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

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

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

#rockbox log for 2007-08-10

00:03:03 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
00:04:58*jhMikeS sees a pattern between some c code of recent importance and some e200 disasseblies - and that's it. )
00:05:12 Nick HellDragon is now known as helldragounette (i=Nocebo@unaffiliated/helldragon)
00:05:38*ender` yawns
00:06:23 Nick helldragounette is now known as HellDragounette (
00:10:20 Join RMenes379 [0] (
00:10:21 Quit RMenes379 (Remote closed the connection)
00:10:57 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
00:11:23 Join LambdaCalculus37 [0] (
00:12:21LambdaCalculus37Ahh, yes, dear friends!
00:12:55 Quit massinissa ("Lost terminal")
00:13:13 Nick HellDragounette is now known as HellDragon (i=Nocebo@unaffiliated/helldragon)
00:13:40 Nick HellDragon is now known as HellDragon` (i=Nocebo@unaffiliated/helldragon)
00:14:01 Quit LambdaCalculus37 (Client Quit)
00:14:14 Join LambdaCalculus37 [0] (
00:14:23 Quit LambdaCalculus37 (Client Quit)
00:14:49 Quit spiorf (Read error: 110 (Connection timed out))
00:15:01 Join My_Sic [0] (
00:15:01 Quit saratoga ("CGI:IRC (Ping timeout)")
00:15:03 Join LambdaCalculus37 [0] (
00:15:14 Join petur [0] (n=petur@rockbox/developer/petur)
00:15:19LambdaCalculus37I think now I should be stablized.
00:15:35 Join spiorf [0] (
00:16:01peturac_away: it won't be for tonight anymore, see ya tomorrow...
00:16:09 Quit petur (Client Quit)
00:17:41 Quit matsl ("Leaving")
00:18:10LambdaCalculus37Am I coming through okay?
00:18:22bluebrotherLambdaCalculus37: yeps
00:18:30bluebrotherloud and clear.
00:18:53LambdaCalculus37I wonder why my IRC client decided to take the 9 off my name.
00:19:14bluebrothermaybe because of a nick lenght limitation of the irc server?
00:19:19LambdaCalculus37Now I'm just the most randomly picked random number.
00:19:41LambdaCalculus37Anyways, I came by tonight to introduce myself on IRC, even though I've been on the forums for a while now.
00:19:51bluebrotherwas there some deep meaning in that number? Looked like random to me anyway ;-)
00:20:10LambdaCalculus37379 is a Chen Prime, and my birthday... March '79.
00:20:28LambdaCalculus3737 is the number most people pick when asked to pick a number between 1 and 100.
00:21:24Soul-SlayerI'd say something more like 50 personally :p
00:21:30LambdaCalculus37I actually stopped by at GodEater's suggestion.
00:21:57bluebrotherwell, you might want to check the freenode documentation about nickname lenght restrictions. IRC servers usually have one.
00:22:08bluebrotherbut apart from that, welcome ;-)
00:22:33 Quit Guile`` ("KVIrc 3.2.0 'Realia'")
00:22:33LambdaCalculus37It has been quite a while since I last visited an IRC channel.
00:22:51bluebrotherI don't think irc changed much since then ;)
00:24:00rasherbluebrother: Does the talk file creating screen need to include that many options? Can't sensible defaults be picked instead? (and the encoder executable searched for in the user's PATH)
00:24:53rasherbluebrother: or perhaps some presets (like "lame - high quality", "lame - low quality")
00:24:53bluebrotherrasher: Domonoky made that, but he talked about moving the options to the configuration dialog
00:24:53bluebrotherbut presets sound like a good idea to me.
00:25:34rasherOf course, I'm merely complaining, not offering to write a patch, so feel free to ignore me
00:26:18rasherI just don't think that level of configuration is needed, and will simply make talkfile creating look difficult and put users off
00:26:22bluebrotherwell, I agree that the window is kinda crowded.
00:26:43rasherHeh, "Current cache size is %1"
00:26:51bluebrotherit's similar to my idea of moving the mountpoint selection to the configuration dialog −− I like it much better this way
00:26:57bluebrotheryeah, that's intended.
00:27:13bluebrother(resemblence of QString::arg())
00:27:34pixelmawhat options are available? Just reminded me that amiconn talked earlier about some limitations about bitrate and working frquencies on Player and Archos in general... (must be in yesterday's log)
00:27:59rasherpixelma: you can enter your own commandline options to lame and the tts, so any options they support
00:28:09bluebrotherfor talk files? I thought this applies only to voice files.
00:28:44bluebrotherentering command line options isn't the best solution for average users though ...
00:28:44pixelmayeah, but I think that the average Archos user (for example me) doesn't know about those...
00:28:49 Join krazykit [0] (
00:29:17bluebrotherit could be a simple configuration dialog showing presets, and an average user can extend the dialog and adjust the values manualls
00:29:21pixelmahmm... I was thinking voice files - though don't talk clips work the same somehow?
00:29:39bluebrothertalk clips are simply renamed mp3 files
00:29:47bluebrotherat least as far as I understood ;-)
00:30:15rasherAbsolutely, which is why I think a few presets is vastly preferable. I don't even think the user needs to be able to specifically set options.
00:30:23 Quit JdGordon (Remote closed the connection)
00:30:43bluebrotherwell, setting the path to the binary might be needed.
00:30:54pixelmaLambdaCalculus37: I thought you were calculating </bad joke> :P
00:30:58bluebrotherjust imagine the localized paths on windows
00:31:07rasherIt just invites people to be obsessive-compulsive about it. They're far more likely to just pick a preset and go on with it if that's all there's available.
00:31:20bluebrotherit's c:\program files vs. c:\programme (in german) etc.
00:31:33rasherbluebrother: those should be able to obtain programatically though
00:31:42LambdaCalculus37pixelma: I got bored. :P
00:31:53bluebrotherhmm. Is there a "standard" way to do so on windows?
00:31:55rasherC:\Documents and Settings\rasher>echo %PROGRAMFILES%
00:31:56rasherC:\Program Files
00:32:27rasherThe windows api have ways of obtaining these paths as well. I should hope Qt offers a bridge to these
00:32:34bluebrothernow you need to obtain the folder with the tts engine
00:33:14rasherbluebrother: anyway, couldn't it be autodetected, and a "select path" only shown if it can't be found?
00:33:32bluebrothermight be possible.
00:33:35rasheror "my tts is not in the list" selected
00:33:41Ishi``ipod : 1.1.1 : over bug not 1.3.1 :)
00:33:49bluebrotherover bug?
00:33:50LambdaCalculus37rasher: Isn't a "select path" the simpler choice to make?
00:34:14Ishi``prefecth abort , undefined constructruction
00:34:18rasherLambdaCalculus37: For the programmer, certainly. For the user, having a simple drop-down list to chose from is simpler, surely.
00:34:21Ishi``is not firmware apple cause 1.3.1
00:34:30bluebrotherafter selecting the path rbutil could try to detect which tts has been selected
00:34:37 Join DanielW [0] (
00:34:58bluebrotherand simply allow to choose between high and low quality preset
00:35:20LambdaCalculus37rasher: True, true... certainly makes the choice a might tough, doesn't it?
00:35:37 Part Quelsaruk ("Time makes no sense")
00:35:47DanielWi got my damn ipod in diagnosis modus again. and now, it says "SDRAM Full Test. Reset to continue." but i tried everything i don't find away to get it to reset
00:35:54bluebrotherhmm, espeak homepage doesn't offer an installer for windows.
00:35:55rasherI guess it's using GNOME influencing me, since they tend to (attempt) to make smart choices for the user and not bother him with the choices if at all possible
00:36:11rasherLambdaCalculus37: I don't understand what you just said
00:36:12 Quit salty-horse ("Leaving")
00:36:14bluebrotherso there is a high chance it's installed just somewhere
00:36:48bluebrotherrasher: that's one of the things I dislike about gnome. It looks like written for stupid users ...
00:36:54bluebrother(well, at least to me)
00:36:57LambdaCalculus37rasher: What I meant to say was, does the coder make the choice for his benefit, or for the end-user?
00:36:58DanielWit's an ipod 5g
00:37:19bluebrotherDanielW: hole Menu + Select
00:37:33LambdaCalculus37bluebrother: So which environment do you prefer? KDE? WindowMaker?
00:37:35rasherbluebrother: kde looks like GNOME written by dumb programmers (to me)
00:37:45bluebrotherI'm kde user.
00:37:58 Quit sergey (Remote closed the connection)
00:38:06rasher"Oh I can't decide how to write this feature, let's just make a gajillion options instead of making a choice"
00:38:13DanielWbluebrother: doesn't work. and this blinking of the backlight is making me crazy
00:38:47bluebrotherhehe ... that's the exact point where I disagree. But we have both environments, so everyone can use the one he thinks is best ;-)
00:38:50rasherIt's a philosophy argument of course, and I just happen to like the way GNOME does it, so that's where I'm coming from.
00:39:05bluebrotherDanielW: flip hold on. flip it off again. Hold Menu + Select
00:39:17bluebrotherand hold it _long_. Can be something up to 30 seconds or so.
00:39:54DanielWbluebrother: oh yeah thanks, now it worked :-)
00:40:34bluebrotherreset needs patience (and I think it's good that way)
00:40:38LambdaCalculus37DanielW: Make a mental note of what bluebrother said, okay? Up to 30 seconds or so if you have to reset.
00:40:55n1sin GNOME I can't do what I want because someone decided that the option is unneccesary, in KDE I can't find the option... :_)
00:41:13rasherbluebrother: consider it my attempt to make sure rbutilqt finds a middle-ground, unlike both kde and GNOME
00:41:36bluebrotherthe funny think is that reset also varies on the same device. I had it sometime needing like 5 seconds, other times 30
00:41:49bluebrotherrasher: I agree that we need something in-between.
00:42:07LambdaCalculus37For the benefit of everyone, coder and end-user alike.
00:42:41*n1s stops staring at talk.c for the night...
00:42:57 Quit atsea- (Read error: 104 (Connection reset by peer))
00:43:02rasherbluebrother: where does rbutilqt keep it's user-config?
00:43:06bluebrotherbtw, if someone with screen reader wants to test the current state: I got it build statically with accessibility support compiled in.
00:43:30DanielWmy rockbox installation is now about 3 month old on my ipod, should i go and make an update?
00:43:31bluebrotherrasher: either in the program folder (for portable installation) or in the "standard" location depending on OS
00:43:47bluebrotheron linux it's ~/.config/
00:43:51LambdaCalculus37DanielW: Absolutely yes!
00:44:06LambdaCalculus37DanielW: You should always keep your build up to date!
00:44:24DanielWbut i am afraid, that i will get problems
00:44:26rasherbluebrother: .config/ didn't see that coming
00:44:51LambdaCalculus37DanielW: What's there to be afraid of? If you encounter a bug that wasn't there before, you come an tell us about it, and we'll look into it.
00:44:56 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
00:45:11bluebrotherit uses QSettings defaults. Look here:
00:45:12rasherbluebrother: Does it not attempt to auto-detect a device on first startup?
00:45:24bluebrothernot yet.
00:45:26LambdaCalculus37DanielW: But you have to tell us what SVN build you found the bug in, of course.
00:45:37bluebrotherexcept that ini files are used on all platforms.
00:45:47bluebrotherautodetection isn't finished yet.
00:45:52rasherbluebrother: Oh well, I read somewhere that .config is "the future", so I suppose I'll see more of it.
00:46:09rasherbluebrother: I can see - just got a segfault trying
00:46:27bluebrotheroh −− haven't had segfaults
00:47:33LambdaCalculus37Oh crap, I've gotta run.
00:47:49bluebrotherbut after my initial checkin Domonoky added quite some stuff.
00:48:01LambdaCalculus37Well guys, it's been fun. But I have some "real world" stuff to take care of.
00:48:10LambdaCalculus37I'll be around again!
00:48:17bluebrotherit's amazing how much is already working given the fact that my initial commit is just about two weeks ago
00:48:24 Quit LambdaCalculus37 ()
00:48:40rasherAt least rbutilqt fits on my screen. The old one didn't (at 1024x768)
00:49:25Soul-SlayerDanielW: What player do you have?
00:50:11DanielWSoul-Slayer: ipod 5.5g
00:50:19 Quit midkay ("Leaving")
00:50:26DanielWipod video 80 gb
00:50:44Soul-SlayerDanielW: I believe that target is still stable. If you do update, make sure you reinstall the bootloader too
00:51:31perrikwpDerPapst: I'm back and ready to test your patch.
00:51:40bluebrotheryou can just make a backup copy of your old installation
00:51:53 Quit ender` (" I believe that God left certain drugs growing naturally upon our planet to help speed up and facilitate our evolution. Okay,)
00:52:20 Part n1s
00:56:10DerPapstperrikwp: ok. apply it and then do what i have quted before from amiconn
00:59:50 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
01:00:11 Join miepchen^schlaf [0] (
01:01:20 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
01:01:41 Join miepchen^schlaf [0] (
01:04:04 Quit bospaadje ("great minds run in great circles")
01:08:34perrikwpDerPapst: LCD R2: 0055, R3: 120C
01:09:34 Quit XavierGr (Nick collision from services.)
01:09:37DerPapstand what is with switching of the display? did it work at all=
01:09:38 Join XavierGr [0] (
01:10:35 Join sarixe [0] (
01:11:30perrikwpthe screen turned off and the backlight stayed on but the registers didn't change
01:11:30rasherDamn. Guess I'll have to shutdown my translate-Rockbox-page.
01:11:35rasherShort run it had.
01:11:44DerPapstperrikwp: that is ok
01:12:00DerPapstbut there were no black lines or such
01:12:06perrikwpnone at all
01:12:13DerPapstgood :)
01:12:29perrikwpglad i could help
01:12:30 Quit ompaul ("later")
01:13:10DerPapstbtw does your ipods contrast change depending on the amount of black pixels being displayed at the same time?
01:13:59perrikwpi don't know, i haven't really notice anything like that before
01:14:56 Quit Juice^ ("Leaving")
01:14:59DerPapstthen you probably don't have that problem
01:15:47DerPapston G1 to G3 the contrast changes depending on the amount of black pixels
01:17:21perrikwpif you need to test anything else with my ipod just tell me. i'm always happy to help
01:17:21DerPapstthen thanks for testing :)
01:17:40pixelmarasher: is it impossible to maintain with the recent changes?
01:18:10DerPapstperrikwp: i guess amiconn will have some more tests for you in the future ;)
01:18:27perrikwpok, i will look forward to that
01:20:07rasherpixelma: yes. I'd pretty much have to set up a build-environment on the server. Which might be possible if it was a linux-server I had ssh access to. But it's a windows server and I have only ftp access. Getting genlang to run at all was no small feat.
01:20:44Soapwhere is colinux.php when you need it? ;)
01:21:34 Quit dan_a (Read error: 110 (Connection timed out))
01:21:38 Join robin0800 [0] (
01:21:40bluebrotherstrange. Seems QHttp has problems using the loopback device
01:21:42rasherIf anyone's interested in hosting it, I'll gladly provide the source.
01:21:46pixelmarasher: pity, it looked very handy
01:21:52 Join hannesd_ [0] (
01:22:01bluebrotheror something is broken in my localhost setup :o
01:22:12rasherpixelma: Yeah, I'm a little disappointed myself.
01:22:12Soaprasher, why do you need to shut it down? TOS violation?
01:22:58rasherSoap: No, practical limitations. I can't run genlang properly without a built environment. And I can't very well set up a build environment with only ftp access. Well, perhaps I could, but I'm not that much of a masochist.
01:23:17rasherOn a windows server..
01:23:23 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
01:23:41Soapone sec
01:23:44 Join miepchen^schlaf [0] (
01:24:26 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
01:24:47 Join miepchen^schlaf [0] (
01:25:23 Quit miepchen^schlaf (Client Quit)
01:27:37Soapnah, SSH access isn't one of the features I can "tack-on" to my hosting.
01:28:13DerPapstGood night everyone :)
01:28:19 Quit DerPapst ("So Long And Thanks For All The Fish!")
01:28:33 Quit robin0800 (Read error: 104 (Connection reset by peer))
01:29:30 Join robin0800 [0] (
01:29:39 Join perrikwp_ [0] (n=chatzill@
01:29:41rasherSoap: it rarely is, unless you have a virtual server-type deal (or real colo).
01:29:55Soapthat's what I assumed, but I needed to check.
01:30:34rasherActually. I'm talking complete nonsense I think.
01:31:04rasherI *don't* need a build-environment, since I'm only running genlang -u, not any of the target-specific modes.
01:31:12rasherSo I'll just have to make minor adjustments, I think.
01:37:11 Quit hannesd (Read error: 110 (Connection timed out))
01:37:12 Nick hannesd_ is now known as hannesd (
01:42:42 Join Entasis_ [0] (
01:46:27 Quit chrisjs169_ (Read error: 110 (Connection timed out))
01:46:29*rasher wonders what the difference between none and "" is in a languagefile
01:46:46 Quit perrikwp (Read error: 110 (Connection timed out))
01:51:11 Join haemmy [0] (n=stefan@
01:51:45 Nick perrikwp_ is now known as perrikwp (n=chatzill@
01:53:50 Quit robin0800 (" HydraIRC -> <- Wibbly Wobbly IRC")
01:56:16 Part pixelma
01:56:59***Saving seen data "./dancer.seen"
01:58:10 Quit spiorf (Remote closed the connection)
01:59:49 Quit Entames__ (Read error: 110 (Connection timed out))
02:01:52 Quit bluebrother ("sleep.")
02:03:04 Quit SliMM ("Chatzilla 0.9.74 [Firefox]")
02:06:49rashern1s: (for the logs) What's the difference between none and "" in a language-file?
02:11:24 Join kubiixaka [0] (
02:12:14 Quit Thundercloud (Remote closed the connection)
02:12:56 Join midkay [0] (n=midkay@rockbox/developer/midkay)
02:25:11 Join barrywardell [0] (
02:25:31 Quit haemmy ()
02:28:48 Quit kubiix (Read error: 110 (Connection timed out))
02:32:32 Quit bdgraue (Remote closed the connection)
02:32:47BigMacHey, anyone mind telling me of a book/website that they found of use when learning ARM assembly?
02:42:48 Quit barrywardell ()
02:46:41 Join barrywardell [0] (
02:47:50jhMikeSBigMac: and DDI0100E_ARM_ARM.pdf
02:50:11 Quit wippeout (Read error: 110 (Connection timed out))
02:50:51BigMacjhMikeS: Sorry if this sounds a little arrogant, but I don't see how reading a dictionary would make me fluent in english
02:51:07BigMacjhMikeS: Unless there is something else you supplemented that with
02:51:36LloreanEnglish has grammar, each word isn't really atomic in the same way individual instructions are moreso.
02:52:19LloreanOne could, in theory, learn C by the O'Reilly pocket reference, and some experimentation.
02:53:39BigMacOh, I see your point. I was thinking of more of the tutorial type route though
02:54:30LloreanHave you just googled "arm assembler tutorial" or something similar then?
02:54:55rasherIt's probably also a matter of whether or not you are comfortable with assembly programming already. If you are, a reference is probably very helpful, but if you're not, it's less likely to be what you need.
02:54:57 Quit XavierGr (Nick collision from services.)
02:54:59 Join XavierGr_ [0] (
02:55:14BigMacYes, but they all seemed to be rather old, I am currently searching
02:55:28BigMacBut coming up with nothing related to arm assembly
02:55:42LloreanWhy does age matter particularly much?
02:56:34BigMacLlorean: Don't programming languages change over time?
02:56:42BigMacdifferent revisions
02:56:48BigMacthingds get depreciate
02:57:01BigMacthings depreciated
02:57:14LloreanARM Assembler shouldn't really be considered a programming language, for one thing.
02:57:26 Quit barrywardell ()
02:57:50LloreanFor another, if you know specifically what version of a programming language (or in this case, which type of ARM processors you're dealing with) you just search for that.
02:57:50jhMikeSBigMac: no, just that basically
02:58:22BigMacjhMikeS: Then kudos to you
02:58:53jhMikeSWell, I just did some simple code to start. the memswap128 function.
02:59:02 Nick HellDragon` is now known as HellDragon (i=Nocebo@unaffiliated/helldragon)
03:24:25krazykitbuh... someone has claimed they're running rockbox on the e280r without modifications...
03:24:28advcomp2019did anyone look at the e200r thread lately
03:24:39advcomp2019i was going to say that too
03:24:41iambenkrazykit did
03:24:58krazykitthat's some hell of timing, eh
03:25:14advcomp2019i was looking there too
03:25:20iambenthe timing on my smartass comment could have been better =(
03:26:05krazykitanyway, i call shens on it. doesn't seem that legit to me. i'd like a photo of rockbox running on it.
03:26:11advcomp2019i wonder if he really have the r version
03:27:04advcomp2019that user is online if i am reading right
03:27:16iambendoes the r version look any different, aside from a tiny little r after the model name on the backplate?
03:27:48advcomp2019in the main menu there is rhapsody channels
03:28:05advcomp2019that is the easiest
03:28:21advcomp2019i would say
03:28:24krazykitphysically, though, the device looks the same?
03:28:40advcomp2019but for the r
03:28:44advcomp2019on the back
03:29:16Lloreankrazykit: As far as we know the hardware is the same.
03:29:26LloreanThe real question would be if it shows the Rhapsody thing during boot, I think
03:29:47advcomp2019that is another way
03:30:56krazykitLlorean, but just installing rockbox like on the non-r has been tried, i assume, and didn't work?
03:31:22Lloreankrazykit: They don't have the hidden partition the same way the e200 does, Sansapatcher wouldn't do anything
03:31:41LloreanMeanwhile, they reject Rockbox .mi4 files
03:31:45LloreanIf you try to use recovery mode.
03:32:45krazykittricky tricky.
03:32:53 Quit DanielW ("Leaving")
03:33:35LloreanWell they require a signed file, iiuc.
03:35:54advcomp2019i wonder if he has a plain version with a r versions battery cover
03:55:11 Join Rob222241 [0] (n=Miranda@
03:55:11 Quit Rob2222 (Read error: 104 (Connection reset by peer))
03:56:22 Join tihoc4n [0] (
03:57:00***Saving seen data "./dancer.seen"
03:57:34BigMacI should probably learn C before assembly right?
03:57:52BigMacIt seems to be referencing C functions a lot in what I am reading
03:59:36 Part BigMac ("Leaving")
03:59:48 Join BigMac [0] (
04:00:21BigMacjhMikeS: Is there a RTE or a simulator for arm, or how did you test your code?
04:06:16jhMikeSBigMac: it's more complicated for arm
04:06:38BigMacjhMikeS: What do you have to do to test your code then?
04:06:54jhMikeSI just tested it in place though. I knew playback.c well enough to get memswap128 to work.
04:07:17jhMikeSYou're writing an interrupt handler in asm?
04:08:03BigMacI haven't written anything yet, I am just trying to figure out how I'd know if my code will work
04:09:12jhMikeSI didn't really know actually. When it bombed-out it I just spent awhile figuring out why and reading the pdf over and over and looking at asm in rockbox.
04:09:52jhMikeSI also implemented the code as C first then made asm out of it.
04:09:53 Quit tchan ("WeeChat 0.2.6-cvs")
04:10:45jhMikeSGiven that the C function worked and what called it worked, all debugging could be in the asm.
04:30:43 Join XavierGr [0] (
04:32:40 Quit Soul-Slayer ("Leaving.")
04:33:36 Join TiMiD[FD] [0] (n=TiMiD[
04:34:03TiMiD[FD]hello sleepers !
04:35:49advcomp2019did anyone find out more info on that post
04:38:47 Quit ctaylorr (Read error: 110 (Connection timed out))
04:40:06 Join Febs [0] (
04:42:19 Join saratoga [0] (i=98039a34@gateway/web/cgi-irc/
04:44:03 Quit XavierGr_ (Read error: 110 (Connection timed out))
04:46:03 Quit XavierGr ()
04:50:52 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
05:01:11 Join TinoM| [0] (
05:13:50 Join Soul-Slayer [0] (n=Administ@
05:19:06 Quit TinoM (Read error: 113 (No route to host))
05:19:44 Join Tino|Home [0] (
05:25:48 Part tihoc4n
05:36:45 Quit TinoM| (Read error: 110 (Connection timed out))
05:53:33saratogaanyone familar with the codec api and still awake?
05:55:06 Join aliask [0] (
05:57:01***Saving seen data "./dancer.seen"
05:57:45saratogahmm i guess i'll go to bed then
05:57:57saratogafor the logs, i managed to accurately seek in my wma test file
06:01:47Soul-Slayersaratoga: Nice job
06:03:16 Quit sarixe ("Peace")
06:06:20 Quit maffe (Remote closed the connection)
06:11:11 Quit saratoga ("CGI:IRC (Ping timeout)")
06:13:54 Quit pabs (Read error: 113 (No route to host))
06:16:19 Join midgey [0] (
06:18:43 Join toffe82 [0] (n=chatzill@
06:20:37 Part toffe82
06:20:55 Join toffe82 [0] (n=chatzill@
06:21:07 Part toffe82
06:21:24 Join toffe82 [0] (n=chatzill@
06:24:39 Quit ToHellWithGA (Read error: 110 (Connection timed out))
06:26:06 Part toffe82
06:26:46 Join toffe82 [0] (n=chatzill@
06:28:39 Join homielowe_ [0] (n=chatzill@
06:33:20 Part homielowe_
06:34:34 Join homielowe_ [0] (n=chatzill@
06:40:33 Join advcomp2019- [0] (n=advcomp2@
06:40:44 Quit advcomp2019 (Nick collision from services.)
06:41:02 Nick advcomp2019- is now known as advcomp2019 (n=advcomp2@
06:43:10 Quit homielowe_ ("ChatZilla [Firefox]")
06:44:12 Part Soul-Slayer
06:44:50 Quit homielowe (Read error: 110 (Connection timed out))
06:45:14 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
06:52:21 Quit elinenbe (Read error: 110 (Connection timed out))
07:04:55 Join n17ikh|Lappy [0] (i=n17ikh@dial-208.r3.ncbrvr.InfoAve.Net)
07:13:34 Join Doomed_ [0] (
07:18:55 Quit n17ikh|Lappy ()
07:22:09 Join n17ikh|Lappy [0] (i=n17ikh@dial-10.r2.ncbrvr.InfoAve.Net)
07:25:30 Quit alienbiker99 (Read error: 110 (Connection timed out))
07:34:33 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
07:48:08 Quit nerochia1o (Read error: 113 (No route to host))
07:50:54 Join Snake [0] (n=Snake@unaffiliated/snake)
07:51:18SnakeWhat do you guys think is the best settings for when you scroll on an ipod to emulate the apple firmware?
07:53:05 Part Snake
07:57:05***Saving seen data "./dancer.seen"
08:01:43 Join terrence1019 [0] (i=d15ed627@gateway/web/cgi-irc/
08:02:32terrence1019greetings, fellow rockboxers
08:03:59terrence1019can the Region name's be renamed or further expanded upon? I live in Trinidad and Tobago, but I have to use "Europe" as my setting :D
08:04:36Lloreanterrence1019: You do realize that people who don't have units with radios, or haven't used the function, would've had no idea what you're talking about, right?
08:04:50LloreanAs well, feature requests generally belong in the tracker.
08:05:00 Part toffe82
08:05:30terrence1019oh, sorry
08:05:52terrence1019I'm referring to the Sansa e200
08:07:06 Join sergey [0] (
08:08:51 Join ddalton [0] (
08:11:03ddaltoncould someone let me know if I have done the indentation right on this patch? ddalton/time_with_setting.diff">
08:12:57 Quit terrence1019 ("CGI:IRC (EOF)")
08:15:41scorcheddalton: your spacing is too needs to be set to 4 spaces
08:15:42GodEaterddalton: you've either got a bit carried away with the spaces, or your editor is inserting tabs instead of them
08:16:33scorcheit seems to be using 16 spaces instead of 4
08:16:39 Join Transience [0] (
08:16:52GodEaterddalton: also, your first if statement after the block of talk_value() functions shouldn't be indented at all, it should be at the same level as the talk_value() functions themselves
08:17:32 Join courtc [0] (
08:18:05GodEaterddalton: in fact, it looks like you've done that all the way down, every time you start an if statement, you indent the if itself, not just it's content
08:18:21ddaltonGodEater: since know one really is interested in this patch so it won't get commited should I just leave as it is? my editer says 20 levels for example. this I think is like 60 spaces. So it should only be on 20. example: if(x==y) { printf("hello"); the printf should be only 4 spaces in but my editor is putting it 16 spaces in. does that sound right?
08:18:56GodEaterthat sounds like what has happened yes
08:19:12 Quit midgey ()
08:19:22 Join pabs [0] (
08:19:37ddaltonso my first if should be 12 spaces in or not?
08:19:54 Join ToHellWithGA [0] (
08:20:04 Join Toki_ [0] (
08:20:07GodEaterddalton: it should line up with the talk_id() call
08:20:14GodEaterhowever many spaces that is
08:20:51 Join nerochiaro [0] (
08:20:55ddaltonOk I will probably look in to linux. (I have got some more info on the screenreader) and I will get a better text editor. what do you think?
08:21:13ddaltonthe voice_current time one?
08:21:13GodEaterddalton: if you're convinced that no-one else wants the patch, then it probably doesn't matter
08:21:21GodEaterddalton: yes
08:21:22ddaltonthe one that would voice "current time"
08:22:12GodEatertalk_id(VOICE_CURRENT_TIME, true); <−− that one
08:23:22ddaltonwell is anyone here interested in a 12 hour clock for the voice? also it can be changed to 24 hour. by going to the time menu (/settings/general settings/system/time/time format) and you can choose how you want the time displayed as well as voiced.
08:24:43ddaltonso for an if in another if how does that work the first if on the left the text inside that 4 spaces in and then another if inside that if 4 more spaces in?
08:24:58ddaltonhopefully you can understand what I am trying to say
08:25:14amiconnFor some reason the 'Rockbox Info' screen lost its ability to announce time
08:25:32amiconnIt already did that before it got reworked...
08:26:17ddaltonok go and try my patch let me just submit it. the indentation is wrong though so do you want to try that?
08:26:17amiconnIt always announced in 24 hour format iirc
08:26:55ddaltonI know I made a setting though so it can say it in 12 or 24 hour depending on the setting.
08:27:15*amiconn silly
08:27:47amiconnIf I do this on a target without RTC, I shouldn't expect the time to be announced...
08:28:00*amiconn obviously has too many targets
08:28:30ddaltonwith no rtc it won't say the time
08:28:49*amiconn isn't fully awake yet
08:29:53 Join Rob2222 [0] (
08:37:08 Quit n17ikh|Lappy ("Roses are red, violets, an honor. Sugar is sweet, WHERE IS SARAH CONNOR?")
08:39:16GodEaterddalton: I tidied up your indenting for you :
08:40:25 Join ender` [0] (
08:42:32 Quit JdGordon ("Konversation terminated!")
08:43:20 Quit Transience ("Leaving.")
08:44:44ddaltonGodEater: thank you very much for doing that. I will have a close look at what you did. also if I got a text editor that did the indentation for me would I have to worry about this?
08:44:56ddaltonalso how do you put code up on that site?
08:45:01GodEaterddalton: probably not ;)
08:45:11GodEateron pastebin ?
08:45:32GodEaterjust go to the front page
08:45:46GodEaterand paste the text you want to upload in into the text box on the front page
08:45:47ddaltonand do I need to sign up?
08:45:56GodEaterit's entirely anonymous
08:46:06ddaltonand how do get a blank page instead with or the links?
08:46:22*GodEater isn't sure he understoof the question
08:46:54ddaltonhow do you get a blank page with out all the links?
08:47:05GodEaterwhen reviewing the post you've done ?
08:47:56 Quit Rob222241 (Read error: 110 (Connection timed out))
08:49:26 Join haemmy [0] (n=stefan@
08:50:51ddaltonyes like your one I went to just then
08:51:21GodEateryou just insert "raw" into the link
08:51:39GodEater is with links
08:51:44GodEater is without them
08:52:13GodEaterit's the same post
08:52:17amiconnThere is also a link in the version with links that leads to the raw version
08:52:31GodEaterwhich may or may not be hard to find with a screen reader
08:52:32*GodEater isn't sure
08:52:49ddaltonok thanks I will look at that
08:53:09*GodEater notes that Universal have decided to go DRM from now until January
08:53:15GodEaterDRM free even
08:55:03*amiconn doesn't know of any text editor whose auto-indent feature gets everything correct
08:55:30GodEatertrue, but it will make ddalton's life much easier
08:55:36GodEaterif not perfect
08:56:05amiconnBraces in C are simple, but one-line indents like in single-statement if-else constructs aren't detected
08:56:13 Join webguest47 [0] (i=c023111d@gateway/web/cgi-irc/
08:56:56ddaltonyeah it would. I still can't work out how it helps. but I am blind so I wouldn't know. Anyway I better follow the rockbox coding rules
08:57:35GodEaterddalton: we appreciate your efforts, even if they make no sense to you at all :)
08:57:44ddaltonbut making small changes here and there would be fine. it is just when I need to do the whole thing.
08:58:35amiconnI think a proper auto-indent needs to know a lot of C rules. ConTEXT handles indenting/unindenting at braces, but neither does it know about single-line indents, nor does it know that it should skip preprocessor instructions when computing indentations
08:58:50GodEateramiconn: my text editor of choice handles single line indents just fine
08:59:17GodEateras well as proper braces rules and the ones for preprocessor instructinos
08:59:37amiconnWhich editor is that?
08:59:50GodEaterit's the KDE one
08:59:56amiconnyes I know
08:59:57GodEaterit also does code folding
09:00:14amiconnIt's the editor that sometimes likes to change text encodings by itself
09:00:24*GodEater hasn't noticed that =/
09:00:39amiconn(reminds me of several early JdGordon commits)
09:00:55markunperhaps it was fixed
09:01:31advcomp2019has anyone found any info on that post with the e200r yet
09:01:52GodEaterwhat *is* our source code encoding ?
09:01:56GodEaterdo we use utf-8 ?
09:01:58amiconnWell, so I still don't know an editor for windows that does proper indentation *and* is as comfortable as context
09:02:13amiconnGodEater: Source encoding is iso8859-1
09:02:35*GodEater is making sure he sets this explicitly in kate, rather than leaving it up to the "KDE default" whatever that is
09:02:49markunonly the manual, language files and credits use utf-8
09:03:03amiconnOn linux I much prefer gnome over KDE
09:03:10GodEateramiconn: in general so do I
09:03:22GodEaterbut Gnome's text editor falls far short of kate's abilities
09:03:25advcomp2019i like gnome too
09:03:34GodEatergedit < kate
09:04:43amiconnWell, gedit does syntax highlighting. Other than that, I just know that it's sloooow on loading, and has the nasty property that it jumps back to the start of the file when finished loading, even though you can scroll down before
09:05:33 Join petur [0] (n=petur@rockbox/developer/petur)
09:05:36amiconnLoading a few MB disassembly text can easily take 30 seconds - that's on an Athlon64 3800+
09:06:30GodEaterthat's very poor
09:11:21*sneakums tries out emacs on a disassembly of itself
09:11:29sneakumsgosh, that was surprisngly fast
09:11:50sneakumsalthough jumping to the end of the buffer the first time was slow
09:12:13sneakumsand the highlighting isn't too fancy
09:12:18sneakumsnot bad for a 750mhz powerpc though
09:12:48GodEateras with most things in emacs, the syntax highlighting is nearly infinitely configurable
09:12:56GodEaterprovided you LISP-Fu is strong
09:13:12sneakumsit's using whatever the maximum the default assembler mode supports
09:13:25sneakumswhich highlights the offset and the first byte of the opcode
09:13:44GodEaterdoesn't sound like much for a maximum
09:13:49GodEaterwonder what minimum does ;)
09:13:50sneakumsi guess emacs 22 was worth waiting for
09:13:56sneakumsnothing, i expect
09:14:28 Join B4gder [0] (
09:14:35sneakumsalso i'm not sure i fed it what it expected, i did objdump −−disassemble into a .s file
09:14:49 Quit sergey (Read error: 110 (Connection timed out))
09:15:26 Join sergey [0] (
09:16:25*GodEater doesn't believe a word of the most recent e200r post
09:16:29sneakumsit does a bit better when it's given real assembler
09:16:54B4gderGodEater: hell no
09:16:57ddaltonGodEater: what do you think of the patch?
09:17:04B4gderI guess it is one of them weird ones
09:17:11ddaltonmy time one
09:17:24B4gderGodEater: there was a guy here before who had a Rhapsody logo on his e280 without it being an R
09:17:28GodEaterddalton: other than I think you need to learn to use switch() instead of lots of if/else stuff - it looks fine ;)
09:17:46GodEaterB4gder: interesting - wonder how that happened
09:18:00*amiconn wonders what's up with the latest linux kernels
09:18:04B4gderyeah, his box even included a bonus/rebate thing for Rhapsody
09:18:11GodEateramiconn: in what way ?
09:18:13B4gderit smelled refurbished or similar
09:18:18amiconnNeither 2.6.21 nor 2.6.22 run on VMware workstation
09:18:57ddaltonGodEater: thanks for that. now I know what to look out for in the next c book i will be reading. is that all you think is wrong with it?
09:19:02amiconnThey just hang at boot, trying to access the scsi controller (after successfully discovering it)
09:19:18amiconn2.6.18 works like a charm
09:20:50 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
09:20:53sneakumslots of changes in the mpt driver since 2.6.18
09:21:05ddaltonGodEater: I ran your patch. It says "patch unexpectedly ends in middle of line " is that a problem?
09:21:08 Join My_Sic [0] (
09:21:20GodEaterddalton: probably - did it say which line ?
09:21:41 Join DerPapst [0] (
09:22:05DerPapstamiconn: have you seen this is the lcd test on a mini G1
09:22:08 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
09:22:09GodEaterah - that might be the lang file actually ddalton
09:22:23*DerPapst gotta run
09:22:27DerPapstwork ;)
09:22:27ddaltonGodEater: no it just says in main_menu.c
09:22:38 Quit DerPapst (Client Quit)
09:22:47amiconnDerPapst: Yes, I just had to scroll back in my client :)
09:23:25GodEaterddalton: have you put this on the patch tracker yet? I can only think that maybe pastebin munged it somehow - my local copy looks fine
09:23:41GodEaterif you haven't put it on the patch tracker, I'll upload my copy there and you can try it from there
09:24:10ddaltonwait let me convert it to unix style line endings. I just saved it as a text file.
09:24:20GodEaterwell don't upload it if it's broken
09:24:25GodEaterI'll just put my copy up
09:24:59ddaltonok p7561 and that didn't work.
09:25:01 Join miepchen^schlaf [0] (
09:25:10 Quit Jon-Kha_ ("Lost terminal")
09:26:16ddaltonit didn't even add the changes to main_menu.c
09:26:39ddaltoncan you upload it (p7561)
09:27:21GodEaterjust did
09:29:49ddaltonyep and it works
09:30:01GodEaterthe pastebin one must have got broken when I uploaded it
09:30:10 Nick _pill is now known as pill (
09:30:12ddaltondon't know what I was doing wrong
09:31:27ddaltonGodEater: thanks for doing that. so do you think if I want to be a good programmer I should be using linux? (mostly programming in c)
09:31:54GodEaterddalton: use whichever OS you feel most comfortable in - the OS doesn't make the programmer ;)
09:32:11GodEateramiconn is a real windows fan, and he's one of the best programmers here (IMHO)
09:32:41GodEaterhe just happens to be in a minority in this group, since most of us use linux
09:32:55amiconnI'm not a windows fan, but I use it most of the time
09:33:18GodEaterthat's really what I meant
09:33:18amiconnIt's just that windows supports several things which aren't supported (or supported well enough) in linux
09:33:30ddaltonwhat do most programmers use. I thought linux has all the compilers and stuff in it. I want to switch now so when I become a proper programmer I won't have to relearn my OS. because I wouldn't call my self a programmer. what do you think?
09:33:52 Join Siku [0] (
09:34:22amiconnThe disadvantage of developing for rockbox on windows is that cygwin is slow (which seems to be not cygwin's fault, but a problem of the windows file API)
09:34:31ddaltonwhat does windows have that linux doesn't. I would be just programming. maybe just looking at a couple of programming websites and receiving email.
09:34:54GodEaterddalton: I'm not sure there's a way to tell what "most" programmers use
09:35:00GodEaterit would depend on what they're programming
09:35:26amiconn3 things I know of so far: (1) windows supports all hardware in my laptop. Linux wouldn't. (2) windows has *stable* USB support
09:35:36GodEaterddalton: the only thing I can say is that if you're planning on doing a lot of work for rockbox, *most* of the programmers here use linux.
09:35:44amiconn(3) I found no equivalent for my preferred editor that runs on linux
09:35:46GodEaterbut that doesn't mean you have to
09:36:01ddaltonso are there good text editors in linux?
09:36:12Lloreanddalton: Depends on what you consider "good"
09:36:13GodEaterddalton: I'm sure you've asked that before :)
09:36:24LloreanFor example, gedit meets all my personal needs at this time, but I fully understand how it doesn't meet amiconn's needs
09:36:31GodEaterddalton: a programmers text editor is a *very* personal choice
09:36:34LloreanAnd ConTEXT is one hell of an editor.
09:36:45ddaltonlike that do the indentation and the unix style line endings.
09:36:50GodEaterddalton: there have long been flame wars on the internet between the fans of vi, and the fans of emacs for example
09:36:52Lloreanamiconn: For #3, is ConTEXT problematic with Wine, or is it general "I'd rather not have to use Wine"?
09:37:22ddaltonwould linux allow me to transfer my rockbox builds to my player?
09:37:25ddaltonand voice files
09:37:30GodEaterddalton: yes
09:37:30amiconnLlorean: ConTEXT runs in Wine, but my linux box is amd64, and I don't want to install 32bit compatibility
09:37:37LloreanMakes perfect sense
09:38:03LloreanIf Linux didn't support all my laptop's hardware, it'd be windows too I think
09:38:41GodEateramiconn: out of curiosity - what hardware in your laptop isn't supported by linux ?
09:38:51amiconnThe modem
09:39:02LinusNwinmodem, i assume?
09:39:04amiconn(at least, as I know linux doesn't support winmodems)
09:39:16ddaltonand will linux hook up to my network?
09:39:19amiconnMaybe other hardware too - I never really tried it
09:39:24LinusNi haven't used my modem for years
09:39:34GodEaterI don't even think I own one anymore
09:39:46GodEaterddalton: yes, linux is a very good network operating system
09:40:04ddaltonand can I modify linux like rockbox.
09:40:05amiconnI use the modem during summer, when there is no broadband connection
09:40:11GodEaterddalton: very much so
09:40:26*GodEater tries to work out where the broadband goes in summer
09:40:31ddaltonAnd would I have to be a better programmer than what I am now?
09:40:33LinusNGodEater: :-P
09:40:38amiconnGodEater: When I'm not at home
09:40:48 Join SliMM [0] (n=chatzill@
09:40:54amiconnThat was an unclear sentence, I know...
09:41:02LinusNi have a USB 3G modem, a real luxury
09:41:15GodEaterddalton: running linux isn't about being a better programmer, you *do* have to more tech savvy than your average windows users
09:41:17LinusNi even got a 3G wireless router
09:41:29LinusNso i can bring my wifi to the summer house
09:41:41amiconnWell, I am also pondering a 3G card, but that will cost money the whole year, even if I don't use it
09:41:42ddaltoni mean modifying linux by editing the source
09:41:55amiconnThe modem is definitely cheaper
09:42:15amiconn...and it's sufficient for SVN and chat
09:42:16GodEaterddalton: I'd be impressed if you could find something you think needs changing the first year you're running it that needs you to modify the source code of the linux kernel
09:42:22 Join atsea- [0] (i=ariel@gateway/tor/x-6a8749e4c8237def)
09:42:42B4gderGodEater: it would then possibly be drivers for hw
09:42:46GodEaterddalton: most things in linux are customisable without you having to modify source code
09:42:51ddaltonok and just one more question how much would a computer that could run linux cost me?
09:43:04GodEaterddalton: the same as one that could run windows, or less
09:43:27ddaltonso $150 off ebay?
09:43:32ddaltonor not
09:43:37ddaltonhaven't looked for a while
09:43:37GodEaterB4gder: I've not had a piece of hardware that didn't work first time on my linux machine for several years
09:43:45amiconnOh, and I forgot another important (for me) windows feature: RDP
09:43:47GodEaterI think you have to hunt hard to find weird hardware no
09:43:51B4gdersound? wireless? graphics...
09:44:00B4gderI have plenty unsupported hw in spanking new hw
09:44:03GodEaterB4gder: everything has worked out of the box for me
09:44:23B4gderlucky bastard! ;-)
09:44:37GodEaterddalton: you're in Australia yes? I've no idea on the exchange rate at the moment, so I've no idea what you could get for 150 AUD
09:44:39amiconnGodEater: I am still struggling with getting a piece of hardware (wifi) to work on another linux box
09:44:53amiconnAnd that hardware even has an open source driver (Ralink RT2500 chipset)
09:45:14GodEatermy wireless card has an aetheros chipset in it
09:45:17GodEaterso I use mad-wifi
09:45:18LinusNamiconn: i have a quite nice 3g deal - it has no monthly fee, but instead 0.2 euro per megabyte
09:45:37amiconnVerrry cheap for no monthly fee
09:45:42B4gdermadwifi is fine
09:45:46GodEaterit's actually more stable than my girlfriend's laptop
09:45:53GodEaterwhich is centrino / windows
09:45:57GodEaterthe signal on that is dreadful
09:46:01B4gderI have the intel chipset wifi in my laptop, which is even better than madwifi
09:46:04GodEaterand she sits closer to the router than me
09:46:31amiconnI can get a 3G card with 200MB/month included for 25EUR/month
09:46:34GodEater"intel chipset" ?
09:46:46amiconn...and every extra MB would still cost me 0.80EUR
09:46:51GodEaterdo you mean centrino ?
09:46:54GodEateror something else ?
09:46:57LloreanYeah, my laptop is intel wifi too, and works remarkably well under ubuntu.
09:47:01B4gderGodEater: hm yeah Intel 3945ABG I meant
09:47:06LloreanGodEater: For me, ipw3945
09:47:11B4gderthat's the one
09:47:23GodEaterI've never tried getting her laptop to run linux - she'd shoot me
09:47:29LloreanI was the only person at DevConW who got a completely consistent signal, I think.
09:47:32GodEaterso I've no idea what the driver would appear as in linux
09:48:06 Join Schwakito [0] (
09:48:06amiconnEven if the card is supported, wifi is a nightmare on linux imo
09:48:19Schwakitohey guys
09:48:20amiconnLots of fiddling with various config files until it works
09:48:22GodEaterit's not exactly "click and go" no
09:48:23B4gdernever been a nightmare for me once the drivers work
09:48:24LloreanGodEater: My laptop advertised it as Intel PRO Wireless 945, I think.
09:48:32amiconn...*if* it works at all
09:48:33kclafdamn i mixed up channels again, im on #openwrt now
09:48:38*kclaf runs...
09:48:46B4gderamiconn: now you're being hostile again... :-)
09:48:50amiconnThe problem I am having with the RT2500 card is that it doesn't want to connect using WPA2
09:48:51Lloreanamiconn: On ubuntu, on first boot, I just had to pick the SSID and type in the key.
09:48:54Schwakitosince 0804 the radio on my sansa doesn┬┤t work as it is supposed to :-(
09:49:09 Join PaulJam [0] (
09:49:19ddaltonso if I bought a win98 computer off ebay could I put linux on it and remove windows. what would it usually cost in america or where ever you are?
09:49:27amiconnIt worked with the old router, which was wpa1 only
09:49:32B4gderand the fact that linux has problems with wifi etc are the fault of the MANUFACTURERS, not anyone in linux
09:49:45GodEaterddalton: not very much money at all - probably about 100GBP, if that
09:49:55ddaltonI think my router said it supports linux but I could be wrong
09:49:58 Quit Schwakito (Client Quit)
09:50:08GodEaterddalton: your router will work fine I'm sure
09:50:24amiconnThe new router offers both wpa and wpa2 (both enabled), and the card doesn't connect at all
09:50:27GodEaterddalton: in general, a router doesn't care what OS the computers attached to it run, as long as they talk ethernet
09:50:35ddaltonis that $
09:50:53amiconnI wanted to try the new rt2x00 driver, but that requires kernel 2.6.22, which wasn't out at the time I tried
09:50:53GodEaterGBP = Pound Sterling, so no - not $
09:51:01 Join pondlife [0] (
09:51:10amiconnNow that debian unstable has 2.6.22, I'll retry
09:51:15ddaltonwhat is that in us $
09:51:25GodEaterddalton: 200 USD
09:51:31amiconnddalton: google can convert currencies
09:51:33GodEatergive or take a few cents
09:52:10pondlifeBagder: I could set up an automated build of Windows voice files. No super-good voices, just Microsoft Mary...
09:52:15ddaltonok but could I get a cheap computer like 64 mb ram 633 mhz and 10 gb hard drive off ebay. would that work? or what ever the specks are?
09:52:21*GodEater finds it ironic that though he works at an investment bank that does currency trading on a daily basis, finding out the current rates on his internal network is nigh on impossible
09:52:24amiconnJust type 'nnn GBP in USD' (also works for other currencies if you know the 3-letter currency name
09:52:35ddaltonpondlife how does 7560 work?
09:52:44B4gderpondlife: would be great!
09:52:52GodEaterddalton: you'd probably want more ram than 64
09:53:12pondlifeddalton: It's in SVN now, just open Cygwin and follow the "Linux" method for building voices.
09:53:33ddaltonon the wiki?
09:53:44ddaltono you just told me sorry
09:53:54pondlifeddalton: Preselect your SAPI5 voice and speed in Control Panel.
09:54:34pondlifeYou may need to do a make clean before you do a make voice to force new clip generation.
09:54:35*GodEater goes to do battle with websphere
09:54:37PaulJamwow, i think i fixed the problem with my h300. looks like a screw that somehow got loose was laying around in the player and causing some kind of short circuit. i really wonder how that screw could have come loose.
09:55:15LinusNi always knew you had a screw loose
09:55:17ddaltonpondlife: do I select normal build?
09:55:26pondlifeNo, Advanced
09:55:39pondlifeAh, the wiki may need updating there...
09:55:40ddaltonso a
09:55:52pondlifeThen V for voice
09:56:05ddaltonwhat about .talk clips can it do that? anyway how do you do that on linux?
09:56:06B4gderLinusN: now that hurt ;-)
09:56:14pondlifeB4gder: How (and when) would I transfer voices to you, to be vaguely synced with the daily build?
09:56:47pondlifeI've got very limited upstream bandwidth (128kbps)..
09:56:48B4gderpondlife: the best approach would be if you'd svn update to the daily build rev and use that
09:56:59B4gderthere's an updated file with that info
09:57:08***Saving seen data "./dancer.seen"
09:57:21pondlifeHmm, will need to write a sooper-dooper batch file.. :)
09:57:27B4gderif only I could remember where...
09:57:41 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
09:57:46pondlifeIt'll be wherever the screw ended up.
09:57:49B4gderpondlife: you could get my buildallvoices perl script if you want to base it on something
09:58:06LloreanWith very low upload, could you like... have the mp3s generated on the slow host, rsync to the fast host (so unchanged mp3s don't need copying at all) then merge them at the destination computer?
09:58:11pondlifeYes, ideally I'd like to do a Cygwin clone of the Linux scripts.
09:58:24LloreanKinda like a remote pool?
09:59:01pondlifeddalton: Is it working?
09:59:23ddaltonpondlife it says flite: Command not found. espeak: Command not found. festival: Command not found. TTS engine to use: (S)API5 (S)? graphic 370 what do I do?
09:59:50ddaltonthe graphic shouldn't be there sorry
10:00:00ddaltondo I need those synths?
10:00:16ddaltonno not yet
10:00:18pondlifeNo, you should have SAPI5... press S
10:01:17B4gderpondlife: this URL will help you sync with the dailies
10:01:43ddaltondo I need voicefont and the other stuff?
10:01:58ddaltonit is going through the c files
10:02:06pondlifeYes, but that should have been build already... did you make your tools?
10:02:14pondlifebuilt already, I mean.
10:02:19ddaltonwhat do you mean?
10:02:50B4gderddalton: you need to run 'make voice', did you?
10:03:09pondlifeB4gder: Does that build voicefont too? I hope so...
10:03:17B4gderi think it doess
10:03:41ddaltonso I need to get all the voice font stuff and then run ../tools/configure press 11 then a then s?
10:04:04ddaltonand then make voice
10:04:23pondlifeA, then V, then S
10:04:27 Join Thundercloud [0] (
10:04:42ddaltonok let me get the stuff
10:06:15 Join spiorf [0] (
10:06:23ddaltonwill I need to compile voicefont.c áor just get the exe version?
10:06:52LinusNddalton: it will be compiled automatically
10:07:03ddaltonso what one do I get?
10:07:40 Quit sergey (Remote closed the connection)
10:07:43pondlifeWhichever build you got from SVN...
10:07:52pondlifeSVN UP will give you the latest.
10:07:57 Join desowin [0] (n=desowin@
10:08:00pondlifesvn up, I mean..
10:08:04ddaltonI just check out svn but do I get voicefont.c á
10:09:16pondlifeIt's in rockbox/tools, along with the voice building scripts.
10:09:18ddaltonthen compile it with gcc?
10:09:30pondlifemake voice should compile it for you
10:09:42ddaltondo I need lame
10:10:06pondlifelame must be in your path
10:11:03pondlifeAnyone know how I can format a numbered list in the wiki to have sublists? i.e. 1) 2) a) b) c) 3) 4)...
10:11:05ddaltonlame isn't there
10:11:11pondlifeYou need to supply that
10:11:27ddaltonok and do I just put it in my build dir?
10:11:36pondlifeAnywhere in your path will do.
10:11:59ddaltonok thanks I will try it
10:12:16Lloreanpondlife: Just more spaces, isn't it?
10:12:25pondlifeddalton: I'd be interested to hear if it works for you.
10:12:32LloreanThree spaces for the first list, six for the second, etc.
10:13:36pondlifeLlorean: Almost, that gives me numbers for the sublist, not a) b) c)...
10:13:57LloreanUse a. instead of 1.
10:14:04pondlifeOf course!
10:14:37pondlifeOK, wiki better now
10:15:13 Join obo [0] (n=obo@rockbox/developer/obo)
10:15:55Lloreanpondlife: e) hit enter when asked for a language?
10:16:00ddaltonwhere will I find my english.voice file?
10:16:08pondlifeIn your build dir
10:16:18Lloreanpondlife: or, e) select a language, or hit enter for english
10:16:20pondlifeLlorean: Yep
10:16:31LloreanI mean, that step's missing
10:16:36LloreanThey're going to be one enter/entry short
10:17:14pondlifeNot any more :)
10:17:18LloreanGotcha. :)
10:17:27LloreanI wasn't sure if maybe the script reacted slightly differently in windows
10:17:43LloreanDetecting your current language, or something
10:17:54pondlifeHmm, sounds very Windows-y
10:18:06LloreanIt seems like something windows might do.
10:18:14LloreanSince you have to pick a language for your SAPI voice anyway, right?
10:18:31pondlifeMaybe the script shouldn't prompt for TTS engine if there's only one available?
10:18:49pondlifeJust like it does for encoder.
10:18:49Lloreanpondlife: I'm not so sure
10:19:13LloreanRight now it doesn't prompt for encoder not because there's only one, but just because we only support one on the players
10:19:33pondlifeHmm, but it still looks for oggenc/speexenc.
10:20:05B4gderwe blame rasher for that ;-)
10:20:38 Quit perrikwp ("ChatZilla [Firefox]")
10:20:50Lloreanpondlife: But I bet if they're found, it still won't prompt will it?
10:21:27pondlifeNo idea
10:21:27 Join n1s [0] (
10:23:36B4gder"Besides being useful directly, the Rockbox project is an excellent introduction to embedded system programming. If you would like to break into that field, but nobody will hire you without experience, here's your chance."
10:23:51pondlifeFrom where?
10:25:14B4gderbtw "This is one of the largest open-source teams in the world, and is in the top 2% of all project teams on Ohloh."
10:25:34pondlifeI don't see those quotes... :(
10:25:42Lloreanpondlife: It's the user review, for the first quote
10:25:59B4gderthe second is shown if you click the "very large team" link
10:26:00n1spondlife: quick talk.c question: It seems to me like talk_menu_disable can be checked while being uninitialized if talk_menus_enabled() is called before talk_init
10:26:42pondlifeHmm, talk_init() must be called before any other talk_ routines...
10:26:53LloreanB4gder: It says 87 developers have contributes. This is the total number of people who've ever had commit access?
10:27:24pondlifeI think the most useful sentence on there is "My only complaint is that the theme it installs with is an embarrassment. Anybody who fails to discover one of the alternative themes soon enough may give up in disgust."
10:27:29peturBagder: the complaint about the default theme is still a valid one... I can't remember what the end result of the last discussion we had on this was.
10:27:31B4gderLlorean: I would even guess that some more has had access but never committed, and possibly some of the 87 people are duplicates (people changed account names etc)
10:27:44pondlifeThe defaul theme competition died?
10:27:47B4gderI agree with that complaint
10:27:48n1spondlife: then I guess some of the recently introduced talking splashes should be reverted...
10:27:51Lloreanpondlife: Sadly, yes.
10:27:59Lloreanpondlife: Someone should make the iAmp change, then.
10:28:01pondlifen1s: Indeed
10:28:19pondlifeWhich ones can appear during init, before talk_init()?
10:28:20Lloreanpetur: I don't like iAmp, but at this point, it's better than what we've got.
10:28:33pondlifeMuch better.
10:28:36Lloreanpetur: I argued for waiting until the end of summer, and we're pretty much there without a usable result from it.
10:28:38peturB4gder: I already merged the duplicates I found
10:28:38B4gderalmost anything is an improvement ;-)
10:28:40*amiconn has to say that he doesn't like any of the shipped themes on 2nd gen
10:28:45B4gderpetur: ah, nice
10:28:52Lloreanamiconn: But "less bad" is still better.
10:28:54amiconnThat includes the default theme this time
10:29:19amiconn(because it has peakmeters, which are bad on PP5002)
10:29:19pondlifeThe rubbish default theme might be why we have so many nice custom themes...:)
10:29:34peturB4gder: of course, I don't know if ohloh took that in their count ;)
10:29:39ddaltonvoice font is not an exe file and it says lame isn't in my patth. where should I put it?
10:29:42pondlifeThe default should not have peakmeters... does iAmp?
10:29:43amiconnI didn't find any of the more graphical themes actually usable
10:30:09Lloreanpondlife: No.
10:30:21Lloreanpondlife: I think it needs a couple fixes though
10:30:24pondlifeddalton: You can put it into \Windows\system32 if you want an easy/nasty fix.
10:30:28n1spondlife: appears to be just one, line 488 in main.c
10:30:35Lloreanpondlife: The volume meter needs an image for >0db, I think
10:30:40pondlifen1s: Revert away then!
10:30:43amiconnpondlife: No it doesn't, but e.g. neither iAMp nor iCatcher or unicatcher show dis activity, and I dislike those coarse graphical battery symbols
10:30:55Lloreanpondlife: Oh, I think I meant iCatcher.
10:30:58LloreanI'm bad with names.
10:31:24Lloreanamiconn: The 'Ultima' theme from the default themes contest actually has both of those points.
10:31:41LloreanIn fact the only complaint *I* have against it (other than it not covering all targets yet I think) is that it doesn't show Play/Pause status
10:31:45n1spondlife: will do :-)
10:31:52amiconnAs long as there is no clever way for putting a hi-res graphical battery symbol into a wps, I'd rather prefer a theme to use the default status bar
10:32:02pondlifeHmm, iCatcher isn't on ...
10:32:07Lloreanamiconn: Why not give it something like the progress bar?
10:32:16LloreanThen you use two images, a frame, and a bar.
10:32:35Lloreanpondlife: It's provided with every build for every target (or nearly every) so there's no reason for it to be.
10:32:43pondlifeI did not know that!
10:32:49*pondlife fires up the sim
10:32:57LloreanIf it's the theme I'm thinking of, that is.
10:33:18pondlifeIt is.
10:33:32ddaltonwhy is it a nasty fix?
10:33:49peturpondlife: it isn't on that site because it is part of rockbox install maybe?
10:33:49amiconnAlso, both iCatcher and unicatcher don't seem to make use of the greylevels on greyscale targets
10:34:00Lloreanddalton: Not the "proper" way of doing things, from some peoples' perspectives. But it's fine.
10:34:17Lloreanamiconn: I thought iCatcher used the 2bpp?
10:34:33pondlifeddalton: Because it's a system directory... it'll work because it's guarenteed to be in your path.
10:34:37B4gderthemes can of course always get improved later on even if we start shipping them as default now
10:34:52LloreanB4gder: Or a new default swapped out again.
10:34:52B4gderit would even motivate people more to fix them
10:35:08ddaltonso is it bad?
10:35:18LloreanI really like the "Ultima" or whatnot from the contest, but I think iCatcher is the most "complete" for across targets.
10:35:20pondlifeddalton: Personally I made a \Program Files\Lame directory and put it in there, then edited my environment. If you know how to do that...
10:35:25pondlifeddalton: Not really, no
10:35:27Lloreanddalton: No, not really
10:35:42ddaltonso nothing bad will happen to my computer?
10:35:57ddaltonI think it is working!
10:36:11pondlifeIt takes quite a while...
10:36:19ddaltonhow long?
10:36:35pondlifeDepends very much on your system.
10:36:39pondlifeMight be 5 mins though
10:37:06ddalton1 gb ram 2.1ghz dule core and vista
10:37:10*pondlife attempts "make -j voice"
10:37:19 Join barrywardell [0] (
10:37:24ddaltondoes it work?
10:37:27pondlifeddalton: Sounds faster than my PC..
10:37:37ddaltonso how long does it take you
10:37:47pondlifeAbout 2-3 miins.
10:38:00*amiconn doesn't know about this 'Ultima' theme, nor any other outcome of the contest
10:38:01pondlifeThe current method isn't optimal.
10:38:02ddaltonand did make -j voice work?
10:38:12pondlifeI'll tell you in 2-3 mins :)
10:38:18ddaltonok then
10:38:28amiconnLlorean: iCatcher has no peakmeters, and that I don't like, except on targets where peakmeters are too expensive atm
10:38:29pondlifeIt doesn't seem to speed it up much
10:38:46amiconn(colour PP targets and the PP5002 targets)
10:39:04*petur wants vertical peakmeters :)
10:39:07pondlifeThat's an opinion thing... personally I don't like peakmeters anyway.
10:39:13Lloreanamiconn:, the little thing in the upper left is an HDD LED
10:39:15 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
10:39:28ddaltonis yours done already?
10:39:30pondlifeWe'll never get agreement on opinion things.
10:39:36pondlifeddalton: Nowhere near
10:39:47pondlifeAm on LANG_TRACK_DISPLAY
10:39:50ddaltonok mine is up to about 250 files
10:39:53ddaltonmp3 ones
10:40:25Lloreanpondlife: It took my computer 20 minutes to produce 5 builds and 5 voices (without using the pool)
10:40:43pondlifeLlorean: Yes, but Cygwin file i/o is super-sucky-slow.
10:40:45ddaltonmine is up to lang_status_bar
10:40:48LloreanAh, right, cygwin ouch
10:41:21pondlifeddalton: Do you use a screen reader?
10:41:29ddaltonI think I am going to investigate into linux. (cygwin takes for ever with everything)
10:41:34amiconnLlorean: Hmm....
10:41:36pondlifeMine finished!
10:41:37ddaltonyes do you?
10:41:50Lloreanamiconn: It's missing peak meters, but it combines "mostly basic" with "nice looking"
10:41:55pondlifeNo. I wonder if that's why it's so slow for you.
10:41:55LloreanIt's not done for all targets yet, unfortunately
10:42:22pondlifeddalton: Under Linux you're limited to the lower quality voices of festival/espeak/flite.
10:42:38pondlifeLower quality IMHO (and many others)
10:42:43ddaltonit shouldn't be I don't know what the screenreader has to do with. it. (they are both totally different voices)
10:43:15pondlifeYour PC is using CPU to read lots of the output, which must slow down generation.
10:43:28pondlifeI'm thinking that the script could display less info
10:43:42amiconnLlorean: Yes, not available for <160x128, and it uses negative display
10:43:44B4gderit talks A LOT
10:44:18pondlifeddalton: Once it's finished, try make clean, followed by make -j voice >nul
10:44:24ddaltonwell it is still going
10:44:40LinusNdamn - my sapi5 refuses to work - "This voice cannot be played. Bla bla..."
10:44:41pondlifeYou should be able to build in about 1 minute
10:45:04pondlifeLinusN: Did you go into Control Panel and set a voice/speed?
10:45:18LinusNthat's when the error message comes
10:45:18n1sLlorean: hah, that theme is very similar in layout to my private theme, which I made around the time backdrops were introduced :-)
10:45:31Lloreanamiconn: "Negative display"?
10:45:34amiconnHmm, seems it's just missing for ipod mini and archos
10:45:43amiconnLlorean: Bright text on dark background
10:45:46ddaltonI am using a voice made on linux it isn't to bad. what do people think?
10:45:50pondlifeLinusN: There's an installer for SAPI5 somewhere on MS site...
10:45:53Lloreanamiconn: Only on color targets though, right?
10:46:05LinusNpondlife: of course i have installed it
10:46:20pondlifeAh, ok... just thinking a reinstall may help.
10:46:30LinusNtried that 3 times now :-(
10:46:35pondlifeHmm, lovely
10:46:37B4gderddalton: most people think the ones made on Windows/SAPI sound better
10:46:53ddaltonare the voices on linux really bad though?
10:47:12pondlifeddalton: Just use the daily built ones and listen. ;)
10:47:24ddaltonok but what do you think
10:47:34pondlifeI think they're pretty bad.
10:47:37LinusNi think they sound like crap compared to the ms ones
10:47:56amiconnAnd ms sounds like crap compared to some commercial voices
10:48:05ddaltoncan someone take a look at p7561 and tell me what they think. is it a worth while patch?
10:48:18amiconn(which often can't be distributed, unfortunately)
10:48:56amiconnddalton: I think the 12 hour announcement can be implemented in a simpler way
10:48:57pondlife"Duplicate of #623325. " seems unlikely - especially after tracker clean week!
10:49:11ddaltonwell I didn't find the one I got to bad LinusN made it for me. so I guess it was on linux.
10:49:28pondlifeddalton: Not necessarily.
10:49:36ddaltonusing the switch statement?
10:49:40pondlifeHow's the build going?
10:49:47*Llorean really doesn't mind espeak, but doesn't depend on voice ever.
10:49:48LinusNddalton: that voice was made with festival on linux
10:49:52amiconnddalton: No, doing the 24->12 reduction in a simpler way
10:49:57*GodEater looks with dismay at his current emerge. When, did I include seamonkey?!?!?
10:50:06ddaltonBut I also wanted it to say for example 8 o'clock pm instead of 8 0 o'clock pm
10:50:31ddaltonok is that one of the ones you think are not very good?
10:50:41pondlifeddalton: That's just a bug by the sound of it (no pun intended). All voices would do that.
10:50:47ddaltonI just subtracted 12
10:50:54ddaltonwhat else can I do?
10:51:19pondlifeddalton: I'm losing track of your conversation.... did the make voice complete yet?
10:52:03ddaltonpondlife: no
10:52:15 Join Nick_Brackley [0] (i=cb57494b@gateway/web/cgi-irc/
10:52:16pondlifeI reckon a make voice >nul will be much faster
10:52:25pondlifeMuch much faster
10:52:28amiconnddalton: You check for the various times of the day, and then talk. I would just precalculate some things and the output the talk clips
10:52:42amiconnbool pm = false;
10:52:49ddaltonso have you got an example
10:53:06amiconnif (hour >= 12) { pm = true; hour -= 12; }
10:53:17amiconnif (hour == 0) hour = 12;
10:53:29amiconnAnd then just talk once
10:53:53 Quit Toki_ (Read error: 104 (Connection reset by peer))
10:54:05pondlifeShould the spoken time be added to the info screen? Or were you thinking of somewhere else?
10:54:05amiconntalk_value(hour, UNIT_INT, true);
10:54:12 Quit webguest47 ("CGI:IRC (Ping timeout)")
10:54:30ddaltonok well it works for me but I will deffinetly try to do that in my next patch. I didn't think anyone was interested in it because they prefer 24 hour time
10:54:33amiconnThe spoken time _is_ in the info screen already
10:54:45*pondlife didn't hear it.... will try again
10:54:45 Quit PaulJam (".")
10:54:47amiconnJust that it's 24 hour mode always
10:55:02ddaltonI know but this does it in 12 hour
10:55:21amiconnddalton: Of course you shouldn't modify the tm_hour entry in the time struct, but copy it before
10:55:53ddaltonin /settings/general settings/system/time/time format you can choose how you want it voiced
10:56:17Nick_Brackleypondlife: did the bug involving holding record for usb charging get resolved?
10:56:29pondlifeNot yet
10:56:44ddaltonyay... it is done
10:56:51Nick_Brackleypondlife: ok thanks
10:56:59*amiconn needs to do a full rebuild for all his targets, and then install a new voice
10:57:05ddaltonare all the mp3 files meant to stay there?
10:57:10pondlifeNick_Brackley: I think the conclusion was that MENU would be used, not REC...
10:57:26pondlifeddalton: OK, you should now have an english.voice
10:57:36Nick_Brackleypondlife: ok that seems like a good solution
10:57:39Lloreanddalton: Yes, they make it take less time next time.
10:57:47ddaltonpondlife yep i do but are the mp3 files are still there
10:58:08pondlifeThey will remain unless you do a make clean, to speed up rebuilds
10:58:32pondlifeHowever, as an experiment, could you do a make clean, then make voice >nul
10:58:55pondlifeThat will rebuild without showing any progress.
10:59:28ddaltonmake voice >nul?
10:59:52pondlifeIt should be quicker, but you'll only know it's done when you get a new prompt.
11:00:28pondlifeTakes about 2 mins on my DualCore AMD 4800+ thingummybob.
11:02:18 Quit Nick_Brackley ("CGI:IRC (EOF)")
11:02:54Lloreanpondlife: That's cygwin time, I hope
11:03:46amiconnpondlife: 2 minutes for sapi voice building on that hardware??
11:04:24LloreanHow fast is SAPI?
11:04:25amiconnMy old script managed to build one sapi voice in <1 minute, on Pentium M (1st generation, Banias) 1.5GHz, and laptop hdd
11:04:43B4gderyour old script was not in cygwin...
11:04:53LloreanIt took me 53 seconds on a 3600+ on my linux box.
11:04:53amiconnCygwin has nothing to do with that
11:05:10ddaltonpondlife I will now do what you wanted me to do. also the voice file works. and the only problem with p7561 I think when using it it doesn't say "a m" correctly
11:05:10B4gderI would guess cygwin eats everything that could run fast
11:05:32amiconnCygwin being slow for compiling rockbox has nothing to do with cygwin itself
11:05:41B4gdercygwin itself is slow too
11:05:44amiconnIt's the file api on windows being slow
11:05:49 Join davina [0] (
11:05:57amiconnCompiling creates tons of intermediate files
11:06:26peturamiconn: then why is compiling with devstudio so fast?
11:06:46amiconnpetur: I would think they do many things in RAM that gcc does on disk
11:06:57GodEateramiconn: does -pipe passed to gcc make any difference ?
11:07:00amiconnBut I really can't compare as I never used it for C
11:07:02GodEateror does that not work on cygwin ?
11:07:02pondlifeCygwin has a slow mapping to Win32 file API, I'd gues.
11:07:09peturmaybe cygwin should use a ramdisk
11:07:26pondlifeI suspect it would be slow still
11:07:35ddaltonis make voice >nul meant to speed it up?
11:07:37peturWin32 file api is fast, I've used it enough in the past
11:07:47pondlifeddalton: Yes, it shouldn't be reading anything out.
11:07:58GodEaterwriting anything surely
11:08:02amiconnpondlife: I tried compiling rockbox (well, the simulator) with microsoft SFU. Same slowness as cygwin
11:08:19GodEaterservices for unix
11:08:20amiconnServices for Unix
11:08:21LloreanCan voices be made jointly?
11:08:29*pondlife googled too slowly
11:08:41 Join kretender [0] (
11:08:43pondlifeHmm, maybe SFU sucks as badly as Cygwin?
11:09:04amiconnI would think microsoft knows how their APIs work?
11:09:19pondlifeVMWare must be using the Windows file API eventually, and that's faster than Cygwin.
11:09:22GodEateryou'd think so wouldn't you...
11:09:26ddaltonit isn't saying anything but it is not creating mp3 files. (I check the build dir) do I need to run configure?
11:09:41pondlifeShouldn't need to.
11:09:41amiconnpondlife: Yes, but VMware opens _one_ file, and then just works on that
11:09:48ddaltonnow it is working
11:10:08amiconnCompiling on cygwin opens thousands of tiny files
11:10:46ddaltonok so I will ask again otherwise I will close p7561 is anyone interested in this patch or should I just close it?
11:11:16amiconnMaybe it's a similar thing like installing a build on Ondio takes several minutes on windows, but a few seconds on linux (but corrupts the disk if you forget to unmount -urgh)
11:11:17ddaltonactually I don't even know if I can close tasks
11:11:39GodEaterddalton: you can't
11:11:41B4gderddalton: don't assume that the whole world of rockbox hackers sit here and responds to your questions on IRC...
11:11:55 Join misarm [0] (i=5546a921@gateway/web/cgi-irc/
11:12:05pondlifeddalton: Seems like a useful patch to me...
11:12:09pondlifeWhy close it?
11:12:16amiconnpondlife: Yes, with some more rework
11:12:28pondlifePersonally, I'd remove the o'clock bit and simplify
11:12:43amiconnAnd if annoucement was removed from 'Rockbox Info' for some unknown reason during a rework, it should be readded
11:12:58pondlifei.e. just change the hour value and add AM/PM.
11:13:06amiconnThe o'clock bit sounds like a nice idea to me
11:13:07ddaltonok then I won't and I can't so doesn't matter but it just seemed like everyone prefered the 24 hour format. I wasn't asking just the developers but everyone. least for pure announcement. Maybe not in the settings screen
11:13:39pondlifeddalton: It should use the configured format really.
11:13:53pondlifeamiconn: O'clock would be ok, but is odd with seconds.
11:14:14amiconnAnnouncement doesn't need to announce seconds imo
11:14:46B4gderthe seconds will have changed by the time you hear them anyway ;-)
11:14:50*amiconn wonders how long the full rebuild for my targets will take
11:14:54pondlifeIs there somewhere other than the info screen then?
11:15:12amiconnpondlife: The time setting screen itself
11:15:30amiconnI don't know whether that annouces the time as a whole though
11:16:14ddaltonok then I will do that after I finish my work and read a bit of my c book. apparently I don't know c good enough to code on rockbox. anyway that is what someone said.
11:16:17pondlifeThe spoken details should match the info screen better IMHO. i.e. say "Battery level 86 percent", not "86 percent battery level"..
11:17:02ddaltonpondlife what configured format
11:17:06pondlifeAnd pressing RIGHT (which refreshes the info) should also re-read.
11:17:27amiconnNumber talking is one thing we still don't have proper localisation for...
11:17:48ddaltonit does I think have you tried the patch?
11:17:49amiconnIt always uses english rules for digit ordering
11:18:05pondlifeddalton: Settings > General Settings > Time And Date > Time Format
11:18:23ddaltonso pondlife your happy for it to say "7 0 pm"
11:18:36 Join Juice^ [0] (n=Juice@
11:18:40pondlifeYes, if configured for 12 hour clock
11:18:50pondlifeOr 19 0 for 24 hour
11:19:01pondlifePersonally, I use 24 hour.
11:19:08ddaltonok and do I need to add an extra setting or just using the one time format that is already there?
11:19:21pondlifeUse the existing setting, definitely.
11:19:49ddaltonok then I will fix it up to morrow first I will get a better text editor cause this one is crap
11:19:55pondlifePlease can someone with a current build (and voice) try something simple for me?
11:20:03ddaltonyes what
11:20:16n1sjust fyi 19 00 would be correct for Swedish but as amiconn said this stuff isn't properly localized
11:20:21pondlifeGo into Settings > General Settings > System
11:20:37ddaltonpondlife then what
11:20:42pondlifeThe screen option is then "Start Screen", but the voice says what?
11:20:50pondlife"Traditional Chinese" here :)
11:21:22pondlifeAnd "Idle Poweroff" becomes "November" :)
11:21:38ddalton"start screen"
11:21:51pondlifeHmm, with your newly built english.voice?
11:22:04*pondlife suspects SAPI5 badness.
11:22:20ddaltonI have the fonts as well
11:22:27n1spondlife: I suspect so too, all that works fine for me
11:22:31pondlifeHmm, I'll rebuild from scratch.
11:22:47ddaltonok and if you are using a h300 get the fonts
11:23:01n1sand the battery level is spoken as "Battery level 54 percent" here and clock anounce works too :-)
11:23:44pondlifeThe clock announce works fine here, but it was putting the percentage first on battery level!
11:23:50pondlifeThis is on the H300 sim, btw
11:24:04*pondlife makes clean and makes
11:24:28amiconnYes, current time announcement works, but includes useless seconds
11:24:39amiconnJust tested on X5
11:24:50n1spondlife: btw FS #7554 was really strange, I had the same bug as the OP yesterday morning and it just went away, and my talk.c fix shouldn't have affected coldfire targets at all, it was ifdeffed for little endian targets...
11:24:52pondlifeProbably H300 vs H300 sim voice files
11:24:59 Join midgey [0] (
11:25:31n1spondlife: yes, the sim needs a different voice file, now possible to build by going to advanced and selecting first sim and then voice
11:25:47*n1s is gone
11:25:48amiconnWhy does the sim need a different voice?
11:26:14pondlifeBecause someone disagrees with me...
11:26:15n1samiconn: because of the features stuff, many things not defined for sims, this could of course be fixed
11:26:23pondlife...that the sim should match the target as close as possible.
11:26:44pixelmamaybe for stuff like the "usb connect setting"?
11:26:45misarmis here anyone, who can help me with pcm_record_data usage in plugin?
11:27:39pondlifeThe sim should be defined to have the same features as the target, then have empty stubs as required..
11:27:46 Join My_Sic [0] (
11:27:52*amiconn agrees with pondlife
11:28:10pondlifeI quite often build a voicefile and test it on the sim before copying to my H300.
11:28:15amiconnAnd sim-only things could be replaced with more appropriate methods
11:28:26amiconn(like the USB (Sim) menu item)
11:28:31pondlifeSim-only things don't need to speak either, IMHO
11:28:46amiconnIt could be replaced by a keypress like the Hold switch
11:29:08amiconnpondlife: No, but e.g. that menu item takes a language ID
11:29:19*pondlife dreams of removing the last #ifdef SIMULATOR
11:29:32amiconnOf course that could be resolved by hard-coding english
11:30:11*petur dreams of a HAL to fix pondlife's dream
11:30:32amiconnpetur: Unnecessary complexity, imo
11:30:48pondlifeYep, stubs will do for now. Sim is NOT a target.
11:30:56 Join webguest47 [0] (i=c023111d@gateway/web/cgi-irc/
11:31:32*petur stops dreaming and resumes work
11:31:43*pondlife too
11:32:03scorchepetur: I can't let you do that...
11:33:07scorcheI'm sorry petur, I can't let you do that...
11:33:22*petur dreams of a nice beer in 6 hours :)
11:34:29*GodEater looks at scorche's baleful red eye
11:34:56*petur just kicked it
11:35:42*scorche starts singing "Daisy Bell"
11:36:10 Join webguest09 [0] (i=534c7a5c@gateway/web/cgi-irc/
11:38:18*amiconn found another pronunciation flaw of festival
11:39:00 Quit webguest09 (Client Quit)
11:39:30scorchesuch a sad scene =( :
11:39:43 Nick Tino|Home is now known as TinoM (
11:39:56misarmPlease help me, what is wrong on this implementation of pcm_record_data? link -
11:41:43misarmCallback allways stops with status < 0
11:43:53*amiconn likes the size of the Player voice file compared to the others...
11:43:57ddaltonwell it worked for me
11:44:00ddaltonthe fonts
11:44:40pondlifeFonts are only used for the display, not the voice at all.
11:44:42 Quit pabs (Read error: 110 (Connection timed out))
11:46:06pondlifeHmm, HAL has quite a decent TTS.
11:46:32GodEaterbit intimidating and creepy though
11:46:50 Join solexx [0] (
11:46:53pondlifeSo is Microsoft Mary
11:46:53GodEaterwith a hint of "hello little girl, do you want to see some puppies?"
11:47:03 Part solexx
11:47:04pondlifeAh, that's Microsoft Mike
11:47:42 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
11:48:31pondlifeGodEater: If you have a Windows PC around, go into Control Panel > Speech, select Microsoft Sam, set speed to Slow-ish and preview with "hello little girl, do you want to see some puppies?"
11:48:46*GodEater goes to do this
11:48:48amiconnThe delete progress screen is broken, at least on recorder
11:48:51ddaltonhas anyone thought about commiting p6138? I just thought it was a good patch. (I didn't write it)
11:49:08pondlifeGodEater: About 3rd notch from the left on the speed slider...
11:49:29pondlifelol, Gurrrl.
11:49:36scorchepondlife: make the bad man go away
11:49:55GodEaterpondlife: roflmao
11:49:59ddaltonwho is the bad man?
11:50:21pondlifeWhy is there no Microsoft Bill (or Microsoft Steve)...?
11:50:26ddaltonwhat are you talking about?
11:50:56GodEaterpondlife: do you have "LH Michael" and "LH Michelle" too ?
11:50:56pondlifeI don't really know anymore. :( back to work...
11:51:15amiconnWhy is there a 'test' file in /.rockbox/rocks ?
11:51:26pondlifeNo, but I've heard LH Michael and he's a bit creepy, right?
11:51:46ddaltonI want to make the quick screen voice. Is this going to be hard?
11:52:02GodEaterpondlife: I think he's worse than Sam
11:52:11pondlifeWhere did you get him? Free?
11:52:25GodEaterdon't think so - he appears to be a standard on ABN's windows build
11:52:40pondlifeAh, your work
11:52:57markunbtw, espeak can also be used with mbrola voices, maybe some better sounding results can be achieved
11:53:02pondlifeNot Association of British Neurologists
11:53:17 Quit TiMiD[FD] ("leaving")
11:53:33pondlifeor Aramaic Broadcasting Network
11:53:46GodEaterpondlife: nope
11:53:57amiconndeutsch.lng is totally borked on player
11:54:20scorche [On Dave's return to the ship, after HAL has killed the rest of the crew]
11:54:20scorche HAL: Look Dave, I can see you're really upset about this. I honestly think you ought to sit down calmly, take a stress pill, and think things over. borked that I can't set it properly from within rockbox
11:54:50*scorche whistles and walks off
11:54:52Lloreanamiconn: Some people on the forums are saying the German language file is not working right
11:55:03webguest47amiconn: do you mean all players with 'player'?
11:55:07pondlifeamiconn: Is it crashing?
11:55:24 Join My_Sic [0] (
11:55:50amiconnI mean the Player with player
11:55:50pondlifeGodEater: Did you get anywhere tracing that voice-file-caused crash yesterday btw?
11:56:12GodEaterpondlife: nope, didn't have time - work :(
11:56:13amiconnpondlife: No, but is wrong strings all over
11:56:21pondlifeIf not, put the bad file up somewhere and I'll look?
11:56:30pondlifeGodEater: ^
11:56:40amiconnThe same applies to english
11:56:55pondlifeDo you still like the size of the Player's voice file?
11:57:10***Saving seen data "./dancer.seen"
11:57:15amiconnThe .lng is what is totally borked. Couldn't even try voice
11:57:34pondlifeAh, that *is * eurgh!
11:58:02amiconnBuilt-in english seems to be ok
11:58:51amiconn...and voice does not work with that
11:58:58 Quit midgey (Read error: 104 (Connection reset by peer))
12:01:14 Quit pixelma (
12:01:14 Quit advcomp2019 (
12:01:14 Quit midkay (
12:01:14 Quit iamben (
12:01:14 Quit lostlogic (
12:01:14 Quit maddler (
12:01:14 Quit YouCeyE (
12:01:14 Quit gxti (
12:01:14 Quit _jz (
12:01:14 Quit mtoups (
12:01:14 Quit dionoea (
12:01:17pondlifeGodEater: Doesn't crash now, just refuses to talk.
12:01:19*amiconn deems rockbox unusable on the Player atm
12:01:26GodEaterpondlife: well that's a bonus then
12:01:38Lloreanamiconn: Because of the .lng thing?
12:01:38 Join midgey_ [0] (
12:01:38NJoinpixelma [0] (i=pixelma@rockbox/staff/pixelma)
12:01:38NJoinadvcomp2019 [0] (n=advcomp2@
12:01:38NJoinmidkay [0] (n=midkay@rockbox/developer/midkay)
12:01:38NJoiniamben [0] (n=ben@
12:01:38 Join lostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
12:01:38NJoinmaddler [0] (
12:01:38NJoinYouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
12:01:38NJoin_jz [0] (
12:01:38NJoingxti [0] (n=gxti@foresight/developer/gxti)
12:01:38NJoindionoea [0] (
12:01:38NJoinmtoups [0] (n=mtoups@HENSON.ISR.CS.CMU.EDU)
12:01:45Lloreanamiconn: It's *possible* that it's broken everywhere
12:02:02Lloreanamiconn: The forum reports say that the german .lng is wrong, and then switching back to english is wrong.
12:02:11LloreanI think, for iPod Nano.
12:02:17pondlifeAh, so only built-in works?
12:02:17amiconnOn other targets I tried it mostly works
12:02:25Lloreanamiconn: "mostly" meaning?
12:02:50amiconnSome strings aren't translated which were translated before, but the ones that are are in the correct place
12:02:59amiconnTried iPod 2nd gen and X5 so far
12:03:02amiconn...and Recorder
12:03:44amiconnAmong the untranslated strings is one that I really don't understand - "Settings"
12:03:50Lloreanamiconn: Would that parallel voices voicing the wrong strings?
12:03:53amiconnIt's a generic string for all targets...
12:04:45amiconnLlorean: Yes, probably. But on other targets I tried it, it works (so far)...
12:05:12Lloreanamiconn: Well, I can confirm voices are wrong on the Nano.
12:05:18LloreanLet me see if I can find the thread.
12:05:30pondlifeamiconn's referring to lngs, not voices.
12:05:40pondlifei.e. display alone
12:05:41*amiconn wants to see that fixed asap, but has no time to fix it himself :(
12:05:55pondlifeWho else has a Player?
12:05:59amiconnpondlife: Probably also voice, but I can't even test voice on the Player
12:06:10ddaltonbe back soon
12:06:14 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
12:06:18LloreanSuggested there that the OndioSP is broken too
12:06:37amiconnOndio SP is still unzipping here
12:06:44pondlifeLlorean: lng or voice?
12:07:01amiconn(and I can't do an extensive test because the thing is suffering from the battery eating bug)
12:07:02Lloreanpondlife: That thread relates to .lng
12:07:31pondlifeOK, it's not totally clear
12:07:32Lloreanpondlife: As far as language goes, I think I'm the only person complaining so far.
12:07:54 Quit spiorf (Success)
12:08:29 Join spiorf [0] (
12:09:25Lloreanpondlife: Darn it. "as far as voice goes"
12:15:36 Quit My_Sic (Read error: 104 (Connection reset by peer))
12:16:07Juice^There should be a couple of predefined naming schemes for the recording filenames in the configuration screen to choose from
12:16:39GodEaterfeel free to add them
12:16:54Juice^i will try
12:17:03amiconnOn Ondio, deutsch.lng is also borked, but not as bad as on Player
12:17:25amiconnMain menu is ok, but sub-menus are mixed up, as is the context menu
12:17:38amiconnVoice speaks whatever the wrong .lng entry is
12:17:54amiconnSo I think there's something wrong with buildlang
12:18:13amiconnThat would explain why switching back to english makes the bug persist
12:18:14 Join My_Sic [0] (
12:18:16 Join advcomp2019- [0] (n=advcomp2@
12:18:33amiconn...because you're then switching to english.lng, which is different from built-in english
12:22:21LloreanWhich is what I assume is used for the english language voices, right?
12:22:45amiconnThe voice uses the same ID order as the .lng
12:22:56amiconnSo that explains why voice is broken on some targets
12:23:18amiconnIf you load english.lng, the menu should show the same wrong strings that the voice speaks
12:23:43amiconnI didn't check everywhere, and the main menu seems to be ok on ipods
12:23:59amiconn...unlike on Player, where even the main menu is all mixed up
12:24:02LloreanOn the Nano, the "System" menu shows it
12:24:24LloreanThe Nano definitely suffers from it
12:25:03amiconn'System' is correct e.g. on 2nd gen
12:25:24LloreanThe "System" option from the main menu?
12:25:27LloreanI realized I forgot to specify
12:26:16LloreanAre the lngs supposed to be the same, or does it have different features?
12:26:24*amiconn wonders what buildlang is doing wrong...
12:26:36amiconnThey differ per target
12:26:36 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
12:26:42amiconnEh, wait
12:27:26amiconnHmm, reloading english.lng on 2nd gen still has the 'System' menu item correct
12:28:38LloreanHm, all I did was reload the .lng, and the splash it gave me said "Unique Only" (which means nothing to me, so I'm thinking even that might've been wrong)
12:28:59pixelmaI think that's related to bookmarks
12:29:11Lloreanpixelma: Meaning it shouldn't splash when loading english.lng. ;)
12:29:25pixelmayeah, guess so ;)
12:29:40amiconnThe splash should show 'New Language'
12:32:46 Quit BigMac (Remote closed the connection)
12:33:16 Join My_Sic [0] (
12:34:56 Quit advcomp2019 (Read error: 110 (Connection timed out))
12:38:28 Quit barrywardell ()
12:40:39 Join ddalton [0] (
12:41:33ddaltonCould anyone sighted answer a question I have?
12:41:46GodEateronce we know what it is perhaps ;)
12:42:16ddaltonwell if you make a bookmark and then go into the bookmarks list what is displayed on the screen.
12:43:08ddaltonGodEater: thanks for doing that indentation for me I think I can understand how the indentation works.
12:44:44GodEaterddalton: if you had no previous bookmarks created, then the first thing you see when you enter the bookmarks menu again is an entry for "List Bookmarks"
12:45:07GodEaterand when you enter that, it will list all the bookmarks you have created.
12:45:57ddaltonGodEater: if you have a bookmark in the list is the name of the file displayed?
12:46:29ddaltonexample: if I bookmark song.mp3 can you tell it is song.mp3 in the bookmarks list?
12:47:12amiconnLlorean: On Ondio FM the main menu is also ok, but 'System' is totally mixed up
12:47:40 Quit midgey_ ()
12:47:41LloreanSo, LangV2 wasn't working quite right after all. =/
12:47:50GodEaterddalton: the menu displays the name of the file you bookmarks, plus a string showing you where in the file you bookmarked it
12:48:04amiconnOr rather, it works properly for the built-in strings, but not for loadable files
12:48:06 Quit desowin (Remote closed the connection)
12:48:23LloreanWhy would they be different?
12:48:32ddaltonGodEater so you could see it was song.mp3 if you bookmarked song.mp3
12:48:58amiconnI think it needs a perl expert here
12:49:04GodEaterddalton: yes, although the bookmark does not display the file extension
12:49:15GodEaterit displays everything up to the .mp3
12:49:20pondlifeamiconn: Does the problem show up in any sim?
12:49:29amiconnI don't know...
12:49:43amiconnStill busy rebuilding targets
12:49:44pixelmaI only ever tested the built-in - there was a bug before for creating the .lng files and that was only worked on in the last few days I believe
12:50:11GodEaterthis actually leads me to ask : "Why are bookmarks created in the parent folder, rather than the current folder?"
12:50:50LinusNbeats me
12:50:51pondlifeGodEater: I've long thought that odd.
12:51:04pondlifeThere are plenty of bookmark bugs by the way...
12:51:12GodEaterI just knew you were going to say that
12:51:13LinusNthere are?
12:51:49pondlifeIIRC, bookmarking and resume use 2 different mechanisms - needs KISSing better..
12:51:57ddaltonok then so it should be fairly easy to voice this?
12:52:25pondlifeLinusN: Yes, mainly regarding the use of long files, it appears. FS has plenty of info.
12:52:35GodEaterddalton: that's a whole different question!
12:53:04ddaltonI know but what do you think?
12:53:18pondlifeddalton: The bookmark stuff is probably harder than the info screen...
12:53:21GodEaterddalton: I've no idea - I've never done anything with the voice code
12:53:35LloreanGodEater: I thought they were in the current directory.
12:53:45LloreanGodEater: But I haven't used bookmarks since the H100 was the only non-Archos.
12:53:48ddaltonok I will try it and see what happens can someone give me a hint on where to look?
12:53:57GodEaterLlorean: well they're definitely not!
12:54:11LloreanI'm saying "Maybe they used to be, and it's not intentional."
12:54:14LloreanBut I dunno
12:54:20pondlifeIt may depend on what sort of playlist you're using :/
12:54:26amiconnOn mini G2 the system menu is ok
12:54:39pixelmaalso about bookmarks: @ 10.53.04 and following
12:55:09*GodEater looks at bookmark.c to see if he can work it out
12:59:43 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
12:59:53 Quit amiconn (Nick collision from services.)
12:59:55 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
13:00:01 Join ctaylorr [0] (
13:00:56amiconnVoice for the small H10 is missing from the download site...
13:03:13LinusNyes, it is commented in the build script, i don't know why
13:03:21B4gderI'm adding it
13:04:38*amiconn wonders what mixes up the IDs in .lng and .voice :(
13:05:11B4gdercan you find the first phrase that is bad?
13:05:18pondlifeID 256?
13:05:31 Join nick89 [0] (
13:05:43 Quit nick89 (Read error: 104 (Connection reset by peer))
13:05:50amiconnSeems to differ per target
13:06:10pondlifeHey, the lng files may borked, but we have a fully green build table! So must be a good time to upgrade, right? ;)
13:07:10amiconnStill building targets... :/
13:07:46pondlifeamiconn: What's an obviously broken string on the Player?
13:07:55*pondlife builds a Player sim
13:08:00 Quit safetydan ("Ex-Chat")
13:08:06amiconnAll of them, right from the main menu
13:08:20amiconnThe main menu becomes totally unusable as soon as you load a .lng
13:08:26pondlifeSo start up and load english.lng?
13:09:48amiconnThe main menu should read Recent Bookmarks, Files, Database, Resume Playback, Settings, Playlists, Plugins, System, Shut down
13:10:57amiconnWith english.lng loaded, it reads Rockbox, Recent Bookmarks, Files, Now Playing, Resume Playback, FM Radio, Playlists, Plugins, Move
13:11:15amiconn'FM Radio' makes no sense at all
13:11:15B4gderone off
13:11:28amiconnNot just one off
13:11:29pixelmaB4gder: is it still the "none" in the lang file, when it doesn't come first (I remember there being a bug like this)
13:11:40B4gderquite possibly
13:12:25ddaltonok I couldn't work out how to voice a splash screen so does anyone know how (gui_syncsplash I am talking about)
13:12:25pondlifeHmm, what happened to the Player sim keymappings? 6=down, 4=up, 8=right, 2=left !!
13:12:39amiconnpondlife: It's like the player buttons are laid out
13:12:48B4gderthe player is... special
13:12:51amiconnThe player has no left/right/up/down cross
13:13:14pondlifeThe wiki is wrong then:
13:13:21amiconnLeft & Right exist (they're called - and +), and hence need to be used for navigation
13:13:34amiconn'Up' us Play and 'Down' is Stop
13:13:35pixelmaB4gder: e.g in "id: LANG_CANCEL_WITH_ANY" the definition for "player: none" comes last
13:13:55*LinusN spots an interesting think in the generated makefile
13:14:01LinusNthing even
13:14:09pondlifeYep, the problem happens in the sim too.
13:14:27LinusNi think i see the problem
13:14:46LinusNor at least *a* problem
13:14:47B4gderI believe the phrases should be put in the order they should match, which means *: should be last
13:15:03LinusNnah, ignore me
13:15:26 Quit My_Sic (Read error: 104 (Connection reset by peer))
13:15:27B4gderbut I'm not 100% sure
13:16:36 Join jgarvey [0] (
13:18:17amiconnB4gder: Iirc it's the other way round, but I'm also not sure
13:18:38pixelmaI think that's it - the phrase I mentioned is quite early in the lang file - which would explain why the player's lang file is so messed up
13:18:38amiconnIf you put * last, it will be used for all targets
13:19:11pixelmalater there is the same with "flash storage: none" being last in a definitions - which would explain the problems on Ondio and Nano
13:19:11amiconnWhy does it put 'FM Radio' in the Player's .lng file?
13:20:11 Part misarm
13:20:11amiconnSo it doesn't ignore IDs marked 'none' for a target?
13:20:32LinusNthe Makefile fails to pass the correct features to
13:21:13 Join iamben_ [0] (
13:21:20B4gdersounds like a plausible reason
13:21:24pixelmathere was a related bug report for the langV2 patch...
13:21:59amiconnLinusN: Then that also applies to voice building
13:22:10pixelmaI only wonder why it works for the built-in
13:22:42LinusNamiconn: no, doesn't build voices
13:22:56B4gderthe built-in is not made with buildzip
13:23:11amiconnNo, but the voices show the exact same mixup as the .lng files
13:26:12LinusNhmmm, the binary generation uses $idmap{} but not the internal one
13:27:10ddaltonwhat does sizeof do?
13:27:37LinusNddalton: it tells you how many bytes a data structure occupies
13:28:22LinusNehum, the features are all wrong
13:28:38ddaltonLinusN: so do you know the bookmarks screen? what function would display that information on the screen?
13:28:56ddaltonthe text like time the bookmark is marked at and file name
13:29:23LinusNi dunno, look in bookmarks.c
13:31:00ddaltonok where abouts do you think do you know what would display the file name would gui_syncsplash?
13:33:30 Join desowin [0] (n=desowin@
13:34:11 Quit iamben (Read error: 110 (Connection timed out))
13:36:12 Join elinenbe [0] (
13:36:31ddaltoncan someone look on line 812 of bookmark.c and tell me if I am in the right spot to find how to voice the file name of the bookmark?
13:37:02 Quit spiorf (Read error: 110 (Connection timed out))
13:37:29 Join spiorf [0] (
13:37:56 Quit nerochiaro ("leaving")
13:38:37*LinusN found a features related bug
13:39:15ddaltonwhat variable holds the name of a bookmark in the bookmarks list?
13:43:21 Join nerochiaro [0] (
13:43:22 Quit SliMM ("Chatzilla 0.9.74 [Firefox]")
13:45:13LinusNregarding the binary language file problem, please test this patch:
13:53:17*amiconn just installed espeak on his linux box
13:53:33amiconnI have to say that it sounds significantly better than festival
13:54:09amiconnIts german voice is also okay
13:55:24*amiconn tries patch with player build
13:56:01*B4gder just reached 10000 commits on ohloh...
13:56:27*B4gder celebrates with coffee
13:57:13***Saving seen data "./dancer.seen"
13:57:29*scorche sets up a fence to keep the adoring fans at bay
13:58:54amiconnpondlife: Btw, the sim should print the button mappings to the console, and if you start it using −−background (which I always do), the background image also shows the button mappings
14:00:11pondlifeamiconn: Thanks
14:01:10 Quit Febs (Read error: 110 (Connection timed out))
14:01:40amiconnLinusN: Very strange.
14:02:01n1sLinusN: with your features fix the player sim seems ok to me
14:02:22amiconnIt doesn't
14:02:33amiconnThe fix just prevents _any_ .lng file from loading
14:02:44LinusNamiconn: yes, saw that too
14:02:51amiconnTry loading deutsch.lng - you'll still have an all-english menu
14:03:03B4gderhow can that fix prevent file loading?
14:03:03LinusNyes, the fix was lame
14:03:08amiconn...even though deutsch.lng does contain german strings (checked with a hex editor
14:03:53n1samiconn: ah, yes forgot to load lng, definately broken :-(
14:03:59LinusNignore that "fix", it was bad
14:04:38amiconnI guess now every ID is the same, and an unused one
14:05:00amiconn(or something like this)
14:05:02LinusNamiconn: id = 0, to be exact
14:05:13LinusNfor every string
14:05:31n1sLinusN: seems wrong :-)
14:05:44B4gderseems like a simpler approach! ;-)
14:06:00*amiconn goes to ditch VMware workstation, install the free VMware server, and make a non-graphical VM start in the background all the time
14:06:14amiconn...which offers samba and sshd
14:06:16n1sLinusN: strange IDs seem fine here...
14:06:45amiconnThen I have to move my build tree there
14:08:01 Quit petur (Read error: 104 (Connection reset by peer))
14:08:03 Join XavierGr [0] (
14:10:44 Join petur [0] (n=petur@rockbox/developer/petur)
14:12:13Lloreanamiconn: Is the samba/sshd VM posted anywhere, or is it not going to be released. I think that might be more usable for people because then they don't see that whole new window and expect to be able to interact with it in ways they can't.
14:13:11amiconnMy webspace has a rather limited download volume/month, so I won't post the link publicly
14:13:24amiconnIirc I also need to fix the keymap so it's english by default
14:13:25B4gderwe can put it on the download server
14:13:41amiconn(right now it's german)
14:14:28ddaltonWho here knows about the id3 viewer?
14:16:57pondlifeHaha, according to we have had no bug fixes since November 2006.
14:17:24B4gderno major ones at least
14:17:34pondlifeMaybe the lng/voice stuff will get a bug fix icon...
14:17:41pondlife(when fixed I mean)
14:18:04LinusNchanges aren't bugfixes, are they?
14:18:30pondlife"X5 LCD remote support" apparently is.
14:18:38B4gderand a bug fix should probably only be counted as a fix if it is for a bug that has lingered for a while, imho
14:18:41LloreanLinusN: It has an icon for bug fixes.
14:18:54pondlifeAll of the bug fixes are new features IMHO.
14:18:55LloreanI would say "3G Battery Life is good" is a bug fix though.
14:19:23B4gderthere's a grey area there, "improvement" vs "bug fix"
14:19:31n1sSomething is definetly wrong... the player sim .lng files have more IDs that there are strings in lang.h...
14:19:35pondlifeI'd say bug fix implies improvement.
14:19:38LloreanBut it's a "doing something wrong" case, vs a "improving something we were doing right"
14:19:58LloreanIt's a gray area
14:20:55rashern1s: What's the difference between none and "" in a language file?
14:21:51 Join chrisjs169 [0] (n=chrisjs@unaffiliated/chrisjs169)
14:22:56LinusNrasher: none is wrong, "" is correct
14:23:25amiconnnone means the ID isn't included for that target. "" means it is included as an empty string
14:23:48 Quit kretender (Remote closed the connection)
14:23:55pondlifeDev Fight!
14:26:12rasheramiconn: what would the point of having something included as an empty string be?
14:26:27B4gdernot break backwards compatibilty
14:26:35rasherHasn't that ship sailed anyway?
14:26:47B4gderwell for this bump yes
14:26:50amiconnRight now yes, but not for the future
14:27:34amiconnAnd since the system is currently broken, all "" could be replaced by none before or with the proper fix
14:27:51rasherThat's what I was thinking.
14:28:06rasherso I haven't gone completely mad
14:28:23B4gdersorry, but we don't have enough evidence to exclude that ;-)
14:32:30 Join bluebrother [0] (i=6iK80xwj@rockbox/staff/bluebrother)
14:32:54pixelmarasher: aren't the "" in the voice part only?
14:33:02 Quit ddalton ("I was using BOFHNet IRC version 1.2 by fmillion - get your copy today from !")
14:36:03rasherpixelma: I'm pretty sure I saw some in the dest part, but I could be wrong. Still, what's the point of including an empty voice clip?
14:36:24amiconnrasher: Not all language entries are voiced
14:36:28 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
14:36:47pixelmaI think it's necessary so that the voice clips are at the same position as the lang entry (at least my understanding)
14:37:18pixelmaif there are non voiced entries between
14:37:48pixelmaand same would apply to the "for voice only" phrases - the different way rouns
14:38:38amiconnVOICE_* ids have their own ID range
14:39:19amiconnBut xytarget: none in the <source> part should be enough to exclude the ID, no matter what's in <dest> and <voice>
14:39:29pixelmaah, ok
14:40:37n1sB4gder: unfortunately it seems there's another genlang bug...
14:40:47 Part ac_away ("Kopete 0.12.4 :")
14:43:55 Join ac_away [0] (n=austrian@
14:44:01 Nick ac_away is now known as austriancoder (n=austrian@
14:44:38austriancoderpetur: so.. I know now which registers are used in the orginal fw
14:46:02 Join barrywardell [0] (
14:49:14 Quit Juice^ ("Leaving")
14:53:14 Quit pill (Nick collision from services.)
14:53:34 Join pill [0] (
14:53:36pixelmabluebrother: hmm... I thought seeking and skipping are related, with both it's necessary to get to a certain position in the buffered data, no?
14:54:00n1sB4gder: anyway the problem is that when we have a target with 'none' after the default case, the string is excluded but there's a gap in the IDs for example LANG_CANCEL_WITH_ANY causes an ID gap for the player, the ordering of the cases is correct however, because we use the last match, so if the player case is above the default case the string is included...
14:55:02Lloreanpixelma: While I haven't looked at the code, seeking requires finding a location a certain amount of time in the data, which requires more calculation than jumping to the beginning or end.
14:55:11 Quit pill (Nick collision from services.)
14:55:34 Join _pill [0] (
14:55:50LloreanI'd think you should always be able to skip tracks
14:57:15pixelmaprobably, but skipping doesn't work with wma yet. I would expect once there is seeking, skipping would work too
14:57:42amiconnSkipping should work with wma
14:58:13amiconnIf it doesn't, it's a bug in playback.c
14:58:20LloreanSkipping shouldn't be codec dependent, I'd think.
14:58:30LloreanSince what you're skipping too isn't even necessarily the same codec.
14:58:41pixelmahmm... indeed. I would have sworn it didn't
14:59:33LloreanWell, it might not, but it'd be a bug rather than a 'not yet implemented'
14:59:46pixelma(it works, I mean)
15:00:43bluebrotherpixelma: I just tried, skipping a wma file works for me.
15:01:12bluebrotherthe following file wasn't wma in my test as I don't have much wma files. But I can skip to the next track
15:03:34pixelmait seems like you can't skip back when you are so far in a file that the first skip would take you to the beginning of the file
15:03:37 Join DerPapst [0] (
15:03:51pixelmaskipping forward always work
15:03:55DerPapstskipping from one wma track to another works for me here
15:04:06*DerPapst hast check back
15:05:01DerPapstyay.. data abourt at boot :D
15:05:02amiconnDerPapst: If you build a G3 bootloader from current svn and install that, it might fix contrast in the OF for you (until it goes into suspend)
15:06:00DerPapstwow.. something is messed up with my rockbox build on my G5.5
15:06:11DerPapstamiconn: will try when i'm home
15:06:38*bluebrother has some progress with themes installation
15:09:53*n1s wishes he understood perl...
15:12:18 Join My_Sic [0] (
15:17:50 Join sarixe [0] (
15:18:57peturaustriancoder: and?
15:19:18austriancoderpetur: only this one are used;
15:19:45B4gdern1s: can you write up a small case/scenario for me that repeats this?
15:20:20austriancoderpetur: and at the moment I play with IDA
15:20:48peturaustriancoder: so only one port, and no control register if I see this correctly
15:21:28austriancoderpetur: yep... and they use ep0, ep1 and ep2
15:21:40peturaustriancoder: maybe put the findings in a wikipage
15:22:29austriancoderpetur: will do, but need to go in a few minutes... with the help of IDA it should be now possible to get tx working
15:23:40Nico_PB4gder: why are you not to eager to see a release ?
15:23:59B4gderbecause I don't think a release is important to us
15:24:23B4gderI think we get along fine without
15:25:16Nico_PI tend to agree but some were saying that a release would help reduce the confusion for some users
15:25:19LloreanB4gder: I think *we* don't need a release, but Users do.
15:25:34LloreanSince we don't really have "Users" right now, but "Testers" or at least "Very daring individuals" ;)
15:25:39B4gderbut that's still why I don't burn for a release
15:26:16LloreanI don't think we need someone cracking a whip at this time, necessarily, but I think we need some target or destination hanging over everyone's heads, just to keep in mind, at the bare minimum.
15:27:46n1sB4gder: sure, I'll put it in the tracker
15:31:14 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
15:31:33 Join My_Sic [0] (
15:35:45pondlifeLlorean: Do you think our users want a release (as opposed to need one)?
15:36:04Lloreanpondlife: I think they both want *and* need one.
15:36:14LloreanMany people are reluctant to try Rockbox without a release.
15:36:41pondlifeIt seems that many users prefer to try the custom builds, which are even less controlled.
15:37:06pondlifeLlorean: More users -> more coders, so that's a good reason.
15:37:09LloreanYes, it really falls into two groups.
15:37:18LloreanYeah, but coders aren't likely to be the ones using only releases.
15:37:42LloreanI think we have "users who don't want things to change" which really boils down to people who try to find one build and stay with it, and maybe only upgrade when they see something *really* shiny
15:38:07LloreanFor those, releases are good because they can wait for the next release (if there's a semi-regular schedule) to get that new feature, and not worry about their music freezing up on a trip (usually)
15:38:18 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
15:38:46LloreanThen there's the "live on the edge" group, who are updating daily (or more often, or less if they have to wait on unsupported builds) and those won't benefit from a release, but also have a different expectation from Rockbox: Constant change.
15:39:13pondlifeI guess the majority of coders are in that second group...
15:39:17LloreanThey like the fast paced development, and post in forums somewhere (often not ours) when there hasn't been a commit for two days asking where everybody went
15:39:27LloreanI think coders fall into another group entirely
15:39:40LloreanI was grouping "users"
15:40:28Nico_Pyeah, I think most users just download a build and stick to it for quite a long time
15:40:30LloreanMy expectation is that coders often update less often than the "live-on-the-edge" group, updating only when they're testing something, since most of the time they're happy with their current features, or they'd be working on the one they're missing.
15:40:45Nico_Pand they don't say anything here or in the forums... I know two of those
15:40:50Nico_Pthree even
15:40:58B4gderI hardly ever update ;-)
15:41:03LloreanI know two people who only update if I tell them "Hey, this new feature came out you might like"
15:41:20pondlifeI guess I started out thinking "why is more users a good thing?" and came up with "because it might result in more coders"... I update weekly-ish.
15:41:24Nico_PB4gder: then you'd be happy with a release, wouldn't you ? :)
15:41:45Lloreanpondlife: I think more users in general has a lot of benefits and costs, and is kinda complicated in and of itself.
15:41:55B4gderfor me _personally_ a release means more work and more administrativa that will not gain me anything
15:42:00n1sB4gder: ok FS #7574 posted
15:42:06GodEaterI'd still love a way to measure the users we have that we never hear from
15:42:20B4gdera release is for the release-hungry user category
15:42:21Nico_PGodEater: I think we have a lot
15:42:37GodEaterI think we do
15:42:41Nico_Pthat's most likely the biggest category of users
15:42:56B4gdern1s: it is now in the top section of my todo-list
15:43:01 Quit desowin ("use linux")
15:43:24*Nico_P whispers "git" in B4gder's ear :)
15:43:24LloreanGodEater: Considering 10,000 registered members in the forums....
15:43:38LloreanNico_P: Even with git, there's administration that must be done.
15:43:39GodEaterLlorean: indeed
15:43:40B4gderNico_P: that too means more work
15:44:07B4gderdamn it, everything is more work! ;-)
15:44:13GodEaterB4gder: are there people you trust here who could help with the server admin ?
15:44:22Nico_PB4gder: don't worry, I've set up which is probably quite enough
15:44:41Nico_PB4gder: though a small script that would push to it on every commit would be welcome :)
15:45:39n1sB4gder: thanks
15:45:41LloreanI think that a "regular" release cycle on a timeline, no matter how small the changes between releases, would be better than trying to get a big release here and there.
15:45:57LloreanJust so you constantly have a "current" release version for users to download, and so they know the other version is the development one.
15:46:14LloreanAnd instead of basing releases on features added, or whatnot.
15:47:27pondlifeWe can always try it and improve the process as we go along (or drop the idea if no fixes occur...)
15:48:09LloreanOr we could always randomly just pick days where Rockbox seems to be working "well enough", call that a "release" and surprise everyone.
15:48:30pondlifeOr do retrospective releases... e.g. 5th Aug was a good one..
15:48:33LloreanConsidering how many people put up with the strange quirks in some of the unsupported builds, you could probably get away with it.
15:49:01pondlifeI don't see any major stability improvements on the horizon, no matter what policy we have.
15:49:08bluebrotherthe most people using unsupported builds do so because of the extra fancyness stuff
15:49:16bluebrothernot because it's moving slower ...
15:49:21pondlifeWell, on the horizon there's Nico_P's stuff, but in the nearer future.
15:49:50Lloreanbluebrother: I meant though, that the stability and consistency difference between "Rockbox on a good day" and "Unsupported builds" is at least somewhat significant.
15:49:51 Join finger11 [0] (i=47b29a40@gateway/web/cgi-irc/
15:50:29pondlifeIncorporating AA (and viewports) might kill off most custom builds..
15:50:34Nico_Ppondlife: btw, have you had time to look at my git log a bit ?
15:50:42pondlifeNo, sorry...
15:50:49pondlifeThat's twice I've failed.
15:50:57pondlifeWell, I read the log, but not the code
15:51:03LloreanI'm always surprised when reading custom build threads. How often they just drop features because they can't sync a patch, etc.
15:51:15Nico_Ppondlife: ok, no worries :)
15:51:23pondlifeLlorean: "There can be only one"
15:51:39n1sWhat do people think of removign the sim only USB entry in the system menu, we already have USB mode bound to the "u" key, it would be one reason less for sim specific lng/voice files...
15:51:59pondlifeMaybe someone should make a stable custom build - i.e. one that doesn't update from SVN.
15:52:11pondlifeThen that would be a "release"
15:52:13Lloreann1s: Or it could be moved to the unvoiced Debug?
15:52:33pondlifen1s: Remove it I say.
15:52:43pondlifeIf the sim becomes more like the target, that is good.
15:53:35*n1s agrees with pondlife and removes :-)
15:53:46*B4gder agrees too
15:54:04pondlifen1s: Oh, and while you're in there could you make the sim use the standard threading code ;)
15:54:14bluebrothermaybe the sim could print all keys upon startup to stdout?
15:54:19pondlifeThen we'll probably be able to repro and fix some of those race conditions.
15:54:22*n1s hides
15:54:23Lloreanpondlife: Clearly a trivial task
15:54:29B4gderbluebrother: it doesn't ?
15:54:37bluebrotherit does?
15:54:38B4gderit used to
15:54:39*bluebrother checks
15:55:57pondlifeToo many "We open the real file.." messages...
15:57:08*amiconn wonders how changing the threading code would help debugging race conditions
15:57:14***Saving seen data "./dancer.seen"
15:57:28*GodEater was wondering that too
15:57:31amiconnThe timing in the sim is always very different from target, regardless of threading code
15:58:37GodEaterI'm not even sure if the sims were true emulators it would help much
15:58:38pondlifeBut increasing the shared code might help.
15:58:52*GodEater makes "grasping at straws" motinos
15:59:11*GodEater gives up with the typing thing for the day
15:59:11*pondlife sends a bag of straw motinos to GodEater
15:59:32pondlife...I've removed the shit first...
16:00:06GodEaterhow will I know when it arrives without that clearly identifiable noise though ?
16:05:35bluebrotherhmm, the sim didn't output the keys.
16:05:55bluebrotherat least on linux
16:06:08B4gderI guess the x11 sim did that and it wasn't added in the sdl version, or something
16:06:26B4gderor possibly I've had too much coffee
16:06:30bluebrotherso −− anyone against adding it again?
16:06:47B4gderI find that very useful
16:06:56B4gdersince they're impossible to remember
16:07:07bluebrotherok, I take this as "no objections" ;-)
16:07:18LloreanYou've got at least three votes for it, too.
16:07:31*amiconn zips up the corrected debbox
16:07:44amiconnB4gder: Just use −−background
16:07:55*bluebrother is out for the weekend but might give it a try
16:08:02B4gderI hate^h^h^h^h don't like background
16:08:07amiconnimo that's better than the printed keymap, because it doesn't scroll away
16:08:09bluebrother−−background makes the sim much slower
16:08:19B4gderand much bigger
16:08:31amiconnThe background is loaded from disk..
16:08:44B4gderbigger as in screen estate
16:08:51amiconnah, that
16:09:01bluebrotheryep. But for some reason it eats up more cpu power. At least on my box
16:09:18peturmaybe just print them in the window next to the display?
16:09:54bluebrotheradd a "?" key that prints the keys to stdout.
16:10:14bluebrotherthat way you can always recall it, and we can have a consistent "help" button on all targets
16:10:37 Quit webguest47 ("CGI:IRC (Ping timeout)")
16:11:11 Join TiMiD[FD] [0] (n=TiMiD[
16:11:21amiconnThe gigabeat sim has the largest background, and even that uses just 2/3 of my screen height and a bit more than 1/4 of the screen width
16:11:51B4gderyes, for a totally useless graphics...
16:11:58amiconnAnd it can be cut smaller
16:12:21bluebrotherbut why not add some text output? You don't need to use it
16:12:29bluebrotherand binary size isn't an issue with the sim.
16:14:36B4gderan example from the old way
16:14:43*amiconn wonders whether SDL button reading obeys keymaps
16:15:41TiMiD[FD]amiconn: yes there are some bugs in the sim that prevent the use of some keys ...
16:16:22 Quit aliask ("ChatZilla [Firefox]")
16:16:35 Join Febs [0] (n=chatzill@
16:17:24 Part LinusN
16:17:53 Join lady_sara [0] (
16:17:56 Part lady_sara
16:23:41 Join My_Sic [0] (
16:28:27 Join desowin [0] (n=desowin@
16:29:29 Quit idnar (Nick collision from services.)
16:29:31 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
16:30:00 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
16:32:39 Quit B4gder ("It is time to say MOOO")
16:34:10 Quit DerPapst ("So Long And Thanks For All The Fish!")
16:34:19 Quit qwm (Read error: 104 (Connection reset by peer))
16:36:59 Quit obo ("bye")
16:40:13 Join saratoga [0] (i=98039a34@gateway/web/cgi-irc/
16:40:38 Join yuan [0] (n=yuan@
16:40:50 Quit ctaylorr (Read error: 110 (Connection timed out))
16:41:33 Join Aicart [0] (
16:42:24 Join BRBLbeta [0] (
16:43:50 Quit chrisjs169 (Remote closed the connection)
16:44:37yuanNow I want to back to rb with my h320,however, I am under the linux how can I patch the FirmWare´╝č
16:44:51 Join chrisjs169 [0] (
16:46:39bluebrothertry rbutil −− it can do the patching for you
16:46:48bluebrotheror run fwpatcher using wine
16:47:27yuanthank you
16:47:56*GodEater thanks that us calling it rbutil all the time, but the wiki page being called RockboxUtility is a recipe for user confusion
16:49:05 Quit finger11 ("CGI:IRC (EOF)")
16:49:35bluebrotherwe could create a RbUtil wiki page that redirects ;-)
16:49:47yuanIt is almost one year that I have not get the information of RB, dose the LCD fade out work?
16:50:10bluebrotheryou mean backlight fading? IIRC that worked a year ago ...
16:51:16yuanI back to the official firmware last summer
16:51:44yuanfrom then on I did not take much attention on the RB, hehe
16:52:30bluebrothercheck the MajorChanges wiki page if you haven't done
16:52:33yuansorry for my bad English, : )
16:52:35bluebrotherit's linked from the front page
16:52:49yuanI am searching
16:53:42bluebrotherand if you haven't found it already
16:54:27yuanokey, thanks a lot~
16:56:41 Join lazka [0] (
16:57:32yuanthere is no version of Rbutil for ubuntu edgy?
16:57:53bluebrotherguess no.
16:58:11 Quit petur ("drink error 666: connection reset by beer")
16:58:16bluebrotherbut you can simply build it yourself −− or try the newer in-development Qt4 version
16:58:29bluebrother(no binaries of that yet)
16:58:36 Quit Aicart (" /")
16:58:47yuanwell, I wil try~
17:02:44bluebrothersaratoga: have you noticed ?
17:04:05bluebrotherok, time for weekend
17:04:27 Quit bluebrother ("back after the weekend")
17:06:27 Join Infinoid [0] (
17:06:55*Infinoid has successfully stuck a 120GB drive into an old archos jukebox 6000 :)
17:07:47 Join BigMac [0] (
17:10:55 Quit yuan ("Leaving")
17:11:27 Part austriancoder ("Kopete 0.12.4 :")
17:14:48 Join Domonoky [0] (
17:22:47nerochiarosaratoga: hi, have a minute ?
17:30:37saratogabluebrother: thanks
17:30:41 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
17:30:42saratogasure whats up
17:30:49saratogai have about 10 minutes before a meeting
17:31:52 Part BRBLbeta
17:32:50 Join amiconn_ [0] (n=jens@
17:33:04 Quit amiconn (Nick collision from services.)
17:33:06 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
17:34:58nerochiarosaratoga: i have to go away soon too. i'm trying to understand the logic in the decode_frame function where it copies the data to the samples output buffer
17:40:08 Join XavierGr_ [0] (
17:40:09 Quit XavierGr (Nick collision from services.)
17:40:12amiconnHmm. Creating binary .lng files only includes the wanted strings, but ID mapping is wrong - it skips ID numbers
17:40:33amiconnIn the player .lng file, the first ID skipped is 0x12
17:40:57n1samiconn: yes I know, and reported in FS7574 Bagder said he will have a look at it
17:41:02amiconnLANG_CONFIRM_WITH_BUTTON is 0x11, LANG_ROCKBOX_TITLE is 0x13
17:41:17amiconnBut the latter should be 0x12, as there's nothing ibetween
17:41:47n1samiconn: yes, it doesn't handle the case where the default case is a string and some target has none
17:42:23amiconnBut only for binary languages - it works for lang.c/lang.h. Very odd.
17:42:29 Join webguest12 [0] (i=c27f0814@gateway/web/cgi-irc/
17:42:43saratoganerochiaro: I'm actually still toying with that code to cut down on the copying
17:42:58 Part Infinoid
17:43:02saratogabut basically, i think that deals with how blocks are overlapped
17:43:04 Join toffe82 [0] (
17:43:09webguest12Hello. Is there any statistics about how many tasks have been closed during the "FS fixing" week? Was it worth it?
17:43:19saratogaso you have to take have the previous block still in memory to decode the next one
17:43:27webguest12I'm asking just out of curiousity.
17:43:29n1samiconn: the same kind of problem have been present in genlang for other cases using none, but since lang.[ch] doesn't use the IDs but are just put in order they work fine...
17:43:35saratogai think it takes blocks off the bottom of that buffer and slides the next one down
17:43:46 Quit TiMiD[FD] (Read error: 110 (Connection timed out))
17:43:57nerochiarobut isn't that an output buffer ? i don't understand what it's sliding
17:44:00saratogathough to be honest, i still haven't gotten my optimized version working, so maybe i'm misunderstanding that
17:44:15saratogathe wma_window function writes blocks into that buffer
17:44:28saratogabut after a block is decoded, its still needed to decode the next block
17:44:59saratogaso its an output buffer, but things in it are used a good bit before they're actually passed out of the decoder
17:47:13nerochiaroso i'm better off allocating a dedicated buffer for that ?
17:47:19nerochiarohow big should that be ?
17:48:27saratogatheres a #define in the code for the size
17:48:53nerochiarook, i'll check it. got to go now, sorry
17:48:55saratogai believe it needs room for two stereo frames
17:49:15saratogaso i guess that would be 2 channels x 4096 samples x 4 bytes a sample
17:49:38saratogayou can see why i'd want to avoid having that in our decoder!
17:49:52saratogawe've only got 48k of IRAM
17:50:24saratogasorry, 2 stereo blocks
17:54:25 Quit webguest12 ("CGI:IRC")
17:54:46 Quit saratoga ("CGI:IRC")
17:56:23 Join ldarby [0] (
17:56:40 Part ldarby
17:57:17***Saving seen data "./dancer.seen"
18:16:10 Quit desowin (Excess Flood)
18:17:12amiconnBagder: ping...
18:18:47 Join My_Sic [0] (
18:25:23 Quit XavierGr_ (Read error: 104 (Connection reset by peer))
18:25:59 Join bdgraue [0] (
18:26:23 Join XavierGr [0] (
18:26:47 Quit haemmy ()
18:31:14 Join petur [0] (n=petur@rockbox/developer/petur)
18:37:00 Quit pondlife ("disconnected has pondlife")
19:04:47 Join pete [0] (
19:05:47petei got a question
19:06:57petei put the rockbox thing on my cowon X5L and have already put a wps on and stuff but now when i want to listen to a song it doesnt play i need to go forward a bit like 3 seconds and then it starts playing
19:07:07peteand for the next song its the same :(
19:07:56petei havent got the jeff rockbox thing i downloaded the latest normal rockbox
19:09:38FebsTerms like "rockbox thing," "wps and stuff" and "jeff rockbox thing" are not very useful to us when trying to diagnose a problem.
19:09:53FebsWhat Rockbox build version are you using?
19:10:01n1spete: you mean that playback stops on every track change?
19:10:16petena it doesnt start
19:10:28peteit only starts if i go further like 3 esconds
19:10:30petethen it runs
19:10:49petebut if the track changes its the same it wont start playing
19:11:16peteunless i fast forward like 3 seconds
19:11:36FebsWhat types of files are you trying to play?
19:12:35FebsDoes this happen for all MP3 files or just certain files? (still waiting for an answer as to which build you're using, by the way)
19:13:05peteah erm all mp3's
19:13:11peteand its the latest build 1 sec
19:13:13n1spete: is there anything special about those files, like embedded album art or other stuff?
19:13:34petena just normal mp3 files in a folder
19:14:02peteiAudio X5
19:15:31n1smp3 works just fine here, h300 and latest build...
19:15:52petelike when u click on it it starts?? and when u change song it starts too?
19:16:14n1spete: could you reset your settings, and try again?
19:18:16petei also wanted ask another thing: theres the Jeff's rockbox thing that is modified and u need to have it to have special wps' things but when i installed that it didnt show a database and thats bit gay bcoz i want a database and a special wps but i cant find the database :(
19:18:53n1spete: we can only help you with the official rockbox builds from
19:19:07petei do have the original rockbox on atm
19:19:14peteit was just b4 i had the jeff thing on
19:19:45petei reset it but its still not playing the song when i click on it it goes to the wps but it wont start only if i fast forward a bit
19:21:06pixelmawhat's the Jeff thing?
19:21:21n1spixelma: unsupported build I suppose
19:21:35peteu need to have it to b able to have kwl wps :/
19:22:18petescroll down to the zen pod wps or something
19:22:22pixelmamaybe there is a rockbox.iaudio in your root folder?
19:22:30peteand theres written "Get required build here for this theme to work. "
19:22:36pixelma(from the unofficial build)
19:23:13petewhat does a rockbox.iaudio do?
19:23:18Febspete, I thought you said you had tested this with r14270? If you haven't, please try the official build.
19:23:27petei do have
19:23:52n1spete: rockbox.iaudio is rockbox' main biary
19:24:12peteb4 i put the other one on bcoz i wanted to have another wps but then ive seen that there wasnt a database and so i put the official release on
19:25:22FebsDo the MP3s play in the original firmware?
19:25:29pixelmayou're hard to understand with this abbreviation
19:25:32 Quit sarixe ("Peace")
19:25:52petenow i put the jeff'S on again and the songs start
19:25:57peteso it works
19:26:03petebut i want a database like with an ipod
19:26:51FebsWas your previous problem occurring only with the database, or with the file browser as well?
19:27:21 Join salty-horse [0] (n=ori@pdpc/supporter/active/salty-horse)
19:27:36pixelmawhere do you have the "Jeff's" build from?
19:27:37petewith everything then 1 min ago i put the "unofficial" release on and it works now but i want a database so i need to use the "official" release or can someone of u tell me how to get the database?
19:28:09peteit was written on the website that i shud use that one
19:28:34n1spete: that places the rockbox.iaudio file in the root, you need to delete this if you install an official build
19:28:41pixelmaas I thought... there is a rockbox.iaudio outside the .rockbox folder in that zip...
19:29:35peteand why is it outside?
19:29:39pixelmacheck in your explorer and delete the one in the root of your drive (outside the .rockbox folder) and install the official build again
19:29:44salty-horseGodEater, here?
19:30:01n1spete: now maybe you understand why we don't support _unsupported_ builds...
19:30:11*n1s is gone
19:30:35peteyeee but i left the file outside the .rockbox folder and it worked perfectly.
19:30:43pete1 second ago
19:31:02pixelmayes, with the rest of Jeff's build
19:31:51petek can u tell me how i should get the ZenPod Vision M wps to work with the official release? here:
19:32:49petebecause if i use the "official" release and that wps it only shows some strange wps which isnt the right one but i tried jeffs thing with that wps and it worked. but i want a database :(
19:33:50 Join pabs [0] (
19:34:59 Join pondlife [0] (
19:37:16GodEater_salty-horse: here
19:37:56salty-horsehi. remember the discussion of tremolo from a few days ago? what's missing to push this forward?
19:38:04salty-horse(maybe you're not the one to ask)
19:38:07GodEater_I'm the wrong person to ask
19:38:22GodEater_as was pointed out at the time though
19:38:35GodEater_our codec already features arm optimisations
19:38:48GodEater_so it's unclear what benefits tremolo would bring (at least to me)
19:38:49salty-horsemaybe there's room for improvement
19:38:53peten1s , pixelma u know an answer?
19:38:58GodEater_maybe there is - but I really don't know
19:39:02GodEater_I'm not a codec author
19:39:20salty-horseIANACA :)
19:39:32GodEater_I don't even know who the original person who ported tremor to rockbox is sadly
19:39:55salty-horsesvn says markun committed it
19:40:06GodEater_I'd talk to him then ;)
19:40:49salty-horseany "seen" bot to use?
19:41:01pixelmapete: all the WPSs that include album art won't work with an official build, maybe you can find one in the unsupported build forum that is newer than Jeff's (from the date of the zip file, it's over a year old...)
19:41:08GodEater_salty-horse: well he's here already
19:41:17salty-horseoh, right
19:41:31petesigh why wont they work :/ whats so different with em?
19:41:44salty-horsei guess I had a typo in the nick to autocomplete
19:41:49GodEater_he reads the logs too
19:42:27peteit says: Get required build here for this theme to work. but that leads me to and i dno what to do there
19:42:30pixelmapete: but for that to work correctly you also have to delete the rockbox.iaudio in the root. And I'd strongly suggest you try with an official one first to see if your playback issues are solved
19:42:41salty-horsewhy is the mandelbrot plugin classified as "game" instead of "demo"?
19:42:48petethey are
19:42:57GodEater_salty-horse: jdgordon got it wrong probably ;)
19:42:57petei got the official on again
19:42:58peteand it works
19:43:57 Join perrikwp [0] (n=chatzill@
19:44:01peteso basicly all i want is a database and the zen vision M wps or the Black Glass 2 theme
19:44:30pixelmaok - is the only thing I can help you with - I don't have an X5 and don't care about unofficial builds...
19:44:50petek ty
19:45:05pondlife18 new H340s available here:
19:45:49petethen ill try that unofficial build
19:47:51 Quit pondlife ("disconnected has pondlife")
19:48:42amiconnn1s: You were lucky that this commit actually saved binary size for archos
19:49:11n1samiconn: I actually made a test build, but please explain :)
19:49:49amiconnYes, 16 bits are enough, but only if you use unsigned short. But when the SH1 cpu loads short or char, it *always* sign-extends, so loading an unsigned short or unsigned char requires an additional instruction
19:50:26n1samiconn: aha
19:50:59amiconnSo you only save binary size if the values are many and the places where they're read are few
19:51:08amiconn(ideally just one)
19:51:56amiconnBtw, another way to save binary size is to optimise struct sizes which are used in arrays to be a power of 2
19:52:16amiconnThis simplifies the pointer calculation
19:52:47amiconnI observed this when fiddling with the solitaire plugin
19:52:51n1samiconn: you mean arrays of structs?
19:53:34n1samiconn: ah, good to know, there's a lot of impact when changing structs...
19:54:16amiconnSolitaire used bitfields for all its struct member, which was quite unefficient
19:54:17*n1s remembers the 1kbyte increase from removing a few members from some metadata struct...
19:54:42amiconnThen I changed the struct to use no bitfields, which made it 5 bytes long
19:55:33amiconnAfter one further change (packing 2 'bool' members into one byte by using bitfields for them) the struct became 4 bytes long - and the plugin several hundred bytes smaller
19:55:48 Quit barrywardell (Remote closed the connection)
19:56:31n1samiconn: sounds like that's something that's worth looking into for the core then... gtg now tho, laters
19:56:48 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
19:57:18***Saving seen data "./dancer.seen"
19:57:29peten1s the theme didnt work with the ColdSphinX csBuild [r13969]
19:57:37peteso far only with the jeffs build
19:57:46petebut i want a damn database :@
19:59:52 Join polluxx2006 [0] (
20:00:51 Join Robin0800 [0] (
20:01:26polluxx2006hello rockbox-devs. after i opened a task in the feature requests i was told that i should discusss my idea here in irc....
20:01:41rasherwhat's the task number?
20:01:47polluxx2006let me see...
20:03:25Domonokyeverybody wants something else for the rec button :-)
20:03:45polluxx2006hmm.. i see. may i explain my thoughts?
20:04:04Domonokysure, but i dont know if it helps :-)
20:05:13polluxx2006its worth the try anyway... ;) i mean.. it doesnt have to be the rec-button. but i really miss a "one-click-after-boot-solution" to browse through the progiles (cfg's)
20:05:22polluxx2006for example:
20:06:01polluxx2006i use my headphones - i choose my setting with eq-things for that headphone
20:06:02Domonokyi have a solution for this.. put the cfg in the root.. then you can browse them with one click :-)
20:06:13amiconnHmm. I'm a little confused:
20:06:47polluxx2006but not if i chose to display only supported files, right?
20:06:52amiconnIf 'Backlight' is set to 'Off', but 'Backlight on hold' is set to 'On', what should happen if I enable hold?
20:07:08rasherI'd expect it to turn on
20:07:14rashersince the latter setting is the most specific one
20:07:27amiconnFrom reading the code I would think it should turn on, but from my own logic I wouldn't expect it to
20:07:37amiconnAnd in practise it doesn't turn on
20:08:04rasherI read it as "I generally want it turned off, but specifically when hold is enabled, I want it on"
20:08:32amiconnI wouldn't expect it to, but it seems to be what the author intended
20:08:40Domonokyyou have more possibilitys if it turns on.. then you can have the blacklight off, and turn it on with the hold button :-)
20:08:41amiconnAnd it doesn't work like this...
20:09:06amiconnI am looking into this because I need to understand the backlight thread in order to extend it
20:09:57 Join Genre9mp3 [0] (n=yngwiejo@rockbox/contributor/Genre9mp3)
20:10:00rasherI don't know, I guess it could go either way, but that's how I'd personally expect it to behave. You could of course flip it around and say that if you set "Backlight off", you're expecting it to never come on in any case.
20:10:24rasherPicking one and documenting it in the manual (under the Backlight on hold setting) should be fine
20:10:41amiconnYes, but that's also not entirely true, as we also have 'Backlight when plugged' (except on Ondio of course)
20:10:56 Join saratoga [0] (i=98039a34@gateway/web/cgi-irc/
20:11:05amiconnIt's just a matter of which setting has priority over the other
20:11:31saratogai'm a little confused about how i'm supposed to seek backwards in a file
20:11:45saratogatheres a buffer advance function, but not one that can go backwards
20:11:48saratogaam i missing something
20:12:07*Domonoky would like to keep this settings independent: so blacklight off only means no blacklight on button inputs..
20:12:28amiconnBut maybe it would be more logical to give 'Backlight on hold' priority, since it should also have priority when it's set to off
20:13:03amiconnInverting priority would be confusing, as it would give priority to a state instead to one of the settings
20:15:00Domonokyhm.. its difficult to unify all those backlight settings..
20:15:13amiconnUnify in what way?
20:15:33petecan someone tell me how i can get the X5 themes in the WPS gallery to work with the lates build?
20:15:38Domonokyso that its logical how they behave..
20:16:10amiconnBtw, I am looking into this because I need to be able to set 2 new states, which take priority over any setting: always on (for shutdown, so backlight goes off last), and always off (for suspend mode, so backlight goes off regardless of setting)
20:16:34amiconnDomonoky: That's what I think it should be like now:
20:16:50amiconn- Normally, the backlight timeout is used
20:16:53nerochiarosaratoga: i've just got back. sorry i had to run. before were you saying that it's better to use a dedicated output buffer since it's also needed for further decoding ?
20:17:06amiconn- This changes when plugged, to the backlight when plugged timeout
20:17:25polluxx2006domonoky: not the nicest way but putting the cfg's into the root-dir is okay ;)
20:17:30 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
20:17:32amiconn- Backlight on hold overrides both
20:18:10amiconnAnd that can be implemented in a simpler way than it is now
20:18:23saratoganerochiaro: i'm working on merging the mdct_ctx and frame_out buffers in order to save memory
20:18:27saratogathis may not matter to you
20:18:41saratogafor us, it may be substantially faster
20:18:43nerochiarono but memory saving is always good
20:18:45Domonokyperhaps we could have a setting with: blacklight : on / normal / off and the settings for the different situations, which are only aktiv in normal mode
20:18:55saratogaparticularly on coldfire
20:19:24amiconnsaratoga: Perhaps (hopefully) also on PP5002
20:19:36saratogai don't know what to think of the pp5002
20:19:47nerochiarosaratoga: also one other thing that i suggest, is that for input we find a way to pass a chain of packet payload addresses, instead of a single continguous buffer with all payloads for the packet, which seems to be needed right now
20:19:56saratogasometimes i put data in IRAM and things get slower
20:20:07amiconnDomonoky: This would be less flexible than now, because of 'backlight on hold'
20:20:10pixelmapete: they probably have small errors in the code which were tolerated by older builds, so you'd have to fix those...
20:20:11nerochiarosaratoga: generally there always seem to be 1 payload per frame, but i suppose there may be more ?
20:20:32saratogayou mean 1 payload per packet?
20:20:39 Join Cardini [0] (i=189fceb0@gateway/web/cgi-irc/
20:20:39amiconnOf course 'backlight on hold' could also have a timer setting instead of just on/off/normal, but that might further confuse users
20:21:04nerochiarosaratoga: so far i've only seen 1 payload per packet, yes
20:21:16saratogafor wma i think that will be typical
20:21:23saratogasince the encoder ca nalso choose the packet size
20:21:33saratogai don't see why the encoder would choose a poor packet size
20:21:41saratogafor files with video it won't always be the case
20:21:43amiconnEspecially since I also want a 'reduced backlight' option on targets which can set backlight brightness, but in too few levels to implement fading
20:21:49amiconn(e.g. H300 and X5)
20:22:00polluxx2006folks, could anyone take a look at my wps (iriver h140)? it worked before but not with the latest build :(
20:22:21*Domonoky understands :-) hm..
20:22:29amiconnThen I would set reduced backlight to minimum brightness, always on, and keep normal backlight like it is now
20:22:38nerochiarosaratoga: indeed, and also for audio if it's legit to have more payloads per frame, it should be possible to handle them more efficently than copying into a big sequential buffer
20:23:00amiconnThis wouldn't need much extra power, but keep the display always readable indoors
20:23:10saratoganerochiaro: thats true
20:23:21 Quit pete ()
20:23:26saratogathough to be honest i don't really understand that part of the code
20:23:45CardiniI have a few questions. I run yesterday's build of RockBox on a Sansa e270. First question: my menus are messed up. "Plugins" contains such submenu items as "Gain Left" and "Gain Right." Each of these contains plugins. The label attached to the screen icon is <All Tracks>, and inside I may find "Display," which is really the font list. It goes on and on.
20:24:00amiconnsaratoga: linuxstb should return these days (don't know extactly, some time between today and Sunday)
20:24:02CardiniIs this a ROckBox problem or should I.....delete rockbox and reinstall?
20:24:15saratogacan we be sure that each block will be entirely contained within a payload?
20:24:40amiconnCardini: It's a known problem, but we don't have a solution yet
20:24:48saratogabecause as long as blocks are only in 1 payload i think its reasonible to do, otherwise we'd have to deal with making the Huffman code aware of payloads, which would be a nightmare
20:24:52amiconnWatch the frontpage for language related commits
20:24:57saratogaamiconn: thanks, i have a list of questions for him
20:25:12 Quit petur ("here today, gone tomorrow")
20:25:34CardiniAmiconn: So...there's nothing I can or should do?
20:25:42polluxx2006could anyone help me with a wps for my iriver h140?
20:26:18nerochiarosaratoga: well, i'm not sure about that
20:26:27 Quit toffe82 (Read error: 110 (Connection timed out))
20:26:33nerochiarosaratoga: not really familiar with WMA format
20:26:41amiconnCardini: You could downgrade to a build from before the langv2 commit for now
20:27:25amiconnDaily build from Aug 5 should be ok
20:27:49amiconnBut then you also don't have plugin categories. Just one big list, like it was before
20:29:18polluxx2006no wps-help then? ok.... :(
20:30:11Domonokypolluxx2006: dont ask to ask .. tell us the problem/error..
20:31:46 Join Ste- [0] (i=56289e82@gateway/web/cgi-irc/
20:31:56polluxx2006well... i have to show the wps-code. hows the best way?
20:31:59CardiniOK, thank you, amiconn.
20:32:55 Quit Ste- (Client Quit)
20:33:23polluxx2006domonoky: i have to show the wps-code. hows the best way?
20:33:34Domonokypolluxx2006: so its a wps you made yourself ?
20:34:11Domonokyif you want to show more then on line.. use or something like it..
20:34:50*Domonoky misses chars :-) ?
20:35:20polluxx2006pastebin? whats that and how does it work?
20:35:34polluxx2006i mean the name says what i does - but how do i use it?
20:36:55Domonokyonly rockbox help here, use google or try it..
20:37:10 Join ptw419 [0] (
20:37:27nerochiarosaratoga: wma_decode_superframe_frame return the number of bytes it has witten into the output buffer, starting from the beginning of it ?
20:38:44 Join webguest79 [0] (i=4c675aa5@gateway/web/cgi-irc/
20:39:05saratoganerochiaro: it looks like it just returns the length of the current frame within the superframe
20:39:09 Quit webguest79 (Client Quit)
20:39:36saratogaso that should be the number of samples decoded
20:44:03 Quit Cardini ("CGI:IRC (EOF)")
20:44:20Domonokypolluxx2006: Failed parsing on line 43 : Invalid conditional char (not in an open conditional )
20:46:13nerochiarosaratoga: well, i need to know how much data was decoded into the output buffer and at what position, since i need to copy that to the soundcard output
20:46:33nerochiarosaratoga: unless i'm missing something
20:49:59polluxx2006domonoky: i tokk a look at it with the wps-manual as a reference but couldt find an error :(
20:51:51ptw419anyone have any idea why the mcu wouldn't acknowledge an irq request even when interrupts are enabled?
20:52:38saratoganerochiaro: it returns the number of samples, so just multiply by the sample size and number of channels
20:52:45 Join perrikwp_ [0] (n=chatzill@
20:54:22 Join My_Sic [0] (
20:55:12nerochiarosaratoga: it's always from the start of the output buffer anyway, right ? the decoder handles that internally
20:56:17saratogai'm not sure
20:56:19saratogai'll take a look
20:56:25Domonokypolluxx2006: there are two | in line 43 which are not inside a condittional, i dont know if they are allowd outside..
20:56:40nerochiarosaratoga: thanks
20:57:42 Quit perrikwp (Read error: 113 (No route to host))
20:57:49Domonokypolluxx2006: if you want to display a | you have to write %| .. :-) i just check the CustomWPS wikipage
20:57:51 Nick perrikwp_ is now known as perrikwp (n=chatzill@
20:58:48polluxx2006oh, okay... thats new.. ;) thanks
20:59:19pixelmapolluxx2006: that isn't new, the old system just tolerated it
20:59:40polluxx2006i see
20:59:45pixelmaand it's also in the manual (see "other tags")
20:59:47Domonokypolluxx2006: its usefull to try the wps in the sim with the −−debuwps option then you can see where the error is..
21:00:47*Domonoky is missing chars again, its −−debugwps
21:02:16 Join advcomp2019-- [0] (n=advcomp2@
21:02:33 Quit polluxx2006 ()
21:04:21saratoganerochiaro: i think this is what happens
21:04:39saratogaframe_out is written by the wma_window function which hanles overlapping the MDCT blocks
21:05:19saratogaat the end of each block decode, the block is copied into samples buffer, and then also copied to first have of the frame_out buffer (where its used to decode the next block)
21:05:52saratogasamples should have the entire frame written to it starting at the 0th byte
21:06:23nerochiarogood, that's probably all i need together with the number of samples from which i can get the length
21:07:35 Join WGC [0] (
21:07:42saratogaamiconn: is it safe to assume that gcc changes divides by 2 to shifts?
21:07:57amiconnNot necessarily
21:08:19amiconnIt will often use shifts, but even if it does, the code will be more complicated it the data is signed
21:08:21WGCIs it possible to get Rockbox to randomly shuffle all of the files in a folder?
21:08:51amiconnThe reason is that integer division rounds towards zero, while arithmetic shift always rounds down
21:09:24amiconnSo if gcc replaces a signed division with shifts, it will add code that changes the rounding behaviour
21:09:41amiconnIf you don't need that, using pure shifts will be more efficient
21:09:53 Join midgey [0] (
21:10:04 Nick advcomp2019-- is now known as advcomp2019 (n=advcomp2@
21:10:58WGCadvcomp, I got the Sansa e250 and installed Rockbox
21:11:24amiconnexample: (-5) / 2 == 2, but (-5) >> 1 == 3
21:11:35 Join midgey_ [0] (
21:11:44amiconnErm, (-5) / 2 == -2, but (-5) >> 1 == -3
21:12:18 Quit midgey_ (Client Quit)
21:12:46 Join Snake [0] (n=Snake@unaffiliated/snake)
21:12:48 Join Attreides [0] (
21:13:03Snakeis there anyway to change/reorder the items that are displayed on the main screen of rockbox?
21:13:14advcomp2019kool WGC
21:13:21WGCHow do I shuffle?
21:13:30SnakeWGC: Hold menu
21:13:34Snakeon the now playing screen
21:13:35AttreidesI use my Sandisk Esansa 270 to play back Napster files
21:14:01AttreidesCan I still do that (via original Firmware) when I installed the rockbox one?
21:15:10 Quit bdgraue (Remote closed the connection)
21:20:02AttreidesWill it play Napster files or not?
21:20:34 Quit advcomp2019- (Read error: 110 (Connection timed out))
21:22:04rasherAttreides: what kind of files are "Napster files"? Do they contain DRM?
21:22:23rasherThen the answer is no. There is no way for Rockbox to support them.
21:22:42AttreidesBut will it still wotk with the original FW?
21:22:49AttreidesWill Napster still recoxnize it?
21:22:56rasherI suspect so.
21:23:11advcomp2019has anyone found any info from the person on the post with the e200r
21:23:17AttreidesI only want rockbox to play Final Fantasy and stuff like that on the player^^
21:28:15AttreidesI'm not going to try it since I don't want to ruin my player
21:28:18 Join bdgraue [0] (
21:28:45Doomed_look on the forums i guess
21:29:06 Nick Doomed_ is now known as alienbiker99 (
21:33:44GodEater_advcomp2019: did you not see Bagder's follow up post ?
21:34:16advcomp2019no i will double check
21:36:51 Join Sodaholic [0] (
21:37:41advcomp2019like homer says "Do'h!"... i did not see the next page
21:38:42amiconnBagder: ping.....
21:40:27advcomp2019by the way read this page in the sansa forums −−>
21:41:44advcomp2019i thought someone said sandisk never contacted you (rockbox) to do the e200
21:42:48 Quit saratoga ("CGI:IRC")
21:44:24advcomp2019i thought so
21:44:53 Join saratoga [0] (i=98039a34@gateway/web/cgi-irc/
21:45:20 Quit saratoga (Client Quit)
21:48:15AttreidesI just wrote a mail to Napster asking if it would still work with the original one^^
21:49:24 Quit Rob2222 (Read error: 104 (Connection reset by peer))
21:49:33 Join Rob2222 [0] (
21:57:19***Saving seen data "./dancer.seen"
21:58:33alienbiker99ermm they wouldnt know because rockbox is unsupported, and plus they only deal with music, not hte hardware and firmware. as for not playing drm you should ask on the forums if nobody knows here. or try to find a channel for the sansa
21:58:58 Quit Siku ()
22:04:39 Quit midgey (Remote closed the connection)
22:05:29 Quit perrikwp ("ChatZilla [Firefox]")
22:07:42 Join WGC_ [0] (
22:07:57 Quit WGC (Read error: 113 (No route to host))
22:07:58 Nick WGC_ is now known as WGC (
22:08:04 Quit iamben_ (Read error: 104 (Connection reset by peer))
22:12:03 Join tihoc4n [0] (
22:17:49 Part tihoc4n
22:20:11 Join fm2 [0] (i=d9b95448@gateway/web/cgi-irc/
22:21:08fm2GodEater: are you here? I have a couple of questions about the code of the shortcuts plugin.
22:23:51fm2Is anybody here at all?
22:24:02Bagder131 persons only
22:24:02advcomp2019i am here
22:25:22fm2Bagder: err.. sorry :-) I should put on my glasses!
22:26:31fm2...Done. But I don't see GodEater. Or, I see but can't hear him!
22:26:40*amiconn notices Bagder around
22:27:02amiconnBagder: 2 things. (1) I know what's going wrong in genlang, but not how to fix it :(
22:27:44WGCIs Rockbox battery efficient?
22:27:47amiconn(2) I wanted to upload debbox.7z to my webspace, but failed twice. How could I send it to you for inclusion on the download server? (~140 MB)
22:29:34BagderWGC: yes in general
22:31:01fm2GodEater: anyway, I trust my eyes more than my ears so I'll just ask. (1) Why do you call sc_init() (and hence load the default file) even if the plugin was called with a parameter, i.e. via the context menu?
22:32:12fm2GodEater: (2) Why are there two different funcs for loading the default and a custom file? A copy-paste rest?
22:32:15Bagderamiconn: regarding #1, you could add that as a comment for the bug entry nls created
22:34:41 Join Robin_0800 [0] (
22:35:04 Quit Attreides ("ChatZilla [Firefox]")
22:36:03 Join WGC_ [0] (
22:38:07 Part fm2
22:41:42 Quit Genre9mp3 ()
22:43:31 Join [1]robin0800 [0] (
22:44:43amiconnBagder: done.
22:44:57BagderI'm reading it now
22:45:03 Join WGC__ [0] (n=chatzill@
22:45:12amiconnI edited it again
22:45:20amiconn(forgot one fact)
22:45:37 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
22:45:41 Join petur [0] (n=petur@rockbox/developer/petur)
22:49:33 Join Cillian [0] (
22:50:02CillianHow exactly does recording on ipod videos work? I assume you plug a stereo line signal into the headphone port?
22:50:39 Quit WGC (Read error: 113 (No route to host))
22:50:46 Join WGC___ [0] (
22:50:49 Nick WGC___ is now known as WGC (
22:52:46 Quit WGC__ (Read error: 113 (No route to host))
22:52:48 Quit Robin0800 (Read error: 110 (Connection timed out))
22:52:48 Nick Robin_0800 is now known as Robin0800 (
22:55:07Bagderamiconn: you're description of how it should work is... how it works, it just bugs
22:55:24Bagderyour even
22:56:11Bagderand I see the bug
23:00:33 Quit WGC_ (Read error: 113 (No route to host))
23:01:02 Quit XavierGr (Nick collision from services.)
23:01:06 Join XavierGr [0] (
23:02:58amiconnShould the H300 non-LCD remote hold influence the main backlight?
23:03:04amiconnI guess it shouldn't
23:03:42amiconnI mean, the purpose of a remote is remote operation, and in almost all cases that happens without looking at the main display
23:04:02advcomp2019remember that forum link i sent earlier they say "Unless there is official word on this from SanDisk, I take what I read on the internet with a grain of salt."
23:04:03amiconnLikewise, the archos remote doesn't influence the main backlight at all
23:07:29 Quit spiorf (Remote closed the connection)
23:08:08Lloreanamiconn: On the H100, the remote and the main unit controls only affected their respective backlights, iirc
23:08:18LloreanOr maybe I just always turned hold on on the main unit...
23:08:25amiconnLlorean: Yes, if you use an LCD remote
23:08:38LloreanOh, right, didn't see the "non-" for some reason
23:08:51amiconnThe non-LCD remote has code in backlight.c that makes its hold switch influence the main LCD
23:09:05LloreanThat seems strange
23:09:13amiconnHmm, in fact that code can't even work, since the non-lcd remote hold isn't detectable
23:10:14amiconnI'll remove it
23:11:04*amiconn already made the backlight code quite a bit simpler & smaller
23:12:35 Quit jgarvey ("Leaving")
23:12:47 Join WGC___ [0] (
23:13:16 Quit WGC (Read error: 104 (Connection reset by peer))
23:13:29 Nick WGC___ is now known as WGC (
23:15:11 Quit HellDragon (Read error: 104 (Connection reset by peer))
23:15:21 Join HellDragon [0] (i=Nocebo@unaffiliated/helldragon)
23:18:33ToHellWithGAnot directed at anyone in particular: do yall like ipod linux? if it has some keyboard-esque interface or a video player more capable than rockbox i might give it a go. i'd like to use it with a (large) database for dict/dictd as a mobile spell checker
23:19:03ToHellWithGAi was thinking i could become king of the scrabble nerds if i had a ridiculously thorough dictionary
23:19:26*amiconn just booted ipl a single time
23:19:54amiconnIt refused to boot a second time. Then I scrapped it and adapted rockbox.
23:20:02amiconn(that was for ipod mini G2)
23:21:44ToHellWithGAi'm on a brand new 30gb ipod video. i love rockbox for playing FLAC so far, but i read that its video decoding did not utitlize the video processing chip on my device
23:21:48LloreanToHellWithGA: Their video player is less useful than ours, I believe.
23:21:54ToHellWithGAoh, bummer
23:22:01 Quit smoo__ (Remote closed the connection)
23:22:13LloreanIf they had information on the Broadcom chip, we'd have that same info anyway.
23:22:23ToHellWithGAoh god, it's broadcom?
23:22:25LloreanThey and we are both in the same boat, for general purposes.
23:22:38ToHellWithGAi can barely get the broadcom wireless chipset on my mom's laptop to work in linux abandons all hope
23:22:45LloreanThe chip the iPod Video uses for decoding video in the Apple firmware is an undocumented Broadcom chip.
23:22:50*amiconn saved 71 lines in backlight.c, saving binary size *and* fixing some quirks at the same time
23:23:08LloreanIf you want video, get a Gigabeat. It's more than fast enough for 30fps.
23:23:13ToHellWithGAamiconn: you are a hero to the common man for making my battery last longer
23:23:32amiconnNot yet, I'm afraid
23:23:59amiconnRockbox has worse battery runtime than the original firmware on all PP502x targets.
23:24:04amiconnThat includes the ipod video
23:24:05ToHellWithGALlorean: i kinda just bought an ipod, mostly for a portable hard drive and just a little for music playback. if i never use the video features i'd still be happy with it
23:24:17amiconn(but rocbox battery runtime is still better than ipl's)
23:24:46ToHellWithGAamiconn: i don't really consider that real though. if i rarely play back mp3 audio, there's not much of a baseline against which to measure battery life
23:25:29ToHellWithGAthe added value of flac and ogg support outweighs battery life by orders of magnitude in my eyes
23:25:31 Quit WGC (Read error: 104 (Connection reset by peer))
23:25:40LloreanToHellWithGA: Well, if the same answers are discovered for the PP502x by someone as were found by amiconn for the PP5002, there's a decent chance we'll outperform the original firmware.
23:25:55peturamiconn: if you want a sansa, there's an e250 (2GB) on ebay, location = Berlin. little less than 2 days left, it's at 54 euro atm
23:26:30amiconnLlorean: It seems that our concept of dynamic boosting/unboosting is actually more efficient than the constant cpu clock most original firmwares set
23:26:44 Join harmattan [0] (
23:26:59ToHellWithGApetur: how many greenbacks are in 54 across the pond funnymoneys?
23:27:08ToHellWithGAwould you say 80 or so?
23:27:18amiconn..and it's also simpler to implement. Most original firmwares have large tables to calculate the necessary CPU clock for each situation (audio codec, dsp features enabled etc)
23:27:27Battousai75ish doesn't have a good grasp on current exchange rates
23:27:38amiconnToHellWithGA: Google '54 EUR in USD'
23:27:39 Join WGC___ [0] (
23:27:39 Nick WGC___ is now known as WGC (
23:27:48ToHellWithGAamiconn: i love google
23:28:13ToHellWithGAi can google things i've overhead in foreign languages and it will suggest correct spellings and help me find translations
23:28:24ToHellWithGAdunno if i'd call those results typical, but they impress me
23:28:25peturarg.. I was watching a cheap 3G ipod with no bids on, the b*stard has withdrawn the object :(
23:28:29amiconnNo really, google does currency conversion with pretty current exchange rates
23:28:44ToHellWithGAugly currency is not allowed then?
23:28:54Lloreanamiconn: I've seen things that suggest the Apple clock is dynamic too.
23:29:16amiconnLlorean: Maybe. Iaudio clock is definitely not dynamic, but table based
23:29:41LloreanWhile researching something, I came across some talk about a way to get the apple firmware to display a number that seemed to correlate with load at least
23:29:42amiconnIt switches when necessary though, but not demand-based
23:29:52LloreanThis number seemed to often be in the range of 29-37 for MP3 playback
23:30:18amiconnIf this is MHz, their MP3 decoder is significantly more efficient than ours
23:30:38 Join WGC___ [0] (
23:30:46LloreanConsidering how efficient ours is on Coldfire, I can imagine that happening if they're running it entirely on the second core, though
23:30:55LloreanBut their UI would have to be pretty efficient too, right?
23:31:06ToHellWithGAwhat is yours? is it mad or mpg123 or some FOSS decoder?
23:31:09 Part harmattan ("Saliendo")
23:31:15LloreanI think it's mad
23:31:28amiconnlibmad with several of our own optimisations
23:32:16ToHellWithGAthat's cool
23:32:24amiconnLlorean: Even with just mp3 decoding on one core, our codec needs more than 37MHz on average
23:32:24ToHellWithGAmp3 doesn't really get my rocks off
23:32:58ToHellWithGAit's cool for auditioning music i don't own at 192kHz, but i prefer lossless formats in case i ever hook my player up to a nice soundsystem
23:33:00 Join WGC____ [0] (
23:33:19amiconnJust run test_codec, and then calculate needed_mhz = 80MHz * 100% / decoding_speed_in_percent
23:33:45ToHellWithGAon that note, can rockbox use the headphone jack on players for coaxial digital output? standard PCM at 16 bit stereo and 44.1kHz sampling rate stuff would be great
23:33:54Lloreanamiconn: True, but they've had years to optimize for ARM (these numbers came from a 1st gen Nano), and are probably using commercial compilers that may be more efficient (possibly)
23:34:20 Quit WGC (Read error: 113 (No route to host))
23:34:24amiconnI guess they're using a hand-optimised codec in arm asm
23:34:36amiconn...and they might trade quality for speed
23:34:40 Quit WGC____ (Read error: 113 (No route to host))
23:34:45 Join WGC_____ [0] (
23:34:46LloreanAlways a possibility.
23:34:47 Nick WGC_____ is now known as WGC (
23:35:15amiconnAt least the iriver mp3 codec (coldfire of course) actually *sounds* to me like it is trading quality for speed
23:35:45amiconnAnd our emac'ified libmad is still very efficient :)
23:35:52LloreanWell, it's a cheaper way of getting better performance than actually putting the hours in. ;)
23:36:23amiconnI cannot easily compare apple OF sound to rockbox
23:36:55LloreanWell, I mean, I'd assume we have a long way to go with MP3 anyway, just because of the difference in the boost ratios for MP3 vs Ogg on Coldfire vs ARM, right?
23:37:02amiconnI would have to install a tool that deals with the itunes db. I learned there is a foobar plugin that is able to do this, but I didn't try it yet
23:37:11WGCRockbox is giving me incorrect battery readings.
23:37:44WGCIs that normal?
23:37:46WGCor rather it's contradicting the original firmware's readings
23:37:51amiconnLlorean: I guess we won't get mp3 much more efficient on arm. We might get ogg more efficient on coldfire though
23:38:17WGCAn ARM processor?
23:39:04BagderWGC: that's normal for several rockbox targets yes
23:39:38Lloreanamiconn: Looks like ~40mhz for 128kbps mp3 according to this test result for 5G
23:41:21advcomp2019wow i see that one forum is going nuts if it is really true or not that sandisk contacted rockbox
23:41:57peturthey did...
23:42:47amiconnLlorean: Yes, that sounds about right
23:43:01advcomp2019yea i know but peter_griffin says "Unless there is official word on this from SanDisk, I take what I read on the internet with a grain of salt."
23:43:04LloreanBagder: I'm honestly surprised Sandisk didn't give us a way to sign our firmware for upload, if they wanted us to upload it.
23:43:19advcomp2019he keeps on saying that
23:43:25Lloreanadvcomp2019: I'm pretty sure there used to be a blurb about it on the Sandisk e200-specific site, but I think that faded.
23:43:42BagderI don't think this was ever very "official"
23:43:46peturthey just used us for marketing
23:44:02peturremember how it went on slashdot & co?
23:44:12LloreanOh, we were definitely just used for marketing.
23:44:14rasherWasn't it also a marketing guy who contacted you in the first place?
23:45:19 Quit Febs (Read error: 110 (Connection timed out))
23:45:54rasherI would say Rockbox having a Sansa "dev board" should be enough proof, even for that peter_griffin person.
23:46:34Bagdernooo, the pics are fake!
23:46:48*amiconn wonders whether that dev board might be useful for researching pp details
23:47:07rasher"e280r, honestly I do not know [how rockbox got a dev board]. But I do NOT see an official statement from SanDisk, so like I said before, I take it with a grain of salt."
23:47:39*petur suddenly thinks of austriancoder and his usb problem...
23:47:39rasherThat's unhealthy skepticism right there
23:48:18BagderI think he'll have to wait for that official statement
23:49:49 Quit WGC___ (Read error: 113 (No route to host))
23:51:28LloreanI imagine if it ever showed up on their site, it'd be "I don't know how the message got on their site but..."
23:54:12peturwhy isn't the plugin code notified of the fact that you're shutting down? Seems like a bug to me unless this has some reason?
23:54:18ToHellWithGAdid yall have coaxial digital output as a goal at any point?
23:54:35ToHellWithGAi reckon it could be a bit of a battery saver as it would rely more on SNR than volume for clear transmission
23:54:59*petur wonders who yall is
23:55:50*Bagder wrestles with genlang
23:55:55peturand GA while I'm wondering
23:57:02LloreanToHellWithGA: Considering the heaphone jack is attached to a DAC, I'm not sure that's really feasible.
23:57:17 Part Cillian
23:57:21***Saving seen data "./dancer.seen"
23:57:43peturToHellWithGA: buy an iriver h1x0, they have optical in/out

Previous day | Next day