what I was wondering... Although doesn't it have vector operations?
16:59:48preglowit does, but that just smacks of hand-optimising with asm
16:59:59linuxstbBut do we really need that much optimisation anyway, unless we're decoding video...
17:00:07Shagnar^markun do you think its doable?
17:00:15preglowon such overpowered targets, i wouldn't really know if it matters
17:00:30preglowbut again, i want to make sure our codecs work in both modes anyway
17:00:35linuxstbNo, I can't think of a good use for a DAP with that much power...
17:01:00linuxstbWe're still struggling to challenge the Gigabeat F.
17:01:16markunShagnar^: maybe LinusN can tell you
17:01:30amiconnPort the dnetc client or sth like that ;)
17:01:34Lloreanlinuxstb: Yes, but we're not playing h.264 ;)
17:01:55linuxstbShagnar^: I'm sure I've seen people make that that mod to their h1x0. Have you tried searching misticriver?
17:03:22Shagnar^markun thanks a lot!
17:04:13linuxstbLlorean: Yes, h264 with AAC audio will challenge the F...
17:08:27markunpreglow: have you given up on ambisonics playback for rockbox?
17:09:43toffe82FYI the remote control of the gigabeat F is working on the S
17:09:59markuntoffe82: did the S come without a remote?
17:10:11preglowmarkun: not given up, just not tried much
17:10:20 Join wookey__ [0] (
17:10:24preglowmarkun: it's not really a huge field of interest for me
17:10:43alienbiker99s didnt come with a remote
17:10:48wookey__hi peeps. Just been reading your wiki which is full of good stuff
17:10:57linuxstbBah, so I'm doomed to life without any gigabeat remotes...
17:11:09wookey__but I couldn;t wpork out if mp2 audio is supported (mpeg1-leyerII, as used in DAB)
17:11:54markunwookey__: it is supported
17:11:54linuxstbwookey__: Yes, mp2 is supported.
17:11:55wookey__I'm wanting to buy a player for the wife, who has a lot of mp2 from DAB. I thougth rockbox was best bet for supporting this
17:12:06wookey__OK. great - on all devices?
17:12:33wookey__brilliant. A -shopping I shall go.
17:12:57markunwookey__: which one are you planning to buy?
17:13:12wookey__yes I've been studying that :-)
17:13:38wookey__FM radio is pretty-much a must so irvier h320 or iaudio X5L look best bets
17:13:42markuna player with DAB support would also be nice
17:14:06markunM5L is also nice if you don't like recharging
17:14:12wookey__are there any players that have an SD slot? (and could read FAT off it)
17:14:14linuxstbDoes the M5L have radio?
17:14:24markungood point :)
17:14:26wookey__indeed - otherwise I'd have picked it
17:14:28*linuxstb reads the buyers guide.... ;)
17:14:32bluebrotherthe sansa has a microsd slot
17:14:48bluebrotherbut IIRC the tuner isn't supported (yet)
17:14:55linuxstbI thought it was?
17:15:02linuxstbThe H10 is the unsupported tuner IIRC.
17:15:04bluebrothermaybe I missed that
17:15:18bluebrotheror I mixed it up ;-)
17:15:32pixelmathe OndioFM has an MMC slot ;)
17:15:35wookey__markun: are there any players with DAB?
17:15:41*bluebrother checks MajorChanges and notices that his memory was wrong
17:16:00wookey__(and do you get reasonable reception anywhere/most places?)
17:16:15 Join austriancoder [0] (n=austrian@rockbox/developer/austriancoder)
17:16:57wookey__DAB seems to be much pickier about reception than FM
17:17:12wookey__pixelma: close, but no cigar :-)
17:17:18 Quit webmind (Read error: 110 (Connection timed out))
17:17:38linuxstbwookey__: I have a cheap MP3 player with DAB (not supported by Rockbox though), and the DAB reception is bad - it almost never receives more than 1 or 2 multiplexes, and never the one I want to listen to (the BBC)
17:18:13wookey__that's about what I expected
17:19:16wookey__It would have been nice to just take SD cards out of DAB box and plug into mp3 player. Can you plug in a USB card reader and have rockbox read that?
17:19:39wookey__bit unwieldy, but still maybe handier than coping everything over via PC
17:20:26linuxstbNot at the moment, but that should happen in the future.
17:20:43bluebrotheryou'd need a player that supports USBOTG too.
17:20:46linuxstbBut you could probably dual-boot into the device's original firmware to do a USB transfer if it supports it.
17:21:21wookey__right - because all the players do usb client, but few do USB host
17:21:48linuxstbCan you use microSD cards in a SD slot?
17:22:13wookey__good question. Maybe an adaptor exists if they are electrically compatible
17:22:54alienbiker99yes, some even come with the adapter
17:22:56pixelmaafaik usually you can use MMC in SD slots... :P
17:23:38wookey__pixelma: sometimes. It doesn't seem to work on our radio (pure 'the bug')
17:23:59linuxstbwookey__: So the Sansa E200 would be a good choice if you can get microSD cards working.
17:25:24Shagnar^ vs - which one is right?
17:25:44amiconnMicroSD cards in an SD slot just require a mechanical adapter
17:25:56wookey__linuxstb: yep - looks good. Thanx everso for the advice.
17:26:21wookey__I look forward to joining the community and if I ever get any spare cycles I can do some hacking
17:26:45wookey__(I do embedded linux for a living so some nice simple rockbox hacking would be light relief)
17:27:30wookey__Although presumably it's like bootload-hacking - hard to tell what broke if it breaks...
17:28:05markunwookey__: depends what you work on
17:29:00wookey__presumably it's developed into its own little OS by now?
17:29:34markunwell, it has always been it's own little OS :)
17:29:45markunwithout the '
17:30:26 Quit Rob222241 (Read error: 104 (Connection reset by peer))
17:30:42 Join Rob2222 [0] (
17:32:23toffe82markun: there is no remote with the S, the only thing is to not connect it when you use the video output
17:33:01toffe82linuxstb: I have a stock of remote if you want :)
17:33:52linuxstbtoffe82: I could be tempted...
17:34:28linuxstbThey don't have an LCD though, do they?
17:34:51toffe82no, but as I bought them on ebqay I will be a littl more expensive that the original vendor
17:35:20toffe82I can give you the site were I find them
17:36:20 Join freqmod [0] (
17:37:47toffe82I am 1$ more expensive :)
17:38:00wookey__looks like you can buy an e250 or 280 but not e200. Only e200 is explicitly listed on DeviceChart but I see comments on others in forum. Are they all supported?
17:38:33alienbiker99e200 is the general name of the device,
17:38:33amiconne200 is the series
17:38:44 Quit freqmod (Remote closed the connection)
17:38:45wookey__ah, OK.
17:38:54*DerPapst wonders why there are so few Gigabeat S'es... and why toffe82 owns half of them
17:39:08linuxstbDerPapst: He's speculating...
17:39:26DerPapstguess so too :P
17:39:55linuxstbwookey__: Make sure you buy a Sansa with FM - not all of them (especially ones officially designated for the EU) don't have FM radio.
17:40:26linuxstbwookey__: With the current exchange rate, it may be a lot cheaper to buy from the US...
17:40:37 Join webmind_ [0] (
17:41:29linuxstbtoffe82: $4.99 for both the earbuds and the remote? No wonder you bought a few...
17:42:04toffe82yes, I bought 10 of them so shipping was less
17:42:11alienbiker99toffe82 did you say you got the F remote working on the s?
17:42:48alienbiker99does it just work or you had to do hardware mods?
17:42:52linuxstbtoffe82: And they only ship to "the 48 contiguous United States"...
17:43:03toffe82if you look the block diagram of the S, the pin on the jack is used by the remote or the video output, so if you don't use the video you can use the remote
17:43:25toffe82linuxstb: I ship to europ too :0
17:43:37alienbiker99oh, the remote doesnt have a screen right?
17:44:23 Join roseen [0] (i=d5f29f80@gateway/web/cgi-irc/
17:45:32roseenHi! Does anyone know where I can find the sources for itdb2tc?
17:46:02linuxstbroseen: It's on the wiki page -
17:46:03roseen(The iTunes to DataBase (formerly known as Tag Cache) Converter)
17:46:37 Join Domonoky [0] (
17:46:48roseenAaaah... now I feel really stupid, it was under the Linix section....
17:47:02roseenOK, thanks.
17:47:06 Part Domonoky
17:47:35linuxstbroseen: Are you intending to update it?
17:48:11roseenWell.... I have had problems with garbage chars showing up in my datyabase...
17:48:18 Join The-Compiler [0] (
17:48:28roseen...but I am not sure, it could be that my iTunes database is f**ed up.
17:48:41roseenAnyway, it would be fun to debug it
17:49:27roseenlinuxstb: Are there people working on it that I should know about?
17:52:03GodEaterroseen: not to my knowledge
17:52:55 Nick datasleep is now known as datachild (
17:53:34Nico_Ptoffe82: hi! have you shipped my S ?
17:54:04roseenOK, I have it now, we'll see if i make any changes worth sharing. See you later!
17:54:34 Quit roseen ("CGI:IRC (EOF)")
17:55:03toffe82Nico_P: sorry I had to work late yesterday and miss it, I send it today for sure
17:55:13Nico_Pok :)
17:55:24Nico_PI'm not in any rush anyway
17:55:44wookey__I notice the sansa port pages say that USB is not working. Does that mean you have to boot into native firmware to upload stuff?
17:55:54austriancoderwookey__: yep
17:56:09preglowaustriancoder: how far'd you say we're away from working file transfer via usb?
17:56:41austriancoderpreglow: tje
17:57:17austriancoderpreglow: the msc class should be programmable in a few days.. at the moment I am cleaning up things and adding some missing stuff
17:57:28austriancodertje... pressed ENTER to early *g*
17:57:50austriancoderat the moment I am fixing fullspeed and highspeed support
17:57:58wookey__so that might well be working in a month or two?
17:58:08wookey__(the birthday I am aiming for :-)
17:58:47preglowman, am i looking forward to that
17:59:03markundo you guys have an idea how we could chose between the various USB modes with the Gigabeat cradle?
17:59:07wookey__ah yes - austriancoder is doing a google SoC, I read somewhere
17:59:35linuxstbpreglow: Do you suffer from slow transfer speeds with the Apple emergency disk mode on your Nano?
18:01:14 Nick webmind_ is now known as webmind (
18:01:39*bluebrother wonders when usb will work on Ipods ...
18:02:10peturaustriancoder: would you manage to get something working before 31/8 (evaluation deadline)?
18:02:37linuxstbbluebrother: Shortly after it works on the Sansa... So far it seems that austriancoder's code is working fine on the ipods.
18:02:55austriancoderpetur: I am working on it... I hope so
18:03:19Nico_Plinuxstb: you fixed the ipod problems ?
18:03:50bluebrotherlinuxstb: well, apart from not working it even doesn't build for mini2g (region IRAM is full)
18:03:50peturaustriancoder: ok, I'll wait with the evaluation until then so you can still earn the money...
18:03:59petur<pressure mode off>
18:04:05preglowlinuxstb: not overly slow
18:04:11austriancoderpetur: thanks
18:04:20linuxstbNico_P: Yes - austriancoder was just a little too violent in removing old code... So I just put it back, and it worked.
18:04:37Genre9mp3petur: Do you have plans of using the sw usb stack later on to implement usbotg for h300 and maybe the x5 and gigabeat as well?
18:04:48linuxstbbluebrother: That's odd - I wonder why the mini has less IRAM free than my ipod color. Is the LCD framebuffer in IRAM?
18:05:31linuxstbaustriancoder: Do you know if there are things that could potentially be removed from IRAM, or do we need to remove existing things to make room for USB?
18:05:43peturGenre9mp3: good question... It would be better to use that stack, but the chips are not compatible at all
18:05:45austriancoderGenre9mp3: Its planned...
18:06:11*bluebrother notices that the wiki doesn't hold information about IRAM sizes
18:06:20 Join bluey- [0] (
18:06:23linuxstbbluebrother: Do you have the 1st or 2nd gen?
18:06:24Genre9mp3petur: not compatible with austriancoder's code you mean?
18:06:53austriancoderlinuxstb: I have tested it and we need all this stuff (arcotg_dcd) in IRAM..
18:06:57peturaustriancoder: did you already reduce iram usage by only putting the buffers in there and not the whole control structs?
18:07:26peturGenre9mp3: not the low-level stuff, no. The rest of the stack should be the same
18:07:27***Saving seen data "./dancer.seen"
18:07:53austriancoderpetur: I have tryed to put only the big buffer in I RAM, but then I got no valid SETUP packets
18:08:46austriancoderpetur: I try to reduce IRAM usage, but I cant promise it
18:08:50linuxstbbluebrother: That's even odder - the mini2g has 32KB more IRAM than the 1G (and ipod Color).
18:08:59peturaustriancoder: do we know what the controller wants to read of it all?
18:09:06linuxstbbluebrother: Although I'm not 100% Rockbox uses the extra 32KB...
18:09:21bluebrotherlinuxstb: well, I'm not used enough to the code to find if lcd buffer is in IRAM
18:10:02peturaustriancoder: I'll check the code tonight, looks like I will have the time to do it
18:10:13austriancoderpetur: fine
18:10:41linuxstbbluebrother: Yes, seems it is - see the start of firmware/drivers/lcd-2bit-horz.c
18:11:14linuxstbbluebrother: You could try removing that IBSS_ATTR and see if it compiles. Although I've no idea how much that will slow down the LCD driver...
18:11:51bluebrotherwell, I could give it a try
18:12:02peturtaking the framebuffer out of iram sounds like a bad idea
18:12:14linuxstbIRAM on PP isn't as important as Coldfire.
18:12:31bluebrotherwhy does the usb stack need that much iram?
18:12:48preglowthe usb controller can't access the normal bus
18:12:53preglowso it needs all its data to be in iram
18:13:01preglowto be expected, really
18:13:03bluebrotherand as idea, could the codec iram be used for that?
18:13:03linuxstbIt's not a caching issue?
18:13:25preglowbluebrother: could, yeah, but what if we want to play music while transfering stuff?
18:13:37preglowbluebrother: anyway, it's a good idea for preliminary work
18:13:38linuxstbOr using a USB serial port at the same time...
18:13:40bluebrotherwell, in that case we simply don't want this :P
18:13:52preglowi think it sounds like a decent solution for now
18:14:14bluebrothera couple of plugins also disable playback because of the buffer.
18:14:14linuxstbAnyone know if we're using all 128KB of IRAM on the PP5022?
18:15:21*linuxstb looks at and doesn't think we are...
18:16:15rasherBagder_: Would it be possible to check that the tarball contains all the needed files from SVN somehow? The tarball gets broken fairly regularly, and it's not something most devs are likely to notice..
18:16:21*austriancoder need to go to band rehearsal... will be back in about 2 hours
18:16:22amiconnlinuxstb: We don't use the extra iram on pp502x (x > 0) yet afaik
18:16:38*bluebrother runs a build with lcd buffer not in iram ...
18:16:45linuxstbrasher: Is it broken now?
18:16:59bluebrothercan't see any obvious difference right now.
18:17:02rasherlinuxstb: I broke it with my voice-building (and fixed it earlier today)
18:17:31rasherAnd the same happened for the usb stuff, etc.
18:17:42linuxstbrasher: There's quite a reliable system in place at the moment - someone breaks it, a user reports the problem, someone fixes it...
18:18:08*linuxstb was joking...
18:18:08bluebrotherhmm, is there some kind of usb debug screen?
18:18:43linuxstbbluebrother: Not that I know of. There's logf output, or the debug output (if any) on your host PC.
18:19:11bluebrotherok, so it's normal I don't see any real difference regarding usb.
18:19:22rasherlinuxstb: It's reliable, but has a high latency unfortunately
18:19:37 Join xigxag [0] (n=xenion13@
18:19:43bluebrotherhmpf. Plugins browser supporting arbitrary folders would be nice.
18:19:48linuxstbCurrently the code attempts a serial port connection - austriancoder told me to do " modprobe usbserial vendor=0x0525 product=0xa4a6"
18:19:52*bluebrother would like to have the lcd test plugin around.
18:20:17linuxstbYou should then see some connection info in /var/log/messages. If you're not running Linux, I can't help....
18:21:00bluebrotherah. Me dumb.
18:21:04linuxstbrasher: Presumably a test of the tarball would need to compile builds for all targets?
18:21:56rasherlinuxstb: or it could compare an unpacked tarball with the svn sources?
18:22:04 Join bluey^ [0] (
18:22:27 Quit DerPapst ("work->home... finally")
18:22:28 Quit Shaid (Read error: 110 (Connection timed out))
18:22:33rasherAnd allo some preset list of files to be missing (I'm assuming there are some files not in the tarball)
18:23:44markunimx31.h didn't really have to be deleted (unless the Gigabeat S port never happens)
18:25:09linuxstbrasher: That could work...
18:25:33linuxstbrasher: Although we would then have two sets of "FILES" files to forget to maintain...
18:28:27*Nico_P is a bit unsure about what to do with id3.elapsed
18:28:43Nico_Pmy gut feeling is that it doesnt belong in the struct mp3entry
18:29:38linuxstbMy gut agrees with your gut
18:30:29 Quit petur ("work->home")
18:30:46Nico_P..thing is the WPS needs access to it
18:30:47 Quit elinenbe (Read error: 110 (Connection timed out))
18:31:38amiconnbluebrother: just enable it in apps/plugins/SOURCES
18:33:28 Quit obo_ ("bye")
18:33:40bluebrotheramiconn: I know that ... I was just annoyed as it isn't present in the plugins list anymore
18:34:31bluebrotheranyway, I'm trying to give talk file presets a shot.
18:34:40bluebrother(in rbutil that is)
18:36:57Nico_Pmoving "elapsed" out of struct mp3entry would break a LOT of things :( If I do it I sure need to be careful
18:37:17 Quit bluey- (Read error: 110 (Connection timed out))
18:38:08linuxstbNico_P: Is there anything else?
18:38:22 Join qweru [0] (
18:38:31Nico_Plinuxstb: anything else than what ?
18:38:43Nico_Pthe WPS ?
18:38:50linuxstbAny other fields that you don't think belong in the mp3entry struct?
18:40:13Nico_Poffset and index seem a bit out of place but less
18:42:50Nico_Pthere are some fileds I'm not really sure about
18:43:28 Join Domonoky [0] (
18:44:01*linuxstb afk
18:45:13 Quit alienbiker99 ("( :: NoNameScript 4.02 :: )")
18:48:43 Join oxygen77 [0] (
18:52:40 Join rogelio [0] (n=rogelio@
18:56:51 Quit Vortex375 (Remote closed the connection)
18:58:25 Join elinenbe [0] (
19:00:25 Join lazka [0] (
19:07:55bluebrotherrasher, Domonoky: current tts configuration screenshot:
19:08:19bluebrotherit has presets but allows user-adjustable presets so the user can edit the options (but only for those presets)
19:10:48Domonokylooks good.. but the tts configuration needs to be extended in the future.. festival needs todo an init and a stop operation..
19:11:09bluebrotherhmm. Why that?
19:11:15 Join petur [0] (n=petur@rockbox/developer/petur)
19:11:19 Quit austriancoder (Remote closed the connection)
19:11:45Domonokyfestival is client/server based.. so you need to start and stop the server..
19:12:28Domonokybut we can leave that for the future..
19:13:35bluebrotherhmm. Festival needs to get started from rbutil? Couldn't the voice script itself do it?
19:13:47bluebrotheri.e. enable the server upon start and disable when finished?
19:14:35 Join austriancoder [0] (n=austrian@
19:16:14Domonokyvoice script ? for talk files, rbutil would need to start the server, bevor generating the first clip, and stop it after the last script.. (of course it could call a script or an exe to init / deinit it..
19:19:02 Join low_light [0] (i=c730190b@gateway/web/cgi-irc/
19:20:30 Quit Febs (Read error: 104 (Connection reset by peer))
19:20:49 Join tumu [0] (n=tumu@
19:21:06*low_light accepts the title of "start-new-ports-without-having-the-previous-in-a-reliable-state guy"
19:21:54low_lightmaking progress :)
19:21:55kkthat's awesome :D
19:22:47amiconnLlorean: Definitely better than plain espeak, but not especially nice either
19:23:21amiconnlow_light: Imo it would make much sense to get the m:robe code into svn asap
19:23:23Lloreanamiconn: I'd say better than flite/festival as well. Still not as good as some of the commercial voices.
19:23:41amiconnI'm undecided whether it's better than festival or not
19:23:51 Join shark_jr [0] (i=4b640637@gateway/web/cgi-irc/
19:24:08amiconnlow_light: Oh, and c200 too of course
19:24:10LloreanFestival's pronunciation of a few things irks me, though I'll admit that could be fixed in the .lang or something.
19:24:53amiconnNot in the lang... that has to stay engine neutral
19:25:13amiconnI noticed that espeak+mbrola doesn't have the ploojins problem
19:25:39 Quit shark_jr (Client Quit)
19:25:43Lloreanespeak doesn't have the ploojins problem
19:25:58LloreanThat seems to be a flite/festival thing
19:26:06*amiconn thinks that both m:robe and c200 are more advanced than elio and av300
19:26:29amiconnLlorean: Not only. The MS voices have the same problem, but not e.g. at&t
19:27:04low_lightamiconn: yes...I need to sync both patches to current svn (my tree is a couple of weeks old).
19:27:45Nico_Plow_light: the pics are unavailable to me...
19:28:32low_lightha ha... exceeded the limit, that was fast.
19:28:49Nico_Plow_light: to sync the patches it might be easier to go back to the svn rev you started with and do svn up
19:30:28*amiconn wonders whether database speed got optimised lately
19:32:01low_lightsyncing it won't be too hard to's just complicated by my devel setup (I'm at work, behind a firewall)
19:34:52 Join Buschel [0] (
19:35:13bluebrotherhmm. Some web server script stuck? the download page shows a different revision than the frontpage
19:35:15markunLlorean: wow, the voice file sounds pretty good
19:35:42low_lightNico_P: and
19:36:58Nico_Plow_light: second one works but not the first... I get the idea though ;)
19:37:28Nico_Plow_light: what doesnt work ?
19:37:33 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
19:37:44 Quit amiconn (Nick collision from services.)
19:37:45 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
19:37:58 Join stripwax [0] (
19:38:00amiconnHmm, the db commit on boot breaks voice on hwcodec (until reboot or probably another event triggering voice reload)
19:38:59*amiconn suspects the commit splash voicing
19:41:32 Join sarixe [0] (
19:42:07low_lightNico_P: audio doesn't work ;)
19:42:43low_lightand it thinks it's plugged in and turns off after 10 or 15 secs
19:43:27amiconnlow_light: That's probably the low bat shutdown (guessing that you didn't implement battery reading yet)
19:44:03amiconnI had that on ipod 2nd gen, fixing it temporarily by faking a sane battery voltage
19:44:04 Quit HellDragon (Read error: 104 (Connection reset by peer))
19:44:19 Join HellDragon [0] (n=Nocebo@unaffiliated/helldragon)
19:44:32 Quit datachild ("Death is better, a milder fate than tyranny.")
19:44:44 Join datachild [0] (
19:45:05 Part bluey^ ("Leaving")
19:45:50bluebrotherok, the dialog is working fine now :)
19:47:11 Part rogelio ("Kopete 0.12.4 :")
19:47:15low_lightamiconn: yes. the battery is read throught the i2c like the e200, but uses a different address
19:49:07low_lightexcept for the lcd & buttons, most of the code for the e200 works
19:49:37low_lightI'm not sure the best way to organize the target tree
19:50:19 Join Entasis_ [0] (
19:53:38 Quit Entasis (Read error: 104 (Connection reset by peer))
19:55:06 Join Vortex375 [0] (
19:56:03 Join bsdfox_ [0] (
19:57:37 Join freqmod [0] (
19:57:43 Quit qweru ("moo")
19:58:22 Join qweru [0] (
20:01:14 Join desowin [0] (
20:03:42 Quit qweru ("moo")
20:04:54 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
20:07:31***Saving seen data "./dancer.seen"
20:08:25 Join obo [0] (n=obo@rockbox/developer/obo)
20:10:20 Quit bsdfox__ (Connection timed out)
20:18:21`sam`hey i have the rockbox.ipod in my .rockbox folder but it says file not found
20:18:47`sam`is it supposed to be located somewhere else?
20:19:03 Quit freqmod (Remote closed the connection)
20:19:20LloreanNo, it belongs in the .rockbox folder, which shouldn't be inside any other folders.
20:20:13`sam`right, the .rockbox folder is in the root of the drive, and in it is rockbox.ipod
20:20:21LloreanIs there anything interesting about the partition it's on, or is it on a 100% untouched (new or recently restored) iPod?
20:20:59GodEater_that CNet video isn't too shabby as reviews/install videos go
20:21:20`sam`i've reformatted it, there's another partition that i had for ipodlinux...
20:21:52Llorean`sam`: So, it's on a fat32 partition, marked as primary?
20:22:27`sam`oh wait, it's fat32, but it might not be marked as primary... let me check that
20:25:36 Quit qwm (Remote closed the connection)
20:25:39 Join bejic [0] (
20:25:39 Join qwm [0] (
20:25:43`sam`how do you make it primary?
20:26:04bejichI i HAVE A QUESTION
20:26:32GodEater_I have one too - can you switch off caps-lock please ?
20:26:34Domonokybejic: just ask, dont shout...
20:26:43Llorean`sam`: Dunno, honestly. I've seen the option, but I don't remember in what tool or where. ;)
20:26:45bejicevery time I connect the sansa to the pc the music folder goes to hidden
20:27:04Llorean`sam`: I just know it's necessary for Rockbox to look on that partition.
20:27:13Lloreanbejic: That's the Sandisk firmware doing that.
20:27:14bejicis there a way to diable this behaviour?
20:27:15GodEater_Llorean: I believe you mean it has to be a primary partition, and be marked as bootable. You can't "mark" it primary. It either is, or it's extended.
20:27:23amiconn`sam`: The partition needs to be primary, not logical. You're probably confusing that with what partition is active
20:27:25LloreanGodEater_: Ah, that's what I meant.
20:27:28Domonokybejic: you could just rename the music folder..
20:27:32amiconnThis is something rockbox doesn't care about
20:27:51`sam`amiconn, yeah i think i was confusing the 2, it is a primary partition, there are no extended partitions
20:27:58amiconnGodEater: The partition doesn't need to be marked bootable (active)
20:28:00bejicThanks domonoky will try
20:28:04LloreanGodEater_: A little tired.
20:28:05GodEater_amiconn: I never said it did
20:28:15Lloreanamiconn: I got everything confused.
20:28:19*Llorean is somewhat out of it.
20:28:28GodEater_Llorean: have you been awake all day ?
20:28:38amiconnRockbox doesn't even care about partition type, as long as it's != 5 (extended)
20:28:45LloreanYeah, couldn't sleep last night for no apparent reason beyond being unable to.
20:28:52LloreanNow I have to make it another 7 or 8 hours.
20:28:53GodEater_that sucks
20:29:10LloreanAnd it looks like we've got a new "Interesting" forum member.
20:29:22GodEater_ooooh - who ?
20:29:31LloreanThis "pink floyd" chap.
20:29:34GodEater_pink floyd ?
20:29:36LloreanI can tell he's going to be a headache.
20:29:45`sam`is there some kind of bootloader config?
20:29:56amiconnGodEater: Hmm, then I misinterpreted your statement
20:30:05`sam`maybe it's looking for it on the other partition
20:30:08GodEater_amiconn: I was trying to work out what Llorean meant
20:30:12GodEater_more than giving it as an instruction
20:30:15Lloreanamiconn: And I was speaking gibberish.
20:30:16GodEater_my wording was poor
20:30:28Domonokysam: no, rockbox just uses the first fat32 partition it finds..
20:30:30 Quit Thundercloud (Remote closed the connection)
20:30:49`sam`well i don't get why this isn't finding it then, because the other partition is ext3
20:31:24GodEater_Llorean: are we going to blame the CNET video for another cluebie ?
20:31:54LloreanI haven't watched the CNET video yet, but it does mean exposure. Good for development, bad for orderly forums. :)
20:32:24Domonokythis CNET Video is nice.. it says read the Manual, several times.. :-)
20:32:34 Join agm3nt [0] (
20:32:46GodEater_the only thing he got wrong was how to get back to apple os
20:32:56GodEater_although his method works, it's not optimal
20:33:13Domonokyjeah, normal shutdown is enough..
20:33:40`sam`the fat32 is /dev/sdb2 and goes from 67 to 912, the ext3 is /dev/sdb3 and goes from 4 to 66
20:33:55GodEater_so they're not in order then ?
20:33:59`sam`and then /dev/sdb1 is empty from 1 to 3
20:34:23GodEater_it's *marked* empty :)
20:34:41Domonokythe empty is the ipodfirmware..
20:34:55`sam`Domonoky, that's what i was thining
20:35:20Domonokybut maybe the bootloader has a problem with the ext3 before the fat32.. but it shoudnt..
20:35:35Domonokyhave you checked the fat32 for errors ?
20:36:10Llorean`sam`: Are you using the Rockbox bootloader, or Loader2 then?
20:36:14`sam`no... how do i check fat32 partition on linux? is there a fsck or something for it?
20:36:30`sam`Llorean, i used ipodpatcher from rockbox site
20:36:42`sam`and selected I to install rockbox bootloader
20:37:06 Join nicktastic [0] (n=nick@unaffiliated/nicktastic)
20:37:20GodEater_`sam` fsck.vfat
20:37:51`sam`i think i'll probably just reformat the drive, i'm probably not going to use ipodlinux anyway, at least somebody can tell me that without a doubt mpd will work on it
20:38:00GodEater_hahahaha - that pink floyd guy is nuts
20:38:55Domonokysam: be sure to leave the first partition intact, or you will have to use itunes restore.. :-)
20:39:12 Join qweru [0] (
20:39:25 Quit low_light ("CGI:IRC (EOF)")
20:39:36`sam`yeah, i'm just going to delete 2 and 3, and then make all that space fat32 and reinstall the apple stuff and rockbox
20:39:40nicktasticUsing rockbox on iAudio X5, every once in a while it tells me "Save failed, no partition" - usually when I skip tracks, or load a new directory. Everything works fine, but it is kind of annoying. Does anyone know why this happens?
20:40:40*GodEater_ gives up. Llorean, there's no talking to him.
20:40:55Domonokynicktastic: you are useing a recent official rockbox build ?
20:41:12*nicktastic hides his eyes
20:41:19nicktasticI withdraw my question
20:41:37GodEater_Domonoky. Scarey man of #rockbox.
20:41:59Domonokyhehe.. i know this error message, but it shouldnt exist anymore in a new rockbox build..
20:42:17nicktasticI see, I'm using maxwen0's June build
20:42:28Domonokythis error message is from the old settings code, i think..
20:42:53*nicktastic will grep around when he has some time
20:43:01GodEater_Llorean beat me to deleting his message
20:43:42 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
21:03:58 Join My_Sic [0] (
21:06:28markunLlorean: this Pink Floy guy is an idiot indeed
21:06:44markun+d somewhere
21:07:03LloreanI think he's got the 'no more nonsense' message now.
21:11:00 Join febs_mobile [0] (i=d809fa6e@gateway/web/cgi-irc/
21:12:25GodEater_famous last words...
21:13:10LloreanGodEater_: I told you I was out of it. I *know* better than to tempt fate like that.
21:13:27GodEater_and yet, there you went ... :)
21:13:28 Quit febs_mobile (Client Quit)
21:14:41 Join donutman25 [0] (n=chatzill@
21:15:37rasherlinuxstb: But at least the new set of FILES would yell and scream at us when we forgot it.
21:17:19amiconnThe web server needs a slight kick...
21:21:39*linuxstb reads the forums, sees hundreds of posts by pink floyd and bangs head on desk
21:21:58GodEater_there there
21:22:05GodEater_Llorean made the nasty man go away
21:22:08*petur goes looking
21:22:09linuxstbIs there an ignore feature on the forums?
21:22:42GodEater_yeah, it's called "banning" :)
21:23:01 Join Febs [0] (n=chatzill@
21:24:23*GodEater_ wonders if we should contact CNet's edtiorial dept. and get them to correct their procedure to restart an ipod in apple os
21:24:41*GodEater_ can just see them refilming the whole thing :)
21:25:21LloreanGodEater_: I've given slightly better diagnostic instructions to the Gigabeat guy (I think). ;)
21:25:31GodEater_Llorean: cool
21:26:07Lloreandir /s > blah.txt gives a much better picture of the folder layout than hoping for a good description, or the screenshots some people like to try to use to answer such a question. :)
21:26:11 Quit The-Compiler (Connection reset by peer)
21:26:12*GodEater_ will put money on him being able to do this
21:26:25GodEater_being able to
21:26:29 Quit inakiniff (Read error: 110 (Connection timed out))
21:26:36LloreanThe last time I went so far as to ask that of someone it worked, and revealed the solution
21:26:52LloreanThe solution being FWIMG01.DAT didn't exist at all
21:27:02GodEater_$5 in the rockbox fund then if you get a result without having to issue further instructions
21:27:26GodEater_I think you need dir /a /s btw :)
21:27:48GodEater_cos otherwise the /s won't go into the system/hidden folders
21:28:15GodEater_and if I catch you editting the post - so help me!
21:28:22 Join Thundercloud [0] (
21:28:42LloreanToo late
21:28:47LloreanI edited it before you warned me against it
21:28:58LloreanBy 17 seconds on my clock
21:29:22LloreanThough I didn't give /a /s last time, just /s and it walked the firmware folders
21:29:56GodEater_changing your post counts as further instructions in my book
21:30:06GodEater_if there are no more such shennanigans
21:30:09GodEater_I will honour our bet
21:30:27GodEater_I win
21:31:07linuxstbYou didn't mention Llorean's penalty for losing though...
21:31:07GodEater_AND I'll let you carry on explaining to him now as part of your prize :)
21:31:23GodEater_$5 from him into the RB fund
21:31:27LloreanI'm curious which step threw him.
21:31:28GodEater_I thought it was implicit
21:31:58linuxstbGodEater_: I read it as you offering $5 if Llorean succeeded.
21:32:00GodEater_I think he got lost at "If"
21:32:15GodEater_indeed, and $5 from him if he didn't
21:32:17LloreanI never actually accepted a bet, but I'll throw in an extra $5 next time I donate anyway.
21:32:38GodEater_can't say fairer than that
21:33:07GodEater_this is going to be a real treasure to watch this thread, I can tell
21:33:26 Quit ompaul ("later")
21:33:31linuxstbDoesn't rbutil support the Gigabeat install?
21:33:44GodEater_I believe so
21:33:45FebsShhhh. You'll spoil the fun.
21:34:13linuxstbSo how many more posts from Llorean are needed to finish this?
21:34:22*linuxstb will take 6
21:34:48rasherbluebrother: The voice config looks really good - is the executable searched for?
21:34:56LloreanWe could have a weekly pool for what the longest new installation thread for that week will be. :-P
21:35:35GodEater_this one wouldn't count though - he's tacked it on the end of an old one
21:36:04GodEater_it won't be six posts if Llorean tells him to use Rbutil
21:36:14linuxstbGodEater_: You sure? ;)
21:36:26 Join Siltaar [0] (
21:36:33LloreanWith this user....
21:36:40GodEater_I'm going with four
21:36:51GodEater_four posts from Llorean that is
21:36:56linuxstbWith rbutil?
21:37:10*Domonoky is optimistic.. only 2.. then llorean will bann him :-)
21:37:25GodEater_Llorean's tired, not grumpy. yet.
21:37:42GodEater_and no cheating Llorean. No one word posts :)
21:37:45 Join DerPapst [0] (
21:37:45LloreanDomonoky: I managed to keep myself from banning Pink Floyd
21:38:05LloreanI convinced myself that the second time he posted offering copyright material he hadn't seen my warning in the other thread yet.
21:38:11*DerPapst wonders what the link is...
21:40:39GodEater_still no reply either
21:40:43GodEater_Llorean's scared him off
21:42:31LloreanI'm hoping for the long shot: "Oh, on reading it again I see what you meant, here's the .txt file"
21:42:53DerPapstbut at least he admits that he's an idiot...
21:43:39 Join fm2 [0] (i=d9b9677b@gateway/web/cgi-irc/
21:43:46amiconnBagder_: around?
21:44:12*fm2 seconds the question
21:44:14GodEater_Llorean: man you really ARE tired if you think he's suddenly going to gain an IQ point or ten
21:44:22*amiconn has a suspicion regarding the webserver not picking up my latest commit
21:44:29LloreanGodEater_: He claims to have "got all that done" but...
21:44:37 Join Dwyloc [0] (i=53688643@gateway/web/cgi-irc/
21:44:37amiconnIt's probably related to today's apache updating
21:44:58GodEater_Llorean: hahahahaha
21:44:58*fm2 needs help with
21:45:24GodEater_Llorean: rbutil it man. Just take the easy way.
21:45:57LloreanI was just curious what he's done.
21:46:07LloreanBut I realized, I'm not sure if he's got a half-finished installation right now or what
21:46:15Lloreanhe may be doing this on a 'clean' gigabeat
21:46:38GodEater_well if he's not showing the back up, it's either clean, or he's REALLY not following the instructions
21:47:29GodEater_I've have linked him straight to the zip file myself
21:47:34GodEater_because you KNOW he'll ask
21:47:53GodEater_anyway - that's post numero uno
21:48:01GodEater_one more for Domonoky to win
21:48:28fm2linuxstb: ping
21:50:14fm2makefile gurus: ping :-)
21:50:42linuxstbfm2: I'll have a quick go now...
21:50:46Domonokyfm2: if just ask.. it will be more successfull. . :-)
21:51:20fm2Domonoky: the situation and the question are in FS #7667
21:52:11GodEater_plus he spoke to both of linuxstb and Bagder_ this morning about it
21:52:15GodEater_albeit in disguise
21:52:59Domonokyah.. i didnt read the log, so i didnt know.. :-)
21:53:07GodEater_hence I told you :)
21:53:21fm2Another question: I created a new folder and (with TotoiseSVN) did an "Add" on it. The folder and all files within it were "added". After that, I produced a patch, and that patch contained the new files twice. What did I do wrong?
21:53:32 Join Zagor [0] (
21:53:59fm2Domonoky: it isn't in the log, not your fault :-)
21:54:10GodEater_fm2: it is...
21:54:30GodEater_in the log that is
21:54:35GodEater_not Domonoky's fault
21:54:41fm2GodEater: that the question is in FS?
21:54:42GodEater_there's a lot of reading in the logs usually
21:54:43amiconnZagor: Looks like there is a problem that might be rlated to today's debian updates
21:54:58GodEater_fm2: no, that you'd already asked this moring
21:55:13Zagoramiconn: ok?
21:55:33amiconnThe frontpage doesn't show my latest commit
21:55:42fm2linuxstb: do you get the same linker errors?
21:56:25Zagoramiconn: ok, checking
21:57:12linuxstbfm2: I haven't tried your version - I can tell it's not going to do the right thing.
21:57:40GodEater_fm2: did you fix the patch file so it's not full of duplicates now ?
21:58:07fm2linuxstb: ok. I just have search&replace on reversi, so if that's not the correct thing to do I'be thankful for a correction
21:58:34fm2GodEater: yes. Manually. I can't produce the correct patch with tortoisesvn (see above)
21:59:20fm2petur, bluebrother: do you know how to add files and folders to svn via tortoise?
21:59:38fm2GodEater: corrected patch is in FS
21:59:41linuxstbfm2: It's correct if you're only creating a single .rock None of the other plugin subdir Makefiles create multiple rocks.
22:00:04fm2linuxstb: yes, we don't go easy ways! :-)
22:00:29Domonokyfm2, normally you just use the "add" function of the svn client, and it should work.. dont know why it doesnt work for you.. maybe aou already added those files before ?
22:00:56peturfm2: right-click, TortoiseSVN->Add
22:02:17 Quit Buschel ()
22:02:28 Join _fm2 [0] (i=d9b9545a@gateway/web/cgi-irc/
22:03:29_fm2Domonoky: I only did "add" once. In the log I saw "Added xxxx" (for each file and the folder)
22:04:33_fm2If I don't do "add" the folder doesn't show up in the item list when creating the patch
22:05:16 Quit fm2 ("CGI:IRC (Ping timeout)")
22:06:06 Nick _fm2 is now known as fm2 (i=d9b9545a@gateway/web/cgi-irc/
22:07:34***Saving seen data "./dancer.seen"
22:09:27 Quit Febs (Read error: 104 (Connection reset by peer))
22:09:58Zagoramiconn: fixed now. thanks for the report.
22:11:22linuxstb_fm2: I've fixed the Makefile. There were also some other problems - you should use DEBUGF instead of rb->debugf, and you shouldn't include the PLUGIN_HEADER macro in shortcuts_common.c
22:11:57 Quit sarixe ("Peace")
22:13:43fm2linuxstb: thanks. Is DEBUGF defined as rb->debugf for plugins?
22:13:45linuxstbfm2: - that's the complete apps/plugins/shortcuts folder. (SOURCES isn't used any more)
22:13:57linuxstbfm2: Yes, but only when debugf is available.
22:14:02linuxstbOtherwise it does nothing.
22:14:11fm2linuxstb: of course
22:14:31linuxstbI was building for a real target - and debugf wasn't available.
22:15:17fm2I seem to have some problems with my file system. I can't access the shortcuts folder.
22:17:03fm2linuxstb: fixed. Compiling...
22:17:24fm2I noticed that makefile also contained many old targets, e.g. win32 sim
22:17:47linuxstbYes, I've removed those. We should probably do that in the other Makefiles now as well. (win32 and x11 sims)
22:20:57 Quit Vortex375 (Remote closed the connection)
22:21:33fm2linuxstb: hmm. Plugins don't work properly. But I'll sort that out. Thank you!
22:23:11linuxstbfm2: Oops, you should remove the "include $(TOOLSDIR)/" from the Makefile (and the comment before it). That's not needed any more.
22:23:42rasherI take it it's not on purpose that the wps/Rockboxed dir (holding images) is not included in the tarball?
22:24:18fm2linuxstb: ok, done. Why don't we use SOURCES anymore?
22:25:29linuxstbBecause SOURCES was designed to list the sources required for one .rock
22:25:54 Quit DerPapst (Read error: 104 (Connection reset by peer))
22:26:27fm2linuxstb: but you still have listed all three
22:26:42fm2SOURCES := ...
22:26:43linuxstbYes, but the OBJS variable is now split into two
22:26:49 Join DerPapst [0] (
22:27:09fm2AH, it was constructed from sources before?
22:27:40rasherBagder_: You left a few ~ files (emacs?) in the directory used to build the bleeding edge tarball: apps/codecs/libfaad/common.h~ and apps/codecs/spc/Spc_Dsp.h~
22:28:07 Part fm2 ("off for fixing")
22:33:46 Quit Zagor ("Client exiting")
22:35:11 Join alienbiker99 [0] (
22:40:42 Part agm3nt
22:41:27 Join HidalgoDC [0] (
22:41:33HidalgoDCcan somebody help me please
22:41:50alienbiker99ask your question soembody maybe able to help
22:41:51rasherHidalgoDC: Just ask your question
22:41:51desowinask instead asking to ask, read rules one day, ...
22:42:43HidalgoDCFisrt how ipod-video-sim-w32 works
22:43:07rasherHidalgoDC: Just doubleclick either the .exe file or the "background.bat" file
22:43:11DerPapstwith double clicking
22:43:14HidalgoDCbecause when i run it and i select browse themes, and i choose my theme
22:43:21 Join Domonoky_ [0] (
22:43:22HidalgoDCit dont show my theme on it
22:43:36HidalgoDCit don't set it..
22:43:44rasherHidalgoDC: Maybe your theme requires a patch, or is broken in some other way?
22:44:04HidalgoDCmaybe, because im noob :p
22:44:26HidalgoDClook that im gonna send you
22:44:29HidalgoDCmy .wps file
22:46:11 Quit petur ("switching...")
22:46:17 Join petur [0] (n=petur@rockbox/developer/petur)
22:46:56HidalgoDCsomebody is there ?
22:47:45Domonoky_where did you got this theme ? and waht doesn work ?
22:48:05HidalgoDCim doing this theme
22:48:10HidalgoDCwhen i choose it in my ipod
22:48:13HidalgoDCit donsent work
22:48:19HidalgoDCit loads correctly but
22:48:25HidalgoDCi see only the icons..
22:48:38Domonoky_so selfmade.. then you probably have an error in the wps file..
22:48:39rasherTry running the simulator with −−debugwps
22:49:24HidalgoDCaccept the file that im gonna send you
22:49:26Domonoky_and please write longer lines..
22:49:28HidalgoDCyou gonna see
22:49:29rasherPlease try to keep your thoughts on one line.
22:49:31HidalgoDCmy backfrop..
22:49:35HidalgoDCokay np
22:50:09Domonoky_try what rasher said...
22:51:12rasherHidalgoDC: I don't know anything about making WPS, and I'm sort of busy. The only help I can give you is to try running the simulator with the −−debugwps flag. That'll give you wps errors in the console-window
22:52:06 Part oxygen77
22:52:08rasherBagder_: you also left some compiled programs in tools/
22:52:24*ender` yawns
22:58:53 Join BigMac [0] (
22:59:06 Join MySic [0] (
23:00:00 Quit austriancoder (Remote closed the connection)
23:00:22 Quit Genre9mp3 (Read error: 110 (Connection timed out))
23:00:39 Quit Domonoky (Read error: 110 (Connection timed out))
23:01:55 Quit MySic (Client Quit)
23:01:58 Quit My_Sic (No route to host)
23:02:21 Join My_Sic [0] (
23:05:38 Quit seablue ("life, death, life, death")
23:05:39 Quit BobShield (Read error: 104 (Connection reset by peer))
23:06:10 Quit Thundercloud (Read error: 104 (Connection reset by peer))
23:06:51 Quit Domonoky_ (Read error: 104 (Connection reset by peer))
23:07:08 Join BobShield [0] (
23:09:56 Quit BigMac (Remote closed the connection)
23:10:21 Quit SirFunk (Remote closed the connection)
23:14:29preglowmokay, it seems like low latency dsp has been broken
23:14:32preglowany obvious culprits?
23:14:53 Join Thundercloud [0] (
23:15:26 Join BigMac [0] (
23:19:29 Quit BobShield (Read error: 104 (Connection reset by peer))
23:19:54 Quit davina (Remote closed the connection)
23:20:50 Quit bsdfox_ (Success)
23:21:15 Join BobShield [0] (
23:21:21preglowseems the eq menu doesn't call soundmenu_callback() anymore, and hence doesn't get low latency
23:21:29preglowi wish i understood how this menu macro stuff works
23:22:54rasherMan, drawing anything in 6x7 monochrome is quite a challenge
23:23:25*petur pats preglow on the back
23:24:40*preglow summons JdGordon
23:25:37rasherI guess I could just generate all 4 trillion possible 6x7 icons and look through them
23:26:49DerPapstand the cool thing you could drop all icons with less than 4 pixels or inverted
23:26:52*bluebrother returns
23:26:54 Quit BigMac (Remote closed the connection)
23:27:16bluebrotherrasher: no, the executable isn't searched yet. My idea was to search for it once the user changes the profile
23:27:31bluebrotherand display a message if the binary isn't somewhere in the path
23:28:55rasherWarning box? I'd rather get a "passive" message, such as a notification icon next to the input box or something
23:28:56 Join darklord [0] (n=darklord@
23:29:17 Quit donutman25 ("ChatZilla [Firefox]")
23:29:29bluebrotheryeah. Not sure if it's a good idea to display some hint in the input box itself and use a red background or something like that ...
23:29:32 Part darklord
23:29:54bluebrotherbut I first want to finish adjusting the talkfile window so I can commit this.
23:30:07 Join BigMac [0] (
23:30:17bluebrotherfurther optimizations can come later ;-)
23:30:59 Join fm2 [0] (i=d9b95442@gateway/web/cgi-irc/
23:31:39fm2Weird. A plugin doesn't get executed when called via context menu.
23:31:53 Quit desowin ("use linux")
23:32:04fm2As in filetype_load_plugin("shortcuts_append", selected_file)
23:32:45fm2If I select it absolutely nothing happens, no splash, no... absolutely nothing. Just returning to the file browser
23:32:49fm2Any hints?
23:34:22fm2And even if I rename the .rock file nothing happens (no message "couldn't load plugin")
23:36:48 Join bsdfox_ [0] (
23:37:34fm2Ah, found the error: an entry in viwers.config was lacking
23:38:24 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
23:38:51 Quit Genre9mp3 (Client Quit)
23:38:51fm2Now it displays "Incompatible model". Why?
23:39:03 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
23:39:37bluebrotherfm2: you get this when the plugin api changed and the plugin doesn't match your build
23:39:57pixelmacompiled for the sim and running on target or vice-versa?
23:40:46 Quit BigMac (Remote closed the connection)
23:40:50 Quit Genre9mp3 (Client Quit)
23:41:06 Join Genre9mp3 [0] (
23:41:47fm2bluebrother, pixelma: no, everything on sim, fresh compiled...
23:42:10bluebrotherdid you run a "make install" after compiling? I tend to forget this ;-)
23:43:01fm2bluebrother: I copy the plugin manually, "install" takes too long
23:43:44bluebrother"Rockbox turns your iPod into a open-source monster" :D
23:45:11fm2The other plugin by me gets executed. Hm...
23:46:17 Join webguest37 [0] (i=5750429c@gateway/web/cgi-irc/
23:46:59 Join SirFunk [0] (
23:47:43fm2Could it be a problem with linking? I.e. something gets linked in at a wrong address and this is seen as the bad version?
23:47:56fm2linuxstb: are you sure the make file is ok?
23:50:14 Join BigMac [0] (
23:50:27linuxstbfm2: No, I'm not sure.
23:50:29fm2Since the version is somewhere in the plugin header
23:51:06 Join jhulst [0] (
23:52:43DerPapstbluebrother: seen the video about rockbox at cnet reviews? he points to the manual several times but he don't know how to reboot the ipod from rockbox... ^^
23:52:59bluebrotherDerPapst: yes, just watched it.
23:53:09DerPapstinstead he performs a reset -.-
23:53:25bluebrotherwell, afaik it is the only way to reboot the Ipod −− otherwise you would need to shut it down first.
23:53:29 Quit Dwyloc ("CGI:IRC (EOF)")
23:53:48DerPapstin rockbox hold down play/pause to reboot the ipod.
23:53:53DerPapstno need to reset it
23:54:10bluebrotherreally? That shuts down my Ipod ...
23:54:21fm2linuxstb: hm... all but 'view' vs 'append' seems to be identical
23:54:28DerPapstyeah. then press any other button and it reboots
23:54:42bluebrotheri.e. during shutdown?
23:55:01DerPapstno, after it's shot down
23:55:07linuxstbbluebrother: I think DerPapst is trying to say that you shouldn't reset (i.e. MENU+SELECT) Rockbox, but do a clean shutdown and power-on.
23:55:08 Quit jgarvey ("Leaving")
23:55:19bluebrotherah, ok.
23:55:20DerPapstthat's it
23:55:42fm2The difference in plugins starts at offset 0x88. Is this beyond header?
23:55:50bluebrotherI thought you meant a real reboot.
23:56:35linuxstbfm2: I don't know how big the plugin header is, but I would expect it to be very small.
23:56:35DerPapstnope.. but anyways. :P
23:56:55 Join ddalton [0] (
23:57:05bluebrotherI would suggest users to do a real shutdown instead of a hard reboot too ;-)
23:57:32*DerPapst goes off for his daily beauty sleep
23:57:42DerPapstGood night everyone :)
23:57:51 Quit DerPapst ("So Long And Thanks For All The Fish!")
23:58:31fm2linuxstb: this seems to be the standard win exe header starting with MZ (I build the sim). RB part is somewhere inside. So comaring the first bytes doesn't make sense
23:58:57linuxstbfm2: Ah, OK. In the sim things are different.

Previous day | Next day