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

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

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

#rockbox log for 2005-03-05

00:03:14Rickweird results from snprintf
00:03:34Rickfirst loop does it fine
00:03:46Rickthen I get weird characters like the buffer isn't set
00:05:40Rick(myguess, char values around 255ish)
00:06:06Bagdermy guess is that your code is wrong ;-)
00:06:16Rickrb->snprintf(update, sizeof(update), "%d", offset);
00:06:26Rickrb->lcd_putsxy(BLOCK_WIDTH * 2, BLOCK_HEIGHT * 2, update);
00:06:35Bagderheh, that is simple snprintf() usage
00:06:45Rickyes, exactly :P
00:07:03Rickoffset() is a simple int that gets a random value each loop
00:07:49amiconnHow large is your buffer, and how many digits can offset have?
00:08:07Rick24, just to be safe
00:08:10Rickat max it should be two characters
00:08:33BagderRick: does this happen in the sim?
00:08:34Rick(1-3, can be negative)
00:08:47 Quit Chamois (" HydraIRC -> <- The future of IRC")
00:08:49RickI don't have the sim built (no xfree in my gentoo build enviornment)
00:08:50amiconnMy guess would be that some other part of your code trashes memory
00:08:57Rickgood point
00:09:02*Rick looks
00:10:18Rickah, I bet I know what it is
00:10:37Rickactually, no that can't actually be it
00:11:17 Join amiconn_ [0] (
00:11:31Rickthat's gotta be it
00:11:52 Quit amiconn (Nick collision from services.)
00:11:52 Nick amiconn_ is now known as amiconn (
00:13:40 Part geoff_o ("Kopete 0.9.1 :")
00:13:45Rickyeah, that was it
00:13:48Rickthanks for the assist :P
00:13:54preglowcodec.h _IS_ included, and still it manages to complain about implicit decl of malloc and free
00:23:22amiconnHmm, bad. However, I have an idea. Maybe malloc() isn't used in a regular function call, but somewhat modified so the redefinition to codec_malloc() doesn't match?
00:23:43amiconnCan I take a look at the source somewhere?
00:23:49preglowi'll have a look
00:23:59preglowit's more than possible i've fucked up somehow, i'm not very motivated right now
00:24:51preglow pow_4_3_t[0] = (POW_TYPE *)malloc( (POW_4_3_MAX * sizeof(POW_TYPE
00:25:08preglowadd )));
00:29:39amiconnHmm, that looks quite regular. Maybe a space just before malloc would help? I really don't know; that's strange
00:30:22Bagderits just a warning, isn't it?
00:31:18preglowbut it's a significant one, it isn't supposed to see malloc at all, just codec_malloc
00:32:22Bagderrun gcc -E and check the preprocessed output
00:32:28Bagderit usually helps to understand
00:32:51 Quit michiel_ ("Leaving")
00:33:56 Join stevenm [0] (
00:34:29stevenmHello, people. Does anyone have the slightest idea as to how to interpret the envelope bytes of a Gravis patch file?
00:34:48preglowstevenm: stripwax found out just how
00:34:58preglowstevenm: in a gus sdk of some kind
00:35:08stevenmpreglow, yeah?? Where is he?
00:35:27stevenmpreglow, woah, I was just about to actually start trial-n-error as soon as I got this stupid burner fixed
00:35:30preglowstevenm: you tell me
00:35:38preglowi have no idea
00:35:42stevenmpreglow, when did he find out?
00:35:47preglowhe mentioned one or two days ago
00:35:53preglowsearch for 'envelope' in irc logs
00:36:02stevenmawesome. why don' I do that
00:36:23stevenmpreglow, also, do you think it would be possible to test this thing on the target? everything but panning and envelopes are done
00:36:44preglowstevenm: 01.03.05, about 20:45
00:36:45stevenmpanning/stereo support is easy to write, but do we want the extra few operations per sample?
00:36:51preglowstevenm: sure
00:37:05preglowstevenm: a few extra operations change nothing
00:37:16preglowonce we've got the data in a register, not much is wasted on doing one more mul
00:37:23stevenmpreglow, for panning, it is like, 2 multiplies, 2 subtracts, and 2 shifts
00:37:45preglowit's just a mul
00:37:57stevenmI guess that works then.. I'll get on it
00:37:57preglowyou precalc the pan factor, of course, since it doesn't usually change that often
00:38:19preglowdon't worry about it, just do it the way you think it has to be done
00:38:25preglowif it can be done better, it'll be fixed some day
00:39:04stevenmwell I guess I can precalculate it as a large number, multiply, then shift. its a bit better than what I was gonna do, but it is a simple enough operation anyway
00:40:09stevenmnow... I was thinking.. would it be possible to have the synth running "in the background" ?
00:40:28stevenmLike, the reason it lags on a PC is that the PC wont execute any more instructions until the sound buffer frees up
00:41:04stevenmbut if we can make rockbox just kinda let it keep syntehsizing data (possibly into a separate buffer, and then copy that into the sound buffer as needed) then that would be ideal
00:41:25stevenmbecause the workload varies a lot as you go through the file
00:41:42amiconnHCl: Lcd modes seem to work just fine (tested with win32 recorder simulator). Now I need to add the switch button....
00:41:44preglowit'll have to run in the general framework
00:42:00preglowstevenm: you need a smaller buffer
00:42:05preglowstevenm: or several smaller buffers
00:42:57amiconnHCl: It now looks veerryy squeezed ;-/
00:43:21stevenmpreglow, how small would it be?
00:43:58stevenmI mean, as long as synthesis can keep up with playback, it would be good
00:45:53preglowstevenm: don't know
00:47:40stevenmaw great. looks like my burner is toast
00:49:23 Join Patr3ck [0] (
01:01:09stevenmyea, it's dead. wonderful
01:02:42preglowlet's hope it was an expensive plextor
01:02:45amiconnHCl: Switching modes with the ON button works perfectly on the recorder
01:03:48stevenmpreglow, nah, not too expensive, but it had to be replaced before. Under warrantym but I paid like, half the price of a new unit in shipping charges. It was glorious.
01:04:08preglowstevenm: they should cover that for you if it was under warranty
01:04:19***Saving seen data "./dancer.seen"
01:05:02Ricki'm a little stumped
01:05:08Rickhow is rockboy supposed to be built?
01:06:03stevenmpreglow, nope.. refused to. called them 7 times or so, racked up enough cell phone charges to pay for the other half
01:06:20stevenmpreglow, and now the new one dies 2 months later. it's wonderful
01:09:32 Quit lolo-laptop ("Client exiting")
01:10:33preglowstevenm: all good, yes
01:13:18 Join Sando [0] (
01:13:36preglowstevenm: which manufacterer is this, btw?
01:13:45 Join XShocK [0] (
01:14:00XShocKare here anyone from US?
01:16:39RickI am
01:17:33XShocKdo you know any big store that sells electronic components? digital, analog ICs, diodes.....
01:17:56*Rick has no idea
01:18:17XShocK"frys" ? what is that? :)
01:19:07XShocKi know only radioshack.. :(
01:19:22Rickmight be a socal thing
01:20:31XShocKthanks.. look at frys now too. :)
01:22:02 Join toolman [0] (
01:22:23 Quit toolman (Client Quit)
01:24:11XShocKmm.. couldn't find one near me.
01:24:44 Part rasher
01:25:01XShocKi need such a small thingy.. :)
01:25:53Rickwhat are you looking to get?
01:26:32stevenmpreglow, this would be I/O Magic
01:27:20XShocKi need a crystal quartz resonator for 32.768khz.
01:27:37coobhaha you wont get one at frys
01:27:40XShocKand a couple of digital ICs.
01:28:11XShocKi figured out that they have also very basic stuff like in radioshack.
01:28:48stevenmXShocK, they used to have a whole WALL of chips. I asked my parents to bring me back a certain chip (was livving in Russia at the time) so they just bought one of each
01:29:18XShocKhehe, i am from russia. :)
01:29:20stevenmXShocK, now they only have like 3 chips... resonators.. try thru their Commercial catalog. that still has some good stuff still
01:29:27stevenmXShocK, me too :)
01:29:44stevenmOkay.. panning works.
01:30:10XShocKso, can you recommend me any good store for such things?
01:30:50stevenmXShocK, all I can think of would be radioshack (commercial catalog, ask for it separately) and sites like Jameco and Digikey
01:31:28stevenmsometimes you get free samples. My friend got a sample of PIC16F628s... which we promptly installed in a bunch of xboxen
01:31:40XShocKso.. radioshack has things, but i need to ask a commercial catalog?
01:32:02stevenmit has like, tables upon tables of various things. all through the mail, though
01:32:02XShocKi got DS1340 chip sample from :)
01:32:26XShocKok... surely will request it.. :)
01:32:35 Quit toolmanwv (Read error: 110 (Connection timed out))
01:38:50XShocKok. gotta go. see you all
01:39:25 Quit XShocK (" HydraIRC -> <- IRC for those that like to be different")
01:42:58Rickwow @ rockboy ;P
01:43:58Rickhmm, weird font issue
01:44:01Rickfor the game
01:44:07Ricktop/bottom gets clipped
01:45:36RickHCl: suggestion, screenshot function ;P
01:46:59 Quit stevenm ("Leaving")
01:47:59preglow16 pixels are clipped at all times
01:48:04preglowthe iriver screen isn't high enough
01:49:01preglowhold button lets you control it
01:49:19preglowahh, you havent got h1x0?
01:49:24 Join stevenm [0] (
01:49:24RickI do
01:49:30Rickbut I don't know what you mean by 'hold button' :P
01:49:30preglowyes, then toggle hold switch
01:49:41Rickthat thing
01:50:01Rickahh, nifty
01:50:31preglowhe just made a mode where you can skip lines every ninth line of the picture, so everything fits
01:50:37preglowbut i guess it'll look like shit, though
01:50:50Rickit looked fine
01:50:53Rickit's just it warps text a little
01:51:38*Rick is trying Final Fantasy Legend III on it ;P
01:51:43preglowgood choice of game
01:51:52RickI lost my copy ages ago
01:52:01preglowi never played the gameboy final fantasys much
01:52:07preglowi've played all the other ones, though
01:52:25Rickrockboy does pretty good without any grayscaling yet
01:52:42preglowmy rockboy has graycale :)
01:53:10Rickyou're not on h1x0 I assume
01:53:14preglowi am
01:53:27preglowhcl posted a grayscale version some hours ago
01:54:12Rickodd then
01:54:30Rickbecause I thought I updated cvs not so long ago
01:54:31*Rick checks
01:54:46preglowno cvs
01:54:54preglowhe posted binaries, it uses a patch that's not in cvs
01:55:10RickI see
01:55:17Rickdoes it require a special rockbox version as well?
01:55:25preglowit contains everything you need
01:55:30RickURL? :)
01:55:32*Rick wants to try
01:55:50preglowsearch irc log for today
01:55:53preglowlook for
01:55:59preglowhe mentions grayscale just before
01:56:05Rickfound it
01:56:08Rickaround 2pm
01:56:14preglowsounds reasonable
01:56:17Rick[01:56:20 PM] <HCl>
01:56:22preglowthat'd be it
01:56:27*Rick pulls it
02:02:18preglowhmm what
02:02:20Rickseems like it's grayscaling the wrong areas
02:02:32preglowyes, some times it is
02:02:57Rickif I had my camera somewhere I would take a shot
02:03:04Rickthat is also missing
02:03:09preglowwhat game?
02:03:12Rickfinal fantasy legends III
02:03:18Rickthe water is supposed to be grayscale, not everything else
02:03:21Rickin the intro of new game
02:03:37preglowi've got problems in zelda
02:03:43pregloweverything is fine apart from the sprites
02:03:51RickI havn't tried zelda yet
02:04:02stevenmcheck the sprite palette bits ?
02:04:13preglowstevenm: he ripped out all palette stuff :7
02:04:33stevenmhow come ?
02:04:47preglowit'll be fixed
02:04:54stevenmah ok
02:04:56preglowi didn't even know gameboy had palette stuf
02:05:01preglowbut apparently it has
02:05:09preglowi managed to toy around with it in no$gbm
02:05:13stevenmgameboy palettes are annoying
02:05:20stevenmyea no$gmb really helps debug that crap
02:05:40stevenmnintendo really tried to hack around with compatibility of gb/gbc too. doesn't make life any easier :(
02:05:42Rickyou know
02:05:46Ricknow that I think about it
02:05:58Rickit's because the contrast is too high that's making it look grayscale in the normal areas
02:06:05RickI think the grayscale is simply too light
02:06:12Rickto be visible
02:06:21preglowRick: the colours are hard coded in the lcd, i think
02:06:26preglowRick: mostly i think it looked good
02:06:36Rickone sec
02:06:48Ricklemme take a snapshot of what the FFL3 intro is supposed to look like
02:07:03stevenmFFL3? It's all about FFA :-P
02:07:12preglowisn't that seiken densetsu 1 ?
02:07:21preglowi love that series
02:08:44stevenmooo that's cool
02:09:01Rickcompared to what you get in grayscalerockbox
02:09:07Rickthe water is wrong
02:09:24stevenmwhat about when the game wants to change the palettes? ie, fade in/out, etc
02:09:38preglowstevenm: is it capable of altering the lcd palette?
02:09:40Rickactually, the dialog box is wrong too
02:09:50stevenmor do any other crazy LCD stuff.. can you actually give the LCD an arbitrary shade of gray?
02:09:50preglowif so, we may have a problem
02:09:58preglowstevenm: that's what i don't know
02:10:01Rickstevenm: I don't know
02:10:11Rickhave to bug HCl and figure out how he does it
02:10:12preglowstevenm: i certainly hope so, and it does make sense
02:10:22preglowRick: no, that's a lcd driver issue, bug linus :P
02:10:34RickI figure HCl modified it
02:10:39stevenmpreglow, I don't think they do THAT much, but I think I remember FFA having a fade sequence.. or maybe they just rotate palettes. Lemme fire up no$gmb and see
02:10:42preglowmarkun did that
02:11:03Rickbecause everything renders fine except water (it's not filled with gray areas) with normal rockbox
02:11:51 Quit Patr3ck ()
02:12:25stevenmpreglow, playing FFA right now. Palette entries definitely change
02:12:36stevenmNow trying to figure out of they are just the same shades of gray in different order, or not
02:13:01stevenmlike.. they swap. but the emulator lets you define a color for each 'shade'
02:13:49stevenmyea. They achieve a fading effect by taking the same 4 shades of gray and assigning them to each color in the palette
02:14:51preglowi've had a look at the data sheet
02:14:57preglowlooks it's capable of setting palette
02:15:32stevenmas far as I can tell, they simply change which shade of gray is assigned to which palette entry
02:16:10stevenmif changing the actual shades depends on the vbk reg, then that didn't come around til GBC
02:17:22stevenmi dont think this is a problem
02:17:58preglowit supports crude palette changes
02:18:43stevenmi guess if they want to fade the game in and out, instead of changing the tilemap or the individual tiles, they just change which of the 4 shades of gray is assigned to which 'color' in the tiles
02:18:55stevenmfaster that way and saves space
02:20:48preglowyes, white and black is hard wired, the two levels in between are user definable
02:21:40preglowwith three bits, it seems
02:22:16stevenmthere's also for gbc, the lower 3 bits of the sprite info is the palette
02:22:30preglowrange is 0x1-0x7
02:22:57Suckahmm, what to rom file extensions have to be to be recognised?
02:23:15Suckahmm, it cant find the file
02:23:30preglowgo to general settings
02:23:35Bonkersset to show supported files?
02:23:36preglowand then file view and .gbc
02:23:39preglowset to all files
02:23:43Ricksupported files
02:23:44Ricklike I did
02:23:45DBUGEnqueued KICK Rick
02:23:45Suckaah ok thanks
02:23:57preglowsupported should work as well, yes
02:24:19Bonkersif supported doesn't show it, thne rockboy isn't installed correctly
02:24:39stevenmman gameboy stuff is friggin weird. there sprites, and then there's this window option too
02:24:46Suckaawesome :D
02:25:51Suckathis is so cool XD
02:27:03preglowstevenm: window has palette as well
02:27:31stevenmpreglow, yea, they're similar to the sprite ones... one of the colors is transparent
02:27:48preglowno$ had four palette entries
02:28:01Suckapreglow: your windows patcher was very nice :)
02:28:18preglowSucka: haha, ok
02:28:25preglowSucka: it works, at least
02:30:29preglowbut yes, don't spread yet
02:30:51Rickyeah it looks like palette needs to be reimplemented
02:30:57Ricksome games use it "backwards"
02:31:02Rickthan how rockboy currently does it
02:32:19Rickor... it's too light ;P
02:32:56 Join ashridah [0] (
02:52:46RickI04:IllInstr at 30025A4C
02:52:48Rickfrom rockboy
02:54:44preglowtry to reproduce it
02:55:36Ricklemme start that rom
02:55:37 Quit Sucka ("a bird in the bush is worth two in your house")
02:55:48 Join Sucka [0] (
02:55:54Rickdidn't do it
03:01:09 Nick Aison is now known as F1^Aison (
03:04:20***Saving seen data "./dancer.seen"
03:06:34 Quit preglow ("flam")
03:12:37 Quit Sucka ("a bird in the bush is worth two in your house")
03:15:52 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
03:58:00Rickwhat are overlays?
03:59:02Ricksomething related to grayscale
04:00:49Rickor not
04:03:23 Quit F1^Aison (Client Quit)
04:05:57 Join QT_ [0] (as@area51.users.madwifi)
04:17:29 Quit QT (Read error: 110 (Connection timed out))
04:20:51 Quit stevenm ("Leaving")
04:45:47 Quit cYmen ("zZz")
04:48:21 Join ze_ [0] (
04:49:21Rickscreendump needs to be modified so it doesn't save it as green
04:49:23Rickon iriver
04:54:56 Quit einhirn (Read error: 104 (Connection reset by peer))
05:01:45 Quit ze (Read error: 110 (Connection timed out))
05:01:45 Nick ze_ is now known as ze (
05:04:22***Saving seen data "./dancer.seen"
05:12:14 Join einhirn [0] (
05:16:53 Quit einhirn (Read error: 104 (Connection reset by peer))
05:22:10 Join beamm [0] (~haydenbea@
05:22:22beammhey all
05:23:13 Part beamm
05:50:26 Join ze_ [0] (
06:01:48 Quit ze (Read error: 110 (Connection timed out))
06:01:48 Nick ze_ is now known as ze (
06:05:58 Join stevenm [0] (
06:26:10Rickwhat's the difference between button_get(false) and button_status() ?
06:27:09Rickahhh, I think I understand, nevermind
06:36:48 Join DMJC [0] (
07:04:25***Saving seen data "./dancer.seen"
07:20:19 Join midk [0] (
07:21:42 Join methangas [0] (
07:25:03 Join ze_ [0] (
07:25:18 Quit ze (Nick collision from services.)
07:25:23 Nick ze_ is now known as ze (
07:36:58 Join wacky_ [0] (
07:37:37wacky_hey guys.. why don't you add my plugin to the CVS ? :) iriverify .. check the Wiki page PluginIriverify
07:37:39wacky_see you later :)
07:37:56 Quit wacky_ (Client Quit)
07:38:48 Join StrathAFK [0] (
07:39:02ashridahwasn't linus looking for him?
07:42:31DMJCfreaky... they ported portage to windows
07:44:03Rickfinally got it to compile without errors
07:44:05CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
07:44:05*Rick is porting a game
07:47:56Rickhehe, menu doesn't work, but that was expected
07:50:01 Quit Strath (Read error: 110 (Connection timed out))
07:54:40Rickyay :)
07:54:43Rickmain menu works
08:02:42 Join ze_ [0] (
08:07:33stevenmRick, whacha compiling
08:08:49amiconnmidk: Now I'm awake (sort of)
08:08:54Ricka game port
08:10:03midkamiconn: i was wondering about a recent commit; what's "lcd switching for recorders"?
08:10:22Rickmidk: what file is it for?
08:10:24midkthough, i guess i've sort of figured it out.
08:10:24Rickprobably for rockbox
08:10:42Rickit allows you to switch the drawing mode
08:10:46Rickso it fits the screen
08:10:51Rick(and others)
08:10:56amiconnmidk: switching the various lcd modes (dropping of lines to make it fit) in rockboy
08:11:31midkhmm.. rockboy looks really cool.. congrats & awesome job to all who worked on it :)
08:11:55midkhow're games played? .rom's?
08:13:01 Join ze__ [0] (
08:13:43Rickyou select a .gb[c] file in the file viewer
08:20:34 Quit Bonkers ("Client exiting")
08:21:29 Quit ze (Read error: 110 (Connection timed out))
08:21:29 Nick ze__ is now known as ze (
08:24:13 Quit ze_ (Read error: 110 (Connection timed out))
08:24:43*Rick kicks his iriver
08:24:50Rickmount already :/
08:24:56 Quit YouCeyE (Remote closed the connection)
08:30:59Ricka nice reboot did the trick
08:31:29 Join YouCeyE [0] (foobar@youceye.user)
08:45:10 Join lostlogic [0] (
08:46:14Ricksomething is seriously fucked with my usb
08:46:44Rickand it's making updating my project really irratating
08:48:33 Join jyp [0] (
08:53:46Rickokay, I don't think it's my linux at all now
08:54:00RickI tried plugging it in on windows and it didn't automount like it usually does
08:54:07Rickwhen I shut the h120 down and started it, it worked
08:54:09Rickodd :<
08:54:47 Quit lostlogic ("Going to the moon")
09:04:29***Saving seen data "./dancer.seen"
09:12:15Rickgoing good again ;)
09:23:37 Nick StrathAFK is now known as Strath (
09:35:30 Join ze_ [0] (
09:40:44HClpain sucks
09:43:41Rick'sup HCL
09:43:45Ricker, HCl
09:43:58HClaggrivated over pain actually worsening when its supposed to get less
09:44:17HClit makes it hard to sleep or do anything for that matter
09:45:13HCland painkillers aren't helping at all
09:46:35Rickwork on something, that usually helps me
09:47:45 Nick kergoth is now known as kergoth`zzz (
09:52:29 Quit ze (Read error: 110 (Connection timed out))
09:52:29 Nick ze_ is now known as ze (
09:52:35Rickwhoo, this rules
09:52:49Rickshould have a 'beta' copy of this port ready within a couple hours
09:53:58HClwhat port?
09:54:39Ricki'm porting a breakout-esq game
09:54:45Rickthat I played so long ago
09:55:01HClah. nice
09:55:07HClthats on our todo list
09:55:19Rickhaving trouble with speed adjustment but i'm sure i'll clear that up later
10:00:12 Quit stevenm ("Leaving")
10:21:03 Join FiZZ [0] (
10:21:15 Quit FiZZ (Client Quit)
10:21:39 Nick QT_ is now known as QT (as@area51.users.madwifi)
10:21:57 Join FiZZ [0] (
10:44:49RickI feel stupid now
10:47:02Rickjust working on a project
10:47:07RickI misunderstood something
10:47:16FiZZi c
10:48:16 Join ghode|afk [0] (
10:48:57ghode|afkhi, does anyone know why i would get the error msg "unrecognised architecture specification '5249'. when i try to build rockbox?
10:49:25Rickghode|afk: you need cvs binutils
10:49:29Ricknot binutils 2.15
10:50:49ghode|afkwhere can i get this from? ><
10:51:12Rickthe link is in the building instructions
10:51:13 Quit midk ("Leaving")
10:51:33ghode|afkcvs -z 9 -d co binutils ?
10:51:56ashridahdagnabit. didn't someone fix the docs to reflect the need for cvs rockbox?
10:52:08ashridahghode|afk: yeah.
10:52:16 Join Patr3ck [0] (
10:52:17Rickyeah, Linus did
10:52:19ashridahget that, then create an empty directory, and run the base configure
10:52:29Rickyeah, you'll need to rebuild gcc too
10:52:44Rickor at least I think so
10:52:49ghode|afkmeh should have asked thislast night. thanks for the help
10:53:02Rickhehe, no problems ghode|afk
10:53:08ashridahRick: i'm not sure you do.
10:53:15ashridahat least, i don't think *i* did
10:53:16Rickashridah: dunno, I did to make sure
10:53:27Rickashridah: do you know anything about the usb screendump ?
10:53:31ashridahi could be mistaken, it's been a while since i built the lot.
10:53:39ashridahRick: uh. no?
10:53:45ashridahwhat usb screendump
10:53:51Rickit's a debug option
10:54:07Rickyou can enable it, then when you plug in usb, instead of doing the hd thing
10:54:10Rickit takes a screendump
10:54:14Rickbut it doesn't seem to work while running a plugin
10:54:34ashridahah. no, sorry, don't know about it.
10:54:41Rickno problem
11:01:22 Quit ghode|afk ()
11:04:33***Saving seen data "./dancer.seen"
11:09:05amiconnRick: Screendump does work in a plugin *if* the plugin event loop is done properly, i.e. uses the default event handler
11:09:57 Join fizzle [0] (
11:10:10 Join ze_ [0] (
11:10:15Rickwasn't aware there was an event handler
11:10:19*Rick will look into that now
11:10:21HClyou probably forget to yield
11:10:45amiconnRick: Most other plugins are good examples of how to use it
11:10:53amiconnyield() is important too
11:11:22amiconnHCl: Rockboy does probably not handle usb?
11:11:51Rickonly a couple plugins yield
11:11:55HClas far as i know someone took screenshots of it with that..
11:12:28amiconnRick: Almost all plugins yield, but sometimes it is hidden within other function calls
11:12:44RickI just did a simple grep yield *
11:12:49Rickin the plugins dir
11:13:03amiconne.g. rb->button_get(true), rb->button_get_w_tmo() and rb->sleep() imply a yield()
11:13:19amiconnbeware that rb->button_get(false) does not
11:13:57*Rick nods
11:14:02RickI think I found a good place to stick yield
11:14:30Ricktwo places, actually
11:15:18 Quit FiZZ (Read error: 104 (Connection reset by peer))
11:18:24 Join ze__ [0] (
11:18:28Rickyep, screendump works now
11:18:34Rickthanks for the tip, again ;)
11:19:36Rickidentify the game and win... nothing I suppose ;P
11:20:19fizzleqnx ?
11:20:45Ricknever heard of it, so no
11:20:49 Quit ze (Read error: 110 (Connection timed out))
11:20:49 Nick ze__ is now known as ze (
11:20:54fizzlejebus staring at those screenies makes me grave vor Metroid II again ;)
11:21:03Rickone sec
11:21:07Ricki have something else then
11:21:15*fizzle smirkes
11:21:20amiconnHmm. (1) I agree, the screenshot colour needs adjustment for iriver. (2) Is it written in a way that allows easy adaption to a different display resolution?
11:21:24 Join ghode|afk [0] (
11:21:41fizzlewhats the res of the H1xx btw ?
11:22:03fizzlealmost half of the h3xx
11:23:24Rickoh damn you
11:23:32Rickh3xx is bigger and has color, :(
11:23:45Rickwhat's the range of color on the h3xx anyway?
11:23:50fizzle18 bits
11:24:08fizzleclaimed in the friendly manual, but I guess 16 bits, really
11:24:38fizzlebtw how bad was the /.ing of the rockbox site ? :) hehe
11:25:22Rickonly problem is you can't really play properly
11:25:24*fizzle curses his H320
11:25:24Rickdue to key limitations
11:25:32fizzlemetroid ? or any game ?
11:25:38Rickyou can't press two keys at once
11:25:41Rickexcept + play
11:25:46Rickand no diagonal
11:25:52fizzleyeah I read that in the logs
11:26:00Rickmakes metroid hard
11:26:06fizzleimpossible eve
11:26:20fizzlecant get down those rocks at the begin
11:26:29Rickyeah, I tried
11:26:33Rickjumping and shooting down
11:26:34Rickcouldnt do it
11:26:43Rickcause it requires holding two keys
11:26:48fizzlethere a way to tweak that up software-wise ?
11:26:55Ricknot yet anyway
11:27:13Rickit would be neat if rockboy had a 'sticky key' feature
11:27:19Rickwhere it "queues" a key
11:27:25Rickthen does both on the next key pressed
11:27:36Rickmaybe hold down play to do it
11:27:56ashridahheh. at least we've got more keys than an ipod's got available
11:28:06Ricknever had an ipod, nor do I want one
11:28:26Rickfizzle: what's the battery like on the h3xx?
11:28:30amiconnRick: did you get my remarks/ question?
11:28:36Rickamiconn: no?
11:28:48amiconn[11:21:19] <amiconn> Hmm. (1) I agree, the screenshot colour needs adjustment for iriver. (2) Is it written in a way that allows easy adaption to a different display resolution?
11:29:02amiconn(referring to your game)
11:29:05Rick1 - yeah, I found the bit that does it, should be easy
11:29:11Rick2 - yes... "hopefully"
11:29:15Ricki've been writing it with that in mind
11:29:17 Quit ze_ (Read error: 110 (Connection timed out))
11:29:20amiconnOf course a di
11:29:23fizzleRick: bat lasts like 14h without too much display activity
11:29:39amiconn+fferent resolution means redesigning the tiles
11:29:42RickI was thinking of dropping manually drawn stuff
11:29:45Rickand going for tiles
11:29:52Rickbecause it's getting really annoying to draw it all manually
11:29:57Rickwe need some sort of vector library
11:30:16Rickwhere it can scale vector bitmaps or something
11:31:00Rickwould be nice
11:31:03amiconnVector graphics are good for scaling, but the usually look ugly when scaled too low, i.e. only a few pixels
11:31:28RickI don't think it'd be much of a problem if the graphics are designed in a small resolution
11:31:35*HCl yawn
11:31:46 Join FiZZ [0] (
11:31:47ashridahRick: it'd be nice, but probably excessive on the archos platform
11:32:02FiZZincredibly stable line today...
11:32:17*ashridah hands fizz the cookie of line noise
11:32:43Rickdunno, I don't have an archos platform to test on
11:33:00FiZZty ashridah :)
11:34:41DeadManhow did Preglow get on with mpegdec? or did he go back to libmad?
11:35:00FiZZdonno, but there were some improvements claimed yesterday
11:35:34amiconnRick: you can build a simulator. The only thing that might be a probelm is speed, but I don't think that it is for this kind of game
11:35:37HClwe didn't really abandon libmad, we're gonna try both, as far as i know
11:36:14DeadManjust need an api to make it work right?
11:36:32amiconnRick: For testing speed with a slow cpu, you can set the iriver cpu speed to 11 MHz via the debug menu, then it will run even slightly slower than the archos
11:39:56Rickgood idea
11:41:19Rickwhat menu option?
11:41:22RickI don't see anything for hz
11:41:27FiZZpll stuff, prolly
11:42:00ashridahactually, iirc, it's the i/o stats or whatever. press up on the joystick while showing them
11:42:04ashridahor down, or something
11:42:10ashridahi'm not sure exactly how it works :)
11:42:51HClio ports, i think
11:42:53HCland press select
11:42:55HClthe top one
11:43:56 Join ze__ [0] (
11:45:11ghode|afkrick do a seacrh of logs, linus explains which press changes which freq.
11:46:17Rickany keywords to lead me by?
11:46:35FiZZpll maybe? *shrugs*
11:46:47ashridahhappy hunting
11:46:58ashridahit cant' have been more than a few days
11:47:46Rickif it was
11:47:52Rickit's not in my local logs under 'pll' at least
11:47:57 Quit fizzle (Read error: 110 (Connection timed out))
11:48:15ghode|afktry i/o :p
11:50:04 Join R3nTiL [0] (~zorroz@
11:50:22ashridahyeah, the actual hidden options were in the I/O info submenu iirc
11:50:52HClits the top option in the debug menu
11:50:57HClthen up/down/select
11:52:17amiconn , ~15:47
11:52:39amiconn(and replace 96 with 120 for current cvs)
11:52:45FiZZkeywords: frequency, clocking, downclocking :)
11:54:51Rickif I managed to do it
11:54:53Rickit plays just fine
11:55:52FiZZif ?
11:56:21 Quit ze (Read error: 110 (Connection timed out))
11:56:22 Nick ze__ is now known as ze (
11:57:51HClohh crap.
11:58:51amiconnHCl: Btw, lcd modes work just fine on archos. Added a mode switch button...
11:59:20HClgood :)
11:59:22HCli hoped they wolud
12:00:01HClcrap, i got a problem on my server
12:00:11HClbrb, hopefully it'll work after it rebooted
12:00:27 Quit HCl (Remote closed the connection)
12:04:39RickFiZZ: I don't really see any confirmation screen
12:05:01Patr3ckI have implemented manual profiling for the iriver / codecs, whom should I give the patch?
12:07:56 Join Aison [0] (
12:08:26ashridahthere's a patch submission system on the rockbox website apparently
12:08:34ashridahpost it there, possibly drop a note to the mailing list
12:08:43ashridahand prod locals
12:09:01 Join rasher [0] (
12:09:09Patr3ckwho is working on the codecs?
12:09:16*ashridah suspects it'll be handy for the Tremor codec
12:09:22ashridahPatr3ck: lots of people.
12:09:46Patr3ckok, I will do as suggested
12:10:18ashridahgod, i keep scaring myself when my code works right off the bat :/
12:10:21Patr3ckenglish not my native language, what is "prod locals"
12:10:54ashridahPatr3ck: oh, sorry. i mean bug people here, and stick around in case they want to ask you questions.
12:11:04 Join ghode1 [0] (
12:11:12ashridahthey might be asleep/away atm, but they drop in from time to time.
12:11:28FiZZPatr3ck: prod = anstossen, stupsen ;)
12:11:40Patr3ckashridah: thanks
12:11:43FiZZa bisserl nerven halt, auf gut deutsch ;)
12:11:57Patr3ckFiZZ: Das war mal verstšndlich ;-)
12:12:06FiZZyou're welcome
12:13:01 Join hcl [0] (
12:13:06 Nick hcl is now known as HCl (
12:13:06*ashridah tosses fizz the cookie of bilingualists
12:13:22ashridahhmm. actually, i'm not sure that's a word.
12:13:25ashridahit looks dodgy
12:14:30FiZZlooks ok to me
12:14:47FiZZbilangualism exists, so this makes sense too
12:15:12FiZZmaybe its just "bilingual" though, whatever
12:16:20 Quit [FlaT]Heidel (Read error: 110 (Connection timed out))
12:17:55HCli rebooted
12:17:57ashridahah, bilinguist.
12:17:59ashridahthat's what i'm after.
12:18:01HCland suddenly i have 5gb more diskspace.
12:18:07HClyou ashridah, explain :x
12:18:18ashridahHCl: what're you running?
12:18:33ashridahgot something that cleans out /tmp on startup?
12:18:45HClyes! thanks XD
12:18:59HClwhy did i have over 5gb of files in tmp o.o;;;
12:19:02*ashridah munches the cookie of the seasoned sysadmin
12:19:08*HCl grins
12:20:05*ashridah morosely watches cygwin chug away
12:20:18 Join muesli- [0] (
12:21:27*ashridah crosses fingers and tests his code on solaris
12:28:36ghode1does anyone know how big the binutils src dir should be?
12:29:09 Quit ghode|afk (Read error: 110 (Connection timed out))
12:29:16ashridahgive or take
12:32:41ghode1crap ><
12:33:21ghode156k sux :(
12:34:03ashridah-z9 compresses the cvs stream
12:34:06ashridahbut yeah, it takes a while
12:34:10*HCl pats 100mbit :3
12:34:25HClhere, you can have 1mbit from my connection, its not like i'll miss it :x
12:34:38*ashridah munches the cookie of jelousy
12:35:13ghode1lol :(
12:35:13 Quit muesli- (Read error: 104 (Connection reset by peer))
12:35:15ashridahghode1: you're looking at about 15-16MB of actual stuff to download, although i'm not sure how close cvs will come to that, should be close
12:35:39ghode1ok that explains it then
12:35:47ghode1now to get gcc ><
12:37:28ghode1i should get 3.3.5 right?
12:38:18HCl3.4 for iriver
12:44:17 Join Renko [0] (
12:52:47 Quit R3nTiL ()
12:57:00rasher*sniff* rockboy doesn't detect usb being plugged?
12:57:19HCli have no idea.
12:59:12rasherIt just continued when
12:59:17rasherI plugged the cable
13:00:01rasherhee, fun
13:00:20rasherrockbox can't shut down if the cable is plugged but a plugin failed to accept that
13:00:58rasherhangs at "shutting down"
13:04:34***Saving seen data "./dancer.seen"
13:09:53 Join sox [0] (
13:10:15soxhoy, could anyone tell me what this error is about:
13:10:17soxMAKE in libmad
13:10:17soxMakefile:21: *** missing separator. Stop.
13:10:17soxmake[2]: *** [libmad] Error 2
13:10:17DBUGEnqueued KICK sox
13:10:17soxmake[1]: *** [build-codecs] Error 2
13:10:19soxmake: *** [all] Error 2
13:10:21soxc-ae3fe255:~/rockbox/rockbox-devel/build svante$
13:10:31 Quit Rick (Read error: 60 (Operation timed out))
13:12:06ashridahis there a line or two above the one about *** missing separator. Stop. ?
13:12:32soxmake[2]: Entering directory `/Users/svante/rockbox/rockbox-devel/apps/codecs'
13:12:52soxso it's something in the makefile for libmad, no?
13:14:13CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
13:14:13*HCl watches the husband of his sister play mario on his iriver
13:14:37 Join cYmen [0] (
13:15:08ashridahsox: odd. i'm not getting that error here. this with a clean checkout from cvs?
13:15:56ashridahOUTPUT = libmad liba52 libFLAC libTremor libwavpack is the line in question here, i don't really see how make could complain about that.
13:16:07 Quit Patr3ck ()
13:16:20soxhrm, i have this problem with makefiles as you know, so no, i havent wiped them all since that means i have to go through them all and manually edit them to make it compile on my mac os x
13:16:49soxso you're right, something must've screwed up when the makefiles were updated from cvs and merged
13:17:01ashridahhrm. so a cvs up bitches that there's conflicts for that makefile?
13:17:22ashridahbecause it should have tried to merge in any changes.
13:17:22soxi guess i can just remove the makefile for libmad and run cvs update
13:17:34soxno, it didnt complain
13:18:10ashridahodd. don't know why it'd fail to compile if it's happy with the changes you made to the makefile and there's no conflicts.
13:19:38soxhm, if I remove the old makefiles and update from cvs i can compile... that's odd, what happened to my old pragma error?
13:19:43soxgone, it seems...
13:21:01ashridahthat particular makefile doesn't do anything with the preprocessor anyway, by the looks of it.
13:21:21soxyou're right
13:22:07soxok so I just needed to wipe the old Makefiles, now it works.
13:22:11soxsorry for wasting your time
13:23:23soxso whats this about: c-ae3fe255:~/rockbox/rockbox-devel/build svante$ make zip
13:23:23soxcp: copying multiple files, but last argument `./wv2wav.rock' is not a directory
13:23:51soxcan't open dir fonts at /Users/svante/rockbox/rockbox-devel/tools/ line 88.
13:24:54ashridahhave you checked out the fonts?
13:25:52hileOK, one more IHP-140 flashed and working ;)
13:27:44DMJChow's codec optimisation going?
13:27:45soxashridah: no, that solved it, except this: cp: copying multiple files, but last argument `./wv2wav.rock' is not a directory
13:28:51ashridahDMJC: apparently FLAC is running at >100% efficiency. have to ask preglow tho, he's the only one who's been doing it, afaik.
13:29:30ashridahsox: that one i don't know about. was the build directory clean when you started the build process?
13:29:48rasherlibmad does >100% for 128kbps mp3 files
13:29:54rasherat least, it did so once for me
13:33:28*rasher tries the viewer patch
13:35:43*DeadMan awaits proper MP3 playback to the headphone/audio out
13:35:53DeadManNot gonna flash until then :)
13:39:42 Join Schee4 [0] (
13:39:47ashridahyou could be waiting a while
13:43:30 Quit Aison ("( :: NoNameScript 3.72 :: )")
13:44:14HCland missing out on gameboy games allll that time #SB
13:44:16HCli mean :P
13:45:33soxashridah: yes, the build dir was clean
13:46:51ashridahi need more roms :(
13:47:07FiZZgb roms ?
13:47:16*HCl has pokemon silver, gold, yellow, and mario
13:47:20HCloh, and zelda
13:47:22HClbut zelda isn't playable
13:47:27FiZZu mean rom images or actual roms ? lol
13:47:35 Quit Schee (Read error: 60 (Operation timed out))
13:47:42ashridahnot with rockboy perhaps.
13:48:01FiZZbtw did you guys read the /. comments about rockboy ? there actually is a mp3 "plugin" for some gameboys :)
13:48:25 Join webguest10 [0] (
13:48:37ashridahand they sound horrible, apparently.
13:51:38rasherI appear to have a whole bunch of Pokemon roms now
13:52:12rasherblue, gold, red, silver, yellow.. also pinball, puzzle challenge and trading card game
13:52:57ashridahi wouldn't mind getting my hands on a metroid one
13:53:39*ghode1 has pokemon crystal \o.
13:53:52HClis crystal for gb?
13:53:55HCli'd like that one, if so
13:54:09HClcan you send it?
13:54:10ghode1er hold on ><
13:54:38rasherwhere are the buttons mapped?
13:54:42ghode1yeah gbc
13:54:59ghode1its 1meg ><
13:55:00 Join AC [0] (
13:55:24webguest10good morninng
13:56:50HClnm, got it :)
13:57:02HCl <- pokemon crystal
13:57:41 Join [FlaT]Heidel [0] (
13:57:45ghode1yey gcc done
14:00:20 Nick Strath is now known as StrathAFK (
14:01:28 Part DeadMan
14:03:33 Part webguest10
14:07:20HCldoes anyone have tetris for me?
14:07:49ghode1i have tetris dx
14:07:52ashridahdidn't someone mention a rom site to this irc channel yesterday or something?
14:10:01HCli put some on my ftp
14:11:12coobhha wario
14:12:37*ashridah leeches
14:14:02fuzzieHCl: possibly not a good idea to leave that unpassworded considering the public logs
14:14:24ACis someboday working on the remote lcd?
14:15:03ACif not, i want to try it
14:18:29HClfuzzie: i don't care, i have a 15gb upload limit per week :P
14:18:45HClof which i used 1.4gb so far
14:20:56rasherHrm, maybe a compressed version of the two dirs would make sense
14:21:01rasherunless roms don't compress well
14:21:26HCli'll just remove it once i hit my limit
14:33:28 Join DeadMan [0] (
14:33:29 Quit ghode1 (Read error: 110 (Connection timed out))
14:37:59*DeadMan installed a tool to get around that annoying problem of vanishing systray icons in win32 when explorer soft crashes
14:38:20DeadManI hate it when that happens
14:38:52HCldoesn't windows xp have that fixed..
14:40:02DeadManYou'd think they'd code that into systray by now
14:40:19DeadManMS programmers are lazy toerags
14:41:45DeadManPS Tray Factory is neat though. It can auto hide inactive icons and only show them when needed or you can hide what icons you like and restore all icons after a explorer crash. Alsi can mess with the transparency of the taskbar
14:41:55FiZZhm whats that tool ?
14:42:05FiZZah that it
14:42:12DeadManyeah it's cool
14:42:35DeadManGives you more taskbar space
14:47:31HClxp does that.
14:47:52HClmaybe you don't have xp prof..
14:49:49ACwho knows how to calcualte the correct value for GPIO1_OUT?
14:52:30FiZZi have xp. 2K. case closed. :)
14:52:45 Join einhirn [0] (
14:53:31DeadManXP only hides icons with that sliding thing
14:53:39DeadManthis hides them proper
14:53:50FiZZdeadman: care to link me up ? ;)
14:53:53DeadManbut it also recovers lost icons in an explorer crash
14:53:57HCloi, the horrid dumb bug in my dynarec
14:54:06 Join edx__ [0] (
14:54:50rasher15 lines of text on my screen.. yum..
14:55:02rasherand reflow is really nice
14:55:09rasherexcept I changed the indenting slightly
14:55:54 Quit einhirn (Client Quit)
14:56:00 Join einhirn [0] (
14:57:27rasherhm, no pillo
14:57:52 Join preglow [0] (
15:02:05preglowgood day
15:02:12 Quit Renko (Remote closed the connection)
15:02:21ACpreglow: do you know who GPIO1_OUT is calculated?
15:02:40CoCoLUSis the latest rockboy version on your ftp?
15:03:09preglowAC: i have no idea
15:03:30ACpreglow: ok.. than i must wait until Linus is here
15:03:51preglowAC: what GPIO1_OUT are you talking about?
15:04:37***Saving seen data "./dancer.seen"
15:04:39ACpreglow: in general.. i want to get remote lcd working
15:05:07ACi have wirtten some code, but now i need to set te corret GPIO1_OUT, GPIO1_,,,
15:05:08preglowthen you need an spi driver
15:05:54 Join redcow [0] (
15:06:14ACcan you give me some more detail?
15:06:42HCli crashed rockbox by plugging in the usb during loading of a plugin
15:06:49ACi thought i only must set GPIO1_... correct and can then write data to the lcd
15:06:54preglowAC: as far as i know, the remote lcd is connected to the cpu by spi
15:08:33AC CS LCD display SPI chip select GPIO34? LCD Controller
15:08:56preglowAC: do we even have specs for the remote cld?
15:09:47ACnot for the same model, but i think it could work
15:11:05 Quit edx (Read error: 110 (Connection timed out))
15:11:34 Nick edx__ is now known as edx (
15:13:38 Join LinusN [0] (
15:13:59AChi LinusN
15:14:17HClmorning linus
15:14:21FiZZLinusN: how much did you pay for that wiggler including the shipping to sweden ?
15:14:28LinusNAC: we have an spi driver, although it's in SH1 assembly
15:14:51LinusNFiZZ: hmmm, $200 in total, iirc
15:15:00ashridahLinusN: Patr3ck has supposedly gotten a rockbox profiling system up and running. i told him to send the patch to the patch pending system on the website, no idea if he's done it.
15:15:05FiZZwaz that in real money ? aka euros ?
15:15:09ACLinusN: so i must convert it and it should work?
15:15:10LinusNashridah: he did
15:15:25FiZZ$ is kinda screwy with the exchange rates atm :)
15:15:29LinusNFiZZ: us dollars is also real money
15:15:43FiZZnot for me in the ego-centric EU
15:15:53FiZZso thats like 150 euros
15:16:10FiZZhow much was the shipping ?
15:16:13LinusNthe P&E company is outside Boston, Ma
15:16:36LinusNthe wiggler is $149, the rest was shipping
15:16:36preglowAC: do you know if the remote lcd is connected to qspi pins on the coldfire?
15:16:46preglowor is qspi != spi ?
15:16:50FiZZthanks LinusN
15:16:51LinusNpreglow: look at the schematics
15:17:24preglowLinusN: i'm not doing it, i'm just trying to give clues :)
15:17:43LinusNi can tell you it isn't
15:17:53LinusNfor some reason
15:18:03LinusNso we have to bitbang it
15:18:05preglowAC: then you have a driver to port!
15:18:16LinusNspi is easy
15:21:40HCldynarec compiled 4 blocks
15:21:46HClthats 1 block more than i had planned.
15:22:20preglowHCl: i read coldfire docs yesterday, and stumbled upon the fact that the code cache has to be manually invalidated, are you aware of that?
15:22:31HClyes, its already doing that
15:23:05 Join webguest72 [0] (
15:23:40LinusNall caches have to be manually invalidated at boot
15:23:44webguest72Does anyone know where / how to download plugin's for rockbox, i cat find them on teh website - thanks
15:23:51LinusNon ever cpu i've encountered so far
15:24:27LinusNwebguest72: all plugins are in the .zip file
15:24:33preglowLinusN: bigger architectures invalidate cache manually on code change, don't they?
15:24:38webguest72k chears!
15:24:50LinusNpreglow: only if they have bus snooping
15:24:57preglowyes, obviously
15:25:00preglowthink x86 does it, at least
15:25:05LinusNand many platforms don't implement the bus snooping
15:25:17preglowwell, it's just as well not to waste silicon on it
15:26:02preglowbut silicon wasting is already so rampant on x86, it probably doesn't matter, heh
15:26:19rasherwell they *are* supporting a 20 year old instruction set
15:26:22redcowwhy is the actual rockbox slower than older release, cause of dynarec?
15:26:44preglowrasher: that's only extarnally! they're all translated to a risc set internally
15:27:19preglowand then you've got clever stuff like the bound instruction
15:27:23preglowenough about that, i want to sleep well
15:28:02*rasher stares at apps/misc.c
15:28:50HClthe good news is
15:28:57HClthe gameboy stack seems to be working properly in the dynarec
15:30:32HClthe bad news is, its getting so far that i'm starting to lose track of where it is in my testing gameboy code :x
15:31:47preglowHCl: did you read yesterdays log? gameboy has a palette, apparently
15:31:49*HCl found it
15:31:54HClpreglow: yes, i read it
15:31:59HCldidn't i talk about it? O.o;
15:32:05HClread the pluginrockboy page
15:32:06preglowyou might have
15:32:13HCli described the problem briefly
15:32:20preglowi was absent portions of yesterday night
15:32:20HCland where to start for anyone who wants to try to fix it
15:32:38HCli know i can't as long as i don't understand the *_scan functions
15:32:46preglowyes, i see so
15:33:05HCland its the least of my priorities, really :/ mostly cause its hard to fix
15:33:07preglowalso seems the lcd is capable of changing two out of the four colours
15:33:08preglowif you need that
15:33:16HClnot needed.
15:33:31HClits just the write routines to the gameboy palette need to be changed to work rather than emulate cgb
15:33:41HCland the sprite routines need to use the proper palette
15:33:41 Join DrRick [0] (
15:39:51 Quit coob ("leaving")
15:41:30HClthe coldfire doesn't have an inc/dec opcode?
15:41:48HClgee. okay.
15:43:01 Join Sucka [0] (
15:43:09 Quit redcow ("quork!")
15:54:48rashergee, this bmpheader in misc.c totally doesn't match any docs I can find
15:55:13ACLinusN: so i must port serial.c to coldfire?
15:55:24rasherI just want to change the colour of dumps on the iriver :>
15:55:31 Join Renko [0] (
15:59:30LinusNAC: serial.c???
15:59:49LinusNno, the SPI drivers are in firmware/drivers/lcd.S
16:00:05LinusNrasher: what is wrong with the bmp header?
16:00:26rasherHrm.. I don't think anything's wrong.. maybe I was just counting wrong
16:02:59rasheris there a define for checking if you're building on iriver? (to change the palette to blue colour)
16:03:13LinusNyou want to change the green color?
16:03:49LinusNchange 0x90, 0xee, 0x90 to for example 0xbb, 0xee, 0xf8
16:04:13rasher0xad, 0xd8, 0xe6 is what the sim uses
16:04:25LinusNuse that then
16:04:40LinusNmake sure you do it with an #ifdef
16:04:40rasherindeed, I was more thinking which define to look for
16:04:48LinusNor do you want me to do it?
16:05:15LinusNi'd use #ifdef IRIVER_H100
16:05:20ACLinusN: Ah.. ok.. now i see where :)
16:05:48LinusNAC: better look at adc.c
16:06:11LinusNeasily ported C version of a SPI-like protocol
16:07:09LinusNcall it firmware/drivers/remote-lcd.c or something like that
16:07:42ACi have choosen lcd-h100-remote.c
16:13:02 Quit ashridah ("sleep")
16:13:33*rasher wonders why his change isn't there..
16:13:42rashercompiling the right filewould help
16:16:01LinusNgotta go
16:16:03 Part LinusN
16:19:44 Quit webguest72 ("CGI:IRC")
16:19:45rasherhah, I did change the colour alright
16:19:49rasherbut it's a beige brown
16:21:35 Quit Sando (Read error: 60 (Operation timed out))
16:22:43 Join nozomiyum1 [0] (
16:23:17nozomiyum1haha - wanna see something funny from the mistic forums?
16:23:25nozomiyum1" is the rockbox for h3xx done yet - should only take a month to complete, right?"
16:24:32crwli don't think the quality of that forum is very high at all
16:24:45FiZZcrwl I second that
16:24:46crwlin fact it's maybe the worst i bother to read even sometime
16:24:56nozomiyum1yeah, it's a shame, because there's some good info
16:25:07nozomiyum1but there's a lot of pollution that ruins it at times -_-
16:25:16crwltoo much zealotry, too much false information, too much funny animated signatures etc
16:25:29FiZZthats the way forums are, these days
16:25:35crwlnot all are, luckily
16:25:49nozomiyum1yup - it's starting to become a pop culture of sorts =P
16:26:11crwlbut there *is* some good info at misticriver...
16:26:36nozomiyum1yeah, there's just a lotta crap to dig through to get to it -_-
16:27:01crwlyeah, that's sad
16:27:03preglownon-programmers easily underestimate the effort needed
16:27:06preglowno surprise
16:27:52 Nick nozomiyum1 is now known as help (
16:28:04 Nick help is now known as nozomiyum1 (
16:28:20FiZZjust dont bother ;)
16:28:30rasherpreglow: I forget, do you have commit access?
16:29:08crwli actually wish that the audio hardware forum in would be more active
16:30:25*HCl has a new stuffed animal :3
16:30:48 Nick nozomiyum1 is now known as nozomikiseki (
16:32:53preglowrasher: yes
16:32:58preglowHCl: what is it?
16:34:05HCla raccoon :3
16:34:55rasherpreglow: very simple patch to change the colour of screendumps from iriver to blue:">
16:35:02preglowhaha, cool
16:35:10preglowrasher: have you tested it?
16:35:27rasherI'm not sure what ther e is to test
16:36:08rasherbut it produces a blue dump
16:36:19rashernot sure what else I can do
16:37:33preglowok, commited
16:39:09HCli suck at math - if i have a reg BC and i need to decrement it, i need to check whether C is zero and if so, decrement B, and i need to decrement C, right?
16:39:44preglowb and c is lower and upper parts of bc?
16:39:54preglowother way around
16:40:11HClC is lower part of BC
16:40:45preglowwell, yes, it they're split internally as well, you have to
16:41:26preglowyou could just decrement c anyway, as it has to be done
16:41:31preglowas long as it is eight bits, at least
16:42:11 Quit gromit` (Read error: 104 (Connection reset by peer))
16:42:23preglowsubq.l #1, %creg \n\t jnc .nocarry \n\t subq.l #1, %breg
16:42:24preglowor something
16:42:45 Join gromit` [0] (~gromit`@
16:42:52HClthey're split internally, just wonder whether i got it right
16:43:11HCl DYNA_TST_b_r(3); // test C
16:43:11HCl DYNA_DUMMYBRANCH(2,0);
16:43:11HCl DYNA_DEC_l_r(2,0); // dec B
16:43:11DBUGEnqueued KICK HCl
16:43:11HCl DYNA_BCC_c(0x6,2,0); //jump here if not zero
16:43:11HCl DYNA_DEC_l_r(3,0); // dec C
16:44:04preglowi thought you said C was the lower part
16:44:07HClit is
16:44:16HClB = upper 8 bits c = lower 8 bits
16:44:20preglowthen surely that's the part that needs decrementing always
16:44:30HClyes, C is always decremented
16:44:40preglowand B only needs decrement if C wrapped
16:44:47HClyea, i have that
16:44:48HCli think.
16:45:04preglowwell, looks to me like you decrement b first and then conditionally decrement c...
16:45:47HClno, the BCC gets in the spot of DUMMYBRANCH
16:45:48preglowanywho, why the test?
16:46:04HClit tests whether C is zero or not
16:46:33preglowwhy do that? doesn't gbz80 decrement if it's zero?
16:46:50HClwell, if C is zero, it'll wrap if decremented
16:46:55HClso B needs to be decremented if C is zero
16:46:59preglowisn't that wanted behaviour?
16:47:08preglowwell, you need to check anyway,. tyes
16:47:10HClnever mind :X :X :X
16:47:10preglownevermind me
16:47:32preglowthe carry from decrementing c would be a better indicator
16:47:43preglowif it just wrapped, the carry will be set
16:59:53CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
16:59:53*AC gives up with the remote lcd..
17:01:29ACi hang at the moment at the spi stuff :(
17:03:40 Quit jyp ("poof!")
17:03:46DeadManAC NOOO!
17:03:59*DeadMan is looking forward to LCD remote working
17:04:39***Saving seen data "./dancer.seen"
17:04:46DeadManThe problem lies in what to display on it in terms of what is displaying on the player itself
17:05:10DeadManYou need a different display type to match what is on the player
17:05:17DeadMansmaller :)
17:05:17ACi know
17:05:31DeadMankinda tricky methinks
17:05:36ACthe first thing i want to get running is a blinking backlight
17:05:50DeadManyeah that would help lol
17:05:53DeadMansee what you are doing ;)
17:06:24*DeadMan got a free Big Mac today and now regrets eating it :P
17:06:42DeadManBig YACK!
17:07:43*AC found some nice infos about spi.. in german :)
17:08:07DeadManSprecken Zee Deutch?
17:08:19[FlaT]Heidelgerman is good :)
17:08:33DeadManGerman programmers are good
17:08:51ACjep.. i speak german..
17:09:09DeadManI know kung fu ;)
17:09:23[FlaT]Heidelshow me
17:09:41DeadMansappy movie Linus wandered off again
17:10:12DeadManwonder what he is working on
17:10:19hileI'm confused about plugins with iriver, where should they really be? .rockbox directory?
17:10:39preglowDeadMan: he's probably not working on something at all
17:10:45preglownot anything rockbox related, that is
17:10:56preglowhile: yes
17:11:14preglowhile: they're in .rockbox/rocks and .rockbox/viewers
17:11:50*DeadMan smacks preglow for destroying my illusion
17:12:10hileah, ok
17:14:21[FlaT]HeidelDeadMan: even linus might have a live besides rockbox
17:17:28ACpreglow: do you know if the remote lcd is conntected via uart or qspi?
17:19:07preglowAC: spi
17:19:25ACi have read the coldfire docs, but there is no spi
17:21:02preglowyes, there is, but the remote lcd isn't wired to it
17:21:07preglowyou have to write your own spi handling
17:22:05 Quit FiZZ (Read error: 110 (Connection timed out))
17:22:35ACis it wired directly to the cpu? So i can use the GPIO stuff?!
17:23:54ACok.. thats fine :)
17:24:10ACso i know now how i will code it...
17:24:29DeadManAC if you get LCD working you be big man on campus :)
17:25:04ACwe will see if i get it blinking
17:25:20DeadManAziz! LIGHTS!
17:26:45HClew, dynarec crashed x.x
17:26:47 Quit Renko (Remote closed the connection)
17:29:09ACsimpsons time :)
17:37:25 Quit DrRick (Read error: 104 (Connection reset by peer))
17:38:20 Quit einhirn (Read error: 104 (Connection reset by peer))
17:42:41 Join webguest57 [0] (
17:42:55 Join einhirn [0] (
17:46:23 Quit webguest57 (Client Quit)
17:48:49 Join Rick [0] (
17:50:15Rickmorning, or afternoon, or evening. :p
18:17:33 Join hjdfsat [0] (
18:18:14 Quit AC ("CGI:IRC")
18:18:47hjdfsatgood day
18:19:04hjdfsati heard about rockbox from the make magazine
18:24:11 Join Renko [0] (
18:29:11 Quit hjdfsat ("CGI:IRC (EOF)")
18:37:56 Join frank_ [0] (
18:58:32*HCl scratches his head
19:03:45 Join webguest89 [0] (
19:04:09*Rick waves at HCl
19:04:40***Saving seen data "./dancer.seen"
19:05:27 Part webguest89
19:09:09 Join jyp [0] (
19:15:49HClhow can i flush a file to disk?
19:15:55HClfdprintf does not seem to do this by default
19:31:03HCldoes it exist?
19:33:23 Join edx__ [0] (
19:34:22Rickhow do you mean by flush?
19:35:05Rickalso, ... I feel kinda stupid, but is there any way to allocate memory?
19:35:25Rickdon't see anything in plugin_api
19:35:39preglowRick: try to avoid it
19:35:45preglowRick: what do you need to allocate for?
19:35:58hileI don't know what functions we have from C-library, just a stupid guess ;)
19:36:11Rickpreglow: reading sprites into memory from file
19:36:36preglowi don't know if they've brought in the malloc yet
19:36:39preglowit wasn't used ebfore
19:37:41Rickwould rather not have to depend on a presized buffer :/
19:37:59hilebtw just flashed my iriver today, should I get any output (wps) when I try to play a flash on iriver by default? I'm just trying to see where we are and can I do something, and playing with it
19:38:13hileplay a wav, of course ;)
19:38:30Rickfrom what I know there isn't any audio playback yet
19:39:24hilewell, you can 'start playing' wav, no sound output but it shows the playing icon, most of the screen is just blank: I thought should there be something shown or not
19:41:06hilemaybe I'll just read the docs and code more... but have to say, even the current non-playing rockbox kicks ass as UI compared to iriver firmware, and for example the startup times or times coming back to use after USB usage is much better... I really can't understand what they have done @iriver ;)
19:41:29Rickpreglow: I guess i'll just do a data[LCD_WIDTH * LCD_HEIGHT] :p
19:41:43preglowRick: yes, just make sure it's a global
19:41:51preglowRick: you haven't got very much stack space
19:42:08Rickreading from the disk every time I want a sprite isn't going to be very good
19:42:38 Quit edx (Read error: 110 (Connection timed out))
19:42:39*Rick wants malloc :/
19:43:38Rickoh great
19:43:46RickI just realized I can't dynamically allocate sprites either
19:43:59Rickthis really sucks ;P
19:46:31 Quit Renko (Remote closed the connection)
19:47:31Rickrockboy uses malloc()
19:47:57 Quit edx__ (Read error: 60 (Operation timed out))
19:48:30Rickrockboy uses the mp3 buffer space for malloc
19:50:17preglowbagder mentioned bringing out malloc again
19:50:21preglowdon't know what came of it
19:50:25Rickdoes it / will it? work with audio playing?
19:51:49 Quit Sucka ("a bird in the bush is worth two in your house")
19:52:37Rickthere's plugin_get_buffer()
19:53:17RickI don't see a way to free buffers though
19:54:27 Nick StrathAFK is now known as Strath (
19:56:58 Join edx [0] (
20:05:23 Quit nozomikiseki (Read error: 104 (Connection reset by peer))
20:07:46RickI guess i'll just wait on my port until I get a more stable way of allocating memory
20:08:43preglowdo you do lots of reallocs?
20:12:30RickI am loading up sprites at startup (and that's it)
20:16:27amiconnRick: For plugin_get_buffer() and friends, you don't need to free it. It is implicitly freed when your plugin exits.
20:16:38RickOh, okay
20:16:41Rickthat's good news
20:17:16amiconn(That's no magic, plugin_get_buffer() simply gives you what's left from plugin ram (which is a fixed size) after loading your plugin)
20:17:55preglowRick: just make your own malloc and free that uses that
20:18:06amiconnTotal plugin ram is 768 KB on iriver atm, and 32 KB on archos
20:18:25preglowshould be really easy if you don't realloc
20:18:28*Rick caculates
20:19:04RickI should be fine RAM wise
20:19:14Rickwhat's the size of the archos lcd?
20:19:31amiconn1 bit (b&w)
20:19:35Ricki'm fine
20:20:32amiconnFullscreen buffer on archos needs 896 byte when bit-packed, and 7 kbyte with 1 byte/pixel (what a waste!)
20:21:36Rickdoes lcd_bitmap support bitpacked?
20:22:47 Join muesli- [0] (
20:24:53amiconnRick: lcd_bitmap in fact only supports bitpacked. The format (which happens to be the internal format of the lcd) is described in lcd-recorder.c / lcd-h100.c
20:25:51Rickcool, thanks
20:29:10 Join stevenm [0] (
20:29:30stevenmhello people
20:30:08Rickamiconn: does this format differ on player?
20:30:43amiconnThe archos player has a charcell based display. Very difficult to run this game on it ;)
20:30:58Rickoh, true ;P
20:31:04amiconnThe recorder and ondio displays are identical
20:31:51amiconnWell, I did write a player gfx library that allows you to do some bitmap graphics, but the max. resolution is 20x14 pixels
20:32:09amiconn(using the 8 user definable characters)
20:32:36Rickwell, does the recorder have the same format as the h100?
20:32:48 Join toolmanwv [0] (
20:33:03amiconnIt has the same format as the h100 in 1-bit (b&w) mode
20:33:13Rickwhich is what I need
20:33:47preglowhas the bootloader been tried on korean and us firmware as well?
20:34:44amiconnrasher: Yes... but I'd suggest to not depend on the format too much. I think the irver display will be switched to 2-bit (4 gray) mode soonish, and other platforms may be using just another format
20:35:05amiconnErr, that was directed to Rick
20:35:10*Rick nods
20:36:09amiconnFor instance, the gmini 120 and SP use the same format as the recorders (just 128x64 instead of 112x64), but the 220 is different
20:36:31amiconnBut then it may be hard to get plugins running on gmini at all
20:37:25 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
20:37:48 Quit edx (Read error: 110 (Connection timed out))
20:40:28RickI guess I should write a sprite format that will hold multiple formats
20:50:12stevenmpreglow, all right.. before I open up a hex editor.. did stripwax ever put up a link to where he found gus envelope info?
20:50:40preglowstevenm: no, don't think that
20:50:54preglowpreglow: but shouldn't timidity have source code for interpreting it?
20:51:51stevenmpreglow, yes..... I have looked at that as well.. and it does not really lend itself to comprehension
20:52:36*HCl yawns.
20:52:50stevenmpreglow, they convert it several times, and the only comments it gives are those of the developers saying they don't know what it means themselves
20:52:53HCldynarec is getting spammy
20:53:23amiconnHCl: Do I understand your last commit correctly? Dynarec is fully implemented, and runs?
20:53:40HClnoo o.o.
20:53:48HCli mean
20:53:51HClthe framework is implemented
20:53:52stevenmpreglow, the commenting is literally like, "what does THIS do?"
20:53:54HCland it runs partly
20:53:59HClbut more opcodes need implementing
20:54:02HCland it has some bugs.
20:54:05amiconnAh ok.
20:54:10HCllike the last block it compiles with my test rom it crashes on
20:54:17HCli added a proper debug system to it
20:55:07 Quit muesli- (Read error: 110 (Connection timed out))
20:56:23*HCl thinks he has a bug in the LDI opcode
20:56:39*HCl yawns
20:56:52amiconnDoes the dynarec core keep track of the emulated cycle count? That seems to be important for proper gb emulation
20:57:00HClyea, it does
20:57:06HClbut it works more in blocks, obviously
20:57:12HClso sometimes it breaks compatibility
20:57:47amiconnIt could update the cycle count on return from a block
20:57:51HClmostly, cause interrupts might happen at a later time than they actually should
20:58:05amiconnAh, the interrupts...
20:58:45amiconnHow much ram does dynarec "eat"?
20:58:51HCli have no idea.
20:58:54HCllet me look
20:59:33HClso far, it uses between 20 and 300 bytes for a block
21:01:30amiconnThe average relation gbz80 code size -> block size is more important imho
21:03:35HClwell, each block has some fixed size overhead
21:03:44HClit gets more efficient with more z80 opcodes per block
21:04:43***Saving seen data "./dancer.seen"
21:05:05amiconnThe overhead is the block return, right?
21:06:29 Join Tipi^ [0] (
21:07:09 Quit Tipi^ (Client Quit)
21:11:24HClyou can check out current cvs, and run it with a rom.. to see
21:11:31HClit writes several files to the /
21:11:37HClone describing whats in the block
21:11:42HClone having the m68k code for the block
21:11:48HCland one having the results of the execution of the block
21:13:13amiconnI can'T (yet), as I have no iriver...
21:13:25HCloh. darn, right :x
21:13:45amiconnI should start with SH1 dynarec soon...
21:14:05amiconnBtw, do you have an idea what the speedup will be like?
21:14:54HCli honestly can't say.
21:15:17HClat the moment i'm more interested in doing an interpreter assembly core
21:15:29HClbut i don't know yet how to do a switch() case: case: case: in asm
21:15:49amiconnJump table
21:17:07HClmhm, i need to read up on asm
21:17:19amiconnEven faster solution (not my idea, but from a zx spectrum emulator for the amiga): Make all instruction emulation blocks equally sized. This wastes space, but you can calculate the jump
21:17:29HCltechnically, it should be able to go faster than the interpreter asm core
21:17:39amiconnThat's very fast if the size chosen is a power of 2
21:17:45HClif the dynarec is designed efficiently
21:17:54HClat the moment, it has a lot of overhead
21:18:10HCldue to movem's before calls to memory functions, and movems to switch to and out of context
21:18:54RickHCl: did you see my sticky keys suggestion that I said many hours ago?
21:19:06amiconnHmm. The memory functions need to be part of the interpreter core to get rid of the additional movem's
21:19:29HClRick: yea, i added it on the todo list of rockboy
21:19:34Rickcool :)
21:19:50amiconns/interpreter core/emulation core/ , meaning either dynarec or interpreter
21:20:47amiconnDoes a movem instruction need more cpu cycles the more registers are stored/loaded ? I guess so...
21:21:03RickHCl: another idea for sticky keys is to make it keep doing the same keys for the second key pressed (as long as play is held)
21:21:27HClamiconn: most likely sol
21:22:19RickHCl: holding play, pressing up then left (does up+left), then afterwards pressing left will do up+left
21:22:25Rickfor as long as play is held
21:23:16*amiconn can do up+left / up+right
21:23:24Rickamiconn: eh?
21:23:29amiconnOn archos...
21:23:33RickOh :P
21:23:37 Join Tipi^ [0] (
21:23:45Rickamiconn: I mean on the iriver
21:24:03amiconnThe archos has a similar button handling as the iriver, but not all buttons are one group.
21:24:45amiconnYou cannot use more than one button of each group at the same time, but you can use buttons from different groups simultaneously
21:25:28 Join Camilo [0] (
21:25:38HCldoes anyone know whether the iriver remote and normal buttons are in different "groups" ?
21:25:51RickHC1: I would assume so
21:26:31amiconnArchos has 4 groups: (1) F1/F2/F3/Up (2) Left/Play/Right/Down (3) On (4) Off
21:26:51amiconnThe iriver main unit has 2 groups: (1) Play (2) all others
21:26:53 Nick kergoth`zzz is now known as kergoth (
21:26:59RickI wish the iriver was seperated more
21:27:08Rickwhy is play seperated anyway?
21:27:22preglowso it can be easily turned on, i guess
21:27:39preglowthe other buttons need an adc to be decoded
21:28:03amiconnThe iriver remote add 2 more groups: (1) Play (2) all others
21:28:38HCliriver remote play is another seperate group?
21:28:40HClthats great.
21:29:02Rickwell, right now play isn't usable
21:29:02amiconnThe Ondio has 3 groups: (1) Up/Down/Left/Right (2) Mode (3) OnOff
21:29:05Rickwell, nm
21:29:08Ricki'm silly
21:29:09HClthat means it will be possible to press a and b at the same time
21:29:20HCland up and left and up and right and down and left and down and right
21:29:21Ricka and b?
21:29:36Rickthe remote doesn't have left/right?
21:29:47HClit has those scroll things
21:29:51HClthat will do fine for up and down
21:29:53Rickyeah, but that's up/down
21:29:58HClthen you can use the base pad
21:30:00HClfor left and right
21:30:07HCland remote play for b
21:30:12HCland base play for a
21:30:13Rickthat'd be annoying
21:30:19HClbetter than nothing.
21:30:27 Quit cYmen ("leaving")
21:30:45Rickis the remote keys handled by the remote or the main unit?
21:30:48Rick*are the
21:30:53HClwhat do you mean?
21:31:00amiconnHCl: I guess you'll need to implement different keypad "modes". Perhaps someone doesn't have the remote....
21:31:07HClamiconn: yae
21:31:17HCldefinately, since some games don't require diagonals at all.
21:31:19Rickyeah, I don't have a remote
21:31:22Rickit's broken
21:31:23HCltetris and pokemon and mario work fine
21:31:33HCldon't need a remote for those
21:32:12Rickthat's why I suggested the sticky keys thing
21:32:14 Join edx [0] (
21:32:35Rickmake a/b select
21:32:44Rickand make a/b (hold for a time) bring up the menu
21:32:53Rickor something like that
21:32:59Rickwhat does select (iriver) do
21:33:15amiconnIf there's a way to automatically detect the presence of a remote, that'd be very user friendly...
21:33:27*HCl nods at ami
21:33:42amiconnThe archos also has a remote....
21:33:52rasherI'd guess there's a way.. since the main unit can tell if the remote has lock on
21:33:53Rickso play - 'sticky'
21:34:07amiconnHmm, probably not suited for playing games though. Way too slow
21:34:24preglowiriver remote rocks, actually
21:34:26preglowi should use it more
21:34:41Rickplay:sticky, stop:a, a/b: b, select:select, rec:start, hold down rect or a/b to bring up rockboy menu
21:34:43Ricksomething like that
21:37:07HClwe're probably gonna have some sort of button layout option
21:37:09 Quit rickst131 (Read error: 104 (Connection reset by peer))
21:37:22HClbut first i'm gonna focus on speed and bugs like that sprite palette bug
21:37:41HClits fast, but not for all games full speed yet.
21:37:51Rickyeah, I noticed some games are very horribly slow
21:37:57HClto be honest, that slightly worries me since that means the 30mhz speed increase didn't do much
21:38:08HClwhich would mean its not dependant on the cpu core speed but other stuff
21:38:31Camilohow is the screen refreshed?
21:38:32amiconnHCl: Plenty of roms on your site. Too bad I can't play all of them
21:38:46HClwhy not?
21:39:03amiconnWell, the 2MB and 4MB roms simply don't fit in memory
21:39:07HClohh. right.
21:39:11HCl :/
21:39:21amiconnI could do the 8MB mod...
21:39:30HClsounds good to me
21:39:54amiconnPerhaps I even will, if dynarec brings it up to somewhat playable speed...
21:40:11HClwell, i think you're gonna need a fair amount of ram for dynarec
21:40:29HCli also think that dynarec won't be especially fast untill the interpreter core is written as well and combined with it
21:40:35amiconnI need to employ some sort of lru cache
21:40:39Rickwhat is dynarec?
21:40:49Rickalso... isn't the mp3 buffer enough?
21:40:50HCleither that, or think of a way to make dynarec work for recompiling code
21:40:58HClRick: on archos, no.
21:41:02Rickoh, yeah :/
21:41:07amiconnAt least the simple roms (64 KB) shouldn't need that though
21:41:25HClno, all the simple roms don't even have writing stuff.
21:41:32HCli need to look at the memory map again...
21:41:37amiconnRick: The total ram on archos is 2 MB
21:42:00 Quit Camilo ("Chatzilla 0.9.67 [Mozilla rv:1.8a6/20050111]")
21:42:32amiconnFrom that, rockbox already takes away a bit (even if the code is running directly from flash rom - rombox - which is what I use), and rockboy takes another ~410 KB
21:43:31HCli have actually fairly high hopes for dynarec
21:43:45HCli think gameboy roms will have barely any self modifying code
21:43:45Rickhow much disk space does the archos have?
21:44:05amiconnAs much as you like. Mine is upgraded to 80 GB
21:44:15amiconnIt uses standard 2,5" laptop drives
21:44:26amiconn120 GB just became available
21:44:31HClNone (32KByte ROM only)
21:44:31HClMBC1 (max 2MByte ROM and/or 32KByte RAM)
21:44:31HClMBC2 (max 256KByte ROM and 512x4 bits RAM)
21:44:31HClMBC3 (max 2MByte ROM and/or 32KByte RAM and Timer)
21:44:33HClHuC1 (MBC with Infrared Controller)
21:44:45HCli'm just kind of wondering how those bigger roms are arranged then
21:44:51HClmaybe my document is out of date
21:45:16RickI wonder how the TI calculator emulators are
21:45:18Rickthat'd be another nice port
21:45:27amiconnThe ondio does of course have less space, it's a flash player. 128 MB built-in, and an MMC slot
21:45:27Rickthen I don't have to port the game i'm writing >:P
21:46:02RickTI calculators have like 30k of ram
21:46:19 Quit stevenm (Read error: 110 (Connection timed out))
21:46:31HClthere's also MBC5
21:46:36HClso prolly also mbc4
21:46:45HClmbc5 supports a max rom size of 64mbit and ram of 1mbit
21:48:03HClthe good news is that gameboy roms are mostly ROM and barely have any ram
21:48:14HClso it simply can't have self modifying code much, if any
21:48:29amiconnMagic rom ;)
21:50:18 Join drumrboy [0] (
21:51:29drumrboyis anyone here?
21:51:57HCland not having to keep account of self-modifying code = nice
21:52:10drumrboyhey every1
21:52:16RickHCl: How are you doing grayscale?
21:58:59HClwith markun's grayscale patch
21:59:51Rickwhat kind of grayscale support is in it?
22:08:27 Join _Lucretia_ [0] (
22:08:54_Lucretia_i'm helping with getting the info for the xclef mt-500
22:08:59drumrboyis there any iriver h320 developmant going on?
22:09:22 Quit drumrboy ("Leaving")
22:10:16_Lucretia_I know that the Gemini has the same cpu TCC730, and according to Dawid, the S3CC410 is similar
22:10:22_Lucretia_but how similar?
22:10:37amiconnHCl: STill looking for that file flush?
22:10:45HClamiconn: sortof
22:11:01amiconnThere is fsync(), but it's not in the api
22:12:02amiconnYou need to add it with that PREFIX() stuff, as there is a special simulator version
22:12:09_Lucretia_is Dawid Ferenczy in here?
22:13:10 Join infamous1 [0] (
22:15:20 Join Bluechip [0] (
22:15:43 Quit jyp ("poof!")
22:17:20 Join jyp [0] (
22:21:09rasheranyone here good with tomcat/servlets ?
22:22:52*webmind runs
22:24:10rasherTell me about it...
22:25:30webmindrather not.. too painful
22:25:38webmindI try not to think about it
22:25:57webmindit runs.. that's all it needs to do
22:29:17*HCl stares at his recompiled code
22:29:21HClthis code makes no sense at all
22:30:14 Quit linuxstb ("Leaving")
22:30:27rasherwebmind: as soon as I'm done with this course, I plan to do the same.. sounds like a good plan
22:32:40HClah, bugs in my debug code.
22:32:54HClits always nasty when the debug code isn't working properly and you end up chasing phantom bugs
22:33:33 Quit infamous1 (Read error: 110 (Connection timed out))
22:33:45Rickhehe HCl
22:34:31Rickyou know a good game to be made?
22:34:38HClyes, its on the todolist
22:34:40HCltwo of them
22:34:42Rickthat one game where the balls bounce around and you fill up the area
22:34:54RickI think there's a gnome version, gbounce or something
22:34:55HClit was a rethorical question
22:35:10Rick-? :p
22:37:08webmindrasher, good plan
22:37:52*HCl bites software -.-
22:40:37HCli can't stand it when things stop working without any good reason
22:43:20 Join midk [0] (
22:44:07Rickthat's the game I was talking about
22:48:28*HCl bites his hacked wireless
22:49:03[FlaT]Heidelhmmm...there can't be much to bite into
22:49:03preglowi made the bloody mpegdec lib compile, btw
22:49:10preglowi'm making a viewer for it now
22:49:44Rick[FlaT]Heidel: eh?
22:50:15[FlaT]HeidelRick: doesn't matter
22:51:57 Quit HCl ("Lost terminal")
22:52:33 Join hcl [0] (
22:52:42*hcl BITES his hacked wireless connection >.<
22:52:47 Nick hcl is now known as HCl (
22:53:57amiconnHCl: Does it also have a bug ?
22:54:20HClno, the guy who owns it is just messing with itt
22:54:22HCland reception is awful
23:00:31*HCl bites gcc and make
23:02:52Rickfound my CD with the TI stuff
23:03:38Rickfirst backup cd I popped in too
23:04:47***Saving seen data "./dancer.seen"
23:09:27Stryke`lucky, whenever i look for something, its always in the last backup CD
23:12:13HCli think i got it.
23:12:24HClamiconn: could it be that when you change the makefile, it doesn't rebuild everything?
23:12:43 Part Bluechip
23:13:01amiconnMaybe. I think I got a bit better at making Makefiles, but still far away from perfect :-/
23:13:17HClkay, well, i think thats my problem
23:13:25HClit was trying to call dynarec when dynarec wasn't linked
23:13:33HCland it was linked with dynarec while it wasn't trying to call it
23:13:41HClbecause it compiled one file with the define, and the other without
23:14:01HClworks now
23:16:05 Quit Strath ("Client closed")
23:17:07HClthere's a big bug in my dynarec :(
23:22:36HClwell, not too big, but. mrf.
23:28:53preglowfirst attempt is bunk
23:29:50HClwhat happneed?
23:30:19preglowit just produces garbage
23:30:23HCl :/
23:30:31preglowand the xxx2wav timer goes mad
23:33:51preglowthis does indeed look promising
23:34:18preglow67% realtime with no opts whatsoever
23:34:29preglowcan't remember what libmad used to run at
23:35:20Bagderthat web page claimed that version played mp3 on a 90MHz coldfire...
23:35:37 Join michiel_ [0] (
23:35:47amiconnAre there coldfires with a data cache?
23:36:05Bagderpossibly they used a better coldfire, yes
23:36:13preglowi don't think they did
23:36:25preglowthe 5249 is one of the highest performing coldfires
23:36:28michiel_rasher: what's with tomcat/servlets?
23:36:32preglowand mind you, the sound i get is garbage
23:36:39Bagderok, then its just strange
23:36:54preglowthat's more than likely just an error on my part, though
23:38:46 Quit rasher ("CGI:IRC")
23:38:47preglowthe source code is a bloody mess
23:38:53 Join rasher [0] (
23:39:14 Join stevenm [0] (
23:39:39michiel_rasher: what's with tomcat/servlets?
23:40:34rasherI had a url-pattern that was acting up.. but I wanted welcome-file, it seems
23:41:14 Quit midk ("Leaving")
23:42:02michiel_those patterns can sometimes suck in new and unexpected ways ;-)
23:42:41HClladadi ladada
23:42:50*HCl wanders around with a bugswatter and swats another bug
23:42:58rasher<url-pattern>/</url-pattern> matches everything..
23:43:01rasherwhich is queer
23:44:09preglowif this isn't just me acting the fool again, i say lets dump this piece of shit
23:44:26michiel_Crystal Waters
23:45:27michiel_sometimes a bit of trial and error will bash it into place
23:45:45amiconnrasher: logo.c with player adaption committed.
23:48:53 Join Camilo [0] (
23:50:56Stryke`anyone know why ctrl+tab no longer switches tabs on firefox since i updated to 1.0.1
23:53:16preglowwell, ok, i got it making sound, still 67% realtime
23:53:29preglowdecode looks decent
23:55:44amiconnWhat was the problem?
23:55:55Bagderamiconn: the new cvs builds don't leave you waiting long ;-)
23:56:19amiconnNah, the build itself takes it's time anyway
23:56:33Bagderbut it checks for new changes every 60 seconds
23:56:48Bagderexcept for when a build is in progress
23:56:54 Quit michiel_ ("Leaving")
23:57:09amiconnBtw, why does the sim makefile try to make the common sim lib twice? Of course the second call doesn't do anything
23:57:27Bagderno reason
23:57:43amiconnAnd imho 'make install' should be independent from 'make all' the same way as 'make zip' is now
23:58:05Bagdersounds logical, yes

Previous day | Next day