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

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

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

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

#rockbox log for 2007-02-15

00:00:53 Quit Thundercloud (Read error: 110 (Connection timed out))
00:01:55preglowi somehow doubt it
00:03:05 Join dpassen1 [0] (
00:03:10 Quit printfXh4 (Read error: 104 (Connection reset by peer))
00:04:10markunpreglow: at least it's good to see people working on other open source compilers
00:04:51 Join miepchen^schlaf [0] (
00:05:04DataGhostanyone in here who can supply me with the identity information for a 30GB 5.5G iPod? :)
00:05:10preglowindeed, i somehow doubt we'll see that project targetted against embedded stuff
00:05:46 Join SirFunk [0] (
00:05:58 Quit midkay ("Leaving")
00:06:06markunpreglow: well, it's a ARM compiler and most ARM CPU
00:06:12 Quit Thundercloud_ (Remote closed the connection)
00:06:16markun's are used in embedded systems, no?
00:07:00amiconn-O2 decreases binary size for arm
00:07:10amiconn-10KB for mini G2
00:07:40preglowmarkun: yeah, mostly embedded
00:07:45 Join Thundercloud [0] (n=thunderc@
00:07:58preglowamiconn: not bad
00:08:06preglowhow'd you fix the error?
00:08:12preglowright, forget it
00:10:13 Join MonkeyTamer [0] (n=chatzill@
00:11:59 Quit entheh ("^~")
00:14:58 Join shoora [0] (i=0cba4104@gateway/web/cgi-irc/
00:15:21*amiconn rediscovered his hack to get rid of -ffunction-sections on arm
00:16:26*amiconn is still unsure whether this hack is good or bad
00:17:04 Quit shoora (Client Quit)
00:18:36preglowamiconn: what hack was that?
00:18:55amiconnfunctions tagged ICODE_ATTR must not be static
00:19:05preglowmarkun: well, try it out :)
00:19:14amiconnCheck the irc logs for 20060505 and 20060506
00:19:59amiconnThere used to be 6 such functions in ata.c (the only ones in all rockbox). Now there are only 4 due to the optimised ata loops
00:20:38amiconnCuts another 12KB off rockbox.ipod for mini2g
00:20:41DerPapstone question about the sim... are there any known problems for running it on win2000? it launches well for me but then the menu is empty and it doesn't react on any keypresses
00:20:45 Quit pabs ("back in a jiffy")
00:20:51 Part Llorean
00:21:30 Quit matsl (Read error: 54 (Connection reset by peer))
00:21:53preglowDerPapst: should work fine on 2k, i think
00:21:59preglowamiconn: hrmr
00:22:08preglowamiconn: i wonder how hard fixing gcc would be...
00:22:11Nico_PDerPapst: the menu is probaby empty because you have no files
00:22:49amiconnNow if I knew how to stop these annoying warnings... there's one more warning in speex.c with -O2
00:23:26amiconn(not the aliasing one this time)
00:24:44amiconnrockbox.ipod: svn 391KB, down to 369KB
00:24:52ArathisIs it planned that with the new system the menu will be themeable as the WPS?
00:25:10DerPapstmmmh...Nico_P but i'm not able to view the plugins or any other stuff. it only displays some widgets (e.g clock and battery-meter)
00:25:32Nico_Phave you done "make install" ?
00:25:56DerPapstotherwise it errors. installation incomplete
00:26:05DerPapstit loads some files
00:27:11amiconnHmm, slightly unfair comparison since my local build doesn't include cue shit, but still...
00:27:31preglowthat just adds 1k5 or something anyway
00:27:36Nico_Pamiconn: cue what ? :p
00:27:44amiconnOn archos it adds 3K
00:27:58preglowyeah, but you weren't talking archos numbers
00:28:15preglowhmm, on second thought, i don't know which target that figure was for
00:28:49amiconnThe -O2 build boots, but doesn't want to play music
00:29:00preglowgo gcc
00:29:05amiconnResuming throws a "Stkov main"
00:29:46 Join pabs [0] (
00:30:00 Quit miepchen^schlaf (Read error: 60 (Operation timed out))
00:30:15 Join miepchen^schlaf [0] (
00:33:53*amiconn wonders why he's redoing this stuff instead of just reading the log
00:36:16preglowi wonder if the people over at #gcc have gotten any more helpful during the last year
00:37:22 Quit fasmaie (Read error: 110 (Connection timed out))
00:38:20 Quit ender` (Read error: 104 (Connection reset by peer))
00:40:23*amiconn wonders what's up with the main stack and -O2
00:40:36amiconnGot a stkov now while just browsing the menu
00:41:32 Nick _pill is now known as pill (
00:42:33preglowchecked what level it's at directly after start?
00:42:36preglowor can't you even do that? :)
00:43:13amiconnI can. 50%
00:45:54 Quit markun (Read error: 104 (Connection reset by peer))
00:46:18preglowwell, that's normal, it seems
00:46:49amiconnYes. The percentage doesn't tell the current usage, but the maximum usage since boot
00:46:53preglowyeah, know
00:46:58preglowjust forgot it for a while, heh
00:47:00amiconnMounting the volume uses quite some stack
00:48:38 Join dewdude_ [0] (
00:49:27 Join MonkeyTamer_ [0] (n=chatzill@
00:50:03 Quit MonkeyTamer_ (Client Quit)
00:50:23 Join markun [0] (n=markun@rockbox/developer/markun)
00:50:32*preglow goes bed
00:52:27 Quit mirak (Remote closed the connection)
00:56:11 Quit Brunellus (Remote closed the connection)
00:56:26 Quit dewdude__ (Read error: 145 (Connection timed out))
00:56:35 Join midkay [0] (n=midkay@rockbox/developer/midkay)
00:56:39 Quit jhMikeS (Nick collision from services.)
00:56:45 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
00:56:45 Quit MonkeyTamer (Read error: 145 (Connection timed out))
00:57:36 Part DerPapst
01:01:33 Join Brunellus [0] (n=luigi@unaffiliated/brunellus)
01:03:41 Quit Mouser_X (Read error: 110 (Connection timed out))
01:07:09 Join Mouser_X [0] (
01:08:35 Quit miepchen^schlaf (Read error: 60 (Operation timed out))
01:08:52 Join miepchen^schlaf [0] (
01:10:31 Join printfXh4 [0] (n=pseudo@
01:13:15***Saving seen data "./dancer.seen"
01:23:42 Quit Mouser_X (Read error: 60 (Operation timed out))
01:24:40 Join dan_a [0] (n=dan-mirc@
01:28:57 Join Mouser_X [0] (
01:36:08 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
01:36:11 Join miepchen^schlaf [0] (
01:36:34 Join blackness [0] (
01:36:38 Quit toffe ("MegaIRC v3.27")
01:36:58blacknessis it possible to hook a mixer up to an ipod and record music?
01:39:03dan_ablackness: It depends on what model iPod. Some have line in on the dock connectors, and you can make yourself a connector.
01:40:23blacknessI have the video
01:40:32 Quit BigMac ("Leaving")
01:40:51blacknessmy friends say they have a mixer with usb connection
01:44:34dan_ablackness: The video is one of the ones where you can record using the dock. At the moment you couldn't use the USB connection of the mixer to record through - but I'm sure their mixer will have a line out too.
01:45:59blacknessis there a tut on making a cable?
01:48:37dan_aHave a look at this:
01:51:51 Quit Nico_P (Remote closed the connection)
01:52:06blacknessso I could just make a line into the headset jack?
01:53:06 Quit miepchen^schlaf (Read error: 60 (Operation timed out))
01:53:17 Join miepchen^schlaf [0] (
01:53:43 Quit Mouser_X (Nick collision from services.)
01:53:54 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
01:53:54 Join Mouser_X [0] (
01:56:20 Part pixelma
01:56:21dan_aNo - the Video is the one we call the 5G (or 5.5G for the new ones.) They can't record through the headset jack, so you need to get a dock connector from somewhere, and then connect a 3.5mm stereo plug to pins 5, 6 and (I guess) 2 of the dock connector.
01:57:00linuxstbdan_a: Any luck with the bootloader on your 4g?
01:57:01w1ll14mdan_a: pin 2 is ground also for line in and line out
01:58:03 Join Thundercloud_ [0] (n=thunderc@
01:58:40dan_alinuxstb: None, but I've not been looking too hard. I've had my parents staying with me, plus I had to learn most of relational theory as quickly as possible, so I've not had too much Rockbox time.
01:58:50blacknessa usb dock adapter wouldn't work?
02:00:14 Join Llorean [0] (
02:00:58dan_ablackness: No. Your friends' mixer will probably look (to computers) like a soundcard, and want to connect to something that looks like a computer. Your iPod looks (to computers) like a hard drive, and also wants to connect to something that looks like a computer.
02:02:37blacknessalright so would a dock connector for a charger work?
02:04:31 Join BigMac [0] (
02:04:37dan_aIf you take it apart and solder the 3.5 mm plug to the right pins, yes.
02:05:24blacknessalright well i gotta go thanks for the help
02:05:37 Part blackness
02:06:35 Quit miepchen^schlaf (Read error: 60 (Operation timed out))
02:06:59 Join miepchen^schlaf [0] (
02:13:44 Quit roolku ()
02:14:32 Nick JoeyBorn is now known as JB_away (
02:15:37 Quit Thundercloud (Read error: 110 (Connection timed out))
02:23:16 Quit YouCeyE (Connection timed out)
02:24:18 Quit miepchen^schlaf (Read error: 60 (Operation timed out))
02:24:40 Join miepchen^schlaf [0] (
02:24:57 Join loenie [0] (n=loenie@
02:25:56 Part loenie
02:26:00 Join loenie [0] (n=loenie@
02:26:04 Part loenie
02:31:42 Join loenie [0] (n=loenie@
02:32:20 Quit Mouser_X (Nick collision from services.)
02:32:29 Join Mouser_X [0] (
02:33:04 Quit |Rincewind| ("Cya")
02:35:49 Quit loenie (Client Quit)
02:45:05 Quit CriamosAndy ("( :: NoNameScript 4.03 :: )")
02:48:42 Part Llorean
02:49:11 Quit x1jmp (Read error: 110 (Connection timed out))
02:50:36 Quit gamkiller (Read error: 110 (Connection timed out))
02:50:45 Join Llorea1 [0] (
02:52:01 Join obo [0] (
02:53:42 Quit Thundercloud_ (Remote closed the connection)
02:54:14 Quit midkay (Read error: 54 (Connection reset by peer))
02:54:48 Join midkay [0] (n=midkay@rockbox/developer/midkay)
02:54:51 Quit obo (Read error: 54 (Connection reset by peer))
02:55:07 Join obo [0] (
02:59:05oboquick question: When building the simulator, I get an error with respect to SDL. I've got libSDL installed, but is there another package that I need?
02:59:49 Join nickv111 [0] (
02:59:51 Quit wooo (Read error: 110 (Connection timed out))
03:00:44nickv111Hey, guys. I'm wondering what you'd think about removing the CORDIC stuff and transcendFunc from calculator.c
03:01:18nickv111Personally, I believe that with the speed of sqrt, an infinite series would work just as well for doing things like logs, exp, and other powers
03:01:34nickv111Same with trig functions: they can be expressed as infinite series
03:03:25JdGordonpatches are always welcome
03:04:26 Quit nickv111 (Nick collision from services.)
03:04:35 Join nickv111 [0] (
03:04:45nickv111Sorry. My internet connection has been a bit sketchy lately
03:05:28nickv111Anyway, I'm wondering what you'd think about replacing CORDIC and transcendFunc with individual functions for things like exp and log. We don't even need a sqrt function: with exp and log, we can quite easily implement sqrt
03:07:54Llorea1nickv111: Are they faster?
03:08:14nickv111I'm not quite sure about that
03:08:14 Quit obo (Read error: 54 (Connection reset by peer))
03:08:32 Nick Llorea1 is now known as Llorean (
03:08:35 Join obo [0] (
03:08:35nickv111But I believe that if we have enough processing power for sqrt, we have enough processing power for infinite series
03:08:42nickv111Besides, it would certainly clean up the code
03:09:06nickv111Also, I don't really like the way transcendFunc is used to handle all transcendental functions
03:09:36nickv111Maybe they should have their own functions?
03:10:28nickv111What do you think, Llorean?
03:10:33LloreanIs that beneficial to the function of the program, or just to reading the source code?
03:11:13nickv111It would help reading the source code, but I'm honestly not sure about performance
03:11:42dan_anickv111: Remember that anything that makes performance worse hurts battery life
03:11:59LloreanYou should test to see if your way of doing things is non-harmful before suggesting it.
03:12:14nickv111Personally, I've had some trouble understanding CORDIC
03:12:24nickv111So I can't really test the difference too effectively
03:12:43nickv111I guess what I'd do is implement my own log and exp functions on my computer, and my own CORDIC, to test performance
03:12:56nickv111I already have log and exp implemented in a C file I wrote.
03:13:01LloreanOr just implement your own log, and compare it to the cordic ones, on-target.
03:13:17***Saving seen data "./dancer.seen"
03:13:20nickv111But the CORDIC functions aren't even implemented yet
03:13:27LloreanTesting on a computer is somewhat pointless since the environment is very different on-target.
03:13:46dan_anickv111: You might be better talking about this between about 08:00 and 23:00 UTC, when the main developers are about
03:13:59Lloreandan_a: Actually, I think he wants to catch jhMikeS for this one, if I recall.
03:15:03nickv111It might help if I understood CORDIC
03:15:10nickv111So far, I don't really understand it much at all.
03:15:28 Join boutmuet [0] (
03:15:45nickv111Mostly, I understand how if you have a complex number in phasor form, and you multiply it by another in phasor form, their angles add and their magnatudes multiply, but that's really trivial
03:16:06nickv111What I don't understand is how functions like sin, cos, tan, and even log and exp are implemented from this. . .
03:16:10LloreanYes, but multiplies are slower than shifts, if I understand, and cordic gets away with just using shifts.
03:16:23nickv111Right, because it's using powers of 2
03:17:03nickv111Anyway, the magnatude in most cases is 1, I believe, because it's the unit circle
03:17:25nickv111But I have no idea how adding angles of complex numbers helps at all
03:17:47nickv111Maybe it has to do with the relation to z=x+iy form and z=r*e^(i*p)?
03:18:28JdGordonaaaaaahhhhhhhh... complex numbers have no place here.. :D
03:18:50nickv111Not according to
03:19:09nickv111Which is an explanation of CORDIC using complex numbers, and CORDIC very much relates to calculator.c
03:19:38LloreanI thought we had some cordic functions already for the equalizer at least, didn't we?
03:19:40 Quit Mouser_X (Nick collision from services.)
03:19:48 Join Mouser_X [0] (
03:20:06nickv111I'll check
03:20:24nickv111First, I'm going to make a patch for calculator.c that at least replaces PI with M_PI and so forth
03:21:22 Join TrueJournals [0] (
03:23:35 Quit miepchen^schlaf (Read error: 60 (Operation timed out))
03:24:17LloreanWhat I don't understand is, if CORDIC's not in use, how are you talking about 'replacing' something that's not there?
03:25:53nickv111CORDIC is there
03:25:55nickv111Just not wholly
03:26:19nickv111If you ever mess with the calculator plugin, you'll find that all of the functions that are transcendental say, "Coming soon"
03:26:20TrueJournalsIs the dictionary plugin not supposed to scroll? Because it's quite annoying not scrolling...
03:27:16nickv111So obviously something's missing
03:27:24TrueJournalsI should learn how to code and make it scroll...
03:27:56nickv111If I understood CORDIC and the algorithms that calculator.c is using, I'd probably know more precisely
03:28:00nickv111Llorean: Do you grok CORDIC?
03:28:52Lloreannickv111: Looking at the article on wikipedia, I could probably pick it up in a few days. Doesn't look like there's anything there outside of what maths I know, but since I've never played with it before I can't say for certain.
03:28:59LloreanBut no, at this moment, I don't. :)
03:30:34 Quit TrueJournals ("User disconnected")
03:33:15Lloreannickv111: Do you understand the basic theory behind what CORDIC is trying to do?
03:33:41LloreanBasically, is it the implementation that's throwing you off, or the theory that leads to it is confusing you making it hard to understand what the implementation is trying to do?
03:33:49nickv111The theory
03:34:12nickv111I know that it tries to approximate an angle
03:34:43Lloreannickv111: Try reading this one:
03:34:43 Quit obo (Read error: 54 (Connection reset by peer))
03:34:48nickv111It does this by subtracting or adding an angle from the current angle
03:34:54LloreanIt speaks of it much more geometrically, without so much talk about rotation matrices and such.
03:34:58nickv111Each angle it adds or subtracts is half that of the previous.
03:35:05 Join obo [0] (
03:35:11 Join safetydan [0] (i=cbca159f@rockbox/developer/safetydan)
03:35:24 Quit Mouser_X (Nick collision from services.)
03:35:34 Join Mouser_X [0] (
03:36:47safetydaneh... I wouldn't look at the cordic function in eq.c as a shining example of cordic use. It's extremely specific to producing sin/cos in the fixed point format the eq wants.
03:37:41safetydanPersonally, I think infinite series should be more than fast enough for the calculator plugin. It's not like you're going to be doing more than, oh say 10, complex functions per second.
03:37:51safetydanJust has to be faster than a human :)
03:38:26Lloreansafetydan: Yeah, but the other thing he's doing is getting rid of constants defined in calculator for ones in math.h, in which case sin, cos, etc might want to be shared as well for future use.
03:38:36LloreanAnd for future use, though ought to be the fastest ones we can leave lying around. ;)
03:38:37 Quit obo (Read error: 54 (Connection reset by peer))
03:39:04 Join obo [0] (
03:39:28safetydanLlorean, the calculator plugin uses double throughout so it has different needs from the core.
03:40:26 Join gamkiller [0] (
03:40:30LloreanAh, well in that case never mind me.
03:41:15safetydanthough if someone wants to expand apps/plugins/lib/fixedpoint.h that'd be great too
03:41:22safetydanthere's only fsincos and fsqrt there
03:41:48nickv111safetydan: I guess battery life depends on performance, though, so do you believe that replacing CORDIC in calculator.c with infinite series would be okay?
03:41:49 Quit obo (Read error: 54 (Connection reset by peer))
03:42:00nickv111safetydan: I'd actually really like to know how CORDIC works
03:42:04 Join obo [0] (
03:42:06nickv111safetydan: I'm still trying to learn the theory
03:42:23safetydannickv111, I wouldn't think that battery life should be (much of) a concern for plugins.
03:42:24dan_anickv111: Ignore what I was saying about battery life if these things are only used in the calculator plugin
03:42:52*dan_a should have been in bed hours ago
03:42:59safetydanThe calculator plugin is in dire need of updating so go ahead and update it however you feel comfortable
03:43:20safetydannickv111, if you understand cordic, you're doing better than I did :) My maths is horrible.
03:43:56nickv111So if performance isn't too much of an issue, I think that infinite series would work just fine
03:44:02nickv111Especially since they may actually /work/
03:44:11nickv111The number of iterations needed for infinite series to work is pretty low
03:46:30nickv111Okay, well, here goes.
03:46:36*nickv111 rolls up his sleeves, and gets to work
03:46:48nickv111Wow, it's been a while since I've coded.
03:46:49 Quit obo (Read error: 54 (Connection reset by peer))
03:47:04 Join obo [0] (
03:49:10 Quit obo (Client Quit)
03:53:49LloreanWow, CORDIC really is quite clever. Now that I've read a bit more, it's one of those things where it's the kind of elegant solution you really feel like you should've been able to come up with yourself, but your brain just doesn't look in that direction.
03:54:06nickv111It's clever, but I don't understand much of it
03:54:17LloreanBut, anyway, yeah for calculator as long as the user can't tell that they're waiting for an answer, it's probably 'fast enough' however you do it.
03:54:54nickv111I believe that you find a solution by rotating a complex number in phasor form
03:55:37nickv111However, the details are still a bit hazy to me
03:56:41nickv111How does one go into the main menu while using the simulator?
03:57:10LloreanDepends on the keymap for the target you're using, I believe
03:57:23LloreanThere should be a bitmap that shows which number pad keys associate with which buttons.
03:57:54nickv111Oh, thanks
03:58:22hcsrun with −−background to display said bitmap
03:59:39 Quit gamkiller (Read error: 110 (Connection timed out))
04:00:01LloreanI will say that the wikipedia article on CORDIC is not very well written, in my opinion. It looks like something I'd try to pass off as a 'showing my work' to one of my highschool teachers. :)
04:01:13LloreanIt's been 5 years since I've really done any non-basic math, but I think I've got a general grip on this cordic thing now. I don't think I could write out an algorithm yet for it, but I see what they're doing.
04:01:37Lloreansafetydan: What was it that fixedpoint.h needed specifically, out of curiosity? It's unlikely I do it, but I've done unlikely things before.
04:03:41safetydanLlorean, nothing as far as I know. It's just that the library looks a little anemic at the moment with only two functions :)
04:05:23 Join JerryLange [0] (
04:06:46 Quit blind (Read error: 104 (Connection reset by peer))
04:06:52nickv111As far as I know, I'm ready to submit a patch for calculator.c that takes care of the whole redundant defines problem
04:09:57nickv111Would it be appropriate for me to possibly post it to a pastebin to see if everyone here is cool with it?
04:10:11 Quit Mouser_X (Read error: 110 (Connection timed out))
04:10:13nickv111The PI function on the calculator still works. That's all I've tested it
04:12:43LloreanM_PI_2 strikes me as a bad name for PI/2
04:12:53LloreanIs that a standard or did someone of ours come up with that?
04:14:13LloreanI apparently don't use half of pi often enough
04:14:38hcsno, you take the whole pi, greedy
04:14:52LloreanPi is good stuff.
04:15:04nickv111Meh. It cuts down slightly on computation
04:15:24nickv111Okay, I posted the new patch
04:15:57nickv111BTW, certain definitions are handy. For example, M_LN10 may exist, but why divide by M_LN10 when you can multiply by M_INVLN10?
04:16:21nickv111For converting bases, it's much more handy to have M_INVLN10
04:16:36LloreanOh, I've no doubt definitions are handy.
04:16:56LloreanI just didn't like the name of that one. Which means I'm going to have to track down some guy somewhere who named it decades ago and complain. :-P
04:17:08BigMacHey on ubuntu how would I install subversion(or whatever I will need to patch and compile for the 5g)
04:17:16nickv111apt-get install subversion
04:18:24 Join fasmaie [0] (
04:19:26BigMacthat all I am going to need?
04:19:50nickv111Are you just trying to check out the code?
04:20:02nickv111To compile it, you'll need to make sure you have gcc and is related components
04:20:14LloreanBigMac: You also need the development toolchain, installed with from the source's /tools/ directory
04:20:17nickv111For the SDL version of the simulator, you'll need the development version of the SDL libs
04:20:22nickv111Right. is incredibly handy
04:20:46LloreanIt makes that whole process a lot simpler.
04:21:08nickv111Formerly, when I had to cross-compile, I'd download binutils, compile it, download gcc, compile it, and download a C lib
04:21:17nickv111Kind of a pan
04:21:46BigMacok so once I svn co sn.... I can cd into the tools directory and sudo apt-get install
04:22:07nickv111No, no
04:22:08LloreanYou just run
04:23:10BigMacoh sorry
04:23:24nickv111Hey, svn stat shows that I've updated calculator.c, and indeed I have, as shows the patch. How do I basically commit the work I've done, without comitting it into the repository? In other words, how do I make it so that it makes a new revision from what I have, but only locally?
04:23:27BigMacand I will look up to compile
04:23:44Lloreannickv111: Why do you need a new revision?
04:24:07nickv111Llorean: So that when I make future changes, I can post a new patch
04:24:38daurnimatordid you guys see that archos firmware is now open source?
04:24:41LloreanDon't you want the new patch to be able to be applied to clean SVN, so someone doesn't have to download *all* of them to be able to apply the changes?
04:24:42daurnimator(for 4th gen)
04:24:58LloreanShouldn't the new patch simply be cumulative?
04:25:14nickv111I guess so
04:25:16nickv111I wasn't sure, personally
04:25:18safetydannickv111, unfortunatley svn doesn't really support that sort of function. If we were using git or darcs or arch maybe.
04:25:30safetydanYou could look at seeing if you could get svk going
04:25:39nickv111That's fine
04:25:43nickv111It's not really that important to me.
04:25:48Lloreannickv111: Or you could always -R your old patch before diffing if you really had to. :-P
04:26:07nickv111Well, I guess I'll just have a cumulative patch
04:26:23safetydangenerally I use multiple rockbox directories if two patches touch the same file
04:26:32nickv111When I post the patch, should I just have a note, "This patch obsoletes x. . ."
04:26:58Lloreannickv111: Well you could just post it as a comment to the same task.
04:27:20nickv111Like, "This patch is obsoleted by y. . ."
04:27:57nickv111BTW, is there a "pow" function?
04:28:31nickv111For using infinite series, it would be handy, though the infinite series only require very simple exponents (in the form a^b, where b is a positive, whole number)
04:29:07nickv111Naturally, I could use a for loop to implement this kind of exponent
04:29:09safetydanI don't think a pow function is in Rockbox
04:29:21nickv111Okay, I can always implement my own simple pow function.
04:29:33nickv111I've got to go take a shower. See you guys later tonight, hopefully
04:30:00daurnimatoranyone want to port to archos 4th gen?
04:30:15Lloreandaurnimator: What is an 'Archos 4th gen' anyway?
04:30:26LloreanAre you talking about the 400-series or something?
04:30:32daurnimatorthe new 4th gen models
04:30:42Llorean4th generation of what?
04:30:44daurnimator - 404,504,604 (wifi),704
04:32:58BigMacLlorean: What would the c lib be called because gcc and binutils are already installed
04:33:18LloreanBigMac: All I personally told you to do was to run the script, right?
04:34:31BigMacYah I know but you seem to be the knowledge base round these parts
04:34:56LloreanYes, but considering you're following instructions other than those I gave you, maybe the problem is that you're trying to do things you don't need to?
04:35:19safetydanBigMac, what have you done so far? The rockboxdev script takes care of pretty much everything.
04:36:56BigMacI haven't run it yet because I was told to install binutils, gcc, and a c lib which I have yet to find
04:37:14BigMacrockboxdev will install it?
04:37:35 Quit Arathis ("Bye, bye")
04:37:54safetydanBigMac, it will install the cross compiler versions, yes.
04:37:55LloreanBigMac: installs all the compilers etc.
04:38:30fasmaieBigMac...just run it as root
04:38:37BigMacok when I connect to svn I should run that right
04:40:10perldiverwhat does ON + PLAY do?
04:40:30fasmaiefirst download the svn archives
04:40:47Lloreanperldiver: Kinda depends on what screen, and what target, I think.
04:41:10safetydanBigMac, yes
04:43:17perldiverso im wondering...
04:44:09 Quit lukaswayne9 ("Ex-Chat")
04:44:15Lloreanperldiver: Well, I've asked what needs to be asked, I think
04:47:15nickv111Good thing is, I already coded exp and log functions, along with necessary functions for those, a simple factorial and a simple power function
04:47:21nickv111So I can basically paste them in
04:47:59 Quit Brunellus (Remote closed the connection)
04:49:44 Join Brunellus [0] (n=luigi@unaffiliated/brunellus)
04:58:19nickv111For a factorial function, what would you guys think of using int?
04:58:24nickv111It may be a little limiting. . .
04:58:32nickv111Maybe a long int would be better?
05:00:53 Join blackness [0] (
05:02:07blacknessso my ipod was working fine earlier after I installed rockbox, I turned it off and came back and I can't get it to do anything is seems dead. Can anyone help me?
05:02:44safetydannickv111, what's wrong with double for factorial?
05:02:53nickv111I guess I could use double
05:02:57nickv111I will. Okay.
05:03:17Lloreanblackness: Have you tried the reset combination?
05:03:46blacknessyeah, nothing but blackness
05:04:06nickv111Which is ironic, given your nickname. Or is that intentional?
05:04:18Lloreanblackness: What exactly did you try?
05:04:51blacknessI did what the manual said for my 5g ipod
05:05:00nickv111Which entails. . .
05:05:13Lloreanblackness: I used the word 'exactly' because I wanted you to describe what you've tried.
05:05:19BigMacwhy am I getting those errors
05:05:35BigMaclike where is my arm elf files supposed to be
05:05:41nickv111BigMac: /usr/local
05:05:50nickv111BigMac: Did you add /usr/local/arm-elf/bin to your path?
05:06:00safetydanBigMac, maybe try doing "apt-get install build-essentials"
05:06:03blacknessI extracted rockbox-ipodvideo to my ipod
05:06:11safetydanLooks like you're missing the glibc dev stuff
05:06:17Lloreanblackness: No, I mean what exactly did you do to try the reset combination?
05:06:36Lloreanblackness: You'll notice I never asked about the installation process.
05:06:36blacknessthe middle button and the menu together
05:06:40BigMacnickv111: Nope it never told me to do that. How do I do that
05:06:47Lloreanblackness: How long did you hold it?
05:06:53nickv111export PATH="$PATH:/usr/local/arm-elf/bin"
05:07:09blacknessnot that long maybe 15 sec
05:07:16Lloreanblackness: And did you turn hold On and Off first?
05:07:46LloreanBecause the official instructions include doing that, which seems to clear up some issues with keypress detection. In _most_ situations you don't need to, but sometimes it seems you must.
05:07:52blacknessah thank you got it working
05:08:06blacknessI was getting worried
05:08:14nickv111blackness: That I can understand ;)
05:08:30BigMacnickv111: do I have to specify anything else or be in a certain directory when I execute that?
05:08:55blacknessdoes it do that a lot?
05:09:53BigMacit still says arm elf is not in my path when I do ../tools/configure
05:10:17Lloreanblackness: No, not normally
05:10:44safetydanBigMac, did you run the rockboxdev script?
05:11:09BigMac./ right
05:11:13BigMacwhile in tools
05:11:22safetydanWhat did it say at the end?
05:11:38safetydanIt should have said something like "Make your PATH include blah"
05:12:12BigMacNormal build selected
05:12:12BigMacUsing source code root directory: /home/mike/rockbox
05:12:12BigMac../tools/configure: 1383: arm-elf-gcc: not found
05:12:12DBUGEnqueued KICK BigMac
05:12:12BigMacWARNING: The compiler you must use (arm-elf-gcc) is not in your path!
05:12:12BigMacWARNING: this may cause your build to fail since we cannot do the
05:12:13***Alert Mode level 1
05:12:13BigMacWARNING: checks we want now.
05:12:15BigMacCreated Makefile
05:12:24nickv111Please do not paste in here, BigMac
05:12:50safetydanBigMac, No, what did the rockboxdev script say at the end?
05:13:20***Saving seen data "./dancer.seen"
05:13:39BigMacoh wait maybe hold on
05:14:29BigMacit told me to Make your PATH include :/usr/local/arm-elf/bin
05:15:04nickv111You probably closed out of your shell after you did export PATH="$PATH:/usr/local/arm-elf/bin"
05:15:05safetydanBigMac, so what does your $PATH look like now?
05:15:39BigMacI am not sure
05:15:47BigMachow can i view it?
05:15:51 Join loenie [0] (n=loenie@
05:16:07fasmaieI just tried built rb from the svn from 10 min ago, and my database is broken
05:16:24fasmaieI cleared and rebuilt it to no avail.
05:16:45fasmaiemp3/aac are reported fine but mpc are not
05:17:30safetydanBigMac, type "echo $PATH" on the command line
05:17:46 Join combrains [0] (
05:21:05 Join miepchen^schlaf [0] (
05:22:04 Quit midkay (Read error: 54 (Connection reset by peer))
05:22:10 Join midkay_ [0] (
05:22:14***Alert Mode OFF
05:29:40nickv111I've got to go. Other responsibilities and all. The problem with this calculator stuff is that not only is there a number involved, but a power
05:29:48nickv111So now I have one more thing to look over, and I'm not totally up for coding tonight
05:29:54nickv111I'll work on it tomorrow. Night, all
05:29:57 Quit nickv111 (Remote closed the connection)
05:32:23 Part loenie
05:32:52 Quit fasmaie (Remote closed the connection)
05:43:03 Part safetydan
05:46:32 Join gamkiller [0] (
05:46:36 Quit Vyrus001 (Connection reset by peer)
05:46:49 Join Vyrus001 [0] (
05:47:34 Join Mouser_X [0] (
05:49:24 Quit rotator ()
05:52:17 Join m0nk_ [0] (
05:52:22m0nk_hey guys
05:52:33m0nk_do any of you know where the videos are stored on the ipod video *using orginal firmware*
05:53:10CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
05:53:10*m0nk_ is about to put rockbox on this bastard device
05:55:18LloreanRockbox doesn't play those videos anyway
05:55:58m0nk_Llorean: i know
05:56:04m0nk_i have rockbox on my gigabeat
05:56:11m0nk_im jsut messing wiht my friend's ipod for her
05:56:23m0nk_otherwise it would have rockbox on it:|
05:56:48m0nk_i have some ipod videos i need to throw on i just dont have Itunes...nor do i have a windows/mac box at my disposal
05:57:25LloreanWell I believe they have to be added to the iPod database to be used anyway, you can't just copy them over.
05:57:35LloreanBut this isn't the right place to be asking about problems with the retail firmware.
05:58:20m0nk_ahh i know i know
05:58:23m0nk_you guys just own:P
05:58:46m0nk_and i know about hte dtabase thign
06:00:09Mouser_XQuestion: Does the VMware stuff run in Win98? For some reason, it runs "not so well" on my WinXP machine. I was wondering if anyone could tell me whether I might have more luck elsewhere.
06:00:29Mouser_X(I want to compile/patch my own versions of Rockbox.)
06:01:13LloreanMouser_X: That would just depend on whether VMWare has a working player for Win98.
06:01:31Mouser_XHmmm. I guess I'll have to look. Thanks.
06:02:19Mouser_XI'm glad SPC was finally commited.
06:02:39Mouser_XNow, when I do a custom build, all I need is the MOD playback patch.
06:03:06Mouser_X(I only got that one recently. HCS was nice enough to build it for me.)
06:06:09 Quit m0nk_ ("Lost terminal")
06:11:29 Quit Febs (Read error: 110 (Connection timed out))
06:16:41 Part blackness
06:17:42 Quit Nibbier (Read error: 110 (Connection timed out))
06:18:30 Join Nibbier [0] (
06:23:22 Quit thegeek (Remote closed the connection)
06:26:16combrainsmouser_X: are you using the vmware server or player?
06:28:11combrainsin either case, you need to install VMware tools inside the VM.
06:28:37scorcheyou dont *need* to, but it helps a ton
06:28:39jhMikeSMouser_X: the player that I could obtain requires xp. one thing I say to everyone is cut the memory down. I set mine to 32MB and it's fine.
06:29:23combrainsI know you dont *need* to but it will solve is problem
06:29:44 Join ptw419 [0] (
06:30:07combrainsI did it on a friend's pc and it helped make his win 98 VM get to a useable state
06:30:20combrainsanyways bbl
06:33:57 Quit XavierGr (Read error: 110 (Connection timed out))
06:37:49 Join Kitt0s [0] (
06:48:12 Join decayedcell_ [0] (
06:51:24 Quit Mouser_X (Read error: 110 (Connection timed out))
07:05:58 Join BHSPitLappy [0] (
07:06:06 Part JerryLange ("all")
07:08:13 Quit decayedcell_ (Remote closed the connection)
07:08:22 Join decayedcell_ [0] (
07:11:20 Part Llorean
07:13:23***Saving seen data "./dancer.seen"
07:21:01 Join Mouser_X [0] (
07:26:49 Join illriginal [0] (
07:27:41illriginalIs it time to upgrade? :D
07:28:59 Join kaaloo [0] (
07:30:38illriginalWhere can i find the FWIMG01.DAT for the toshiba F series?
07:30:46illriginalOn your site to download that is.
07:31:56Mikachudo you want to install rockbox?
07:32:46illriginalyeah, what happened was, i turned up my toshiba mp3 player while holdin the up button on the side (the volume) and it replaced the original FWIMG01.DAT with the rockbox one.
07:32:55illriginalso basically i need to get the rockbox version.
07:33:19illriginalwait... i said that really wrong.
07:33:19Mikachui don't know how to install on those
07:33:32Mikachubut you have read the install instructions in the manual?
07:34:16illriginali turned ON my mp3 player, while holdin the volume up botton, and it went into a screen where it found my original FWIMG01.DAT file, and it replaced the rockbox FWIMG01.DAT version.
07:34:21 Part decayedcell_
07:34:38illriginalyeah im tryin to find it on your site.
07:35:11Mikachui think most are asleep right now, try waiting a few hours
07:35:56Mikachuis this it?
07:36:27Mikachuyou probably want to start here though
07:38:32illriginalthat's what i need, the bootloader
07:41:03illriginalwoohoo thanks a lot Mikachu
07:41:25illriginalis there a page with the latest updates since jan?
07:41:28Mikachuall i did was click the Manual link on the web site
07:41:57illriginalyeah i did... i was lookin for the bootloader, but it gave me this page:
07:42:06illriginalthank you mikachu
07:44:18illriginalhey Mikachu... a Cuesheet is the table of contents file right? the .m3u file that when you open it'll play all the songs in the folder instead of draggin all of the files?
07:44:44Mikachuno idea, i don't use cue files
07:45:18 Quit Mouser_X (Read error: 110 (Connection timed out))
07:46:45 Join Mouser_X [0] (
07:50:24illriginal`<<, records FM
07:50:40 Quit Mouser_X (Nick collision from services.)
07:50:51 Join Mouser_X [0] (
07:52:18 Quit jhMikeS (Nick collision from services.)
07:52:24 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
07:52:49 Join JerryLange [0] (
07:52:55 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
07:54:26 Quit ackbahr (Read error: 60 (Operation timed out))
07:57:39illriginalisn't the manual supposed to tell you what every option does?
07:57:55LloreanThat's the goal.
07:58:01LloreanIs there a particular one that's been left out that you noticed?
07:58:03Mikachuthere are things added very often, so very new items might not be there yet
07:58:12Mikachuor in the wrong section
07:58:44 Quit scorche (Read error: 54 (Connection reset by peer))
07:58:46illriginalno no i remember there was a page where it tells you what each option means and what it does for the Toshiba
07:58:53illriginalim tryin to find that.
07:59:06 Join scorche [0] (n=scorche@rockbox/administrator/scorche)
07:59:07illriginali wanna learn about this fm recording and also cuesheets.
07:59:11Mikachuare you looking at the toshiba manual?
07:59:23illriginalyes sir
07:59:26LloreanWell, the Gigabeat doesn't have an FM chip, so FM recording isn't exactly a feature.
07:59:46Mikachucue sheets were added yesterday
08:00:03LloreanThey also don't have much in the way of a related option? Isn't it just enable/disable?
08:01:30illriginalit is
08:01:45illriginalalso.. what option lets my mp3 player play every song in my folder
08:01:52LloreanJust play a song.
08:02:00illriginalit just repeats the same album.
08:02:07LloreanDoes that folder just have that album in it?
08:02:31LloreanYou did ask how to play all the songs in a folder...
08:02:38illriginalyeah like i have the folders spllit up my genre, then alphabetical order of artist, then their albums
08:02:58illriginali want to be able to play every single song possible from artist A to Z
08:02:59Mikachuif you open the context menu on a folder, you can add it recursively
08:03:26Lloreanillriginal: Look up the "Insert" playlist option, as well as the option to recursively add subdirectories
08:04:21illriginalor could it be the auto change directory?
08:04:52LloreanAuto Change Directory will work, in a way.
08:04:59LloreanBut if you turn on shuffle, it'll only shuffle the songs from one album
08:05:06illriginalnah i dont like shuffle.
08:05:42illriginalbasically if im listening to artist A and im on his last album, last song, the next song should start on artist B with his first album.
08:06:00illriginaland first song of that first album.
08:06:28LloreanWell, you can try auto change directory, but I believe it causes more rebuffers than simply using playlists properly
08:06:47illriginalrebuffers as in it takes time to load?
08:07:16LloreanMaybe a second or two between albums, but it also just uses a bit more battery (assuming I'm right, I'm not sure on this one, but I seem to recall mention of it)
08:07:37illriginalah i see..
08:07:53 Quit dpassen1 (Read error: 110 (Connection timed out))
08:07:54Mikachuit would probably rebuffer the start of the album when you're at the end, then throw that away again when it switches to the next album
08:08:14LloreanMikachu: I think it just doesn't rebuffer beyond the end of the album, so you get no next song info.
08:08:24illriginalso recursive does the same exact thing yes? And if so, does this option mean that I need to load every single track by every single artist onto this recursive playlist?
08:08:30LloreanBecause you have to have 'repeat none' for it to work
08:08:45Lloreanillriginal: No, you just click the music folder, and choose "Insert" and you're done.
08:08:47illriginalyeah i have repeat completely off.
08:09:03illriginalcan i just click on the genre?
08:09:13LloreanJust make sure recursively adding is enabled.
08:09:19illriginalyeah doin that right now
08:09:36illriginalhow may i inster?
08:09:45LloreanIt's in the context menu.
08:09:45illriginalplaylist catalog?
08:09:48LloreanPlease, just read the manual
08:09:57LloreanYou could've searched it for "Insert" in like, half a second. :-P
08:10:07illriginallol that's why im speakin to you... the manual im lookin for.. i can't find.
08:10:14LloreanYou can't find the manual?
08:10:23illriginalim on a manual.. hold on ill show you.
08:10:34LloreanThe PDF manual?
08:11:07illriginal-.- nevermind im on it... i must have right clicked open in new tab and not realized that i didn't check it.
08:14:22illriginalno it doesn't say how to insert it
08:14:57LloreanTry using the PDF manual so you can simply search for it in the whole thing...
08:15:10illriginaloh the PDF has more information?
08:15:15LloreanNo, it has just as much
08:15:20LloreanBut you're looking in the wrong place.
08:15:30LloreanSince the PDF is one big file, you can search the whole thing using the search function.
08:15:36LloreanWhich would save you a lot of time
08:17:04illriginaldo i need to create this playlist?
08:17:22LloreanThe first thing you insert automatically creates a playlist in memory
08:17:25LloreanYou can save it if you'd like
08:18:37illriginalim sorry... i have the option set to ON
08:18:46illriginali read both the html and the pdf of this section: 5.8. Playlist Options
08:19:25illriginalnothin reads anything about how to insert the folders into this playlist so that it will play all folders.
08:19:30*Llorean sighs
08:19:41LloreanIs there some reason that you're ignoring what I'm telling you to do, and doing something different instead?
08:19:51illriginali click create playlist, then view current playlist... it says i have no playlist.
08:19:52LloreanI said specifically to use the search function, in the PDF.
08:20:03illriginalim ill search for insert.
08:20:32LloreanInstead you reread the same section that I told you already was not the right section, after telling you that the PDF has just as much information as the HTML.
08:20:51illriginalso the manual isn't organized correctly?..
08:20:59LloreanNo, the manual is organized fine.
08:21:41LloreanBut it'd be very hard to insert a song into a playlist from anywhere but the filetree.
08:21:53LloreanSpecifically the file context menu
08:22:00LloreanWhich was mentioned to you very much earlier in this conversation.
08:22:12LloreanBut you were also told explicitly to use the search function
08:22:31illriginalok what is the keyword i need to search for
08:22:35illriginalit's not Recursively nor insert.
08:22:41LloreanNo, there's definitely insert.
08:22:45LloreanAre you searching the PDF manual?
08:22:48LloreanOr just one page of the HTML one?
08:22:58illriginalthe PDF ;\
08:23:02LloreanIt's there.
08:23:05LloreanI went and checked myself
08:23:31LloreanThe word insert shows up a LOT in the manual.
08:23:47Llorean22 times, specifically
08:24:58Mouser_XJust to note: I almost always use the insert feature to make my playlists.
08:25:25LloreanIt's probably one of the most used features in Rockbox, I'd imagine.
08:25:44combrainshell yeah
08:25:44Mouser_XRarely do I select a specific song, unless I'm in my OCR directory (there's over 1500 files in there. This makes it difficult to go through in one day).
08:26:08LloreanIt features rather heavily in the whole "Working with Playlists" section of the manual
08:26:11combrainsI like being able to make an OTG playlist
08:26:28combrainscant do it with many things
08:26:32LloreanMouser_X: I just leave it on shuffle and resume it several times over a period of months to get through OCR.
08:26:48illriginalcombrains i gotta register.
08:26:49LloreanOf course, OCR mixed with my SPC collection, now that's a long playlist.
08:27:02Mouser_XI started some FF music. 12 hours later, it was still digging through the FF music...
08:27:20Mouser_XAll of which was from OCR.
08:27:36Mouser_XI *very* rarely use the shuffle.
08:27:46 Join GodEater_ [0] (i=c2cbc95c@gateway/web/cgi-irc/
08:27:48combrainsillriginal, i mean register your nic on the irc server - then I can help you in a pm instead of clogging the main channel
08:27:53Lloreanillriginal: Look, just see section 4.4 of the manual.
08:28:10LloreanBut I strongly suggest learning how to properly use a search box. It'll help you greatly in the future.
08:28:35Mouser_XMy method of searching requires patience.
08:28:35combrains*spot the sarcasim*
08:28:52Mouser_XI just go through the whole thing, and keep an eye open for what I'm looking for.
08:29:18Mouser_X(There is some truth to that, but that's certainly not how I do most of my searching.)
08:29:42LloreanHonestly, the only parts of the manual I've read, are the parts I check up on immediately before telling someone to go read the manual.
08:29:53combrainsone of the things that I have been wondering about lately, what are these 2 new codecs in rb and what do they do?
08:30:02combrainsie speeks and scp
08:30:09Mouser_XSPC is GREAT!
08:30:13Mouser_XIt's SNES music.
08:30:21Lloreancombrains: Speex is primarily a voice codec. You can encode audiobooks at less than 24kbps with still perfectly suitable quality
08:30:36combrainsah i see
08:30:43Lloreancombrains: SPC is the format of SNES music on the hardware, and the codec actually emulates the SPC chip I believe to play it back fairly accurately
08:31:01illriginali clicked on the folder root, and it goin through every song and saying codec failure.
08:31:09Mouser_XIf Rockbox supported compression, I could have 20,000 SPCs in less than 100 MB.
08:31:33Lloreanillriginal: That usually means an improperly installed Rockbox. Have you played any music since your last update?
08:31:46illriginali just updated like 30 minutes ago...
08:31:54LloreanThat doesn't answer my question.
08:31:54illriginalall i did was overwrite the songs.
08:31:58illriginali mean
08:32:01LloreanYour last update of Rockbox.
08:32:02illriginaljesus it's late
08:32:14illriginali just updated the buid
08:32:20illriginaland the ftw
08:32:23illriginalor whatever
08:32:31LloreanWell, the error suggests that you improperly updated the build.
08:32:36LloreanAnd I have _no_ idea what you mean by 'ftw'
08:32:48LloreanThere's not really an abbreviation even close to that which relates to Rockbox.
08:32:51illriginalthe bootloader.
08:33:04illriginalall i did was overwrite the two files.
08:33:10LloreanWell, Rockbox also has a folder.
08:33:15LloreanWhich kinda needs to be updated as well
08:33:28LloreanWhich is why we tell you explicitly to simply extract onto the device.
08:33:57combrainssorry if this is to far off topic but does anyone know of a good place to find audiobooks?
08:34:07Lloreancombrains: Project Gutenberg has some.
08:37:50 Join mirak [0] (
08:38:29illriginaloh boy...
08:39:00illriginalthe status bar.. that tells you how far into the song it is (graphical) it completely off now.
08:39:15illriginalthe time is off
08:39:33Mouser_XThat happens for MOD files, for me.
08:39:43Mouser_XOtherwise, it's works fine for everything else.
08:39:44illriginalit used to be all good.
08:39:46illriginalthe GUI that is.
08:40:30Lloreanillriginal: What theme are you using?
08:42:48illriginalipod real
08:43:39 Quit BHSPitLappy (Read error: 104 (Connection reset by peer))
08:44:00 Part Llorean
08:44:15illriginaloh but now the mp girl is working :P
08:44:35boutmuetim sorry how do you start the 3gen ipod in disk mode?
08:44:52boutmueti tried last time rockbox froze on rebooting
08:47:13 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
08:47:15illriginali just had to restart my mp3 player and now all themes work
08:47:24 Join ender` [0] (
08:47:32Mouser_XSo, you didn't restart, after updating?
08:47:46Mouser_XThat was probably one of the first things you should have done...
08:47:52boutmuetupdating what?
08:48:00illriginali had to restart twice
08:48:02Mouser_XSorry, that was @ illriginal
08:48:03boutmueti just installed and that was it
08:48:45illriginalafter the extraction into the mp3 player, i restarted but themes were messed up, then i restarted again and now themes are good.
08:49:13 Join Rob2222 [0] (
08:52:00combrainsquestion - what variations of the mod file format does rb support?
08:52:17combrainsif its in the manual then just tell me and ill read up
08:52:38LloreanMOD isn't a supported format, outside a patch
08:53:00combrainsa WIP I take it?
08:53:23LloreanI dunno how much work is being done on it.
08:53:26LloreanIt's not one I follow
08:53:33combrainsok cheers
08:53:42combrainsill read up in the tracker
08:54:26boutmuetwooo flac!
08:57:49 Quit illriginal ("Leaving")
08:59:58boutmuethow does one put their music on the rockbox ipod?
09:00:03boutmuetspecifically flac
09:01:12LloreanAny way you want?
09:01:19LloreanDrag and drop is the most basic
09:01:33boutmuetto what folder on the ipod?
09:01:46LloreanWherever you want them
09:01:47boutmuetoic the main
09:02:15boutmueti understand the 3g ipod is a little slower with rockbox but im having trouble navigating through the menus
09:02:55LloreanThere's a reason the 3G isn't in the official list of supported targets on the front page.
09:03:19boutmuetso yes basically?
09:03:26LloreanYou didn't ask a question...
09:03:31LloreanSo I'm not sure what I'd be saying 'yes' to...
09:04:08boutmuetmy question then is, menu navigation is slow because 3g ipods arent an "offical"
09:04:46LloreanMenu navigation is slow because the 3G iPods have a broken cache, and we haven't come up with a good way to work around it.
09:04:57boutmuetic, thats my question :)
09:05:22boutmuetdoes it have anything to do with original ipod firmware
09:05:37LloreanRockbox has nothing to do with the original iPod firmware
09:05:40 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
09:05:54boutmuetoh ojk
09:06:53 Quit Rob222241 (Read error: 110 (Connection timed out))
09:13:27 Join himitsu [0] (n=himitsu@
09:13:27***Saving seen data "./dancer.seen"
09:16:00markunLlorean: except the USB mode
09:16:14Lloreanmarkun: I suppose.
09:16:27LloreanBut there's still a clear line, you aren't in Rockbox any more at that point.
09:17:31markunbut saying that rockbox has nothing to do with the OF when we have to rely on them for USB..
09:17:54JerryLangewhen you enter the ipods usb function when inserting the usb what happens after you remove the usb? does it go back to rockbox? reset then go to rockbox?
09:17:56boutmuetthis 3g ipod is painfully slow with rockbox! :(
09:18:08markunLlorean: I wonder what this guy doesn't understand about the error message:
09:18:17Lloreanmarkun: Well, I meant it in a sort of no-code-sharing sense.
09:18:45markunLlorean: hey, if I would have claimed the same you would have corrected me as well :)
09:19:55LloreanYou do have a point though. :)
09:21:08 Quit BHSPitMonkey (Read error: 110 (Connection timed out))
09:21:53JerryLangeor is it called "disk mode"
09:22:38LloreanJerryLange: When you unplug the USB cable after disk mode, if it's the disk mode it automatically goes to from Rockbox, it should automatically reboot after.
09:22:41 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
09:23:59 Join sweKing [0] (i=50d835fc@gateway/web/cgi-irc/
09:24:06JerryLangeLlorean: thanks for clearing that up
09:24:10boutmuetbut seriously this kicks ass i love the levels when you listen to music, great work guys really :)
09:25:12sweKingey... How is it going with the new iPods? 80Gb? havent heard much
09:25:29LloreansweKing: You'll hear about it when it's done, at least.
09:25:53sweKingyes! of course!
09:26:17LinusNsweKing: it's on my table, really
09:26:28sweKingAA!! its done then?
09:26:35LloreansweKing: That means it's being worked on
09:26:41LinusNi am working on it, but it is far from trivial
09:26:51LinusNand my spare time is limited
09:27:22sweKingok.. The big problem. spare time
09:27:43JerryLangejust a quick question. how long did it take before a stable version of rockbox was first made on the archos players?
09:27:44directhexspare time? i donated fourty whole dollars! that's enough for full-time work!
09:28:28JerryLangei wouldnt want to be hired by you directex :p
09:28:38LloreanJerryLange: That wouldn't really be able to be used for predicting it on other players, especially the iPods which have some very big hurdles still remaining
09:28:46*petur calculates how many minutes that pays for
09:29:06LinusNJerryLange: you mean how long before we released 1.0?
09:29:13JerryLangeyeah i figured as much, but i just wanted to know. just curious
09:29:15sweKingI think im going to donate sometime. Its a nice projekt
09:29:28JerryLangeLinusN: sure
09:30:24JerryLangeor when you first said "we got sound" on the archos
09:30:47LinusNJerryLange: roughly 6 months
09:31:32sweKingse ya.. keep working ;-)
09:31:36 Quit sweKing ("CGI:IRC")
09:31:43JerryLangewere you fairly new at coding for daps then?
09:32:42LinusNfirst time
09:34:20LinusNfirst time for daps, but i have worked with embedded systems for many years before that
09:34:51JerryLangeits alot easier now isnt it? all the players use the same code for the firmware. the hard part is the bootloader when first coding for a dap?
09:36:05JerryLangethe players that are currently ported
09:36:15Mikachuuh, the firmware code is incredibly different
09:36:49JerryLangeok i was thinking of it a different way.
09:38:36LinusNyes, it is simpler in many ways because you reuse most of the application code
09:39:06LinusNbut it gets more and more complicated at the same time, since we have so many targets to care for
09:39:37LinusNevery change in the main code must work on all supported targets
09:41:30 Join My_Sic [0] (
09:43:20JerryLangeive neer heard of any other alternative firmware this big before. Also i like the discussion of how the name of rockbox was created. I'm glad it wasnt Orpheus.
09:44:08 Quit Mouser_X (Read error: 110 (Connection timed out))
09:45:22 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
09:47:18LinusNactually, i haven't heard of *any* alternative firmware that attempts to replace the original firmware
09:47:34LloreanLinusN: There's a group working on one for the s1mp3 based players
09:48:15*LinusN is browsing trying to find out what it actually does
09:48:33JerryLangethere was one. i forgot what it was called. but then to me it sounded like they gave up because rockbox was working on it. a big argument i remember in the mailing list. if not on the mailing list it was on there site.
09:49:01LloreanJerryLange: Odd, if they were going to give up, why not just combine efforts?
09:49:59JerryLangei wish i knew the site. it was a site that had beter batterties for the h120s and 140s. you guys had most of the same stuff anyways.
09:50:05 Quit combrains (Read error: 110 (Connection timed out))
09:50:06LinusNthere are numerous projects that either tries to port linux, or just run some games on the dap, but not a complete replacement
09:51:07 Quit directhex ("Leaving")
09:54:03LloreanBagder: We need an actual bootloader rom from an R series right? There's nothing useful I could get from one if I say, knew someone I might be able to borrow one from for a day?
09:54:31 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
09:54:32daurnimatorLinusN: depends on which player
09:54:44daurnimatorLinusN: join #archopen now - the head of project is there
09:54:48linuxstbLinusN: From what I've gathered, archopen provides a very complete kernel (similar to our firmware/ directory), which is used directly by the larger applications (Doom, various emulators), and there is also what they call a "shell" application which is used to launch smaller applications (.med files). I think the aim is to move all the standalone apps to .med files.
09:54:57linuxstbThere's no audio playback system though afaik.
09:55:50LinusNthe archopen faq is...sparse :-)
09:55:56linuxstbdaurnimator: I saw you link to's GPL page in the logs - have they actually released any of their own software/drivers, or just the source to pre-existing GPL'd software that they should have been distributing the source for all along?
09:56:16daurnimatorreleased their modifications for GPL code
09:56:26daurnimatornot everything - ie, no signature stuff :(
09:56:34linuxstbSo there is some interesting code?
09:56:50daurnimatorhaven't looked - i have downloaded it though
09:57:01linuxstbAh, so they give you the source, but no way to install modified versions on your hardware? (as usual...)
09:57:20daurnimatoryep :(
09:58:52LinusNnot very surprising
09:59:14 Join BHSPitMonkey [0] (
09:59:58LinusNJerryLange: are you talking about the iriverlounge guys?
10:00:45 Quit BigBambi (Read error: 104 (Connection reset by peer))
10:00:57JerryLangeLinusN: yes i believe so. thanks for reminding me i was searching misticrivers forums to find it but was having no luck
10:02:03LinusNthey weren't happy about rockbox, because they felt we "stole" their research
10:02:09 Quit dan_a ()
10:04:26JerryLangeyes i remember the discussion.
10:06:28LinusNwas quite an experience, i rarely get to play the Big Bad Evil one
10:07:18JerryLangestrange i cant even find the iriverlounge anymore.
10:07:38LinusNnope, it's gone
10:09:10JerryLangethats too bad. did they even try to help or just give up and say you guys stole there research?
10:09:20boutmuetare you guys planing on releaseing a rockbox for the zune?
10:10:42scorchewe arent not planning on it
10:10:54scorcheports happen when people who want to do them come forward
10:11:03 Quit BHSPitMonkey (Read error: 104 (Connection reset by peer))
10:11:06JerryLangeboutmuet: read here
10:12:15 Join BHSPitMonkey [0] (
10:12:45scorchemmmmm...swedish chocolate
10:13:45LloreanBah, the most important thing is where the beans come from.
10:14:14GodEater_Isn't there a swedish chocolate called plöp or something ?
10:15:17boutmuetif i decide to get a new mp3 player, which one is generally best supported by rockbox?
10:15:25LinusNGodEater: Plopp
10:15:26scorchei am eating marabou apparently
10:15:51 Join MonkeyTamer [0] (
10:15:53Lloreanboutmuet: Do you want the option to record, and or FM Radio, or is the only really important thing Music?
10:16:07boutmuetonly music
10:16:16boutmuetand space
10:16:23boutmueti have a lot of lossless music
10:16:30LloreanProbably the Gigabeat F-series then
10:16:58boutmuet:) thanks
10:19:39GodEater_or X, if you can find one
10:19:53LinusNGodEater: it will be interesting to see what princessblacklung comes up with... my guess is that it will be another "Simple guide to... with loads of Windows screenshots"
10:20:09GodEater_Yeah I was thinking the same :)
10:20:25GodEater_what a horrific nick name to choose for yourself too
10:21:01JerryLangemaybe hes smokes alot?
10:21:21MonkeyTamerthat or self-proclaimed heir to a coal mine..
10:21:27LinusNsometimes i think having difficult instructions isn't such a bad thing, to weed out the cl00bies
10:21:49LinusNbut we really need to rename .rockbox
10:22:07LinusNthe mac silliness is killing me
10:22:09GodEater_because OSX is dumb
10:22:45amiconnWhy care about macos when most devs don't care about archos? </sarcasm>
10:22:46LinusNthose mac cl00bs don't know how to unzip, and OS X isn't exactly helping them out
10:22:52LinusNamiconn: lol
10:23:03GodEater_oooh I felt the claws on that
10:23:17*amiconn fails to see the fun in it
10:23:26LinusNi couldn't care less about the mac losers, but they are a pain in the ... because they are seeking support all the time
10:23:52JerryLangewhat would you change it to then if you did?
10:24:00amiconnI like the fact that .rockbox is hidden
10:24:05LinusNamiconn: me too
10:24:17GodEater_it could still remain hidden on the player though
10:24:23JdGordonIts good hidden on the player, but hidden on the comp is annoying
10:24:27LinusNwe *could* of course create some kind of installer app
10:24:48JerryLangei agree with amiconn and LinusN
10:25:01LinusNthe problem is that mac users don't get to choose where to unzip files
10:25:19JdGordonforce mac users to use rbutil?
10:25:59 Join fejfighter [0] (
10:26:18JerryLangewould be nice if it was like most installers. There are check marks for certin stuff you want.Tthat would be really nice.
10:27:12MonkeyTamerI wish all installers had that option... I personally think HP should be banned from making drivers or installers for that reason
10:28:15LinusNamiconn: regarding the lack of care for archos, we tend to care most for the platforms we actually use. i haven't used my archos for years.
10:28:55LinusNmost of the devs here have never even seen an archos player in real life
10:29:41LinusNso we need a plan for how we want to continue the work to make rockbox run on the archos
10:29:59GodEater_I thought it did run, I thought the only issue was rombox ?
10:30:09 Join B4gder [0] (
10:30:12LinusNwhich imho will involve stripping features sooner or later
10:30:36LinusNGodEater: yes, but the RAM usage matters too
10:30:50GodEater_that's really low too huh ?
10:31:05scorcherebuffers quite often
10:31:59GodEater_at the risk of being branded "heretic", would it be worth conducting some sort of poll of our user base (those that bother to talk to us) and see what the numbers are on most used players ?
10:32:12GodEater_or I guess we could do something with the download stats of the images ?
10:32:55GodEater_it would seem, erm, "frivolous", to me to expend lots of effort on something which not a great deal of people use
10:33:07*GodEater_ hides from amiconn
10:33:21B4gderwell, this is not a democracy
10:33:25scorcheoptimizing helps all targets
10:33:31LinusNyes it does
10:33:52safetydanGodEater_, I think the idea was floating around that we get rombox working again (though various means, lang cleanup, menu recoding) and then tag the source at that point
10:34:27LinusNBagder and I suggested a fork of the Player some time ago
10:34:28GodEater_but is there anyone actually *doing* that? (apart from JdGordon with his menu clean up)
10:34:30safetydanFrom there... well I guess a separate "low mem" branch could be made
10:34:39*safetydan raises hand
10:34:51safetydanwell I started doing the lang cleanup anyway. Been distracted by speex lately
10:34:54GodEater_I guess I just have low visibility of that then :)
10:36:07LinusNthe problem is that optimizing memory usage of a player that you don't even own is often not as fun as doing stuff that enhances the player that you do own
10:36:43LinusNso what will you spend your precious spare time on?
10:36:45B4gderand the amount of devs caring for the Archos models are now very fast approaching: 1
10:36:59GodEater_that's been my observation
10:37:10scorchedo i count? =(
10:37:20B4gderscorche: shhh, you break my math
10:38:04 Quit jhMikeS (Nick collision from services.)
10:38:06 Join Vyrus [0] (
10:38:10 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
10:38:16linuxstbI don't think we've reached the end of rombox yet though. There's the menu/settings cleanup, the langv2 work, plus the bootbox shrinkage which should give rombox back. At that point, I think we should at least make a new Archos release.
10:38:28LinusNi agree
10:38:44LinusNhow far away is langv2 btw?
10:39:12safetydankind of hard to tell really
10:39:17 Quit Vyrus001 (Read error: 104 (Connection reset by peer))
10:39:18LloreanB4gder: We need an actual bootloader rom from an R series right? There's nothing useful I could get from one if I say, knew someone I might be able to borrow one from for a day?
10:39:36B4gderLlorean: we need to get e200tool to run against one
10:39:52B4gderusing that, we might be able to extract the BL
10:39:59safetydanLinusN, there's one bug with the lang scripts that's sort of holding things up a bit
10:40:18LloreanB4gder: So, me being 'relatively clueful', if I could borrow one would this be something I could likely try to do?
10:40:28B4gderLlorean: yes
10:40:33*LinusN is slightly amused by his last incoming PM in the forums: "i am really ??? i realli need help and major i dont understand i followed evry thing it dad but it still didnt work help"
10:40:43LinusNand that was all
10:40:44 Join barrywardell [0] (
10:40:46LloreanB4gder: Alright, I'll try to get my hands on it then.
10:40:48B4gderdad? ;-)
10:41:08 Join tucoz [0] (i=528612c1@gateway/web/cgi-irc/
10:41:11 Quit pill (Nick collision from services.)
10:41:26LinusNi did some research, and it is another mac luser failing to unzip rockbox
10:41:31 Join _pill [0] (
10:41:36B4gderLlorean: I guess you've seen that others (well at least one person) have failed to run e200tool on the R model
10:41:55tucozI remember i filed a bug report, maybe a year ago about A-B repeat not working on ipods. I can not find this bug anymore. Is it fixed?
10:41:56LloreanB4gder: Yeah, but the worst that can come of it is that I'll fail too.
10:42:26Lloreantucoz: I believe so. Try it. :)
10:43:02tucozsorry. no ipod here :)
10:44:37tucozi think i missed that fix. wonder if the manual is updated then. I think it was that the menu item was present, but it was unpossable to set the markers
10:44:42*petur wonders why certain OSses need to use '.name' to hide a file/dir when there's a 'hidden' attribute in the filesystem
10:44:57linuxstbBecause there isn't a hidden attribute?
10:45:02B4gderpetur: because it makes it work without fs attributes
10:46:05LinusNin the mac case, the "." is not the real problem
10:46:30LinusNthe problem is that the unzip applications don't let you choose where to unzip the files
10:46:35 Quit MonkeyTamer ("Chatzilla 0.9.77 [Firefox]")
10:46:35linuxstbLinusN: Another Mac/Finder issue seems to be that dragging one folder over the top of another will delete the original folder, not replace the changed files.
10:47:06safetydanIt'd be nice if Rockbox support configs outside of .rockbox
10:47:12*petur wonders why mac users laugh at windows
10:47:20safetydanI mean browsing for configs outside of .rockbox
10:47:22LinusNi wouldn't expect the OS to do any of those
10:47:28GodEater_petur: because they don't know any better
10:47:52LinusNi would expect the os to copy the folder to the one you dragged it onto
10:48:18tucozsafetydan: isn't that possible? i have a config in the root that i can use after an update
10:48:47LinusNsafetydan: rockbox doesn't care where the cfg files are
10:49:18B4gderbtw, I think there's still some requirements about certain files being in certain dirs for them to survive reboots
10:49:20tucozLlorean: is this information correct then?
10:49:40tucozthe A-B repeat actions
10:49:48LinusNB4gder: yes, that is true, but config files aren't remembered in that way, are they?
10:49:54Lloreantucoz: Honestly, I don't know.
10:50:10B4gderLinusN: no,this wasn't directly related to that issue
10:50:21B4gderI was more thinking we shouldn't need to do this
10:50:28B4gderwith the text based configs
10:50:46LinusNthat is true
10:50:51safetydanLinusN, tucoz, I actually meant reading fixed.cfg from outside of .rockbox
10:50:56*safetydan should learn to type what he means
10:51:10Lloreansafetydan: fixed.cfg seems like one of the two .cfg files that really should be hidden in there.
10:51:20LinusNwell, the location of the initial cfg file must be fixed
10:51:33linuxstbHopefully rbutil will help mac users - it does the unzipping directly. But I haven't yet managed to compile it successfully in Mac OS X.
10:51:39tucozAnyone with an ipod that can test if A-B repeat works? (following the instructions from
10:53:44JerryLangeanybody watch the video of that firmware on ? they previewed everything but what i wanted to really see.
10:54:13LloreanSoap: Doesn't seem to work for me.
10:54:19LinusNme neither
10:54:41barrywardelllinuxstb: what language is rbutil written in?
10:55:05GodEater_C with WxWidgets I believe
10:55:07tucozgreek ;)
10:55:07barrywardelland what have you tried wrt getting it to compile on OS X
10:55:26 Join norbusan [0] (
10:55:33 Part norbusan
10:55:39amiconnLinusN: I don't know how you are using your various rockboxes, but for me rockbox on archos is still far more usable than on any swcodec target.
10:56:00amiconnAnd since I don't really need the extra features of the swcodec targets, the choice is most often obvious
10:56:01B4gderhm, jdgordong added a warning
10:56:06LinusNamiconn: in what respect?
10:56:10B4gder64bit related I think
10:56:18LinusN(i mean how is it more usable)
10:56:19amiconnLinusN: It's stable
10:56:46amiconnStable when using it for its primary purpose: playiung music
10:57:08B4gderI have no problems with Rockbox on the swtargets I use
10:57:12LinusNi find my iriver and x5 quite stable too when playing music
10:57:12GodEater_nor me
10:57:31amiconnThere is still this really annoying boost issue
10:57:35GodEater_I can't remember the last time the iRiver did something "unstable"
10:57:44amiconnBoost count is 1 after boot, and sometimes even 2 or 3
10:58:30directhex|worki think i managed to crash my gigabeat with a dodgy plugin, but that's about it
10:58:51tucozLlorean: <Llorean> Soap: Doesn't seem to work for me. <−−- was that for me?
10:59:12GodEater_Gigabeat is an immature target though directhex|work
10:59:15Lloreantucoz: Yeah
10:59:16amiconnApart from stability issues like this, the archos sound is simply better
10:59:18LloreanI apparently can't type
10:59:20tucozok. thanks :)
10:59:25LinusNamiconn: i see that as an argument to fix the bugs in those targets, not an argument for working on the archos targets
10:59:41amiconnI'm not saying this is a rockbox problem, it's a hardware problem of the dacs in the newer devices
10:59:41tucozthen i'll open up another bug then
10:59:44directhex|workGodEater_, hey, how many times do you think my karma crashed with supposedly stable firmware? i'm not complaining
11:00:01LinusNi agree that the sound on the archos recorder is really good
11:00:03GodEater_directhex|work: I used to own a karma too - so I know exactly wha tyou mean
11:00:07B4gderdarn, my size table fixes didn't work
11:00:13amiconnLinusN: If I would understand enough of the code in question, I would do so
11:00:15LinusNi don't think anyone has said that the archos is bad
11:00:17directhex|workGodEater_, truth be told, i really don't give a shit about plugins
11:00:50directhex|workthe archos looks ugly!
11:00:53B4gderI plan to make each build round display how the sizes change from the previous commit
11:01:03LinusNB4gder: only delta?
11:01:28B4gderwell, I'll store the sizes for each build round in their own files
11:01:43B4gderso you will be able to see them
11:01:59B4gderbut the delta is what I want to show in the table
11:02:16LinusNi want to see how far from rombox we are
11:02:27B4gderI can't show that atm
11:02:30 Join thegeek [0] (
11:02:31amiconnAnd I have the impression that people who do understand enough of that prefer working on esoteric features over fixing the basics :/
11:02:33LinusNthen do that! :-)
11:02:37B4gdersince that would require the map file and filtering of that
11:02:42 Quit barrywardell (Remote closed the connection)
11:02:50 Part tucoz
11:02:55LinusNamiconn: yes, it's called "rubbish" :-P
11:03:17LinusNamiconn: and i do share your opinion
11:03:52directhex|workam i the only one whose archos suffered disk failure after a couple of years?
11:03:59amiconnThere's another limit for archos I'm beginning to worry about
11:04:18LinusNthe voice buffer?
11:04:44amiconn(recorder v1 and a bit later player)
11:04:50LinusNdirecthex|work: i haven't had any disk failure on mine
11:04:51amiconnNo, the 200KB limit, again
11:04:58LinusNamiconn: aha, yes
11:05:04 Join JerryLange_ [0] (
11:05:34amiconnSpeaking about the player, I asked you to try something...
11:05:39B4gderwell, there's some 40K left to 200
11:05:51 Join t0mas [0] (n=Tomas@rockbox/developer/t0mas)
11:05:54t0mashi! :D
11:05:58t0masfinally back
11:06:07B4gderhey who are you? :-P
11:06:40t0masyour worst nightmare ;)
11:08:04GodEater_that's a bold claim, I have some pretty bad nightmares....
11:08:19LinusNamiconn: ah, yes, the ata poweroff
11:08:27LinusNamiconn: forgot about that one
11:08:33 Quit boutmuet ()
11:09:04LinusNwill try tonight
11:09:25amiconnShouldn't take long, I think
11:09:32LinusNguess so
11:09:38LinusNi just have to remember to do it
11:10:09amiconn..and I will probably get rid of one of my (atm) uncommitable changes
11:10:18LinusNah yes
11:10:40 Quit My_Sic ("Miranda IM! Smaller, Faster, Easier.")
11:10:44*amiconn has #define HAVE_ATA_POWEROFF uncommented in config-player.h
11:11:15scorcheamiconn: do you have any idea of what would cause (codewise) a v1 to take 3 seconds to change a song? (larger amount of time when you have a folder with a lot of files in it)
11:11:39scorchei have experienced that on 3 different v1s....archos firmware does not have that effect
11:11:45amiconnSisk spinup?
11:12:07B4gdersounds likely
11:12:08scorchepossibly....but why would the OF not have that gap?
11:12:22scorcheit still has to spin the disk up
11:12:26amiconnI never experienced this- My largest folder contains 200something tracks
11:12:53scorcheso when you skip a track, it immediately goes to the next?
11:13:28***Saving seen data "./dancer.seen"
11:13:30amiconnIt does unless it has to spinup the disk first
11:13:49scorcheyeah....i know it is related to spinup
11:14:11scorchebut the thing is, i dont knwo why the gap is there in rockbox, but not near as large in OF even when it has to spin up
11:14:46scorcheand it seems directly connected to the number of files in the dir
11:15:03amiconnThere shouldn't be a gap in rockbox. It should keep playing the old track until the new one is ready
11:15:14amiconnI don't skip often, so I'm not 100% sure
11:15:38 Join BigBambi [0] (
11:16:03linuxstbbarrywardell: (If you read the logs) - rbutil is C++ using wxwidgets 2.8.0. I've installed wxwidgets, but I'm getting errors compiling rbutil. It could be because the Mac port of wxwidgets is still relatively new, and the widgets rbutil uses aren't implemented. Or I could have just configured wxwidgets wrongly (i.e. not compiled all the needed modules).
11:16:06scorchei dont seem to remember it to keep on playing
11:16:33amiconnI'm not using rockbox that much atm - the radio in my new car doesn't have aux; I need to do something about this...
11:16:35scorchebut is there possibly something that we are doing when finding the file inside the dir that is taking longer than it should?
11:17:38directhex|workthe lack of aux in my car is maddening
11:17:45directhex|workat least fm transmitters are legal now
11:18:28*B4gder has a new car with a fine aux... :-)
11:18:48 Join markun_ [0] (i=c2fe7883@rockbox/developer/markun)
11:18:51*GodEater_ uses a casette convertor in his car
11:18:55JerryLange_in my older car i used the input on my amp as my aux for my mp3 player.
11:19:08B4gderI did the casette convertor before, it worked quite fine
11:19:22GodEater_yeah - I think it's pretty near perfect
11:19:38GodEater_fm transmitters are crappy by comparison
11:19:47markun_Does anyone run a ssh server on port 443 and can give me an account? (just to start my own sshd on that port)
11:20:19markun_stupid proxies..
11:20:21amiconnGodEater: It's a CD radio.
11:20:51GodEater_markun_: PM
11:20:59amiconnI need to send in my Blaupunkt dab radio for repair, and then put that one in my new car
11:21:26directhex|workmy car has no aux connector, no tape hole, and can't be replaced with any other cd unit. bah.
11:22:34peturjhMikeS: around?
11:24:01 Quit JerryLange (Read error: 113 (No route to host))
11:24:27 Join Mouser_X [0] (
11:25:40JerryLange_woohoo JerryLange finally left
11:25:45 Nick JerryLange_ is now known as JerryLange (
11:26:02GodEater_JerryLange: you know you can kill with nickserv's GHOST command right ?
11:26:14 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
11:26:31JerryLangeno i didnt. how? i think i tried it once but failed at it.
11:27:03 Quit markun_ ("CGI:IRC 0.5.7 (2005/06/19)")
11:31:45GodEater_JerryLange: /msh nickserv help ghost
11:31:50GodEater_ /msg even
11:34:18JerryLangeThank you GodEater
11:35:04LinusNamiconn: can you tell me more about that boost issue?
11:39:24pixelmaI don't know if it was talked about earlier but there is still a 64-bit warning left from JdGordon's commit in the sims (I have to admit that the 2 others that come from the faulty gcc hide that...)
11:39:31 Join wooo [0] (
11:42:18amiconnLinusN: As saoon as a voice file is present (voice doesn't need to be enabled), the CPU is boosted after boot
11:42:26LinusNaha, ok
11:42:55amiconnMost of the time this is a single boost (count == 1), but I also observed higher boost counts more than once
11:43:29markunI had a strange problem with my Gigabeat yesterday. I was looking at the audio thread while playing some SPC files and noticed the PCM buffer only filled for a few %
11:44:09LinusNmarkun: low latency?
11:44:14markunI thought it was normal, but after that I played some Ogg Vorbis files and it started skipping. I looked at the audio thread and did the same thing.
11:44:20directhex|workmarkun, did you get the url for my settings reset patch?
11:44:20markunAfter a reboot all was fine.
11:44:25 Quit JerryLange ("Leaving")
11:44:35markunLinusN: Is there a way to trigger it by accident?
11:44:49LinusNi dunno
11:45:04markundirecthex|work: it's in the logs somewhere, right?
11:45:24directhex|workmarkun, should be. look for gigabeat_reset
11:45:56markunLinusN: but yes, sounds like the low latency mode. I will investigate it a bit more.
11:51:33 Join Thundercloud [0] (n=thunderc@
11:53:20markunLinusN: perhaps if there is a way to leave the sound menu without triggering the ACTION_EXIT_MENUITEM
11:53:52markunlet me try
11:55:16markunno, that didn't work
11:56:35 Quit Tom0473 ("[G]Script - v3.3 ©daGroove - -")
11:56:37markunah yes, it did :)
11:56:38 Join Tom0473 [0] (
11:57:19markunLinusN: I though just entering the sound menu was enough, but it's if you enter one of the subitems and then insert the USB cable
11:57:33LinusNmaybe there are other ways to exit without ACTION_EXIT_MENUITEM
11:58:28markunIt was already very strange that the 300MHz ARM9 couldn't keep up with the SPC playback :)
11:58:44directhex|workmarkun, seems that zune port is needed - 540mhz!
12:00:48markun(or Gigabeat S)
12:01:10preglowjhMikeS: so the wavpack updates from about five weeks ago makes encoder compression ratio suck? i'd tell david bryant about that if so
12:01:14 Quit kaaloo ("Leaving.")
12:01:43 Join kaaloo [0] (
12:01:43 Quit fejfighter ()
12:02:16directhex|workmarkun, same difference
12:02:23 Join MonkeyTamer [0] (
12:02:34directhex|worki still need to find some kind of case or pouch for my gigabeat, it's getting filthy :/
12:03:04*GodEater_ has an iskin for his gigabeat in his coat pocket.
12:03:19GodEater_it's a bit of a tight fit though, which makes the cross sensor on the front SUPER sensitive
12:03:34markundirecthex|work: would be nice if we could find a digital camera or mobile phone battery with some more mAh to put inside..
12:03:42 Quit wooo (Read error: 110 (Connection timed out))
12:04:20markunGodEater_: could you take a picture of it?
12:04:22directhex|worki've got no battery life complaints. should i?
12:04:34markundirecthex|work: no, but more is always nice :)
12:04:40GodEater_markun - certainly
12:07:13 Join webguest02 [0] (i=52e53337@gateway/web/cgi-irc/
12:07:37 Quit webguest02 (Client Quit)
12:08:44*GodEater_ tries to get bluetooth working
12:09:21directhex|workGodEater_, working in what way?
12:09:37GodEater_working so I can transfer the pics I just took from my phone to my pc
12:10:31Llorean GodEater_ SUPER sensitive? I kinda felt the cross was fairly insensitive naturally
12:10:42directhex|workaptitude install gnome-bluetooth && gnome-obex-server
12:11:00directhex|workymmv. ianal. bbq.
12:11:28GodEater_Llorean: I don't find that when the gigabeat is out of it's case - I find the cross about as responsive as I'd expect - in that I don't find it difficult to interactive with
12:11:42GodEater_but putting the case on makes it respond to the slightest brush of your fingers
12:11:45GodEater_which is a PITA
12:11:59LloreanGodEater_: See, I expected it to respond to brushes, and I feel like I nearly have to press them like buttons.
12:12:13GodEater_Llorean: I just tap it...
12:12:26LloreanI'd have to say more than a tap, but less than a press for me.
12:12:30directhex|worki'd settle for some kind of pouch. like the ones minidisc players (and karmas) always used to come with.
12:12:38LloreanBut my fingers don't seem to pick up on touch-sensitive surfaces very well at all
12:12:43markunThere were plans to calculate the pressure by the number of sensors pressed at the same time
12:13:16markunGodEater_: maybe then we can add a sensetivity level setting for you
12:13:44GodEater_that would be cool
12:14:19 Quit B4gder ("Rockbox TCP exception #04123 - connection reset and user kicked out")
12:14:29GodEater_hmm haven't tried bluetooth under linux - looks like I'm going to have to - the windows machines here are all XP SP1, so the new devices don't work with them :(
12:16:11directhex|workbluetooth under windows is unending pain
12:16:32directhex|workthanks to hard work by broadcom, mainly
12:16:42GodEater_Llorean: it seems to work just as well as a touchpad mouse on laptops do
12:16:46GodEater_which I get on with just fine
12:17:06LloreanGodEater_: I definitely don't. Some days I can't get them to respond at all.
12:17:07 Join Nico_P [0] (
12:17:48directhex|workdo you have greasy, wet, or cold fingers? touchpads tend to dislike any of the above features
12:18:01GodEater_you must have an unusually low body electric field ;)
12:18:22Lloreandirecthex|work: Actually, they tend to be quite dry.
12:18:33*GodEater_ has sweaty paws
12:18:37LloreanSometimes dampening them, and rubbing them together for a bit helps.
12:19:00directhex|workodd. import a chinese gold farmer and get him to press buttons for you?
12:19:07LloreanHahaha, good plan.
12:19:30 Part Llorean
12:19:35 Join Febs [0] (
12:19:53GodEater_neat - I appear to have already included bluetooth support in my kernel. I must be psychic
12:20:23LinusNamiconn: on my x5 the boost counter stays at 1, but only if Voice Menus is On
12:20:45directhex|workGodEater_, did you compile it yourself?
12:20:56LinusNfor me, it is not sufficient to have the voice file present
12:23:10LinusNthere is a call to trigger_cpu_boost() last in the cpu boost log, which is never unboosted by the kernel
12:23:34GodEater_directhex|work: of course - I'm a gentoo man ;)
12:24:24directhex|workGodEater_, i thought the point of self-compiled kernels was to avoid wasting disk space with support for hardware you don't have at that particular instant? O_o
12:26:24GodEater_"wasting disk space" is a bit much on a machine with an 80GB hard drive when you're talking a few hundred kb for a module or two...
12:27:12GodEater_also (whilst I wouldn't quite call myself a gentoo ricer) you DO get to make sure the kernel is compiled for your specific hardware arch - rather than just "i386"
12:28:06markunGodEater_: I have just though up a way to do the pressure calculation, I'll let you know if I have a patch and if it works well.
12:28:13GodEater_neat :)
12:28:35directhex|workGodEater_, i work on the basis that $distro's opteron is unlikely to be wildly different from my athlon64, other than things that are determined at runtime anyway
12:28:41 Join XavierGr [0] (
12:31:48 Join Thundercloud_ [0] (n=thunderc@
12:34:09 Quit Thundercloud (Read error: 60 (Operation timed out))
12:34:36 Quit safetydan ("Ex-Chat")
12:38:39GodEater_markun - got the pics on the PC now - where would you like me to send them ?
12:39:08markundo you have a webserver running?
12:40:08GodEater_tell you what
12:40:14GodEater_I'll copy them into your home directory ;)
12:46:30GodEater_not bad for mobile phone pics, even if I say so myself
12:47:02 Join Vague_Rant [0] (n=opera@wikipedia/vague-rant)
12:47:15Vague_RantOK, this isn't really a Rockbox question, and I apologize for that.
12:47:25Vague_RantBut how the hell do I get to the menu on the original H300 firmware?
12:47:33Vague_RantAs in the bit where I choose video, text, etc.
12:47:42directhex|workhold the navi button iirc
12:48:09Vague_RantThat seems to take me to the settings.
12:48:25 Join mathgl [0] (
12:49:38Nico_PVague_Rant: i think it's with the REC button
12:49:46Nico_Pprobably long press
12:50:00Vague_RantAh, that did the trick. Thanks very much.
12:50:08Vague_RantLucky Rockbox is more intuitive.
12:50:13Vague_RantAppreciate it.
12:50:44directhex|worki've found most korean players i've used to be extremely counterintuitive
12:51:35Nico_Pmarkun: does the gigabeat's CPU have a FPU ?
12:54:25*JdGordon is back
12:55:07 Part Vague_Rant
12:55:24 Join lukaswayne9 [0] (
12:56:23*JdGordon wonders why that warning isnt elsewhere also
12:57:22Nico_PJdGordon: the one in the ondio FM and mini 2G builds ?
12:57:26pixelmait's on 64-bit servers?
12:57:46JdGordonyeah, but im sure i use that elsewhere..
12:58:06Nico_Ppixelma: doesn't look like it
12:59:02JdGordonyeah, the ondio build is on amiconn's server.. thats one of the 64bits isnt it?
12:59:38Nico_Pwhat about ?
13:00:08JdGordonno idea
13:00:20Nico_Pprobably 64 bit too
13:01:53Nico_PJdGordon: does the number of header files included in a C file have an effect on binary size ?
13:02:04JdGordonI dont think so
13:02:48JdGordonmarkun: is there a problem with the sound menu callback and setting the low latency?
13:05:33*JdGordon isnt sure what the least-hacky way to fix this warning is...
13:05:34 Quit Mouser_X (Nick collision from services.)
13:05:43 Join Mouser_X [0] (
13:07:27 Join Xerion_ [0] (
13:07:27JdGordoneveryone off to lunch are they?
13:08:22 Join BigBambi_ [0] (
13:08:44LinusNnot me
13:08:59amiconnJdGordon: I think it's enough to cast to intptr_t instead of int
13:08:59 Join barrywardell [0] (
13:09:11 Join pondlife [0] (
13:09:32JdGordonamiconn: if you get a chance, could you test that please?
13:09:55 Quit pondlife (Client Quit)
13:10:02amiconnLinusN: That's tsrange. I get this boost even when booting with all voice options disabled, just my 2 usual voice files present
13:10:14 Quit Xerion (Read error: 104 (Connection reset by peer))
13:10:14 Nick Xerion_ is now known as Xerion (
13:10:22*JdGordon noticed there has been a boost at boot for a while now
13:10:59JdGordonenable logf build and check the boost tracker in debug... the stray boost is somehwere not very helpful irrc
13:11:16LinusNi get the permanent boost just by enabling Voice Menus
13:12:06JdGordontypical... i just checked and its not boosted... but voice is enabled
13:12:21JdGordon(not that it should make a difference.. but this is with the root menu patch)
13:13:19amiconnI get this boost at boot practically always
13:13:19directhex|workfairly offtopic, but has anyone ever used an iaudio u3?
13:13:30***Saving seen data "./dancer.seen"
13:13:54amiconnCan't check atm, but y'day I got a boost count of 3 (!) on my mini2g after boot
13:14:13LinusNdirecthex|work: i'm sure *someone* has used one, it's a big world after all
13:14:46directhex|workLinusN, okay, "anyone with standards when it comes to portable players, i.e. current channel residents"
13:15:35directhex|workactually, forget the "standards" bit, it's full of ipod users in here! badum-tish!
13:23:06 Quit jhMikeS (Nick collision from services.)
13:23:12 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
13:23:48 Quit BigBambi (Connection timed out)
13:24:01 Join Bagder_ [0] (
13:28:25LinusNdoes anyone know why the voice thread boosts when receiving Q_VOICE_PLAY?
13:29:03 Join Surger [0] (
13:29:13GodEater_do we have a part of the wiki dedicated to custom bootsplashes in a similar way to WPS themes ?
13:30:55JdGordonwe have custom bootsplashes?
13:33:37 Quit secleinteer (Remote closed the connection)
13:34:26 Quit kaaloo ("Leaving.")
13:34:39 Quit GodEater_ ("CGI:IRC (Ping timeout)")
13:35:19MikachuJdGordon: gigabeat added them without asking anyone iirc
13:35:34Mikachu(only to gigabeats)
13:36:20 Join secleinteer [0] (
13:38:08 Quit jhMikeS (Nick collision from services.)
13:38:14 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
13:40:32SurgerIs it my imagination or do MP3 player manufacturers deliberately cripple the recording quality in most players?
13:41:19 Quit Bagder (Read error: 110 (Connection timed out))
13:42:19 Nick Bagder_ is now known as Bagder (
13:46:30preglowas in how?
13:50:22SurgerWell for starters I cannot find an MP3 player that records at more than 160 kbps and this seems to be a firmware limitation - not a hardware limitation.
13:50:43SurgerI mean 160 kbps MP3
13:50:58preglowiriver did allow that, at least
13:51:01SurgerAnd even that ones that support WAV recording only go up to 32 kHz
13:51:11preglowthen it is without doubt deliberate, yes
13:51:38SurgerMost of iRiver's new products no longer have line-in.
13:52:28SurgerAnd the Sony Minidisc players can record high quality but don't allow you to dump it in digital form to a PC - you have to record with analogue from the device.
13:52:48SurgerScrew the MPAA, RIAA and all those bastards
13:53:02SurgerI'm trying to record music lessons ...
13:54:02 Join bluebrother [0] (i=810d4899@gateway/web/cgi-irc/
13:56:00Nico_PLinusN, amiconn: what do you think of the progressbar-y-coord patch ?
13:57:09directhex|workSurger, how does the recording industry know that? safer to assume you're illegaly recording songs from the radio
13:57:12LinusNmy general opinion on WPS patches nowadays is "sure - go ahead" - until we implement viewports
13:57:30petuririver 320kbps recording sound worse than 128kbps lame encodings (for me)
13:57:45LinusNhowever, i wish someone could have a go at theming the rest of rockbox, for example the status bar
13:57:57Surgerpetur: Yeah I've heard that. :-(
13:58:14directhex|worki'd still like themable icons next to menu entries. but i'm an eye candy whore
13:58:19LinusNi'm sure rockbox mp3 encoding is even worse
13:58:25Nico_PLinusN: ok
13:58:36LinusNdirecthex|work: i want that too
13:58:42Nico_Phow could the statusbar be themeable ? custom bitmaps ?
13:58:53LinusNNico_P: but i haven't seen that particular patch, though
13:58:56JdGordonLinusN: I themed the list widget.. but had a problem getting text to display over a bmp
13:59:03LinusNNico_P: yes, custom bitmaps
13:59:03JdGordonso gave up and deleted the changes
13:59:17amiconnJust bitmaps won't work
13:59:17LinusNwhy over a bmp?
13:59:30bluebrotherwas there some discussion about the plugins sorting thing recently?
13:59:35JdGordonbecause I had the ability to display a background image for each line of text
13:59:39LinusNbluebrother: check the wiki
13:59:39amiconnThe volume and battery status icons are drawn
14:00:02bluebrotherI did, I was just wondering if there was some more recent discussion here ;-)
14:00:06Nico_Pit's a rewrite of gl's patch ( FS #4783)
14:00:08*JdGordon wishes patches adding lots to the bin size would wait for rombox to get under control again...
14:00:11bluebrotherI might try the folder based approach
14:00:26LinusNNico_P: looks ok to me
14:00:28*pixelma thinks the same as JdGordon
14:00:36bluebrotherat least if I find time over the weekend :)
14:00:43LinusNJdGordon: i agree
14:01:29bluebrotheras it adds only one more check when filtering this is the best solution imo regarding code size (and I don't see any advantages of the other proposals)
14:02:35*JdGordon is doing the display menu now.. so hopefully we get a decent bit back
14:02:48directhex|workassign metadata to plugins, and allow browsing by metadata!
14:02:53Nico_PLinusN: i think i'll commit it and add y coords to the rockboxed WPS progressbars... it will clean up the bitmaps a bit (and fix the bad markers with cuesheets)
14:02:59LinusNthe voice boost problem is an effect of the system where the kernel handles the unboost
14:03:15LinusNNico_P: what bitmaps?
14:03:33bluebrotherJdGordon: the new icons in the menus are toggled by the setting in file view / show icons. I think that should get moved to display / show icons now
14:04:03preglowLinusN: isn't it more a case of something trying to boost/unboost manually and forgetting to unboosT?
14:04:04bluebrotherat least it doesn't make too much sense there anymore ;-)
14:04:06*amiconn wonders where [IDC]Dragon vanished again
14:04:13preglowLinusN: i don't see how the kernel can "forget" to unboost
14:04:15Nico_PLinusN: the progressbar bitmaps ("pgln-*.bmp") are bigger than needed to have better alignment
14:04:16amiconnHe was working on the cut-down bootbox
14:04:30LinusNpreglow: no, the actual boosting is done with a trigger_cpu_boost() call
14:04:38preglowthen hooray
14:04:51preglowSlasheri: time for the weekly "update RockboxKernel" nag :)
14:05:08JdGordonbluebrother: where in the display menu order?
14:05:27JdGordonpreglow: hehe weekly.. thats at least the 3rd ive seen since monday
14:05:43LinusNthe problem is that the voice thread uses queue_wait() when it has finished playing a voice, but queue_wait() doesn't unboost, only queue_wait_w_tmo() does that
14:06:00preglowLinusN: ah, yes, in these cases boosting should be done manually, i think
14:06:14preglowLinusN: at least that's what slasheri said
14:06:44pregloweither that or queue_wait should unboost, but there was a pretty good reason for that not doing so
14:07:02LinusNi don't understand why the voice thread would need to boost at all
14:07:12markunNico_P: the S has a FPU, but not the ones rockbox runs on
14:07:21Nico_Pmarkun: ah
14:07:29directhex|workNico_P, what do you want to do with an FPU?
14:07:36preglowhave an easier life?
14:07:53Nico_Pdirecthex|work: nothing in particular, just the perspectives ;)
14:07:54preglowit's good for hairloss
14:08:02LinusNpreglow: assuming you have a life to begin with, of course :-)
14:08:36markunJdGordon: yes, did you read about the problem with the low_latency setting and USB?
14:08:53directhex|worki wonder whether quake would run on suitably high-spec targets
14:09:08markundirecthex|work: I think it will
14:09:13preglowme too
14:09:17preglowbut it will need an fpy
14:09:18JdGordonmarkun: yeah, I think (hope) that unsetting that if usb comes in set_cound() would fix it...
14:09:51JdGordonmarkun: assuming set_sound is used for all the settings that is a problem with...
14:09:52directhex|workpreglow, funny thing is, there's a playable port to the gamepark gp32 handheld, (based on pocketquake) which is a 320x240 133mhz arm9 device
14:09:57 Quit SirFunk (Read error: 60 (Operation timed out))
14:10:00Nico_PLinusN: you said you were in favor of WPS patches until viewports arrive... is this also true for scrolling margins ?
14:10:11LinusNguess so
14:10:23directhex|workgive or take some mhz
14:10:24LinusNbut......... maybe we should consider code size a little
14:10:44JdGordonwe were soo close... :'(
14:10:48bluebrotherJdGordon: I don't think that matters much ... maybe directly above the "scrolling" entry?
14:10:57LinusNand i am not the only one with a vote in this matter
14:11:05pixelmaNico_P: pleasa calm down a bit and consider what JdGordon said... :/
14:11:14pixelma*please too
14:11:30Nico_Pi know, i wasn't intending on committing it, it's too big
14:11:45Nico_Pthe y coord one is very small
14:11:59bluebrotheris the scrolling margin that big?
14:12:18LinusNnot big, but it adds to the code size
14:12:20Nico_Pbluebrother: well the patch file is big :) not sure about the effect on bin size though
14:12:20preglowi'd rather we not commit those now, really, too much of a chance the current wps system will end up staying even though everbody thinks it should be redesigned
14:12:23pixelmathere are two: scrolling margins and partial-scroll
14:12:48JdGordonif anyone is bored and wants to help on the filesize front... please grab the root_menu patch and fix the dirborwser() and radio screen functions to work nicer... should save a few bytes with them fixed
14:12:53preglowbut of course, the alternative might be nothing happens with it at all...
14:13:49pixelmathen don't demotivate amiconn too much ;)
14:13:52*LinusN recalls the metadata-in-buffer situation
14:13:57JdGordonmidkay_: you still around?
14:14:03JdGordonhaha pixelma
14:14:30LinusNwe can't have viewports - it adds too much to the binary size :-P
14:14:43LinusNwhat a dilemma :-)
14:14:55bluebrothercut the wps for it! :)
14:15:31pixelmaJdGordon: it's not too funny... I myself don't too much motivation to even try standard SVN on my OndioFM...
14:15:39pixelma*don't have
14:16:11markundirecthex|work: yes, I also saw that quake port with fixed point optimizations
14:16:17markundo you want to give it a go?
14:16:24preglowthat's one reason to not use it! lots of fancy stuff happening over on the ipod/iriver side here ;)
14:16:26*JdGordon looks around for someone who can draw an icon for the lcd remote for the menus
14:16:40directhex|workmarkun, i hate c, but i'll give it a quick tinker.
14:17:06markundirecthex|work: hate is suck a strong word :)
14:17:06preglowyou could always do it in asm
14:17:09markunsuch even
14:17:26*bluebrother is willing to work on the binary size font if someone donates the time :o
14:17:38markundirecthex|work: well, forget about it then, which language would you like to use?
14:17:57Nico_Pwhat exactly needs to be done in order to shrink bin size ?
14:18:05JdGordonremove waste code
14:18:12Nico_Pie what are the priority areas ?
14:18:17directhex|workmarkun, none that are appropriate for embedded devices. hurrah \o/
14:18:18markunfile browser?
14:18:19JdGordonanything bloated
14:18:21bluebrotherrewrite inefficient code
14:18:29markundirecthex|work: ok, but which one then?
14:18:32pixelmapreglow: first I don't have a different target and second: I really like it and none of the currently supported other targets suits my needs this much
14:18:32amiconnLinusN: I don't think viewports will add that much to the binary. And if the other parts of rockbox make use of it, I expect the size to go down
14:18:44LinusNamiconn: that is probably true
14:18:56LinusNamiconn: do you have a general idea how it would be best implemented?
14:19:10directhex|workmarkun, i prefer working with high-level languages like c#. i learnt to hate c during my degree
14:19:13amiconnViewports already exist in my head (yes, still)
14:19:40amiconnAnother thing I'll add in the process will be proper bitmap masking
14:19:50Nico_Pamiconn: don't you want to write it down (eg on the wiki) so that we can have an idea how it will work and maybe even help you implement it ?
14:20:57Nico_Pare there any parts of the code that are known to be inefficient i can try to improve ?
14:21:10JdGordonthe radio
14:21:16markunThe playback code also maybe?
14:22:03pixelmamarkun: what playback code are you referring too?
14:22:37Mikachuif i'm updating my arm-elf-gcc, should i use 4.0.3 or 4.0.4? has anyone tried 4.0.4 successfully?
14:22:37*JdGordon wonders if the list widget could work with less code
14:24:01directhex|workmarkun, this looks like it's beyond what i have the time & current skillset to do. i've got to stop taking things onboard i don't have the time to do properly
14:24:20daurnimatorJdGordon: goto bed!
14:24:27daurnimatorand give me your iriver
14:24:58JdGordondaurnimator: im in a horrible sleep patern atm... 2.30-11.00 ish... im coding for 30more min then going
14:25:00JdGordonand no :D
14:25:03JdGordon.. or soon
14:25:20daurnimatoroh well
14:25:22daurnimatorbed for me
14:25:39*JdGordon found a bug in the current remote scroll step setting
14:26:24 Quit Mouser_X (Read error: 110 (Connection timed out))
14:27:21pixelmaJdGordon: did you try the root menu in an Ondio-sim? I just wondered how the button mappings work there...
14:27:35pixelma(haven't tried yet)
14:27:51JdGordonpixelma: i havnt... it uses the same buttons as the rest of the menu code.. so if that works you should be ok
14:34:42 Join Criamos [0] (
14:39:02preglowhave anyone seen gcc skip a none-volatile asm() that had output constraints?
14:39:11preglowi think the volatile keyword might interfere with code generation
14:42:16 Join himitsu_ [0] (n=himitsu@
14:44:23preglowmaking the ones in dsp.h non-volatile shaves something above 100 bytes off
14:44:47 Quit Thundercloud_ (Remote closed the connection)
14:44:48 Join GodEater_ [0] (i=c2cbc95c@gateway/web/cgi-irc/
14:45:21Mikachucompiling with -funit-at-a-time seems to save a few kB, but i seem to have made a note that it was unstable for the vorbis codec
14:45:23GodEater_woohoo - PS3 pre-ordered!
14:45:35 Join Thundercloud [0] (n=thunderc@
14:45:39preglowMikachu: what does it do, really?
14:45:52Mikachuit considers the whole file before starting compiling
14:45:55Mikachusays the man page anyway
14:46:04Mikachuit also suggests that passing all .c files to gcc at the same time is better
14:46:11preglowahh, yes
14:46:34amiconnpreglow: Using volatile forces gcc to not discard the asm block, and not move it around
14:46:51preglowMikachu: isn't that enabled by O1 anyway?
14:47:06amiconnIf an asm block has only inputs and no outputs, volatile is assumed automatically
14:47:12preglowamiconn: yeah, i remember it's needed for asm blocks with no output constraints
14:47:16Mikachuit says O3 i think
14:47:35amiconnBut if a non-volatile asm block has outputs and these outputs aren't needed by subsequent code, the block might be discarded
14:47:38preglowMikachu: depends on the gcc version i guess, 4.1 does it at O1
14:47:54Mikachui don't know if it would save the same amount on archos, if it's to do with saving function calls
14:47:59Bagderpost-commit summary
14:48:01amiconnpreglow: It's _not_ needed for asm blocks with no outputs
14:48:25amiconnFor blocks with outputs, it depends on what these outputs do
14:48:40JdGordonBagder: nice... can you add a delta to that?
14:48:53BagderI intend to read out and present the delta in the table
14:48:53preglowamiconn: what? if you don't have output or input/output constraints, gcc _will_ throw non-volatile asm blocks away, i'm 100% certain of that
14:49:36JdGordonBagder: awesome :)
14:50:04BagderI'm not sure how to present it though
14:50:14JdGordonthe ipodvideo bin is huge compared to the others :p
14:50:18 Join dddd [0] (
14:50:18preglowwhy is ipodvideo so bloody big?
14:50:23Mikachui would probably make a second table
14:50:28 Quit dddd (Remote closed the connection)
14:50:38Bagderyeah, it might make sense to have a separate one
14:50:38JdGordonBagder: could you do it in a tooltip for the compile log number?
14:50:47preglowBagder: as long as its visible on the same page
14:50:57Mikachutooltips wouldn't be so good to get an overview
14:50:58preglowa bit of the point is forcing people to see what they just did to the binary size :)
14:51:05BagderI'll wait for a few more commits and then work on getting something to show you
14:51:24Bagderit just now started to work on a per-commit basis
14:51:34JdGordonBagder: wha ever happened with the idea of rotating the table 90 deg?
14:51:42Bagderits still there
14:51:47Mikachui have a javascriptlet for that
14:51:57*Mikachu checks i f it works
14:51:57Bagderideas are plenty, time is less
14:51:59JdGordonjust an idea? or is it actually going to happen?
14:52:02JdGordonah ok
14:52:06preglowamiconn: if it's assumed automatically, that's got to be late addition, i distinctly remember wondering what the hell gcc was up to one time it threw away my asm blocks
14:52:24preglowamiconn: or perhaps that was a block with neither inputs or outputs, but then that definitely should be assumed volatile as well
14:53:31MikachuJdGordon: transpose tables there
14:53:46Mikachuit sort of doesn't work though, because the pngs are so tall
14:53:46 Join fasmaie [0] (
14:54:23BagderMikachu: yeah, when rotated we wouldn't have pngs like that
14:54:32Mikachuand the times would be hard to print
14:54:44Mikachuunless firefox suddenly supports rendering vertical text
14:54:57Mikachuthe table is HUGE when rotated :)
14:55:01BagderI figure the times would be just numerical 1 2 3 4 5 ... and we have a "tooltip" for the actual time stamp
14:56:16 Quit himitsu (Read error: 110 (Connection timed out))
14:57:38Bagderstill, I'll leave that idea for later
14:57:49 Quit XavierGr (Read error: 110 (Connection timed out))
14:58:10JdGordonBagder: was the sizes you pasted from current svn?
14:58:13bluebrotherthat numbers could be the svn revision number ...
14:58:27BagderJdGordon: from the most recent svn zips, yes
14:58:52Bagderbluebrother: still they are 5-digit and would become pretty wide when used in the table
15:00:05preglowamiconn: but anyway, you agree that removing the volatiles in dsp.h should be a-ok? i'm running a build without them now, and all sounds good
15:00:05peturwith <br> inbetween each digit
15:00:18JdGordonBagder: ajbrec.ajz is the compressed bin isnt it? is it posible to get the uncompressed one instead?
15:00:19*preglow stares at the sound menu
15:00:26fasmaieHello, I just built rockbox from svn, and now I have lost the tags on all my mpc encoded music
15:00:35BagderJdGordon: yes it is, and since this is from the zip getting something else is tricky
15:00:42fasmaieOggs, mp3, aac are all fine
15:00:54preglowwhat's with the icons?
15:01:00preglowcrossfeed has a cross, eq has a filled dot
15:01:07MikachuBagder: do you have any means to update the build script on the build servers?
15:01:10JdGordonbecause eq hasnt been done yet
15:01:14Mikachuwould be easy to plop in a ls -l then
15:01:36Mikachuor just have the Makefile do it
15:01:38 Join unleet [0] (i=d0b7e10b@gateway/web/cgi-irc/
15:01:42JdGordonpreglow: I can change it to a dot to + to make everyone happy... but it also means i wont forget about that menu
15:01:49unleetWhat up guys?
15:02:02preglowJdGordon: sure, i don't mind, just wondered about why
15:02:18unleetDoes the H10 firmware have the fm tuner in it yet? the version i have does not have it.
15:02:22preglowmy, oh my, spc's sound empty with no echo enabled
15:03:26 Quit unleet (Client Quit)
15:04:22JdGordonamiconn: more good news... 900bytes gone and im barely 1/3 of the way through the display menu :)
15:10:37 Join Winchester345 [0] (
15:11:13 Join pondlife [0] (
15:13:32***Saving seen data "./dancer.seen"
15:18:01pondlifeAnyone else have problems viewing today's IRC log on the website? i.e.
15:18:28pondlifeIt hasn't worked for me since it was redone - although past days work fine.
15:18:38JdGordonlooks fine here
15:18:52pondlifeI know nothing about HTML, but the first thing displayed is "−−Delimiter Content-type: text/html "
15:19:12pondlife(IE6 on Win2K)
15:19:39peturyes, there's something wrong in IE, firefox is ok
15:20:16pondlifeOK, as long as it's not just me.
15:20:17MikachuJdGordon: i can't enter the "Repeat" setting in the Playback Settings menu
15:21:06preglowargh! it does it again, i remve volatile from dsp.h macros, and now suddenly apply_gain has no emac code
15:22:16JdGordonMikachu: i fixed that eariler today... update to svn
15:22:22 Join SirFunk [0] (
15:24:34MikachuJdGordon: another thing i find a bit confusing, why does Idle Poweroff have a hollow circle and Sleep Timer a filled circle?
15:24:54JdGordonIs it better to wait to finish the entire display menu before commiting (might not happen till monday), or commit what ive done (all except the lcd and remote lcd sub menus and the peak meter submenu)
15:25:11JdGordonMikachu: hollow is settings, filled is functions
15:25:22Mikachuin this case they appear identical to the user
15:25:43JdGordonhmm... ill change it then
15:27:42Bagderhm, it's hard to make a readable table with the targets and their binary sizes
15:28:00Bagderit either gets very wide, or unreadable
15:28:21Bagderthat's with <br> between each digit to make it narrow
15:28:26 Join XavierGr [0] (
15:28:39Mikachumaybe kB and bytes on tooltip?
15:28:55BagderI'll try
15:28:59Mikachuor delta, and size on tooltip
15:29:00JdGordonBagder: that looks ok
15:29:33JdGordonoh wait... the verticle number the filesize not the delta..
15:29:45Nico_PBagder: would a graph be hard to do ?
15:29:47Bagderyes, there's only one table available
15:29:54Bagderdelta is hard then...
15:30:08JdGordonhehe graph would be nice
15:30:22BagderI'll leave the graphing to someone else ;-)
15:30:46JdGordonMikachu: or delta, and size on tooltip <- would be best
15:31:18BagderI need another commit now to get some deltas to work with... :-)
15:32:16preglowi was planning on doing one, but gcc is a slut
15:32:37directhex|workBagder, since your data points (file sizes) are long rather than thin, what's the reason for putting them in the chart vertically?
15:33:15JdGordonI can do one
15:33:36Bagderdirecthex|work: I want the table to be not be too wide
15:33:42JdGordonactually better not
15:33:52Bagderhm, minus a few extra words
15:35:33bluebrotherBagder: are the values itself available somewhere? Like as a csv file?
15:36:26Bagdernope, but I guess I could make them
15:36:26JdGordonok Bagder, commit sent
15:37:10 Quit pondlife (Read error: 104 (Connection reset by peer))
15:37:23bluebrothermaybe create some kind of log ... <revision number> followed by the values sorted by the target id? First is 0, and unused are simply empty
15:37:27preglowdsp.c doesn't include config.h ......
15:37:38Bagderevery size "dump" is available in this style:
15:38:43directhex|workBagder, ah, i suppose i don't notice these things on a 1920x1200 screen
15:38:52preglowthere, almost a kilobyte off
15:39:18Bagderdirecthex|work: browsing in fullscreen is bad for you ;-)
15:40:08amiconnpreglow: Even gcc 3.0.x has it:
15:40:26amiconnAbout 3/4 through the text
15:40:36*JdGordon notices the table fits almost perfectly into 1024x768 without the timestamp
15:41:31 Join dnaf[working] [0] (
15:42:51dnaf[working]hi there. I've got my iPod nano freeze. something like RockBox, loading. and that's all.
15:43:11JdGordongah, bloody system bell and stereo on full! scared the shit out of me
15:43:21 Quit Rob2222 (Read error: 104 (Connection reset by peer))
15:43:23 Quit printfXh4 (No route to host)
15:44:15amiconnpreglow: Regarding ipod binary sizes... I think that the 4 functions in ata.c don't need to be in iram for arm, so we can keep them static
15:44:41amiconnI'll test my theory tonight, measuring disk speed. If that works, we can get rid of -ffunction-sections
15:44:57amiconn...which in turn gets rid of many long calls
15:45:32amiconnWith just -O, the resulting binary works fine
15:45:44amiconnIt's -O2 which breaks it, causing the stkov
15:45:58preglowi wonder what's going on
15:46:04Nico_PBagder: how to we access the different dumps ?
15:46:39Nico_P(size dumps)
15:46:53Bagdercurrently they're available as links from the sizes table
15:46:57preglowamiconn: and you're absolutely sure that move.b and move.w leaves the upper bits of the destination untouched, yes? i could have sworn i read it somewhere myself, but can't find it again
15:47:23amiconnpreglow: Yes I'm 100% sure
15:47:24Bagderbut I can provide them in an easier way too if you like
15:47:47JdGordonthats the deltas?
15:47:55LinusNno, kilobytes
15:48:05Bagderno deltas yet
15:48:14JdGordonoh ok
15:48:22preglowamiconn: good, but i've still got the nice problem here. i remove volatile from the dsp.h macros, and boom, gcc ignores them
15:48:33Nico_PBagder: it looks fine to me
15:48:42preglowamiconn: at least the ones in apply_gain it ignores
15:48:52Nico_Pcould you add a link to the regular build table towards this one ?
15:49:14Bagderyes, I will do that soon
15:49:34*JdGordon going to bed, cyaz
15:49:37 Quit JdGordon ("Leaving.")
15:50:50amiconnpreglow: Of course it does. Those asm block *do* have output constraints whose values are unused
15:50:59Mikachu30 more commits and it's number 12345 :)
15:51:19amiconngcc treats asm blocks as volatile which do *not* have output constraints
15:52:47preglowso i see
15:53:03preglowi should just start writing everything in asm to not have to worry about this
15:53:24 Quit Surger (Read error: 145 (Connection timed out))
15:53:43Bagderkind of interesting
15:54:02preglowseems like only FRACMUL_8_LOOP is used now too
15:54:07Mikachumaybe color green on -, red on + and not on no change?
15:54:21Bagdergood idea
15:55:35markunBagder: and can you choose the green and red from the wiki? They are visible to me and XavierGr :)
15:55:37 Part dnaf[working]
15:56:08*amiconn wonders what's special about the wiki green/red
15:56:11Bagdermaybe I should even cut off the total
15:56:12directhex|workmarkun, you're colorblind?
15:56:13 Join Lynx [0] (
15:56:36Mikachu(can you have tooltips on things which are not links or images?)
15:56:50BagderI think there's some way to make it work
15:56:57directhex|workMikachu, not AFAIK, but you could do some clever css hax to achieve the same effect
15:56:58markundirecthex|work: yes, and probably some more people here
15:56:59Bagderalthough I'm not the master of that stuff
15:57:07amiconnah ok
15:57:10Mikachui basically only know <br>
15:57:26amiconnMikachu: Tooltips on standard text is possible
15:57:37Bagderyou know how?
15:59:39markunBagder: with <span title="..">some text</span> maybe?
15:59:59amiconnYes, title="Tool tip text"
16:00:27markunI'm not sure if IE and FF use the same tag for tooltips :)
16:00:59markunbut they you can include both
16:01:52Bagdertry now then
16:01:58markunah no, it works with both
16:02:00amiconnShould work on nearly all html elements, like paragraphs
16:02:46bluebrotherBagder: how about adding the actual size at the top of the table and only giving deltas in the list itself?
16:03:26Bagderbut how would the deltas be compared to the size in the top?
16:03:44Bagderright now, the last line has the total only because it has no delta
16:03:54Bagderi.e two logs => one delta
16:04:07amiconnBagder: Putting title= should work directly in the <td>, no need for the extra <span>
16:04:50Bagderand now on the build table page...
16:05:28 Quit Criamos ("( :: NoNameScript 4.03 :: )")
16:05:42 Join Criamos [0] (
16:05:44markunBagder: you could put the delty between the lines maybe?
16:05:57markunat least also show the latest binary size
16:08:36preglowamiconn: so asm("nop"); is also equivalent to asm volatile("nop"); ?
16:08:50amiconnIt should
16:08:57bluebrotheronly use delta values to the last build, and put the latest size at the top?
16:09:09bluebrothershould be sufficient to get the change itself.
16:10:20directhex|workdon't include the bottom row, use it only for calculating the bottommost delta, show only deltas w/ tooltips for full sizes
16:11:02directhex|workthereby avoiding a random wide row, and making the obvious thing you want to see (deltas) the primary chart focus
16:11:42 Quit Lynx_ (Read error: 110 (Connection timed out))
16:11:42 Nick Lynx is now known as Lynx_ (
16:13:37 Quit Winchester345 ("Download Gaim:")
16:14:40Bagderand now some colors
16:16:10Mikachuis the delta in bytes?
16:17:03 Join Surger [0] (
16:17:12pixelmawhat about the standard rockbox website layout on this page... :)
16:17:56pixelmathat's the one I mean (especially the left side menu) ;)
16:18:10markunBagder: nice
16:18:16Bagderit uses the same css, just not any menu
16:20:44peturBagder: any reason the target order is not the same as the build table?
16:20:57 Join midgey|web [0] (i=8dd33f4d@gateway/web/cgi-irc/
16:20:59Bagderyes, it has no knowledge of that order
16:21:02pixelmais there a reason that it doesn't have the menu? (because I don't like it stick on the left side of my monitor)
16:21:18Bagderno other than that it already is very wide
16:21:44*petur hands pixelma some cleaner to remove the sticky bits
16:21:51directhex|workalign="right" for pixelma!
16:22:07linuxstbIf we want to measure codesize, shouldn't we try and sync all build servers to the same gcc/binutils versions?
16:22:21 Join perl|work [0] (
16:22:22pixelmadirecthex|work: then it sticks at the right side... doesn't make it better ;)
16:22:37peturpadding !
16:22:47directhex|work24" tft for pixelma!
16:22:53peturor margins
16:24:43 Join webguest30 [0] (i=3f7cde02@gateway/web/cgi-irc/
16:24:49pixelmadirecthex|work: nothing against that... :P (but I guess it'll look even worse)
16:26:48 Part LinusN
16:26:57webguest30Hi, I just registered myself for the Twiki in order to post some battery tests I've run on my Nano. Can I be granted write permission for posting my results? My Wiki Name is DavidBinder.
16:29:00peturwebguest30: done
16:33:34 Quit Febs ("Chatzilla 0.9.77 [Firefox]")
16:37:54webguest30Thank you very much!
16:38:14pixelmapetur: I wanted to ask you something about the properties plugin... I think it has a rather strange way to display the filesize wrt accuracy (<10MB it displays Bytes, >10MB it only displays full MB). Why does it have to be different to what "Rockbox Info" does?
16:38:39 Join digen [0] (n=digen@unaffiliated/digen)
16:38:39 Quit webguest30 ("CGI:IRC (EOF)")
16:40:40peturpixelma: I just wrote it like that, haven't looked at rb info code. I wrote it as compact as possible first to include into main code, then moved it to a plugin
16:43:12pixelmawouldn't reuse of code that's already there be easier?
16:44:03 Quit Thundercloud (Remote closed the connection)
16:44:03peturI tried, and found my code to be more compact than reusing. But now that it's in a plugin, it doesn't matter anymore
16:44:26GodEater_join #gigabeat
16:45:09 Join matsl [0] (
16:52:13amiconnpetur: You could just use output_dyn_value(). Probably needs to be exported
16:54:12peturamiconn: I didn't like that function, can't remember why.
16:56:49 Quit crwl (Read error: 110 (Connection timed out))
16:57:44 Quit lostnihilist (Remote closed the connection)
17:00:18 Join dpassen1 [0] (
17:01:09 Quit midgey|web ("CGI:IRC (EOF)")
17:03:26 Quit barrywardell (Remote closed the connection)
17:05:05 Join Rondom [0] (
17:06:30 Quit fasmaie (Remote closed the connection)
17:10:50amiconnpetur: It would be helpful to know why. This function was coded for this very same purpose: outputting wide-range numbers with (neraly) constant precision
17:10:57amiconnIt also supports voicing
17:11:58peturok, note to self made, will try to remember tonight ;)
17:13:37***Saving seen data "./dancer.seen"
17:13:41preglowsome of the asm() statements should have earlyclobber outputs, i think
17:13:45preglowit's a wonder they have broken yet
17:13:51preglowhaven't, that is
17:15:07preglowperhaps gcc doesn't alias input regs onto output regs that often anyway
17:19:43 Join n1s [0] (
17:29:35 Join fasmaie [0] (
17:30:31 Join lostnihilist [0] (
17:31:17fasmaieSlasheri: The changes made to use new endian caused the tags on my mpc files to break
17:31:42 Join Arathis [0] (
17:31:53fasmaieSlasheri: At least, when I reverted to earlier versions of the files you changed, all worked fine
17:31:58 Quit DataGhost (Nick collision from services.)
17:32:04 Join DataGhost [0] (
17:32:08fasmaieSlasheri: Just thought I'd tell you
17:34:17 Join toffe [0] (
17:38:48 Quit GodEater_ ("CGI:IRC (Ping timeout)")
17:42:17 Quit bluebrother ("CGI:IRC")
17:43:53 Quit t0mas (Read error: 104 (Connection reset by peer))
17:44:20 Join GodEater_ [0] (i=c2cbc95c@gateway/web/cgi-irc/
17:45:58 Quit HardDisk_WP ("Disconnecting from stoned server.")
17:46:08 Join HardDisk_WP [0] (
17:46:31 Join |Rincewind| [0] (
17:48:49 Part hcs
17:49:41preglowamiconn: you don't know of any mechanism that allows you to specify which input and output operands can't be aliased in an asm constrain list, do you?
17:50:00 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
17:50:30 Join Wiwie [0] (
17:50:42 Quit petur ("stkov")
17:52:25 Quit matsl (Read error: 60 (Operation timed out))
17:54:43 Quit lukaswayne9 (Remote closed the connection)
18:00:45 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:01:46 Quit GodEater_ ("CGI:IRC")
18:04:21preglowLear: do you have any use for FRACMUL_8 planned, or can i just remove it?
18:04:50LearNo dsp work planned, no...
18:04:55preglowFRACMUL_8 was used once upon a time, i think, but no more
18:05:09LearWasn't it used in gain calculations?
18:05:48preglowLear: maybe? the only place any of them are used now are in apply_gain, and there only FRACMUL_8_LOOP
18:06:09LearAh, yes.
18:07:15preglowi'm applying a minor opt to it, so i'll just remove the cruft in the same go
18:15:16 Quit mathgl ("Quitte")
18:17:36 Quit BigBambi_ ("Leaving")
18:19:11 Join nicktastic [0] (n=nick@unaffiliated/nicktastic)
18:19:41Slasherifasmaie: could you send the file for me to try?
18:20:09Slasherithat's very possible because i couldn't test the metadata changes, only the voice ui part was tested
18:21:18amiconnHmm, who added the speex patch?
18:21:32amiconnI wonder why speex is compiled with just -O
18:22:09Mikachulooks like dan according to svn log
18:22:15Mikachur12241 | dan | 2007-02-09 11:06:53 +0100 (fre, 09 feb 2007) | 1 line
18:22:15MikachuAdd Speex playback support. Patch from FS #5607 thanks to Frederik Vestre.
18:23:41 Quit tipi^ ("javisst")
18:28:06 Join knight4led [0] (i=be56320d@gateway/web/cgi-irc/
18:28:54knight4ledAnyone here know how to compile a sim for OSX?
18:28:57 Join Domonoky [0] (
18:30:56 Quit Wiwie (Read error: 104 (Connection reset by peer))
18:32:16 Quit knight4led (Client Quit)
18:49:05 Join blackness [0] (
18:49:27blacknessis it possible to do a dual boot with the ipod's original firmware?
18:50:12preglowat least i can do so
18:50:13blacknessis there a tut that you can point me toward?
18:50:25Mikachureboot and turn on hold
18:50:26preglowdunno if the installer changes of late has changed that
18:51:11 Join spiorf_ [0] (
18:51:18 Join BigBambi [0] (
18:53:06blacknessis it standard or do I have to do something on install?
18:53:15linuxstbblackness: Which ipod do you have?
18:53:29blackness5g 60
18:54:17linuxstbThen there are no problems - just do what Mikachu said. Dual-boot is standard.
18:54:32blacknessthat kicks ass
18:56:22blacknessif I backed up then reinstalled my the original firmware should I have to back it up again?
18:58:16linuxstbYou've lost me... Backup what? Your music or the firmware?
18:58:24blacknessmy original ipod firmware
18:59:00linuxstbIt doesn't hurt, but it's not suggested as part of the install instructions any more. The latest installation tools are very reliable.
18:59:32linuxstbAnd you can always restore the original firmware with itunes, or simply download it from Apple via
18:59:51blacknessokay, I was just scared of never getting it back
19:00:42linuxstbBTW, the instructions here are newer and easier than those in the manual.
19:03:43 Join blueworm [0] (
19:03:47*linuxstb wonders if he's alone in thinking the sleep timer isn't a setting
19:04:57 Join tipi^ [0] (
19:06:49preglowhe's not
19:07:19 Join linuxstb_ [0] (i=5343d4aa@gateway/web/cgi-irc/
19:08:59 Join Juddy [0] (
19:09:10 Join Soul-Slayer [0] (n=Jonno@
19:09:18Juddyhey prob a stupid question
19:09:28Juddybut i forget where i need to put bootsplash?
19:09:38Juddyand where is the original boothsplash available
19:09:48Juddyfor the gigabeat i mean!
19:10:50fasmaieSlasheri: I just used the previous versions of each file
19:12:01toffeJuddy :
19:12:09toffeJuddy : end of the page
19:12:40toffeJuddy : it goes there \GBSYSTEM\FWIMG
19:13:00toffeJuddy : how did you fix your system error ?
19:13:39***Saving seen data "./dancer.seen"
19:13:47Slasherifasmaie: yeah, but that is not helping to fix the bug at all
19:13:53Juddyhey. i just kept fiddling with it. constant formatting and reformatting then..
19:14:04Slasherifasmaie: i meant that could you send the problematic mpc file to me?
19:14:05Juddyi put the 3.0 US on by itself
19:14:15Juddystill didnt work
19:14:27Juddybut i added the rockbox files individually
19:14:32Juddyand that started to load
19:14:35Juddybut with errors
19:14:40Juddyso i just kept on doing it
19:14:49Juddyand now it works consistantly
19:15:19Juddybut original firmware i still cant use. oh well
19:15:23Juddydont like it anyway
19:15:48toffestrange pb
19:19:47SurgerI can't believe what you peeps are doing. :)
19:20:54 Join wooo [0] (
19:22:15 Join crwl [0] (
19:22:15 Join dan_a [0] (n=dan-mirc@
19:23:04 Join matsl [0] (
19:25:35blacknessdoes it seem like most of the fonts are really small to anyone else?
19:26:13Domonoky*just reading the logs* if we get rbUtil to compile for MacOSX (dont know if linuxstb had success in building it) all those unzipping problems would be gone..
19:26:47preglowblackness: most of our targets have way smaller screens than the ipod5g, remember
19:28:53blacknessso will any fonts I want on there work?
19:30:40linuxstb_Domonoky: Keep reading the logs...
19:31:02Domonokyah.. found it it errors..
19:31:20preglowblackness: you can use whatever fonts you want, as long as they're the correct format
19:31:27preglowie. no truetype fonts
19:31:59blacknessthats the only kind I know, what is the extension?
19:32:37blacknessnm got it
19:33:05preglowtruetype fonts can be converted to be usable, but i don't know how
19:33:38 Quit digen (Read error: 60 (Operation timed out))
19:33:54linuxstb_blackness: You could look here -">
19:35:26 Join blithe [0] (
19:37:21 Join CriamosAndy [0] (
19:41:33 Join My_Sic [0] (
19:42:57 Quit MonkeyTamer ("Chatzilla 0.9.77 [Firefox]")
19:44:07 Join directhex [0] (
19:44:16 Join digen [0] (n=digen@unaffiliated/digen)
19:44:24 Quit My_Sic (Client Quit)
19:45:56directhexmarkun, lcd power down has GOT to go. it makes the remote pretty much useless. any idea off the top of your head of what i can comment out for my own builds?
19:46:53linuxstb_Why does lcd power down interfere with the remote?
19:49:12preglowi've got a nice error going in rockbox here now...: "couldn't load /.rockbox/codecs/ .codec"
19:49:16preglowsplashes again and again
19:50:19linuxstb_The invisible codec...
19:50:44preglownot a hanger, though, just pressed stop, found something else to play, a couple of loud clicks, and hey, we're on again
19:51:21preglowalso, i quite like the new menu stuff
19:51:37linuxstb_Tried the root menu patch?
19:51:46preglownope, just talking about svn
19:51:55preglowwhat with the icons and all
19:51:58 Join hardeep [0] (
19:53:21 Quit Criamos (Read error: 110 (Connection timed out))
19:55:07 Join pilot000 [0] (i=550262a9@gateway/web/cgi-irc/
19:55:11preglowLear: isn't the replaygain preamp granularity a bit high? who ever used 0.1 dbs?
19:55:59LearWell, 1 dB is a bit coarse, isn't it? But I guess it could be reduced a bit... What does the eq use? 0.5?
19:56:59preglow0.5 db is much more sensible
19:56:59 Quit atsea-196 (Remote closed the connection)
19:57:13preglowthe list code makes using 0.1 db a bit of a bother since you have to scroll for ages
19:57:46LearFeel free to change. Makes sense to have the same as in the eq...
19:58:08 Join atsea-196 [0] (i=ariel@gateway/tor/x-1cbed1a99d0c7f68)
19:58:45preglowi wonder what happens to users who already have a setting that'll end up incompatible
19:58:48pilot000Hi, I have a problem with my iriver H140 (flashed Rockbox). When I will boot the unit, it shows "No partition found. Insert USB cable and fix it." When I try to format the disk with Swissknife, I get a read/write error. Is there a other way to reset the partition table and reformat the unit?
19:58:55 Join My_Sic [0] (
20:00:00preglowpilot000: well, short of taking the disk out and plugging it into a computer with an adaptor, no, that's it
20:00:17preglowpilot000: but anyway, if it doesn't work like that, it does sound very hardware failure-ish
20:02:10pilot000preglow: I modded my player last week with the RTC chip. But I had before also problems with the disk.
20:02:38pilot000I seems like I need a new disk
20:02:39preglowpilot000: and you can't boot iriverfw?
20:03:02preglowit probably pukes when it doesn't see a partition as well
20:03:06Arathispilot000: If it's something with the partition table testdisk could help:
20:03:18preglowthis reminds me i'm having a problem with my nano too....
20:04:48pilot000preglow: no change to boot to iriver fw. I said goobye to it before a long time (bootloader 7-pre3 and flashed everything)
20:06:35pilot000Testdisk does not see the iriver :-(
20:06:47preglowthen it does sound like a new disk would be good
20:08:09Arathishas the H140 something like an UMS mode or how does it show up as a device?
20:08:44directhexlinuxstb_, for some reasons markun can probably explain better than me, the remote behaves... oddly... when it's enabled
20:09:58 Part blackness
20:10:22pilot000Arathis: I don't see it in the windows explorer. I see it in swissknife as an USB device. It shows me a TOSHIBA harddrive. But thats all. No serial or so ...
20:10:31preglowlinux throws scsi errors at me when i try to write to my nano, and windows claims it's not formatted
20:10:35preglowthis doesn't really look too well
20:11:02directhexpilot000, as we say in the cluster business, sounds buggered
20:11:04preglowit can read itself fine, though
20:11:12Arathispilot000: I think that's why testdisk can't find it, but I'm not sure. I'm not familiar with flash players
20:11:29pilot000this is a hd player
20:11:36Arathisoh :D
20:12:32 Join lee-qid [0] (
20:12:33pilot000I think it's time to order a bigger HD, because I won't loose my player forever
20:13:30 Join funky [0] (n=repulse@unaffiliated/funky)
20:13:39preglowi think i'm gonna try giving it some ipod restorer magic
20:13:50 Join printfXh4 [0] (n=pseudo@
20:13:54preglowghah, or do i remember right when i say that you need itunes to restore an ipod now?
20:14:51Mikachuwhen i tried to use it, it didn't even find the ipod, i had to do it manually in the end
20:15:25preglowi'll be bloody annoyed if it's broken
20:15:44preglowhow do i go about doing so manually, btw?
20:15:52preglowor are you just talking about a plain repartition?
20:16:07Mikachuyeah, partition, mkfs, copy boot partition
20:16:11Mikachuwhat else is there to do?
20:16:58Mikachui don't remember if it needed itunes, did it on a friend's computer with windows
20:17:00preglowthe boot partition is available as a plain image?
20:17:38Mikachui think this is it
20:17:46Mikachujust unzip it first
20:17:55pilot000This Disk seems the one I need (for iriver H140): TOSHIBA MK-6008GAH ?
20:18:07preglowsounds familiar
20:18:38Mikachui think i have my backup lying around too, 2GB model if that matters but i don't think so
20:18:56preglowwell, cool, in case everything blows up, but i think i'll manage
20:19:04preglow2 gig here as well anyway
20:21:11Mikachudon't forget the boot partition needs type empty set
20:21:56preglowanything fancy i should pass mkdosfs?
20:23:26linuxstb_Just -F 32 afaik.
20:25:21preglowthe apple screen with a progress bar?
20:25:32Mikachuit's probably flashing then?
20:25:41 Join anathema [0] (
20:25:51preglowok, i didn't think they did that
20:26:16linuxstb_Yes, any time you install a new firmware partition.
20:26:18amiconnpilot000: Depends. 1.8" HDDs are available with 2 different connector types
20:26:26linuxstb_(as opposed to a backup of one)
20:26:39Mikachuwhat happens if you write an older?
20:26:45linuxstb_No idea...
20:27:08linuxstb_I think it reflashes it though.
20:27:14pilot000amiconn: but I think i need a toshiba one ?
20:27:16Mikachuis it possible the newer ones fix the slow emergency mode?
20:27:42preglowi seriously doubt it
20:28:25linuxstb_rbutil for the Mac comes in at a lightweight 9MB...
20:28:35pregloweverything statically linked? heh
20:28:35Mikachuwith code for 300 different processors?
20:28:51linuxstb_Yes, everything static...
20:29:03linuxstb_And universal binary, so just two processors.
20:29:11amiconnIt doesn't strictly need to be a Toshiba, it's just that (nearly?) all 1.8" HDDs are made by Toshiba
20:29:25amiconnStill, the Toshiba drives are available with 2 different connector types
20:29:27Domonoky@ linuxstb is it striped ?
20:30:02preglowlinuxstb_: ipodinstallationbeta link to amd64 binary is quite dead
20:30:23Domonokyon windows uclpack is used which puts the size from ~6 MB down to 2
20:30:24preglowlinuxstb_: permission problems
20:30:42pilot000amiconn: you mean a connector type like this
20:30:46*linuxstb_ pings Bagder and LinusN
20:31:19Domonokylinuxstb_ so you got rbUtil to compile on MacOsX ?
20:31:28pilot000then I need the first one, because it looks exactly as the one is in my iriver
20:31:29linuxstb_Domonoky: I'm not sure if it's stripped, I'm still getting to grips with the Makefile... But it can be distributed as a compressed .dmg, so that will help by about 50%.
20:31:31preglowlinuxstb_: please, man, this has become too easy...
20:31:45Domonokythats good :-)
20:32:04preglowi didn't even have to bloody tell it the device name
20:32:04amiconnpilot000: Both offers are examples where the image doesn't match the text
20:32:09linuxstb_Domonoky: Yes, I didn't realise that wxwidgets is in fact many libraries, and I wasn't including a needed one in the Makefile (which I copied from a sample application).
20:32:32linuxstb_preglow: And now your PC will boot Rockbox...
20:32:51Domonokybut there are still lots of things todo in rbUtil..
20:33:01preglowanyway, it seems my nano is back on track
20:33:03preglowkudos to everyone
20:33:34linuxstb_Yes, and it needs adapting for Mac OS a little... /usr/local/share/rbutil/rbutil.ini isn't a useful location to look in.
20:33:45linuxstb_(Or in Linux for that matter)
20:34:13Domonokyhm is rbutil in /usr/local/share/rbutil ?
20:34:23amiconnThe MK6008GAH seems to be available with ZIF connector only, which would be the correct one for iPod G5 but the wrong one for irivers
20:34:28Domonokyi think it look for the ini in the programfolder.
20:34:55preglowMikachu: still slow as hell
20:34:59Mikachuokay, thanks
20:35:03linuxstb_Domonoky: Nowhere near. It's still in the source directory.
20:35:29Domonokyok, i will have a look into the code
20:35:51 Join redbreva [0] (i=56956084@gateway/web/cgi-irc/
20:35:55pilot000hmmm. what I need then, a MK6006GAH , if I want a 60GB replacement drive ?
20:35:56linuxstb_I'm not sure what the convention is on Mac OS X, but something like ~/.rbutilrc would be more Unix-like
20:37:18preglowpilot000: how big was the drive you had?
20:37:32pilot00040GB dual patter
20:37:39preglowpilot000: then you have room for a 60 gig, afaik
20:37:52amiconnLooks loike it. The MK6006GAH connector is shown in this pdf page 8. That's the correct one for irivers. The MK6008GAH's ZIF connector looks like here : much smaller
20:38:00Domonokyhm, i think the place where the executable sits would be the best, the programm doesnt change the ini..
20:38:51preglowlinuxstb_: btw, do you know if barry or someone else have mentioned if the usb chip in the ipods seem to have some power down state worth trying?
20:39:27pilot000MTF: 300'000, I used my dap a lot then ....
20:39:48pilot000thanks amiconn, MK6006GAH is the one I need
20:40:12 Join niskel [0] (
20:40:13amiconnYou could also try to find a MK8007GAH
20:40:21 Quit blueworm (Remote closed the connection)
20:40:24linuxstb_preglow: I think he said it's already powered down.
20:40:28LearFeel free to change. Makes sense to have the same as in the eq...
20:40:32preglowbummer for us
20:40:34LearSorry, wrong window. :)
20:40:43preglowLear: you meant to say that in another window? :P
20:41:04LearI meant to press "up" + "return" in another window, yes.
20:41:59preglowahahah, cool, now i have a fat driver panic
20:43:47 Quit wooo (Read error: 110 (Connection timed out))
20:43:53pilot000I found a MK8006GAH in switzerland for 221 CHF ~145 Euro
20:44:18pilot000But there are nor 8007 on stock somwhere
20:44:25 Quit redbreva ("CGI:IRC")
20:44:40preglowMikachu: hmm, 92 megs were copied in 20 secs
20:44:55Mikachuthat is including sync/unmount?
20:45:12preglowMikachu: i timed both the cp and the eject
20:45:37Mikachuwas this one big file or many small? iirc only the latter is slow
20:45:47preglow13 files totalling 92
20:45:55preglowi can see how long copying the rockbox tree takes
20:46:05preglowor can you think of any better benchmarks?
20:46:19Mikachui remember when i tried that, i thought it had locked up
20:47:28preglowMikachu: about nine secs in total for both unzip and eject :>
20:47:58Mikachuhm, well that is an improvement
20:48:08preglowi'd say so, yes
20:48:12preglowat least i can live with it now
20:48:29 Join bronze--- [0] (
20:48:39Mikachui could almost remove the ipod firmware then
20:48:54Mikachuyou don't happen to know if the emergency mode charges as quickly as OF?
20:49:20bronze---Just nned to ask one thing: Does anyone know how long until a working version of Rockbox for iPod 80GB?
20:49:37*linuxstb_ installs the bootloader and on his ipod with rbutil/MacOSX
20:51:01Domonokyi found the error with the position of the ini file. the wrong wxwidget function was used :-)
20:51:14preglowMikachu: no idea
20:52:01 Join blueworm [0] (
20:52:02linuxstb_Would anyone object to fwpatcher being moved from tools/ to rbutil/ ? (so it can be assimilated)
20:52:09preglowgo ahead
20:52:26preglowdon't know if i'd want to assimilate that code, though...
20:52:41Domonokyits already assimilated, and it works (tested by me)
20:52:55preglowit's just a really cut up and beaten scramble/descramble and mkboot
20:53:02 Quit bronze--- (Client Quit)
20:53:13Domonokyonly the code in iriver.c and md5.c plus the sums headers are used
20:54:46amiconnpilot000: The Toshiba site doesn't seem to know the 8006
20:55:14preglowhmm, how do i make a fonts zip?
20:55:18Mikachumake fontzip
20:55:31preglowit is...
20:55:49Mikachumake <tab> if you have zsh :)
20:56:59preglow"scroll speed setting example" isn't long enough on ipod :)
20:57:05 Join redbreva [0] (n=redbreva@
20:57:26amiconnpreglow: Depends on your font
20:57:36Mikachuand language
20:57:37preglowamiconn: yes, surely, but i'm talking the default one
20:58:05pilot000amiconn. what a typing fault. I meant MK6006GAH
20:58:12Mikachuthere's also make help
20:58:36Domonokyhm, about bootloader uninstallation with rbUtil, should i integrate it into the normal "rockbox uninstallation", or should i make an extra button ?
20:59:38 Quit blueworm ("Leaving")
21:00:06dan_aDomonoky: I would say an extra button - or possibly a question that gets asked, in case someone has IPL on an ipod
21:02:24Domonokyan extra button would be easier, because for bootloader uninstallation you have to specify your device, for normal rockbox unistallation the place in the filesystem is enough..
21:03:34dan_aIn that case, possibly pop up a notice after normal rockbox uninstallation suggesting that people remove the bootloader too.
21:04:29Domonokyjeah , i try to implement it :-)
21:04:33pilot000amiconn, preglow, thanks for the help. I just ordered a MK6006GAH. I should receive it tomorrow before lunch.
21:05:06dan_aGood work, by the way, Domonoky :)
21:05:36amiconndan_a: COuld you test something on your ipod G3?
21:05:47dan_aamiconn: Certainly
21:06:39amiconnI would like to know whether there is a speed difference in disk access if you remove the ICODE_ATTR from the 4 functions in ata.c, and if so, how much slower it becomes
21:07:06 Quit Rondom ("Ex-Chat")
21:07:07dan_aI'll try it, and report back.
21:07:21amiconnHmm, does the G3 use the optimised transfer loop?
21:07:37amiconnIf not, then there are 6 functions where ICODE_ATTR should be removed
21:07:56dan_aIt doesn't use your ASM function - that failed on the PP5002
21:08:22 Quit Surger ("zzz")
21:08:28amiconnAh ok, so 6 functions to change (there are only 6 functions with ICODE_ATTR in ata.c)
21:09:36 Quit pilot000 ("CGI:IRC")
21:12:43 Join Rondom [0] (
21:13:01Nico_Pi've got a WPS tokenizer that seems to work well... the only thing is i'm having difficulties adapting the rest of the WPS code to it
21:13:20Nico_Pmaybe i should just try to rewrite the whole thing
21:13:42***Saving seen data "./dancer.seen"
21:13:50Nico_Pany ideas on what could be improved ?
21:14:18linuxstb_.bmp files loaded from a tar file....
21:14:54amiconnThere are 2 bugs in the new menu code on the player....
21:15:48Nico_Plinuxstb_: you had a patch for this, didn't you ?
21:16:17Nico_Pah, found it
21:16:50Nico_Pbut actually i was rather thinking about the displaying code
21:17:14Nico_Pi'm kinda lost in gui_wps_refresh()
21:19:47 Join perpleXa [0] (i=perpleXa@unaffiliated/perplexa)
21:21:09 Quit toffe (Read error: 131 (Connection reset by peer))
21:21:20 Join webguest98 [0] (i=3e2fb9c7@gateway/web/cgi-irc/
21:21:35 Join toffe [0] (
21:22:18preglowamiconn: is there a player sim?
21:23:20amiconnThe only difference between swcodec and hwcodec sims is that the hwcodec sims don't play music (but they can show the wps and "pretend" to play)
21:23:38preglowsure, just wondered because of the charcell
21:24:11preglowthe rockboxed wps background colour is horrible
21:24:13preglowplain and simple
21:24:15preglowat least on nano
21:24:23amiconnThe charcell lcd is fully simulated, including the separate icons
21:24:39amiconnOne bug relates to the icons, btw
21:26:11 Quit perpleXa ("leaving")
21:29:00 Quit niskel (Remote closed the connection)
21:29:16 Join petur [0] (n=petur@rockbox/developer/petur)
21:29:58XavierGrpreglow: I agree!
21:30:39preglowamiconn: btw, i'm gonna start coding the treble/bass filters now, i just need to figure out one thing so that i have enough registers
21:30:50preglowfor the asm optimised routines, that is
21:31:38Mikachuheh, my boot partition is too small for the updated firmware
21:32:12 Quit webguest98 ("CGI:IRC (EOF)")
21:32:38amiconnpreglow: nice :)
21:33:29amiconndan_a: Oh, if you do the test I asked you for, you could optionally also try to build with -ffunction-sections removed from the main Makefile
21:33:57amiconnThat's why I am doing this - getting rid of wasted space and some unnecessary long calls on arm
21:34:27amiconn(full rebuild required to see the complete effect of removing -ffunction-sections)
21:34:30 Join wooo [0] (
21:34:43dan_aamiconn: I'm just about to do the first test now (ICODE_ATTR removed.) Then I'll take out -ffunction-sections and then do a control test
21:35:00Mikachudon't you have to add -ffunction-sections to the codecs etc makefiles too?
21:35:16amiconnWe want it removed, not added
21:35:28amiconnAnd the global $(GCCOPTS) are used everywhere
21:35:28Mikachuthey don't have any code in iram then?
21:35:38Mikachu(on arm)
21:35:39amiconnThat's a different story
21:35:48Mikachui thought ffunction-sections was for iram
21:36:12amiconnRemoving -ffunction-sections doesn't remove all long calls, but it removes long calls to static functions
21:36:21amiconn(-mlong-call has to stay)
21:36:54amiconnata.c is the only file where iram functions are also declared static - and gcc goofs and thinks it could use short calls for those
21:36:54dan_aDidn't linuxstb try getting rid of -mlong-call?
21:37:14 Join Rob2222 [0] (
21:37:30Mikachuas i understand, it's impossible for the compiler to know if it should generate a long call since it doesn't know where the linker is placing the function
21:37:33amiconnSo for arm, iram functions must not be static, or vice versa, static functions must not go into iram
21:38:29amiconnMikachu: Yes, however, the gcc manual states that gcc will use long calls as soon as it sees a section attribute
21:38:44amiconn...which is obviously not true if it also sees a 'static' qualifier
21:38:54Mikachubut that doesn't really help, does it?
21:39:23Mikachuwell, you can ignore me if you want, it won't help that i understand it :)
21:39:39dan_aOooh, icons! Pretty!
21:39:45amiconnI don't understand what part you don't understand
21:39:46peturamiconn: regarding output_dyn_value(): It has an issue with large values (like if you do dir properties of your whole music tree (33MB here)), I noted this when making my first patch:
21:40:05Mikachui thought long calls were needed when jumping from iram to normal ram, or vice versa
21:40:27Mikachubut gcc will only see a section attribute on iram functions, so it would put long calls when going into iram, and when going inside iram, but not out of
21:41:32amiconnpetur: The comment doesn't say what output_dyn_value can't handle
21:42:09amiconnWorks like a charm in all places where it's used, and I tested it across 9 orders of magnitude (the whole 32 bit integer range)
21:42:51amiconnThe rockbox info screen uses it, among other places
21:42:51peturI missed there, my tree is 33GB and that is > 32bit
21:43:26amiconnWell, if you count in bytes, the it will overflow 32 bit ints
21:43:36amiconnJust convert to KB first, and all is well
21:43:48peturwhat's the point then
21:44:01Mikachucan't it just take a long?
21:44:08amiconnA long *is*
21:44:10preglowlongs are 32 bits?
21:44:12amiconn32 bits
21:44:13preglowso no point
21:44:15Mikachua long long? :)
21:44:22amiconn(unless you're talking sims on amd64)
21:44:30 Quit YouCeyE (Read error: 104 (Connection reset by peer))
21:44:59amiconnpetur: The point is that it still scales to show KB, MB or GB with 3 significant digits
21:44:59peturit made the code bigger than doing it myself
21:45:30amiconnCheck how the info screen does it - it's dead trivial
21:46:00 Quit merbanan ("Leaving")
21:47:29peturamiconn: what file is that in
21:50:42peturah, info never goes down to bytes, only KB and up
21:50:57amiconnThat's what I said...
21:52:01amiconnfat_size() returns KB, insofar it's a bit easier here
21:52:45 Quit anathema (Read error: 110 (Connection timed out))
22:11:22LearHm, just managed to shave some 450 bytes of code and 350 bytes of bss from bookmark.c (coldfire). Seems to still work too. :)
22:12:04LearNo compiler set up yet...
22:12:50amiconnGetting rid of -ffunction-sections reduces code size for nearly every rockbox binary on arm (rockbox.ipod, *.rock, *.codec)
22:13:12amiconnFor some reason, wav2wv.rock grows quite a lot though
22:13:16preglowsmall wonder
22:13:28amiconnI guess it's due to inlining
22:14:12*amiconn should make a comparison list
22:15:48LearOops, filename display is wrong. :)
22:17:25directhexcoo @
22:17:55Bagderthey all release the same code anyway
22:18:22Bagderat least they have done in the past
22:18:36amiconnIt seems that getting rid of -ffunction-sections also reduces boost somewhat (*iirc* that is)
22:18:43Bagdermost often that is u-boot and linux kernel, with no fun drivers included
22:19:19amiconnI get 37% boost with 128kbps mp3 now
22:20:12amiconnMaybe that without -ffunction-sections -O2 will be better, as one would expect
22:20:21amiconnFor libmad, that is
22:22:34 Quit BigMac (Read error: 60 (Operation timed out))
22:23:56Nico_PBagder: how come there are small random variations in code size ? different compilers ?
22:24:05BagderI guess
22:24:11Mikachumaybe you should trigger the colors on >32 bytes different or something
22:24:20Bagderyeah possibly
22:24:24Mikachuor if you're bored, stronger color on larger diff :)
22:24:36Bagderwe'll see that better when we get some bigger changes happening
22:24:44Nico_Por have the same compiler on all build server ?
22:26:17 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
22:26:36Nico_Pwould i get yelled at if i were to commit the progressbar-y-coord patch ?
22:26:50Nico_P(very small one)
22:26:53LloreanNico_P: I think there was talk about that already.
22:27:37LloreanDo the y-cord patch stop it from taking up a line of text?
22:27:38 Quit printfXh4 (Connection timed out)
22:27:57Nico_PLlorean: what do you mean ?
22:28:18*Lear notes that it is a good idea to reboot before testing new code. :)
22:28:19LloreanAt the moment, the progress bar takes up one whole line of text, right?
22:28:29 Join entheh [0] (
22:28:33LloreanAs in, it offsets all text after it one line.
22:28:38LearAnd it is centered on the line in question.
22:28:53LloreanLear: Not necessarily, with the margins you can make it appear to be uncentered can't you?
22:29:08LearNot vertically, IIRC.
22:29:11Nico_PLlorean: good question
22:29:15LloreanOh, yeah, not verticaly
22:29:34 Join kubiix [0] (
22:29:45LloreanNico_P: Though you could always make that the very last line of text, 'offscreen' as it were, so it doesn't have to offset later lines.
22:30:24Nico_PLlorean: it probably eats the line but yes what you say works
22:30:27Nico_Pi'll check though
22:30:41LloreanBut basically, the progress bar should be able to work like it does now (the patch doesn't break existing WPSes, including those with non-graphical PBs), but when using the Y-Cord can it be essentially positioned like any arbitrary image?
22:31:24Nico_PLlorean: that's it. it doesn't break WPSes that don't have the y-coord and adds the possibility of having one
22:32:34Nico_PIt's a patch that's very used in custom builds because WPS creators love it
22:32:47LloreanYeah, but how used a patch is isn't always an argument for it.
22:32:53Nico_Pi know
22:32:53LloreanBut in this case, I think the patch is a good thing
22:33:14LloreanThe progress bar is now a graphical element, not a textual element, so arbitrary positioning makes sense.
22:33:30LloreanWe could already do it on the x-axis, but it was mostly pointless since you couldn't put anything on the same line with it.
22:33:40LloreanWith this patch, you can put it on the last line, and make it appear to be on the same line as text, right?
22:34:04amiconnHmm, 44% boost with -O2 and 45% boost with -O3. Quite the opposite from what I expected...
22:34:11Nico_PLlorean: probably. i never tried that
22:34:31Nico_Pi'll try
22:34:42amiconndan_a: Hmm, so it gets slower... that's not good
22:35:03LloreanNico_P: While I'm in favour of it, it seems to me that if done properly the patch should allow PBs to mimic the behaviour of images. :)
22:35:13amiconnThe other option is to un-static these functions for arm. It's a bit ugly but it shouldn't cause slowdowns this way
22:42:21 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:42:33 Quit amiconn (Nick collision from services.)
22:42:35 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:44:14blitheHooray for cuesheet support! Thanks Nico. :P
22:45:05*directhex wonders what use cuesheets are when one has gapless playback
22:45:15Nico_Pblithe: you're welcome and i'm gload you like it :)
22:45:29blithedirecthex: Think live concerts.
22:45:29 Join barrywardell [0] (
22:45:38Nico_Pdirecthex: live DJ mixes
22:45:39blitheNico_P: Exactly!
22:46:03preglowyou don't _need_ cue for that
22:46:06*petur cuts his live recordings up in separate files - no problem
22:46:12preglowmp3 and no cue works just as nice
22:46:29preglowmp3 + cue is a hack
22:46:35peturpreglow: Exactly!
22:46:39preglowthat was needed because media player coders were braindead
22:46:46Nico_Ppreglow: true, and i used to do that too but it's nice to not have to bother with cutting
22:46:54*petur hugs mp3directcut
22:47:01Nico_Palso my goal is to add chapters support for MP4
22:47:14Nico_Ppetur: i could never get mp3directcut to cut at the right places
22:47:19preglowNico_P: sure, i'm mainly just talking for new stuff here
22:47:28preglowthere's no point in making new mp3 + cue files
22:47:30Nico_Pprobably because i was using it with VBR mp3
22:47:41*amiconn doesn't know whether cue sh*t support even works properly on archos
22:47:46amiconnNo material to test with
22:48:34preglowamiconn: there's really no need to be derogatory
22:48:35dan_aamiconn: Would any slowdown in the disk be compensated for elsewhere by having more IRAM available? It also reduces the size of rockbox.ipod by 14k
22:48:39directhexpreglow, there's no point making new mp3 files! yay for vorbis! and i understand live sets tend to come in shn or flac format
22:48:40preglowpeople have yet to call your commits shit
22:49:05blitheWell, a lot of Electronica DJ sets get distributed as MP3 with a cue sheet.
22:49:27blitheSo it's nice to have support for it.
22:50:06directhexi wish i knew how to get ahold of recordings of concerts i've been to
22:50:18blithedirecthex: Not on
22:50:57directhexblithe, nope
22:51:21Nico_PLlorean: having the progressbar on a text line works well... it just has to be on a line that follows the text so that it overlaps it
22:51:25Nico_Plast line is fine for that
22:51:42petur directhex: record them yourself: get an iriver h1x0 or h3x0 with rockbox on, a good pair of mics and maybe a battery box or pre-amp. check for details
22:52:27peturor an x5 of course
22:53:03amiconnpreglow: I was referring to this .cue file idea. I can't see the point in it except maybe for cd copying (cue sheet does not only store track boundaries but also track indices). But in this case the .cue/.wav pair is temporary
22:53:23preglowi can't see the point in it at all, apart from being able to play files other people have made
22:53:27preglowof which i have a couple
22:53:41preglowmainly some dj sets
22:53:42amiconnI've never seen a single .cue/.mp3 pair myself
22:53:55blithePersonally I prefer to have my DJ sets as a single MP3.
22:54:02blitheI guess it all comes down to preference and usage.
22:54:16amiconnI ripped my dj set cd into multiple tracks. Plays perfectly gapless even on hwcodec
22:54:19Nico_Pi agree it's probably not very useful/elegant in the priciple but it does get used a bit
22:54:22preglowi don't really care, but i'd rather have several files than one
22:54:24amiconn(I used lame -nogap of course)
22:54:25Mikachui've seen cue+ape and cue+tta
22:54:30preglowif i want to listen to just one segment: copy one file, done
22:54:32Mikachuwhich i promptly converted to something sane
22:54:35Nico_Pand like blithe i prefer to keep DJ sets in one single file
22:54:37perl|worki have tons of .cue+mp3 as well
22:54:49perl|worksome .cue+flacs too
22:55:07amiconnAnyway, it would be useful to know whether cue support works on hwcodec
22:55:22Nico_Pamiconn: i'd be happy to provide you with material so you can test :)
22:55:27peturiiuc, mp3directcut can open a .cue and split it up into files
22:55:57LearStrange. Build of sh1 cross compiler fails. as complains about a '-big' option...
22:56:17amiconnWhat binutils and gcc version did you try?
22:56:40Lear4.0.3 and 2.16.1
22:57:04amiconnHmm... those are the recommended ones.
22:57:10Bagderyou on 64bit?
22:57:16LearNo, Cygwin.
22:57:42amiconnLear: as or sh-elf-as?
22:58:14amiconnMaybe you forgot to set the path after installing binutils?
22:58:25amiconnHappened to me too once...
22:58:43LearIt says "as" in the output. Path should be right...
22:58:53 Join BHSPitLappy [0] (
22:59:24LearHm, seems I've installed the wrong binutils in that path though... :)
23:00 :-)
23:00:14 Join webguest18 [0] (i=53c73228@gateway/web/cgi-irc/
23:00:22LearNah, "by hand" is the only way. :)
23:00:28*amiconn should try gcc 4.0.4 some time
23:00:43webguest18hello ! tplease there is a rockbox for ipod nano 2nd gen ?
23:00:43Bagderhehe, clearly
23:00:44amiconnPerhaps even 4.1.2 (except for coldfire)
23:00:55Bagderwebguest18: no
23:01:24webguest18thats sad, and rockbox for 1st nano is ok for the 2nd ?
23:02:31Lear4.1.3 is out, I think.
23:02:36directhexwebguest18, if it were okay, he wouldn't have said "no"
23:02:36*preglow suddenly understands that he can use the existing filtering routines
23:02:43preglowsometimes, people, it pays to think before coding
23:03:17directhexwebguest18, if that worked, he'd have said "yes, you can use the same one as the generation 1 ipod nano". but he said "no" instead
23:03:33 Quit kubiix ("Miranda IM! Smaller, Faster, Easier.")
23:03:39webguest18okay okay be cool, Im french, i dont understand all your subtilities
23:03:45webguest18thank you
23:03:53 Quit BHSPitMonkey (Read error: 145 (Connection timed out))
23:04:14Mikachuhow is "no" subtle?
23:04:32Bagderits not _what_ I said ut but how! ;-)
23:05:18 Join lukaswayne9 [0] (
23:08:50 Join BHSPitMonkey [0] (
23:13:43***Saving seen data "./dancer.seen"
23:14:13LearFirst bit in, but there's probably more that can be done (I have an idea or two to explore another day).
23:14:39 Quit SirFunk (Read error: 110 (Connection timed out))
23:14:59 Part Llorean
23:16:07 Part perl|work
23:17:56preglowwhy the dither in the size table?
23:18:02preglowdifferent compiler versions?
23:18:08preglowif so, we really should start unifying compiler versions a bit
23:18:22 Quit Arathis ("Bye, bye")
23:18:35Mikachua table with some red color in it can motivate you guys to do anything
23:18:55 Quit n1s ("Leaving.")
23:18:57directhexyou got red on you!
23:19:16LearOh, hadn't noticed the size table.
23:19:30 Quit BHSPitMonkey (Remote closed the connection)
23:20:22preglowi wonder what happened to it right now...
23:20:28Nico_PLear: nice one on the sizes :)
23:20:40preglowthere you go
23:20:44Nico_Pi'm gonna get a red one
23:21:10 Join Juice^ [0] (n=Juice@
23:21:19 Join Arathis [0] (
23:21:23preglowi'm starting to think perhaps grouping the sims at the end of the table would be wise, too
23:21:24pixelmathanks Lear :)
23:21:30preglowright now it's too easy to ignore a warning
23:21:53preglowhaha, and here comes a new feature dead on its heels to reclaim the space
23:22:13Bagdernice space deltas!
23:22:51Nico_Ppreglow: it probably won't add much
23:23:11Nico_Pat least i hope so... it's just a rewrite of existing code with very little added
23:23:29dan_aLear: Well done!
23:24:12Bagderthe new table does indeed at some new excitement while waiting ;-)
23:24:19Juice^2007-02-14: Icons in the menus - is this for all targets with a color screen?
23:24:46dan_aJuice^: It's on greyscale targets too
23:24:47Juice^or is it bw for all targets
23:25:07dan_aI don't know if the icons can be colour
23:25:46Juice^oh ok
23:25:54directhexwhat's the ifp7xx sim?
23:26:05Nico_PBagder: are you going to add a check to avoid displaying colors on very small size changes ?
23:26:10*amiconn wonders when Zagor will be around again
23:26:20BagderNico_P: probably
23:26:38amiconndirecthex: Iriver iFP 7xx series
23:27:08amiconnThe target is not yet up to a level where providing a current build would make sense
23:27:15amiconnOnly one dev is working on it
23:28:40directhexhm, random midrange flash-based player.
23:29:17directhexi wonder why fewer people seem to work on flash targets
23:30:17 Join midgey [0] (
23:31:33dan_aWe could do with someone having a look at People keep complaining that it doesn't work on IE, and then breaking it for Firefox by saying "- -Delimiter Content-type: text/html"
23:31:41 Quit dpassen1 (Read error: 110 (Connection timed out))
23:32:22LearOuch. Don't try perl programming via voice recognition. ;)
23:32:40midgeydan_a: safari isnt a fan of reading the current day log either
23:32:45midgeyno problems with older days
23:32:54amiconndan_a: That's why I was wondering about Zagor coming around. It's his work
23:33:11preglowLear: would almost sound like klingon
23:33:21dan_amidgey: Is the last thing in the log Pondlife saying that all he sees is ''?
23:33:21Bagdera red delta line there
23:33:21amiconnThe current log also has (slight) problems in firefox
23:33:25directhexhow does one access's source?
23:33:32LearThere's a video on youtube of one trying. Pretty funny...
23:33:36midgeydan_a: thats how it is in firefox
23:33:38Mikachuone asks zagor for it probably
23:33:49Bagder print "Content-type: multipart/mixed;boundary=Delimiter\n\n";
23:33:49Bagder print "−−Delimiter\n";
23:33:49Bagder print "Content-type: text/html\n\n";
23:33:53Bagderis the key for what people see
23:33:54midgeyon safari it just downloads
23:34:11*midgey likes the binary size table
23:34:13Nico_Phow's my delta ? not too bad ?
23:34:21Mikachu+100 bytes on average
23:34:28amiconnBagder: I guess IE wants \r\n after the header
23:34:47amiconnBut that's not the only problem - IE moans about the javascript
23:34:56Nico_Pi did manage to ged a green one :)
23:35:11midgeyNico_P: doesn't seem too bad
23:35:33midgeyany plans on committing the slider bitmap progress bar?
23:35:41amiconnNico_P: Yes, that's due to the recorder binaries being self-extracting, so the script checks the compressed size only
23:36:25Nico_Pmidgey: why not but i'll have to check people aren't stringly opposed to it and i'm waiting for a possible bug report to clear
23:36:27amiconnAnd Rec-8MB lost less size than Rec, although the code size change itself was identical on all archoses
23:36:36amiconnI mean in th eprevious commit
23:37:39midgeyare we using the same gcc and binutils on all build servers?
23:38:51preglowbut i think we should
23:38:57*midgey agrees
23:39:09amiconnI think we do for the targets
23:39:19amiconnFor the sims this will be difficult to achieve
23:40:01peturbut sims just have to build w/o errors
23:40:07preglowreally? then how do the size differences happen?
23:40:14preglowfor sims it will be hard, yes, i care more about targets
23:40:54Mikachuthe size table doesn't even show sims
23:41:00amiconnI observed an interesting effect yesterday: Building for the same target (Ondio FM) using the same binutils (2.16.1) and the same gcc (4.0.3 rockbox patched) yielded slighly different binary sizes on cygwin vs. debian-amd64
23:41:21Bagderthey do differ with the same gcc version
23:41:23preglowthen that's probably it, then
23:41:34preglownothing we can do about that
23:41:41amiconnThe difference was 16 bytes iirc (one section alignment step)
23:41:42dan_aWill it depend on what gcc gcc was compiled with, or something like that?
23:41:57amiconndan_a: That's the only reason I could think of
23:42:02Mikachumaybe even what cflags it was compiled with
23:42:08Mikachuphase of the moon..
23:42:08amiconn...but afaui that shouldn't happen...
23:42:23preglowit really shouldn't
23:42:32BagderI'll make the colors require >8 or < -8
23:42:34preglowbut i think it can with cross compiles
23:42:50preglowBagder: probably +/- 16 would be better
23:43:13Mikachui think only gcc 4.3 and later added self bootstrapping on installing, all older are built with the host compiler, at least some parts, afaiu
23:43:15*amiconn would really like to know how to get rid of the silly section (and string) alignment on sh
23:43:19Bagderok, 16 it is
23:43:39 Quit Juice^ ("Leaving")
23:44:00 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
23:47:24dan_aI remember when I tried to upgrade to the brand new Linux 2.0 kernel, I had to upgrade libc, which involved upgrading gcc, and step 1 was to compile gcc, and step 2 was to compile it again using your just-compiled gcc.
23:48:27Mikachuthen there's the crazy gentoo users who will recompile every package on the system after they update gcc
23:48:31 Quit miepchen^schlaf (Read error: 60 (Operation timed out))
23:49:38preglowdan_a: i believe it's common for gcc to bootstrap itself when not doing crosscompiles
23:50:11Mikachuprobably some parts requires gcc extensions :)
23:50:30 Quit Lear ("Chatzilla 0.9.77 [Firefox]")
23:50:33 Join miepchen^schlaf [0] (
23:53:08 Quit lex (Nick collision from services.)
23:53:12 Join lex [0] (
23:53:14 Quit lex (Nick collision from services.)
23:54:50directhexMikachu, you forget: compiling things yourself is cool
23:55:01preglowyes, very
23:55:54 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
23:56:37 Quit Tom0473 ("[G]Script - v3.3 ©daGroove - -")
23:56:41 Join Tom0473 [0] (
23:57:20Nico_PBagder: it's much nicer that way
23:57:39Nico_Pmaybe the total delta column needs a special value ?
23:57:49Mikachuwouldn't average delta make more sense than total delta?
23:58:12 Join lex_ [0] (
23:58:54Bagderyes I guess

Previous day | Next day