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

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

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

#rockbox log for 2007-02-21

00:03:39 Part n1s
00:06:08Shaiddataghost was working on a non-submitted version with 80g support.
00:08:27 Part LinusN
00:08:48 Quit funky ("leaving")
00:12:44 Quit printfXh4 (Read error: 145 (Connection timed out))
00:14:17 Join printfXh4 [0] (
00:15:37 Quit petur ("Zzzzz")
00:15:37*preglow wonders why use of bcd seems so popular with rtc chips
00:17:57Wincheste1hey this seems to be a recent bug....
00:18:05Wincheste1when I connect the USB cord to the ipod
00:18:14Wincheste1the usb logo comes up and just stays there
00:18:19Wincheste1without going to disk mode
00:18:43bluebrotherWincheste1: does the usb cord actually has a usb connection or is it only power?
00:18:51Wincheste1actual USB
00:19:14 Part redbreva
00:19:35pixelmathat seems to happen recently with some daily builds...
00:19:48Wincheste1using latest SVN
00:19:57 Quit midgey ()
00:20:05bluebrotheroh. Haven't heard of that yet
00:20:29 Quit Febs ("It's been a long day.")
00:21:31Wincheste1I'll just assume it exits and not press it....
00:21:45preglowcorrectly assumed
00:22:11Wincheste1not on GAIM
00:22:31pixelmabluebrother: I think linuxstb and amiconn are aware of that
00:22:53Wincheste1so that will get fixed?
00:23:00pixelmaso maybe a few svn builds are also affected by it
00:23:10Wincheste1if it helps I noticed it about 2 days ago
00:23:37Wincheste1compiled from latest source.... still there
00:24:11pixelmaoh? Maybe it is something different than what I thought
00:24:12 Quit Mouser_X (Read error: 110 (Connection timed out))
00:26:08*preglow goes ahead and plays around with the pcf wake-up fields
00:26:20bluebrotherpixelma: wasn't aware of it yet, but I'm quite busy with RL atm ... makes it a bit hard to keep up to date ;-)
00:26:58bluebrotherhmm, the recent svn log on the front page doesn't cope with utf8 characters while the changes page for the commit itself does.
00:27:01webguest75I can't figure this out. I'm using cygwin to build my sim. Up until the about the 15th everything was working fine. Since then, when I load my WPS, it loads the wrong background and doesn't show any of the WPS tags
00:27:09webguest75Has anyone encountered this yet?
00:28:58pixelmabluebrother: I was forced to notice because I was affected by it when someone lend me his Mini for a few days ;)
00:29:12Wincheste1also... running make doesn't seem to recompile a plugin with changed source
00:29:29pixelmathough it was alright when compiling my own
00:30:14 Join Thundercloud [0] (n=thunderc@
00:30:17B123and i got it all working now
00:31:56 Quit Juice^ ("Leaving")
00:31:59x1jmpWhat can cause a ATA error -11?
00:32:38 Quit desowin ("use linux")
00:33:02 Quit Vulc4n ("Leaving")
00:33:35Soapbluebrother: I would consider 85EUR high for a Mini 1st gen, and about right for a mini 2nd gen, assuming you get the cable, the dock, good condition, etc. 85EUR for a Mini 2nd is quite good if there is a recent replacement battery.
00:34:06 Join pearldiver [0] (
00:34:13SoapI base this off of US prices, specificly Craigslist where I have been hunting for a Mini#2 myself to swap to CompactFlash.
00:35:11preglowin a fine twist of irony, thunderbird thinks christopher woods' latest mail to the ml is spam
00:35:49preglowfirst ml post it has ever marked as spam
00:38:07Soapahh, bluebrother, now that I do the euro to dollar conversion in google instead of my head, I'd say 85 is pushing the high side of reasonable.
00:38:50 Join Dryden [0] (
00:39:55SoapI got a iPod 5th gen 30GB for $80 yesterday, dead battery, no cables. Figure 20 for a battery delivered 10 for a cable and that's right about 85 euros. But I think that is not the fairest example as I consider that 5th gen a really good deal.
00:39:56 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
00:40:32Kasperlebluebrother: :D
00:41:06 Join SirFunk [0] (
00:41:22Soappreglow: my thunderbird thinks the email is a scam.
00:41:39SoapI think it's the URL.
00:42:47 Join midgey [0] (
00:43:23 Join Mouser_X [0] (
00:45:26 Join Winchester [0] (
00:45:51Winchesterhmm... dodgy net
00:46:08Winchesterso... is there a page about bmp2rb?
00:47:06Winchesteror am I disconnected again?
00:47:52JdGordonbmp2rb has usage info if you run it with no args
00:48:47bluebrotherSoap: thanks. Will think about it, but it looks like ipods are still quite expensive here (germany)
00:49:05bluebrotherbut now first out for some sleep
00:49:20 Quit bluebrother ("leaving")
00:50:35Winchesteryeah but.... for greyscale what method does it use.... does it try to blend from shade to shade?
00:51:23 Quit dewdude (Read error: 145 (Connection timed out))
00:52:13Winchesteroh and what I mentioned before about RB not going into disk mode..... that was more of a linux problem 'cause it wasn't unmounted properly so nvrm that
00:53:25preglowhrm, looks like the bootloader actually enabled the watchdog
00:58:01 Quit Thundercloud_ (Connection timed out)
00:58:29 Quit Wincheste1 (Read error: 110 (Connection timed out))
00:58:38B123now that i figured this all out lmao
00:58:46B123i dont see why anyone would use the original mac OS
00:58:48B123on the ipod
00:58:51B123this ones just so much better.
00:59:20Soapbattery life, itunes store, audible books, car headunits
00:59:41B123i suppose.
00:59:42B123car headunits?
00:59:52Soapones that control the ipod./
00:59:59B123like...a itrip?
01:00:19Soapno, actual car stereos which control the ipod.
01:00:26B123Oh gotcha
01:00:34Soapsome factory ones which allow you to control the ipod from the steering wheel.
01:00:40B123I was thinking that the itrip would still work pretty good with this.
01:00:44B123Yeah i know what you mean now.
01:01:03 Join midkay_ [0] (
01:01:31 Quit midkay (Nick collision from services.)
01:01:33 Nick midkay_ is now known as midkay (
01:01:50 Quit Dryden ()
01:01:54 Quit B123 ("KVIrc 3.2.4 Anomalies")
01:02:29Soapany version of the iTrip where you program the frequencies from the ipod won't work.
01:02:39Soapwe have a wiki page _started_ on the issue.
01:02:58 Join dewdude [0] (
01:03:55 Join fasmaie [0] (
01:06:59 Quit ender` (" Where there's a will, there's an inheritance tax.")
01:07:02 Quit midkay (Read error: 104 (Connection reset by peer))
01:07:35 Join dezgot [0] (
01:08:01JdGordonhehehe"> ... not quite working :p
01:11:05 Quit entheh ("^~")
01:11:17 Join webguest69 [0] (i=4a84afb0@gateway/web/cgi-irc/
01:11:46 Quit webguest69 (Client Quit)
01:15:49 Quit busa_blade ("Leaving")
01:16:33***Saving seen data "./dancer.seen"
01:17:23 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
01:17:38 Join miepchen^schlaf [0] (
01:18:02 Part dezgot
01:20:59preglowjhMikeS: that's a pretty nice emac specific hack you've commited, sir :>
01:21:25preglowfracmul only saturates on coldfire, on arm, it wraps around
01:22:29jhMikeSI don't think there any danger since the sample only approach from the 24 bit side
01:22:58 Quit Mouser_X (Nick collision from services.)
01:23:07 Join Mouser_X [0] (
01:23:16jhMikeSJust plopped it in that way...should I change it so as not to deceive?
01:23:25preglowif it works, keep it
01:23:48preglowbut i'm not very keen on platform specific codec functionality, but then again, dsp.c is full of that, so...
01:24:23jhMikeSIt would have to be an awfully loud spc to wrap that mult ;)
01:24:23preglowi would have given an arm (pun intended if you don't think it too lame) for an arm with saturating math instructions
01:24:44jhMikeSThey are nice. But that emac does some nice tricks
01:24:52preglowi like the emac
01:24:59preglowit's fast and does what's needed
01:25:43preglowwell suited to dsp
01:26:03preglowas opposed to the arm stuff which really just does generic long long math
01:26:07 Part toffe
01:26:08preglowand is slowish
01:27:09jhMikeSI can't say it enough...what were they thinking by not included bitrev and byterev in ISA_A ?? That one ticks me off a lot too.
01:28:00preglowwould be nice with an arch that has what i want for a change
01:28:40jhMikeSpreglow: at least it's more like Intel's instructions with multiple registers
01:28:42 Quit printfXh4 (Remote closed the connection)
01:29:00preglowit would be pretty cool if they coudl saturate
01:29:03preglowcould, yes
01:29:11preglowthat my major nitpick
01:29:33preglowhaving to do major register extraction work just to get an extended result with the emac tends to bore me too
01:29:47preglowespecially since it's so with the actual calculations
01:29:57preglowso fast, that is
01:30:01 Join daxxar [0] (
01:30:27daxxarI'm looking at getting a small (flash-based?) mp3-player, 4-8gb. Battery time is a great asset in my eyes.
01:30:36*preglow is curious to see if his nano will refuse to switch on again now
01:30:40jhMikeSI hate those durn extension bytes. And then you gotta work handle 2 accs at once which is bs
01:30:49daxxarDo you guys have any suggestions? I'm looking at the iRiver E10, but it doesn't seem to be rockboxed (yet)
01:30:55daxxarThe E10 has great reviews afact.
01:31:16preglowwell, it's not rockboxed yet
01:31:31preglowi haven't heard any talk of it being rockboxed either
01:32:09preglowthings have been slow on the iriver front since h10
01:33:22 Join midkay [0] (n=midkay@rockbox/developer/midkay)
01:34:17preglowfor a second there, i thought i had bricked my nano
01:34:27Mikachuplaying with the wakeup registers?
01:34:46jhMikeSpreglow: i don't understand the platform specific code aversion in codecs. imho it's called for wherever it's beneficial enough. :\
01:34:49preglownothing too fancy, but you never know
01:35:42preglowjhMikeS: i don't mind platform specific code, i mind platform specific behaviour, like clipping happening on coldfire platforms, but not arms
01:36:08preglowright now there's nothing keeping anything in dsp from overflowing on arm :/
01:36:14preglowalso, everything is one bit less accurate on arm
01:36:23preglowdoesn't much matter, but i still don't like it :>
01:36:29*preglow intends to wake by nano tomorrow
01:36:31 Quit midkay (Client Quit)
01:36:37 Join midkay [0] (n=midkay@rockbox/developer/midkay)
01:36:57*jhMikeS predicts preglow's gonna be late for something tomorrow :D
01:37:13Mikachuheh, i would use a backup-alarm the first time
01:37:16preglowi'm gonna have a backup
01:37:23 Quit midkay (Client Quit)
01:37:32Mikachuputting the beer opener on a timer?
01:37:35preglowi always have one of those in case i'm so tired i just switch whatever my first means are off and go to sleep again without remembering it
01:37:56 Quit Winchester (Read error: 110 (Connection timed out))
01:37:57 Join printfXh4 [0] (
01:38:07preglowone time i did demolish an alarm clock without remembering it
01:38:25preglowi'm going to be pissed if i wake up to a crushed pair of headphones with sprinklings of ipod nano on top
01:39:00jhMikeSpreglow: well, I _really_ can turn off the saturation if you want them the same :) Maybe you'll sleep better? :D
01:39:13preglowjhMikeS: haha, no, i'm fine with it
01:39:28preglowlike i said, i don't really know how to make arm do that efficiently, and saturating is the better way
01:40:11preglowit's a bit too nazi to nerf all platforms down to the common denominator
01:40:21Mikachusaturating in this context means something like 0xff+1 == 0xff ?
01:40:25preglowMikachu: aye
01:40:36preglowit's generally the way you want to mix audio
01:40:45Mikachudoesn't it cause clipping?
01:40:45preglowsince overflowing sounds very, very bad
01:40:50preglowMikachu: it _is_ clipping :)
01:40:56Mikachuheh, indeed
01:41:03preglowbut it sounds better than overflowing
01:41:15 Quit Lynx_ (Excess Flood)
01:41:31preglowsometimes you want overflow in the intermediate stages of calculation since the result can always end up in the correct range, but for output, you always want saturation
01:41:48 Join Lynx_ [0] (
01:42:34pregloweverything dsp can do saturating math instructions for free, but portalplayer didn't care too much about that
01:43:03 Quit jhMikeS (Nick collision from services.)
01:43:09 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
01:44:30JdGordonto make a 1d array 2d (and then get a cell) I would do array +(width*y) + x right?
01:45:45*jhMikeS has to take a side step and figure out the deal with the x5 lcd backgound flashing white at startup while booting an old version with current bootloader doesn't have that symptom :\
01:45:51 Join Winchester [0] (
01:45:54Mikachuusually array[width*y + x]
01:46:04Mikachubut the result is sort of the same
01:46:49Winchestergot disconnected again.... "Winchester: hmm.... how come rockbox doesn't seem to use the %p conversion character?"
01:46:51jhMikeS...and the current lcd code isn't changed from before either. hrm
01:47:06JdGordonMikachu: ta :) so Im not doing the math wrong at least..
01:47:59 Quit Mouser_X (Nick collision from services.)
01:48:08 Join Mouser_X [0] (
01:48:44*jhMikeS suspects the contrast setting flopping all over the place since the bootloader text remains during the flicker
01:48:49MikachuWinchester: look in firmware/common/sprintf.c
01:49:08preglowWinchester: rockbox' printf implementation isn't exactly fully standards compliant
01:49:26preglowjust use %x
01:49:34Mikachui think all pointers on rockbox targets are 32bit
01:49:38preglowthey are
01:49:45Winchesterthere's a printf?
01:49:52preglowthere's an snprintf
01:49:59 Join midkay [0] (n=midkay@rockbox/developer/midkay)
01:50:00Mikachui only saw sprintf
01:50:02*Mikachu runs ls again
01:50:32preglowthen have a look inside the file :)
01:50:48Mikachui meant .c
01:50:49 Quit webguest75 ("CGI:IRC (EOF)")
01:50:54Mikachui can't be held responsible for bad naming :)
01:51:06Winchesterisn't snprintf used to fill variables or something?
01:51:49preglowWinchester: well, yes, in the same way that everything in c is
01:52:11Nico_Pamiconn: is the fix for the AMD64 warnings ?
01:54:09WinchesterI don't get it.... but %x it is then
01:54:35Mikachusprintf is like printf but fills a *char instead of stdout
01:54:41Mikachumaybe you were thinking of scanf
01:54:46Mikachu(there is also sscanf)
01:55:19Winchesterdo you mean snprintf or is there a sprintf?
01:55:31Mikachuthe n just limits the size of the *char
01:56:04 Quit linuxstb ("CGI:IRC")
01:56:10Winchesterand what's sscanf for?
01:56:35Mikachuwell, scanf reads from stdin and sscanf reads a *char :)
01:56:40Mikachuit's all very logical
01:57:14Mikachuit fills in all the %i and whatnot from a string, instead of the other way like printf does
01:57:45Mikachuyou can make very bad config file parsers with it for example
01:58:07Winchesterso.... instead of using splash or putsxy straight away I should use sprintf first?
01:58:45preglowdepends what you want to do, if you just have a string, you can splash it straight away
01:58:57preglowif you first want to make a string from some other variables, then first you need to snprintf
01:59:46 Join griphiam [0] (i=d03ba035@gateway/web/cgi-irc/
02:00:02Winchesterwhat's the reasoning behind that?
02:00:16griphiamCan anyone answer a few questions on how the Database *.tcd files work?
02:00:20jhMikeSpreglow: so what happens on ARM with large sample depths (like the 28 output by most of the codecs) and huge filter boosts?
02:00:31preglowthe alternative would be having all display functions take printf style parameters, which would bloat code
02:00:32griphiamWhat is the difference between the tag*.tcd and database*.tcd
02:00:46griphiam(in the .rockbox folder of course)
02:01:44Winchesteryeah I was kind of just using it all as I would with printf.... seemed to work fine >.>
02:01:54preglowjhMikeS: right now, nothing special
02:02:02preglowjhMikeS: and now that you mention it, it really should overflow
02:02:15preglowthe three extra bits are not enough to accomodate 24 decibels of gain
02:02:33preglowthat might yield a max amplitude of a bit less than sixteen
02:02:44preglowwhich would need 4 int bits to accomodate
02:02:54Winchesterwell anyway..... thanks for answering my stupid questions.... >.> I'll try to cut down on those for a while.
02:03:31jhMikeSI figured that one out already. :) 24db is enough then eh for everything?
02:04:02jhMikeSpreglow: thought you had four for arm with the 1bit reduction
02:04:49preglowjhMikeS: hmm? the output formats are the same, which is what matters
02:04:50JdGordongriphiam: you can delete the tag*.tcd files... they were renamed to database.tcd
02:05:17preglowjhMikeS: and to make matters worse, the mac instructions on arm can't even bloody flag an overflow
02:05:39preglowif they did, then clipping manually would actually be pretty easy
02:05:42preglowand fast
02:06:06griphiamJdGordon: Is there any way to get rockbox to recreate the database*.tcd files from the tag*.tcd files other than removing the database*.tcd files and reiniting?
02:06:36preglowjhMikeS: but i've been meaning to check this out some day, if we can't depend on having saturating instructions available, we might want to start using even more guard bits than the three we currently do
02:06:38JdGordonjust rename them
02:06:46jhMikeSpreglow: arm has mac instructions too then but they're just a cheap wannabe imitation of cf? :)
02:06:55JdGordongriphiam: i dont think the actual file format was changed recently
02:06:59JdGordonjust the name
02:07:16jhMikeSpreglow: Then output 27 bits for arm and leave my cf bits alone ;)
02:08:11preglowjhMikeS: sure they do, smull does a 32 x 32 -> 64 mul, and smlal does that plus an accumulate, both give a full precision 64 bit result which i shift one bit left, yielding the same answer as on coldfire, sans clipping
02:08:18griphiamI'm just trying to develop a workflow using Floola... Since Floola modifies the iTunes database, I need to rerun itdb2tc again to reexport the data
02:09:17jhMikeSpreglow: why doesn't the dither stage just do the final sample output instead of doing a second pass?
02:09:37 Quit Mouser_X (Nick collision from services.)
02:09:44 Join Mouser_X [0] (
02:09:56preglowjhMikeS: laziness
02:10:05 Join frag [0] (
02:10:23preglowi wasn't even sure if it was worth keeping, so i didn't groom is as much
02:10:32fragLo All!
02:10:33preglowbut then people didn't say "omfg, wtf, this is a waste of space, remove it" so i kept it
02:10:37 Join toffe [0] (
02:11:14jhMikeSnot very much code or size in that
02:11:43 Nick frag is now known as [CK]Fragula (
02:11:57 Part pixelma
02:12:02 Join Kret [0] (
02:12:15Krethello all
02:12:23[CK]FragulaLo Kret!
02:12:27*jhMikeS sees dsp_cf.S taking on more responsibility
02:12:28Kretanyone here know what SVN means?
02:12:34preglowKret: subversion
02:12:51preglowjhMikeS: as it should be
02:13:06preglowdsp traditionally is a very asm heavy part
02:13:19Kretpreglow: subversion as in an alternate or other than main version? <:)
02:13:19jhMikeSin the Roman empire: SVN = SUN ;)
02:13:50KretjhMikeS: yeah..that's true..
02:14:02preglowKret: as in the source code management system
02:14:16preglowgoogle will fall over out of eagerness in helping you
02:14:41[CK]Fragulawhich is a bity biased i guess
02:14:56 Nick jaebird_ is now known as jaebird (
02:15:02[CK]FragulaThere are /lots/ of flavours of subvsions
02:15:14preglowtortoisesvn is nice enough
02:15:17Kretpreglow: point taken...was just curious as to what it was :P
02:15:46preglowKret: i don't mind, just making sure you know i won't be explaining to you what it is in detail, heh
02:16:03griphiamIs there a document interface for the database format?
02:16:07[CK]Fragulaidd. I have to use Windows<spit> at work due to being a Steinberg pilot, but it keeps me in touch with the sauce tree
02:16:19preglowbhargh, i try to install cmake, and apt-get wants to bloody install emacs with it
02:16:28preglowthe lord save me from emacs
02:16:36Kretpreglow: well, whatever you're willing to give i'm grateful for...hmm...
02:17:04preglow[CK]Fragula: steinberg, as in cubase steinberg?
02:17:14[CK]FragulaYeh... :-(
02:17:19*preglow shudders
02:17:22*preglow crosses himself
02:17:25*[CK]Fragula cries
02:17:38Kretso how likely is it for that spc and nsf player to be added to CVS?
02:17:46preglowKret: they're both in svn right now
02:17:52preglowKret: playing just nicely too
02:18:40jhMikeS...which apparently in Latin has no meaning at all
02:18:49Kreti'm using kosh's build and he's been busy it seems..hasn't gotten around to updated his release...guess its time to start doing my own builds..
02:19:31preglowwouldn't know anything about any custom builds, no
02:20:18Kretthanks for the info preglow...i'm heading out and getting started..i know this is gonna take all night to get going...*wave*
02:20:28 Part Kret
02:20:54 Join webguest98 [0] (i=47cdc81c@gateway/web/cgi-irc/
02:20:56 Join Vulcan [0] (
02:21:21 Quit Mouser_X (Nick collision from services.)
02:21:31 Join Mouser_X [0] (
02:21:44webguest98what happens if you try to hook up a rockbox ipod to an ipod compatible car radio?
02:22:51webguest98scorche: oh so the radio wont see the disk?
02:23:01Vulcanhow would it?
02:23:22VulcanI guess its possibl it could work booted into disk mode
02:25:06griphiamwhat is supposed to be stored in database_6.tcd and database_7.tcd ?
02:25:13griphiamthey look pretty empty
02:25:15webguest98Vulcan: i guess that would be ok if i could find a receiver that can play ogg and flac
02:25:38jhMikeSpreglow: I read in the logs XavierGr was having difficulty...just didn't get the oops fix?
02:25:46Vulcanwebguest98, theres about zero chance of that
02:26:04Vulcanyour best bet is to just run to the line in on the head unit
02:26:23webguest98Vulcan: there is just phatbox and something called the ural that looks discontinued since 2003
02:26:23Vulcanand it will work with everything
02:26:34jhMikeSXavierGr: well if you're it ok now?
02:26:41Vulcanwebguest98, probably junk.
02:26:42preglowjhMikeS: with what?
02:26:49preglowjhMikeS: ah, yes, that, it's ok now
02:26:57Vulcanthe only other option I can think of is a carputer or soemthing
02:27:06webguest98Vulcan: like a mini-itx
02:27:41Vulcanyou can get a cheap micro itx board
02:27:44Vulcanand a pico psu
02:27:50webguest98Vulcan: it just gets so cold here though and hot
02:27:51*jhMikeS wants to compare the hardware with the em and will dig out the SNES now
02:28:01Vulcanadd in a hard drive and some sort of screen and you'd be set
02:28:26Vulcanwebguest98, I think the only device I would be concerned aobut in temp extremes is the HD
02:28:34Vulcanhard drives don't do well in the cold
02:28:40webguest98Vulcan: which is a concern
02:28:52Vulcanyeah, don't know of a good solution to that
02:28:57VulcanI'm sure people have come up with some
02:29:05webguest98Vulcan: solid state hd? :)
02:29:18Vulcanmaybe in 5 years
02:32:02 Quit fasmaie (Read error: 110 (Connection timed out))
02:32:07griphiamI'm not having very much luck greping through the source code... is the itdb2tc tool controlled in the rockbox svn tree?
02:32:08Vulcansomeone buy my zuuune
02:32:10webguest98Vulcan: seems like the easiest thing would be to just get a head unit and plug in a laptop
02:32:21Vulcanmaybe webguest98
02:32:25Vulcanor use a portable hard drive
02:32:36Vulcanthat you can take inside when it gets cold
02:32:55Vulcanwell, really its just you don't want to use it while its cold
02:33:24webguest98Vulcan: or get a mac mini
02:33:39Vulcanmac mini is pretty cool
02:33:46 Quit x1jmp (Read error: 110 (Connection timed out))
02:33:59webguest98would still need an lcd though for that i bet
02:34:27webguest98Vulcan: i dont need to do spreadsheets in the car just play music
02:35:37 Join CpuWhiz [0] (
02:36:22JdGordongriphiam: no, it2rb is seperate
02:37:08 Quit griphiam ("CGI:IRC")
02:39:36 Quit [CK]Fragula ("Chatzilla [SeaMonkey 1.1/2007011111]")
02:42:18Soapscorche: since Febs and LL are gone I'll ask you.
02:42:32SoapDid axlgreasetires lock his own thread? can people do that?
02:45:00 Quit YouCeyE (Remote closed the connection)
02:46:27scorcheSoap: i think they can
02:46:55jhMikeShmmm...a little noisier (to be expected with the all the linear interp and quietude)...but everything's dead on otherwise. :)
02:49:19preglowi've heard some problem with noise
02:49:27preglowas in wind effects and so on
02:49:32preglowbut otherwise it's surprisingly good, yes
02:49:45 Join V3Hero [0] (i=931acbee@gateway/web/cgi-irc/
02:50:17jhMikeSat least the anomalies in the echo match the hardware (town in Actraiser)
02:50:41preglowthey're probably not anomalies
02:50:47preglowdamn, you made me want to play actraiser now
02:51:39jhMikeSwhen the music first starts in "town" it's got a pulse that dies out
02:51:53jhMikeSguess I should continue that game I save a decade ago ;)
02:52:12 Quit inversions (Read error: 110 (Connection timed out))
02:52:23 Join bonbonthejon [0] (n=jon@
02:54:07 Quit V3Hero (Client Quit)
02:56:49preglowi did complete it
02:56:54preglowbut that's about what i remember of it
02:57:00preglowapart from it being a pretty fun game
02:57:21 Quit Nico_P (Remote closed the connection)
02:59:50 Quit Thundercloud (Remote closed the connection)
03:08:29preglowthe lang cleanup boys should exchange "id3" for "tag" for the platforms where it applies
03:12:29preglowanywho, high time for sleep
03:12:48 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
03:13:15 Join miepchen^schlaf [0] (
03:13:26 Join Demonio [0] (
03:13:34DemonioAnyone here?
03:16:37***Saving seen data "./dancer.seen"
03:18:11VulcanI'm here
03:18:22*Vulcan is trying to milk idiots on ebay
03:27:22 Part Demonio
03:27:35XavierGrgod damn I am jealous of you jhMikeS, owning a real snes!!
03:28:07XavierGrthough I didn't have a problem with spc even without the fix
03:28:20XavierGrit was just that the new spc files I got had two versions of the same song
03:28:28XavierGrone with echo and another one without
03:29:05XavierGrso I was confused and thought that the spc codec had problems with the new echo addition
03:30:18 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
03:33:21 Quit JdGordon ("Leaving.")
03:33:22jhMikeSXavierGr: :) It seems to have picked up a video problem just sitting in the closet...vertical lines to the right of sprites sometimes :\
03:34:23Vulcanclean your contacts
03:34:31Vulcanand blow out the cartridge
03:34:34Vulcanthe usual stuff man
03:35:33jhMikeSdid all that and no's got funky screws though and I think should give look inside
03:35:50Vulcantorx a hemisphere with six notches
03:36:44jhMikeSnot quite reverse torx
03:43:01 Quit YouCeyE (Remote closed the connection)
03:44:57*jhMikeS tenitavely blames aging caps
03:45:20Shaidtheyre the special nintendo screws.
03:45:27ShaidIve got a bit that undoes them
03:45:36Shaid(not that it came with my gc modchip or anything)
03:46:01jhMikeSI don't. :) I'm sure I can make something up.
03:46:11ShaidYou can
03:46:21Shaidwith the plastic bit from outside a cheap biro
03:46:33Shaidjust melt the end, put it over the screw, let it resolidify, and bam
03:46:45Shaidgo read up on opening a GC, youll find the method :P
03:47:09hcsI've tried that a few times with an N64, just wound up destroying several pens
03:48:37jhMikeSlook like there's enough room around each to notch out a slot screw driver and grab two of the notches
03:51:01 Quit SirFunk (Read error: 110 (Connection timed out))
03:51:04jhMikeSIf the melt technique doesn't work, I can also spray parting film inside and cast it with epoxy weld
03:51:33hcsI took apart an SNES with a hammer once.
03:52:55jhMikeShrmph...I'm afraid that option is out :)
03:54:09 Join YouCeyE [0] (n=YouCeyE@unaffiliated/youceye)
03:57:25XavierGractually blowing is what I did with my gameboy all the time
03:57:34XavierGrbut I've read somewhere that it is actually bad
03:57:49XavierGrdue to corrosion on contacts from breath moisture
03:58:20Mikachucompressed air should be the best, but maybe overkill
03:58:34XavierGrand yeah I remember those bloody nintendo screws, I still can't get my Game Boy open because I can't find a damn screwdriver that can turn them
03:58:34Mikachuyou can always polish the contacts if you have to
03:59:09 Join JerryLange [0] (
03:59:24XavierGrsomeone crazy enough should port rockbox to gameboy and run the gameboy emulator from it :P
04:00:06 Join ep0x [0] (
04:00:12 Nick ep0x is now known as w00t (
04:00:17w00tWhere are my songs @ ?
04:00:19JerryLangehaha that would be funny.
04:00:25 Nick w00t is now known as weet (
04:00:38weetI just installed RockBox
04:00:38Mikachumaybe you should stop changing your nick every 5 seconds
04:00:42weetcant find my songs
04:00:44weeteh sorry
04:01:27weetbut where are my songs ? :/
04:02:25weet5g Video
04:02:49XavierGrwell I am not sure about the ipod rockbox installation but you better kept a copy of them :P
04:03:10weetyes i did lol
04:03:21weetthe problem is, theirs 2 Gig missing in the Ipod
04:03:26weetso the songs are still their sumwheres
04:03:47 Quit JerryLange (Remote closed the connection)
04:03:53XavierGrsorry can't help you more, I don't have an ipod and I am not sure how the installation works
04:04:57weetwell its installed
04:05:00 Part CpuWhiz ("Leaving")
04:05:03weetworks flawless
04:05:09weetits getting the songs back
04:05:15weetwich i dont understand
04:05:15weetin the tutorial
04:05:57XavierGrdid you read this on the link: How do I access music that I have transferred to my iPod using iTunes
04:06:02 Join dan_a_ [0] (n=dan-mirc@
04:06:23XavierGrtake your time
04:07:05weetI will right now
04:07:12weetlol its kinda hard to take my time im a student
04:07:38 Part dan_a_
04:08:18 Join l00pcereal [0] (
04:10:19esp8any other packs of games to throw on my rockbox?
04:11:24l00pcerealit says i have to activate " view all hidden folders "
04:11:28l00pcerealthats a nice feature :)
04:11:42l00pcereal me = weet
04:12:29XavierGralso read the manual for keymaps
04:12:48l00pcerealwell i think i will convert
04:12:52l00pcerealmy Ipod Database first
04:13:52l00pcerealthink im gonna put Doom
04:14:32l00pcerealanother question quick
04:14:36esp8yeah ive got doom on there but i was looking for rockboy and i dunno other games other then what commes already on the rockbox site
04:14:41l00pcerealhow do we, apply patches ?
04:15:14XavierGryou have to learn to compile the source first
04:15:25XavierGrthen you apply the patch to the source code and recompile
04:15:31l00pcerealoh god
04:15:39l00pcereali don't have c# knowledge
04:15:41esp8only apply patches if you have problemes
04:15:43 Quit weet (Read error: 145 (Connection timed out))
04:15:46XavierGrwell patches are not meant for users anyway
04:15:54esp8same as for any patch.. only if you have problemes
04:16:02XavierGrthough you don't need to know programming to set up a build enviroment and compile
04:16:42XavierGresp8 for rockboy games you must search the internet
04:16:46XavierGrgoogle is your friend on this
04:17:02esp8yeah for the rockboy games i know that ive got some roms on cd... but for the rockboy emu....
04:17:16esp8and other emu's or nice appz to throw on my rockbox :)
04:17:24XavierGrthe rockboy emu comes with the build
04:17:39XavierGryou don't have to download anything more
04:17:52XavierGrif you run rockbox on your device you most probably have the emulator too
04:18:18esp8oh... cant find it...
04:18:27XavierGryou just play the rom file
04:18:32XavierGrand the emulator will start
04:18:43esp8any .gba pretty much
04:18:51XavierGrnot .gba
04:18:55esp8what extensions? and .gbc
04:19:09XavierGrgba is for game boy advanced
04:19:17esp8yeah so it dont run a GBA? :P
04:19:18l00pcerealARE YOU SERIOUS ?
04:19:25l00pcerealtheirs a built-in GB emulator ?
04:19:25esp8yeah cereal :P
04:19:42XavierGryes it CAN'T run gba
04:19:50 Join toffe_ [0] (
04:19:59l00pcerealit can run .gb files ? :o
04:20:07 Quit Mouser_X (Read error: 110 (Connection timed out))
04:20:14esp8can it run nes files? :D
04:20:19XavierGrcomeon weet read the FAQ
04:20:25XavierGrall the answers are there
04:20:42XavierGresp8: nope only nes music files
04:20:52 Quit webguest98 ("CGI:IRC (EOF)")
04:21:09XavierGr(add coma after nope)
04:21:31esp8:( so it has built in name and gb emu no other emu's can run on a ipod?
04:21:41 Nick l00pcereal is now known as weet (
04:21:44weetI will
04:21:46weetdon't worry
04:21:48XavierGrwell there is the doom emulator too
04:21:49weeti will read everything
04:21:53XavierGr(if you can call it emulator)
04:21:58XavierGrand some other minor ones
04:22:08XavierGrlike chip8 and z80 (IIRC)
04:22:19weetthis rockbox thing is better then Ipod Linux
04:22:21esp8yeah ive got that
04:22:58esp8chip8? z80? it can run old shit like off a commodore :P
04:23:23hcsc64 had a 6502
04:23:46esp8so it would't run?
04:24:03hcsc64 tunes (sid) run
04:24:08hcsnes tunes (nsf) run
04:24:12hcssnes tunes (spc) run
04:24:17hcswhat more could you want?
04:24:34esp8why is it only the tunes?
04:24:46esp8couldt you play willy beamish on your ipod?
04:24:47hcsbecause rockbox is an audio player
04:24:55esp8yeah but its got appz too
04:25:00esp8and video player...
04:25:00XavierGrsorry replace z80 with ZX Spectrum 48k emulator
04:25:46XavierGrdamn I've never fiddled the ZX emulator it should be fun
04:25:53 Quit toffe (Read error: 60 (Operation timed out))
04:30:04weetI created the Doom folder
04:30:07weetand it wont appear
04:30:15weeton rockbox
04:31:13XavierGrwhere did you created?
04:31:21XavierGrinside .rockbox folder?
04:33:15esp8do i need to extract the .GB files?
04:33:20jhMikeSthis stuff won't work for sega genesis? would be cool to have that if possible.
04:33:49esp8NES would be at least nice :P
04:34:05jhMikeSI'm just talking the tunes part actually
04:34:59jhMikeSbesides, if the sound emulation is so cpu intense, good luck on running another cpu emulator and the graphics!
04:36:08hcsjhMikeS: performance may be a little better with echo on ipod, but still not full speed on Dai Koukai Jidai 2
04:38:33jhMikeShcs: I've had no trouble even running it on x5 which is about the slowest coldfire target due to the lcd slowness. Runs about 4% more boost. Coldfire will run the c echo but at 10-15% more boost than emac.
04:40:06jhMikeSpreglow told me arm has multiply accumulate instuctions (not with saturation though). not quite as nice a set as cf but could probably help a lot.
04:40:26hcsgood, let's pester him about it
04:40:58jhMikeShehe...I pestered a little and he said he was gonna do up the resampling in the dsp core. I don't know when he'll get around to.
04:42:48 Join Mouser_X [0] (
04:45:43jhMikeShcs: how much time is spent in the cpu emulation as opposed to dsp?
04:46:26hcsjhMikeS: not sure, haven't reprofiled in a while. definitely better performance now, though, some tracks in Secret of Mana that skipped before run fine now
04:51:21weeti did the file conversion
04:51:24weetfor my songs
04:51:26weeti dont understand
04:51:30weetdoesnt seem ike it worked
04:51:47 Join gandhijee [0] (
04:52:14gandhijeehey, does Rockbox work with the car adapters and such for the iPods?
04:52:37jhMikeShcs: one thing I'm wondering about is reversing things and doing the full run of samples for a single voice at a time...things would be more coherent that way. Any reason that shouldn't work?
04:53:06hcsgandhijee: I have experience with it working with the iTrip and Bose sound dock
04:53:36gandhijeeguess i'll give it a shot with mine
04:53:45gandhijeeand see if it works with my car stero
04:54:00hcsjhMikeS: well, you need the whole prior sample for echo
04:54:53gandhijeeis there firewire support yet?
04:54:53hcsbut i guess you could do all the samples unprocessed and then run another echo pass over it
04:55:10hcsgandhijee: one thing to note is that anything with external controls won't actually control the ipod
04:55:50gandhijeewell thats no good =/
04:56:45weetI DONT Get it!
04:57:26hcsthat doesn't work
04:57:32 Join DavGerm4 [0] (
04:57:46jhMikeSnot all voices are included in the echo its mix down is only the voices it's assigned to, correct? will think...maybe a limited number of samples per pass until the full output size is generated. "\
04:58:01DavGerm4I have a question about Rockbox
04:58:26weet<hcs> that doesn't work
04:58:26weet<hcs> anymore
04:58:28weetwhy ? :(
04:58:39hcsdatabase has changed a lot since I wrote that
04:58:58DavGerm4If I install Rockbox on my ipod Video, will I be able to still update my ipod with music and stuff
04:59:23 Quit gandhijee ("Leaving")
05:01:07weetI wanted to convert my songs
05:03:18weeti get a " database not ready " error
05:03:37SoapDavGerm4: yes
05:03:44jhMikeShcs: I did improve the c resampling in the core doing the same thing. It was quite a bit faster for me. Any improvements for you? I'm guessing yes, but it's a guess.
05:03:45DavGerm4but the thing is this
05:04:20hcsjhMikeS: I'm not sure precisely what is responsible but I can play more tracks with echo than I could before
05:04:41DavGerm4I want to do the firmware thing on my PC, and all of my songs and crap are on my Dad's MAC. If I do the firmware thing on my PC, will I still be able to update it on the MAC?
05:05:20jhMikeShcs: then I don't doubt that c optimization had a good positive effect there.
05:06:01SoapDavGerm4: I _think_ as long as your don't let your dad's copy of iTunes "update" or "restore" your iPod all will be well.
05:06:23weetHOLY SHIT
05:06:25weetthe tool worked
05:06:25jhMikeShcs: was also wondering about running gaussian and using the brr cache. It looks like that could be fast for cf but I'm not sure what the result will be.
05:06:32weetHcs : it still works!
05:06:32Soapbut if your iPod is alread a MacPod (HFS+ formatted), you will need to convert to FAT32 before Rockbox can work.
05:06:40weetits just not working with album photos
05:06:58Soapit never did.
05:07:05DavGerm4@Soap, are you talking to me?
05:07:13hcsweet: hmm, I'm sort of surprised it worked at all...
05:07:18SoapDavGerm4: regarding MacPods, yes.
05:07:46weetyeah it worked
05:07:48weetwhat i had to do
05:07:53DavGerm4So how do I convert it to FAT32
05:07:58weetis go to Database
05:08:06weetand i placed " update Database
05:08:08SoapDavGerm4: that is covered in the wiki, if not the manual.
05:09:09DavGerm4So if I convert it to FAT32, will the MAC still be able to update it?
05:09:36hcsweet: that has nothing to do with the tool, then, it just reread the metadata from the files (if I understand you correctly)
05:10:50SoapDavGerm4: as for if a Mac can sync files (you do NOT want the Mac to "update" it) I do not know.
05:11:16weetits not so great i admit do ;)
05:11:22weetvery buggy wiht the interace
05:11:46SoapDavGerm4: you don't need to use iTunes with Rockbox, and you can not play music from the iTunes Store through Rockbox.
05:11:48DavGerm4No, I want the MAc to update it, because all of my movies and songs are on there
05:11:57SoapDavGerm4: wrong word.
05:11:57DavGerm4my PC doesn't have enough Memory to store all of it
05:12:13SoapUpdate = upgrade firmware = will get converted to HFS+
05:12:26SoapSync = add music, movies, photos.
05:13:13SoapRockbox can not play Apple movies.
05:13:18SoapYou can dual boot, though.
05:13:41DavGerm4My movies were converted from DVD to MP4 format
05:13:42weetdual boot the ipod ?
05:13:51SoapDavGerm4: rockbox can not play those.
05:13:53Soapweet: yes
05:13:58weetWOAW how do you do that?
05:14:01weeti wanna do that
05:14:03Soapweet: tripple boot if you so desire.
05:14:30weetexplain plz!
05:14:52SoapI believe it is covered in the manual, if not we'll add it.
05:15:09weetlol k
05:15:15DavGerm4So I can't play my movies with RockBox? Is there any way for them to work?
05:16:40***Saving seen data "./dancer.seen"
05:16:42SoapDavGerm4: convert them to MPEG2
05:17:17Soaprockbox does not use the broadcom chip, so all movie decoding is done on the main CPUs, thus much slower performance at this time than apple OS.
05:19:10weetI don't wanna lie do but i'm disapointed wiht the glitches
05:19:19weetI cant wait for a stable version on ipod do!
05:20:28 Quit Vulcan ("Leaving")
05:20:47weetlike in the themes the text often isnt centralised like it should be
05:20:48 Quit Mouser_X (Read error: 110 (Connection timed out))
05:20:58weetor sumthing will be over sumthing that shoudlnt be
05:20:58Soapyou skipped part of the install instructions
05:21:03Soapyou didn't install the fonts.
05:21:22weetor ?
05:21:32Soapdid you install the fonts?
05:21:52SoapI'll take that long pause as a no?
05:22:24weetyes i did
05:22:43Soapahh - my bad ;)
05:23:00weetelse i wouldnt be complaining :P
05:23:06weetabout glitches
05:23:15SoapOR - you are using a WPS which needs patches you don't have applied to your build. - Like album art or scrolling margins.
05:24:31weetyes that!
05:24:31weetand idk how to patch lol
05:25:06Soaplots of wiki tutorials on patching, and a whole forum section on it.
05:25:22weeti dont know how to use C++
05:25:23weet :/
05:25:58Soapand "idk" is unappreciated here. Many people here do not speak english as their primary language, and non-technical acryonyms are not appreciated.
05:26:15SoapRockbox is C and Assembler.
05:26:46SoapAnd you don't need to know /why/ you are doing certain steps to patch. You just need to be able to follow the instructions.
05:27:12SoapRockbox has a most excellent set of patching instrucitons. The best I've ever seen on a project like this.
05:27:57weetwhy wont the devs just patch it for the user ?
05:28:16weetwhy wont the developers*
05:31:14hcsit's enough work supporting what's in SVN, plus there are too many possible combinations of patches to build them all (and rest assured someone will request the oddest combination imaginable)
05:31:43jhMikeSweet: would you rather we develop or spend all our time applying patches ;)
05:32:17bonbonthejonwould it be possible to have an option to turn off cpu throttling?
05:32:27weetturns out theirs a patchning software..
05:32:51weetsommeone mentioned above that we had to compile the patches
05:33:20bonbonthejonweet: yes, you apply patches, then need to compile the code
05:33:38jhMikeSshould be in the debug menu if you have frequency scaling. Just set the boost count to 1 manually.
05:34:13bonbonthejonjhMikeS: can I disable scaling? I have an ipod and I've heard that clears up some problems
05:35:53jhMikeSit's not an option that's persisted but it can be turned off
05:36:00weeti installed the GNU patchning software
05:36:15weetall there is in the directory is a .pdf file
05:36:20*jhMikeS thought that wasn't enabled for iPods anyway. hrm
05:37:18 Quit DavGerm4 ()
05:37:28bonbonthejonjhMikeS: I watch the audio audio thread debug item, the frequency switches between 30 and 75 Mhz
05:38:07midgeyyou have to recompile with scaling disabled
05:38:56bonbonthejonmidgey: i figured, but I was wondering if there was a way that maybe from the debug menu you could set it disabled
05:41:04 Join perldiver [0] (
05:45:50 Quit pearldiver (Read error: 60 (Operation timed out))
05:56:12 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
05:56:50weeti checked out
05:57:09 Join combrains [0] (
05:59:39JdGordondoes RoLo not work on any targets?
06:01:30 Quit combrains (Client Quit)
06:01:42scorcheit works on archos
06:01:46scorcheand irivers i believe
06:02:03midgeyworks fine on H300
06:02:09midgeyjust tested
06:07:55scorcheit technically works on iPods, but it will only RoLo the current running image
06:09:32 Quit xNibbler (Read error: 60 (Operation timed out))
06:10:53JdGordonhmm. yea i thought i remembered it not working properly on at least one target
06:12:57 Quit Sikkan ("( :: NoNameScript 4.02 :: )")
06:17:24 Join xNibbler [0] (
06:28:14 Quit rotator ()
06:33:09 Join Llorean [0] (n=Llorean@rockbox/administrator/Llorean)
06:34:53 Join Mouser_X [0] (
06:35:11LloreanJdGordon: The Gigabeat port _really_ doesn't need optimization. I suspect that user just didn't have Fade on Stop/Pause disabled.
06:35:27JdGordonoh ok.
06:35:48JdGordonjust coz its cpu is redicoulsly powerfull doesnt mean it doesnt need optimization tho
06:36:01LloreanYeah, but for responsiveness and stuff.
06:36:25LloreanI meant it doesn't need it in "there's no user-noticeable lack of responsiveness, as far as I can tell"
06:36:37LloreanClearly having everything as optimized as possible is still idea. :)
06:44:02 Quit Mouser_X (Nick collision from services.)
06:44:11 Join Mouser_X [0] (
06:58:54 Quit XavierGr (Read error: 110 (Connection timed out))
06:59:04 Quit jaebird (Read error: 60 (Operation timed out))
07:01:26 Join pearldiver [0] (
07:01:44amiconnscorche: Afaik rolo on ipods has been fixed
07:02:27 Quit perldiver (Read error: 110 (Connection timed out))
07:03:31Lloreanamiconn: I thought it was limited still. Like, same version of Rockbox can be rolo'd, but that's it, or something
07:04:00 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
07:04:18 Join jaebird [0] (
07:04:23 Join miepchen^schlaf [0] (
07:04:27LloreanI know I can rolo the IPL binary, but not the apple_os yet.
07:04:54WinchesterWhat's the point of rolo if thee is Loader 2?
07:05:36LloreanWinchester: A) I like Rockbox to boot as quickly as possible, B) It allows easy testing of alternate binaries, C) It allows Apple_OS to be loaded from disk without having it in the firmware partition so NO bootloader is needed.
07:05:51LloreanWhich means on my nano Rockbox can boot ridiculously quickly once ROLO allows booting of the apple OS.
07:06:19 Join ArrowStomper [0] (
07:06:33ArrowStomper# ircettiquette
07:06:55Winchesterwell without C..... since Loader 2 an load the apple firmware image
07:07:22LloreanWinchester: You missed the whole point of C.
07:07:45LloreanThe point of C was that Rockbox can be loaded from the firmware partition, without having to go through a bootloader.
07:07:54LloreanCutting boot time a few more seconds.
07:08:00LloreanMaking it nearly instant on the Nano.
07:08:20WinchesterHow is Nano diferent?
07:08:26LloreanNo disk spin up time
07:08:37LloreanOn a disk based player, it'll be 'nearly instant after the disk spins up' instead
07:08:48ArrowStomperis rockbox better than the original nano firmware?
07:09:04LloreanArrowStomper: other than battery life, yes. And since I don't use it for more than 4 hours at a time, it's fine.
07:09:24ArrowStomperI'm trying to tell my friends why it is better
07:09:30WinchesterOk... guess I was under the impression RoLo was just for loading the image files through the file browser >.>
07:09:36ArrowStompercan you play games?
07:09:41LloreanWinchester: Besides, I don't want to use loader 2 anyway. I'm sorry but I don't want to boot into an unnecessary menu every time, when only about .05% of the time I'm not going to be loading Rockbox.
07:10:00LloreanWinchester: It's for loading _any_ image file. IPL's bin, Apple_OS.ipod if you've extracted it, etc.
07:10:23JdGordonfreeking woot :D I got resumin in the browsers working correclty again with the root patch
07:10:36midkayJdGordon: bravo :D what's left to fix/add?
07:11:03Winchesterok guess I missed the point of rolo..... still don't get it... but.. ok.
07:11:04JdGordonthe stuff on the wiki.... keymaps mostly iirc
07:11:31LloreanWinchester: The point is that you can launch any firmware binary from within Rockbox without having to shut down and boot back up.
07:12:00Winchesteroh..... so when that binary closes.... will it return straight to rockbox?
07:12:19LloreanYou're in that other firmware.
07:12:28LloreanSo, with the Apple_OS you'd have to do the normal reboot.
07:12:39Winchesterthen.... I don't see how that's different from Loader 2.....
07:12:43LloreanWith linux, you'd have to shut down unless they implement an in-linux way of launching another binary.
07:12:57LloreanWinchester: Because Loader 2 slows your boot time EVERY time you boot.
07:12:59LloreanEvery single one.
07:13:49Winchesterhmm... having the timeout on 1 second guess I don't really notice it
07:14:08amiconnjhMikeS: The cause of the flashing background is known, and LinusN wanted to fix it...
07:14:22amiconnIt doesn't only happen on X5, but also (at least) on H300
07:14:28LloreanWinchester: It still takes longer to boot than the Rockbox bootloader, and the Rockbox bootloader takes longer than having no bootloader at all, and just using RoLo to launch other firmwares.
07:14:52Winchesterwould rolo be able to take parameters?
07:15:07LloreanWhat do you mean, parameters?
07:15:33Winchesterlike linux.bin /mnt/iboy
07:15:33LloreanYou click on a .ipod (or .iriver, or .whatever) file, and it launches.
07:16:20LloreanRoLo isn't really designed around the idea of managing several firmwares.
07:16:31LloreanEspecially ones that require arguments.
07:16:41***Saving seen data "./dancer.seen"
07:17:05Winchesterok then... Loader 2 just suits me better that way since I use iboy once in a while.
07:17:08pearldivermaybe im just tired, but i cant find the option to add a shuffled folder to the end of playlist
07:18:14Lloreanpearldiver: There's not really one. There's Insert Last which isn't shuffled, and Insert Shuffled, which isn't last.
07:18:39pearldiverthats what i figured
07:19:11pearldiverit seemed like a useful option to have
07:19:36 Part Winchester
07:19:54Lloreanpearldiver: I can't imagine that I'd ever use it, honestly. Why do you want just that folder shuffled among itself?
07:20:09pearldiverwell, an example
07:20:16pearldiveryou're listening to an album
07:20:27pearldiverthe one you love and always listen from start to end
07:20:53pearldiveryou're in the middle of it and want to create a playlist for the next few hours
07:21:31pearldiveryou got this really big folder of your favourite artist you'd love to come up next
07:22:26LloreanHonestly, I wouldn't end up in that situation I don't think.
07:22:42LloreanI would've created the shuffled playlist first, than inserted the sequential album.
07:22:52LloreanI can see how it'd be useful, to a limited extent, though.
07:23:46pearldiveryeah it seems the workaround here is to create preshuffled playlists
07:27:49*amiconn doesn't like the last commit
07:29:51JdGordonwhy not?
07:30:10Lloreanamiconn: It does make it fit the other lines on the screen in style
07:31:23amiconnOn the player, this line has to scroll even with no prefix at all. And on the bitmapped archoses, the prefix makes it scroll
07:31:47amiconn...with default font
07:32:19LloreanPerhaps it could be removed for those?
07:32:24amiconnI think that it's ok on bitmap, but the player should stay with no prefix
07:32:40JdGordonhe didnt change the line on the archo's
07:32:49amiconnHmm, and it's a case where the lang feature patch would really simplify things
07:32:57amiconnJdGordon: He did.
07:33:17JdGordonplayer,recorder,fmrecorder,recorderv2,ondiosp,ondiofm,gmini120,gminisp: "%d%% %dh %dm"
07:33:40JdGordonfrom *: "%d%% %dh %dm"
07:34:09amiconnErm, somehow it looked to me like there were only 2 lines, wrapping :\
07:34:23 Quit ArrowStomper ("Leaving")
07:34:29JdGordonhehe, yeah, i tripple checked that before pasting
07:34:32amiconnWe need the lang features patch
07:34:44JdGordonbug in viewVC then :)
07:36:01amiconnBagder: I think those bad-gcc build servers shouldn't be used for sim builds until fixed
07:36:34amiconnBecause some amount of yellow is expected from them, the 64 bit warnings obviously got undetected
07:36:50 Quit Mouser_X (Nick collision from services.)
07:37:00JdGordonshouldnt the ownsers be nagged to update gcc?
07:37:01 Join Mouser_X [0] (
07:37:50amiconnyes, that too
07:38:25amiconnJdGordon: Btw, the 64 bit warnings are yours (eq_menu.c)
07:40:09JdGordonoh.. didnt realise I added them
07:40:28JdGordonis that the same problem as in settings_menu.c or whever it was last time?
07:40:36amiconnThat's what I mean with the expected jumpy yellow hiding actual warnings
07:40:52 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
07:41:12 Join miepchen^schlaf [0] (
07:42:29JdGordonwhat was the int type to use again in the cast?
07:42:44amiconnIt's one of the usual 64 bit problems. Casting between a pointer and int directly (either direction) gives a warning, because unlike on 32 bit archs these types are no longer the same size
07:43:03amiconnYou either need to double cast, or use some implicit trickery to do the rest
07:44:22amiconnThe latter is possible here, I just tried it and will commit. Casting to intptr_t first (which actually is a long on amd64), and let the assignment to the int do the rest
07:44:31 Part toffe_
07:44:54JdGordonok, ta
07:55:08amiconnHmm. Why do the archos binaries become larger if the string isn't changed for them?
07:56:50JdGordonisnt the size of the archos all the compressed bin? so the compression could be different slightly...
07:57:30amiconnOnly recorder v1 is compressed atm
07:57:51amiconnI guess it's the order. *: needs to come first iiuc
07:58:41JdGordonI thought *: had to come last?
07:59:14JdGordonalthough, going first does make sense
07:59:16amiconnOther places in english.lang have *: first
07:59:42amiconnThe gmini can be ignored, but the ifp should use the no-prefix string
08:04:55daurnimatori moved isps
08:05:02JdGordonwho you with now?
08:05:16daurnimatorgot that job too
08:05:22JdGordonamiconn: you said ccache in the vm's slows it down yeah?
08:05:25JdGordonnice :)
08:05:37amiconnNo, on cygwin it slows things down
08:05:40JdGordonare they looking for more ppl?
08:05:56daurnimatordoubt it
08:06:01JdGordonvmware and colinux no change or speed up?
08:06:06JdGordonor oyu dont know?
08:06:11daurnimatorits pretty trampy at the office
08:06:14daurnimatorbut, i work from home
08:06:47amiconnNo idea, didn't try ccache on vmware, and didn't try colinux at all
08:07:02JdGordonSoap: you around?
08:08:11JdGordonguess not
08:08:55JdGordonmidkay: around?
08:09:26JdGordonsome guy I know is after someone to do icons for his app... interested?
08:09:51midkaysure. :)
08:10:41JdGordonok, you might get a pm then :)
08:12:00midkayindeed :)
08:12:56 Quit Mouser_X (Nick collision from services.)
08:13:05 Join Mouser_X [0] (
08:16:28 Join ender` [0] (
08:16:34 Join XavierGr [0] (
08:17:40 Quit Mouser_X (Nick collision from services.)
08:17:48 Join Mouser_X [0] (
08:21:30 Join cynicalliberal [0] (
08:22:07 Quit cynicalliberal (Client Quit)
08:26:19JdGordonhas anyone got distcc working with the crosscompilers?
08:27:03 Join decayedcell [0] (
08:28:20 Part decayedcell
08:30:51 Quit miepchen^schlaf (Read error: 113 (No route to host))
08:31:06JdGordonhaha using distcc makes the compile log funny
08:31:16JdGordonactually, I tihnk its the -j8 part that makes it funny...
08:34:55 Quit printfXh4 (Read error: 148 (No route to host))
08:35:02 Join printfXh4 [0] (
08:36:21 Join Soap_ [0] (n=Soap@rockbox/staff/soap)
08:37:32 Quit Soap (Read error: 104 (Connection reset by peer))
08:41:58 Quit JdGordon ("Leaving.")
08:42:01 Quit Mouser_X (Read error: 110 (Connection timed out))
08:43:00 Join safetydan [0] (n=dan@rockbox/developer/safetydan)
08:49:02 Join Rob222241 [0] (
08:55:41 Join miepchen^schlaf [0] (
09:02:32 Nick lex_ is now known as lex (
09:05:21 Join kaaloo [0] (
09:06:24 Join himitsu [0] (n=himitsu@
09:06:32 Quit Rob2222 (Read error: 110 (Connection timed out))
09:08:24 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:08:51 Quit GodEater (Client Quit)
09:09:05 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:09:27 Quit GodEater (Client Quit)
09:09:49 Quit safetydan (Remote closed the connection)
09:09:58 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
09:09:59 Join petur [0] (i=d4efd6a6@gateway/web/cgi-irc/
09:10:34 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
09:13:19 Join PaulJam [0] (
09:16:38 Join tick [0] (i=c27f0812@gateway/web/cgi-irc/
09:16:43***Saving seen data "./dancer.seen"
09:18:05 Join JdGordon [0] (i=dced3920@gateway/web/cgi-irc/
09:18:53tickIn RB, there are some functions of type get_xxx() and set_xxx(newVal). Sometimes, the old value is temporarily changed and then restored. This is done with get_xxx() with immediately following set_xxx(). Wouldn't we save a couple of bytes if we'd let the set_xxx() to return the previous value?
09:25:42 Join B4gder [0] (
09:27:20tickLinusN: the last commit of spc.c. Ok, it's a codec but I think there are also places in the core code.
09:27:55LinusNplease find an example for me
09:27:58 Join linuxstb__ [0] (
09:30:03 Quit linuxstb_ (Read error: 110 (Connection timed out))
09:32:31tickLinusN: do you mean an example from the core? Since spc.c is such an example (but not from the core).
09:33:33LinusNi am mainly interested in saving code space in the core, the codecs are not important
09:35:05 Join rp- [0] (i=rp@
09:39:07B4gderbloody spammers
09:39:23B4gder4 recent users regged as LanXing*
09:39:39*B4gder runs for coffee
09:40:21peturmaybe they're all family?
09:42:01GodEaterCAPTCHA time ?
09:42:09midkayshanghai cleaning service. haha.
09:42:24midkaynice place to advertise..
09:44:24peturwhat's the menu button on h1x0?
09:44:42PaulJamJdGordon: are you here?
09:45:13GodEaterpetur: A-B
09:48:59 Quit pearldiver (Read error: 110 (Connection timed out))
09:52:12 Quit GodEater ("CGI:IRC")
10:01:00 Join bluebrother [0] (i=TAsGWj3f@rockbox/staff/bluebrother)
10:01:20tickLinusN: dsp.c(843), (possibly) debug_menu.c(773), plugin.c(614), gui\scrollbar.c(135). More to come (maybe)
10:02:15 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
10:02:29tickLinusN: gui\splash.c(139)
10:03:23 Quit Presence (
10:03:23 Quit gtkspert (
10:03:23 Quit NuclearDog (
10:03:23 Quit secleinteer (
10:03:23 Quit spug (
10:03:23 Quit Overand (
10:03:30NJoinsecleinteer [0] (n=scl@
10:03:30NJoingtkspert [0] (
10:03:30NJoinspug [0] (
10:03:30NJoinNuclearDog [0] (
10:03:30NJoinOverand [0] (i=overand@pdpc/supporter/active/Overand)
10:03:30NJoinPresence [0] (
10:05:33tickLinusN: that's it for now. Not much actually...
10:06:47LinusNtick: but every byte counts...
10:11:06LinusNlcd_getmargins would have to return two values though, so it would probably not be worth it
10:12:13*petur always uses poiters for returning values, the return value = success/fail
10:12:21peturpointers even
10:12:40B4gderyeah, that's a good practise imho
10:14:50peturdoes original C also ro references or was that added later? I'm always confused...
10:15:16LinusNpointers would certainly work, but i doubt that it would save us the code space we wanted
10:20:18LinusNi wonder which svn version JdGordon based his last rootmenu patch on
10:20:37amiconnI also doubt that pointer would give the desired effect here, because of the extra function parameters
10:21:10amiconnExtra function parameters can increase code size quite much - see Slasheri's endian conversion stuff
10:21:26amiconnOne extra parameter - +700 bytes in the calling code on archos
10:22:12LinusNholy cow
10:22:19peturwow, didn't know that
10:22:35 Join Mouser_X [0] (
10:22:45amiconnYes, the endian conversion code itself is only ~400 bytes
10:23:40amiconn...but ecwrite() and ecread() have one more parameter than write() and read()
10:24:38linuxstb__Llorean, amiconn: The issue with rolo on the ipod is that it only restarts the main CPU, not the COP. So it will load IPL (as long as you don't use any IPL applications that use the COP) and the SVN Rockbox, but not the Apple firmware. The "only restarting the same Rockbox" problem was a caching problem which dan_a fixed.
10:25:04Lloreanlinuxstb__: But, it won't work with COP builds then, I assume?
10:25:21LloreanThat would explain why it didn't work when I tested it. :)
10:25:25LloreanI thought I'd just done something screwy.
10:26:49amiconnlinuxstb__: Hmm, then we should make it restart the cop as well
10:27:20amiconnOn a related note, rolo is still a hack. It outputs messages from firmware code, and those messages aren't localised
10:30:16linuxstb__amiconn: Of course. I think it's somewhere on both dan_a's and my todo lists...
10:31:43linuxstb__Could we just move it entirely into apps/ ?
10:32:18linuxstb__Or maybe have a small function in the target-tree code that does the actual execution, after the firmware is loaded from disk.
10:34:56printfXh4Why, that's a very good question, linuxstb__.
10:36:26bluebrotherLinusN: seems to be r12394 from the patch info −− at least it applied cleanly to that revision for me
10:36:48*LinusN is blind
10:37:09*bluebrother just downgraded to that release to test
10:37:56amiconnlinuxstb__: The copy & start code is target specific, but the high level stuff should be in apps
10:38:24LinusN12391 it seems
10:39:17bluebrother12394 for keymap-h10.c
10:39:34 Quit Presence (
10:39:34 Quit secleinteer (
10:39:34 Quit NuclearDog (
10:39:34 Quit gtkspert (
10:39:34 Quit spug (
10:39:34 Quit Overand (
10:40:12LinusNit's unfortunate that he so often forgets to do "svn up" before he creates the patches
10:40:19bluebrotherindeed >*
10:40:40bluebrotherthat release is 3 days old
10:41:51NJoinsecleinteer [0] (n=scl@
10:41:51NJoingtkspert [0] (
10:41:51NJoinspug [0] (
10:41:51NJoinNuclearDog [0] (
10:41:51NJoinOverand [0] (i=overand@pdpc/supporter/active/Overand)
10:41:51NJoinPresence [0] (
10:42:09amiconnLinusN, petur: Perhaps the increase from the extra parameter is so big because it pushes the number of parameters above the regparm limit for SH
10:42:36LinusNsounds likely
10:43:14 Join desowin [0] (
10:45:15 Join BigBambi [0] (
10:47:02 Join Nico_P [0] (
11:01:44 Quit Mouser_X (Nick collision from services.)
11:01:52 Join Mouser_X [0] (
11:03:57 Join lee-qid [0] (
11:12:02 Join sp3iel [0] (
11:13:42 Join inversions [0] (
11:16:04 Quit esp8 ()
11:16:46***Saving seen data "./dancer.seen"
11:17:45 Quit GodEater ("CGI:IRC (Ping timeout)")
11:20:01 Join BHSPitMonkey_ [0] (
11:25:56 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
11:26:16 Join miepchen^schlaf [0] (
11:27:22 Quit BHSPitMonkey (Read error: 145 (Connection timed out))
11:31:23*preglow longs for summer
11:31:50Mikachuwe had snow today :(
11:32:03 Quit Mouser_X (Nick collision from services.)
11:32:04B4gder-15C here this morning. It bites
11:32:05LinusNi love snow
11:32:12 Join Mouser_X [0] (
11:32:44preglowsnows here now
11:33:08preglowwhich i'm ok with, but i seem to have developed a fondness for freezing
11:33:37Mikachuyeah, snow is fine, but i wish it could snow at +15°C :)
11:34:10preglowi usually like both snow and cold, but the cold doesn't sit too well with me right now
11:34:21preglowgetting old, i suppose
11:34:38preglowat least waking up to the nano was a success story i'll repeat
11:35:32peturhow much work to get the h3x0 to do that too?
11:35:38preglownot much at all
11:35:44peturdouble yay
11:35:51preglowi THINK, btw
11:36:16preglowbut it uses the pcf too, so i don't see why it should be too different
11:36:31LinusNpetur, preglow: the bootloader is not rtc-wakeup aware
11:36:33preglowbut right now the driver for that pcf and the ipod pcf are split, so they need rejoining in some way
11:36:38preglowLinusN: ahh, right
11:36:40preglowthere was that
11:36:56preglowLinusN: but does it wake up at all?
11:37:12preglowwell, then it should work
11:37:12LinusNbut the bootloader turns it off again
11:37:16preglowahaha, right
11:37:25preglowthen it won't work :)
11:37:26peturchecks the on button, right
11:37:39preglowthe ipod bootloader also seems to erase the interrupt status
11:37:47preglowso i use another method to detect if we got an alarm
11:39:02preglowprimarily just comparing current time and alarm time, then seeing if the alarm wakeup bit is set
11:39:08preglownot foolproof, but should work most of the time
11:41:51linuxstb__On the ipods, I imagine people might want to leave the hold switch on after setting the alarm (e.g. if they're sleeping on a train/plane with their ipod put away in a pocket), it could be nice to disable the loading of the apple firmware in the Rockbox bootloader and settings-reset-on-hold in Rockbox itself if we're waking from an alarm.
11:41:54 Join anathema [0] (
11:42:01 Nick linuxstb__ is now known as linuxstb (
11:42:24preglowretailos is loaded on hold????
11:42:46Mikachuyeah, someone changed that
11:43:00Mikachuyour settings would be reset otherwise anyway
11:43:20Mikachuah, i did my old "only read the bottom-most line and reply" trick
11:45:16preglowand some trick it is
11:45:51*preglow throws Mikachu a biscuit
11:45:54Shaidthe setting reset with hold on got me for a while
11:46:05Shaidcause I always keep my ipod on hold when its off, so it doesnt get turned on accidently.
11:46:13Shaidand I wondered why my settings kept getting reset. :P
11:46:19*petur get hungry and looks at the time
11:46:57preglowpetur: you can't eat the clock
11:47:01preglowgives you digestive problems
11:47:28Mikachuit'll give you ticks
11:54:32bluebrothergrrr. Why can't I interrupt a search in Altium Designer?
11:58:41 Quit inversions (Read error: 110 (Connection timed out))
11:58:46 Join x1jmp [0] (
11:59:01*bluebrother joins petur getting hungry and leaves for food
12:06:27 Join Ribs [0] (n=ribs@
12:06:34 Join inversions [0] (
12:11:06preglowLinusN: why disable low latency mode there?
12:11:54luckzdesowin: wonderful app, really
12:12:05 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
12:13:28luckzI'm squirming in delight due to how well it works ;p (of course a "recalc date/time to be after last-submitted entry" would still make it even better)
12:13:44LloreanWell, as if JAVA wasn't bad enough:
12:14:31peturno wonder the gigabeat S needs 400MHz
12:15:14*linuxstb wonders what people are learning in school about computing nowadays...
12:15:19 Nick Soap_ is now known as Soap (n=Soap@rockbox/staff/soap)
12:16:03jhMikeSpreglow: I'm getting lockups on my H120 too but was playing an mp3. It happened when the drive spun down a few seconds in. :\
12:16:32LinusNpreglow: because 1) it is only needed when tweaking the settings, and 2) a flaw in the menu code makes it not disable it again
12:16:59preglowjhMikeS: probably the same bug, then
12:17:00preglowjhMikeS: that's nasty
12:17:08preglowLinusN: ouch
12:17:17preglowLinusN: and it only applies to crossfeed?
12:17:37LinusNother submenus could suffer from the same bug
12:18:22SoapJdGordon: still need something?
12:18:57LinusNthe menu bug is that the callback is called with ACTION_MENU_EXIT only when leaving the root menu in a menu hierarchy
12:19:37LinusNand with ACTION_STD_MENU or ACTION_STD_CANCEL in other cases
12:20:35LinusNi didn't bother fixing that, since the crossfeed submenu isn't supposed to activate the low latency mode anyway
12:20:36 Quit Mouser_X (Nick collision from services.)
12:20:45 Join Mouser_X [0] (
12:24:02 Quit anathema (Read error: 110 (Connection timed out))
12:30:30 Join kubiixaka [0] (
12:30:30 Quit kubiix (Read error: 131 (Connection reset by peer))
12:30:46amiconnLinusN: Hmm, didn't you want to fix the lcd glitch on startup (h300, x5, possibly other colour targets)?
12:31:05LinusNyes i wanted to, but i want to sync it with slasheri
12:31:28jhMikeSoh, this lcd glitch thing is a common problem? didn't know.
12:32:05jhMikeSwhat's going on there? looked like contrast going wild
12:32:23LinusNthe lcd is updated with garbage
12:32:56jhMikeSok, then that's not the one I thought. haven't gotten that one.
12:33:49LinusNmaybe we should just revert slasheris change and let him deal with it
12:34:31preglowwhat change did he do?
12:35:37JdGordonPaulJam: midkay: im back
12:35:56JdGordonSoap: i was gonna ask about colinux, but dw
12:36:04 Quit lee-qid ("aufwiederbyebientotsayonara")
12:36:10LinusNhe turns on the backlight too early, and the lcd is updated with a framebuffer filled with garbage
12:36:18midkayJdGordon: welcome back, i was gonna offer you a keymap or two, dunno where they go in the patch nor am i familiar with the 'new' button system thing.
12:36:31 Join lee-qid [0] (
12:37:16JdGordonLinusN: whats the flaw int he menu system?
12:37:28jhMikeSwhen did this start? I'm just seeing a background flashing a light color real contrast maxed. :\
12:37:53PaulJamJdGordon: i just wanted to report the issue with the low latency mode in the crossfeed submenu. but it is fixed now.
12:38:13LinusNJdGordon: do_menu() only calls the callback with ACTION_MENU_EXIT if you leave the top menu
12:38:51LinusNso the callback can't rely on ACTION_MENU_EXIT
12:38:53JdGordonOh? I thought I fixed that..
12:39:03JdGordonmaybe I did and forgot to commit it :p
12:39:24Soapoh, I worry! ;)
12:39:24LinusNalso, the crossfeed menu used the callback for no reason
12:39:39JdGordonsoap: hehe :)
12:39:41LinusNno need to enable low latency mode in the submenu
12:39:42 Quit Mouser_X (Read error: 145 (Connection timed out))
12:40:07LinusNJdGordon: could you resync the root menu patch?
12:40:12JdGordonusing the callback was most probably an oversight on my part.. but if the EXIT action isnt being called then that needs fixing
12:40:26JdGordonin a while... in windows atm
12:40:39*JdGordon setting up a distcc farm
12:40:55LinusNthe callback is called, but with the action that triggered the exit, i.e ACTION_STD_CANCEL or ACTION_STD_MENU
12:41:34*amiconn checked an ooold log (20060112) regarding regparms on coldfire
12:41:49amiconnXavierGr: I think I know why the test didn't work at all on H1x0
12:41:57JdGordonhmm... thats odd.. the code looks like it calls it with the correct callback
12:41:59PaulJamlittle question, is there a way to view the version number a little bit longer than in the credits plugin?
12:42:12JdGordonoh, no, your right
12:42:16LinusNJdGordon: line 624
12:42:22amiconnlostlogic adjusted memcpy and memset, but the H1x0 also uses asm for the lcd write, which also needs to be adjusted
12:42:23JdGordonthe else got me confused
12:42:38LinusNJdGordon: a comment or two regarding the menu stack would be nice
12:42:48LinusNexplaining the idea
12:43:15LinusNJdGordon: any reason why get_menu_callback() is called so many times?
12:43:25JdGordonthe order of callbacks in is the wiki... and partial explanation, I'll try to rememer to put it in the code
12:43:51JdGordonbecause I put it in as an afterthouught and havnt optimized it...
12:43:58LinusNi see
12:44:21JdGordonand the actual calling is even possibly called too much
12:45:08JdGordonback in a few min
12:45:08 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
12:45:35amiconnLinusN: Why should crossfeed not enable low latency?
12:46:03LinusNamiconn: it is only needed when you actually change a setting, not in the submenu
12:46:45amiconnAh, yes
12:47:52 Part daxxar
12:49:02 Quit JdGordon ("CGI:IRC")
12:49:12 Join JdGordon [0] (i=dced3920@gateway/web/cgi-irc/
12:49:40PaulJambtw, shouldn't the low latency mode discard the already decoded audio in the PCM buffer? now the actual change of a setting in this mode still takes a few seconds to be audible imediately after entering the low latency mode (especially when crossfade is enabled => larger PCM buffer).
12:50:09JdGordonLinusN: get_menu_callback() is called often because menu changes often
12:50:33JdGordonI think it has to be called once every time through the loop... but ill check out to make sure
12:51:28JdGordondamn! I only have 300mb free on my win partition... back in a min again
12:51:32 Quit JdGordon (Client Quit)
12:52:17preglowPaulJam: discard? why? that would result in a terrible glitch
12:52:19 Join Obsys [0] (
12:53:13LinusNmidkay: the clock plugin draws the hour hand wrong when hour == 12
12:54:20preglowamiconn: doesn't gcc require a patch for regparm on coldfire?
12:54:25midkayoh, is that when? i encountered some kind of bug like that at one point, never saw it again... thanks, i'll look into it (probably gonna need to rewrite the mess of hour variables and their handling)
12:56:33LinusNmidkay: i have committed a fix
12:56:43midkayLinusN: gee, thanks, that was quick :D
12:56:51 Quit linuxstb (Read error: 113 (No route to host))
12:57:20midkayyou also reminded me that there was a bug i never fixed.. in digital mode, or plain mode too probably, at 12:00am it shows as 00:00.. probably related.
12:57:20LinusNmidkay: classic mistake :-)
12:57:23 Join nls [0] (
12:57:47midkayLinusN: haha. confusing code is what it is, i swear!!!! :)
12:58:30B4gderit was about time that was fixed!
12:59:26PaulJampreglow: i meant that the discarded audio should be decoded again, but with the low latency. But i don't really have a clue about this stuff.
13:00:00preglowPaulJam: not exactly an easy trick to pull off
13:00:09 Join JdGordon [0] (n=jonno@rockbox/developer/JdGordon)
13:00:13preglowi'd rather we make the audio engine low latency throughout
13:00:19*JdGordon hates wifi in linux
13:00:31preglowthis can be done by storing the pcm buffer in full precision and only applying dsp late in the chain
13:00:37B4gderwifi works fine in my linux
13:00:48preglowwifi doesn't work fine in any os or setup i've tried
13:01:09preglowit always has me moaning for a cable after a couple of hours
13:01:44JdGordonB4gder: have you had any ideas how to send the zips later so people with slow net can help out? Im getting a distcc farm going here and it may as well be used :p
13:02:23B4gdersending the zips later is not really a possibility
13:02:36B4gderI mean, when builds are done immediately after each other
13:02:40JdGordonwhat about using bittorrent for the builds?
13:02:57B4gderhow would that help/work?
13:03:13preglowbittorrent isn't really suited for that
13:03:30JdGordonsetup a tracker on and add each build as its done?
13:03:36amiconnpreglow: Yes it does. On 2006-01-12, lostlogic played with it
13:03:45preglowamiconn: i'd love to use that
13:04:13preglowi wonder if it works well, though, a lot of the m68k based patches are crap
13:04:14amiconnIt looks like the speedup was minimal at best
13:04:19preglowamiconn: what about size?
13:04:22amiconnCheck the mentioned irc log
13:04:25B4gderJdGordon: we still need to complete the transfer before the next build
13:04:33B4gderso it wouldn't help
13:04:34amiconnSizes weren't mentioned
13:05:03preglowbtw, what is the point with the .size directive in asm functions?
13:05:10amiconnB4gder: Why can't the zips be transferred in the background, in parallel to a running build?
13:05:19B4gderthey can
13:05:23amiconnIt's not that the transfer needs much cpu power
13:06:58B4gderit's just that adding another thread for each zip transfer add complexity
13:07:23JdGordondoes it make any sense to not allow the user to leave a menu?
13:09:37JdGordonLinusN: I was wrong.. the callback code can be simplified a fair bit :p will commit in a shortly
13:10:11 Join barrywardell [0] (
13:12:14JdGordonB4gder: either way, do you want to add my server again for the sims, and remove them faulty ones?
13:13:24amiconnB4gder: One thread for each? Or perhaps one transfer thread that queues the transfers?
13:14:03JdGordonqueing the transfers will keep it slow tho...
13:14:25 Quit tick ("CGI:IRC")
13:15:17JdGordonB4gder: once the build scriopt knows a build is done, could you get it to rcp it in the background? that doesnt add too much complexity does it?
13:16:51***Saving seen data "./dancer.seen"
13:23:09preglow- Dette kan nok forsvare en plass innunder kunstbegrepet. Men vi kan stille spørsmål ved om dette virkelig er kunst og ikke minst om det er kvalitetskunst, sier kunsthistoriker Tommy Sørbø. Han legger til:
13:23:18preglowgnome and pasting :/
13:23:24preglow int volume; /* audio output volume: 0-100 0=off 100=max */
13:23:29preglowthat isn't exactly right anymore is it?
13:23:57linuxstb_Some people would say it's right :)
13:24:27 Join Obsys_ [0] (
13:24:50preglowdon't mention it, please, i think it's _finally_ been forgotten about
13:25:00linuxstb_You brought it up...
13:25:14preglowno i didn't!
13:25:17linuxstb_You need to hide it in a different commit...
13:25:24JdGordonLinusN: and the root menu patch.. I uploaded the wrong file :p Ill resync and upload soon
13:25:48*JdGordon feels stupid :(
13:27:49 Quit Obsys (Read error: 110 (Connection timed out))
13:27:49 Nick Obsys_ is now known as Obsys (
13:29:37 Quit kaaloo (Read error: 113 (No route to host))
13:35:59JdGordon:) clean up a bit of code, drop 50 bytes... nice :)
13:36:06 Join web-taz [0] (
13:36:39JdGordonhow does adding an else decrease codesize?
13:37:14Shaidless confusion for the compiler/linker
13:37:18Shaidor something
13:37:25*Shaid is taking stabs in the dark.
13:39:01*petur waits for 'the dark' to login and stab back
13:39:18peturforehead did....
13:39:46amiconnJdGordon: Queueing the transfers in the background won't be any slower than starting multiple transfers at once if bandwidth is the limiting factor
13:40:12preglowand it's a cleaner approach
13:40:38JdGordonupstream is the problme, not downstream... if you que all transfers then you still have to wait for the slow servers to finish uploading before starting the next one..
13:41:18amiconnIt's the build results what counts for the devs, the transfers then just run a little longer
13:41:41JdGordonyour saying get them all at the end?
13:41:43amiconnYou mean for the next build round?
13:41:50amiconnNo, not at the end
13:42:19JdGordonI'm confused
13:42:31amiconnBut with background transfers, build servers would do builds back-to-back, without putting the transfers in between
13:42:50amiconnSo it would save time *within* the build round
13:43:04*JdGordon thinks we are saying the same thing
13:43:28amiconnThen what do you think would be the problem with queueing?
13:43:34 Join atsea-145 [0] (i=ariel@gateway/tor/x-eceb1d453615cd3b)
13:43:41JdGordonbecuase we still want to get the builds in quickly
13:44:09amiconnAbove all we want to get the results quickly.
13:44:12JdGordonand downstream bandwidth is huge so doing 21 downloads wont suffocate it..
13:44:19JdGordonwe can do both
13:44:50amiconnWhat I am confused about is that you think queuing would somehow change that??
13:44:52 Join relaxed [0] (n=relaxed@unaffiliated/relaxed)
13:45:09JdGordonforeach $build (@builds) { `scp build{server}:build-$target ./ &` }
13:45:37amiconnIt's just that queueing would only need one additional thread per build *server*, as opposed to one thread *per build*
13:45:44amiconn...with no speed disadvantage
13:45:48JdGordonbecause it might take 3 min to upload the zip from my server, and 3 min to upload 5 files of other servers... do thema t the same time and its 3 min... que them and its 6min
13:46:10preglowthat _can't_ be right
13:46:14JdGordonOH.. ok we are confused
13:46:26JdGordonyour saying que the ones coming from the individul severs?
13:46:31JdGordon not all in one que?
13:46:52amiconnA send queue would be at the sending side, can't imagine how that could be different
13:47:00 Quit atsea- (Remote closed the connection)
13:47:12JdGordonok, yeah that would work... I was thinking from the recieving end
13:47:44B4gderthere's already one thread for each server running
13:48:56amiconnI mean on which side?
13:48:58 Join nicktastique [0] (n=nick@unaffiliated/nicktastic)
13:49:01B4gderin the buildmaster
13:49:24amiconnWell, would the build master also need an extra thread per server with backgrtound transfers?
13:49:44amiconnI thought the upload happens using curl or sth like this
13:50:13preglowscp, probably
13:50:15B4gderthe sending of zips from build servers are requested by the master using scp
13:50:24B4gderas ssh is the only access method it has
13:50:46amiconnYes, but couldn't the master just tell the build server to upload the file using curl?
13:50:56preglowbetter to use scp anyway
13:51:01pregloweasier and better authentication
13:51:04LinusNcurl can do scp :-)
13:51:36 Quit nicktastic (Read error: 60 (Operation timed out))
13:51:40amiconnIt just seems to me that initiating the transfer from the build server would be better
13:52:01amiconn...because the build server knows when the build finished, so no polling necessary
13:52:16amiconn...and it could queue the uploads without any extra thread on the master
13:52:21B4gderthe build master knows that too
13:54:27 Join Rondom [0] (
13:54:34preglowthe master server just sshs to the build server and tells it what to do directly, yes?
13:54:55LinusNit logs in and executes a script on the build server
13:55:19preglowshould be fairly trivial to download a build while performing a new one in that case
13:55:41*preglow kicks firmware/
13:55:41LinusNwell, you need to start a new thread
13:55:55JdGordonthreads are cheap...
13:57:04JdGordonanywho.. B4gder, want another server? just need to fix up the account...
13:57:08LinusNJdGordon: yes, but keeping track of them in a perl script is not entirely trivial
13:57:10 Join Mouser_X [0] (
13:57:35pregloweverything is trivial in perl :)
13:57:53Mikachuwho writes scripts in perl anyway? :)
13:58:06JdGordonok, it just clicked why you actually need to keep track of it :p
13:58:08preglowpeople who vomit when seeing python
13:58:12LinusNJdGordon: it is not about processing power, it's about complexity and if it is worth spending time on developing
13:58:54LinusNmostly the latter
13:59:03 Quit barrywardell (Remote closed the connection)
13:59:17LinusNspare time is a scarce resource
14:00:27 Quit midkay ("Leaving")
14:05:33B4gderand the fact that the buildmaster script is more fun when it works and debugging huge number of threads is not cheap
14:06:36preglowit's more fun to solve them problem with more servers anyway :>
14:06:54 Join JdGordo1 [0] (
14:07:18LinusNi'd rather spend my time on developing rockbox
14:07:19preglowjdgordoone enters
14:07:31 Quit JdGordon (Nick collision from services.)
14:07:35 Nick JdGordo1 is now known as JdGordon (
14:08:08B4gdersome people think we'd better spend the time obfuscating email addresses :-P
14:09:30LinusNit's funny how microsoft think sometimes: they create a product (Vista) that nobody cares about and costs too much, then they blame the poor sales on piracy :-)
14:10:26JdGordonwell.. you cant blame the need for the latest computer to run it... everyone upgrades every 18months!
14:10:57preglowdon't much see the need for vista, no
14:11:12JdGordon4gb of ram is perfect for wista aparently..
14:11:18LloreanSo I hear.
14:11:40LloreanI still haven't managed to find out what Vista actually offers me, as a home user, that's worth $150
14:11:52JdGordoneye candy!
14:13:04Shaidwait, Vista isnt selling well?
14:13:17LloreanWell, Rockbox isn't selling well either.
14:13:19ShaidYou mean you cant just put shit in a fancy box and expect gullible people to buy it?
14:13:28LloreanClearly we too have a piracy problem.
14:13:40ShaidMaybe we need a fancy box to put it in!
14:14:07LinusNyeah, we put it in a box and call it... Rockbox! naah...
14:14:08JdGordonare there any forks of rockbox? (other than the custom builds on the forums and gliters) ?
14:14:18LinusNJdGordon: yes, one
14:14:22LinusNJdGordon: openneo
14:14:30B4gderfrom many years ago
14:14:39JdGordondoes that still resemble rockbox at all?
14:14:46LinusNnot much
14:15:39LinusNthey have rewritten major parts of it, partly because they wanted it to work differently, and partly because they want it to look like they wrote it
14:16:06 Join FOAD_ [0] (
14:16:28LinusNtheir first fork wasn't a fork at all, but plain theft
14:16:41LinusNthen we reminded them about the GPL and the law
14:17:01JdGordon:) damn that part... always getting in the way of innovation!
14:17:18B4gderit still isn't quite GPL clean afair
14:17:26B4gderwith their voice stuff etc
14:18:06Llorean"The Open Neo Project would like to thank the Rockbox Project for their help."
14:18:06LinusNwell, it's that grey area of what is linking and what is not
14:18:25LinusNLlorean: yes, now it says that
14:18:34LloreanI like the wording.
14:20:21 Quit Mouser_X (Read error: 110 (Connection timed out))
14:20:33PaulJamanother question about the low latency mode: is there a reason for the behavour in this mode that the codec buffer doesn't get filled completely once it is empty resulting in the harddrive constantly spinning? or is this a bug?
14:20:43JdGordonhmm... they have customizable keys... I wonder how difficult that would be to add with the action system
14:21:30JdGordonPaulJam: my guess is it does it on purpose to rebuffer more often so changes happen quickly?
14:22:29LinusNPaulJam: that should only apply to the pcm buffer
14:22:32PaulJami meant the compressed audio buffer, i don't think the compressed audio is modified by dsp stuff
14:22:48LloreanLinusN: I have noticed the PCM buffer struggling to fill while in low latency mode.
14:22:56LloreanLinusN: Err, compressed, not PCM.
14:23:04LinusNsounds like a bug to me
14:25:47LloreanIt was stuck in low latency mode though, so it could've just been part of whatever bug that caused that state.
14:25:50preglowalmost sounds like a priority bug
14:26:01preglowSlasheri: weekly RockboxKernel update reminder time
14:26:36 Join pearldiver [0] (
14:28:58JdGordoncan anyone with a archos player please tell me if the jumpscroll setting works?
14:30:11preglowreally, why do we have a math.h ?
14:31:28Mikachuit's a bit strange that it has #ifdef __cplusplus though
14:32:19 Quit FOAD (Read error: 110 (Connection timed out))
14:33:31preglowLear: for the logs, get_replaygain_int doesn't have an error of more than 0.2 dB at -48 for argument
14:33:40preglowwhich isn't too shabby, really
14:33:48peturMikachu: because we change ported code as little as possible?
14:37:27 Join midkay [0] (n=midkay@rockbox/developer/midkay)
14:39:18JdGordondoes distcc need the binutils/libc6/others? version on all servers to be the same? or just gcc?
14:41:19preglowjust gcc, afaik
14:41:28preglowthe distcc servers get the preparsed code
14:41:31Mikachuyes, it only compiles, no linking
14:41:34pregloweh, preprocessed
14:42:17preglowJdGordon: i don't know how well distcc will work for rockbox, though, i think it's primarily a nice thing for bigger files
14:42:43preglowLear: also for the logs, it also seems to handle gains as far as up to 48 dB with no bigger errors than 0.01 dB
14:43:09JdGordonI was getting 25s make bin time without it (with ccache), with both its 17s.. so tiny diff
14:43:24JdGordonthis is over 100mbit LAN so network doesnt cause a bottleneck
14:43:33Mikachuit makes no sense to benchmark distcc with ccache active
14:43:54Mikachuwith no changed code
14:44:28JdGordonok, ill run one without ccache
14:44:52 Join wLLm [0] (
14:46:00 Quit w1ll14m (Read error: 104 (Connection reset by peer))
14:46:01JdGordonhmm.... sfa diff with or without distcc
14:46:17 Join webguest72 [0] (i=839b6d84@gateway/web/cgi-irc/
14:46:21preglowJdGordon: like i said, distcc works the very best on big .c files
14:46:41 Join Obsys_ [0] (
14:46:49webguest72Hello, can I be granted access as a twiki-user?
14:46:50JdGordonim adding 2 more comps tomorow so ill see for sure then..
14:46:56webguest72I is PaulClaassen
14:48:51 Join pilot000 [0] (i=c31ce03b@gateway/web/cgi-irc/
14:48:58 Quit pilot000 (Client Quit)
14:49:08 Join pilot000 [0] (i=c31ce03b@gateway/web/cgi-irc/
14:49:25 Quit pilot000 (Client Quit)
14:50:26 Join pilot000 [0] (i=c31ce021@gateway/web/cgi-irc/
14:51:00*preglow leaves that to someone who knows how
14:51:36pilot000JdGordon are you here ?
14:51:37bluebrotherwebguest72: done
14:51:55bluebrotherpreglow: you just need to add the wiki name to TWikiUsersGroup
14:52:14JdGordonpilot000: yeah
14:52:25 Join thegeek_ [0] (
14:53:38pilot000Theres a bug in "General Settings" "Display" "LCD Settings" "Line Selector". If i choose "Bar (Inverse)" the LCD mode will be set to Inverse too.
14:53:57JdGordonwhich target?
14:54:13 Join Obsys__ [0] (
14:54:44pilot000iriver H140
14:54:49 Quit Obsys (Read error: 60 (Operation timed out))
14:54:55 Nick Obsys__ is now known as Obsys (
14:54:58bluebrothersame problem here (h120)
14:55:41bluebrotherand regarding the rockbox menu patch: the browse xyz problem still exists
14:55:49pilot000And one more: in the same menu, the item "Contrast" can not be changed
14:55:59webguest72Can I be granted access as a twiki-user? My name is Paul Claassen.
14:56:15LinusNwebguest72: as far as i can see, you are added
14:56:24bluebrotherwebguest72: I added you to the list a couple of minutes ago. It should work.
14:56:26webguest72OK, I'll try it then
14:56:58JdGordonbluebrother: hmm... I tried it in the sim and i either did it wrongly or i thought fixed it
14:57:05bluebrotherJdGordon: and fade on stop from rockbox menu doesn't work
14:57:17bluebrotherfade on stop works fine from the wps
14:57:25JdGordonpilot000: can you mention these bugs in the forum thread?
14:57:30LinusNbluebrother: that is an old bug
14:57:39pilot000no problem
14:57:51JdGordonim going to bed soon so ill forget otherwise
14:57:54bluebrotherit is? I thought that was resolved for the current file browser view
14:58:01webguest72Thanks! It works.
14:58:07LinusNbluebrother: maybe i'm wrong
14:58:14pilot000btw, good job with the new menu
14:58:29*bluebrother goes downloading a current build
14:58:29 Quit webguest72 ("CGI:IRC (EOF)")
14:59:01pilot000JdGordon, where is the forum thread ?
14:59:11JdGordonunder anouncments
15:00:33bluebrotherhmm. New bug with usb? Just disconnected, and the player freezes :(
15:00:34JdGordonhmm... the menu and file browser should respect party mode when pressing stop, shouldnt they?
15:00:40 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
15:00:42PaulJamJdGordon: a little issue: i think the scroll start delay had a default value of 1000ms before the menu change, now it defaults to 100ms.
15:01:07 Quit GodEater ("CGI:IRC (Ping timeout)")
15:01:14bluebrotherok. Stop from the file browser fades
15:01:26JdGordonand it will fade from the menu in 3 min also
15:01:27bluebrotherStop from the Rockbox Menu doesn't
15:02:06JdGordonbah, you got me confused
15:02:14 Quit thegeek (Read error: 110 (Connection timed out))
15:02:14JdGordonit does in the root patch, not svn
15:03:33JdGordonPaulJam: fixed
15:03:35preglowamiconn: does the exp in gray_core.c handle negative numbers?
15:04:13JdGordonbluebrother: did it crash with usb with svn or root menu?
15:04:30 Quit himitsu (Read error: 111 (Connection refused))
15:04:45bluebrotherit was the root menu version.
15:05:07 Quit petur ("stkov")
15:06:33JdGordonbluebrother: ok, so its not just me... Im not sure why it does that.... MENU_ATTACHED_USB should return a safe value
15:06:43JdGordoncan usb connection be simulated in the sim?
15:06:53LinusNi believe so
15:07:02LinusNnot sure though
15:07:32bluebrotherJdGordon: I added my latest findings to the FS entry
15:07:53bluebrotherthe line selector thing is unrelated to the root menu (unsurprisingly)
15:08:19bluebrotheras it's also present in current svn
15:08:53JdGordondo you know what the config string for that setting is?
15:09:30LinusNfound the bug
15:09:44JdGordoncopy paste i imagine?
15:09:48LinusNsettings_list.c, line 434
15:09:57JdGordonah, just found it
15:11:12PaulJambtw, is it just my imagination, or is the menu a little laggy with a font larger than 12 pixel ?
15:11:41preglowif you're implying there's something wrong with rockbox, clearly your imagination
15:12:28amiconnJdGordon: Only the usb screen can be simulated. There's a menu item just for this in the sim
15:12:29 Quit Obsys_ (Read error: 110 (Connection timed out))
15:13:10JdGordonOK, I tihnk ill add a key for it then to debug this, should I commit that change if it works?
15:13:34LinusNJdGordon: sure
15:13:34JdGordonwas I beaten to the inverse mode fix? or shall i commit?
15:13:41LinusNyou do it
15:13:55amiconnpreglow: Yes it does: if (x < 0) x += 0xb1721, y >>= 16;
15:14:26LinusNJdGordon: or remove the bar cursor setting and enter the Rockbox popularity contest :-)
15:14:37 Join O112358 [0] (
15:14:50*O112358 cries his heart out
15:15:08JdGordonLinusN: ? ppl do or dont like the bar?
15:15:15*LinusN feels bad for O112358
15:15:30O112358thanks :)
15:15:44LinusNJdGordon: ancient rockbox joke
15:15:45PaulJami like the bar
15:16:06 Join himitsu [0] (n=himitsu@
15:16:16O112358my.. *sniff*.. batter.. broke :'(
15:16:31O112358my batter is tasty as usual..
15:16:36LinusNO112358: which player?
15:16:52***Saving seen data "./dancer.seen"
15:16:53LinusNwell, then you can just buy a new one
15:16:55O112358got a battery off ebay.. i think it was in too tight..
15:17:08O112358it leaked..
15:17:13O112358smells bad..
15:17:13JdGordonLinusN: ah ok....
15:17:22PaulJamyou can replace it with an ipod g1/2 battery, but you need to change the polarity
15:17:26 Quit BHSPitMonkey_ (Read error: 60 (Operation timed out))
15:17:43O112358but luckily it didn't get all over my player.. so that's good at least.. and it still works when plugged into the mains..
15:18:05preglowamiconn: and it handles all fixed point numbers with a fractional part of 16 bits?
15:18:08SlasheriO112358: it was a li-ion battery? you should get a li-poly
15:18:13 Join kaaloo [0] (
15:18:16Slasherithat should not leak
15:18:41O112358it was li-poly..
15:19:21O112358is there any way to get into rockbox when plugged into the mains?
15:19:41LinusNnot without a battery, i'm afraid
15:20:48O112358is it possible that because the battery was a bit squashed by the tight fit it leaked?
15:21:00 Quit wLLm (Connection timed out)
15:21:03O112358it had such a nice battery time as well :(
15:22:23PaulJamif you flash the firmware the device turns off, even when connected to the charger. maybe then rockbox can be started. but i'm not sure if it is a good idea to flash the firmware everytime you want to start rockbox.
15:23:22O112358hehe.. i think i'll just have to get a new battery :(
15:23:23amiconnpreglow: It handles all numbers which don't cause an overflow
15:23:46O112358i just want to make sure that the same thing doesn't happen again.. but i don't know what caused it..
15:26:53 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
15:28:17 Quit himitsu ("Leaving")
15:28:36 Join tri170391 [0] (n=tri17039@
15:31:08tri170391I'm trying to modify the jpeg viewer plugin to view jpeg ebooks
15:32:23LinusNjpeg ebooks? interesting concept
15:33:15preglowamiconn: i wonder if errors of 0.3 decibels are going to be audible in the prescaling code...
15:33:19JdGordonwe have working usb simulation in the sim now.. now to see if it helps debug this crash ::p
15:33:48tri170391They are just ebooks translated to series of jpeg images
15:34:00preglowcolour books, then?
15:34:06JdGordonit does crash the sim :)
15:34:18preglowok,then i guess i at least see the point
15:34:39tri170391but i think the jpeg viewer controls theme is not suitablr for ebook viewing
15:34:51preglowprobably not, no
15:35:07tri170391i mean ipod
15:35:16tri170391click wheel ipod
15:35:25tri170391(nano 5g ...)
15:36:00 Join miepchen^schlaf [0] (
15:36:00tri170391the zoom button is too near the move button so it is easy to zoom image accidentally
15:36:15preglowhmm, i wonder how hard it'd be to implement an api for the plugins that makes the core reserve a buffer area in the mp3 buffer for stuff like buffering images
15:36:31tri170391sorry for my bad english because it isn't my native language
15:36:37preglowlike making the mp3 buffer smaller for the duration of the plugin's life
15:37:16JdGordonbluebrother: fixed the usb crashing bug... it was stupid and I should have got it even without simulating usb in the sim :p
15:37:30LinusNpreglow: the tricky part would be to handle it while the buffer is used
15:37:56preglowLinusN: indeed
15:38:17LinusNbut it would surely be a nice thing
15:38:19bluebrotherJdGordon: nice.
15:38:27preglowLinusN: it would require being able to throw away part of the already used buffer
15:38:38preglowwithout further buffering suffering from it happening
15:38:46preglowsuffering as in bugging out
15:39:50preglowwhich again means being able to backtrack in playlists, i guess
15:40:07JdGordonbluebrother: I just ried the folder bug again and cant reproduce it
15:40:32JdGordonwhat setting have you got show title to?
15:43:03bluebrother"Show path" is set to "Current dir only"
15:43:24 Join tanawana [0] (
15:43:41LinusNi can trigger it easily
15:43:59LinusNi just enter Language and then Left to exit -> blank screen
15:44:04JdGordonI did exaclty like the comment said
15:44:49JdGordonexiting the lang screen with left puts me in the menu like I expect
15:45:04LinusNnot me (h100)
15:45:34JdGordonh300 sim... shouldnt be, but maybe thats the problem?
15:45:38bluebrotherok, I just rolo-ed again and tried ... worked as expected for the first time, resumed playback, and now I can reproduce the bug
15:46:40JdGordonah,,, odd
15:46:49JdGordonyes, with playback enabled i get the bug
15:46:57JdGordonwtf? :p
15:46:59LinusNsame here
15:50:04JdGordonhmm... if the wps was entered at all if causes the bug
15:50:31JdGordonand fixed...
15:51:06JdGordonI was letting it exit the dirbrowse function properly in those browsers instead of using return...
15:51:31JdGordonwhich means there is heaps of dead code at the bottom there
15:52:19JdGordon... possibly
15:52:32JdGordonill update FS and then im going to bed
15:52:40 Quit tri170391 (" HydraIRC -> <- The future of IRC")
15:53:05LinusNJdGordon: sleep tight
15:53:16 Join Domonoky [0] (
15:53:36 Quit JdGordon ("zzzzzz")
15:55:16 Join Arathis [0] (
15:59:00 Join shoumy [0] (
16:00:55 Join Mouser_X [0] (
16:03:46 Quit shoumy ()
16:04:43 Quit pilot000 ("CGI:IRC")
16:06:07 Join BHSPitMonkey [0] (
16:07:04 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
16:19:25 Quit GodEater ("CGI:IRC (Ping timeout)")
16:19:34 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
16:20:16 Join miepchen^schlaf [0] (
16:24:50 Quit Mouser_X (Read error: 110 (Connection timed out))
16:25:22 Quit PaulJam (Read error: 104 (Connection reset by peer))
16:25:29 Join PaulJam [0] (
16:30:03 Join perl|work [0] (
16:30:59perl|workrandom question, is ebay down for everybody? works here
16:31:33 Quit Xerion (Read error: 131 (Connection reset by peer))
16:31:39 Join Xerion_ [0] (
16:31:41hcsperl|work: works here
16:31:41O112358nope fine
16:31:46O112358all fine
16:32:03perl|worksome US parts cant access it at the moment it seems
16:32:06perl|workwhich is weird
16:33:25 Nick Xerion_ is now known as Xerion (
16:33:36perl|workalso, my player powered up with "no system found on hdd" today
16:33:40perl|workgigabeat target
16:33:59perl|workbattery flip broight things back to normal
16:34:03perl|worknever happened before
16:36:28 Quit rp- ("leaving")
16:40:28 Join Jonashn [0] (i=d90a1a7e@gateway/web/cgi-irc/
16:41:34 Join SirFunk [0] (
16:42:45preglowjhMikeS: do you think perhaps moving dsp and eq to firmware/ would be a wise thing?
16:50:48 Quit Obsys ("Chatzilla 0.9.77 [Firefox]")
16:50:57 Quit PaulJam (".")
16:53:23 Quit B4gder ("Rockbox TCP exception #04123 - connection reset and user kicked out")
16:54:12 Quit Jonashn ("CGI:IRC (EOF)")
16:54:52preglowamiconn: does the prescaling ever push the final real gain value above 0 db if possible? looks like it to me
16:55:41 Join GodEater [0] (i=c2cbc95c@gateway/web/cgi-irc/
16:57:12 Part perl|work
16:57:35amiconnpreglow: ?
17:00:43 Join perl|work [0] (
17:01:20preglowamiconn: if the prescaling code ever results in a gain over 0 db for the targets where that's possible
17:02:11 Part LinusN
17:02:16preglowlooks like it does, and i guess there's nothing bad with that
17:04:09amiconnPrescaling just compensates for positive treble or bass gain, and in itself never goes positive
17:04:27amiconnIt's just that prescaling is reduced when main volume hits the upper limit
17:05:46preglowthe main volume is what i was asking about
17:06:27preglowjust wondering if the upper limit it's allowed to hit is 0db or even higher if possible, but i see it's the latter, so forget it
17:09:21preglowi think i'll just implement a dsp_set_volume to make prescaling work correctly here
17:09:32 Join nickv111 [0] (
17:11:01 Join st8outtacompton [0] (
17:11:20st8outtacomptondoes ipod linux give the ipod photo video
17:11:48GodEaterwhy ask here ?
17:12:05 Join dmdfan [0] (
17:12:19st8outtacomptonhow do i join it
17:12:42GodEater /join #ipodlinux
17:13:52nickv111You know what's a pain? You can't shift a double.
17:14:06nickv111That makes me angry, because CORDIC requires shifting
17:14:07st8outtacomptonk got it
17:14:18tanawanaspaces get ya huh?
17:14:35hcsyou could shift a double, but it wouldn't do what you want...
17:15:00nickv111If I do 1.0>>1, I want it to give me 0.5. Very simple.
17:15:27hcssometimes you actually have to do division
17:15:54nickv111But that's a pain. In CORDIC, the whole benefit is that you don't ever have to do multiplication
17:15:54 Join norbusan [0] (
17:16:05 Part norbusan
17:16:06 Join norbusa1 [0] (
17:16:07hcswell, maybe you should be using ints?
17:16:08 Part norbusa1
17:16:19nickv111But 1>>1 is 0
17:16:29hcsfixed point ints?
17:16:33 Join norbusan [0] (
17:16:42 Part norbusan
17:16:45preglownickv111: the whole point with cordic is also using straight integers
17:16:53***Saving seen data "./dancer.seen"
17:16:57preglownickv111: if not, the shift property doesn't hold
17:17:05nickv111preglow: What do you mean?
17:17:14preglowyou can use floats, but you need to multiply instead of shift
17:17:37nickv111Well, if we're multiplying instead of shifting, then we might as well be using infinite series
17:17:49nickv111Which, IMO, may not be a bad idea for calculator.c
17:17:57hcsliterally shifting the floating point representation of 1.0 will result in a very small number...
17:18:05 Quit ender` (Read error: 131 (Connection reset by peer))
17:18:25nickv111hcs: Why is that?
17:18:28preglowdepends on the shift direction, but shifting the exponent will yield drastic results for sure
17:18:35nickv111Oh, I see
17:18:46nickv111A double has an exponent and a base, right?
17:18:56preglowit has a sign, an exponent and a mantissa
17:18:59preglowthe base is always two
17:19:04hcsaye, and 1.0 only had a single bit set for the exponent to make it 2^0
17:19:27 Join ender` [0] (
17:19:31nickv111preglow: So the numbers are basically in scientific notation, except with a base of 2 instead of 10?
17:19:44preglownickv111: correct, and they're always normalised
17:20:01preglownickv111: that is, there is always only one digit before the point, and that is always 1
17:20:19preglowsince that's always one, it's also not actually encoded in the double itself
17:20:27nickv111Okay, in calculator.c, check this out. There's always a double, a base, and an exponent. That base is 10
17:20:33nickv111So that's redundant.
17:20:41preglowwell, ok, then they use base 10 floating point, it sounds like
17:21:09hcspreglow: always normalised?
17:21:09preglownot a bad idea for a calculator
17:21:09preglowhcs: yes
17:21:23 Join akaidiot [0] (
17:21:27 Quit SirFunk (Read error: 110 (Connection timed out))
17:21:32nickv111preglow: Wouldn't it be better to convert to scientific notation at the last minute, though, instead of storing the power and all?
17:21:37hcspreglow: but I like my crazy small but fewer significant digits values...
17:22:05preglownickv111: what do you mean? if you use floating point you are always in scientific notation
17:22:19nickv111preglow: Well, you're in x*2^y, right?
17:22:28nickv111preglow: Instead of x*10^y
17:22:44preglowhcs: so what? they'll still be represented accurately, i think you misunderstand what normalised means
17:22:59preglowhcs: normalised in this context means that there is always one number before the fractional point position
17:23:08preglowhcs: so the exponent will always have to be adjusted to make that so
17:23:23hcspreglow: I thought the standard allowed for denormalized values
17:23:25preglownickv111: well, both are scientific notation
17:23:34 Join xandl12 [0] (
17:23:56hcspreglow: when the exponent field = 0
17:23:58preglowhcs: and it does too, when the exponent is encoded a certain way, then there is no longer an implicit 1 in front of the fractional point
17:24:06preglowhcs: but you should never use denormal number anyway
17:24:25nickv111preglow: Okay, what I mean is it's probably best to convert it to base-10 scientific notation at the last minute. As in, leave it as a double, and then use a simple algorithm to print it out as base 10
17:24:40preglownickv111: that defeats the entire point of using base 10 in the first place
17:24:57preglowmidkay: the reason for using base 10 is that humans use base 10, and a calculator always prints its results
17:25:06preglowmidkay: sorry about that
17:25:08preglownickv111: read that :)
17:25:22 Quit st8outtacompton ()
17:25:39preglownickv111: if you use base 2, you will run across numbers that cannot be accurately represented all the time, like 0.1, which you can't express accurately in base 2
17:25:57hcspreglow: just looking for a chance to show off my mostly useless knowledge
17:25:59preglownickv111: and i think you can imagine how surprised someone would be if they wrote 0.1 + 0.1 and got 0.20000003 for an answer
17:26:01nickv111preglow: That doesn't matter. Precision isn't our issue here.
17:26:18preglownickv111: yes it is
17:26:23preglowprecision and representation
17:26:27nickv111preglow: So you're saying you should entirely use base 10?
17:26:35nickv111preglow: In that case, calculator.c still fails
17:26:37preglownickv111: most calculators do, afaik
17:27:13 Join SirFunk [0] (
17:27:17preglowlooks like it just uses doubles to me, yes
17:27:41nickv111preglow: Let me tell you how calculator.c fails in this accord. You, like I've said already, have a double (base 2) and an exponent. As in, the double is multiplied by 10^power. When you multiply these, you are multiplying the doubles and adding the powers
17:27:49nickv111preglow: So calculator.c still suffers from that problem
17:28:55zylchenot as bad as microsoft calculator 3.129, the one where 3.01-3 would get you 0
17:29:01nickv111preglow: Now, I was thinking of maybe having a fixed-point type, and an integer power. These are in scientific notation, and within a struct. As in, I would typedef a struct.
17:29:31 Quit O112358 ()
17:29:46preglownickv111: for the purpose of doing cordic?
17:30:22hcsnickv111: note that most decimal values won't be precisely representable in binary fixed point
17:31:24hcs(or in floating point for that matter)
17:32:45preglowwhy not?
17:33:24hcsassuming his integer power is a power of 2
17:33:34preglowmost integers can be represented accurately
17:33:40preglowdepending on floating point format
17:33:43hcswell, yeah, but 0.1 ?
17:33:49preglow0.1 can't, no, not in base 2
17:33:52preglowbut in base 10 it can
17:34:04hcsof course
17:34:32hcswhich is why I qualified it with "binary"
17:34:46preglowi try to make a habit of skipping random words when i read
17:34:53preglowmaking for more fun ;)
17:36:08hcsjust backing up your point that it should be done in some representation of decimal
17:36:42 Quit ender` (Read error: 131 (Connection reset by peer))
17:38:28nickv111preglow: Okay, since I'm out of ideas, what would you do?
17:39:16preglowi don't even know what you're trying to do
17:40:14 Quit hcs ("Download Gaim:")
17:40:26nickv111I'm trying to get calculator.c working well. I was thinking of fixing the CORDIC, but it might also work to implement infinite series and forget about CORDIC. I doubt the speed difference would be significant, given that in CORDIC we'd still have to multiply if we're using doubles
17:40:57preglowactually, i think the shifts might be implemented as additions/subtractions to the exponents
17:41:07nickv111That's what I figured
17:41:24nickv111Didn't say it, but maybe that's a good idea. . .
17:41:24preglowsubtracting one from the exponent is the same as dividing by two which is the same as shifting right by one
17:42:03nickv111How do you subtract from the double's exponent?
17:42:19 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
17:43:07 Part hcs
17:43:09 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
17:44:29 Join busa_blade [0] (
17:45:33preglowmask it out, subtract, mask it back
17:45:35preglowdinner time
17:45:37 Join robin0800 [0] (
17:45:54 Join _FireFly_ [0] (
17:45:54 Join ender` [0] (
17:46:14 Part _FireFly_ ("Kopete 0.12.3 :")
17:49:18 Nick Everybody|away is now known as Everybody (
17:49:36 Part tanawana
17:51:26 Join linuxstb__ [0] (
17:56:18 Join allstar5449 [0] (
17:58:13 Part allstar5449
18:02:45 Quit linuxstb_ (Read error: 113 (No route to host))
18:06:34 Join Mouser_X [0] (
18:11:12 Join nickv111_ [0] (
18:11:13 Quit nickv111 (Nick collision from services.)
18:11:23 Nick nickv111_ is now known as nickv111 (
18:13:30 Nick Everybody is now known as Everybody|away (
18:16:06 Quit Arathis (Remote closed the connection)
18:17:08 Join TheSphinX^ [0] (
18:17:22 Quit ender` (Read error: 131 (Connection reset by peer))
18:23:00 Join Arathis [0] (
18:27:28 Join tim66 [0] (
18:27:29 Part madduck
18:27:43 Quit bonbonthejon (Read error: 131 (Connection reset by peer))
18:29:26 Quit lee-qid ("aufwiederbyebientotsayonara")
18:29:49 Join lee-qid [0] (
18:34:47 Quit pixelma (" bbl")
18:46:43 Join stoffel [0] (
18:48:43 Quit Mouser_X (Nick collision from services.)
18:48:52 Join Mouser_X [0] (
18:49:36 Quit GodEater ("CGI:IRC (Ping timeout)")
18:50:00 Join redbreva [0] (
18:50:14 Quit xandl12 (Read error: 110 (Connection timed out))
18:51:01redbrevaAdmited, my sim is a few days old now, but has the screendump facility got lost (or moved) in the recent menu re-org?
18:52:04 Part tim66
18:52:10 Join tim66 [0] (
18:52:54bluebrotherredbreva: should be independent of the menu rework
18:57:14 Quit akaias (Read error: 60 (Operation timed out))
19:02:57 Part sp3iel
19:06:16 Nick Everybody|away is now known as Everybody (
19:07:03redbrevaSim from 26th Jan has Debug (Keep Out) but sim from 17th Feb has lost that entire menu
19:07:57 Join akaias [0] (
19:08:19 Quit midgey ()
19:09:21redbrevausing Rasher pre-built sims, I am in the process of rebuilding my PC, so cannot compile my own at the moment
19:11:27 Quit busa_blade (Read error: 145 (Connection timed out))
19:12:23 Join toffe [0] (
19:13:48amiconnredbreva: You don't need the debug menu for screendumps in the sim...
19:14:06 Part kaaloo
19:14:25redbrevaHas that changed?
19:15:25amiconnScreendump in the sim is always enabled, back to when screendump was introduced, in hwcodec-only times
19:16:05redbrevaOK... Alway asumed it had to be switched on - hence the menu option
19:16:22redbrevaor always even
19:16:56***Saving seen data "./dancer.seen"
19:17:30amiconnThe menu option shouldn't even exist in the sim. If it does, it's a bug
19:17:58redbrevaIt doesn't any more - since the rework I assume - it was always there before
19:20:41redbrevaAs mentioned earlier, the entire Debug (Keep Out) option is no longer in the sim
19:21:23 Nick linuxstb__ is now known as linuxstb (
19:21:41 Join Thundercloud [0] (n=thunderc@
19:21:51amiconnThe debug menu was a fairly recent addition in the sim. For hwcodec it made no sense at all (only hardware stuff that doesn't apply to th esim)
19:22:07amiconnScreenshot is F5 and can't be disabled
19:22:53redbrevaf5? I always used 0 on the numeric keypad
19:24:10redbrevahmmm, something old to learn new every day ;-)
19:24:28amiconnOk, it's both F5 and numpad 0
19:24:40*amiconn prefers non-numpad keys
19:24:45amiconnLaptop keyboard.
19:25:02 Join busa_blade [0] (
19:25:18redbrevaAhh - understood
19:30:28 Quit Ribs ("eh eh ehhhh!")
19:31:34 Quit Mouser_X (Nick collision from services.)
19:31:41 Join bonbonthejon [0] (n=jon@
19:31:43 Join Mouser_X [0] (
19:32:46 Join Juice^ [0] (n=Juice@
19:35:00 Quit Juice^ (Client Quit)
19:35:13 Join Juice^ [0] (n=Juice@
19:37:47 Join CpuWhiz [0] (
19:40:21 Join redbreva_ [0] (
19:44:42 Quit Mouser_X (Nick collision from services.)
19:44:51 Join Mouser_X [0] (
19:49:21 Quit kubiixaka ("Miranda IM! Smaller, Faster, Easier.")
19:54:17 Quit redbreva (Read error: 110 (Connection timed out))
19:59:01 Quit CpuWhiz ("Leaving")
19:59:27 Join funky [0] (n=repulse@unaffiliated/funky)
20:01:13*Domonoky would like to commit his rbutil work .. any objections ??
20:05:15nlsgo, go, go!
20:05:31Domonokydoes anybody know why the new bootloader (and bootsplash.raw) isnt on ?
20:06:02Domonokybootloader for Gigabeat ^^
20:06:08nlsI think it's because that whole thing with the splash faced some resistance...
20:06:35Domonokydoes the old Bootloader on work ?
20:06:57nlsyes, it's just the splash stuff that differs
20:07:36Domonokyah, thats good, then i can enable bootloader install for Gigabeat in rbutil :-)
20:08:14amiconnDoing the bootsplash properly would require introducing parameter passing from bootloader to core
20:08:35amiconn...because if there is no splash, the core should show the rockbox logo
20:08:53amiconnAfaik it doesn't do that on gigabeat currently
20:09:17Domonokyok, rbutil will offer the old one :-)
20:09:45Nico_Pwhat does the old one display ?
20:10:06Domonokyanother question, why is the guy who ported chip8 to Rb not in the credits ? (he is mentioned in the source)
20:10:38amiconnIt would be useful to know how much the splash code slows down boot, both with present and absent splash file
20:11:23amiconnNico_P: Afaik the old one displays the same things as other swcodec bootloaders
20:11:29 Join linuxstb_ [0] (i=5343d4aa@rockbox/developer/linuxstb)
20:11:32*amiconn doesn't have a gigabeat
20:11:59Nico_Poh i might try that
20:12:07linuxstb_raw bitmap files also seem very inconvenient for users - I wonder how much slower a .bmp file would be.
20:12:28Nico_Phaving the bootsplash is a bit weird... i'm still used to the text
20:12:53amiconn*Imho* the bootloader shouldn't display anything unless something goes wrong
20:13:09linuxstb_I agree with that.
20:13:43amiconnThe standard swcodec bootloaders remind me of linux booting... lots of cryptic stuff scrolling fast
20:14:16amiconnPerhaps interesting for technicians, but certainly not for users
20:14:21*TheSphinX^ is wondering about the patch worked but since some svn-updates ago the new menu entry is no longer showed in the menue :(
20:14:52linuxstb_amiconn: One problem is if the bootloader crashes - no chance to display errors...
20:15:11amiconnHow would that be different from now?
20:16:17linuxstb_We now know approximately where the crash is.
20:16:44linuxstb_But bootloaders shouldn't crash, so...
20:18:36Nico_Pamiconn: about the AMD64 warnings... would be the solution ?
20:18:47amiconnWhich warnings?
20:19:18amiconnThe AMD64 warnings happen when building the sim and someone wasn't careful.
20:19:51amiconnNico_P: That's also a sim, and it's *not* a 64 bit warning
20:19:58Nico_Pah... i thought it was
20:20:17Nico_Pwhat is it then ?
20:20:20amiconnThis build server is 32bit. It uses a known bad snapshot of gcc 4.1.2 for its *native* gcc
20:20:39amiconnStrange distro that uses a snapshot from half a year back...
20:20:57amiconndebian had that snapshot last summer... for around 3 weeks iirc
20:21:09Nico_Pso should the tracker entry be closed ?
20:21:38 Join x1jmp_ [0] (
20:21:41amiconnThe backport of the 64 bit fixes for m68k-elf-gcc should probably be incorporated in the gcc-rockbox-patch
20:22:21amiconnThey're probably better than the hackaround currently used (although even that hackaround doesn't cause problems)
20:22:56amiconnI didn't try it yet; still have the hackish one running
20:22:58 Join petur [0] (n=petur@rockbox/developer/petur)
20:24:08amiconnThe issue with m68k-elf-gcc isn't warnings, but m68k-elf-gcc (at least up to 4.1) can't be built on amd64 without patching
20:26:00 Join entheh [0] (
20:26:02Nico_Pcan't godeater update his gcc snapshot ?
20:26:08amiconnThe hackish fix is here: , last file attachment
20:27:10amiconnHe should, as soon as a fixed one becomes available for his distro
20:28:03 Quit funky ("leaving")
20:28:08 Join funky [0] (n=repulse@unaffiliated/funky)
20:29:49amiconnGodEater_: around?
20:30:43 Join redbreva__ [0] (
20:30:49 Nick redbreva__ is now known as redbreva (
20:31:54 Quit BigBambi ("Leaving")
20:38:29XavierGrwow amiconn where did you digg that up? I had completely forgotten about this conversation.
20:38:58 Quit x1jmp (Read error: 110 (Connection timed out))
20:39:09amiconnI was just googlich for mregparm (without the minus, b/c that means "not" in the search mask)
20:40:58XavierGryeah full ntfs support for linux
20:41:02XavierGrversion 1 is out
20:42:59 Quit x1jmp_ (Read error: 110 (Connection timed out))
20:43:44TheSphinX^now i can say "yay i knewed it before XavierGr!" :D
20:44:16XavierGrwell I just saw that they declared the rc1 into stable
20:44:27XavierGrso it is now finally official
20:46:04TheSphinX^but i think we are in different timezones, on my location it's 8:45pm and i saw it today in my newsfeeds :o)
20:46:55 Quit redbreva_ (Read error: 110 (Connection timed out))
20:48:17 Quit Mouser_X (Read error: 110 (Connection timed out))
21:03:54 Quit funky ("leaving")
21:04:50 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
21:06:14 Join Ribs [0] (n=ribs@
21:11:12 Quit stoffel ("leaving")
21:12:54*Domonoky updated the rbUtil patch, it now supports all Targets (without Sansa) should i commit it ?
21:13:41dan_aDomonoky: Go for it!
21:14:09hcsADX only plays the right channel...
21:17:01***Saving seen data "./dancer.seen"
21:17:19 Join x1jmp_ [0] (
21:17:34 Nick x1jmp_ is now known as x1jmp (
21:17:37hcsis STEREO_INTERLEAVED broken?
21:18:31 Join darksoulk [0] (
21:19:12preglowit might be, jhMikeS touched that code a couple of days ago
21:19:27preglowi'll check it out in half an hours time if no one else does
21:19:30Domonokyhm, it seem i dont have commit access..
21:20:20dan_aDomonoky: Bagder is the person to talk to about that
21:20:32 Join pixelma [0] (i=pixelma@
21:20:42DomonokyBadger !!! :-)
21:20:52Domonokyhh Bagder !! :-)
21:21:16Domonokysomeone should wake him up :-)
21:22:39LinusNDomonoky: have you had svn access before?
21:22:43VakioMushroom, mushroom?
21:22:57Domonokyno, but i was invited via email..
21:23:03 Join Thundercloud_ [0] (n=thunderc@
21:23:15LinusNdid you reply with your username/password?
21:23:22 Join GreyFoux [0] (
21:23:40Domonokyjeah i send a mail to daniel @ rock ...
21:23:49 Quit GreyFoux (Client Quit)
21:24:47 Quit nickv111 (Read error: 104 (Connection reset by peer))
21:26:10preglowsweet, sweet lord
21:26:15preglowwhy is gnome so slow
21:28:34*Domonoky made his first commit.. :-)
21:31:36*LinusN awaits Domonoky's first "Ooops" commit
21:32:26 Quit Thundercloud (Read error: 60 (Operation timed out))
21:32:39*Domonoky waits for testers, so he can say "Ooops" :-)
21:32:45*amiconn wonders why the build system still doesn't know about directories where changes shouldn't trigger a rebuild
21:33:08redbrevashouldn't it be full name in the 'who' commit table?
21:33:40LinusNamiconn: simple - we haven't cared enough to spend time on fixing that
21:34:05LinusNredbreva: yes
21:35:28amiconnRebuilding all cores and sims if no relevant code changed seems like a waste to me
21:35:43amiconnIn cvs times the build system knew the difference
21:36:25LinusNi agree that it is a waste
21:37:06LinusNbut fixing that is on my low-priority list
21:37:10 Quit Shaid (Read error: 110 (Connection timed out))
21:37:11 Nick darksoulk is now known as Shaid (
21:37:16Domonokycan i request Wiki write access? i like to updated the rbutil page
21:37:26LinusNDomonoky: wikiname?
21:38:35LinusNDomonoky: done
21:40:11linuxstb_Anyone have suggestions about where to put (beta for now...) binaries of rbutil? I can build a Mac version.
21:41:25pixelmalinuxstb: you adviced the guy in the forum to google for disk mode - isn't it already in the first post a few lines above? :)
21:41:45Domonokywiki ?
21:42:11linuxstb_pixelma: :) He obviously didn't see it there... (like me!)
21:42:48*amiconn wonders whether it's worth trying gcc 4.0.4
21:43:14linuxstb_It just annoys me when people post questions like "how to I put my ipod in disk mode" when typing that question into google gives 100 pages wth the answer on...
21:44:05preglowamiconn: i wouldn't bother...
21:44:28 Nick Everybody is now known as Everybody|away (
21:45:30pixelmalinuxstb: yeah, that too - but obviously he is one (of the pitily many) who don't really read instructions...
21:47:34 Quit TheSphinX^ ("Miranda IM! Smaller, Faster, Easier.")
21:49:09perl|workdevices with recording capabilities, what shows up on the screen in recording mode?
21:50:56perl|workWPS wise, should anything exctra be included for them?
21:50:57linuxstb_A peakmeter and gain settings, plus a few other things IIRC (I'm sure someone else will reply properly...)
21:51:40linuxstb_The recording screen contains controls, it doesn't just display things.
21:52:30 Quit Rondom (Read error: 113 (No route to host))
21:53:32pixelmathe recording screen looks something like this (status bar isn't standard, file info,...) can also look a bit different for triggered recording for example
21:54:11preglowhcs: think i've found a bug, yes
21:55:08 Quit Juice^ (Read error: 104 (Connection reset by peer))
21:55:11 Join Juice- [0] (n=Juice@
21:59:53preglowanyone know where i can get adx files? i can't seem to find any
22:00:24 Join bombastic1 [0] (
22:01:51linuxstb_Domonoky: On the second page of the bootloader wizard (after I choose "ipod" from the list of target), it says "Rockbox utility has detected the following location:", but I need to press SCAN before it will detect it.
22:02:41 Quit darkless ("Leaving")
22:02:48 Join darkless [0] (
22:02:58weethow do i put back my original firmware ?
22:03:59 Quit bonbonthejon (Remote closed the connection)
22:04:07linuxstb_Domonoky: Also, Mac OS X automounts the ipod, and the user needs manually unmount it before ipodpatcher can write to the disk. It could be worth testing that rbutil has write access to the device before downloading the bootloader, and if it fails, tell the user what to do (or we could even run the unmount command).
22:04:32linuxstb_weet: Which player are you talking about?
22:04:50weetIpod 5G
22:05:10preglowdoes rbutil need any particular version of wxwidgets?
22:05:14linuxstb_Are you asking how to dual-boot, or how to uninstall Rockbox?
22:05:15weetI know last night i was pressing random buttons and the original firmware poped out
22:05:21weethow to uninstall :/
22:05:24linuxstb_preglow: 2.8.0 - it was released a few weeks ago.
22:05:30linuxstb_weet: "ipodpatcher -d"
22:05:39linuxstb_Or use itunes to restore your ipod.
22:05:42Domonokylinuxstb i will look into it..
22:05:58preglowlinuxstb_: small chance i'll find it in the ubuntu repos, then
22:06:11bluebrotherDomonoky: rbutil gives a windows error if the ini file isn't found.
22:06:27bluebrotherwouldn't it be better to just create a new one in that case?
22:06:32bluebrotheror is that not feasible?
22:06:56Domonokythen i would have to include the ini file somehow..
22:07:05linuxstb_Or download it from a permanent URL.
22:07:26weetin Ipodpatcher
22:07:29weetwhat do i type
22:07:32Domonokythats also possible..
22:07:33 Join Winchester [0] (
22:08:01Domonokybut i dont see the problem with distributing the exe together with the ini..
22:08:02linuxstb_weet: You run ipodpatcher with the "-d" option from the command prompt.
22:08:51preglowlinuxstb_: wouldn't it have been wiser to use an older version? or is it easy to statically link wxwidgets?
22:09:42linuxstb_It was Cassandra's choice, but (at least for the Mac), wxwidgets is still under heavy development, so I think it makes long-term sense to use the latest version.
22:10:27Domonokyat the moment i just link statically
22:10:27preglowsure, i'm all for using newer versions of stuff myself, really, but in this case it makes life harder for me :P
22:11:05Domonokyi will upload a linux binary shortly the we will see if there are bad depencies :-)
22:11:28preglowif you don't link statically, there will be for sure
22:11:40preglowalso, i'm on 64 bit linux so i'm practically begging for trouble
22:11:56Domonokyits statically linked, at least i told wxwidget so.. :-)
22:11:58Mikachuwouldn't it be better to use a library that's used by more than like 2 projects? :)
22:12:07Domonokyjeah that doesnt work on x64
22:12:11weetmy ipod is fixed :D
22:12:17Domonokywe need extra binarys for x64
22:12:37weetthx linuxstb_
22:12:44Mikachuweet: no, now it's broken ;)
22:12:50Mikachuinstalling rockbox fixes it
22:13:19 Join midgey [0] (
22:13:49preglowMikachu: like what?
22:14:02Mikachugtk (or qt)?
22:14:07preglowgtk is hell in windows
22:14:09preglowqt is a better choice
22:14:22preglowat least personally, i'm fed up with gtk
22:14:41weetMikachu : no wai! my power drained in 3 hours
22:14:52weetMikachu : with rockbox
22:14:59Mikachuthen you did something strange
22:15:12weetlike ?
22:15:14Mikachudid you play doom constantly for 3 hours?
22:15:28weeti booted it like 2 or 3 times
22:15:30Mikachudid the backlight turn off?
22:15:35weetnever did
22:15:36 Part [shodan] ("Leaving")
22:15:46Mikachuso instead of turning the backlight off you decide to throw it away?
22:15:57Mikachusee how long the apple firmware lasts if you set the backlight always on there
22:16:00weetlol yeah I dint see any option to turn the backlight off
22:16:17weetlike on the ipod firmware
22:16:20Mikachui think it's sneakily hidden away under settings/display/lcd/backlight
22:16:38weetis their an option like
22:16:45linuxstb_Mikachu: What would you suggest instead of wxwidgets to create nice-looking native GUI binaries on win32, Linux and Mac OS X?
22:16:47weet" turn off after 30 seconds of no-touching "
22:17:20Mikachui don't know, but wxGTK doesn't really look that nice either
22:17:34*bluebrother likes Qt
22:17:41Mikachuis there a mac version of qt?
22:17:53desowinlinux, mac and windows
22:17:55 Join amiconn_ [0] (n=jens@rockbox/developer/amiconn)
22:18:00*bluebrother got beaten
22:18:05 Quit amiconn (Nick collision from services.)
22:18:07 Nick amiconn_ is now known as amiconn (n=jens@rockbox/developer/amiconn)
22:18:09desowinsince qt4 it's even opensource on windows
22:18:11linuxstb_But isn't qt it's own gui, rather than compiling to native widgets?
22:18:12weetI dont know why
22:18:20bluebrotherand Qt Designer is quite nice for quickly building up dialogs
22:18:23desowin(it's dual licensed)
22:18:24weetbut i had the impression playing songs took more battery life then playing Doom
22:18:53preglowlinuxstb_: qt uses theming apis now
22:19:00preglowlinuxstb_: if you want it to, that is
22:19:02bluebrotherQt used emulation on Windows mostly. On Mac it uses the OS functions, and also on Windows Luna
22:19:29Domonokyqt would also be a good choice for rbUtil, but i dont wont to rewrite rbUtil.. (i only extended it)
22:20:04linuxstb_I guess I'm just prejudiced against Qt because of the early licensing conditions.
22:20:13*desowin likes idea of basing rbutil on qt
22:20:17Bagderpersonally, I find qt a pain
22:20:33Bagderbut then I guess everything gui and c++ would be pain to me
22:20:34Mikachui'm prejudiced against Qt because it takes like 6 hours to compile and is a lot more than just a toolkit
22:20:34bluebrotherI recently started using Qt on windows. Too bad I don't have time for looking into a Qt rbutil version
22:20:55Thundercloud_Mikachu: Just compile the sub-tools and it's a lot quicker.
22:21:09desowinanyone started work on qt's rbutil ?
22:21:33linuxstb_What would changing to Qt gain us?
22:21:33Domonokyit would be a nearly complete rewrite..
22:21:46Mikachuit would work for preglow :P
22:21:47Domonokyat practically no gain..
22:21:58linuxstb_And I'm sure preglow needs a GUI installer...
22:22:03desowinor make 2 versions, one for wxGTK and one for qt ....
22:22:06bluebrotherwx looks awful on linux
22:22:14linuxstb_Looks good to me...
22:22:17Thundercloud_I'm a KDE man
22:22:19Thundercloud_Qt ftw
22:22:24Thundercloud_Can't wait for Kde4 :P
22:22:26desowinI'm xfce man, but qt is my choice
22:22:35bluebrotherif the functionality is separated out of the GUI I don't think it would be too much work
22:22:37preglowi'm considering applying for a job at trolltech, so i may be biased :P
22:22:44Thundercloud_preglow: Ahh
22:22:47Thundercloud_GNOME MAN EH?
22:22:53preglowi hate gnome...
22:22:57preglowuse it, but hate it
22:22:59Thundercloud_preglow: That was a joke :(
22:23:08preglowtoo tired for that :/
22:23:08Thundercloud_Wait, why do you use it if you hate it? KDE ftw man
22:23:17preglowcan't be bothered
22:23:23MikachuThundercloud_: you need to work on your argumentation skills
22:23:24Thundercloud_KDE is Linus Torvalds Approved (tm)
22:23:27preglowwe'll see the next time i feel like mucking about with aptitude or installers
22:23:35Domonokyfunctionality is not really seperated and everything uses things like wxString for Stirngs..
22:23:44Thundercloud_Mikachu: My arguing skills are fantastic.
22:23:55bluebrotherhmm. I liked the way I can handle strings with Qt.
22:23:57Thundercloud_Except when they're not
22:23:59Thundercloud_Which is always
22:23:59Mikachuyeah, subjective argument right there, nice going
22:24:24bluebrothermaybe I'll give it a try when I find free time ... which may happen in a couple of years :(
22:24:28DomonokywxStrings are also very nice.. didnt miss anything against qtstrings.. :-)
22:24:46preglowhcs: can you recommend me some decent test files among them?
22:28:37 Join Mouser_X [0] (
22:28:56*Domonoky attached a zip with linux binary of rbutil to the RockboxUtility wikipage, would be nice if someone tests it..
22:29:27hcs is stereo
22:30:01linuxstbDomonoky: It could be worth writing the SVN revision of rbutil in the attachment comments (and maybe the filename).
22:30:33linuxstbDomonoky: I've just tried compiling it myself, and get a segmentation fault during the scan for ipods...
22:31:38linuxstbIn wxStrlen...
22:32:06Domonokyin which Os ?
22:32:44linuxstbDebian 32-bit.
22:33:36 Join JustJohnny68 [0] (
22:34:47linuxstbDomonoky: And your ".zip" file is a 7zip file....
22:35:38 Quit desowin ("use linux")
22:36:23amiconnHmm, how am I supposed to build rbutil?
22:36:41Domonokytype make in the rbutil folder
22:36:48Domonoky(if on linux)
22:36:54amiconnHmm, seems I am missing something
22:37:04amiconnwx-config: command not found
22:37:24Domonokywxwidgets2.8 isnt correctly installed..
22:37:43JustJohnny68Nico_P: Thanks for apply my patch to the Rockboxed theme the other day!
22:37:43amiconnNo wxwidgets 2.8 in debian unstable
22:37:51amiconnLatest is 2.6.3
22:38:01Domonokyyes, only selfbuiling..
22:38:03Nico_PJustJohnny68: you're welcome :)
22:38:10Nico_Pdid i do it right ?
22:38:17amiconnSo no amd64 rbutil for now
22:40:05JustJohnny68Nico_P: far as I can tell :)
22:40:38preglowamiconn: as linuxstb said, that release seems to be very new
22:40:48preglowso don't expect to it see it in package managers soon
22:41:17preglowhcs: got stereo back now, jhmikes made a little nasty in dsp.c ...
22:41:25Nico_PJustJohnny68: in your patch the changes i made before to the 220x176x16 version were reverted... i corrected that, assuming that you diffed against an old svn revision that didn't have these changes
22:42:37JustJohnny68Yeah, that's what happened. I haven't had a chance to run the simulator with your changes yet.
22:42:56linuxstbAnother problem will be that most distros (I think) ship a Unicode build of wxwidgets- rbutil doesn't compile with that. You need to use an ansi version (which is the default if you build yourself).
22:43:24preglowjhMikeS: there?
22:43:27Winchesterhey... when porting something to RB.... what's the best source to look for... DOS/Linux/Windows?
22:43:32bluebrotherbut why doesn't rbutil work with unicode? For later localization we should imho
22:44:11linuxstbBecause Cassandra didn't use the right macros for her strings... It needs someone who understands how unicode works in wxwidgets to fix rbutil.
22:44:34amiconnRelease more than 2 months ago, hmm...
22:44:39bluebrotherthere are different macros for strings?
22:45:04linuxstbDon't ask me, I don't understand unicode in wxwidgets....
22:45:10 Quit Mouser_X (Nick collision from services.)
22:45:10petur_T ?
22:45:17linuxstbBut I _think_ it's related to the macros.
22:45:19 Join Mouser_X [0] (
22:45:41bluebrotherI've seen that ... but as far as I understand the docs it should work for both
22:45:50bluebrotherbut I barely know wx
22:46:25linuxstbI just remember running into that problem and the obvious things (which I think was adding _T) didn't seem to work.
22:46:46 Join Thundercloud__ [0] (n=thunderc@
22:46:49preglowjhMikeS: check the fix i did now in dsp.c, that endian specific code below my fix looks dubious to me. is it really necessary?
22:47:01linuxstbBut I now don't have a unicode version of wxwidgets to test it with.
22:47:13 Join mattzz [0] (
22:47:41 Part JustJohnny68
22:49:15preglowjhMikeS: it's not like the channel order is endian specific, now isit
22:49:48*Domonoky is just reading about Unicode in wxWidgets, should be easy to change..
22:53:52 Join funky [0] (n=repulse@unaffiliated/funky)
22:54:01 Join bospaadje [0] (
22:55:25tim66my h140 gets very confused why I skip between ogg and mp3 albums a few times
22:56:04tim66it often ends up displays two menus over each other in a flickering/strobing kind of way
22:56:51 Quit Juice- ("Leaving")
22:57:16tim66anyone else seen this or have any idea why it might be happening?
22:58:38bluebrothertim66: no, but have you tried resetting your settings?
23:03:59 Quit Mouser_X (Nick collision from services.)
23:04:08 Join Mouser_X [0] (
23:10:48 Quit redbreva ("Chatzilla 0.9.77 [Firefox]")
23:11:03 Quit Thundercloud_ (Connection timed out)
23:16:07 Part perl|work
23:17:05***Saving seen data "./dancer.seen"
23:19:13preglowNico_P: there?
23:19:20 Part nls
23:19:29Nico_Ppreglow: yes
23:19:56preglowNico_P: i got some updated dumb sources from the guy who did the foobar plugin once and he told me lots of stuff was fixed in it
23:20:05preglowit's at if you want to have a look
23:20:43tim66bluebrother: will do, i tried using a build from about a month ago but it just hangs now
23:20:58preglowNico_P: it's quite old, but so is the latest dumb release, i see
23:21:07 Part tim66
23:21:49Nico_Ppreglow: you sure this is for me ?
23:22:44preglownot at all
23:23:12preglowseems you just edited the flyspray entry, yes, ignore me
23:23:16Shaidwhats dumb?
23:23:17preglowi'll post a comment there instead
23:23:27Shaidother than me for not knowing
23:23:29bluebrotherShaid: a mod decoder
23:23:36Shaidoh, nice.
23:23:56ShaidI was half thinking of writing a s3m codec, but I havent actually checked if someone is already doing/done one
23:24:09BagderShaid: you do know that your apostrophes are "weird", right?
23:24:15Nico_Ppreglow: i edited it because it was in the wrong category
23:24:23Shaidweird how?
23:24:32Bagderas in weird utf8
23:24:40preglowNico_P: yeah, i just gleaned fast over it and saw you're name so assumed you posted it :>
23:24:40Bagderthey are plain boxes to me
23:24:46Shaidhow odd.
23:24:48bluebrotherthey show up as inverted R for me
23:25:02preglowlook light highlighted R's here
23:25:06preglowand i am on an utf8 console
23:25:07amiconnBagder: They're not utf8, but not "proper" apostrophes either
23:25:10Shaidas cool as that is...
23:25:16Bagderok, was just a guess
23:25:18amiconnThey're iso8859-1 accents
23:25:21bluebrother... but an apostrophe is in ASCII, so there shouldn't be any problem here
23:25:26*bluebrother suspected that
23:25:57Shaidnever been told this before
23:26:02Shaidand I use them all the time
23:26:07Bagderand after having started that off-topic talk, I'll go to bed ;-)
23:26:10preglowBagder: not on an iso8859-1 compatible client, eh!?!?
23:26:15Bagderyes I do
23:26:24BagderI use iso8859-15 atm actually
23:26:27bluebrotherthere is a reason why these strange characters don't show up immediately (at least on windows) *sic*
23:26:34Bagderwhich is why amiconn's comment surprises me
23:27:23*amiconn re-checks
23:27:24scorchei can see amiconn's backtics, but i still cant see Shaid's "apostrophies"
23:27:32bluebrotherbut something is still weird, amiconn's apostrophes show up correctly for me
23:27:53ShaidIm using a Mac keyboard on my pc, but I dont really see how this causes it.
23:27:55pixelmathe logs are in ISO-8859-1, this is what I get too
23:28:06ShaidIm pushing the normal apostrophe key, I swear...
23:28:21amiconnHeh, they are Win1251 accents
23:28:44Shaidquick fix?
23:31:03 Quit Mouser_X (Nick collision from services.)
23:31:11Domonokysomeone with Flyspray superpowers can close FS #6643 please..
23:31:12 Join Mouser_X [0] (
23:31:38 Join Thundercloud [0] (n=thunderc@
23:33:19bluebrotherDomonoky: done
23:33:57bluebrotherIIRC Bagder can add you to the developer list in FS so you can close tasks
23:34:25preglowbluebrother: seems we cooped on that one
23:37:23SoapI see Shaid's "apostrophes" just fine.
23:39:39scorchei see them fine on my windows client
23:39:49amiconnMee too, but they're not correct apostrophes. They're slanted
23:40:10 Quit Ribs ("eh eh ehhhh!")
23:40:20 Join ackbahr [0] (
23:40:20scorchei see them as the "elegant" apostrophes
23:41:47Shaidtheyre standard straight ones when they leave here... :(
23:42:02 Quit funky ("leaving")
23:42:50midkayit's a subtle slant.
23:43:05Soapdepends on the font
23:43:06 Join fasmaie [0] (
23:43:19Soapsome fonts you won't see a difference at all.
23:43:28scorchei use the standard x-chat monospace
23:44:16Soapmicrosoft sans serif 11 shows them with a cute curl down and to the left.
23:44:25scorchewhat keyboard mapping do you have Shaid?
23:44:54*midkay uses lucida console and sees them.
23:44:56scorcheSoap: that sounds like what i am calling the "elegant" apostrophe
23:45:30midkaywhy not just call it slanted or curled?
23:45:42dan_aYay! Using GCC 4.1.1 saves about 8kb on a G4 iPod
23:46:02scorchebecause it is the one where it is like a circle up top and then curls down to the left
23:46:23preglowamiconn: proper apostrophes are supposed to be slanted
23:46:29preglowat least typographically
23:46:34bluebrothersounds like the english closing single quote
23:46:40preglowi guess terminals use some kind of typewriter bastardised thing
23:46:42bospaadjewhen did shaid use apostrophes? im not seeing any
23:46:44Soaptypographic bigots, the lot of you!
23:47:02amiconnpreglow: Yes, sure, but the standard ascii apostrophe is not, at least not in the font I am using here
23:47:02scorche<Shaid> theyre
23:47:30bluebrotherthe ASCII apostrophe is different to the typographical one ...
23:48:40amiconndan_a: I know a way that'll save 40KB+ on ipods...
23:49:04amiconn...but it requires going through many source files, adding ICODE_ATTR to declarations
23:49:09preglowstill, that's not bad at all for just a compiler upgrade
23:49:09dan_aamiconn: Getting rid of -mlong-calls?
23:49:30 Quit mattzz ("Leaving")
23:49:39amiconnIn fact I was thinking about it, and it looks like it can be done incrementally
23:49:51preglowamiconn: hmm? i thought gcc hatet section attributes on declarations
23:50:02amiconn...flipping the actual switch in svn when all files are converted
23:50:18preglowamiconn: i'd probably be happy to help with it
23:50:20amiconnpreglow: Section attributes don't matter
23:50:34dan_aamiconn: Sounds like a good plan. I'll help.
23:50:43amiconnThe point is that __attribute__((long_call)) obviously counts as part of the function type,
23:51:08amiconnso it has to be applied both to (all) declarations *and* the definition, consistently
23:51:49amiconnFirst I had to find out how to add attributes to the definitions, because
23:52:16amiconnvoid myfunc(int myarg) __attribute__((myattr)) { do_stuff }
23:52:19amiconndoesn't work
23:52:25preglowindeed it doesn't
23:52:30amiconnIt must be written as:
23:52:44amiconnvoid __attribute__((myattr)) myfunc(int myarg) { do_stuff }
23:52:55 Quit ackbahr (Remote closed the connection)
23:52:55preglowthose gcc boys sure enjoy their liquor
23:53:13amiconnFor the declaration it works in both places
23:54:20preglowso that way we can be rid of that stupid extra declaration we currently use?
23:54:21amiconnSo basically all functions that use ICODE_ATTR need to have that on all declarations and the definition, consistently, obeying the selectable iram features stuff
23:54:59preglowbut sure, i'll help
23:55:11amiconn...and then ICODE_ATTR would be defined as __attribute((section(".icode"),long_call)) for arm targets, *except* the iFP
23:55:34amiconnSorry, __attribute__((section(".icode"),long_call))
23:55:55 Join ender` [0] (
23:56:41preglowso ok, what to do: remove all the extra declarations we currently have above some function definitions and prefix the function name with ICODE_ATTR, then also do the same if there are any declarations other places?
23:56:53preglowmostly .h files, etc
23:56:55amiconnAnd the we can finally remove -mlong-call, leaving only iram functions as long call targets
23:57:12 Quit pearldiver (Read error: 60 (Operation timed out))
23:57:12 Quit Thundercloud__ (Read error: 110 (Connection timed out))
23:57:16 Quit Winchester (Read error: 110 (Connection timed out))
23:57:25 Join rotator [0] (n=e@rockbox/developer/rotator)
23:58:13 Join Acorn [0] (
23:58:21 Nick Acorn is now known as o112358 (
23:58:22amiconnpreglow: The extra declarations within the .c files can go away when tacking ICODE_ATTR to the definition. And yes, we also need to change the .h files
23:58:28 Quit lee-qid (Read error: 104 (Connection reset by peer))
23:58:28 Nick o112358 is now known as O112358 (
23:58:42amiconn...which is a bit nasty for the standard includes
23:58:42 Join lee-qid [0] (
23:58:48O112358hmm.. would there be anyway to change the firmware on something like this? :
23:58:53 Quit lee-qid (Client Quit)

Previous day | Next day