Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2005-12-19

00:10:11 Join webguest97 [0] (
00:10:51webguest97hi all
00:11:32 Quit actionshrimp ("a bird in the bush is worth two in your house")
00:11:58 Quit webguest97 (Client Quit)
00:12:09 Join actionshrimp [0] (
00:26:18 Join memmem [0] (
00:26:56 Join ryan [0] (
00:28:12 Quit ryan (Client Quit)
00:34:54 Quit Strath (Read error: 104 (Connection reset by peer))
00:42:34 Join DJDD_ [0] (
00:47:28 Join _DangerousDan [0] (
00:50:07 Quit ender` (" People are more violently opposed to fur than leather because it is safer to harrass rich women than motorcycle gangs.")
00:58:45 Join RotAtoR [0] (
01:03:38 Quit DangerousDan (Read error: 110 (Connection timed out))
01:05:42 Quit mirak ("Ex-Chat")
01:18:29 Quit Kohlrabi ("Leaving")
01:25:53 Quit Moos ("Happy Birthday Rockbox")
01:28:57 Join Mirthoneist [0] (
01:33:42 Quit dpassen1 ()
01:39:42 Quit Subterranean ()
01:42:34 Join elinenbe [0] (
01:43:13 Quit San (Read error: 110 (Connection timed out))
01:45:05elinenbewho is working on the ipod port?
01:46:14***Saving seen data "./dancer.seen"
01:54:09Cassandralinuxstb and preglow.
01:54:17CassandraI'm hoping to help at some point.
01:57:38 Quit memmem ("ERC Version 5.0.4 $Revision: 1.726.2.19 $ (IRC client for Emacs)")
02:00:19elinenbeDo you know the status?
02:05:28 Quit Light0 (Remote closed the connection)
02:22:04 Quit Mirthoneist ("Chatzilla 0.9.69 [Firefox 1.5/2005111116]")
02:36:05 Join Hooligan [0] (
02:41:53 Join Paul_The_Nerd [0] (
02:54:45 Quit Paul_The_Nerd ("Chatzilla [Firefox 1.5/undefined]")
02:55:17 Join Paul_The_Nerd [0] (
03:02:41 Nick Lost-ash is now known as ashridah (
03:07:26 Quit PaulJ ("CGI:IRC")
03:09:37 Quit DrumRBoy320|away ()
03:15:24 Quit Hooligan (Read error: 104 (Connection reset by peer))
03:19:00 Join San [0] (
03:24:21 Join AllenMockery [0] (
03:26:16AllenMockeryI'm a blind person trying to figure out how to get an mp3 player that I can use .So, like, I like Rockbox because of that.
03:26:25AllenMockeryThe problem is... well. Um.
03:26:35AllenMockeryIs there a single mp3 player that's supported I can actually buy from amazon or whatever?
03:26:41 Join mixin [0] (n=me@
03:26:58AllenMockeryBecause i've been searching all day, and it's really annoying me.
03:27:24 Join DrumRBoy320 [0] (
03:28:44 Quit Paul_The_Nerd (Read error: 104 (Connection reset by peer))
03:31:37NibblerAllenMockery: hmmmzzzz when i bought mine, it was on amazone
03:31:41Nibblerbut thats long time ago
03:32:02AllenMockeryYeah, that's the problem I'm running into. All of hte Archos stuff on amazon is discontinued.
03:32:13AllenMockeryAll of hte mp3 players are anyway.
03:32:52AllenMockeryAt least, the ones I can find. I haven't looked at iriver on amazon yet, but I dont' see 1X0 or 3X0 on their website, just a H10 with 20, 6 or 5 GB limmits.
03:33:00AllenMockery(Obviously I'm getting 20.)
03:33:20Nibblerall unavailable, stoopid
03:33:29Nibblercant you get another dealer?
03:34:04AllenMockeryI live in a teeny tiny island in Alaska. I don't have that many choices locally.
03:34:39 Quit mixin ()
03:34:43Nibblerah, i c :|
03:35:11 Join Paul_The_Nerd [0] (
03:35:31 Quit DJDD_ (Read error: 104 (Connection reset by peer))
03:35:55NibblerAllenMockery: i'd try ebay
03:38:03 Join DJDD_ [0] (
03:39:01 Part DJDD_
03:39:43 Quit _DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
03:40:07 Join Zak1392 [0] (
03:40:31 Quit Zak1392 (Client Quit)
03:45:58 Join Hooligan [0] (
03:46:15***Saving seen data "./dancer.seen"
04:01:12 Quit AllenMockery (Read error: 110 (Connection timed out))
04:18:21 Join Leperkawn [0] (
04:21:31LeperkawnOk, just making sure I was in.
04:21:52LeperkawnCould anyone here give me the latest on the iPod video Rockbox?
04:22:03LeperkawnI just saw it in the CVS a few moments ago.
04:22:20 Quit Hooligan (Read error: 104 (Connection reset by peer))
04:22:21ashridahit doesn't play music or anything useful, last i heard
04:23:32LeperkawnIs there a wiki page for it or anything? I couldn't seem to locate one.
04:24:46ashridah and
04:24:54ashridahi don't know why they're not directly attached to the front page
04:25:11 Quit ehntoo ("Leaving")
04:25:14LeperkawnThankee sai
04:25:23LeperkawnThankee sai.
04:25:26dwihnoashridah: Do you know if it is possible to use the ipod mass-storage style?
04:25:34LeperkawnWhoops, double posted.
04:25:39LeperkawnI do it with my video
04:25:42ashridahdwihno: no
04:25:45ashridah USB handler 0%
04:25:51ashridahsounds unlikely
04:25:53Leperkawnyeah you can
04:26:03Leperkawni drag and drop onto my video just fine
04:26:14Leperkawnor.. are you talking about the port?
04:26:16dwihnowithout additional drivers or anything?
04:26:30Leperkawnhmm, you got me there
04:26:38ashridahdwihno: i don't know if it just uses usb-storage-on-a-chip or what
04:26:38LeperkawnI just know that i can from normal ipod use
04:26:47LeperkawnI'll shut up now >_>
04:27:00ashridahLeperkawn: that doesn't tell us if it's software-usb or hardware. i'd be inclined to assume the latter
04:27:18dwihnoI hope it's hardware
04:27:37ashridahbut yeah, ipods work as a normal hard drive if so required, it just doesn't let you add music to itunes that way
04:27:37dwihnoBut assuming is bad, so I'll just sit tight and wait for information ;)
04:28:03 Quit San (Read error: 110 (Connection timed out))
04:28:43dwihnoThe ipod port will be great anyways!
04:29:08dwihnoRemoving the need for itunes would also be a great accomplishment.
04:29:44ashridah(not easily anyway)
04:30:20LeperkawnAs long as an itunes similar program still works
04:30:31LeperkawnI kind of liked the manager as opposed to drag and drop.
04:31:21dwihnowith mass storage compliance, there are unlimited ways to sync your music
04:31:30Paul_The_NerdWell, Rockbox isn't really designed around a manager concept, though I suppose someone could create one compatible with the TagDB.
04:32:08LeperkawnI could live without it though. I did before =P
04:32:31LeperkawnWell, I just realized something. I can still use iTunes/other manager, then just copy the music over to rockbox.
04:32:50Leperkawncopy the music manually*
04:33:15Paul_The_NerdWell, iirc the iTunes database is mostly known, so support for it could be created theoretically.
04:33:40*ashridah pats his copy of rsync
04:34:00ashridahtwo-way management, none of this 'To-device-only' junk :)
04:34:06LeperkawnThere's some other programs out there that are said to be better than iTunes (faster too), but I /just/ got an ipod recently so I have not used them yet.
04:34:30LeperkawniTunes is indeed shit in the music transferring category.
04:34:49dwihnoashridah: rsync wouldn't sense file name changes?
04:35:42ashridahdwihno: no, it ends up copying the file again after deleting the old one, but i rarely do that nowadays.
04:37:21dwihnoSo how does it know which is the master?
04:37:39dwihnoBad formulation.
04:37:48ashridahdwihno: directionality
04:37:59ashridahi have a 'toplayer' and 'fromplayer script
04:38:18dwihnoscript? isn't rsync operated like a command tool?
04:38:32ashridahjust got to be careful not to use the wrong one. i might scrounge around and find/make better system eventually.
04:38:39ashridahdwihno: yeah, so?
04:39:00dwihnoashridah: are there a plenthora of options?
04:39:13ashridahnot really, but i'm lazy :)
04:39:28dwihnocare to share? I might test it under windows
04:39:28*ashridah comes from the "do it once and save it in a script" mentality of system administration
04:40:36lostlogicashridah: do you use the script tool to make the scripts?
04:40:38 Join Zoric [0] (
04:41:56ashridahlostlogic: for a one liner? not really, no
04:42:03 Quit Leperkawn ("CGI:IRC")
04:42:13lostlogicdidn't notice that it was a oneliner ;)
04:43:15ashridahbut generally, i don't use script, no.
04:43:15 Join Hooligan [0] (
04:43:32lostlogicme neither... I was just wondering.
04:43:39ashridahif i'm anticipating something large, i'll typically start off by writing the script first
04:44:46 Quit Zoric (Read error: 104 (Connection reset by peer))
04:59:13 Quit actionshrimp ("a bird in the bush is worth two in your house")
05:13:15dwihnoashridah: I just found an utility named unison... Have you tried it out?
05:18:07 Join Midgey34 [0] (
05:25:08Paul_The_NerdIs anyone encountering "sound.c:68: error: parse error before '<<' token" when compiling from CVS?
05:28:06Paul_The_NerdOh, never mind
05:28:14Paul_The_NerdWait, yeah, it's still doing it.
05:30:34 Quit Rob2222 (Read error: 110 (Connection timed out))
05:30:37 Quit Hooligan (Read error: 104 (Connection reset by peer))
05:30:57 Quit TCK (Read error: 104 (Connection reset by peer))
05:46:19***Saving seen data "./dancer.seen"
05:58:08 Join Hooligan [0] (
06:03:20 Part Midgey34
06:05:00 Join San [0] (
06:17:39 Join Rob2222 [0] (
06:25:26 Quit Hooligan (Read error: 104 (Connection reset by peer))
06:36:05 Quit RotAtoR ("zzzzzz")
06:46:19 Join DJDD_ [0] (
07:06:35 Join Hooligan [0] (
07:14:46 Quit San (Read error: 110 (Connection timed out))
07:43:25 Quit DrumRBoy320 (Read error: 110 (Connection timed out))
07:45:56 Quit Hooligan (Read error: 104 (Connection reset by peer))
07:46:23***Saving seen data "./dancer.seen"
07:47:30Paul_The_NerdOkay, for the log, nevermind about the question I asked earlier. A clean cygwin install helped a lot
07:48:45 Join Membrillo [0] (
07:49:10Membrillolinuxstb: I have my ipod video with me now
07:51:28 Quit phaedrus961 ("Leaving")
07:51:48Membrilloi can test out the build when you want
07:55:12 Join phaedrus961 [0] (
07:59:52 Part Sando
07:59:55 Join Sando [0] (
08:01:23 Join Bger [0] (n=Bager@
08:01:53Paul_The_NerdI got it all fixed.
08:02:04Bgerit seems your compile error is caused by conflicts...
08:02:18Bgeri mean conflicts between the cvs version and your local one
08:02:28Paul_The_NerdI tried a clean CVS download and it still didn't work.
08:02:38Bgerhms ?
08:03:43Bgerok, just wanted to help
08:03:51Paul_The_NerdAs I said, I got it working.
08:04:21Paul_The_NerdA script to help download the toolkit for GAIM apparently caused it, though I haven't worked out exactly where yet.
08:05:19Paul_The_NerdTo top it all off, the script was outdated enough that it didn't matter.
08:05:51Bgervery strange ...
08:07:25Paul_The_NerdI was going to try to compile guifications against 2.0, but the script automatically downloaded a bunch of 1.X specific GAIM files, and actually broke my GAIM installation (replaced some DLLs.) Well beyond "useless" and into "irritating"
08:11:01Bgerblah :(
08:11:05 Quit Bger ("brb")
08:11:48 Join Bger [0] (n=Bager@
08:23:40CassandraBleeding nora. Is that the time?
08:24:00CassandraCurse my obsessive compulsive personality.
08:26:31 Join ender` [0] (i=ychat@
08:41:16 Quit DreamTactix291 (Read error: 104 (Connection reset by peer))
08:41:36 Join AllenMockery [0] (
08:47:07 Join DreamTactix291 [0] (
08:50:12BagderCassandra: sounds like a neat tool!
08:52:22BagderCassandra: I'm quite sure you'll get questions about how your tool differs from Membrillo's
08:53:09 Join San [0] (
08:54:05Membrillosay what?
08:54:12Membrilloive heard my calling
08:55:23 Join _FireFly_ [0] (
08:55:39 Join Zagor [0] (n=bjst@pdpc/supporter/sustaining/Zagor)
08:55:41Membrilloso its like a hardcore version of mine?
08:55:51Bagderto me it sounds like that
08:56:06Bagderbut I'm not on windows so I've only read the descriptions of your projects
08:56:31Paul_The_NerdMembrillo: From the looks of it, it's also a replacement for the Archos installers.
08:56:58Bagderyes, it should support all rockbox targets as I've understood it
08:57:37Membrilloso im being bumped off am i :)
08:57:55Paul_The_NerdAll this means is that you need to make yours even better. :-P
08:58:12Membrillowell i think mine will win on simplicity :P\
08:58:38BagderI can't see any harm in having multiple tools around
08:59:03CassandraMembrillo's? What's that?
08:59:14 Quit elinenbe (Read error: 110 (Connection timed out))
08:59:40Membrillogo to MR −−> h1xx or h3xx −−> rockbox downloaders
09:00:01CassandraDoesn't suppord iPod yet, since they don't seem to be in the build system properly.
09:00:05Paul_The_NerdCassandra: Your tool doesn't work on my computer
09:00:15CassandraBut it's just a matter of editing a config file when they come in.
09:00:21CassandraWhat happens, Paul?
09:00:38Bagderright, there are no daily ipod builds yet
09:00:54Paul_The_Nerd01:57:52: can't open file 'C:\Documents and Settings\DarkkOne\Application Data\rbutil\download\' (error 3: the system cannot find the path specified.)
09:00:55Paul_The_Nerd01:57:52: Download URL: Can't create output stream (wxSTREAM_WRITE_ERROR) writing to download /C:\Documents and Settings\DarkkOne\Application Data\rbutil\download\
09:00:55Paul_The_Nerd01:57:52: Install: Unable to download
09:01:53CassandraPaul_The_Nerd, that's kind of weird. It's supposed to make the directories itself (and does on my PC).
09:02:44Paul_The_NerdClearly is simply does not love me.
09:03:55 Quit AllenMockery (Read error: 110 (Connection timed out))
09:05:15CassandraI tell a lie, Paul. It's stopped creating the download directory.
09:05:21CassandraBetter fix that.
09:05:52 Quit nathanh (Read error: 104 (Connection reset by peer))
09:05:59 Join nathanh_ [0] (
09:06:06Membrilloit looks cool cassandra. More user friendly than mine, although a few more clicks :P
09:06:20 Join einhirn [0] (
09:07:36 Quit nathanh_ (Read error: 104 (Connection reset by peer))
09:08:23CassandraIt's particularly aimed at the kind of people who can't cope with unzipping, although I think others'll find it useful too.
09:09:28BgerCassandra the same problem here
09:09:46Membrilloyeah i took the same approach, although, i tried to aim for doing it quickly as well. For mine i just open the exe, select daily or bleeding from radio buttons then click go and it downloads, unzips to the root of the exe, then deletes the zip
09:09:46CassandraYeah. I've figured out why.
09:09:56Bgerokay :)
09:10:00CassandraNever assume C++ strings are 0 terminated.
09:10:07CassandraGive me a few minutes to fix.
09:11:59_FireFly_Cassandra why should c++ strings not 0 terminated ??
09:15:12Bger_FireFly_ i think she means something like "class string"
09:16:36CassandraFirefly - because they can contain \0. They have a length field instead.
09:18:30_FireFly_Cassandra afaik all strings are null terminated
09:19:06_FireFly_the length field in a string-class is only a kind of replacement for the strlen()-fn :) afaik
09:20:13Bger_FireFly_ only if the string is not unicode ....
09:21:03Bgeri.e. unicode strings can contain \0 as part of the string itself
09:23:04 Join DrumRBoy320 [0] (
09:24:34CassandraRight. New version up. Should work now.
09:25:10Bgerthe same link ?
09:26:16Cassandra(Current known bug is that it won't cope well if you try to download a daily before it's been put up on the website.)
09:26:24 Join nathanh [0] (
09:26:43Cassandra(I had a workaround for this, but it hangs the app rather badly for 10 secs, so I disabled it.)
09:27:05CassandraActually it will cope. It just won't work.
09:28:17 Join Kohlrabi [0] (
09:28:47 Join Jungti1234 [0] (n=jungti12@
09:28:59 Part Sando
09:29:01 Join Sando [0] (
09:29:09linuxstbCassandra: Do you need someone to work on a Mac version of your new utility?
09:29:18nathanhurl for utility?
09:29:37BgerCassandra one error
09:29:39linuxstbnathanh: See the post at
09:29:54linuxstb(General Forum)
09:29:58Bgerthe h300 build's link is in build-h300 dir
09:30:03Bgernot build-h320
09:30:13 Quit Membrillo ()
09:30:32Bgeralso for the daily
09:30:36Bgerh300, not h320
09:31:25Bgerand the filename is, not *-h320-*
09:32:18linuxstbdwihno: To answer your question from 5 hours ago, yes, the ipod acts as a standard UMS device. It's just that Apple's firmware will only play music that has been added to the itunesdb on the device.
09:32:29CassandraThey use the same string.
09:32:39CassandraUploading an updated version now.
09:33:04CassandraIn the final release, hopefully the config file will be self-updating.
09:34:06CassandraOK. Done.
09:34:07linuxstbCassandra: Will your utility also incorporate the fwpatcher functionality for iriver?
09:34:22BgerCassandra also, are you sure your utils really does *full* uninstall
09:34:29Cassandralinuxstb, that's something that's high on the To Do list, yes.
09:34:59CassandraBger, it wipes the .rockbox dir and checks for all possible files in the root.
09:35:04 Quit Kohlrabi (Read error: 104 (Connection reset by peer))
09:35:10Bgerand registry ?
09:35:13Cassandra(rockbox.* ajbrec.ajz archos.mod
09:35:27Bgeri mean the utility's uninstall
09:35:32CassandraOh, right.
09:35:37CassandraNo I don't think it does.
09:35:54CassandraThat's standard practice.
09:36:00linuxstbCassandra: Something the ipod port needs is a way for Windows users to install the bootloader. It just so happens that ipodlinux has a wxwindows based gui installer for Windows which does a similar job :)
09:36:22Cassandralinuxstb, I'll happily incorporate it.
09:36:25BgerCassandra ah, at last :)
09:37:07CassandraIt worked?
09:37:07Bger(registry) that's somethink that really annoys me in most uninstalls
09:37:36Bgeryes :)
09:37:41CassandraBwahaha! My two enormous registry keys will suck up your resources for ever more!
09:37:51linuxstbCassandra: The installer is here:
09:37:59Bgerdidn't try it on the device
09:38:16CassandraI did. Updated to the latest daily on my H120.
09:38:19BgerCassandra ok, you're right, but the problem is that nearly everyone thinks this way
09:38:45 Join Membrillo [0] (
09:38:57CassandraBger: On the other hand, some people regard it as a convenience. You can't win, really.
09:39:09Bgeranyway, i downloaded it just for bugtesting ...
09:39:22Bgerhm, "make it option" :P
09:39:38CassandraIf I can be arsed I wiil.
09:39:48CassandraDepends how easy it is to do in Inno Setup.
09:39:48 Quit perplexity ("*pop*")
09:40:05linuxstbCassandra: Why do you need to install your installer? It should be possible to create a standalone .exe file statically linked to the wxwidgets libraries.
09:40:41Membrillolinuxstb: i have my ipod video with me now
09:40:56linuxstbThat way, you can just keep the .exe file on your device and upgrade from anywhere.
09:41:03Cassandralinuxstb: because of the external config file.
09:41:10_FireFly_linuxstb or a simple zip file with al neede files in it
09:41:35linuxstbMembrillo: Sorry, I don't have time now - need to start my real work. I should be able to help this evening (European time).
09:41:46Membrillook sounds good
09:42:01Membrilloits 741pm now
09:42:04linuxstbWhat's in the external config file?
09:42:08Membrilloill be up till about 2am
09:42:10Membrillowill you be around?
09:42:25CassandraPlatform details etc.
09:42:47_FireFly_Cassandra I would put it in a zip file
09:42:58linuxstbMembrillo: That's probably too early - I'll still be working. Maybe I'll catch you when you wake up tomorrow.
09:43:04 Quit DJDD_ (Read error: 110 (Connection timed out))
09:43:26Cassandra_FireFly_, I don't see why the installer is such a big deal, really.
09:43:43_FireFly_why use an installer for so a small util ??
09:43:44linuxstbIt means you can't install it and run it from your device.
09:44:08_FireFly_why to mess up the registry if it isn't really needed
09:44:30Jungti1234hey all
09:44:31Jungti1234I wonder.
09:44:43dwihnoThe registry is the tool o' the devil!
09:44:49Cassandra_FireFly_, because I find it annoying to have to enter the same details over and over again.
09:45:01Jungti1234Day before yesterday OGG Q10 of build that I make was played normally, but Q10 of Rockbox's build isn't played normally today.
09:45:14Jungti1234Why is it?
09:45:18_FireFly_Cassandra which details ??
09:45:20linuxstbCassandra: You could just store the config details in a text file in the same directory as the exe is located.
09:45:35CassandraLocation of player and what type it is.
09:45:39linuxstbwxwidgets supports ".ini" style text configuration files.
09:45:45CassandraThat's all that's stored in the registry.
09:45:58linuxstbAh, the windows way...
09:46:03Cassandralinuxstb, wouldn't be portable.
09:46:05_FireFly_to make it slow
09:46:10CassandraYou can't do that in Linux.
09:46:12Jungti1234Does no one know reason?
09:46:19linuxstbCassandra: How can a text file not be portable?
09:46:19_FireFly_Cassandra with a config-file you can
09:46:24***Saving seen data "./dancer.seen"
09:46:29linuxstbYour registry method is the non-portable option.
09:46:46CassandraTry sticking your text config file in /usr/bin along with your executable.
09:46:57CassandraSee how well that works.
09:47:07linuxstbOn Unix, you just create it in $HOME/.rockboxinstaller
09:47:21CassandraThat's what the current program does.
09:47:26 Join MrShlee [0] (
09:47:34_FireFly_Cassandra no it doesn't at least under windoes
09:47:56_FireFly_and not all will be install this installer under linux into /usr/bin
09:48:04MrShleeHey, is there a way to trigger a plugin everytime a song loads? im guessing there isnt a simple trigger
09:48:08_FireFly_they will create a dir in his/her home-dir
09:48:14CassandraThat's because wxWidgets uses the native way of the underlying OS.
09:48:21linuxstbCassandra: I'm just thinking that the program will be even more useful if you can install it on your device - so the installation is specific to that device, not specific to a certain computer.
09:48:31Cassandralinuxstb, you can.
09:48:47_FireFly_no you can't when registry entries are needed
09:49:05CassandraThey aren't *needed*. They just store your default choices.
09:49:08_FireFly_because if the device is used on a windows-pc whre this entries are missing
09:49:25CassandraThe program creates them on the fly. Not the installer.
09:49:27_FireFly_why don*t you put the default values into the program
09:49:29Jungti1234Does no one know reason?
09:49:29CassandraGive me some credit.
09:49:42CassandraBecause they're per user defaults.
09:49:56_FireFly_under winxp there is also an home-dir
09:50:06_FireFly_so no need for an registry entry
09:50:43CassandraLook, that's the way wxWidgets implements it. Go argue with them about their design philosophy.
09:51:12_FireFly_i don't think so that they depends on using the registry under windows
09:51:27_FireFly_when storind "default" config-values
09:52:00linuxstbCassandra: So you can't be persuaded to ditch the installer? :)
09:54:14Cassandralinuxstb, the installer isn't necessary.
09:54:36linuxstbI've tried your program under Wine, and it crashes with a fatal error "Can't create data directory" (0.1a)
09:54:49CassandraCopy rbutil.exe and rbutil.ini into the same dir on the player, and it should just work.
09:55:04 Join yngwi [0] (
09:55:19linuxstbCool. So we could offer two different versions - a zip file with the exe and ini, and the full installer package.
09:55:31CassandraI have no problem with that.
09:56:15linuxstbDo you need someone to work on the Mac port?
09:56:16CassandraI can even provide an option from within the program to install itself onto the MP3 player.
09:56:35CassandraThe code needs cleaning up first though.
09:56:49linuxstbI can help with that, but it won't be until after the new year.
09:57:02Paul_The_NerdWould it be possible to add a "don't prompt me in the future" option?
09:57:09CassandraThat's OK. I'm not likely to do much before new year.
09:57:26Paul_The_NerdLike, if I *always* want the daily build for h120/140, I could just double click the .exe and no more input is needed (after the first run, where I've set that.)
09:57:36CassandraPaul_The_Nerd, probably.
09:57:47Paul_The_NerdIt seems if it's gonna go on the player, a fire and forget version might be ideal.
09:58:11CassandraWell, future versions should offer more than just install.
09:59:02 Quit MrShlee ("Changed by popular Request - "I am the heart of your darkness, you stinking oompa-loompa morons!"")
09:59:33CassandraEssentially I released it because it's a useable skeleton of the final app.
09:59:35 Join DJDD_ [0] (
09:59:46CassandraDefinitely 0.1 rather than 1.0.
10:00:03_FireFly_Cassandra you can always use wxFileConfig to have only a config-file
10:00:16_FireFly_so no need to use the regestry
10:00:29CassandraI know. That's how the ini file works.
10:00:40 Quit Vladoman (Read error: 110 (Connection timed out))
10:00:41_FireFly_This class is used by default for wxConfig on Unix platforms but may also be used explicitly if you want to use files and not the registry even under Windows.
10:01:05CassandraI really don't see this as as much of an issue as you do.
10:01:39 Join Vladoman [0] (
10:01:41_FireFly_Cassandra forgive me but i hate tools which use the registry if it isn*t really needed
10:02:45CassandraI might change it just to make you happy.
10:02:54CassandraBut it's really not a priority for me.
10:03:13Bgerthis'll make me happy too ;)
10:03:34_FireFly_the problem , which i see, is that your programm will create the registry-entrys upon start when they didn't exist
10:03:38_FireFly_am i right ??
10:03:39 Quit yngwi ("Chatzilla [Firefox 1.5/undefined]")
10:04:01linuxstbIs it possible to create the installer packages of your utility under Linux (i.e. on the build server)? Or can it only be done in Windows?
10:04:29_FireFly_so then you will clutter the registry on PC where no uninstall-programm for this exist, when the programm was ran from the device
10:04:34Cassandralinuxstb, it's the same installer we use for the old Windows installer. It can be done in the build system.
10:05:16Cassandraconfig file, registry - it's all clutter somewhere.
10:05:39_FireFly_Cassandra not at all for config-files if they are in the same dir as the app
10:06:05linuxstb_FireFly_: You don't like the registry, do you? :)
10:06:05_FireFly_especially when the program is located on the target-device
10:06:14Cassandra_FireFly_, that's really not a sensible thing to do if you're trying to write for Unix.
10:06:34_FireFly_linuxstb not really they make windows slow when it's get bloated from crappy programs and uninstall-programms
10:06:44nathanhanybody else having difficulty getting to
10:07:09linuxstbnathanh: Just tried it - loads perfectly.
10:07:28_FireFly_i have also no problems
10:07:37_FireFly_maybe a isp problem at your location
10:07:43_FireFly_or routing-problem
10:07:48linuxstbI am logged into sourceforge if that makes any difference.
10:07:51nathanh works
10:08:05nathanh gives me a "connection dropped"
10:08:08nathanhill try another browser
10:09:39_FireFly_linuxstb the registry-idea it self is not so bad but how M$ had it implemented in windows is crap
10:10:20_FireFly_and also why uninstall-programms thinks not to remove the entries when the program gets uninstalled
10:10:31CassandraRight. I'm off to bed.
10:10:41_FireFly_Cassandra good night
10:10:54CassandraMorning. :(
10:11:08linuxstbI see David Bryant has made some ipod fixes for wavpack...
10:12:00Bagdertime to get more coffee!
10:13:06Paul_The_NerdI strongly suspect to have an iPod after Christmas (or by January 12th). You guy are gonna audio working by then, right? Right? wink wink, nudge nudge. (Hey, someone has to pester with the "when's it gonna get dones" to justify the Topic now that H3X0 is doing pretty well.
10:14:52linuxstbPaul_The_Nerd: Which ipod?
10:15:28Paul_The_NerdThe Nano
10:15:37linuxstbThe Nano and Color/Photo will have audio support first - because audio is working in ipodlinux. The Video is an unknown beast.
10:15:41Lynx_awyhmm, are there known problems with the h340 usb mode, with original firmware? i just copied something, and then windows complained about a full disk, and the drive disappeared. the h340 still shows 'connected' though
10:16:03Lynx_awyoh, and the disk is not full, of course
10:16:07Paul_The_NerdI'm really happy with my iHP120, but I was craving a flash-based player.
10:16:07 Nick Lynx_awy is now known as Lynx_ (
10:16:29 Join Verbal91 [0] (
10:18:26Paul_The_NerdOn an unrelated note, how many people in here play games of one sort or other on their Rockbox-equipped player, and if so, how long do you spend on them, and do you think you'd be interested in more complex games? (Like, a platformer, or RPG?)
10:18:53nathanhpaul_the_n: not me, zero, no
10:19:35BagderI played bejeweled once on Linus h320, that looks nice
10:19:47linuxstbThe "old" or "new" version?
10:19:53Bagderthe latest
10:20:46BagderI'm a lousy rockbox user, I just play music on my players
10:20:53Membrilloim less worried about games. More anticipating getting full colour functionality of the LCD
10:21:37Membrilloand hopefully line in recording by febuary 17th!!!
10:21:46Membrilloi want to bootleg a James Brown concert
10:27:31preglowi want to sleep
10:28:45nathanhi want a blonde
10:29:22Paul_The_NerdI want an XBox 360 running Linux.
10:29:30Paul_The_NerdY'know... while we're all being unrealistic.
10:29:38linuxstbChristmas is coming...
10:29:49BgerPaul_The_Nerd why unrealistic ? :))
10:30:01Paul_The_NerdWell, unrealistic any time *soon*
10:30:07Bgeror PS3 running linux ? :)
10:30:14Paul_The_NerdI'm not sure which I'd prefer
10:30:46preglowlinuxstb: yo, you tried wavpack after bryant's fixes?
10:30:54Paul_The_NerdWord is the 360 actually gives you more flexibility in some areas, whereas the PS3 is just flat out faster. John Carmack claims the 360 is actually better for his purposes, or did at some point, for what little it's worth.
10:31:05linuxstbpreglow: No. I wonder what made him do that?
10:31:25Bgerwho's John Carmack ?
10:31:33preglowhe's mailed me, and he said he suspected the ultra fast decode speed of wavpack was because of some decoding error
10:31:43preglowwavpack will zero out a block and move on if it detects an error
10:31:49Paul_The_NerdHe used to be a Hotshot Game Designer back in like... '92ish. :-P
10:31:50preglowand he knew of spots in wavpack that would be bothered by unsigned ints
10:32:12linuxstbI'm hlad he's still following Rockbox.
10:32:13preglowBger: lead coder/tech guy/whatever in id software
10:32:29preglowso he says it might be quite a bit slower now
10:32:53_FireFly_but without decoding errors :)
10:33:31Bger10x ... it's obvious that i don't follow the game industry
10:33:36BgerPaul_The_Nerd link ?
10:33:55nathanhhe's still hotshot
10:34:12 Quit San (Read error: 110 (Connection timed out))
10:34:45preglowdavid bryant might get a nano :-P
10:35:19Paul_The_NerdBger: The statement was made something like 3 or 4 months ago. I wouldn't know where to find it any more. =/
10:35:46Bgerno problem
10:35:54Membrilloreading up ^^^^ my xbox runs linux
10:36:38Membrilloim not sure if the 360 has been attacked htough
10:36:56Paul_The_NerdThey're working on it.
10:37:25Membrilloi guess theyre already running unsigned code?
10:37:31Paul_The_NerdI don't think they are yet.
10:37:45linuxstbpreglow: Just done a quick wavpack test, and 60 seconds of audio plays in 46s
10:37:54Membrillowere microsoft stupid enough to leave bios override ports open like the xbox?
10:38:14Paul_The_NerdI don't really know, but it sounds like it's a *lot* tougher.
10:38:33Paul_The_Nerdlinuxstb: Did you guys ever figure out why ogg and mp3 were so slow?
10:38:36Membrilloin comparison to modding an xbox, anything is easy
10:38:44linuxstbWell, the hacking of the original xbox was well-documented - so I'm sure Microsoft have learned from their mistakes.
10:38:45Membrilloanything is hard i mean
10:39:12Membrillowell, leaving bios override ports open, was the main problem
10:39:35linuxstbPaul_The_Nerd: No, but I expect the codecs are not running at full speed - the playback code will be slowing them down artificially. But I still don't understand that code properly.
10:39:46preglowlinuxstb: and mp3 and ogg still isn't realtime for you?
10:40:01 Quit _FireFly_ ("IceChat - Chillin with the Best of em")
10:40:09preglowi wonder why the hell it's so fast here
10:40:21linuxstbIt could be the disk accesses slowing it down for me.
10:40:32linuxstbBut then again, the disk stops...
10:40:37linuxstbSo it's not that.
10:40:39 Quit Jungti1234 ("Bye~ -")
10:41:49 Join AllenMockery [0] (
10:42:06preglowthe ipl people had encountered no cache problems
10:42:29preglowthat is, josh, which deals with dynamic libs, has only got a 4g
10:42:38preglowbut he says nano users use it without problems
10:42:54preglowbut then again, they do stuff differently to us
10:43:03preglowthey might not even load stuff at the same address all the time
10:43:03linuxstbMaybe it's a timing problem. If the cache is enabled, then it runs slower.
10:43:10linuxstbI mean disabled.
10:43:11 Part AllenMockery
10:43:17preglowi can't imagine what kind of timing problem that could be
10:43:30preglowthere are some cache invalidate regs specifiec
10:43:40preglowperhaps i should just play with them a bit
10:43:47preglowor i could dig out ida pro and have a day of fun instead
10:44:16linuxstbDo you know anything about audio on the 5g? Does it go via the broadcom chip?
10:44:41preglowthey don't know
10:44:50preglowi can't imagine why it should, though
10:44:53preglowthey're just using it for audio
10:45:07preglowand i don't think apple is in the business of extending their products too far
10:45:40linuxstbIt's a similar WM codec chip, so it could just be a simple change.
10:46:33 Join nathanh_ [0] (
10:46:34 Quit nathanh (Read error: 104 (Connection reset by peer))
10:52:43Membrillolinuxstb: are you available now for help with the video rockbox>
10:53:22linuxstbWe can give it a go, but I might have to disappear.
10:53:30Membrillook no probs
10:53:35Membrilloi can work it out
10:53:43Membrilloim not a complete retard
10:53:50Membrillobut i might need help in places
10:54:02linuxstbAre you in Linux with your video plugged into usb?
10:54:22linuxstbI'm just preparing some binaries for you now, but there is some preparation you'll need to do.
10:54:29 Quit DJDD_ ("Trillian (")
10:54:44Membrilloi have colinux but im struggling to work out how to use it
10:55:24 Join nathanh [0] (
10:55:24 Quit nathanh_ (Read error: 104 (Connection reset by peer))
10:55:34linuxstbYou need to open up a terminal (shell) window.
10:56:55Membrillohow do i do that
10:57:08Membrilloits not as simple as opening the exes is it
10:57:33linuxstbHave you started Linuix?
10:57:51Membrilloim in windows
10:59:08Membrillodo i have to boot up colinux, or do i run it under windows
11:00:34linuxstbYou need to follow the instructions that came with colinux and somehow start linux. I can't help with that.
11:03:22Membrilloill find a readme
11:04:31 Join Polo_o [0] (
11:15:37 Join nathanh_ [0] (
11:15:37 Quit nathanh (Read error: 104 (Connection reset by peer))
11:16:41Membrillowould this be ok
11:19:00markunMembrillo: no, I don't think so. You could run vmware with linux if you want
11:19:09 Part Paul_The_Nerd
11:27:06Membrilloi dont understand why i need linux at all
11:28:03BagderI'm sure it can be done in plain windows too
11:28:12Bagderbut then you're even more on your own I'd guess
11:30:11amiconnlinuxstb: Your iPod video scan is all black except the menu symbols ?? :(
11:30:30amiconns/menu symbols/button symbols/
11:30:34 Join perplexity [0] (
11:34:24perplexityHrm.. can't boot my h340 from the h300-non-lcd remote as it tells me the hold switch is on :(
11:34:51nathanh_wheres a good place to buy the h300 lcd remote
11:35:00 Join LinusN [0] (
11:35:42 Join yngwi [0] (
11:36:10preglowamiconn: well, it's a black ipod
11:38:48amiconnStill it's way too dark. I dont't see anything apart from the button symbols. The nano image I used is also a black one...
11:42:40 Join _FireFly_ [0] (
11:46:28***Saving seen data "./dancer.seen"
11:46:53linuxstbamiconn: I agree the scan is dark, but on my monitor I can see (faintly) the outline of the LCD and the clickwheel.
11:47:01linuxstbI tried googling for a good image, but couldn't locate one.
11:48:41perplexitycan you just load it up in gimp and change the gamma of the image?
11:49:35nathanh_ooo, top gear on telly
11:49:38*nathanh_ disappears
12:04:55linuxstbAm I right in thinking that Rockbox organises the memory as follows: [rockbox.bin, bss, Audio Buffer, Codec Buffer, Plugin Buffer]
12:05:16 Nick ashridah is now known as Lost-ash (
12:05:39 Join Zak1392 [0] (
12:05:52Zak1392hey guys
12:09:31Zak1392thanks for the warm welcome!
12:12:53 Join Moos [0] (
12:13:14LinusNlinuxstb: sounds familiar
12:14:25linuxstbI'm just wondering how we could support the fact that video ipods come in 32MB and 64MB versions (the 30GB and 60GB versions).
12:14:44linuxstbIs there a reason the audio buffer isn't the last thing in memory?
12:15:20linuxstbAs memory size is (I think) the only difference, it would be nice not to have to add another target.
12:15:55Bagderwe did that to have the plugins at a fixed address
12:16:01LinusNi believe the reason is that the plugins don't need to be relinked for every rockbox update
12:16:04linuxstbYes,it's just clicked.
12:16:16 Join San [0] (n=test@
12:16:47 Quit Moos (Client Quit)
12:16:53linuxstbSo it seems we have no choice but to have separate builds for 32MB and 64MB players.
12:16:53 Join Moos [0] (
12:17:05 Join Jungti1234 [0] (n=jungti12@
12:17:20LinusNlinuxstb: looks like that
12:19:27linuxstbUnless the codecs and plugins went at the start of RAM, before rockbox.bin. I think that would work on the iPods.
12:20:01 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
12:20:06 Join einhirn [0] (
12:20:19Bagderand you need to be able to autodetect amount of ram
12:20:50linuxstbIf all else fails, that can simply be based on the hard disk size - 30GB = 32MB, 60GB = 64MB.
12:21:18linuxstbOr a config option.
12:22:05Bagderfood time
12:23:28nathanh_swoit, i have a semi-working SDL version of the sim
12:33:11LinusNnathanh_: wee
12:40:07 Quit edx (Read error: 110 (Connection timed out))
12:42:03 Join Paul_The_Nerd [0] (
12:42:46Bagdernathanh_: you talked to bazz ?
12:43:07nathanh_who is bazz?
12:43:16BagderI believe it was he who also works on an SDL version
12:43:54linuxstbnathanh_: Are you working on Windows or Linux?
12:44:10linuxstbThat could be useful then - I think bazz was working with windows.
12:46:16BgerLinusN did you try initializing the h300's lcd according to the datasheet?
12:46:28 Quit Zak1392 ()
12:51:35*preglow pushes for a relocatable plugin format!
12:52:11BagderI won't stand in your way ;-)
12:52:31preglowit will yield bigger code and bigger plugins, though
12:52:32LinusNBger: no
12:52:36preglowbut i for one think it would be nice
12:52:44nathanh_double swoit, i have h340 sdl simulator now
12:52:51nathanh_finally, no more plugging and replugging to test my code :-)
12:53:06preglowohh, i look forward to just one sim target
12:53:08preglowthat will be sweet
12:53:24Bagderit'll be a cheerful day
12:53:34preglowand a working target as well
12:53:40preglowi still can't get sound out of any of the sims
12:56:15preglowbut yes
12:56:36preglowrelocatable plugins might be useful, i'll try to research how hard it'll be
12:56:53preglowbut now i need to run some errands, bbl
12:57:05*nathanh_ nudges preglow with an elbow... "psst, while you're at it, research how hard it will be to slip a malloc past the censors"
12:57:47nathanh_hrm, ive got colour issues, whats the packing format for RGB565 unswapped?
12:57:58*preglow swats the malloc fly
12:59:06amiconnpreglow: Relocatable plugins will add complexity to the loader, and I don't see the need...
12:59:13linuxstbnathanh_: It's simply 5-bits Red, 6-bits Green and 5-bits Blue stored in an unsigned short - Blue is the least-significant 5 bits.
12:59:30nathanh_thats what i thought, but all my colours are coming out wonky
12:59:35nathanh_the remote looks fine, 1bit depth
13:00:34 Quit Polo_o (Read error: 113 (No route to host))
13:01:19Bagdernathanh_: compare with the win32 sim code
13:01:23Bagdersince that works
13:01:26nathanh_thats what im doing
13:05:49LinusN^BeN^: yes?
13:06:02 Quit Verbal91 (Read error: 113 (No route to host))
13:06:22^BeN^the configfile_save
13:06:28^BeN^doesnt work for me
13:06:41^BeN^it returns an error
13:06:47nathanh_wee, colors good now
13:07:31LinusN^BeN^: oh?
13:07:58^BeN^returns error
13:08:17^BeN^i checked it in win32 sumilator
13:09:09LinusN^BeN^: can i see the code?
13:09:34^BeN^all the source
13:09:39^BeN^or only this part
13:10:38preglowamiconn: the need is that we one day might want to load more plugins than just one, plus the fact that the current system of having to statically linking the plugins to the end of memory obviously clutters the build system
13:10:52preglowamiconn: plus, it should add veyr little complexity, i'll have a look at it later
13:11:16LinusN^BeN^: i'd prefer the entires source, so i can compile it
13:11:26^BeN^ok moment
13:11:46preglowi'm talking a custom format here, not elf or anything
13:11:54preglowas minimal as can be
13:13:15nathanh_might be worth picking a minimal standard like COFF, rather than reinventing the wheel
13:13:26 Join amiconn_ [0] (
13:13:39preglowcoff is more complex than what i'd like
13:13:52preglowi don't want to add that kind of complexity to the loader
13:14:01preglowi'd like sectioning and fixups, that's that
13:14:07 Join slowcoder [0] (
13:16:18LinusN^BeN^: there seems to be a firewall problem here
13:16:27^BeN^ok moment
13:16:32^BeN^i upload it to lycos
13:17:05preglowbut ok, i got stuff to take care of, alter
13:17:06preglowlate <-
13:17:08preglowlateR <-
13:20:09 Part Sando
13:22:47 Part Paul_The_Nerd
13:23:04Membrillodoes anyone hear no Visual Basic?
13:24:11markunMembrillo: no, didn't hear it :)
13:25:06linuxstbmarkun: I'll say it again - "no visual basic!"
13:25:47Membrilloim not thinking what im typing
13:25:56Membrilloi had two typos in one sentence... thats shocking
13:26:25Membrillolinuxstb: shutup :P
13:26:29linuxstbmarkun: Does the Toshiba gigabeat firmware do anything that demonstrates a high number of frames-per-second displayed on the LCD?
13:26:52markunlinuxstb: It has some visualisations, but didn't really try them out.
13:27:09Membrillowell, does anyone know how i can edit a string in another form?
13:27:11markunI will take a look when I get home.
13:27:17linuxstbI'm wondering if there is a hardware reason for no video.
13:27:34markunI think it's just lazyness
13:27:41nathanh_mebrillo: strtok?
13:27:54linuxstbOr not wanting to deal with the tech support nightmare of people trying to convert their own videos.
13:28:20nathanh_withheld feature... its a marketting trick
13:28:30nathanh_release hardware without video, then 3 months later release with video and charge $100 more
13:28:41markunlinuxstb: is the screen update on the ipod faster than on the H300?
13:28:43linuxstbYep, that's more likely. Unless Rockbox gets there first....
13:28:58Membrillonathanh: whats strtok
13:28:59linuxstbmarkun: I've never seen a h300, but from the comments I've read, I think it is.
13:29:00nathanh_i see the xvid screen grabber patch
13:29:03nathanh_up on sourceforge
13:29:28^BeN^LinusN: check praivte
13:29:48markunWould be interesting to see how fast xvid decoding on the ipod is.
13:30:03markunAlthough it's not optimized at all
13:30:17linuxstbI tried the xvid decoder mirak posted, and it seemed to do about 2fps before crashing after 7 frames.
13:30:37linuxstbThat was with a large (640x352 I think) test file.
13:30:45markunwow, not too bad
13:31:03nathanh_on the h300?
13:31:06 Quit amiconn (Read error: 110 (Connection timed out))
13:31:07 Nick amiconn_ is now known as amiconn (
13:31:08nathanh_im impressed it showed any frames at all
13:31:29Slasherilinuxstb: that's great. I think the problem might be with the mallocs..
13:31:30markunlinuxstb: does it dump the frames to file?
13:31:37 Join Viper115 [0] (
13:31:42SlasheriSo the memory would overflow very soon
13:32:06linuxstbmarkun: I don't think so. I think that part was commented out. I just tried it very quickly last night though.
13:32:52linuxstbI could be tempted to take it forward, but I don't know if xvidcore will be the best starting point. ffmpeg has an xvid decoder as well, and on past experience, ffmpeg decoders are very efficient.
13:33:55nathanh_ffmpeg is truly awesome
13:35:18linuxstbnathanh_: The xvid decoder plugin doesn't display the frames at the moment - it just displays messages saying they have been decoded. But I think it can also write the frames to disk as bitmaps.
13:35:51 Quit perplexity (Read error: 104 (Connection reset by peer))
13:44:19 Join Zoric [0] (
13:44:51 Part Zoric
13:44:52 Quit Rob2222 (Read error: 104 (Connection reset by peer))
13:46:29***Saving seen data "./dancer.seen"
13:50:00nathanh_checkout the sdl patch, now on sourceforge
13:54:33 Join Verbal91 [0] (
13:59:56 Join perplexity [0] (
14:01:01 Quit Membrillo ()
14:02:34 Join webguest31 [0] (
14:02:41webguest31Cassandra: around?
14:07:01 Join Polo_o [0] (
14:10:37 Join muesli- [0] (n=muesli_t@
14:13:43 Part LinusN
14:17:54webguest31Cassandra: I'm wondering what the installer for the utility is doing? You say that the utility itself creates the config, then why not just distribute it as a standalone app?
14:21:30 Join edx [0] (
14:23:15 Quit perplexity ("*flibble*")
14:28:24 Join LinusN [0] (
14:29:38BgerLinusN is there any reason why the brightness patch is not in cvs ?
14:30:34LinusNBger: yes, i occasionally have to work and sleep
14:30:49LinusNand spend time with my family
14:31:07Bagderor rather the problem is that there aren't that many committers with colors
14:31:23Bgersorry, just asked is there something wrong with the patch itseld
14:31:46 Join mirak [0] (
14:32:04*Bger suggests himself :P
14:35:04linuxstbBagder: Have you got a few minutes to discuss copyrights?
14:35:14*amiconn is still looking for a H340
14:36:10linuxstbSpecifically, the ipodlinux people (quite rightly now I look at our code) are complaining that we've put our own names in the (C) message at the top of the source files.
14:36:26*slowcoder is here for comments if needed
14:37:00linuxstbI was just wondering what convention we should use for cases where we have source files either ripped directly from ipl, or based heavily on their code.
14:37:14linuxstbMy feeling is to just keep whatever (C) messages are in the original ipl files.
14:37:18Bagderimho, we should keep their name and their copyright mentions
14:38:14webguest31linuxstb: I noticed you have an ipod video.. if you need to use video-out, have a look at this page (to avoid having to buy an apple cable):
14:38:27Bagderonly if we use shorter pieces or just get inspiration from their code we should consider doing anything else
14:38:32 Join Kohlrabi [0] (
14:39:07webguest31Maybe add something below it "adapted to Rockbox by ..."
14:39:30Bagderwell of course lots of text could be added
14:39:38Bagderthe main point is the actual copyright line
14:39:50Bagderwe don't want anyone to feel hurt or miscredited
14:40:55linuxstbA good example is probably button.c - it contains button drivers for all targets, and the code preglow wrote is not a verbatim cut and paste from IPL.
14:41:11linuxstbBut of course, it is heavily based on ipl.
14:41:53webguest31Sounds like that just needs at most "based on information from ipodlinux" or something
14:43:06 Quit thegeek ("( :: NoNameScript 3.81 :: )")
14:46:19linuxstbslowcoder: I'll make the obvious changes now - for files which have been copied more or less verbatim from ipodlinux, I'll put Bernard Leach's name as the sole copyright holder.
14:47:37slowcoderianal, but the (c) stands even if it's derived work afaik
14:48:06linuxstbBut the only doubt I have about that is that he didn't actually write the file in Rockbox - I want to credit him, but not blame him for any mistakes I made :)
14:48:17linuxstbSo maybe a "Modifed for Rockbox by XXX" line as well.
14:49:01Bagdersounds fine to me
14:49:03slowcoderYou can have dual copyrights
14:49:21slowcoderAs long as you agree on a license (which you seem to have)
14:49:56Bagderits all GPL
14:50:10linuxstbI'm not worried about claiming credit for myself if that's what you're thinking.
14:50:54slowcoderI'm not
14:50:54*linuxstb has just read clause 2a) of the GPL
14:51:04slowcoderBut credit should be where credit is due
14:51:57linuxstb2a) says that modified versions of GPL code should state that the files have been changed and when. So I'll just add a line to that effect.
14:52:42slowcoderYea, but that's for the license to be valid.. It has nothing to do with copyrights
14:53:18linuxstbI know - it's a different issue. I'm happy with Bernard being the copyright holder.
14:56:00markunlinuxstb: yes, ffmpeg might be better. It was a bit difficult to find out which code was needed for MPEG-4 decoding only.
14:58:59 Join georgeblunt [0] (n=georgebl@
14:59:04mirakmarkun: maybe you can try to port it, maybe it's faster
14:59:52markunmirak: I'll leave for Italy in 2 days. Maybe when I get back (I only stay a week)
15:00:45mirakmarkun: what file was it ? Maybe I can give it a look
15:01:42mirakIt's unbelievable that it's the same guy that did ffmpeg and qemu
15:02:00mirakhe is french by the way
15:04:28linuxstbslowcoder, Bagder: Any comments on this suggested text:
15:04:59Bagderlooks fine
15:07:08 Join Rob2222 [0] (
15:07:23slowcoderFeedback from bern:
15:07:31slowcoder"sure, though if dave claims any copyright on it then he should add one"
15:08:20linuxstbIs that feedback on the lcd-ipodvideo.c file I just posted?
15:09:55markunmirak: There is a mpeg4_decoder function in h263.c
15:10:08slowcoderlinuxstb: Yep
15:10:29 Join _FireFly__ [0] (
15:11:04slowcoder"...either express or implied." <−− Isn't that "expressed" ?
15:11:54slowcoderMkay.. Just sounded a bit wierd to me..
15:12:10BagderI agree, but all licenses use that phrase that way
15:12:15Bagderand they always did
15:12:28linuxstbslowcoder: I'll go ahead and apply that text to all the IPL-derived files now.
15:13:47 Join sdfsdfsdf [0] (
15:13:53 Part sdfsdfsdf
15:14:20 Join sdfsdfsdf [0] (
15:14:22slowcoderlinuxstb: Much appreciated
15:14:41linuxstbNo problem. Apologies for the mis-appropriation.
15:15:14sdfsdfsdfcan someone help me?
15:15:36mirakwho made the video player plugin ?
15:15:59 Part sdfsdfsdf
15:16:10 Join sdfsdfsdf [0] (
15:16:17 Part sdfsdfsdf
15:16:39nathanh_how sweet does the colour bejeweled look!
15:16:50 Join tuxianer [0] (
15:17:05webguest31mirak: [IDC]Dragon I believe
15:17:09 Quit dwihno (Read error: 110 (Connection timed out))
15:17:20Bagdernathanh_: it is indeed very ncie
15:17:23Bgerhow much is sleep(1) ?
15:17:25miraknathanh_: I find it a bit slow
15:17:41BagderBger: 1 tick
15:17:54Bger1 tick = ? ms ?
15:18:02Jungti1234Can't I see video as Rockbox in H300?
15:18:09Bger10x :)
15:18:18Jungti1234Is it working?
15:18:23BgerJungti1234 no, it's not
15:18:27Bgerfirst steps
15:18:41Jungti1234Is it developing?
15:18:43tuxianeris there a german documentation?
15:18:50Bgerbut i suspect there will be a long way before having usable video
15:18:59LinusNtuxianer: no
15:19:23tuxianercan i suffle all my music files?
15:19:29nathanh_tuxi: yes
15:19:35webguest31Sure, just add it all to a playlist
15:19:46mirakBger: very long
15:20:13tuxianerit is possible without a playlist?
15:20:17nathanh_linusN: you have commit privs?
15:20:45nathanh_linusN: could you take a look at the sdl uisim patch up on sourceforge
15:20:50tuxianerand how?
15:21:13LinusNnathanh_: when i have time, i'm at work now
15:21:25webguest31tuxianer: No. Everything is a playlist.
15:21:40preglowmy, fluxbox is limiting
15:21:45markunmirak: will you give it a try with ffmpeg?
15:22:01tuxianeri have a h120 how can i create a playlist?
15:22:11webguest31preglow: what's it limiting? I quite liked it back when I used it
15:22:12mirakmarkun: don't know. I wich there is some more people working on that
15:22:53webguest31tuxianer: press and hold the joystick on the folder you want to play choose "playlist > insert"
15:22:57mirakmarkun: it would be maybe be better, since it's probably smaller. Like I said actually I have time, so if I am bored ...
15:23:01 Part Polo_o
15:23:56 Join Polo_o [0] (
15:24:03mirakmarkun: what's the difference between xvid and ffmpeg decoder ?
15:24:18preglowwebguest31: for one, i can't bloody resize windows
15:24:23 Quit _FireFly_ (Read error: 110 (Connection timed out))
15:25:14webguest31preglow: there are "handles" in the bottom corners
15:25:18webguest31or you can alt+middleclick anywhere iirc
15:25:19 Join thegeek [0] (
15:27:21tuxianerand where i can get plugins?
15:27:27webguest31okay, the last one was for evilwm
15:27:36webguest31tuxianer: menu > browse plugins
15:27:40markunmirak: I don't know the difference. Just 2 implementations of the same thing I guess.
15:28:18preglowlinuxstb: feel free to change whatever copyright stuff in any of the files i've added
15:28:21preglowi don't care much about it
15:30:13tuxianeri mean get new plugins...
15:31:02Bagdertuxianer: official rockbox plugins come with the rockbox packages, other plugins you get from the people that provide them
15:32:13markuntuxianer: or write some new plugins..
15:32:48linuxstbpreglow: I've just committed the (C) changes. Let me know if I've forgotten any of your code - I don't know how much you took from IPL, and how much you wrote yourself.
15:33:25preglowonly thing i can think of is pp5020.h
15:33:35preglowthose are my names :P
15:33:39linuxstbYes, I wasn't sure about that - it's just a list of names and numbers.
15:33:43preglowwell again
15:33:55preglowi don't give a shit what copyrights are mine, so just change them if in doubt
15:35:12tuxianerwhen i go to my musik folder and press the joystick−−>playlist insert it do but when i navigate to playlists in A/B and then show playlist...nothin hapens
15:38:26tuxianerwhat can i do?
15:38:34webguest31You should turn on recursive insertion
15:43:02 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
15:44:17tuxianerok it works fine thx
15:44:17 Quit _FireFly__ ("IceChat - Keeping PC's cool since 2000")
15:44:34^BeN^why i cant use variable from string type/?
15:44:35tuxianerand ist there a wma codec?
15:44:51Bagder^BeN^: C has no string variable type
15:44:55Bagdertuxianer: no
15:45:43tuxianeri have some music in wma
15:45:52webguest31A lot?
15:46:03Bgertuxianer : port ffmpeg's wma :P
15:46:20tuxianerone band
15:46:21webguest31I was going to say "transcode to lossless", but that is indeed a better option
15:46:30***Saving seen data "./dancer.seen"
15:47:00linuxstbtuxianer: I am sure wma will come to Rockbox in the future, so don't throw away those files. But as far as I know, no-one is working on it at the moment.
15:47:34CassandraPersonally I'm against it philosophically.
15:47:40linuxstbUnless they are DRM'd, in which case Rockbox will never play them.
15:47:50CassandraI don't believe we should be supporting crippled formats.
15:48:03webguest31Then mp3 is right out
15:48:09webguest31and shorten (no tagging)
15:48:23webguest31and of course alac
15:48:27CassandraI meant in the sense of DRM crippled.
15:48:28webguest31which was also re'd
15:48:44webguest31wma is no more crippled in that sense than aac
15:49:03LinusNwe were talking about drm
15:49:32LinusNdrm'ed wma will not be supported
15:49:53webguest31of course
15:50:01webguest31But Cassandra said it about just wma
15:50:14LinusNno she wasn't
15:50:26LinusNshe didn't, i mean
15:50:27webguest31No one mentioned drm before she said she was against it
15:50:40LinusNlinuxstb: did
15:51:36webguest31[15:46] <Cassandra> Personally I'm against it philosophically.
15:51:57webguest31That's why I figured she was against all forms of wma support
15:52:01linuxstbCassandra: What did you mean? :)
15:52:02preglowit does hint at it...
15:52:03webguest31Which seems a bit silly
15:52:07preglowbut anyway
15:52:10preglowi'm for support wma
15:52:13preglowbut nothing with drm,
15:52:21mirakmarkun: wma suck
15:52:21preglowi want to stay as far away of that as possible
15:52:23CassandraI believe I clarified here: <Cassandra> I meant in the sense of DRM crippled.
15:52:33LinusNah, i think she figured wma==drm
15:52:36webguest31Oh Christ
15:52:43BagderI'm not really for it, although I'm not against it
15:52:45webguest31I missed that in my rambling
15:53:06markunmirak: was I talking about wma?
15:53:10LinusNi fail to see the reason to use wma
15:53:20 Quit tuxianer ()
15:53:21mirakmarkun: wrong nick sorry
15:53:25preglowLinusN: because it's cd quality at 64kbs !!!
15:53:36*webguest31 dies laughing
15:53:40Bagderthe only reason to use WMA is for all those stolen music that you can re-encode to something else
15:53:46markunAnd if you transcode your mp3's to wma they will sound better!
15:53:59linuxstbThere are a lot of webcasts in wma - that is the only reason I would use it.
15:54:01webguest31Or because you were silly enough to use WMP to encode cds you borrowed/lost etc
15:54:06linuxstbSimilarly for RealAudio.
15:54:09webguest31Or stream-dumping
15:54:37webguest31though streams are usually in such bad quality that transcoding them to q5 oggs don't change much
15:55:15markunSaab rider told me most arab music is being shared in real audio format. Very low bitrates and ripped from audio casettes :)
15:55:34webguest31Oh boy, audio casettes. Haven't seen one of those for years
15:55:37Bagderdo we support casettes yet? ;-P
15:56:01 Join actionshrimp [0] (
15:56:40preglowif we support shorten, then we should support wma :P
15:56:48webguest31Cassandra: Now that you're around, I'm curious about why you're using an installer for the Rockbox Utility. I saw the discussion earlier, but didn't find an answer
15:56:52preglowi can see even less reason to support that
15:57:18webguest31One lossless format should be enough for everyone.
15:57:36preglowwell, at least lossless can be transcoded
15:57:59webguest31Yeah, that's what I meant. Keeping shorten files around is just lazy
15:58:02linuxstbI don't even think the person that ported the shorten codec actually uses it....
15:58:17Cassandrawebguest31, to make it as simple as possible.
15:58:26Bagderwell, I've done lots of rockbox stuff I never use...
15:58:31mirakwma is a plague like wmv
15:58:38CassandraSince there's been a request, I'm lilkely to provide a zip file as well.
15:58:44CassandraIt'll work either way.
15:58:45preglowwhich reminds me, i should finish the eq
15:58:46linuxstbBut some live music traders are obsessed by md5sums of Shorten files. Which is why they are still around.
15:58:47mirakthe format suck if you use linux
15:58:56webguest31It does mean you can not download it on a school/work computer and use it there
15:59:09preglowlinuxstb: why obsessed with md5sums applied to that particular format?
15:59:27webguest31They should use md5sums of the uncompressed version
15:59:33preglowthat's what they do
15:59:35linuxstbBecause it relates to particular recordings that were first traded when Shorten was the only available format.
15:59:39preglowthat is
15:59:40preglowwavpack does that
16:00:16linuxstbYes - nowadays, people use the md5sums of the uncompressed audio data, but there are still databases of the old style md5sums
16:00:16webguest31Cassandra: It does mean you can not download it on a school/work computer and use it there (unless you installed it at home first and copied the files over etc.)
16:00:38webguest31linuxstb: Lazy I tells ya!
16:01:10CassandraAs I said, I'm happy to provide installer and ZIP file.
16:01:45preglowCassandra: would be nice to integrate fwpatcher into that as well
16:02:00CassandraIt's on my todo list.
16:02:08 Join tucoz [0] (
16:02:13webguest31Cassandra: Sounds great, I was just being curious and making clear that there were really good reasons for preferring a zip
16:02:17CassandraToday's release is just a bare bones skeleton.
16:02:23preglowjust one rb tool for all platforms would be smashing
16:02:23linuxstbWhat about automatically downloading the iriver firmware from iriver's website? Is that a good idea?
16:02:25Bgerhm, better name for HAVE_UNREADABLE_SCREEN_WITHOUT_BACKLIGHT ??
16:02:45preglowHAVE_H300 ?
16:03:09CassandraBut it's already much more useful than the old Windows installer, so I decided to let people play.
16:03:14tucozsorry for the ot, Just curious, anyone here using for free music? I'm having a hard time finding anything of interest.
16:03:25Bgerpreglow do you want to tell me that u can read the ipods' screen without backlight ?
16:03:38preglowi can
16:03:45Bgerreally ?
16:03:54preglowi do it all the time
16:04:00BagderI believe the x5 one is readable too
16:04:09preglowit's more or less just as easy as on my h120
16:04:11Bagderbut I'm not sure
16:04:14Bgerthis is stupid ...
16:04:32linuxstbIt depends what you mean by readable. I wouldn't say either the iPod Color or iPod Video screens are readable without backlight.
16:05:03Bgerok, on h300's screen you can't see anything without backlight ...
16:05:04linuxstbBut you _can_ read them if you try hard enough.
16:05:04preglowwell, it all depends on environment
16:05:10preglowi think h1x0 is pretty unreadable as well
16:05:26preglowi can read nano easily in a normally lit room
16:06:52Bgerso, we don't need another define ?
16:06:57 Quit muesli- ("ich will Kühe!!!")
16:07:42*Cassandra wonders if it's time we started doing iPod daily builds?
16:07:56preglowi'm against it for now
16:08:00pregloweven basic operation is flakey
16:08:17CassandraFair enough.
16:08:27tucozanyway, now that there are a few people with cvs-access here. Do you think the LaTeX-proposal is ready for commit? Or does it need more work?
16:08:38Jungti1234Bye all
16:08:40Jungti1234good night
16:08:41preglowbut i'm not firmly against it either, if you feel it's time, go ahead
16:08:51preglowtucoz: yeah, i think so
16:08:52 Quit Jungti1234 ("bye")
16:09:37tucozok, what do the others say. I believe a new module should be added in that case. (manual perhaps).
16:09:48BagderI say we add dailies if it adds a purpose to anyone
16:10:21preglowyou've still more or less got to be a dev to run it
16:10:22preglowbut hey
16:10:31preglowinstalling the bootloader is pretty hard
16:10:41Bagderperpaps we could have the new latex docs as a subdir of docs/
16:10:55tucozBagder, good idea.
16:11:05preglowi don't think so
16:11:12tucozas long as it's not included in the build
16:11:14preglownormal developers wont be interested in this
16:12:11tucozhmm, the problem is that the docs will contain a substantial amount of images which will bloat a dev-checkout.
16:12:18preglowyes, that too
16:12:30preglowi think the manual is an important enough part to merit its own cvs root
16:12:36preglowor something
16:12:37Bagderok, so we do a new module
16:12:40preglowmodule, yes
16:12:49linuxstbGets my vote as well.
16:13:46linuxstbIs it possible to build different manuals for different targets yet?
16:14:24tucozlinuxstb, it is possible. I included info in the README
16:15:01linuxstbtucoz: Are you happy that the current file/directory structure will cope with target-specific manuals?
16:15:02mirakthat's steve coleman, try some stuffs, I love that but that's maybe to esoteric
16:15:17webguest31WARGH BLINK!
16:15:28tucozmirak, ok, I'll check it out. Thanks
16:16:18linuxstbIMO, the multiple targets is something we need to sort out before committing to CVS - so we aren't restricting our choices in any way.
16:16:25lostlogic2 strange things: The assembly optimizations that I've been working on are not faster than baseline alone. My combined optimization patch profiles 3-4% faster than baseline however.
16:16:36tucozlinuxstb. Well, I think so. But, I would like some thoughts on that as well.
16:16:37preglowlinuxstb: hmm? how?
16:16:59miraklostlogic: what's baseline ? plain c ?
16:17:27tucozlinuxstb, you can add \opt for target-specific includes
16:17:29lostlogicthe second is that I know why iRiver doesn't use 124mhz −− battery life. 124MHz draws heavily from the battery in bursts which can artificially lower the voltage and cause the unit to die, especially on aging batteries, or in cold weather
16:17:58lostlogicmirak: baseline for the purpose of this is Tremor as it currently is in CVS plus inlining oggpack_look and oggpack_adv
16:18:08preglowlostlogic: well, no surprises there
16:18:16preglowlostlogic: if you keep up the work we wont have to boost :)
16:18:26lostlogicpreglow: in my dreams.
16:18:43*Bger compiles and hopes everything to be OK
16:18:46preglowi think a major speedup could be had in rewriting the floor1 reconstruct
16:18:55 Quit georgeblunt (" HydraIRC -> <-")
16:18:57preglowto use asm, that s
16:19:24miraklostlogic: what brings inlining ?
16:19:25tucozlinuxstb: \opt{iriver}{\input{iriver.tex}} in rockbox.tex will check if iriver is defined, and include that file if that is the case
16:19:38lostlogicthe inline keyword in...
16:19:46miraklostlogic: it means the function is coded each time it's called ?
16:19:59linuxstbtucoz: How do you define "iriver" when running the latex command?
16:20:01lostlogicmirak: yeah
16:20:07tucozyou get a prompt
16:20:22miraklostlogic: what's the interest ? Aren't branching fast already ?
16:20:25tucozcould probably pipe it or stdin it or something :)
16:20:40linuxstbCan you avoid prompts by defining it in a .tex file?
16:20:57lostlogicpreglow: floor1_inverse1 and floor1_inverse2 combined only use 95k out of 2million ticks in the decode process...
16:21:10mirakhow much gain using iRam can provide in general for a codec ?
16:21:21amiconnmirak: Branching is relatively fast, calling subroutines is quite a bit slower
16:21:24lostlogicmirak: depends how much _data_ there is to support the codec
16:21:24linuxstbtucoz: So is there one prompt for every \opt command?
16:21:29tucozlinuxstb, I am not sure, I haven't really investigated that much on that part
16:21:35tucozlinuxstb, yes
16:21:58linuxstbSo it seem to be hard to automate.
16:22:15tucozlinuxstb, but preglow talked about something with letting the preprocessor handle that
16:22:18miraklostlogic: X2 X5 ?
16:22:18preglowlostlogic: then what part of tremor scales so badly with bitrate?
16:22:22mirakhow much ?
16:22:24mirakin genereal
16:22:32preglowmp3 handles bitrate increases far better
16:23:01linuxstbtucoz: I think the alternative is to use the C pre-processor instead of \opt
16:23:10preglowwhy is that hard to automate?
16:23:13lostlogicpreglow: I think it's the oggpack stuff still that's the problem −− I'll profile some different songs with and without bitrate spikes and see which parts scale worse when I have a chance.
16:23:18preglowwe already do automation like this for the daily builds
16:23:30linuxstbBut the questions in the configure script don't change.
16:23:37lostlogicmirak: near impossible to estimate.
16:23:44linuxstbIn the manual, the opts will change everytime someone commits.
16:24:24tucozhmm, why is that?
16:24:56linuxstbMy understanding is that every \opt in the file prompts the user to answer a question. The answer to the question may not always be "iriver" - it may be something like "fm radio" or "recording"
16:26:07amiconnWouldn't it be possible to just make one include depend on \opt ? Then the prompts will only change if we add a new target
16:26:09tucozI see,
16:26:18Bgerhm :(
16:26:35Bgerit's not so quick ...
16:26:39amiconnThe included file could define further things similar to the config-*.h files do for the sources
16:27:01tucozlinuxstb, latex "\def\UseOption{opta,optb}\input{file}"
16:27:02preglowevery opts prompts?
16:27:07preglowthat's bad
16:27:19tucozlinuxstb, it is possible to tell latex on commandline what options to use
16:27:35preglowthere's got to be some other mechanism for this :/
16:27:37preglowdinner time
16:27:38tucozif I understand that correctly
16:28:11linuxstbSomeone needs to actually try and implement this....
16:28:24markunCassandra: will your installer use 7zip?
16:28:41tucozpreglow, I don't think the prompts are needed at all.
16:30:40linuxstbtucoz: Yes, that seems to do the job perfectly.
16:30:51tucozok, good
16:31:38tucozas long as the file structure looks good, I think things like this should be patched later on
16:32:54Cassandramarkun, afailk it's not supported by wxWidgets yet.
16:33:09tucozbut the optional.sty suggests having a different main-file for each document-type
16:34:23 Quit Zagor ("Client exiting")
16:34:51linuxstbtucoz: That's only when the documentclass etc are different. In Rockbox's case, it will be the same.
16:35:03tucozOh, I see
16:37:21linuxstbtucoz: Do you have a URL to your latest version?
16:37:32linuxstbIs it in your forum thread?
16:37:50tucozI haven't changed much since the last post
16:38:17 Quit Viper115 ()
16:40:09webguest31"(edit: I had by mistake pasted a link to the european curling championships, so here is the not so short guide )" heh, I love those kinds of mixups
16:40:56Bgerok, "turn off LCD on h300 when backlight is turned off" submitted to patch tracker
16:42:59tucozwebguest31, I was lucky it wasn't my creditcard number and expiration date
16:43:41webguest31Those mixups are indeed less amusing
16:44:41mirakBger: nice :D
16:45:56Bgermake[1]: *** [/home/Bager/Rockbox/rockbox-devel/build/apps/rockbox.elf] Error 1
16:45:56Bgermake: *** [all] Error 2
16:46:27Bgerhome/Bager/Rockbox/rockbox-devel/build/apps/settings.o: In function `settings_:
16:46:39 Join leftright [0] (
16:46:50Bgerhome/Bager/Rockbox/rockbox-devel/apps/settings.c:900: undefined reference to `'
16:47:18LinusNBger: looks like bad macro
16:47:59leftrightslasheri. the transition between tracks when skipping tracks, (xfade is off) is not clean, it has artifacts
16:48:04BgerBuild LDS file
16:48:04BgerLD rockbox.elf
16:48:13Bgerthis error is from LD ...
16:48:17Bgeror i'm wrong ?
16:48:38mirakh263 is mpeg4 or is that different ?
16:49:36BgerLinusN any ideas?...
16:49:45Bgerplease ...
16:50:10linuxstbtucoz: I've looked at your manual a little, and I'm thinking we should maybe go with subdirectories for each chapter - to store any images, and also to allow us to move target specific text into separate files.
16:50:20LinusNwhat is on line 900 in settings.c?
16:50:55Bger#define RTC_BLOCK_SIZE 44
16:51:04Bgerthis is on line 900 in settings.c
16:51:16Bgerops, line 90
16:51:44mirakmarkun: I am looking the code. There is h263dec.c seems dedicated to decoding
16:51:51tucozlinuxstb, ok. That is fine by me. Anyway, by defining \opt{archos}{\input{archos.tex}}, and \opt{iriver}{\input{iriver.tex}} in rockbox.tex
16:52:22linuxstbtucoz: Yes, that's the idea.
16:52:25tucozit is possible to generate different docs with latex "\def\UseOption{archos}\input{rockbox.tex}" and latex "\def\UseOption{iriver}\input{rockbox.tex}"
16:52:48tucozSo I guess the optinal package do what we want
16:53:10tucozwe could define sections as \opt{iriver,archos,ipod} for instance
16:53:58LinusNBger: so, is there a "brightness" in the global_settings struct?
16:54:05LinusNBger: maybe a "make clean" helps?
16:54:16Bgercd build; rm -rf *
16:54:20Bgerand after that
16:55:43Bgerconfigure ...
16:55:52Bgerand then
16:56:13 Part leftright
16:56:42Bgeri'll check for brightness in glob.settigns
16:56:56amiconnBger: *lcd*_set_brightess() ? I would expect backlight_set_brightness() ...
16:57:14tucozlinuxstb, what do you suggest we call the subdirs? same as the main_files?
16:57:55amiconn(or is that a brightness function of the lcd that doesn't have to do with backlight?)
16:58:12Bgeramiconn u're right ... :)
16:58:22Bgeri didn't named it
16:58:42linuxstbtucoz: I tried adding \opt{recorder,ondio}{\verb+ajbrec.ajz+} but the verbatim formatting goes wrong. Have I done something stupid? (I've forgotten all the LaTeX I once knew)
16:59:10tucozwhy do you do \verb?
16:59:16linuxstbtucoz: Yes, just call the chapter directories the same as the main .tex file for that chapter.
16:59:19 Join XavierGr [0] (
16:59:23linuxstbtucoz: Because that's what was there before.
16:59:43tucozah, I see
16:59:51tucozhmm, let me check
17:00:00linuxstbIt's in the getting_started.tex file
17:00:18Bgerstrange ...
17:00:45tucozlinuxstb, % As usual, "\verb" commands and verbatim environments cannot be used
17:00:45tucoz% in the argument to "\opt". For this purpose there is a variant form
17:00:45tucoz% of "\opt" called "\optv" (optional verbatim) which may have a limited
17:00:45DBUGEnqueued KICK tucoz
17:00:45tucoz% class of verbatim material in the argument. It can do so by leaving
17:00:45tucoz% the braces around the argument, which may have undesired side effects.
17:00:46***Alert Mode level 1
17:00:46tucoz% For an "\optv" argument to be successfully ignored, the verbatim material
17:00:48tucoz% must have balanced braces etc.
17:00:50tucozsorry for the flood
17:01:18solexxlinuxstb: do you really need the {} around \verb?
17:01:30solexxi think the +es are the delimiters
17:01:33tucozyou need to use \optv I guess
17:02:14solexxah, forget it
17:02:20Bgerok, i don't get it :(
17:02:52Bgerapps/settings.h :
17:02:54Bger#ifdef HAVE_LCD_BRIGHTNESS
17:02:54Bger int brightness; /* iriver h300: backlight PWM value: 2..15
17:02:54Bger (0 and 1 are black) */
17:02:54DBUGEnqueued KICK Bger
17:03:21LinusNshould be HAVE_BACKLIGHT_BRIGHTNESS imho
17:03:21Bger#define HAVE_LCD_BRIGHTNESS
17:03:32Bger. /* define this if you have an adjustable LCD brightness */
17:03:42Bgeryes, u're probably right
17:03:51LinusNBger: "."???
17:04:00Bgeri wrote "."
17:04:09Bgerbecause the stupid bx ...
17:04:16Bgertreats it as command
17:04:51webguest31Try putting a / in front.. like "/ /* define etc."
17:05:09linuxstbtucoz: Yes, optv does the job.
17:06:58Bgerany suggestions ?
17:06:59tucozlinuxstb, to change to using subdirs is as simple as creating the dirs, and adding the relative path to the files in rockbox.tex
17:07:14tucoz...and moving the files of course :)
17:07:39linuxstbI've just tried \opt with \newcommand - so you can do something like:
17:07:42tucozlike \input{getting_started/getting_started.tex}
17:08:36tucozcool, that is great
17:08:40 Join DangerousDan [0] (
17:08:49linuxstbThat could be useful for lots of things - like button names.
17:08:58tucozYes, that is splendid
17:09:40tucozWhat do you think? Should I do any more on this at the moment?
17:09:53CassandraOh, btw, what do people think of the wizard picture on the new installer?
17:10:29linuxstbSo we can follow the Rockbox convention of having a "config.tex" file which defines general commands and also includes the appropriate "config-target.tex" file which defines commands specific to the targets.
17:10:29preglowCassandra: you fixed the blind user issue?
17:10:48***Alert Mode OFF
17:10:50CassandraNot yet. I imagine if I stick some tooltips on the button that'll do it though.
17:11:02preglowi'd guess so
17:11:02tucozlinuxstb, that sounds perfect
17:12:05preglowCassandra: btw, what do you use for downloading, file handling and such?
17:12:51tucozlinuxstb, would you like to fix the remaining stuff?
17:12:53preglowi expect any other answer than 'curl' is going to get you kickbanned :>
17:13:12linuxstbtucoz: I would like to, but I have 101 things to do before Christmas, and then I'm going away for a couple of weeks.
17:13:22tucozsame here
17:13:31linuxstbBut I might be able to give it a go tonight.
17:14:17linuxstbI only said "might" :)
17:14:26tucozI got that. hehe
17:15:28tucozBut I think you know better how to organize the files than me.
17:19:16linuxstbWhere did your LaTeX.Rules file come from?
17:20:02preglowi've asked if it's redistributable, and he says it is
17:20:29Cassandrapreglow, wxWidgets functions.
17:20:37Bgerlinus, amiconn, any hints ?
17:20:41preglowCassandra: cool, i didn't think it did that much
17:21:20CassandraIt's pretty cool, although a bit flaky sometimes.
17:21:22linuxstbpreglow: It does everything - bitmap loading, http, html, audio, threads, ....
17:21:34preglowdamn, nice
17:21:39preglowi thought it only did windowing
17:22:06CassandraString handling too.
17:22:36CassandraI don't think there's a single standard C++ library call in the program.
17:22:51tucozlinuxstb, the LaTeX.rules come from a link you can find in the README
17:22:55 Join rayo [0] (n=rayo@
17:23:48linuxstbtucoz: Which README? There isn't one in
17:23:57tucozhmm, weird
17:24:03amiconnmarkun: What I wanted to ask - playlists are always utf-8 now?
17:24:28CtcpIgnored 2 channel CTCP requests in 2 minutes and 9 seconds at the last flood
17:24:28*Cassandra is particularly proud of the wizard photo though. I took it myself.
17:24:29markunamiconn: They should be, but I didn't check
17:24:40rayodoes a wps file work with 16-bit images? (for the iriver h3x0 series) because i'm interested to run my iriver h3x0 with rockbox
17:24:55linuxstbrayo: No, not yet.
17:25:08Bgerguys, any hints ?
17:25:27amiconnThere are some problems reported in the forum related to playlists... Afair, nobody mentions how these playlists are created
17:25:50rayook then i must wait for it, but a wps file for the h1x0 work with the h3x0?
17:26:28CassandraYes. The H100 wpses work on the H300, although they may look a little odd.
17:26:50amiconnIiuc if someone saves a playlist e.g. on german windows it will be Win-1252 encoded (extended ISO8859-1). This will cause problems with rockbox now...
17:26:51LinusNBger: not really
17:27:24amiconnBger: That's with the brightness patch?
17:27:44Bgeri'll correct it tomorrow
17:27:45 Join PaulJ [0] (
17:28:07markunamiconn: Should we use the default_codepage setting for playlists?
17:28:21CassandraYou know, if a WPS dimension is smaller than the screen dimension, we ought to center it on the screen if possible.
17:28:21markunAnd for the tag db too?
17:28:31Bgeri mean, i'll correct the patch to substitute lcd with brightness
17:28:43CassandraAlthough I suppose that still won't fix scrolling.
17:28:54Bgeror backlight
17:29:04rayobut the lcd driver for the h3x0 work? (in 16-bit mode)
17:29:16LinusNrayo: yes
17:29:59tucozlinuxstb, download this
17:30:29amiconnmarkun: That's the big question - (1) it would require that rockbox saves playlists using the same code page as it uses for loading them. (2) It would make 'multi-language-playlists' impossible
17:31:15amiconnmarkun: For the tagdb, I think the tools should be changed to always write utf-8 into the db.
17:31:15 Join DrMoos [0] (
17:31:41 Quit Moos (Read error: 104 (Connection reset by peer))
17:32:22amiconnCassandra: The low level work has to be done first.
17:32:41markunamiconn: if Slasheri's tag thing replaces the current db then we don't need to worry about it.
17:32:54amiconnmarkun: Maybe the tagdb will be replaced by Slasheri's tag cache soon...
17:32:59amiconnhehe :)
17:34:00tucozlinuxstb, there is a README in btw
17:34:04preglowi'm looking forward to seeing it, at least
17:34:04amiconnLinusN: Afaik, our CPU should be strong enough to run old PalmOS. Old palms had a dragonball CPU which is also m68k based - at 33 MHz
17:34:15amiconnThe problem would be the LCD resolution
17:35:34linuxstbtucoz: I've found it now - I've been a victim of firefox renaming the file to "". I had an old
17:36:48rayomhh in lcd-h300.c ist the function lcd_blit empty, so the displaydriver just runs but won't display anything, correct?
17:37:19amiconnlcd_blit() isn't needed for 8bit+ displays
17:37:45amiconnThe H300 display driver is working 100%, it's just still a bit slow
17:38:01LinusNamiconn: what about touchscreen and lcd resolution?
17:38:23amiconn[17:34:15] <amiconn> The problem would be the LCD resolution <=
17:38:33rayoist it done directly with the framebuffer?
17:38:36LinusNamiconn: i'm blind
17:39:22preglowthen rockbox is perfect for you!
17:41:03amiconnlcd_blit() is for low depth displays, allowing access to the lcd from an external buffer, to do some tricks like showing more grayscales than natively possible
17:41:09LinusNpreglow: :-)
17:41:32amiconnIt has nothing to do with the standard drawing
17:41:45BgerXavierGr ?
17:42:46tucozlinuxstb, please let me know if there is anything you want to be changed with the docs.
17:43:03 Join DJDD_ [0] (
17:46:32***Saving seen data "./dancer.seen"
17:47:16 Quit DJDD_ ("Trillian (")
17:47:49tucozLinusN, are you aware of this forum?
17:48:01SlasheriHmm, tag cache engine should be almost ready now (at least the core functionality is). Now i still have to implement some settings, and develop the browser
17:49:02Bgernite all
17:50:46LinusNtucoz: now i am :-)
17:51:16LinusNi liked the last post in that thread :-)
17:52:32amiconnLinusN: Btw, old PlamOS has fixed resolution 160x160 pixel. monochrome 2bit or 4bit
17:52:52amiconn8MB RAM was standard...
17:53:01tucozLinusN, post #75 was quite funny as well
17:54:21 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
17:54:30webguest31Whoa, there was a post from cowon
17:56:43preglowwhat does it say?
17:57:41LinusNtucoz: lol
17:57:58LinusNamiconn: interesting
17:58:23tucozpreglow, it says the cowon-engineers are interested in the porting. They had not heard of rockbox.
17:58:51 Join _FireFly_ [0] (
17:58:58mirakI have a good one. I started the H300 in the metro with my ear plug, and I was like what the hell is that, I am receiving radio or weird frequences on radio ? I then realised that in fact it always start in record mode, so it always record from internal microphone. I was earing amplified what the other were saying
17:59:15preglowmirak: hahaha, i've tried using it for that too
17:59:17preglowmirak: it actually works
17:59:24mirakYou can spy
17:59:35mirakand I think the hardware amplify voice
17:59:42preglowand amplifies everything
17:59:44mirakyou can hear better with it
17:59:55mirakyes but the mic focuses on voice or something
17:59:56preglowif they knew of a way to amplify just voice, they would be rich
17:59:59preglowricher than they are
18:00:05mirakI think it does
18:00:21preglowinternal mic?
18:00:30mirakI have read that in the manual
18:00:30preglowit's not unlikely the mic has a freq response suited for voice, no
18:00:33preglownot at all
18:00:43preglowfor dictaphone use, it makes sense
18:00:47mirakthe ship that convert sound maybe
18:01:08preglowtucoz: cool, it would rock actually having the manufacturer on the team for once
18:01:23mirakwell I have read in the manual something super special about voice ;) preglow
18:01:30LinusNpreglow: i've been in contact with cowon
18:01:37preglowLinusN: what do they say?
18:01:40LinusNno go
18:01:41preglowgl hf?
18:01:46preglowno go on what?
18:02:05LinusNthey aren't interested in helping
18:02:14LinusNhardly surprising
18:02:32preglowwell, at least most of the components have specs
18:02:36preglowgod, how i hate apple
18:02:38preglowi will kill them
18:02:38mirakLinusN: you think you could sell it ?
18:02:39tucozI've tried to use the iriver to listen to the baby's heartbeat in my gf's tummy. Even though I can hear if I put my ear next to it, I cannot hear anything but the other stomach sounds. Maybe in a few weeks.
18:02:52LinusNmirak: sell what?
18:03:05mirakLinusN: (rockbox ...)
18:03:21 Quit Verbal91 (Read error: 104 (Connection reset by peer))
18:03:26preglowtucoz: it works better with just using your head in that case, the sound will propagate better through flesh than through iriver
18:03:31linuxstbpreglow: How are apple different to any other DAP maker?
18:03:41LinusNmirak: i can't sell rockbox, it isn't mine
18:03:52preglowlinuxstb: s/apple/portalplayer/
18:03:58mirakLinusN: it's wom ?
18:04:05linuxstbpreglow: OK, I agree then.
18:04:16LinusNmirak: it belongs to all contributors
18:04:22tucozpreglow, ok. That is probably the case. But, I can hear my own heartbeat if I put it on my chest.
18:04:45mirakLinusN: yep, but you get the money and resdistribute it
18:04:48preglowsure, and you'd probably hear it better if you could bend your own head to your chest ;)
18:05:05LinusNtucoz: ultrasound.rock :-)
18:05:10tucozhehe, probably
18:05:27*tucoz starts coding
18:05:52LinusNi wonder why austriancoder gave up
18:05:57preglowwho knows
18:06:05mirakLinusN: do you if cowon use their own compiler or if they have gcc ?
18:06:11linuxstbReal Life?
18:06:20LinusNmirak: i guess they use gcc
18:06:35preglowi don't there are too many other good 68k compilers
18:06:41webguest31somone really should have words with the m68k maintainer of gcc
18:06:42preglowperhaps metrowerks have got something
18:06:43LinusNlinuxstb: maybe, but he sure as hell didn't tell us
18:06:48tucozwhat about the metrowerks?
18:06:56 Join ratpack91 [0] (
18:06:58linuxstbSurely Motorola must have one?
18:06:59preglowwebguest31: well, is there one?
18:07:07webguest31no idea
18:07:15preglowwebguest31: sure as hell doesn't look like it
18:07:22preglowthe m68k backend is old and crummy
18:07:41webguest31maybe he considered it "done" and left
18:07:53tucozor she
18:07:57preglowat least it happened long ago
18:08:21preglowcan't remember ever having seen a name that looked vaguely female on the gcc lists
18:08:32preglowthey probably have more sense than men
18:08:54tucozok, he then
18:09:10webguest31linuxstb: sick!
18:09:11preglowthat's pretty recent as well
18:09:15webguest31as in, attack.
18:09:30webguest31hrm, that's not the right use of the word. Nevermind.
18:10:13preglowsomeone mail him and tell gcc for 68k sucks, quick!
18:10:37webguest31I'm wondering if he isn't aware of it, myself
18:10:38LinusNthat's constructive criticism for sure :-)
18:11:01webguest31Well, lostlogic has found some specific examples iirc
18:11:02linuxstbHe's got a email address - so maybe that means he's getting paid.
18:11:04preglowbut no, many of the issues are probably cf related
18:11:15preglowyes he hs
18:11:24preglowlike no addq and subq use
18:11:25LinusNi wasn't aware that it sucked that much
18:11:57amiconnpreglow: Uh? I think I saw addq and subq in the disassemblies...
18:12:18preglowi've seen them too
18:12:25preglowbut lostlogic has seen some pretty stupid spots where they weren't use
18:12:51preglowit might be more of a register allocator issue than enything else, though
18:12:55preglowi haven't seen the code in question
18:13:10preglowoh, holy god
18:13:23preglowthe freebsd box i use for rockbox coding has gcc 2.95!
18:13:27preglowso much for using it for other coding
18:13:45amiconnIf you compare gcc asm output to the C source, the quality varies a lot. Gcc sometimes generates brilliant code, other times it generates the most stupid s**t
18:14:02preglowi'm really curious as to how gcc4 code is
18:14:06preglowthey've changed a ton of internals
18:14:06amiconnI observed the exact same behaviour for SH1
18:14:11preglowbut only for some platforms, i think
18:14:31preglowm68k code is generally worse, they say
18:16:00 Nick DrMoos is now known as Moos (
18:16:03tucozHmm. there are lot's of nice stuff at the freescale site:
18:16:16tucozToo bad they have a restrictive license
18:17:01preglowthe dsp stuff at least didn't properly use the emac unit
18:17:08preglowas a matter of fact, it didn't is it at all
18:17:12preglowjust the older mac unt
18:18:10tucozok, what about the PORTASM68K? It says that it's free, and able to produce optimized coldfire code from 680x0 and CPU32 assembly code.
18:18:41preglowwhat do we need that for?
18:18:51 Quit YouCeyE ("Leaving")
18:19:07preglowbtw, what happened to the mp3 encoder?
18:19:18tucozporting 680x0 code
18:19:30preglowtucoz: well, i don't know of any 68k code we want
18:19:46preglowapart from pegase, perhaps, but i don't know how many people will be interested in layer1/2 encoding
18:20:06tucozpreglow, not me either at the moment. But, if one found something written in that it could be used.
18:20:09LinusNthe shine encoder is the hottest candidate
18:20:23LinusNbut it needs a *lot* of work
18:20:32preglowLinusN: i know, and someone ported a optimised version to coldfire, why isn't it in cvs?
18:20:41tucozIsn't that in cvs yet? I thought it worked quite fine
18:20:48tucozthat is it *works*
18:20:49LinusNpreglow: where is it?
18:21:06preglowgimme a sec, i think rasher has it
18:21:29preglowhe seems to have removed his build page...
18:21:34 Join memmem [0] (
18:21:34LinusNwe would need to extend the codec api
18:21:45preglowwell, yes
18:21:49preglowwe already have a wavpack encoder
18:22:02tucozLinusN, is it this?
18:22:08preglowthere's no need to wait for the mp3 encoder for that
18:22:40preglowthat's it
18:22:44preglowone file of pure bliss
18:23:06 Join muesli__ [0] (
18:23:13tucozthe link in the comment?
18:23:58tucozWell, that should serve as a good start anyway it seems.
18:24:04memmemHi LinusN, are you working on the X5 port now? I think chances are small for turning the device into a brick as the boot loader seems to be in the lower 64 KB of the firmware which are not contained in the firmware file, so it probably will work even if the rest of the firmware is broken.
18:25:12muesli__could any english speaking native give me some help with some sentences?
18:25:58LinusNmemmem: interesting
18:26:08LinusNmemmem: what does the x5 bootloader do?
18:27:07memmemLinusN: Don't know, I haven't seen it (no BDM). The flashing code in the firmware file seems to be unused.
18:27:28memmemI guess it looks for the firmware file and flashes it when found.
18:28:53preglowdoes this hellmann hang on irc?
18:29:07PaulJmuesli__: i read the post in the german forum about the clicking an popping sounds. i think the reason is tha the files have a samplingrate of 32kHz
18:29:22tucozpreglow, did you see the viewer in the comment?
18:29:39muesli__PaulJ ok. but it should be fixed, isnt it?
18:29:40preglowtucoz: what about it?
18:29:42preglowmuesli__: no
18:29:55muesli__why not?
18:30:03preglowbecause i haven't got around to it?
18:30:15memmemLinusN: Could you please give me the lower 64 KB when you manage to connect the BDM? (The code of the firmware file gets loaded to 0x10000.)
18:30:18tucozpreglow, maybe it works better than the patch
18:30:36muesli__it wasnt a complain ;)
18:31:00preglowbesides, i'd more like to throw away the entire resampler and make a new one
18:31:08preglowbut eq first
18:31:15preglowperhaps i should make the eq work now
18:31:26LinusNmemmem: sure
18:31:36preglowwill anyone shoot me if i add eq.c ? i don't want to make dsp.c a playback.c like monster
18:32:22LinusNpreglow: i won't kill you
18:32:25memmemLinusN: Fine.
18:37:05_FireFly_can unions be used in rockbox ??
18:37:34preglowsure, why not
18:37:53_FireFly_ah ok i thinking about a new wps-handling-code
18:38:56_FireFly_which will be token-based (if it doesn't use to much memory)
18:41:19amiconnpreglow: We should either make stereo width work, or remove the options for swcodec.
18:42:02preglowwell, i can implement it as i did playback speed
18:42:21preglownot very nice, but it works
18:43:48lostlogicxnathanh_: how are your display lists? :)
18:44:02_FireFly_display lists ??
18:44:45 Part tucoz ("Leaving")
18:45:09lostlogicx_FireFly_: new WPS handling allows for giving parts of the screen more attributes (ie color) and eventually even alpha blending between different Z levels.
18:46:16_FireFly_i hope he has in mind the low-memory-device when designing these display-lists
18:47:20lostlogicxThey shouldn't use any more memory than current WPSs for a WPS that works on the existing devices, but for color devices, they will allow specification of colors and stuff. Actually if memory really is a concern, they could use less memory by completely preparsing the WPS into sets of bitfields rather than chars or something creative like that.
18:47:30 Part Polo_o
18:47:56preglowdecode it into a huffman codec bitstream!
18:48:25 Join RotAtoR [0] (
18:48:50_FireFly_yeah that would decrease the memory-usage for the code-itself but you need extra-memory for the dict
18:49:30lostlogicxdictionary of who?
18:49:53_FireFly_which defines which huffman-code represents which tag(in the case of the wps)
18:50:35_FireFly_the defalte algo for example uses an huffman-code to compress data
18:51:01 Join linuxstb_ [0] (
18:51:04preglowi was joking
18:51:12amiconnForget huffman. It would be sloo-oow
18:51:19 Quit linuxstb (Read error: 110 (Connection timed out))
18:51:53preglow_FireFly_: deflate does not use huffman primarily, afaik
18:51:57preglowcompress however, does
18:52:13_FireFly_the compression it self is a kind of lz77
18:52:25_FireFly_and a combination with huffman
18:54:33 Join Lear [0] (
18:54:50_FireFly_or was it zlib which uses huffman
18:55:14 Join Mmmm [0] (
18:56:39_FireFly_deflate: This specification defines a lossless compressed data format that
18:56:40_FireFly_ compresses data using a combination of the LZ77 algorithm and Huffman
18:56:40_FireFly_ coding
18:57:06preglowokokok, anyone, what TERM should i set to make freebsd understand i can take colours without crying
18:57:11preglowxterm-color does not work
18:57:26preglow_FireFly_: zlib == deflate
18:57:29_FireFly_term-emulation or
18:57:36_FireFly_preglow: not at all
18:57:46_FireFly_zlib uses currently the deflate algo to compress data
18:57:47preglow_FireFly_: it sure as hell can decompress deflate data, at least
18:57:49 Join perplexity [0] (
18:58:09preglowi use zlib for decompressing zip files in my ansi viewer
18:58:13_FireFly_but it also defines an kind of fileformat
18:59:00_FireFly_i think i would be hard to port it to rockbox (if it will be ever needed)
19:08:07 Join ep0ch| [0] (n=ep0ch|@
19:08:18ep0ch|preglow: which shell are you using?
19:10:00 Join Polo_o [0] (
19:10:34 Part Polo_o
19:10:39ep0ch|seems to work for me
19:10:44preglownot for me
19:10:55preglowwhen i enable syntax highlighting in vim, i just get underlined words
19:11:00preglowand no colours
19:11:04preglowperhaps it's a vim problem
19:11:33_FireFly_in my terminal-emulation(urxvt) the TERM is set to xterm
19:12:25_FireFly_preglow is COLORTERM set ??
19:12:33preglow_FireFly_: no
19:12:51_FireFly_maybe this helps you:
19:13:06XavierGrI read somewhere that Ubuntu uses .db packages, does that mean that .rpm packages will not work for Ubuntu?
19:13:19preglowubuntu is a debian based distro
19:13:26preglowit can use rpm, but it isn't the primary format
19:13:30preglowubuntu uses .deb
19:13:34ep0ch|preglow: do you have CLICOLOR set?
19:13:43XavierGrso it will not accept other packages right?
19:14:13preglowep0ch|: no
19:14:18preglowexport CLICOLOR=1 ?
19:14:24webguest31XavierGr: Using rpm is a baaaad idea.
19:14:33 Quit linuxstb_ (Read error: 110 (Connection timed out))
19:14:38webguest31(on Debian-based distros such as ubuntu)
19:14:42ep0ch|well mine is just CLICOLOR=, but i think it should be CLICOLOR=YES
19:14:57_FireFly_CLICOLOR= then it isn't set
19:14:59 Join linuxstb_ [0] (
19:15:08_FireFly_or has an empty string
19:17:55_FireFly_preglow try to set COLORTERM=yes
19:18:15preglow_FireFly_: neitehr colorterm nor clicolor works
19:18:48ep0ch|i have TERM=xterm and CLICOLOR set
19:18:54ep0ch|i get color
19:19:11_FireFly_preglow: do you have color in the terminal itself ??
19:19:47preglow_FireFly_: yes
19:19:56preglowso it's prolly a vim problem
19:19:59preglowi'm asking some people now
19:20:09_FireFly_it seams
19:24:58webguest31I'm going to blame FreeBSD
19:25:51ep0ch|you done ':syntax on' in vi/vim?
19:26:33preglowyes, of course
19:26:41preglowi think it's a configure time error
19:26:43preglowbloody freebsd
19:27:38 Quit muesli__ (Read error: 110 (Connection timed out))
19:30:11*preglow shoots freebsd and desecrates the corpse
19:31:02 Nick ^BeN^ is now known as Paprica (
19:33:00LinusNPaprica: cool
19:33:12markunpreglow: I have this in my bashrc: export TERM=xterm-color :)
19:33:22preglowmarkun: well, it doesn't work anyway
19:33:26Papricacan you add it to the next daily build?
19:33:33preglowonly TERM i make colour work with is 'ansi'
19:33:34Papricait support h1xx too
19:33:37preglowand then page up/down doesn't work anymore
19:33:44LinusNPaprica: where's the patch?
19:33:57Papricano patch
19:34:01Papricaonly source
19:34:15LinusNok, where's the source?
19:34:17markunpreglow: hm, now I see I've commented it out. Don't know why it works then.
19:34:26Papricamm moment =]
19:34:27 Join muesli__ [0] (
19:34:30_FireFly_markun: echo $TERM ??
19:35:02markun_FireFly_: It's xterm-color in xterm and 'screen' when I use screen.
19:35:39_FireFly_then the xterm set TERM to xterm-color
19:36:47preglowmarkun: hah!
19:36:50preglowTERM=screen works
19:37:14markunscreen is a nice program, you should use it anyway :)
19:37:27preglowi do
19:37:36preglowi'm writing through screen as we speak
19:38:56markunme too
19:39:19 Quit muesli__ (Read error: 104 (Connection reset by peer))
19:39:27preglowonly irc client worth keeping
19:39:43markunxchat was also ok, but I prefer irssi now.
19:39:51preglowi never could stand xchat
19:46:36***Saving seen data "./dancer.seen"
19:51:05 Quit memmem ("ERC Version 5.0.4 $Revision: 1.726.2.19 $ (IRC client for Emacs)")
20:04:36 Join muesli__ [0] (
20:05:35 Join tvelocity [0] (
20:05:47 Quit San (Read error: 110 (Connection timed out))
20:08:24CassandraHmmm. What time does the daily build get updated? I'm trying to work out a good algorithm for when to use today's date to grab the latest daily, and when to use yesterdays.
20:08:25 Join merbanan [0] (
20:08:40BagderCassandra: it starts building at 06:00 CET
20:09:01Bagderusually completes before 06:20
20:09:03CassandraWhat's that in real time?
20:09:11Bagderthat's GMT+1
20:09:23CassandraSo 5 UTC then?
20:10:02Bagderunless during daily savings time ;-)
20:10:16Bagderain't time lovely
20:11:01 Quit Mmmm ()
20:12:26CassandraMaybe I should go back to the original plan of testing for rockbox-<today> on the server.
20:12:40CassandraJust seems like a waste of server resources though.
20:12:48webguest31Just do a HEAD
20:12:52webguest31that's practically free
20:13:21CassandrawxWidgets HTTP support is kind of high level for that.
20:13:34CassandraHowever, I'll ask on #wxwidgets.
20:14:03BagderCassandra: I would say that you should check tomorrow, today and yesterday...
20:14:33Bagderunless you speak GMT
20:14:57Bagderif you want me to, I could make some special file somewhere with meta data
20:15:03CassandraI think the time wxWidgets returns is UTC/GMT unless you tell it you want a specific timezone.
20:16:14linuxstb_Bagder: I was about to suggest that. A simple text file with the date of the latest daily builds would be enough.
20:16:29preglowwhy not just url?
20:16:46linuxstb_The url is different for every target. Just the date changes.
20:16:55preglowanywho, i'm trying to commit a quick working eq here
20:17:15preglowanyone have any suggestions on what they want the settings to look like?
20:17:48CassandraLike an explosion in a jam factory?
20:17:57CassandraOr were you looking for serious suggestions?
20:18:13preglowof course not
20:18:26lostlogicx*cheers for screen +irssi users*
20:19:01*Cassandra looks at lostlogicx funny.
20:19:21lostlogicxreferenced an old conversation.
20:19:36CassandraBagder - yeah - special file sounds like a good idea.
20:21:09CassandraSaves me having to piss around with timezones.
20:21:24Bagderok, I'll make one just after all builds are done
20:21:56muesli__Cassandra thx 4 your your utility :)
20:24:28preglowBagder: perhaps only update the file if the build isn't red?
20:24:48Bagderthe daily build system doesn't know about stuff like that
20:25:57 Join petur [0] (
20:26:03Bagdergood enough?
20:26:59CassandraLooks cool. Can you make it in windows ini file style format? wx has code for parsing them.
20:27:20Bagdershow me and I'll adjust
20:27:36_FireFly_value = "date"
20:27:41_FireFly_for example
20:27:51CassandraThat'd be it.
20:28:11preglowfeels dirty
20:28:27CassandraIf we can also store the latest version number of rbutil.ini in there, I can use it for the update mechanism too.
20:28:46linuxstb_Or could that URL just be rbutil.ini ?
20:29:03Bagderwell, I feel better having this whole file generated by the autobuild system
20:29:46Cassandralinuxstb_, but then you'd download a new rbutil.ini every day.
20:30:01peturLinusN: would it be ok to move the calling of create_numbered_filename() from pcm_recording.c to pcm_record.c?
20:30:02CassandraI'm not sure that's such a great idea.
20:30:22CassandraBadger - I agree.
20:30:27peturLinusN: calling it spins up the disk, so recording effectively only starts after disk spinup
20:30:32Bagder now then?
20:30:38linuxstb_Cassandra: I was just thinking aloud...
20:31:15LinusNpetur: not very popular, since the create_numbered_filename() function is in apps/
20:31:20CassandraThere's no 'y' in dailies, and I think you have a spurious space before date, but yes.
20:31:25peturI know
20:31:38peturBut that disk-spinup really s*cks
20:31:44_FireFly_Cassandra: it shuldn*t matter if the implementation is correct
20:31:51LinusNpetur: yes it does
20:31:54_FireFly_for ini-files-parsing
20:31:58Bagderfixed that too
20:32:12preglowstarting recording spins up disk?
20:32:13preglowthat's bad
20:32:16Bagderthis way we _can_ add other info to this file later on
20:32:35CassandraCan I have a CVS module for the utility, btw?
20:32:39peturIt has to spin up to create the file, but it spins up first to find out what numbered file it can make
20:32:48BagderCassandra: sure!
20:32:49LinusNpetur: we could of course move the create_numbered_filename() function to firmware/ as well
20:33:03peturwould like to do that
20:33:24peturalso, pre-recording is implemented, needs debugging
20:33:30LinusNpetur: or even better, move the recording code to apps/
20:33:39LinusNwell, not better
20:33:59peturnot better at all
20:34:10LinusNin fact, we should consider moving some parts of it to apps/
20:34:23preglowwhat guidelines do we currently have, as to what goes where?
20:34:26peturlike? creating the file?
20:34:59LinusNpreglow: there was once a thought about separating lowlevel and highlevel code
20:35:09peturI'll finish pre-recording first, then we can see
20:35:15LinusNi think the buffering code could be in apps
20:35:33preglowisn't it in apps?
20:35:55peturnot the DMA handler?
20:36:15LinusNthe dma handler should probably be in firmware/
20:36:39preglowi'm thinking most stuff an app would want to re-use should be in firmware
20:36:49LinusNsomething like that
20:36:56preglowwhich includes the entire playback engine...
20:37:25LinusNover time, the firmware/apps division has become more of a nuisance than anything else
20:38:01LinusNi suggest a common directory and a target/ subtree with the target specific code
20:38:11LinusNsome day...
20:38:16preglowwhen we go svn...
20:38:18peturI'm thinking of making the recording interface like this: If you pass it a filename, it uses that one, if you don't, there's the auto-naming that habppens now. Or,
20:38:49preglowat least the file name logic shouldn't be executed until the first buffer write
20:38:51peturnah, no or...
20:39:27peturI want to start the recording before that, yes
20:39:39peturbecause the spin-up takes long
20:40:15peturbut auto-naming causes spn-up atm at a higehr level :(
20:40:29preglowthe disk noise will also disturb the recoridng
20:40:42_FireFly_maybe disk-cache can be used (if enabled)
20:41:21peturoh, you mean create the file if it doesn't exist, which would be at the first savepoint
20:41:32preglowi'm tempted to say the disk cache should be default...
20:41:37SlasheriHmm, tagcache needs ~55 KiB of memory to build all unique indexes for ~5000 songs. I wonder if we can get that much of memory on archoses too (at least plugin buffer would be too small)
20:42:04peturbbl, got some other stuff to do...
20:42:49 Quit DrumRBoy320 ()
20:42:51preglowyou mean like working memory?
20:43:17Slasheripreglow: ram, yes. But that is only needed when building / updating the cache
20:43:32Bagderand Archoses have MUCH bigger disks with larger music collections...
20:44:27preglowokay, i don't think we'll have any trouble with people getting their hands on this particular rockbox target
20:44:33CassandraBadger I think you have to set up a new module (although it could just be that I can't work CVS.)
20:44:37Slasheriof course we could use audio buffer for that, but then the operation would be less transparent (unless we reserve a portion of that buffer when building the db)
20:45:14BagderCassandra: anyone with cvs access can do that. I'd suggest you use "import" in a totally empty dir just to get the module created
20:45:30Bagderthen check it out, then start adding files
20:46:00Papricablah, when i play music and going to change theme or change wps, rockbox stuck in and i need to restart it.
20:46:13Papricasomeone have this problem too?
20:48:48amiconnSlasheri: What's stored in the working RAM for tag cache generation? Your numbers tell it's 12 bytes per entry?
20:53:57lostlogicxBger: have you done battery usage comparisons with LCD turn on and off
20:54:09amiconnpreglow: Performing the file name logic could be delayed until the first buffer flush if we find a clean way to handle that - but only on non-RTC targets
20:54:46Slasheriamiconn: index_entry (4 bytes containing tag length) and the tag data. That is used to create a unique index for certaing tags (artist, album and genre)
20:54:51amiconnThe RTC filename logic doesn't cause a spinup anyway
20:55:12preglowamiconn: why not?
20:55:32amiconnIt composes a filename from time & date
20:55:52Slasheriamiconn: oh, in fact that is tagfile_entry (2 bytes for length), but it seems that it's aligned as 4 bytes
20:56:21preglowamiconn: that's not exactly a guarantee for the filename not already existing
20:56:26amiconnThe archos recording has delayed buffer write to avoid the exact same problem - spinup disturbs recording from internal mic - as long as possible
20:56:36Slasheriamiconn: or never mind, it is 2 bytes as i checked now.. header seems to be long aligned only
20:57:04amiconnpreglow: It is, unless you manage to start recording, stop and start again within one second
20:57:38preglowamiconn: or the file already exists for some reason
20:58:13amiconnThe delayed buffer write doesn't prevent the spinup at start for non-RTC units though, but this wasn't a problem on archos. The only target without RTC that can record is the Ondio FM, and that doesn't suffer from spinup noise ;)
20:59:25preglowthe english lang file is really a jumble of different capitalisation styles now
20:59:58amiconnpreglow: That's very unlikely. Example filename: R051219-210005.mp3
21:00:31preglowbtw, did we remove all the deprecated strings for 2.5?
21:00:47preglowamiconn: yeah, it is unlikely, i never said it wasn't ;)
21:00:52preglowwhy not?
21:01:08amiconnYou can't simply remove them, we need to bump the .lng version
21:01:16preglowand why didn't we do that at 2.5?
21:01:17amiconnI'm in the process of cleaning up...
21:01:22preglowahh, ok, that's what i expected :P
21:01:35amiconnIn fact I was waiting for l10n v2 ;)
21:01:41Bagderand possibly because I was tempting the masses for the v2, yes
21:01:58amiconn...but now unicode is a good reason to do it
21:02:24amiconnI noticed voice files don't contain the .lng version, that's not good
21:04:02amiconnpreglow: I removed all deprecated strings, combined those that were split in two just because there was no splash() back then, and now I am in the process of grouping strings
21:04:06amiconnQuite some work...
21:04:31preglowi can imagine
21:04:49preglowbut yeah
21:04:57preglowwhat's the rule for capitalisation in english?
21:05:16BagderYou Captilize Each Word In Titles, Don't You?
21:05:30preglowOnly Titles?
21:06:51 Join qwisp11 [0] (
21:06:52preglowthere's a lot of non-capped words as well
21:07:44*preglow finally gets to use the sims!
21:09:50nathanh_the #1 rule in English is that there are lots of exceptions
21:09:53preglowisn't it a bit wrong for string "Gain" to be called LANG_RECORDING_GAIN ?
21:09:55amiconnpreglow: I think the idea for english.lang was to capitalise each word in the menus & settings, but to use standard rules elsewhere (like for popup messages)
21:10:43amiconnpreglow: Nope.
21:10:53preglowok, so the id shouldn't reflect what the string is?
21:11:02preglowso i have to use LANG_RECORDING_GAIN identifier in my eq code?
21:11:09preglowsounds logical
21:11:22amiconnThis string is called LANG_RECORDING_GAIN because it is the string for the gain in the recording screen
21:11:28preglowyes, i figured
21:11:32amiconnIt just says 'Gain' to save space
21:11:33preglowbut that string can be shared all over rockbox
21:11:41preglowit's very generifd
21:11:56amiconnLess generic than you might think
21:12:10preglowok, so i should add an identical gain string for my eq as well?
21:12:33amiconnYou could use it everywhere you need the string 'Gain' alone, but not in combination with other words
21:13:23 Join TCK [0] (
21:14:23preglowbut you're still saying i should add my own gain string for the eq?
21:15:13amiconnI'm not sure. The string could say something else than the equivalent of 'Gain' in other languages
21:15:41 Quit muesli__ (Read error: 110 (Connection timed out))
21:16:11preglowit is possible, yes
21:16:16preglowi'll just do that for now
21:17:53preglowand i should in general do this with every string, even though they might be shared?
21:19:35preglowwell, this'll be nice...
21:19:48preglowhow do you suggest i handle multiple bands?
21:20:09preglowthe bands will all have identical strings, apart from a number
21:20:26preglow1: Center Frequency \n 1: Q Factor \n 1: Gain
21:20:27preglowfor example
21:20:35preglowthen 2: Center Frequency, etc
21:21:13LearQ factor?
21:22:08preglowLear: bandwidth
21:22:11pillwill rockbox for ipod allow to get rid of itunes?
21:22:20pilli.e. use the ipod as an UMS?
21:22:21preglowLear: higher q factor = narrower band
21:22:27LinusNpill: yes
21:22:54preglowi might replace Q with actual bandwidth, but then i need to code a fixed point log function
21:22:57preglowwhich i can't be bothered with now
21:23:09pillmeans ipod is to be reconsidered as a multi codec jukebox/UMS
21:23:18pillmeans the nano turns out a kickass option
21:23:22preglowthe usb code will be extremely tricky, though
21:23:47pilli guess
21:23:49preglowin the mean time we'll just reboot to disk mode or apple os, let you transfer your files, then let you play them in rockbox
21:24:02preglowipod will be our first target that hasn't got hardware controlled usb
21:24:10preglowat least it doesn't seem like it's got hardware controlled usb
21:24:16LinusNpreglow: i thought it did...
21:24:33pillgood luck on this one guys
21:25:05preglowLinusN: common consensus seems to be it's software controlled
21:25:54preglow21:25 < aegray_> theres a hardware usb ehci controller which is controlled from software
21:26:25LinusNlostlogic: there?
21:26:31preglowno datasheets.............
21:27:29preglowit's part of the portalplayer core......................................
21:27:39preglowgod, that's going to be a nightmare
21:27:49LinusNindeed................................................................................................................... :-)
21:31:35 Quit qwisp11 ()
21:32:43mirakpreglow: the ipod dosn't have a 5.1 sound :p
21:33:49Learpreglow: what kind of log?
21:34:04preglowLear: like logarithm
21:34:11preglowthe natural one
21:34:15 Quit XavierGr ("Trillian (")
21:34:23LearOk, I was wondering about the base...
21:34:38preglow1/Q = 2*sinh(ln(2)/2*BW*w0/sin(w0))
21:34:42preglowi need to implement that formula
21:34:53preglownow wait
21:34:55preglowi just need sinh
21:35:26_FireFly_ln(2) is a fixed value :)
21:35:41preglowyes, i noticed
21:35:59preglowsin i have, so i just need sinh then
21:36:10_FireFly_what is sinh ??
21:36:17LearBut I remembered wrong; replaygain does an exp (but based on code that I'm quite certain contained a log too).
21:36:17dpassen1hyperbolic sine
21:36:37_FireFly_ah thanks dpassen1
21:37:40preglowsinh is basically just a couple of e^x
21:38:09dpassen1(e^x - e^(-x))/2, iirc
21:38:10preglow(e^x - e^(-x))/2, actually
21:38:31preglowat least we agree
21:38:38dpassen1parentheses and all
21:41:54 Join goa [0] (
21:43:00 Join San [0] (
21:45:41preglowdamn annoying function, fixed point wise
21:45:57preglowi chose to not use tan() for a reason
21:46:01preglowand now it returns here to bite me
21:46:16nathanh_oh, fantastic, tops work to bger for the h300 lcd fix
21:46:37***Saving seen data "./dancer.seen"
21:55:15peturyes, now we just need somebody to commit that and the brightness patches :D
21:55:24wubblanathanh_: lcd fix?
21:55:43peturturn off lcd when backlight is off
21:55:58Papricahow can i get random numbers???
21:56:20wubblapetur: but what is this good for? saving battery life?
21:56:20LinusNPaprica: use rand()
21:57:08LinusNfirst seed it: rb->srand(*rb->current_tick);
21:57:41LinusNthen x = rb->rand() % RANGE;
21:57:57LinusNwhere RANGE is 10 for numbers between 0-9
21:58:09peturwubbla: yep
21:59:12wubblapetur: but i assume that this lcd-issue is not the reason for the weak battery life of rockboxed h300's, right?
22:00:48peturgive it some time... first we need to get all the stuff work ok...
22:01:56 Join DocterD`idle [0] (
22:01:58preglowok, any ideas on how i'm going to fix this x bands with identical settings problem?
22:02:27preglowjust spam english.lang with almost-duplicate strings?
22:03:15wubblai think i'll give lcd_h300_turn_off.diff a try...
22:03:39wubblaif it's considered to be "safe"...?! :-)
22:06:12preglowor perhaps include a different sub-menu for every eq band?
22:06:39preglowthat'll be the cleanest, but very cumbersome
22:06:55peturI'd rather have it like recording settings... no?
22:07:05peturbelow each other?
22:07:07LinusNpreglow: will you have a fixed number of bands?
22:07:14 Quit tvelocity (Remote closed the connection)
22:07:19preglowLinusN: well, if not, i'll have hell with settings
22:07:27preglowLinusN: so i was going for five
22:07:30LinusN5? 7?
22:07:50 Join tvelocity [0] (
22:07:59preglowi'm almost aching to see how this eq will perform
22:08:02LinusNi can imagine a graphical setting
22:08:24preglowi'm _not_ bloody coding any gui thing here
22:08:33 Join Mmmm [0] (
22:08:39preglowsomeone framework it for me, an and sure
22:08:53preglowi hate gui code with a red hot passion
22:09:10peturI've done some in the past
22:09:52preglowif we're going to start supporting graphical settings, we might as well contrive a scheme that'll work independently of lcd sizes first
22:10:07 Quit solexx (Read error: 104 (Connection reset by peer))
22:11:44_FireFly_are pointers on the targest 16bit or 32 bit ??
22:11:57 Join solexx [0] (
22:12:02_FireFly_Bagder: thanks
22:12:04peturLinusN: can I do the filename creation in the recording part? and do something like pcm_get_filename() if the status screen needs to know the file...?
22:12:06 Join XavierGr [0] (
22:13:03 Quit XavierGr ("Leaving")
22:13:33markunI was taking my Gigabeat appart to make some scans but I got scared and put it all back together :)
22:13:55linuxstb_Do you have any spare bits left?
22:14:10_FireFly_uses a boolean var one byte in ram ??
22:14:48preglowat least
22:16:30LinusNpetur: i was wondering, does create_numbered_filename() spin the disk when the disk cache is enabled?
22:16:30markunBagder: Hey, where's your Gigabeat?
22:16:42mirakmarkun: I consider almost impossible to use ffmpeg mpeg4 codec
22:16:58peturprobably not... will try.. just a sec...
22:17:19mirakmarkun: even the decoder function uses the encoder struct
22:17:22amiconnLinusN: It shouldn't, that is, if Slasheri didn't overlook something ;)
22:17:24markunmirak: ok. Then I'll give it a go when I have time.
22:17:33mirakmarkun: good luck
22:17:52markunAt least we have xvid decoding.
22:18:18mirakmarkun: I tried the other approach trying to add methods, but it will take ages, and there is useless structs
22:19:15peturaargh... can't tell as it opens the file right after that, and that definitely spins up the disk
22:20:29mirakwho included flac codec into rockbox ?
22:20:41peturI have still something wrong with recording start. Using internal mic, I should hear the disk spin up in the recording, but only hear ir spinning - starts recording too late :(
22:21:21 Join XavierGr [0] (
22:21:57XavierGrGreat! I just lost all my permission rights in Ubuntu. Now I can't log on at all. :x
22:22:33mirakhow did you did that ?
22:22:42mirakXavierGr: boot with -single option
22:22:58mirakat boot time edit the bootloader options
22:23:05XavierGrI changed my home to root and now whenever I try to log on it says that my home isn't permitted
22:23:37mirakwhat did you do !!!!
22:23:46_FireFly_yeah what did you do
22:23:48webguest31Well that was silly, but do as mirak says and boot with -single
22:23:59XavierGrsingle means?
22:24:04webguest31and change your homedir back
22:24:14_FireFly_mirak: but why he should edit his bootloader settings
22:24:15webguest31edit the line in grub and add "single" after the options
22:24:22XavierGrok thanks
22:24:34webguest31That'll dump you to a single-user root mode
22:24:41mirak_FireFly_: mmm there is maybe a safe mode
22:24:46webguest31Good point
22:24:48mirakyou are right
22:24:50webguest31Use the safe mode
22:24:52XavierGrI have a really hard time with Ubuntu. I don't have any permission on my machine.
22:24:53markunmirak: linuxstb did the ffmpeg flac I think.
22:24:57XavierGrEverthing is locked up
22:25:02_FireFly_mirak: singe is an safe mode
22:25:06preglowi'm the official #ipodlinux paster
22:25:07webguest31XavierGr: what do you want to do?
22:25:08XavierGrI mount a drive and then I cant browse it
22:25:34Bagdernow _that_ is bricking
22:25:37webguest31Yeah, you need to mount it with the right options
22:25:41XavierGrIt mounts it but then the owner of the dir is the root
22:25:54XavierGrand when I open the folder there is nothing there
22:25:55webguest31That is indeed a bit troubling
22:26:34XavierGrthen I logon as root (when I permitted it) and can't see the files of my real account
22:26:42XavierGrI thought that root can see all the files
22:26:57webguest31He can. You must be doing something wrong there :)
22:27:18_FireFly_XavierGr: what partition did you mount in which dir ??
22:27:23webguest31But you should very rarely need to be root
22:27:31XavierGran ntfs partition on /media
22:27:43_FireFly_and this brokes you home-dir ??
22:27:44XavierGrI used the sudo
22:28:01XavierGr_FireFly_: No I broke my home dir myself
22:28:15XavierGrafter I realised that I don't have acces to folders created by root
22:28:30_FireFly_that was the wrong approach for this "problem"
22:28:35XavierGrI was really annoyed and changed the homedir to "/"
22:29:14XavierGrI wasn't aware that it is so easy to brake your account
22:29:28webguest31it's not broken as such
22:29:31webguest31you just can't log in
22:29:40webguest31disabled, you might say
22:29:46_FireFly_you should edit the /etc/fstab file to include the ntfs-partition whith proper mount-options set you can then mount the partition as normal user
22:29:47XavierGralso why when I play some music from cd loud parts get distorted?
22:30:08XavierGr(while in windows is normal)
22:30:37webguest31Turn down the volume :p
22:31:08mirakmarkun: ok so I let you do ffmpeg ...
22:32:06XavierGrAnd I thought I new about PCs. I can't do nothing in linux
22:32:20preglowanyone know of a nice standalone xwindows server?
22:32:22preglownot cygwin, no
22:32:41Bagderyou mean on windows I take it
22:32:42XavierGrI kept reading to guides and information but still I don't get some parts of it...
22:32:54XavierGranyway I will reboot and see if -single works out.
22:33:02*petur is confused
22:33:22webguest31single, not -single
22:33:36peturwhat's the difference between apps/pcm_recording.c and apps/recorder/recording.c?
22:33:43XavierGrso I will add a "single" command to grub?
22:33:58webguest31First use the "rescue mode" or whatever it is
22:34:04webguest31that should work
22:34:15preglowBagder: yea
22:34:24 Part XavierGr
22:34:25_FireFly_preglow: the only one which i just found is an comercial one
22:34:38Bagderon Windows I've only used the cygwin X server the last couple of years
22:34:38preglowi suddenly remembered i can't be bothered to make a mingw cross compiler as well
22:35:00Bagderthere's a mingw debian package!
22:35:06mirakwhy do you want a standalone server ?
22:35:08preglowthis is a freebsd server :////////////
22:35:16LinusNpetur: recording.c is the archos code
22:35:21preglowwhich means nothing useful will work on it, ever
22:35:27preglowespecially not debian packages
22:36:07mirakdrop it
22:36:13peturweird: if I modify pcm_recording, make doesn't see it, but it does if I change recording.c? This I must confirm...
22:36:15mirakwhy don't you put linux on it ?
22:36:32miraklinuxstb_: are you around ?
22:36:33preglowmirak: if it was my server, it would not be running freebsd in the first place, believe me
22:36:41linuxstb_mirak: Yes.
22:37:02miraklinuxstb_: markun told me you implemented ffmpeg flac in rockbox is that right ?
22:37:04LinusNpetur: sorry, i'm an idiot
22:37:22LinusNpetur: apps/pcm_recording.c is obsolete
22:37:34linuxstb_mirak: Yes. I can guess your next question...
22:37:37peturjust confirmed this ;)
22:37:46miraklinuxstb_: we will see that, try :)
22:38:01linuxstb_You want to know how to rip out the xvid decoder.
22:38:06 Quit Moos ("Happy Birthday Rockbox")
22:38:08miraklinuxstb_: ?
22:38:25miraklinuxstb_: there is no xvid decoder into ffmpeg, it uses xvidcore
22:38:27 Join Moos [0] (
22:38:38miraklinuxstb_: or encoder whatever
22:38:57miraklinuxstb_: that's just an interface I think
22:38:57markunmirak: the mpeg4 decoder decodes xvid..
22:39:03linuxstb_What's h264.c ?
22:39:10mirakmarkun: yeah I realised that
22:39:29preglowh264 is reaching a bit high
22:39:33miraklinuxstb_: h264 is the new format for blue ray
22:39:40mirakisn't it ?
22:39:48linuxstb_What's xvid then?
22:39:50mirakit's the thing that apple is promoting
22:39:54*linuxstb_ is confused...
22:40:03preglowh264 is the new mpeg4 based video codec
22:40:06preglowalso called avc
22:40:07miraklinuxstb_: xvid is an anhanced encoder I think, I am not sure
22:40:12preglowit's a very hard deocde
22:40:26mirakyou need 2ghz cpu to decode it
22:40:35preglownot exactly
22:40:39preglowbut it's very intensive
22:40:59linuxstb_But isn't that because it's used for HDTV resolutions - e.g. 1280x720 @ 60fps
22:41:07 Quit Mmmm ()
22:41:29preglowit's plain and simple because it's more advanced
22:41:38miraklinuxstb_: well we were looking with markun how to extract the mpeg4 decoder out of ffmpeg, but well seems there is a lot of work. A lot more than xvid
22:41:53preglowfines motion prediction modes, deblocking filter, more coef prediciton
22:41:59preglowgod, how i write like swine
22:42:12linuxstb_h264 is what the video ipod plays.
22:42:14miraklinuxstb_: I was wondering what approach you used to extract flac code
22:42:20preglowthat and ordinary mpeg4 video
22:42:27miraklinuxstb_: in 320*240 at best
22:42:39linuxstb_No, in 640x360 or 480x480
22:42:49miraknot for h264 I have read
22:42:57mirakmight be wrong
22:42:57linuxstb_I've played such files.
22:43:43mirakso we are wondering if it's worth to implement ffmpeg mpeg4 decoder instead of xvid one, in case it's faster
22:43:55preglowjust have a look at it
22:44:04linuxstb_But anyway, ripping things out of ffmpeg is a nightmare - unless you are very familiar with the file format of the codec you are dealing with.
22:44:41miraka nightmare ok, if even you, you say it
22:45:08preglowit is a nightmare to rip stuff out of ffmpeg
22:45:11linuxstb_Maybe I'm exaggerating, but there is a _lot_ of code in ffmpeg, and the problem is identifying the minimum you need.
22:45:14pregloweverything is very intertwined
22:45:16mirakthe framework is nice, but not very modular
22:45:44preglowi've got to go
22:45:47preglowsee y'all later
22:46:07_FireFly_preglow: cu
22:46:17miraklinuxstb_: I used the approach by adding methods instead of ripping, but that would take ages as well
22:46:40mirakI started with just one method, trying to had types etcetera
22:46:59linuxstb_One good thing about ffmpeg is that the actual decoder will probably just be one or two .c files.
22:47:18mirakwhat do you mean ?
22:47:39BagderI think he means that the actual decoder will probably just be one or two .c files
22:47:53mirakand you what do you mean ?
22:48:17linuxstb_I mean in libavcodec/ there will probably just be one or two .c files that implement the decoder. But those files will use general-purpose functions from elsewhere in ffmpeg that you will need to track down.
22:48:20miraklinuxstb_: ?
22:48:21markunmirak: that the actual decoding is only in h263dec.c
22:48:45mirakyes or general types
22:48:51mirakand structs
22:49:21_FireFly_ this is the first design approach for a token-based wps-handling
22:49:43mirakmarkun: will you do it ?
22:49:57markunI will try, but I don't have time now.
22:51:33peturLinusN: just calling rec_create_filename(path_buffer) without actually starting recording spins up the disk (with dircache enabled) :(
22:53:53peturthats on H300 (CONFIG_RTC not defined)
22:57:44mirakmarkun: maybe there is cleaner code in mplayer or vlc, I think they have the codecs in the source code too
22:58:26_FireFly_afaik mplayer uses ffmpeg
22:58:59LearInteresting... ffmpeg has its own vorbis decoder, using an fft. Not fixed point though...
23:00:08 Join Strath [0] (
23:00:18linuxstb_Lear: Have they implemented it themselves (like the FLAC decoder), or is it derived from libvorbis?
23:00:46merbananlinuxstb_: ffvorbis is implmented from specs
23:01:08linuxstb_Shame it's not fixed point then.
23:01:27merbananis tremor so bad ?
23:02:26 Join San||Away [0] (
23:02:54linuxstb_I'm just going by the startling difference between libFLAC and the ffmpeg FLAC decoder - so I'm prejudiced in favour of ffmpeg always being best.
23:03:12 Join muesli__ [0] (
23:04:15merbananffvorbis is abit faster then libvorbis, but thats when using sse
23:04:32Learffvorbis doesn't do floor0 files either, it seems.
23:07:03lostlogicxmerbanan: how does ffvorbis compare to tremor on PC hardware, out of curiosity?
23:08:32 Quit San (Read error: 110 (Connection timed out))
23:09:04merbananlostlogicx: no idea, the only thing I heard was that it was 10% faster then libvorbis (not tremor)
23:10:19mirakI have read that he mplayer dev said xvid decoder is as fast as ffmpeg if not faster on a 2002 post on xvid mailing list.
23:11:13linuxstb_That was three years ago though... You could check the ffmpeg CVS to see what changes have been made.
23:11:30mirakmmm ...
23:11:31merbananI think they are almost as fast with ffmpeg a littlebit faster
23:12:20merbanananyway most of the speed is from simd code
23:12:28mirakffmpeg code have the advantage of having arm assembly optmisations
23:12:43mirakfor some ipod colors it can be worth
23:13:10linuxstb_I'm just trying to get audio first....
23:13:25markunand can also give us a clue how much speed can be improved on H300 by rewriting that code into asm
23:13:43markunlinuxstb_: get on with it :)
23:14:26markun.. with the voice of an entire army from quest for the holy grail :)
23:15:08mirakmarkun: there is x86 assembly optimisation in xvid code also. we can know what it's worth to optimize at least
23:15:36mirakyou know what the last time I did 68k assembly it was ten years ago.
23:15:40mirakon an atari st
23:15:53mirakI cracked a game
23:17:06mirakthe ridiculest protection ever. It was a protection by symbol. So there was a manual with 30 symbol and it asked randomly one among 30. I just to change the rand by a move to always have the first symbol
23:17:29mirakI am rpoud of that
23:18:26 Quit Lear ("Chatzilla 0.9.69 [Firefox 1.5/0000000000]")
23:19:25_FireFly_good night everybody
23:20:25 Quit _FireFly_ ("Leaving")
23:20:40lostlogicxlsr shifts zeros in at the left, asr shifts MSbs in right?
23:22:42 Quit Kohlrabi ("Leaving")
23:31:04 Quit rayo ()
23:43:10 Join SnokeekonS [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
23:44:04 Nick SnokeekonS is now known as AliasCoffee (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
23:46:40***Saving seen data "./dancer.seen"

Previous day | Next day