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

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

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

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

#rockbox log for 2005-03-03

00:04:23 Join Sucka [0] (~NNSCRIPT@host81-156-214-74.range81-156.btcentralplus.com)
00:06:59amiconnHCl: Please read http://www.rockbox.org/docs/contributing.html No offense...
00:07:46 Quit ashridah ("Leaving")
00:09:09HCloh.
00:09:15HClwell, i don't follow the comments bit
00:09:22HClbut ok
00:09:38HCloh, and no tabs, ok :x
00:09:51preglowset expandtabs
00:09:59preglow:]
00:10:22[IDC]Dragoncan somebody call my nick, please (testing highlighting)
00:10:30lImbus[IDC]Dragon,
00:10:32lImbus[IDC]Dragon
00:10:36[IDC]Dragonno go
00:10:43[IDC]Dragonbut thanks
00:10:58 Join amiconn_ [0] (~jens@pD95D1DC6.dip.t-dialin.net)
00:10:59lImbusmhmm. I stopped fighting the same problem
00:11:39 Quit amiconn (Nick collision from services.)
00:11:39 Nick amiconn_ is now known as amiconn (~jens@pD95D1DC6.dip.t-dialin.net)
00:11:58amiconn[IDC]Dragon, lImbus: Highlighting in Hydra is not really complicated.
00:12:26[IDC]Dragonperhaps, but not working here
00:12:42amiconnIt's only complicated to explain that, wanting to look a the options dialog ant typing in at the same time.
00:12:53amiconnThe options dialog is unfortunately modal
00:13:08lImbusand neither he is very verbose...
00:13:14lImbuss/he/it
00:14:12amiconnIt is in fact very flexible
00:14:37[IDC]Dragonit's enogh for me if it triggers on IDC and Dragon
00:14:38amiconnI'll make some screenshot
00:14:44amiconn+s
00:14:55lImbusnice
00:15:02[IDC]Dragon;-)
00:15:55HClby the way
00:16:00HCli now get an mmc icon on my iriver
00:16:12[IDC]Dragonhaha
00:16:28[IDC]Dragoniriver has no LED?
00:16:36HClit does.
00:16:58[IDC]Dragonnot supported yet?
00:17:06*HCl wonders wth he's unable to check for hold...
00:18:20HCleh, the lcd is hard wired to the hdd
00:18:25HClyou couldn't disable it if you wanted to, i think
00:18:32HClerr
00:18:33HClledd
00:18:39HCltypo, i swear >..
00:21:06preglowhahaha
00:21:09preglowuseful lcd
00:21:23amiconn[IDC]Dragon, lImbus: There are 3 windows where something needs to be adjusted if you want proper notification.
00:21:25amiconnOne of them is for (global) highlighting of your nick/ realname.
00:21:53amiconnThe other 2 are for when you want some special people grouped and notified if they join/leave
00:22:27preglowhydrairc looks pretty good
00:22:49*HCl pets irssi
00:22:56*preglow hugs irssi
00:23:14pregloweverything that has a perl engine built in is bound to be good
00:23:43geoff_oLike dpkg?
00:23:56HClthis is really puzzling me.
00:23:57amiconn(1) In preferences/Buddy Groups, I created a group "Rockbox" for the "FreeNode" network. Then I added my special people list in Members. See amiconn.dyndns.org/buddy_group.jpg">http://amiconn.dyndns.org/buddy_group.jpg
00:23:59*HCl slaps gcc.
00:24:12HCli hate it when my code looks absolutely correct yet it doesn't work.
00:24:28preglowHCl: i've got one of those right now
00:25:36lImbusamiconn: I like that "host matching not yet implemented"
00:25:38lImbus.-)
00:25:49[IDC]Dragonamiconn, I have a connection problem
00:25:55[IDC]Dragonto your site
00:25:56amiconn(2) To get notified about those special peope, select the newly created group "FreeNode / Rockbox" in Buddy Groups / Notifications. Then add the desired events, and what Hydra should do. amiconn.dyndns.org/channel_notifications.jpg">http://amiconn.dyndns.org/channel_notifications.jpg
00:26:10[IDC]Dragonah, now
00:26:18amiconnYeah... just wait some mintes
00:26:29amiconnDyndns needs to pickup my recent IP change
00:26:48HClpreglow: the worst part is that the code to set my variable to 1 is getting executed, but simply not setting it to 1
00:26:56HClif i initialize the same variable to 1 from the start
00:26:57HClit works
00:27:00HClbut i can't set it to 0 tehn
00:27:01HClthen*
00:27:21 Join Counter89234 [0] (~MB@ool-43509614.dyn.optonline.net)
00:27:47lImbus[IDC]Dragon: take this :- ) http://217.93.29.198/buddy_group.jpg
00:27:56amiconn(3) For highlighting your nick /realname in all networks, select the group "All Networks / Default", then add the desired events. The match text uses regexp. See amiconn.dyndns.org/global_notifications.jpg">http://amiconn.dyndns.org/global_notifications.jpg
00:30:30 Quit Chamois (" HydraIRC -> http://www.hydrairc.com <- State of the art IRC")
00:31:34HClthis is just so odd.
00:31:37 Quit Counter89234 (Client Quit)
00:31:38*[IDC]Dragon configures
00:31:42*HCl slaps stuff :/
00:31:54*lImbus configures too. and feels good while doing so
00:34:15 Join XShocK [0] (~XShocK@pcp09492659pcs.nrockv01.md.comcast.net)
00:35:40pabsdoes the h100 not have a rtc or something?
00:35:44 Join Sando [0] (kekekek@CPE-147-10-21-132.vic.bigpond.net.au)
00:36:12[IDC]Dragonok, done, can somebody nick me please?
00:36:13lImbusamiconn: I found my problem. It's actually too flexible. I thought regular expressions would be an option. so I only wrote my name there...
00:36:34lImbus[IDC]Dragon of course. could you now call me ?
00:36:42 Join Camilo [0] (~chatzilla@userca029.dsl.pipex.com)
00:36:47[IDC]DragonlImbus, yeah!
00:36:52lImbusYAY
00:37:05*[IDC]Dragon needs to find a color
00:37:06lImbuscOlOrEd, even !
00:37:26HClpabs: no, it doesnt
00:37:47amiconnHCl: New driver is working on the recorder and in the win32 simulator. The x11 simulator behaves rather odd with the new driver...
00:37:58HClhm :/
00:38:26pabsHCl: :(
00:38:28HCli'm working on adding nice defines around the grayscale stuff
00:38:43lImbusamiconn: thanks for the assistance
00:38:51[IDC]Dragonyes, thanks
00:38:58 Quit Sucka ("a bird in the bush is worth two in your house")
00:39:31HClamiconn: shall i wait with giving you the grayscale stuff till some grayscale stuff is actually in the cvs?
00:39:57amiconnHCl: Well, the x11 simulator button handling itself is a bit odd. I have some code fragments to improve on this, but I need to get it running stable
00:40:12HCl*nods*
00:40:17amiconnCurrently it crashes after a while
00:40:32amiconnThe xlib doesn't like multithreading
00:41:10Camiloamiconn, calling X from more than one thread?
00:41:17amiconnyup
00:41:22Camilobad!
00:41:33amiconnMy first shot at improving the button handling did this
00:41:41preglowany 68k gurus here?
00:41:53HClwell, i'm not a guru at all. but ask :x
00:41:53[IDC]Dragontoo long ago
00:41:58preglowhaha
00:42:41preglow"movq.l #8, %d3 \n cmp.l %d3, %d2 \n jgt .mylabel" should jump to newlabel if d2 is larger then d3, yes?=
00:42:42 Quit methangas (" HydraIRC -> http://www.hydrairc.com <- The future of IRC")
00:44:21preglow'cause that somehow doesn't seem to happen :VV
00:45:05*[IDC]Dragon says goodnight!
00:45:10 Part [IDC]Dragon
00:46:09lImbusamiconn: I not just got the notification that J鰎g quit. yay, it works :-)
00:46:27lImbussad I could not give him a bed time kiss :-%
00:49:56amiconnPooh, huge commit done.
00:50:24*HCl waits for it to show up
00:50:32HClami, shall i send you the grayscale patch...?
00:51:17amiconnNo, not yet
00:51:45amiconnMaybe I should have a look at the Z80 flag generation
00:52:05HClokay
00:52:20amiconnI simply can't believe it'll take 14 instructions to emulate the Z80 flags for add/sub
00:52:21HClyou have the dynarec version right? i'm not sure if you have an updated one..
00:52:25HCleh..
00:52:26HCl30
00:52:28HClnot 14
00:52:29HCl :X
00:52:42amiconnIt must be possible in < 10
00:52:49HCli hope so
00:52:56 Quit Renko (Remote closed the connection)
00:52:58HClmaybe looking at the i386 asm core might provide clues
00:53:28amiconnWell, zero and carry are easy. Only the halfcarry needs more work
00:53:34HClyup.
00:53:44HClcarry is simply bit 8
00:53:48HClzero is zero
00:53:54HClhalfcarry is, well. igh :x
00:54:08amiconnI believe this can be done with some nice xor'ing
00:54:23HClmaybe.. the current algorithm already uses xor..
00:54:58amiconnIf I am correct (this is only from imagination), it may work as this:
00:56:56amiconn(when adding b into a): ((a+b) ^ a ^ b) & 0x10 Halfcarry should then just be bit 4
00:57:18HClit already does that :/
00:57:38HCl(FH & ((A ^ (n) ^ LB(acc)) << 1))
00:57:41HClis the current code
00:57:52amiconnReally? Why it uses so many instructions then?
00:58:06HCli'm not sure.
00:58:18HClit needs an additional bitshift and then the and you had earlier
00:58:46HCli don't have many calculate registers to spare either..
00:58:47amiconnYeah, the bitshift is probably to move the halfcarry to the correct position in the flags register
00:58:52HClyea, it is
00:59:10HClFH = 0x2
00:59:13HCl0x20
00:59:14HClmy bad
00:59:15amiconnYou only need one additional register (additional to the actual 2 that are added)
00:59:41HCleh.. i need a var to store the result in
00:59:50HCli need the sources, unchanged
00:59:53amiconnYes
00:59:55HCland i need the flags
01:00
01:00:04amiconnNot separately
01:00:04HClwhile only really having one register to spare
01:00:15Bonkerstime to test out rockboy
01:00:17HClwhat do you mean not seperately?
01:00:36amiconnYou have the flags assigned to one of the m68k registers, don't you?
01:00:39HClyes
01:00:42HCld7
01:00:49amiconnOkay.
01:01:08amiconnI can only talk sh1 asm here, but m68k should be similar
01:01:16HCl*nods*
01:02:08amiconnI assume A == r1, B == r2, flags = r7, r0 is my spare (you know, the special one)
01:02:20HClyup.
01:02:29HCli have d0 as my calc register too.
01:02:50amiconnmov r1,r0
01:02:50HCli've been thinking of storing two gameboy regs in 1 register
01:02:54HClusing the swap instruction
01:02:58amiconnadd r2,r1
01:03:08amiconnxor r1,r0
01:03:14amiconnxor r2,r0
01:03:19***Saving seen data "./dancer.seen"
01:03:24amiconnshll r0
01:03:31amiconnand #0x20,r0
01:03:37amiconnor r0,r7
01:03:42amiconn**done**
01:03:45HCluh
01:03:46HClno..
01:03:58 Part Camilo
01:04:02HClr7 has to be set to 0
01:04:03amiconnThe addition plus the halfcarry are done
01:04:06HClyea
01:04:29HClcarry and zero are fairly trivial from that point
01:04:43HClthe only problem is that with a compare, i mustn't change the value of r1 eventually
01:04:53amiconnBtw, just in case this differs from 68k: for sh asm the left register is source, the right one is dest
01:05:06stevenmwoohoo, midi drums work
01:05:10HClit doesn't, its the gcc syntax that defines that
01:05:12HClnice :)
01:05:15HClbut yea, you're right
01:06:23HClyea.
01:06:32HClbut next to that comes the zero check...
01:06:55amiconnThe zero check is very simple on sh....
01:07:33amiconnAh, no, it's not *that* simple if the value is not in r0... still not hard
01:08:07HCllet me try to rewrite it..
01:11:56HClyou're better at asm than me xD
01:12:55HCljust need to add carry and zero
01:13:52amiconnCarry is also simple. One way would be (note that the sh can only shift by 1, 2, 8 or 16 bits per instruction, iirc the 68k can do all n)
01:14:04preglowyes
01:14:06amiconnAfter doing the above for add & H
01:14:21amiconnmov r1,r0
01:14:27amiconnshlr2 r0
01:14:31amiconnshlr2 r0
01:14:39amiconnand #0x10,r0
01:14:43amiconnor r0,r7
01:14:51preglowahhh
01:15:09preglowgod, it's nice to find code backed up when you thought you hadn't backed it up
01:15:10amiconn(The two shlr2 can then be combined on 68k)
01:16:11amiconnI might be that a test/branch is faster here, I'd need to count cycles)
01:19:03Bonkersamiconn: rockboy works pretty well here on my recorder just so you know, except of course for the speed ;)
01:19:24HCl17
01:19:36HCl& H ?
01:19:36amiconnBonkers: I know how it works on a recorder, but thanks :)
01:19:47amiconnhalfcarry
01:20:14HClohh. right.
01:20:40HClyea, i did that.
01:20:43HCland i added zero too
01:20:48HCli got to 17 instructions now
01:20:51HClrather than 26
01:20:56amiconnNot bad
01:20:58HClyea
01:21:00amiconn:)
01:21:02HCland sub got smaller too
01:21:57HCllet me give you the final numbers after i altered it some more
01:23:11HClsub is 16 instructions, cmp is 18
01:23:30HClslashed it almost in half :)
01:23:32amiconnI suspected these to be similar
01:23:33HClsee, i suck at assembly xD
01:24:08HCllet me do add in the same way, it was too big too
01:25:39preglowarghhh
01:25:42preglowi want a simulator
01:26:26HClamiconn: are you still interested in making a assembly interpreter core? i'm somewhat getting the feeling that an interpreter core in asm is enough for iriver..
01:26:32HClarchos probably needs dynarec either way
01:27:37HCligh.
01:27:39*HCl forgets the time.
01:27:52HClgotta go sleep, bad dentist appointment tomorrow >.<;
01:28:47amiconnHmm, gotta go to sleep too
01:28:54amiconnNight
01:29:06preglownight
01:29:37 Part amiconn
01:31:50 Quit preglow ("a")
01:35:57 Part geoff_o ("Kopete 0.9.1 : http://kopete.kde.org")
01:37:39 Quit Aison ("( www.nnscript.de :: NoNameScript 3.72 :: www.XLhost.de )")
01:45:23 Join elinenbe [0] (elinenbe_@207-237-225-9.c3-0.nyr-ubr1.nyr.ny.cable.rcn.com)
01:45:42elinenbeHCl: does rockboy really compile and work on the archos?
01:48:54Bonkerselinenbe: I just ran it on my archos recorder
01:48:56lImbussomebody said it's slow, but works on the recs
01:48:56Bonkersa bit slow...
01:49:13Bonkersmaybe like 10-15% the original speed
01:49:16Bonkersby rough guess
01:53:08elinenbewhat is the speed on the iriver?
01:56:25Bonkersno idea
02:00
02:00:27 Quit mecraw ("Trillian (http://www.ceruleanstudios.com")
02:01:20HClnear full speed
02:02:20HClbut not with current cvs.
02:03:55HClmy last version has 2bit grayscale, cpu core in iram.. and working on different lcd modes thats malfunctioning at the moment, oh. and ofcourse the dynarec in progress
02:04:46 Quit DjMnG ()
02:06:24*HCl wants to see some pics of archos running rockboy
02:08:30 Quit rickst131 ("‹‹UPP››")
02:12:52Bonkersit's not terribly interesting
02:13:03HClinteresting enough, really
02:13:04BonkersI could tell enough of what was going on to run forward and jump on the first goomba and stuff
02:13:15HClmainly cause there are no pics of it as of yet
02:14:10Bonkerslets see what I can manage
02:18:02lImbusBonkers, you can let rockbox make screenshots for you
02:18:08Bonkersya, I'm doing that currently
02:18:14lImbuskk
02:18:29HClhm
02:18:36HCli more kind of meant real pictures :/
02:18:38HClbut ok..
02:19:31Bonkerswell I don't have a camera
02:19:36HCl*nods*
02:19:37HClits ok
02:19:44HCli need to go to sleep anyways
02:19:48Bonkersheh
02:19:53Bonkerswant the screens first?
02:19:57HClyea
02:20:08HClgot somewhere to put them? or use dcc..
02:20:25HClgotta remember to tell amiconn to commit the speed increase stuff
02:21:34Bonkersputting on a web page in a sec
02:23:37Bonkershttp://127.0.0.1/rockboy/
02:23:39Bonkerswhoops
02:23:57Bonkershttp://terrapin.shacknet.nu/rockboy/index.html
02:24:03HCl
02:24:05HCl :)
02:24:16HClahh.
02:24:17Bonkersfirst one worked from here ;)
02:24:19HClinteresting
02:24:46Bonkersinterestingly those thumbnails are actually blown up images...
02:26:05HClhehe P
02:26:05HCl :P
02:26:05HClblown up?
02:26:19HClwell
02:26:21HCli can't say its pretty
02:26:27HClbut. its something, at least
02:26:31Bonkersclick on the pictures to see the actual size
02:26:34HClyea
02:26:36HCli noticed
02:27:34HCli really gotta go sleep now
02:27:35HClbyebye
02:27:39HClthanks
02:27:39Bonkerscya
02:39:16 Join kergoth_ [0] (~kergoth@ip71.churchill-park.dfw.ygnition.net)
02:39:50 Nick kergoth_ is now known as kergoth (~kergoth@ip71.churchill-park.dfw.ygnition.net)
02:39:54 Quit kergoth (Nick collision from services.)
02:40:59 Join kergoth [0] (~kergoth@ip71.churchill-park.dfw.ygnition.net)
02:43:29 Quit XShocK (" Like VS.net's GUI? Then try HydraIRC -> http://www.hydrairc.com <-")
02:47:04 Quit kergoth (Read error: 104 (Connection reset by peer))
02:47:15elinenbeHCl: the iriver port now supports grayscale?
02:47:28elinenbehow many bit grayscale does the iriver LCD support?
02:47:50rasher2 bits
02:48:03 Join kergoth [0] (~kergoth@ip71.churchill-park.dfw.ygnition.net)
02:48:06rasherand no, it doesn't.. not in cvs at least
02:48:14rasherbut the code exists
02:48:15elinenbehow many bit grayscale does the gameboy support?
02:48:20rashersortof at least
02:48:24elinenberasher: sounds nice...
02:48:55elinenbes, is two bit grayscale 3 colors −− off, gray, black, or is it four? is 01 the same as 10?
02:51:23BonkersI think gameboy is 3 colors + white
02:52:55HClrasher: not sortof, heh. it exists.
02:53:09HClgameboy is 2 bit grayscale too.
02:55:38HClhttp://titania.student.utwente.nl/rockbox.zip <- should-be-working grayscale rockboy
02:55:46HClfor iriver
02:57:07Bonkersis the recorder ever expected to have sound, as in is it even possible?
02:57:11HClthe pokemon picture on the wiki is a pic with grayscale, though its not one of my best pictures
02:57:24HClfor gameboy?
02:57:56HClas far as i know archos have mp3 decoder chips and won't eat anything else but mp3s, so.. no
02:58:31HClnot without encoding it into mp3 before sending it to the chip
02:58:39HCland we definately don't have the cpu for that, heh
02:58:51Bonkersya, that's what I figured had to be done
02:58:58fuzzieyou could do it in non-realtime. :)
02:59:14fuzziespend an hour or so buffering a song, play it, etc :p
02:59:21BonkersI'm not terribly familiar with the mp3 format, but there's no easy way to encode sine waves possibly?
03:00
03:00:06HCli have no idea
03:00:10HCli just do iriver :x
03:00:35elinenbeHCl: looks nice
03:00:51elinenbeHCl: how is the dynarec? did it give a nice speedup?
03:00:58HClnot finished yet
03:01:02elinenbeand, what frequency are you running the plugin at?
03:01:08HClits still running the interpreter core
03:01:10HCl96mhz
03:01:27elinenbeand what is the speed of the current rockboy (on that core?)
03:01:54HClnear-full speed
03:02:12HCli'd say 90%
03:02:23elinenbenice.
03:02:46elinenbecan you bump the plugin up to 144 (is that full speed of the cpu?)
03:02:47HClyea, didn't expect it to work that well at alll
03:03:19elinenbeHCl: I do expect (in time) for rockboy to work in the background behind an mp3!
03:03:22***Saving seen data "./dancer.seen"
03:03:29elinenbeI have faith in you guys...
03:03:32HClgheheh.
03:09:35HClwell, since mp3 decoding isn't even at realtime yet
03:09:40HCli don't think we'll manage that
03:10:29stevenmHow fast can you run the CPU without overheating it?
03:10:36HCli don't know.
03:11:01stevenmAre there any thermal specs we can get? And someone with an opened device.. then again, closed it will heat up mpre
03:11:33stevenmAnd would it be feasale to write an MP3 decoder purely in ASM ?
03:18:43HCli think we tried one that was written in pure asm.. and it didn't even get to full speed..
03:18:47HCldunno though.
03:19:23HClyou really have to ask linus about cpu speed
03:20:08 Quit lImbus (" zzzZZzZZzZZZzzZzZZ")
03:26:02 Part rasher
03:30:18 Part fuzzie
03:30:46 Join fuzzie [0] (~fuzzie@cl-109.lon-01.gb.sixxs.net)
03:42:09 Quit stevenm ("Leaving")
03:44:55 Join Sutrannu [0] (~Sutrannu@c68.117.45.139.rose.mn.charter.com)
03:45:59 Part Sutrannu
04:00
04:06:00 Join QT [0] (as@area51.users.madwifi)
04:06:00 Quit QT_ (Read error: 54 (Connection reset by peer))
04:08:54 Join kergoth_ [0] (~kergoth@li11-226.members.linode.com)
04:09:02 Quit kergoth (Nick collision from services.)
04:09:18 Nick kergoth_ is now known as kergoth (~kergoth@li11-226.members.linode.com)
04:33:00 Join ze__ [0] (ze@adsl-69-231-231-17.dsl.irvnca.pacbell.net)
04:34:47 Quit cYmen ("n8")
04:44:24 Quit ze (Read error: 110 (Connection timed out))
04:44:24 Nick ze__ is now known as ze (ze@adsl-69-231-231-17.dsl.irvnca.pacbell.net)
04:52:28 Join ze__ [0] (ze@adsl-69-231-222-84.dsl.irvnca.pacbell.net)
05:00
05:03:23***Saving seen data "./dancer.seen"
05:05:51 Quit ze (Read error: 110 (Connection timed out))
05:05:52 Nick ze__ is now known as ze (ze@adsl-69-231-222-84.dsl.irvnca.pacbell.net)
05:08:00 Join lostlogic [0] (~lostlogic@node-4024215a.mdw.onnet.us.uu.net)
05:22:38 Part pabs
05:22:47 Join pabs [0] (~pabs@xor.pablotron.org)
05:30:43 Quit ze (Read error: 104 (Connection reset by peer))
05:35:15 Join ze [0] (ze@adsl-69-231-206-215.dsl.irvnca.pacbell.net)
06:00
06:07:36pabsno rockboy in the simulator?
06:29:35 Join ashridah [0] (ashridah@220-253-121-6.VIC.netspace.net.au)
06:39:02BonkersI think I heard the key input code was broken?
07:00
07:03:25***Saving seen data "./dancer.seen"
07:17:44 Join ze__ [0] (ze@adsl-69-231-236-196.dsl.irvnca.pacbell.net)
07:21:04 Join StrathAFK [0] (~mike@dgvlwinas01pool0-a239.wi.tds.net)
07:23:27 Join LinusN [0] (~linus@labb.contactor.se)
07:27:22 Quit Strath (Nick collision from services.)
07:27:28 Nick StrathAFK is now known as Strath (~mike@dgvlwinas01pool0-a239.wi.tds.net)
07:30:37 Quit ze (Read error: 110 (Connection timed out))
07:30:38 Nick ze__ is now known as ze (ze@adsl-69-231-236-196.dsl.irvnca.pacbell.net)
07:30:53LinusNwow! rockbox on the iFP5xx doesn't seem impossible at all...
07:31:04Strath?
07:33:57LinusNan iriver flash based player
07:34:36Strathso should be an easy port :)
07:35:53 Join DMJC-C [0] (~cba66be7@220-245-162-47-sa-nt.tpgi.com.au)
07:37:20Strathrelativly anyways...
07:37:28LinusNhmmm, maybe not...
07:37:56LinusNlooks like we have another case of secret data sheets :-(
07:38:43DMJC-Cso are any of the sound codecs running at realtime yet?
07:38:51Strathuh oh...
07:39:01 Quit DMJC-C (Client Quit)
07:43:33pabserr
07:43:44pabsLinusN: yo, what's up with the commit of rockboy in cvs?
07:43:59LinusNwhat do you mean?
07:44:10pabsit's ass slow, if i had to guess, it looks like either an older version, or it's got a bunch of debugging code
07:44:42pabs(i realize it's not your deal, but i was hoping you might know what's up with HCl)
07:44:58*pabs doesn't want to patch against something that's obsolete
07:45:59LinusNpabs: did you increase the cpu frequency?
07:46:20pabsLinusN: i haven't done anything to it really
07:46:45LinusNthe cpu is running at 11MHz by default
07:46:53pabsLinusN: all i know is i had a rock from HCl's ftp earlier this mrning that was like 80% or 90% normal speed, then i compiled the rock from cvs this evening and it's slow as poo
07:46:56LinusNyou need to manually boost it in the debug menu
07:47:07pabsk
07:47:30LinusNpabs: did you install a new rockbox.iriver too?
07:47:35pabsyup
07:47:44LinusNand you rebooted?
07:47:50pabsyeah
07:48:11LinusNdid you boost the frequency last time you tried?
07:48:32pabsLinusN: nope
07:48:50pabsLinusN: i think he's got a bunch of debugging code in his rockboy snapshot or something
07:48:55pabsat least, that's my best guess
07:48:59LinusNperhaps
07:49:17pabsi haven't mucked with the cpu frequency or anything like that
07:49:20LinusNanyway, you can boost the freq in the "view i/o ports" debug screen
07:49:24pabsk
07:49:39LinusNgo there and press UP, the STOP to exit again
07:53:42dwihnoyay! friday! ... wait a sec! it's still thursday! boooo!
07:54:00pabsLinusN: yup, that did it
07:54:14pabsLinusN: and i commentd out his framecount code too
07:55:13pabsLinusN: how did it get set back to 11MHz?
07:55:28LinusNback?
07:55:51pabsi guess my question is, rockboy was running at this speed this morning
07:55:59pabsi never set the freq to 96MHz
07:56:27Bonkersit's certainly not the latest rockboy
07:56:32LinusNi guess his private version boosts the frequency
07:56:33Bonkersthings like grayscale aren't in at the moment
07:56:38pabsgotcha
07:56:57Bonkersdon't know what else, and it's like 10% speed on the recorder if that even
07:57:21pabsBonkers: doh
07:57:24LinusNrockboy on the archos seems pretty pointless to me
07:57:47Bonkersmaybe some recompliation tricks could make it work
07:57:57Bonkerslike recompiling before you load the rom onto the recorder
07:58:01Bonkersit's still faster than the original gameboy
07:58:27LinusNbut the screen is half the size
07:58:34LinusNand monochrome
07:58:36Bonkersit's not that bad still though
07:58:42Bonkersisn't there a grayscale trick?
07:58:50Bonkersmonochrome only misses 2 colors of the gameboy anyway
07:58:59Bonkershttp://terrapin.shacknet.nu/rockboy/index.html
07:59:02Bonkersthat's what it looks like
07:59:17Bonkersthumbnails are actually scaled up, if you click on them you get the real size
08:00
08:00:29LinusNsaw those
08:01:40pabson the plus side
08:01:49pabsmulti-key button support seems to work in this one
08:01:52pabswhy, i don't know
08:02:01pabsso i can play bionic commando again
08:02:03pabs:>
08:02:59LinusNwhere do you find the roms?
08:03:17pabsme?
08:03:27Bonkersyou can read them off carts decently easily, there are tons of homebrew gameboy roms out there though
08:03:45Bonkersthere used to be a pretty decent sized gameboy hacking seen out there, don't know about now
08:03:54Bonkersthe gameboy camera grew out of the old mailing list
08:03:55ashridahpabs: HCl got basic multi-key support (ie, joystick and play/stop at the sametime) going iirc
08:04:04pabsashridah: k
08:04:14pabsashridah: maybe this is a recent build then, and he just left some debugging support in there?
08:04:30pabsi saw the code for multi key stuff in there, i was like "if this is an old build, then why is this here?)
08:05:25Bonkersya, multi key certainly works on recorder from cvs
08:05:44pabsit's slow as ass if you don't commnt out the framecount stuff
08:05:49pabsi believe it's in emu.c
08:06:03BonkersI'll try taht
08:06:13ashridahdoes it still only use the single file in the root?
08:06:18ashridahfor the rom?
08:06:52Bonkersit's 2 files
08:07:00pabsthe other changes i made were minor, i converted the is*() libc wrappers to macros, and i switched the startup splashes to lcd_puts()
08:07:10pabssince i got tired of waiting 1 second for each splash
08:08:04pabsLinusN: is rb->memcmp() reasonably fast?
08:08:14ashridahpabs: where in emu.c does one disable the framecount?
08:08:16LinusNpretty fast
08:08:20pabsLinusN: alright
08:08:31pabsi noticd he's updating the entire lcd each frame
08:08:49pabsi thought i'd maintain a lastframe, and try to memcmp() blocks of the screen, to shrink the update rect
08:09:00pabsi don't know how much of a speed increase that'd be, but it seemed like something to try
08:09:10LinusNupdate_rect is faster than memcmp
08:09:20LinusNon the iriver
08:09:23pabsk
08:09:45pabsashridah: let me double-check where i made the change, hold on a sec
08:10:12ashridahthe snprintf(meow,499,"%d",framecount); through to rb->lcd_update_rect(0,0,LCD_WIDTH,8) bit at the end?
08:10:38Bonkershardly any speedup on the recorder with no framecount
08:10:44pabsashridah: yup
08:11:11pabsBonkers: it's the sprintf, lcdupdate that's slowing it down tehre, that looks like some debugging code he lfet enabled on accident
08:11:18pabsashridah: i was about to paste that
08:11:48pabsLinusN: is there a chart that has relative costs of api calls like that anywhere?
08:12:04LinusNno
08:12:06pabsLinusN: k
08:12:20pabsi googled around for some coldfire documentation earlier and didn't reallly find anything useful
08:12:30pabsi stumbled across the m68k faq, which hasn't been updated since 1996
08:12:35LinusNdocs about what?
08:12:40pabsthe coldfire section says "coming soon, see my web site!"
08:12:54pabsLinusN: well, two things, 1. speed of api calls, 2. i was just looking for more processor info
08:12:59LinusNwhat coldfire section?
08:13:00pabstwo separate things
08:13:12pabsLinusN: in the m68k faq on faqs.org
08:13:19pabsLinusN: it says "coming soon!" (last updated, 1996)
08:13:32LinusNpabs: look in the rockbox documentation page, under "data sheets"
08:13:38pabsLinusN: so i checked the url at the top of the page, and it hasn't been updated since 1995
08:13:42pabsLinusN: k
08:14:14Bonkerspabs: I only left in the yield there
08:14:26LinusNcoldfire is nice, but it has several real major flaws
08:14:37LinusNthe bus controller is one
08:14:47pabsBonkers: i doubt there's ay way you'll be able to make it bearable on the archos, but LinusN is the man to ask
08:14:52pabsLinusN: why?
08:15:03LinusNit can't run internal and external accesses in parallel
08:15:13LinusNand no delayed writes
08:15:29LinusNwhich means that the pipeline stalls on every external access
08:15:47 Join ze__ [0] (ze@adsl-69-231-206-145.dsl.irvnca.pacbell.net)
08:16:26pabsLinusN: :(
08:16:37pabsLinusN: as long as i ca nplay my bionic commando i'm happy
08:16:58LinusNif the bus controller could access external memory in parallel, you could benefit from interleaving the instructions
08:16:59BonkersI'd probably never play a game on my recorder anyway, I'm jsut looking on for the coolness factor
08:17:22pabsLinusN: what kind of overhead is there to the thread calls in the plugin api?
08:17:39pabsBonkers: it seems liek stuff is actually playable on my iriver
08:17:45Bonkersya, from what I hear
08:17:46pabsBonkers: i mean it's a little slow, but it's not too bad
08:17:49 Join AC [0] (~5078751e@labb.contactor.se)
08:17:51pabsi think that can be fixed
08:17:53AChi
08:18:03BonkersI'd work on it if I didn't have my hands in too much already
08:18:06AChas somebody a gameboy rom for me?
08:18:09pabsi've bee npeeking through the code and it looks like there's s a bunch of stuff htat can be done
08:18:14ashridahuh. what the hell are the keys for rockboy?
08:18:14pabsi might not necesarily be the guy to do it
08:18:16pabsbut
08:18:21Bonkershttp://www.zophar.net/roms.phtml?op=show&type=gb for starters
08:18:42Bonkersashridah: arrows are arrows and f1 was a and f3 was start, don't know the others
08:18:46pabsashridah: on the iriver? record = start, ab = select, stop = b, play = a
08:19:32pabsLinusN: i'll spend some more time reading the wiki stuff, rather than pestering you with idiotic questions
08:19:49pabsLinusN: and i'll play around with the thread api and figure out how it feels
08:23:26ashridahrofl. can't tell where the holesi n the ground are ;)
08:24:38 Quit AC ("CGI:IRC (EOF)")
08:25:04pabsLinusN: actually, one more idiotic question
08:25:14pabsLinusN: how bad is it if an app doesn't yield?
08:25:29pabsor, if it only yields periodically?
08:27:50 Quit ze (Read error: 110 (Connection timed out))
08:27:50 Nick ze__ is now known as ze (ze@adsl-69-231-206-145.dsl.irvnca.pacbell.net)
08:32:03BagderI believe I fixed the "no daily ondio fm build" problem just now
08:32:04 Quit elinenbe (Read error: 104 (Connection reset by peer))
08:32:21 Join elinenbe [0] (elinenbe_@207-237-225-9.c3-0.nyr-ubr1.nyr.ny.cable.rcn.com)
08:34:15LinusNpabs: it is very bad to not yield
08:34:15 Quit shx ("CGI:IRC (EOF)")
08:34:24LinusNBagder: welcome home
08:34:30Bagderthanks
08:34:34Bagderbeen a busy week I see
08:38:11BonkersLinusN: what are you yielding to, mp3 playback code?
08:38:21Bonkersand possibly screenshot stuff too
08:38:40LinusNplayback code, power management, backlight handling, ...
08:39:04LinusNdisk spindown
08:39:16Bonkerswell since rockboy won't ever do mp3 playback at the same time on recorder at least, couldn't you yield every 5 seconds or something?
08:39:31Bonkersinstead of every frame like it's doing right now
08:39:42LinusNevery frame is not necessary
08:39:58Bonkersso maybe that would speed it up
08:40:03LinusNbut i'd like it to yield *at least* every second
08:40:03BonkersI'll try that I guess
08:40:23Bonkerswell right now on the recorder, it's maybe just under a second per frame
08:42:04pabsBonkers: i tried that, didn't help much
08:42:08pabsjust now
08:42:12Bonkersnot yielding at all?
08:42:15pabsyup
08:42:26pabsLinusN took too long to respond ;D
08:42:35Bonkersheh, well saves me effort then
08:42:38pabsyeah
08:42:59pabsstill should only be like once every 15 frames or something
08:43:05LinusNyielding doesn't cost that much
08:43:21LinusNas long as the other threads don't have anything to do
08:43:24 Quit einhirn (Read error: 104 (Connection reset by peer))
08:43:27pabsLinusN: k
08:43:47pabshis ui polling doesn't look like it's got any significant overhead either
08:43:49BonkersLinusN: is there anything like a TLB on this thing that gets cleared on thread switches?
08:44:00Bonkersobviously not a tlb
08:44:02LinusNno mmu
08:44:35Bonkersso all "cooperative" addressing/allocation and multitasking?
08:46:28 Join preglow [0] (thomj@s183a.studby.ntnu.no)
08:48:50LinusNBonkers: yes
08:48:56LinusNaddressing/allocation?
08:49:33Bonkersas in no memory protection and one memory space
08:49:38LinusNyes
08:49:57Bonkersfun
08:50:06LinusNeasy at least
08:50:53LinusNextremely simplistic
08:52:37preglowwow, didnt' know that about the bus controller
08:52:39Bonkersya, in the same way PIC assembly is easy
08:52:53Bonkerscan't get too confused with 200 bytes of ram and 36 instructions
08:52:59preglowa bit flawed
08:53:08Bonkersbut then when you want to even multiply 2 numbers, the going gets rough
08:55:15LinusNpreglow: yes, very annoying to be so dependent on the internal ram
09:00
09:01:51 Join einhirn [0] (Miranda@bsod.rz.tu-clausthal.de)
09:03:28***Saving seen data "./dancer.seen"
09:06:08preglowahoy, linus, i tried putting __attribute__ ((section(".icode"))) on a bunch of stuff in libmad, but it really didn't do anything
09:06:27LinusNoh?
09:06:30LinusNhuffman stuff?
09:06:37preglowno, mostly the asm stuff
09:06:49preglowso it should work
09:06:55LinusNi'm afraid it won't do much unless the data is in iram too
09:07:14preglowsounds reasonable, but i would have expected at least a small boost
09:07:15preglowanyway, shower
09:07:18LinusNme too
09:07:32LinusN(expected a boost, that is :-)
09:08:44Bonkersso you're not showering with him? ;)
09:09:34LinusN:-)
09:15:52 Join shx [0] (~a4810127@labb.contactor.se)
09:27:26preglowhrmf
09:27:40LinusNpreglow: you say that a lot :-)
09:27:52preglowthe default case for the flac lpc routine isn't incredible fast
09:28:32preglowbut i can't continue with the unrolling strategy that far
09:30:00preglowand flac supports having two functions internally, one for orders under 8, and one for those over, right now i have both in one routine, so flac will do a irrelevant check before calling
09:30:07preglowbut that beats having to duplicate some function code, i hope?
09:31:12LinusNi don't think that matters
09:31:15preglownor do i
09:38:58 Quit midk (Read error: 104 (Connection reset by peer))
09:39:10 Join midk_ [0] (~midk@c66-235-14-120.sea2.cablespeed.com)
09:45:54 Join Chamois [0] (HydraIRC@champigny-5-82-226-182-23.fbx.proxad.net)
09:46:07Chamoismorning
09:46:50LinusNmorn
09:47:27 Quit shx (Read error: 110 (Connection timed out))
09:47:28*Bagder drowns under heaps of email
09:47:43ChamoisLinusN: are you always interested into the battery levels ?
09:48:06LinusNall the time :-)
09:48:14Chamoisok I'm working on it
09:50:59preglowexcellent
10:00
10:11:31preglowholy shit
10:11:36preglowflac runs well above realtime here
10:11:45preglow140% on a real life example
10:12:26preglowdecreases, but still 115%
10:12:36LinusNkewl!
10:14:40Bagderrocking preglow
10:14:52Bagderrockpreglowbox
10:15:12preglowbit identical ouput
10:15:14*preglow smiles
10:16:17preglowmy screen is flickering like hell
10:16:30preglowif i suddenly shut up, don't wonder about it
10:18:48preglowi wonder why my test sample was so slow in comparis
10:19:03preglowperhaps flac doesn't like c64 music :)
10:19:24LinusNit has no taste
10:19:42LinusNhow can it *not* like c64 music!!!
10:19:51preglowi don't know
10:20:13preglowi fed it a goto80 sample, and it ran 70% realtime
10:21:19 Join elinenbe_ [0] (elinenbe_@207-237-225-9.c3-0.nyr-ubr1.nyr.ny.cable.rcn.com)
10:21:19 Quit elinenbe (Read error: 104 (Connection reset by peer))
10:21:23 Nick elinenbe_ is now known as elinenbe (elinenbe_@207-237-225-9.c3-0.nyr-ubr1.nyr.ny.cable.rcn.com)
10:25:51preglowit might be the fact that it's ripped from vinyl, somehow
10:30:35 Quit einhirn (Read error: 104 (Connection reset by peer))
10:32:38 Join einhirn [0] (Miranda@bsod.rz.tu-clausthal.de)
10:37:17 Join Rick [0] (rick@Rick.user)
10:37:21pabsdammit
10:37:29pabsi finally got this menu finished, and theusb on my machine freaks out
10:47:32preglowcan anyone think of what gas thinks is wrong with the following?
10:47:34preglowmac.w %a6.l, %d7.l, (4*4, %a2), %a6, %acc0
10:47:54preglowmac.w is listed in m68k-opc.c, so it has to be something with the registers
10:50:00preglowdoesn't work with uppercase L either, as per coldfire programmers reference
10:50:14 Nick Strath is now known as StrathAFK (~mike@dgvlwinas01pool0-a239.wi.tds.net)
10:50:49 Join [IDC]Dragon [0] (~d90a3255@labb.contactor.se)
10:51:17 Join jyp [0] (~jp@181.18-136-217.adsl.skynet.be)
10:54:56LinusNdoes mac.w take that many parameters?
10:55:12preglowyes
10:55:19preglowit takes as many as mac.l
10:56:03LinusNMAC.<size> Ry.<ul>,Rx.<ul>,<shift>
10:56:20preglowthat's the one with no parallel load
10:57:25LinusNand macl is with the load?
10:57:41preglowboth have parallel loads, they're the same instruction, only a flag bit is different
10:58:37preglowthat'd be the size flag, which can be 1 or 0 for either long or word mode
10:58:37LinusNi'm just looking in the data sheets
10:59:05preglowno
10:59:08preglowi'm in error
10:59:18preglowthe .w version actually has no parallel load
10:59:23preglowhow sucky
10:59:49LinusNcould you describe the difference between mac.l and macl.l?
10:59:51preglowhmm, no, here i am, talking out of my ass again, the programmers reference clearly states it has
11:00
11:00:00preglowthere is not macl.l, only mac.l
11:00:19LinusNthe data sheet says there is
11:00:24preglowreally?
11:00:28LinusNmacl.w too
11:00:39preglowthat's probably the load version, then
11:00:41 Join ze__ [0] (ze@adsl-69-231-203-220.dsl.irvnca.pacbell.net)
11:00:44LinusNmac is "multiply and accumulate"
11:00:55preglowgas autodetects which i use from how many parameters i pass it
11:00:58LinusNmacl is "multiply and accumulate with register load"
11:02:10preglowi've been using both variants as simply "mac.x", so it does work
11:03:31***Saving seen data "./dancer.seen"
11:06:40preglowoh, bah, it's not very important anyway, i just saw that flac only passes 16 bit wide coefs to my routine, so i can use the mac.w form which is only two cycles
11:06:43preglowbut it can wait
11:12:11 Quit ze (Read error: 110 (Connection timed out))
11:12:12 Nick ze__ is now known as ze (ze@adsl-69-231-203-220.dsl.irvnca.pacbell.net)
11:12:24LinusNlooks like gas doesn't understand the .l and .u extensions to the source fields
11:12:51LinusNbetter hurry with a patch, binutils 2.16 will soon be branched
11:14:23preglowcan you handle that? i have no idea how to fix it
11:14:53LinusNpreglow: found it, it's not %d1.l it's d1l
11:14:56LinusNskip the "."
11:15:08preglow...
11:15:30preglowok, i'll try that then
11:15:46LinusNmac.w %a6l, %d7l, (4*4, %a2), %a6, %acc0
11:18:27LinusNlunch time
11:18:45preglowso, let's see if this helps
11:20:37preglownah, impossible to tell. the read stalls the instruction too long with sdram data
11:27:13 Join jyp_ [0] (~jp@14-177.240.81.adsl.skynet.be)
11:29:19 Quit jyp (Read error: 110 (Connection timed out))
11:38:33 Join ze__ [0] (ze@adsl-69-231-203-50.dsl.irvnca.pacbell.net)
11:42:45 Join Aison [0] (~hans@zux166-181.adsl.green.ch)
11:51:33 Join Renko [0] (~jon@host217-43-110-62.range217-43.btcentralplus.com)
11:52:11 Quit ze (Read error: 110 (Connection timed out))
11:52:11 Nick ze__ is now known as ze (ze@adsl-69-231-203-50.dsl.irvnca.pacbell.net)
11:53:17 Part Renko
11:55:51 Join Renko [0] (~Renko@host217-43-110-62.range217-43.btcentralplus.com)
12:00
12:11:57HClgahhhh
12:11:57HClhi
12:12:08*HCl got two of his wisdom teeth pulled >.<
12:14:22 Join webguest80 [0] (~dcf0e22d@labb.contactor.se)
12:14:50preglowhaha
12:14:52preglowlucky you
12:14:53 Quit webguest80 (Client Quit)
12:15:05HCl :/
12:15:06preglowa dentist did it, i assume? :P
12:15:10*HCl can't talk
12:15:13HClum.. jawsurgeon
12:15:25 Join guarsaris [0] (~dcf0e22d@labb.contactor.se)
12:15:26preglowi've had me share of teeth problem
12:15:28HClmuch better than dentist
12:15:37HCldidn't feel a thing aside from the needle to sedate it
12:15:45HClbut now my whole left side of my mouth doesn't feel anything
12:15:53HCland i get the feeling i'm sometimes chewing on my own cheek
12:15:56preglowyes, it's good fun
12:15:57HCland not noticing
12:16:00preglowi always feel like i'm drooling
12:16:34preglowtalking in that state has you sounding like a heroin addict
12:16:36HCli'm just waiting for this really really annoying sedation to go away
12:16:44HClxD
12:16:45HClyea
12:16:52HCli can't even smile properly
12:17:14HCli wish i could feel anything in there >.<
12:17:26HClofcourse it'll be replaced with pain
12:17:28HClbut blah.
12:17:32preglowhaha
12:17:41HCli think i can take pain
12:17:45HClthough it may make me cranky
12:17:47preglowi hate tooth pain
12:17:51preglowall other pain is ok
12:17:53*HCl shrugs
12:17:58HCli just mostly think its annoying.
12:18:02RickHCl: Put a large cotton wad between your cheek and your teeth, :P
12:18:03HClits not as bad as backpains
12:18:07HClor emotional pain
12:18:18HClRick: heh...
12:18:27preglowsure, but i meant physical pain
12:18:39preglowtooth pain just annoys the hell out of me
12:18:41HClwell, backpains are far worse
12:18:44HClyea, its annoying.
12:18:51preglowi can't get shit done
12:19:03RickWell, it's better than not knowing your teeth are rotting to hell. ;)
12:19:10HClheh.
12:19:20 Quit guarsaris (Client Quit)
12:19:31RickI mean, if your tooth hurts like hell, then doesn't anymore. It's usually a sign something bad has happened. :P
12:19:40preglowhaha
12:20:42Rick(aside from possible pus expulsion)
12:20:55 Join hgoslek [0] (~chatzilla@static-45.226.240.220.dsl.comindico.com.au)
12:24:36ashridahhm. Output: '聦i猫路酶" is not a substring of "酶玫每驴脿[每路麓E没路". probably not a good sign, those were normal characters going in :)
12:26:00ashridahrofl. had the src and dest the wrong way around on a strncpy
12:26:01*ashridah smacks head
12:26:29Rickhah
12:26:34Ricki've done that before myself
12:30:41 Quit [FlaT]Heidel (Read error: 110 (Connection timed out))
12:31:30Rickhmm
12:36:06Rickhow can I check out a copy of the uisim? cvs didn't pull it when I checked out rockbox
12:37:23ashridahrockbox-devel is an alias that includes the uisimulator.
12:37:26ashridahjust cvs co uisimulator from the checked out rockbox directory
12:37:27Rickah, okay
12:37:37ashridahit'll get the extras
12:37:44Rickthanks :)
12:44:47Rickscary!
12:44:51Rickuisw32 - 799 error(s), 14 warning(s)
12:44:52*Rick grins
12:46:54ashridahsounds like you're missing some development headers
12:47:10Ricki'm working through the errors now
12:47:27Rickyeah, appears so
12:47:29Rickmissing lang.h
12:47:44Rickseem to be missing some sort of config as well
12:47:55Rickas undefined #DEFINE's
12:49:34Rickyeah
12:49:37*Rick kicks self ;)
12:53:18LinusNRick: are you sure you do this the right way?
12:53:39preglowargh
12:53:41LinusNdid you use the configure script?
12:53:51LinusNpreglow: not "hrmf"?
12:53:52preglowhow do i make cvs forget my anonymous login?=
12:54:02LinusNcheck out in a fresh dir
12:54:12preglowthat's perhaps clever anyway
12:55:11RickLinusN: wasn't aware that was needed for win32 building.
12:55:29LinusNit's the same as for target builds
12:59:12preglowsystem.c is built for simulators as well?
12:59:41 Join cYmen [0] (~cymen@nat-ph3-wh.rz.uni-karlsruhe.de)
12:59:53preglowahh, of course, CONFIG_CPU will block most of it out
13:00
13:01:19Rickhmm
13:01:23Rickthere's a language.h, but no lang.h
13:03:33***Saving seen data "./dancer.seen"
13:04:27 Join shx [0] (~a4810127@labb.contactor.se)
13:05:00LinusNRick: lang.h is generated
13:05:53Rickah
13:06:41RickThen i'll muck about getting linux back on my second harddrive tommorow, still havn't done that yet
13:10:18 Quit jyp_ ("poof!")
13:13:25 Quit preglow (zelazny.freenode.net irc.freenode.net)
13:13:25NSplitzelazny.freenode.net irc.freenode.net
13:14:02NHealzelazny.freenode.net irc.freenode.net
13:14:02NJoinpreglow [0] (thomj@s183a.studby.ntnu.no)
13:14:14preglowexcellent...
13:17:27LinusNreally?
13:18:00preglowok, new flac lpc decode in cvs
13:18:17preglowlets hope i didn't broke anything
13:22:24preglowi hope the id strings are fixed automatically, i completely forgot about them
13:23:04 Join markun [0] (~markun@bastards.student.utwente.nl)
13:23:07ashridah$Id$ you mean?
13:23:11ashridahcvs takes care of them
13:23:54markunLinusN: do you remember the mail I send you about gcc for m68k not compiling with binutils from cvs on FreeBSD?
13:24:00LinusNyes
13:24:27LinusNi forgot to reply, didn't i?
13:24:56dwihnohmm
13:24:58dwihnonew charging code!
13:25:13markunLinusN: I figured you didn't have time. Maybe I can send it to the binutils or gcc mailing. Which one?
13:25:41LinusNit's funny, because i had the same problem the other day
13:27:19preglowseems a couple of tabs has sneaked in
13:27:46*preglow edits .vimrc
13:27:56markunThe strange thing was that gcc compiled fine with the same binutils under linux.
13:28:00LinusNi fixed it by removing all references to unneeded targets in gcc/libgcc.mk
13:28:25LinusNthis was with cygwin
13:28:46markunLinusN: Did you also have fmovem instructions in your .s file?
13:29:06LinusNyes
13:29:23LinusNnothing wrong with that, that file is for 68040
13:30:22markunI was just wondering why the .s file under linux was so different.
13:30:26preglowLinusN: i suppose i should continue using tabs in files that already use tabs?
13:30:40LinusNnah
13:30:42 Quit hgoslek ("Chatzilla 0.9.67 [Firefox 1.0/20041107]")
13:31:48preglowit'll look stupid for people with eight space wide tabs :V
13:32:29LinusNtabs should always be 8 spaces imho
13:33:02preglowi ordinarily use tabs in my code, so i don't like them that wide
13:33:04LinusNso many things break if tab != 8
13:33:20LinusNlike printouts with lame printer drivers
13:33:27HCloh. um.
13:33:29preglowthere's a duplicate line in there as well, so i'll fix it
13:33:34LinusNor if you view it in a browser
13:33:35HClincidentally, i'm not entirely sure who ta ask
13:33:41LinusNtabs are evil
13:33:42HClbut. might it be possible for me to get cvs access...?
13:35:02markunLinusN: What do I need to keep in libgcc.mk? Everything starting with libgcc/m680xx ?
13:36:00LinusNremove all dependencies to cpu32 up to the end of the line
13:36:31LinusNthe only thing you really need is the 5200 stuff
13:42:35preglowcould anyone check out my commit to make sure i didn't break anything before i go out for a minute?
13:42:58LinusNpreglow: http://www.rockbox.org/daily.shtml
13:43:07preglowLinusN: good idea
13:43:30preglowi'll take that as a yes...
13:43:32 Nick Lynx_awy is now known as Lynx_ (HydraIRC@134.95.189.59)
13:45:22preglowwhy is it even compiled for the sim? is CONFIG_CPU set for the sims?
13:45:36preglowwell, apparently it is
13:48:00preglowlinusn: i haven't got the x11 headers at the moment and need to run soon, it's just matter of adding && !defined(SIMULATOR) in libFLAC/SOURCES, yes?
13:48:09LinusNyes
13:48:35LinusNand make sure the C version is compiled instead
13:49:57preglowit's always compiled at the moment
13:50:58LinusNah yes
13:57:29 Quit Schee (Read error: 110 (Connection timed out))
13:58:52 Quit [IDC]Dragon ("CGI:IRC (Ping timeout)")
13:58:58 Quit shx ("CGI:IRC (EOF)")
14:00
14:01:06 Quit DeadMan (Read error: 110 (Connection timed out))
14:01:42 Join Schee [0] (~SeeSchlos@ARennes-352-1-10-114.w83-195.abo.wanadoo.fr)
14:05:40markunLinusN: I removed all the lines with cpu32, but it didn't compile. I was stupid enough to do a make clean, so now I have the old .mk file again.
14:05:56LinusN:-)
14:06:15markunLinusN: Was that the only thing you had to change to get it to compile?
14:06:50markunOr is removing all the lines not the correct thing to do?
14:07:19 Join preglow_ [0] (thomj@s183a.studby.ntnu.no)
14:08:16LinusNi removed the dependencies for cpu32 and up in the all: and install: targets
14:09:02LinusNand also the lines that install them in the install: target
14:09:07 Part preglow_
14:11:22 Join Nuxator [0] (~chatzilla@abo-131-236-68.guy.modulonet.fr)
14:11:29Nuxatorhi
14:12:00NuxatorI tried to buils win32 simbuild for iriver
14:12:13Nuxatorit seems that last comit for libflac broke somthing
14:12:26*HCl prods preglow
14:12:38 Join R3nTiL [0] (~zorroz@83.69.98.216)
14:13:01NuxatorCC coldfire.S
14:13:02Nuxatorcoldfire.S: Assembler messages:
14:13:04Nuxatorcoldfire.S:24: Error: junk at end of line, first unrecognized character is `,'
14:13:06Nuxatorcoldfire.S:28: Error: invalid character '.' in mnemonic
14:13:10preglowgreat
14:13:34LinusNNuxator: update CVS and rebuild
14:13:41Nuxatoroops jusr see that it was updated
14:13:44preglowi hope this is the sim
14:13:53Nuxatorsorry
14:14:00preglowno need to apologize
14:14:04preglowi'm the fool around here :)
14:16:29 Quit lostlogic ("Going to the moon")
14:17:19 Quit pabs (Remote closed the connection)
14:18:16markunShall I commit my grayscale stuff to cvs tonight? It makes 2-bit bitmaps a lot slower because I now draw every pixel.
14:18:20HClLinusN: do i need to keep account of archos when adding iram optimizations..?
14:18:33LinusNyes
14:18:48LinusNarchos only has 4k, and it's nearly full
14:18:51HCl*nods*
14:19:09HClis there a global define yet for the attribute iram yet..?
14:19:49preglowthe sim is still bloody broken
14:19:49preglowarghh
14:21:02LinusNi suspect that you forgot to change the #ifdef in stream_decoder.c?
14:21:27 Quit Sando (Read error: 54 (Connection reset by peer))
14:21:52preglow#if CONFIG_CPU==MCF5249 && !defined(SIMULATOR)
14:21:54preglowthat ?
14:22:09preglowno, the error is the same, it still tries assembling the .S file
14:23:14 Join pabs [0] (~pabs@ip68-100-248-22.dc.dc.cox.net)
14:24:28LinusNhm
14:24:48 Quit Renko ("Leaving")
14:25:03markunLinusN: I get the same error again while building libgcc/m68040/_fixunsdfsi.o
14:25:23LinusNthat shouldn't be built
14:26:24 Quit Nuxator ("have to gp")
14:26:39markunI removed the mcpu32 parts from all: and install: in libgcc.mk
14:26:42ashridahgod i love subversion sometimes.
14:26:56LinusNmarkun: i said "mcpu32 and up"
14:27:09LinusNmeaning cpu32, 68020 68040 etc
14:27:20markunLinusN: ok, sorry. My cpu history is not so good :)
14:27:28ashridahnear instantaneous diffs, even tho my repository is stored on the other end of a congested network through an ssh tunnel
14:30:04 Join DrRick [0] (DrRick@81-86-89-192.dsl.pipex.com)
14:31:28 Join tnc [0] (~c1bec102@labb.contactor.se)
14:33:30 Part tnc
14:34:04fuzzieashridah: it can be much less fun when someone's decided that webdav is 'good enough' for your repos access.
14:36:01preglowmy god, x11 is a big compile
14:36:09ashridahhm. last time i used webdav it seemed okayish. bit too fiddly to setup auth tho
14:36:19ashridahpreglow: compile openoffice
14:36:38markunLinusN: Thanks, it worked.
14:37:12ashridahwth. why did my latest build of rockbox nuke my config?
14:37:34preglowashridah: i remember compiling gnome once...
14:37:46LinusNashridah: the settings data structure has changed
14:38:15ashridahah
14:38:37preglowLinusN: i have to vanish for a couple of hours in about ten minutes, i don't even think x will be done compiling by then, so can't fix it
14:38:49LinusNi'll do it
14:38:54preglowthanks a lot
14:39:11*HCl swallows more painkillers :/
14:39:41ashridahanyway, time to crash.
14:39:42 Quit ashridah ("sleep.")
14:40:57HClgeeze.
14:40:59HClthats really fast.
14:41:17HClrockboy at about 95%..
14:44:22 Nick preglow is now known as pregone (thomj@s183a.studby.ntnu.no)
14:49:19 Join bobTHC [0] (~foo@l03v-42-151.d1.club-internet.fr)
14:52:30bobTHChi mates!
14:53:44 Join edx__ [0] (edx@pD9EAAC82.dip.t-dialin.net)
14:57:26markunCompiling libfaad gives me some nice errors: syntax.c:1999: error: insn does not satisfy its constraints: (insn 1344 878 1346 81 syntax.c:1987 (set (reg:QI 12 %a4) (mem:QI (plus:SI (reg/f:SI 15 %sp) (const_int 67 [0x43])) [0 start_coef_bits+0 S1 A8])) 33 {*m68k.md:826} (nil) (nil))
14:58:52HClhmm
14:59:02HClrockboy looks uglyish without grayscale :/
14:59:35markunHCl: why would you NOT use grayscale?
14:59:46HClmarkun: cause its not in cvs yet
15:00
15:00:28markunok ok, I'll commit it soon. Is it a problem if things don't work as fast as they used to?
15:01:05HClwell, i don't mind, but i don't know about other people
15:02:12 Join lImbus [0] (~manuel@kernel.cycos.net)
15:02:16lImbushi all
15:02:21HClhey
15:02:41lImbusdid I badly miss something ? rockbox_flash.rock is not build anymore (local cvs and daily builds)
15:03:34***Saving seen data "./dancer.seen"
15:04:09LinusNlImbus: which target?
15:05:42lImbusrec v1
15:05:45HClHCl received pikachu
15:05:51HClyay :3
15:08:26LinusNlImbus: it's in the daily build zip
15:09:57 Quit webmind (Read error: 104 (Connection reset by peer))
15:11:27 Quit edx (Read error: 110 (Connection timed out))
15:13:05 Join webmind [0] (~random@217-195-236-172.dsl.esined.net)
15:13:19HClwb
15:15:54 Nick edx__ is now known as edx (edx@pD9EAAC82.dip.t-dialin.net)
15:18:07 Join elinenbe_ [0] (~elinenbe_@65.115.46.225)
15:20:56 Join DeadMan [0] (Rori@deadman3000.plus.com)
15:21:41 Join lolo-laptop [0] (~lostlogic@68.251.84.226)
15:22:23 Join jyp [0] (~jp@14-177.240.81.adsl.skynet.be)
15:23:25 Join ze__ [0] (ze@adsl-69-231-215-34.dsl.irvnca.pacbell.net)
15:23:30HClo.o
15:25:28jyp@.@
15:26:26DeadMan?.?
15:26:31jyp~.o
15:26:37*HCl prods markun and other people
15:26:44DeadManwot
15:26:46HCltoo bad amiconn isn't around
15:26:55HClhe had some thoughts about grayscale support..
15:27:14markunYes, maybe I should talk to him before I commit anything
15:27:46 Quit R3nTiL ()
15:28:05HClyea
15:28:24DeadManI think my postal service must have got blown up. No mail all week.
15:28:59DeadManand why does XP give you a BSOD with no usefull info whatsoever? :P
15:29:30DeadManI knew it was my Nokia phone driver but still...2nd time I installed it it booted up fine
15:30:19DeadManand Bill Gates recieved a knighthood from the Queen of Britain the other day so I heard
15:30:28DeadManfor what? being a total scumbag?
15:30:35HClhe probably bribed her
15:30:44DeadManmaybe :)
15:31:08jypHe has information about Charles the Queen doesn't want made public ... :P
15:32:28DeadManOne cannot simply bribe their way into Mordor!
15:33:39DeadManany 4chan /b/tards here? :)
15:36:14 Quit ze (Read error: 110 (Connection timed out))
15:36:14 Nick ze__ is now known as ze (ze@adsl-69-231-215-34.dsl.irvnca.pacbell.net)
15:37:02 Quit DrRick (Read error: 54 (Connection reset by peer))
15:37:54 Join methangas [0] (methangas@0x50a43276.virnxx10.adsl-dhcp.tele.dk)
15:45:12 Part LinusN
15:50:42 Join LinusN [0] (~linus@labb.contactor.se)
15:51:03LinusNwow! super mario in 120MHz is very playable!
15:51:22lolo-laptopDo we know what speed the iRiver firmwares run maximum?
15:51:43LinusNi believe it's 96MHz
15:52:19lolo-laptopLinusN: and is that the highest speed the ATA controller works at or have you gotten better than 96 w/ ATA working?
15:52:35LinusNi seems ata has nothing to do with it
15:53:00lolo-laptopoh, I must be on old data; cool.
15:53:05LinusNthe cpu gets too hot
15:53:17lolo-laptopor not cool... heh.
15:53:24LinusNheated by the constantly spinning hard drive and the cpu frequency
15:53:34HClwhat if you turn the hdd off though
15:53:35HCl?
15:53:35lolo-laptopthat was actually my guess that I ventured in #misticriver yesterday −− power and temp reasons led them to underclock...
15:53:39HClrockboy does not need hdd at all
15:54:19dwihnoHCl: audio playback during gameboy emulation...
15:54:21LinusNHCl: the problems occur when running the codec benchmarks, not in rockboy
15:54:52LinusNit probably won't even happen with normal playback since the hd doesn't spin that much
15:55:02jypLinusN: What api should I implement for cpu freq. changing on the gmini ?
15:55:02HClLinusN: so technically, it might be possible to add part to the api that would allow rockboy to run faster than 96mhz..
15:55:14HCldwihno: why would i need hdd for audio playback..? or do you mean mp3?
15:55:20LinusNjyp: check out system.c, cpu_boost()
15:55:24jypshould cpu_boost be moved to arch. independant code?
15:55:35LinusNHCl: yes
15:55:41HClcool.
15:55:56LinusNjyp: cpu_boost() would be independant, yes
15:56:03jypok
15:56:06*HCl drinks more painkiller
15:56:11LinusNgotta run
15:56:13LinusNcu guys
15:56:15 Part LinusN
15:56:16HClbyebye
15:56:19jypcu
15:57:21dwihnoHCl: yeah.. audio (codec) playback during play
15:57:27HClwell
15:57:40HCli doubt we'll have cpu left for that, considering current mp3 decoding speed
15:58:06dwihnoah, okay
15:58:12dwihnowhat about sound emulation then?
15:58:12HClgod, i hattttte teeth being pulled
15:58:28HCldwihno: should be trivial to add, its already working, just not hooked up into the iriver's audio interface
15:58:45HClthere's simply no audio driver linked with rockboy at the moment
15:58:45dwihnoHCl: neato
15:59:07dwihnowould be really nice to be able to listen to music while playing tetris or some kind of puzzle game
15:59:16HClyea, i know
15:59:22HClthe only hope of getting that to work is using dynarec.
15:59:32dwihnohas there been any progress?
15:59:40HClbut at the moment, i think i'm getting more interested in writing an asm interpreter core for iriver
15:59:46HClsince interpreter seems to be fast enough
15:59:51HCland its always more compatible than dynarec
16:00
16:00:00HClyea, a little
16:00:17HClwith the help of amiconn we managed to get a sub/cmp/add instruction down from 30 m68k instructions to 16
16:01:50HClyea.
16:02:04HClwell, just dynarec might be fast enough
16:02:14HClbut the interpreter will never be fast enough to play mp3 too
16:21:21 Join tnc [0] (~c1bec102@labb.contactor.se)
16:22:43 Nick pregone is now known as preglow (thomj@s183a.studby.ntnu.no)
16:23:43preglowwell, woot
16:23:53tncHi guys, it's the first time i join the irc channel, but i've been reading logs for a while now
16:23:59preglowi would not have been able to fix that bug :V
16:24:25tnci have to tell you you're doing a great job :-)
16:24:32 Join shx [0] (~a4810127@labb.contactor.se)
16:25:04tncit made me interested in some stuff i never was before
16:25:31preglowlike what? :)
16:25:38tncsuch as cross compiling under cygwin, and how emulators work
16:26:06tnci had a hard time setting a working m68k binutils/gcc under cygwin
16:26:12preglowi can imagine
16:26:16preglowi just gave up and used linux
16:27:18tncyou have to use stable m68k binutils to build gcc, then "switch" to cvs binutils to build rockbox
16:28:11tncand even after that, i ran into problems, my 68k build of rockbox had no errors, but all the bins crashed the iriver
16:29:09tncit seemed to comme from m68k-elf-gcc, i was using version 3.3.5, had to switch to 3.4.1
16:29:25preglowseveral people have had problems with 3.3.x
16:29:54tncNOW i can test some stuff on the "real" iriver (i had the win32 sim working for a while, and wrote some test plugins)
16:30:10 Join wacky_ [0] (~wacky@modemcable040.196-203-24.mc.videotron.ca)
16:30:16wacky_hey guys
16:30:25wacky_I was wondering if you wanted to add my plugin to the CVS ? Remember the 'iriverify' thing ?
16:30:55tnchad i built the bootloader myself with gcc 3.3.x, i'd have a great brick now :-|
16:31:03markunpreglow: I got some internal compiler errors while compiling libfaad. Do you want to take a look at it?
16:31:18preglowmarkun: don't think there's much i can do about internal errors
16:31:50tncah, more gcc fun
16:31:57markunMaybe I should report it to gcc?
16:32:14preglowif you can make a simple test case, then yes, do so
16:32:19markunIt only happens with -m5200
16:33:02tncmaybe someone could help me, i have a problem at linking a plugin i made, due to size constraints...
16:33:17tncbut i'm not that good at understanding the linking process
16:34:12tnciiuc, building plugins for iriver target gives me 500kb plugin space, right ?
16:34:13preglowtnc: error message, please
16:34:58tnci meant more than 500, but can't remember right now
16:35:10tncthe PLUGIN_RAM full message :-)
16:36:25tncbut the .o file is "only" 200kb, how can it more than triple size when transformed to .rock file ???
16:36:31wacky_anyone with CVS write access ?
16:36:54Ricktnc - debug?
16:36:57preglowtnc: tnc because of data i guess
16:37:02wacky_Rockbox is so fast that it *is* really useful to create playlists.. and play them on the original firmware for the iRiver
16:37:05preglowtnc: could i have a look at the source?
16:37:19wacky_so I though it would be a good addition, don't you guys think ?
16:37:58preglowwacky_: what addition?
16:38:05wacky_the iriverify.c plugin
16:38:18wacky_a tiny plugin that translates the .m3u files into something readable by the original iRivier firmware
16:38:31wacky_making on-the-fly playlists a reality, even with no sound support :)
16:38:39tncsure, i'll put the c file somewhere an give you the url
16:38:50wacky_a thing iriver people have been waiting for for so long
16:39:35HCltnc: what does your plugin do?
16:41:03tncsomething you inspired :-) : i tried to port an open source vectrex emu
16:41:15HClah.. whats vectrex..?
16:41:20tnci had one such beast when i was young
16:42:04preglowHCl: i think your button_hold addition broke the sim build
16:42:15tncit was a console that appeared in the 80s, that had vector-based gfx instead of bitmaps
16:42:16HClick. how..?
16:42:22preglowHCl: well, have a look at the dailies
16:42:26*HCl checks
16:42:30RickI was thinking about porting Lua to iriver as a plugin just to see if I could do it ;P
16:42:47HClack
16:42:48HClmy bad
16:42:52HCllet me fix
16:43:11preglowgood to see i'm not alone in breaking a build on my first day of cvs access
16:43:20Rickhehe
16:43:23wacky_hey dudes ?!? what about my plugin ?
16:43:32*wacky_ cries for his plugin to be added!
16:43:44bobTHClol
16:43:47*wacky_ has inserted a trojan horse to take control of your houses into it :)
16:43:58preglowwacky_: i'd like to hear with the big boys first
16:44:22wacky_right :) ok I'll come and cry again later on :)
16:44:30preglowor at least someone with a bit more authority on what should be added than i have
16:44:30wacky_btw, thanks for your awsome job
16:44:33preglowheh
16:44:59HCli don't understand why x.x
16:45:18wacky_ok, I already have set a wiki page PluginIriverify.. so you can start from there if you meet the big guys :)
16:45:19wacky_see you later!
16:45:35HCldo simulators have different .h's?
16:45:44 Quit wacky_ ("that box rocks!")
16:46:29preglowHCl: oh, most certainly, sims are complex
16:47:09tncpreglow: if you still want to take a look at the source, it's there : http://users.skynet.be/bk240411/rocktrex.c
16:47:53tncit's somewhat ugly, as i put all the different parts of the emulation in one file :-$
16:48:13*HCl goes to try to fix, has terrible toothache x.x
16:48:49tncit's my first try at a bigger plugin, so i may be doing really stupid things i shoudn't do
16:49:05tncit works great in the win32 sim though
16:49:49tnc(except the gfx are a bit...er... compressed? due to the resolution)
16:50:02HClok, got a temp fix for it
16:50:40preglowtnc: hmm, can't see anything obvious
16:50:52preglowtnc: rockbox should handle this code size well, rockboy is bigger
16:51:49tncthat's what i was thinking too, but i thought that maybe i forgot something (change the Makefile or other params...)
16:52:28tnci just had to add the .c filename to the SOURCES in my apps/plugins dir
16:53:03preglowthat should be enough
16:53:32tnceven in the .map file, it states that plugin size is c0000 ... so i don't know what to do
16:54:39tncmaybe my m68k gcc setup is still buggy...
16:55:21HClfixed
16:57:23preglowHCl: goodie
16:57:42HCltemp fix, but its a fix.
16:58:37preglowyes, just so long as it build
16:58:37preglows
16:58:45HClit should
16:59:01preglowyes, it does
16:59:02preglowi just did it
16:59:10preglowanyone know of an adequate x server for windows?
17:00
17:00:02 Join mecraw [0] (~mecraw@69.2.235.2)
17:02:07tncpreglow: you're running linux? if you put my .c file in the apps/plugins dir, add the file to the SOURCES, does the iriver target build work ?
17:03:10HCli run linux, i can try it
17:03:36***Saving seen data "./dancer.seen"
17:03:39tncpreglow: at work, we're using a windows X server called XWinPro v5.3 (i think it was renamed for version 6)
17:04:39preglowtnc: it does not
17:04:56preglowyes, bss, you have too much data some place
17:04:59preglowuninitialized data
17:05:23tncah, ok, at least the problem is consistent (better than those dreaded internal errors)
17:05:38tnci'll look into it
17:05:43tncthx a lot
17:08:28 Join webguest00 [0] (~d463c4a8@labb.contactor.se)
17:08:32preglowi'll go fry myself some food
17:08:34preglowbbl
17:09:48 Quit webguest00 (Client Quit)
17:16:43tnci wonder if the vars declared "anywhere" (not respecting c90) could confuse the linker about the sections...
17:18:59HClcan someone explain to me why the build table just jumped from 14:40 to 16:00 ?
17:19:31preglowno changes?
17:19:43HClhm
17:19:46HClokay
17:19:47preglowit doesn't build unless it has to
17:19:50HClah.
17:20:11preglowand if you think that's strange, you had too much painkiller :)
17:20:26HClactually, they're not working much >.<
17:21:05preglowthey seldom do
17:21:39jypHit your own chest till you oly feel the pain from your blows...
17:21:52HCldoesn't work.
17:22:00HClits hurting too much for that.
17:22:05jyp:P
17:22:19preglowarghh, all out of hp sauce
17:22:30jyphewlett packard ?
17:22:44HClgo to your printer and grab the ink cartridge
17:23:01preglowso, ink is good on eggs?
17:23:02preglowfancy
17:28:19tncpfff, can't understand this linker problem, i have to go, see you later, keep up the good work
17:30:19 Quit bobTHC ("( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )")
17:30:31 Part tnc
17:39:53 Join Tang [0] (~chatzilla@ARennes-252-1-14-86.w83-195.abo.wanadoo.fr)
17:43:02Chamoishi tang
17:43:11TangHi Chamois
17:43:13TangCa av?
17:43:17TangJe lis le topic iHP
17:43:19Chamoisca va
17:43:30Tangbravo et merci
17:43:34Tangpour tes contributions
17:43:38Chamoisnormal
17:43:38Tangje me sentais un peu seul
17:43:42Tang:)
17:43:44Chamoislol
17:43:51Tanget en plus je ne suis pas la svt
17:43:52preglowenglish, please
17:44:00Tangok sorry
17:44:02Chamoissorry
17:44:06Tanghi preglow
17:44:08preglowhello
17:44:21Tangwe were
17:44:31Tangtalking bout a thread on a french board
17:44:41Tangabout iHP1xx
17:45:01Tangusually i used to maintain the information specially about Rbx
17:45:14Tangbut since i've no personnal connection anymore
17:45:19Tangi was less active
17:45:44Tangi thanked Chamois to make the job very actively and efficiently
17:45:45Tang:)
17:46:49TangSeems there were cool progress recently for mp3 decoding according to Chamois
17:46:59Tang50% real time?
17:47:16Chamoisthanks to frequency improvment
17:48:40preglowmp3 decoding is 50% realtime
17:48:44preglowi'll try to speed it up more afterwards
17:48:47preglowflac is above 100% realtime
17:49:03Tang100%?
17:49:06TangVry nice
17:49:23preglowit's around 130% realtime for some tracks
17:49:31preglowi've seen it go as high as 150% realtime
17:49:32TangAh 130% okay
17:49:40preglowand there's still lots to be done
17:49:55Tangmaybe there will be more CPU freq step than the two actual?
17:52:31preglowperhaps
17:53:03TangI'm reading the wiki about this
17:53:05Tang:)
17:53:55 Nick kergoth is now known as kergoth`bbl (~kergoth@li11-226.members.linode.com)
17:58:53TangOkay i understand better
17:59:42Tangthe frequency changing delay is about to be reduced or it's linked to the CPU architecture?
17:59:54HCllinked, afaik.
18:00
18:00:08TangOkay :)
18:00:13TangThanks HCl
18:00:24preglowif you mean the 10ms, then yes, that's unavoidable
18:00:27Tang@Chamois:
18:00:30preglowthat's the time the pll needs to sync to the clock
18:00:38TangYes was talkinga bout the 10ms
18:00:39Tang:)
18:00:49Tang@chamois: are you still here?
18:00:52Chamoisyes
18:01:19Tang:)
18:01:23TangOkay this image:
18:01:31Tanghttp://chamois94.free.fr/IMG_0487.jpg
18:01:50Tangit's a logo or a jpeg viewed with a Rbx plugin?
18:02:06Chamoisa jpeg file
18:02:20Chamoisnamed einstein.jpg on my root
18:02:31Chamoisinto my iHP
18:03:51Tangpkay
18:03:53Tangthanks
18:04:09Tangseems to be a nice shot for a start logo anyway
18:04:16Tang:)
18:19:57 Join pabs_ [0] (~pabs@xor.pablotron.org)
18:21:32 Part shx
18:23:03 Quit pabs (Nick collision from services.)
18:23:07 Nick pabs_ is now known as pabs (~pabs@xor.pablotron.org)
18:25:14pabswelcome to the junkgle
18:25:15pabs-k
18:25:23pabsHCl: you around?
18:26:13 Quit YouCeyE (Remote closed the connection)
18:27:09pabsalright well i have a patch for rockboy, was wondering where to send it
18:27:35markun I have a strange problem. When I use the macro DRAW_PIXEL from lcd_bitmap the bitmaps look all wrong, but when I call the function lcd_drawpixel (which only does DRAW_PIXEL) everything looks good again.
18:28:07HClpabs: yea..
18:28:15HClpabs: what kind of patch?
18:28:20pabsHCl: wrote a menu for rockboy
18:28:26HCloh. sweet
18:28:27HClgimme?
18:28:28pabsHCl: like i was telling you bout hte other day
18:28:34pabsHCl: k, email or what?
18:28:36HClyea
18:28:39HClum, dcc or ftp
18:28:44HClftp://titania.student.utwente.nl/incoming
18:29:06pabsdon't have dcc access easily (i'm firewlaled)
18:29:12 Join [FlaT]Heidel [0] (~h@pD9530F75.dip.t-dialin.net)
18:29:16HClnor ftp then?
18:29:32pabsi can ftp, i'd prefer to send a diff via email though
18:29:39HClmrf.
18:30:01HCli don't really have an email i can use >.<
18:30:10HClftp would be much handier for me :/
18:30:25HCli still have to rewrite that cgi i once had that allowed people to upload files through http
18:31:02pabsthey won't right now either
18:31:03pabserr
18:31:03pabsww
18:31:10pabssorry, wrong window
18:31:28pabsHCl: hw are you committing to cvs and following the dev list?
18:31:31pabs(w/o email?)
18:31:48 Nick Lynx_ is now known as Lynx_awy (HydraIRC@134.95.189.59)
18:31:49HCli'm not following the developer mailinglist
18:31:55HCland you don't need email for cvs :3
18:31:59pabsi know
18:32:33HClhrm..
18:32:43HClcan you try to use ftp? :/
18:32:50pabsyeah, lemme just clean it up here
18:32:54HClk
18:43:18 Nick kergoth`bbl is now known as kergoth (~kergoth@li11-226.members.linode.com)
18:43:47 Join Sucka [0] (~NNSCRIPT@host81-156-214-74.range81-156.btcentralplus.com)
18:52:30pabsHCl: sorry, got sidetracked for work
18:52:31 Quit markun (Read error: 54 (Connection reset by peer))
18:52:31pabsback now
18:54:09HClok
18:54:12HClthats okay
18:54:14HCli got sidetracked too xD
18:54:16HClupload?
18:56:12pabsHCl: no, i got sidetracked right after i started to clean it up
18:56:20HCloh, ok.
18:56:24pabsHCl: i want to do one more test too, since i jus renamed some functions and deleted some cruft
18:56:28HCl*nods*
18:56:31HCltake your time
18:56:44HClwhats in the menu so far?
18:58:28pabsstatic const char *main_menu[] = {
18:58:28pabs "Back to Game",
18:58:28pabs "Load State...",
18:58:28DBUGEnqueued KICK pabs
18:58:28pabs "Save State...",
18:58:28pabs "Options...",
18:58:30pabs "Quit RockBoy"
18:58:33pabs};
18:58:36pabsonly quit an dback to game are functional
18:58:42pabsthere are stubs for load/save state
19:00
19:00:14HClok, nice
19:00:21HCli'll implement the load state and save state
19:02:07preglowwe need a key mapping feature
19:02:26preglowfor times when you need to hold a button as well as move
19:02:44HClwhat?
19:02:47preglowworks well with play button, so should be able to swap them around
19:03:01preglowlike in zelda, you need to press the shield button while you move
19:03:09pabsHCl: i saw the stubs for save_sram/etc
19:03:13preglowand that's fine if the feature is mapped to the play button
19:03:34preglowbut if it's mapped to another button, you quite simply can't do it, so you should be able to remap the keys while you play
19:03:38***Saving seen data "./dancer.seen"
19:06:03pabsyeah
19:09:54Bonkersanyone else with a recorder had that volume control on the headphones die on them?
19:15:56 Quit einhirn ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
19:31:34 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.umbc.edu)
19:41:11 Join amiconn [0] (~jens@pD95D1DC6.dip.t-dialin.net)
19:41:18HClhey ami
19:41:36amiconnhi
19:41:36HClmarkun wanted to talk to you about the grayscale stuff
19:42:10HCloh. and i got this patch for different lcd modes and i can't get it to work and i'm absolutely puzzled by it, think you can take a look at it?
19:46:17amiconnI saw that you got cvs access today.
19:46:22amiconnCongrats :)
19:46:24HClthanks :)
19:46:49amiconnI presume you'll commit the dynarec stuff yourself then?
19:46:54HClyea
19:46:57HCli dunno
19:47:04HClam i supposed to commit work in progress?
19:47:12HCleven if its not working yet?
19:47:20HClobviously, disabled with a define
19:47:41amiconnI don't think so. The only reason would be to let someone else work at it in parallel
19:47:45HClits probably easier if i do, cause we'll be able to work on it together through cvs
19:47:49HClyea
19:48:08amiconn...e.g., adding the SH1 implementation
19:48:10HClwell, it is development i guess.. hold on..
19:48:10HClyea
19:48:31amiconnI guess I can reuse much, I merely have to adapt the asm parts
19:48:49HClyea
19:48:53amiconn(Block calling and the actual instruction implementation)
19:49:49HClthe only moral problem i'm having is that the dynarec require changes to the interpreter to make them work together
19:50:15HClunless i store registers 2-per 1 m68k register
19:50:25HCland those changes make the interpreter considerably slower
19:51:00amiconnCouldn't you make those changes selectable by the define as well?
19:51:10HCli can try.
19:51:16HClit'll take me a while
19:51:22HClcan you take a look at that diff file i sent you?
19:51:34HClcan't get it to work but have no idea why it won't work
19:52:15amiconnIn what way it doesn't work?
19:53:57 Quit Chamois (" HydraIRC -> http://www.hydrairc.com <-")
19:54:17HClamiconn: it won't set my variable to 1
19:56:21amiconnWell, that's obvious.
19:56:38amiconnfb.mode is an int, while rb->button_hold() returns a bool
19:57:44HCli tried it with if(rb->button_hold()) fb.mode=1; else fb.mode=0; too
19:57:46HCldidn't work either.
19:59:10amiconnThat should work imho, only that I would write fb.mode = rb->button_hold() ? 1 : 0;
19:59:29amiconnHmm.
19:59:37 Join default_user01 [0] (~435f4332@labb.contactor.se)
20:00
20:00:58default_user01Hey, I have a quick question: I'm getting an Archos Player 20, and I have read that shuffle works only in the directory you are currently in.
20:01:20default_user01Does RockBox fix this and allow for Shuffle to be used across the drive, in all folders?
20:01:48default_user01Or would I need to make a playlist of all the files first, then shuffle from that to allow this to work?
20:12:10pabsHCl: alright, how do you want this? as a patch, should i jsut tar up my rckboy directory?
20:13:23HClpatch against current cvs
20:13:29HClif you have checked out from cvs, cvs diff -u
20:13:41HClbe sure to run cvs update first
20:14:03pabsyeah, i know how to use cvs, thanks
20:16:24default_user01Does RockBox fix this and allow for Shuffle to be used across the drive, in all folders?
20:16:27default_user01Does RockBox fix this and allow for Shuffle to be used across the drive, in all folders?
20:16:39default_user01Sorry, didn't mean to post that twice.
20:17:21 Quit gromit` (Read error: 104 (Connection reset by peer))
20:17:45 Join gromit` [0] (~gromit`@ras75-5-82-234-244-69.fbx.proxad.net)
20:20:07ScheeI just rediscovered my broken fm recorder... I'll see if I can do something for it
20:20:08default_user01Anyone?
20:20:27preglowdefault_user01: someone will reply if they know the answer, don't worry
20:22:32pabsHCl: headed your way, waht's the public ftpa gain?
20:22:43pabsactually
20:24:52Scheehmm... do anybody know if 3 years old LiIon batteries that have inflated still have a chance to work ?
20:25:32amiconnSchee: I'd better not even try
20:26:24pabsHCl: http://pablotron.org/download/rockboy_menu-20050303.tar.gz
20:26:33pabsHCl: it's in there (menu.c and the patch for the rest of the source)
20:26:41HClokay, thanks
20:26:57CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
20:26:57*HCl is struggling with vim
20:27:02ScheeI know... but I have this broken unit here, and I don't even know why it doesn't work
20:27:13 Join [IDC]Dragon [0] (~idc-drago@pD9E346DB.dip.t-dialin.net)
20:27:20amiconndefault_user01: Rockbox shuffle is per directory or per playlist. To shuffle-play the whole drive, you need to create a playlist. This can be done from within rockbox
20:27:31Schee(well it could very well be only a problem with the batteries...)
20:28:40amiconnSchee: If there is no warranty anymore (I guess so), I'd open it up and connect a lab power supply to the battery contacts to see whether the device itself is working
20:29:12amiconnIf it is, then you just need to get replacement batteries. Iirc they are available from somewhere
20:29:15lolo-laptopamiconn: when you do per dir shuffle, does it let you select a dir to shuffle? or does it do that shuffle the dir this song is in bs that iRiver does?
20:29:17[IDC]DragonHi Jens et al
20:29:30amiconnhi [IDC]Dragon
20:29:51[IDC]Dragonmy Ondio got no Rockboy...
20:29:58amiconnNo
20:29:58Scheethe batteries have opened it up already, when they inflated :/
20:30:06pabsHCl: did you try it out? :D
20:30:07amiconn[IDC]Dragon: There are too few buttons :(
20:30:12[IDC]Dragon(just teasing)
20:30:14HClpabs: hold on, adding dynarec first
20:30:23pabsk
20:30:33Scheebut having a lab power supply would be useful, indeed
20:30:46[IDC]Dragonwe should sell an upgrade MMC with some buttons on it ;-)
20:30:52amiconn[IDC]Dragon: Gameboy emu needs 8 buttons
20:31:36amiconnToo bad, it would even be a bit faster on Ondio...
20:31:43[IDC]Dragonyes
20:32:44amiconnAs you committed the charging patch: I found another (minor) issue.
20:33:02[IDC]Dragonwhich is?
20:33:13amiconnThe debug info short & long deltas seem to be no longer computed. They show as 999999
20:33:39amiconnI'm just recharging my recorder with the new algorithm
20:33:46[IDC]Dragondo you mind posting that?
20:34:11amiconn(standard charger, 2200 mAh cells)
20:34:54default_user01Amiconn: thank you. That's what I needed to know
20:35:49pabsHCl: holler when you get a chance to merge it in
20:35:57pabsHCl: i'm going to send it to the mailing list too
20:35:57 Quit default_user01 ("CGI:IRC")
20:36:18HClok
20:37:03amiconnlolo-laptop: The default per-dir shuffle is done when selecting a track within that dir. There is an option to set whether you want to play the selected track first, the others afterwards in shuffled order, or you want all tracks shuffled, including the one you selected
20:37:39lolo-laptopamiconn: what I meant was: how would I shuffle all files in a when I select a/b/c.mp3 ?
20:38:16amiconnYou mean, shuffling all sub-dirs of a higher-level subdir?
20:38:25lolo-laptopright
20:38:47amiconnThis can be done.
20:39:18amiconnFirst, you need to set 'insert directories recursively' to yes
20:39:27lolo-laptopnod
20:39:37lolo-laptopplaylistwise
20:39:51pabsHCl: oh i fixed a couple of other minor things as well, i changed your is* libc wrappers to faster defines, and i swapped most of the rb->splash calls for faster putsxy() calls
20:39:53HClok
20:39:55HCloops
20:40:00HClokay
20:40:00pabsHCl: so it loads quicker too
20:40:13amiconnThen you would position the cursor on dir 'a', then do a long press of play (iriver: select?), and select playlist->insert from the menu
20:40:36amiconn(while having shuffle activated of course)
20:41:13lolo-laptopnod −− was hoping there was an easier way than that... as it is probably the most common thing I want to do and which the iRiver firmware doesn't support...
20:41:24HClalmost done with adding dynarec
20:41:29HCljust need to test it
20:42:07amiconnlolo-laptop: I think this is quite simple
20:43:29lolo-laptopamiconn: hehe, not bad... much bettre than say iRiver's way of just not supporting it.
20:44:25HClok, done
20:44:28HCllet me look at the menu now
20:45:04Bonkerswhen exactly is dynarec done, as the emulator hits new code or all at once at startup?
20:46:17amiconnI think it has to be done as the emulator hits new code. Otherwise it might run out of ram...
20:46:46amiconnAn lru cache might also be necessary then
20:47:06HClas the emulator hits new code, since you can't follow a programs codeflow otherwise
20:54:58HClok, menu is in cvs - looks really nice, i must say
20:55:18amiconn[IDC]Dragon: Did you already inspect your pile of archoses a bit?
20:55:34 Quit Sucka ("a bird in the bush is worth two in your house")
20:57:53pabsHCl: thanks!
20:58:11pabsHCl: now it won't quit when i'm playing games :D
20:58:30HCl :P
20:58:38HCli had already moved the quit button to mode
20:59:08pabsyeah, i noticed, swapped it back because it seemed to be more intuitive there
20:59:18pabsbut it'll be easy enough to add mappings for everything in the menu
20:59:34pabsthere's an options slot, i thought keybindings and stuff could go in there
21:00
21:02:05 Nick StrathAFK is now known as Strath (~mike@dgvlwinas01pool0-a239.wi.tds.net)
21:02:10BonkersHCl: I think you forgot to commit menu.c
21:02:13HClack!
21:02:14HCli did.
21:02:15HClhold on
21:02:43HCldone.
21:03:41***Saving seen data "./dancer.seen"
21:03:43Bonkerssys_rockbox.c:84: error: structure has no member named `button_hold'
21:03:57HClhuh?
21:04:06Bonkerslemme trakc that down more
21:04:48HCloh, are you trying to build for archos?
21:04:51Bonkersyep
21:04:55HCllet me fix that.
21:04:57HClforgot an ifdet
21:05:00HClifdef
21:05:03Bonkersok
21:06:13HClfixed.
21:07:22Bonkersmenu.c:290, you've got a BUTTON_SELECT reference
21:07:48Bonkerslooks like you need the #if stuff from sys_rockbox and change that to BUTTON_PAD_SELECT
21:07:57Bonkerserr, ROCKBOY_PAD_SELECT
21:08:39 Join hubble [0] (hubble@h68n2fls302o1033.telia.com)
21:08:40HClits ok, i can set it to BUTTON_OFF
21:09:17 Join matsl [0] (~matsl@1-1-4-2a.mal.sth.bostream.se)
21:09:43HCllet me testcompile, then commit
21:09:45pabstha's my fault
21:09:47pabssorry
21:10:54HClfixed.
21:11:53Bonkersworks here now too
21:12:12HCljust teaches me to test archos builds too before committing stuff that seems to work flawlessly on iriver
21:12:28Bonkersheh, but what else is that error/warning checker for?
21:12:49HCltrue
21:12:54HClbut
21:13:02HClits meant for people who want to know how stable the current cvs it
21:13:02HClis
21:13:10hubblewhat
21:13:20hubblewhat's holding linus back to commit the uda1380 driver?
21:13:52HCli dunno, did you give it to him?
21:15:34 Join rickst131 [0] (UPP@resnet-51-215.dorm.utexas.edu)
21:17:33amiconnHCl: Mapping the menu to button_off is not very intuitive.
21:17:40amiconn(for recorders, that is)
21:18:01amiconnI think about remapping the recorder rockboy buttons a bit
21:18:11HClsorry, i was kind of desperate to quickly fix cvs to get it to build again :x
21:18:14amiconnMenu is F1 almost everywhere
21:18:50BonkersHCl: are there more than 3 things on that menu?
21:19:07amiconnSo perhaps F1 should be menu, pad A and B moved to F2 and F3, and The start button could theb be ON
21:19:20HClBonkers: return to game, load state, save state, options, shutdown
21:19:21amiconnOff could then still quit immediately :)
21:19:25HClonly the first and the last works
21:19:45Bonkerswell only 3 are visible on the recorder
21:19:55HClcan't you scroll down?
21:20:00[IDC]Dragonamiconn: I'm checking the pile a bit
21:20:06Bonkersnope, selection just disappears
21:20:10HCldarnit.
21:20:17Bonkerssave state... also overlays the bottom of the box
21:20:42 Join bg_ [0] (~chatzilla@216.183.244.90)
21:21:07*HCl sighs.
21:21:19HCli'll make the menu iriver only for now >.<
21:21:19Bonkersheh, take your time
21:21:26amiconn[IDC]Dragon: I could use a player with old lcd, you know. http://funmp3players.com/faq/archos/#hwfw:hwversions could give you some hints which player may be that old, if the serial numbers are still readable
21:21:46amiconnI guess only 6000's are old enough
21:22:40[IDC]Dragonyou want a 3.18?
21:23:08HClurfl
21:23:09amiconnWell, the old lcd is in everything < 4.53
21:23:11*HCl flops on the floor
21:23:12HClokay.
21:23:16HCleverything should be working
21:23:29pabsHCl: yo
21:23:30HClplease try it x.x;
21:23:35pabsHCl: is load/savestate what i'm looking for
21:23:36pabs?
21:23:43pabssaving and loading games?
21:23:43amiconnHCl: You could build an archos simulator
21:23:44HClyes?
21:23:59HClamiconn: yea, but i don't have vnc access at the moment.
21:23:59pabsassuming it is, then i have aother patch for you
21:24:04HClso i can't actually run / see it
21:24:10HClpabs: kay, gimme..
21:24:10amiconnHuh?
21:24:11pabs(that implements the load/save)
21:24:13pabsHCl: alright
21:24:31HClamiconn: i compile on my server, i don't have the internet connection to use vnc at the moment
21:24:47HClso i can build it, but i can't check whether its working properly
21:25:01amiconnAh ok
21:25:06amiconnBagder: r u there?
21:25:39preglowgoddamn, rockboy is starting to get good
21:25:45[IDC]Dragonamiconn: I think this numbering scheme doesn't match
21:25:59[IDC]Dragonmost of mine start with 03xxx
21:26:22BonkersHCl: should I test it again now or wait?
21:26:43HClBonkers: you can test it again, it should not have the menu on archos anymore at the moment till thats fixed
21:26:49Bonkersok
21:26:55preglowehh
21:26:59HClsince i got my wisdom tooth pulled and its still hurting a rather lot, i'm not really up to looking at it now.
21:27:08preglowHCl: can you press select and move the stick at the same time?
21:27:12amiconn[IDC]Dragon: I think these are even newer. The page stops at rom 4.53, but we know that there are several 5.0x roms
21:27:12BonkersHCl: did you get any vicodin?
21:27:14HClpreglow: no
21:27:18bg_HC1: didnt you get any cool pain killers?
21:27:21Bonkersthat's what really saved me
21:27:24HClbg_: they don't work
21:27:32HClbut its okay
21:27:35bg_HC1: what did they give you?
21:27:35Bonkersthey give you those crappy big tylenol?
21:27:37HClits far from as bad as this afternoon
21:27:39HClibuprufen
21:27:44HCland lots of it
21:27:47bg_blah
21:27:53bg_you need opiates
21:27:54bg_hehe
21:28:02preglowhahaha
21:28:05preglowwho doesnt
21:28:22bg_my girlfriend was nice enough to share hers with me when she got hers pulled
21:28:37bg_i guess ill have to do the same for her when i get mine done...
21:29:23preglowHCl: goddamn, man, rockboy is super smooth
21:29:44preglowhow long since we sat here discussing this the first time?
21:29:51preglowi'd never believed it'd work this fast then
21:29:52HCldunno
21:29:53HCland yea
21:30:00BonkersHCl: works
21:30:05HCli split the changes the interpreter needed for dynarec
21:30:07HClinto the dynarec defines
21:30:11HClthey made it slower
21:30:40amiconn[IDC]Dragon: There must be something wrong with the new powermgmt code on Ondio. I get the 1-minute '?' period when the battery type is set to NiMH, but not when it is set to Alkaline
21:36:31preglowlibmad 60% realtime
21:36:38HClnice
21:36:39HCl :)
21:38:01preglowonly a small change as well
21:38:07HClmm?
21:38:42bg_at what point will you guys consider rockbox for iriver good enough for use by the general public? Like, how many codecs running over 100% realtime?
21:39:02preglowbg_: all of the ones that are included, of course
21:39:09preglowbg_: and they'll have to be well over 100%
21:39:13preglowbg_: more like 200%
21:39:17bg_oh
21:39:25bg_so you got a ways to go yet...
21:39:29preglowof course
21:39:35preglowbut yes
21:39:39bg_what does that include? flac, ogg, mp3... anything else?
21:39:45preglowcodec api and a couple of codecs, and we'll start talking
21:39:56preglowwe probably won't include every codec we've got now
21:39:57preglowwe'll see
21:40:00preglowmp3 is a no-brainer
21:41:00 Quit Tang ("Chatzilla 0.9.66 [Mozilla rv:1.7.5/20041108]")
21:45:04Bonkerswhat's the deal with the iriver? no hardware decoder chip or no docs on it?
21:45:26HClno decoder chip.
21:46:08Scheewhat formats does the original firmware play ?
21:46:25HClogg, mp3... wma..? not sure
21:46:47preglowwav, vorbis, mp3, wma, asf
21:47:21 Join XShocK [0] (~XShocK@pcp09492659pcs.nrockv01.md.comcast.net)
21:47:47preglowi've got libmad running at 82% realtime
21:47:58*HCl prods pabs
21:48:02HClnice :)
21:48:22preglowthe best thing is i've barely done anything at all
21:48:27HClwhat did you do? :P
21:48:28preglowput in my opts, put some code and data in iram
21:48:29preglowthat's that
21:48:32HClah.
21:49:05 Join frank_ [0] (~frank@p54A1689A.dip.t-dialin.net)
21:49:20 Quit Ka_ (Connection reset by peer)
21:49:34HClamiconn: dyna is in cvs, by the way, if you want to take a look at it
21:49:47amiconnSaw that already :)
21:51:23 Join Ka [0] (~tkirk@pcp0010733332pcs.howard01.md.comcast.net)
21:51:26preglowgod damn
21:51:35preglowthere has to be a better way of doing this huffman decoding :/
21:52:46lImbuscould someone explain me in a few words what dynarec is ? i googled a bit around, but did not get it. I see loooong defines with something like interpreting and emulating opcodes :-%
21:53:53 Join XavierGr [0] (no@vdp03936-noc01.cos.internet.gr)
21:56:23BonkerslImbus: translates GBZ80 instructions to native instructions and stores them instead of manually emulating each and every instruction
21:57:59preglow86% realtime... so close...
21:58:48XavierGr86% realitime in what preglow? :)
21:58:52amiconnHCl: Hmm, it looks like adapting dynarec is more work than I thought. There are many 68k instructions that have no equivalent on SH.
21:58:57Bonkersdynarec = dynamic recompilation
21:59:02amiconnHowever, I think it is doable.
22:00
22:00:03preglowXavierGr: libmad, the mp3 decoder
22:00:15XavierGrwooow nice!!!
22:00:50preglowi'm just experimenting to see where the gains are to be found
22:00:52XavierGrSo can you actually hear anything or it is just the encoding process in the cpu?
22:00:59preglowjust decoding in cpu
22:01:02preglowit's probably a bit faster as well
22:01:06preglowit's writing to the disk now
22:01:11 Join redcow [0] (~redcow@host120-14.pool80183.interbusiness.it)
22:01:27redcowhi
22:01:47amiconnHCl: Do I understand correctly that you defined a function for every possible 68k asm instruction to add this instruction to the block, and then combine them in dynamic_recompile() (partly using macros) ?
22:02:34hubblepreglow: aa, using dma to transfer to sound chip takes no cpu compared to writing to disk i guess?
22:02:47Bonkerswhy every 68k instruction and not every GBZ80 instruction?
22:02:51preglowhubble: don't think so, but i'm not sure at all
22:02:55Bonkersor is that just for portability?
22:03:03preglowhubble: hard disk is in pio mode, i think
22:03:35preglowhmm
22:03:38preglowi'm getting weird errors here
22:05:07preglow#define IRAMSIZE 0x8000
22:05:16preglowthis has to be wrong, that's just the first iram block
22:05:25*preglow summons LinusN
22:07:12amiconnpreglow: Iiuc this is all the IRAM the plugins may use, the other part is used by rockbox itself
22:08:08preglowhrmf
22:08:15preglowi hate being this dependant on iram
22:08:19redcowum rockboy with dynarec seems a lot slower than before
22:08:44pabsHCl: i have state loading and saving working from the menu now
22:08:54pabsHCl: that's the patch i was talking about, i'll send it in a second
22:09:13preglowamiconn: i really hope the requirement isn't going to be that large
22:09:51preglowbut yes, the codecs will probably just have around 32kb iram each
22:09:52preglowgah
22:09:53preglowi hate this
22:09:57hubblepreglow: dma transfer takes virtually no cpu, but the ide driver might use dma too?
22:10:03preglowhubble: i have no idea
22:10:06HClokies
22:10:34amiconnHCl: Got my comments/ question about dynarec?
22:10:45HClno
22:10:46HClhold on
22:10:47HClsorry
22:10:50HCli was playing mario
22:10:50HClxD
22:10:59HCli got to level 3!
22:11:09HClamiconn: yes
22:11:19HClthe DYNA_* instructions actually do the encoding
22:11:23redcowHCl, with the current rockboy speed?
22:11:28HCland you can pretty much use them as if you're coding in assembly
22:11:30redcowrespect =)
22:11:31HClredcow: yes
22:11:36HClredcow: um. its really fast.
22:11:41HClalmost near full speed
22:11:53coobcool
22:11:56HClohhh. piramid
22:11:57redcowwith dynarec?
22:11:58coobnow get sound working :)
22:12:02HClwithout
22:12:17HClmhz increase + moving to iram
22:12:40redcowah yeah :), you got that release on your ftp? cuz i overwrite my old one with the dynarec version and now its slow as hell
22:12:51amiconnHCl: Okay. Then there will be significantly less DYNA_* functions for sh.
22:12:58redcowand have to catch some pokemons :p
22:13:06HCldynarec version doesn't even run gameboy roms yet.
22:13:17HClrockbox.zip on my ftp contains a current cvs build of it
22:13:21 Part hubble
22:13:26DeadManThis is funny. A music player getting faster development for gaming than for music :)
22:13:42coobyou've obviously not heard of the ipodlinux project :)
22:13:55DeadManlol
22:13:57redcowwhat about the batterie status code, so difficult or nobody cares ? there are no changes made
22:14:01coobmore games and useless graphics demos for it than actual music stuff...
22:14:18 Join LinusN [0] (~linus@labb.contactor.se)
22:14:20HClcoob: thats cause they can't get mp3 to work properly ;/
22:14:26LinusNpreglow: you called for me?
22:14:27DeadManI don't mind people working on that stuff if that is their specialty
22:14:34HCldoh
22:14:38preglowLinusN: haha, i didn't know summoning works in real life :P
22:14:38amiconnHCl: (1) No differentiation between address & data registers (2) No opcodes for 8 and 16 bit operations in registers. All register operation are 32 bit
22:14:41coobHCl: yeah dual cores with odd memory access :(
22:14:42DeadManI can't do nothing so I can't complain :)
22:14:55preglowLinusN: i was just wondering about the iram limit, and thought maybe you had made a mistake
22:14:59preglowLinusN: but that's probably not the case
22:15:13LinusNyou mean #define IRAMSIZE 0x8000?
22:15:17preglowyup
22:15:22LinusNin plugin.lds?
22:15:24preglowyes
22:15:32HClnot seeing the floor in mario is really a handicap
22:15:34LinusNthat's intentional
22:15:51LinusNit's 64k for the application and 32k for the plugins
22:15:54DeadManand it's free. not like they are iRiver devs or anything. so I have doubly less to complain about. however that does not stop me from wishing the music side of things was moving faster but that's just me being selfish hehe
22:16:00LinusN96k in total
22:16:02amiconnThat's not fair ;( The iriver plugins may use an iram space the same size as the *whole* plugin space on archos...
22:16:09preglowLinusN: we probably won't have much more than 32kb for each codec anyway
22:16:27preglowLinusN: but yes, luckily, placing the huffman tables in iram doesn't seem to do much, that's a relief
22:16:32LinusNpreglow: this solution is temporary anyway
22:16:49preglowbut i've got it at 86% realtime placing other important stuff in iram
22:16:51LinusNthe plugins will be separate from the codecs later on
22:17:10LinusNnice
22:17:21LinusNi'm running in 120MHz, no crashes so far
22:17:23preglowLinusN: and are you certain the __attributee blabla(.icode) scheme works? i can't seem to measure any difference
22:17:39preglowwith .idata i get a measurable difference straight away
22:18:05amiconnThe .icode stuff may be less important because of the instruction cache
22:18:10preglowgood point
22:18:12LinusNyou can check for yourself in the .map file
22:18:37HCldarn, game over
22:18:45LinusNremove the "static" keyword, and the functions appear in the map file
22:18:52preglowyes
22:18:57amiconnMaybe using the iram for data rather than code on the iriver might be the better usage
22:19:04LinusNit depends
22:19:15preglowamiconn: mostly, i'd say you're right
22:19:31LinusNif the code is unrolled, iram might be better
22:19:55preglowi'm only unrolling where i have to now
22:20:14DeadManLinusN did I hear you got iRiver Coldfire running at 120Mhz without crashing thus far?
22:20:19preglowlike where i can fit entire coef sets in a register set
22:20:19LinusNyes
22:20:22amiconnWell, only if it's unrolled so much that it is single-pass. I doubt that a fully unrolled codec would fit...
22:20:25DeadManCongrats! :)
22:20:35DeadManThat's pretty cool
22:20:53preglowit'l be a nice performance boost to have available
22:21:20LinusNthe plan is to have only two frequencies, NORMAL and MAX
22:21:29amiconnLinusN: You would need to check whether the 120 MHz are stable at 40 degrees as well. May be difficult at the current season...
22:21:37DeadManso it can go to max as and when needed
22:21:40preglowhaha
22:21:49preglowi can check if it's stable in -15C for you
22:21:57DeadMan;)
22:21:58LinusNi'm charging at the same time
22:22:05DeadManit's cold here too but not that cold...yet
22:22:23DeadMancharging too? that would be warm
22:22:23LinusNsuper mario just flies
22:23:23preglowbut yes, putting the main libmad structures in iram did not surprisingly provide a very handsome boost
22:23:40DeadManWill there be a preview release for the n00bs once it's running stable and libmad is up and running? Or are you only gonna release something once it's most of the way there or even complete?
22:24:03HClpreview release == daily build
22:24:05LinusNDeadMan: we are constantly releasing, every day
22:24:23preglowoh, and by the way, while i remember to ask
22:24:28preglowwhat features should our patcher have?
22:24:28DeadManyeah I know. but people say not to use it unless you want to be on the cutting edge or a dev
22:24:36preglowright now it patches and checksums
22:24:40preglowthat's that
22:24:44DeadManI don't wanna risk a bad flash and also I don't know how to compile it anyhow
22:24:47 Quit elinenbe_ (" HydraIRC -> http://www.hydrairc.com <- IRC has never been so cool")
22:24:58HClyou can't really risk a bad flash since only the bootloader is flashed
22:25:07LinusNDeadMan: when we feel confident, we will release an "official" boot loader
22:25:21DeadManyeah that's what I meant
22:25:38lImbusBonkers: dynarec: so it's solely for speed ? is dynarec the name of the technology or is it the name of the gnu-project it's been stolen (tm) from?
22:25:51LinusNright now, it would be a burden for us to let the crowds try rockbox
22:25:55preglowlImbus: it stands for dynamic recompilation
22:26:02HClthe name of the technoligy, what gnu project
22:26:04DeadManI can dig that
22:26:06lImbusyeah, I just saw that line, thanks
22:26:21DeadManI would rather the remote was up and running too to be honest
22:26:31preglowyes, we get more than enough "status requests" as it is
22:26:35LinusNpreglow: patch and checksum sounds enough, did you have anything else in mind?
22:26:54HCldownload iriver firmware
22:26:59 Quit redcow ("Verlassend")
22:26:59preglowLinusN: not really, no, someone requested automatic download of firmware, but i really don't want to spend time coding that
22:27:10LinusNabsolutely not
22:28:10 Nick coob is now known as coobert (pen0r@host-84-9-63-253.bulldogdsl.com)
22:28:23DeadManI won't say anymore since it's not really worth asking. It's done when it's done. :)
22:30:18 Nick coobert is now known as coob (pen0r@host-84-9-63-253.bulldogdsl.com)
22:32:12 Quit XavierGr ()
22:33:34 Quit preglow (Read error: 54 (Connection reset by peer))
22:34:37 Join preglow [0] (thomj@s183a.studby.ntnu.no)
22:34:47amiconnHCl: Concerning your latest commit - if you would place the opening brace of a block on a separate line, you would only need to #ifdef one line, instead of two
22:34:49preglowi will kill that switch one day
22:35:08HClwhat do you mean..?
22:35:21amiconnYou did:
22:35:28amiconn#ifdef BLAH
22:35:36amiconnif (blurp) {
22:35:38amiconn#endif
22:35:40 Join michiel_ [0] (~michiel@heren.demon.nl)
22:35:44amiconndostuff();
22:35:48*HCl nods
22:35:49amiconn#ifdef BLAH
22:35:51amiconn}
22:35:53preglowopening brace on a separate line is an evil practice and should be abolished
22:35:53amiconn#endif
22:35:54HClyea
22:36:05HClo.o.
22:36:10amiconnIt would look better this way:
22:36:15amiconn#ifdef BLAH
22:36:19 Join ashridah [0] (ashridah@220-253-121-78.VIC.netspace.net.au)
22:36:20amiconnif (blurp)
22:36:23amiconn#endif
22:36:24amiconn{
22:36:26HClohh.
22:36:27HClright.
22:36:28amiconndostuff();
22:36:29amiconn}
22:36:37HClokay
22:37:31amiconnpreglow: I actually think the code looks much cleaner with the opening braces on a separate line. I always do it that way, and the majority of rockbox code agrees with me
22:37:42preglowhaha
22:37:44preglowi don't like it
22:37:56preglowi used to write code like that about eight years back or so
22:38:35preglowbut bah, doesn't matter
22:38:39BonkersI think it makes finding the opening brace way easier
22:38:47amiconnyup
22:39:02preglowindentation makes that more than clear enough for me
22:39:02amiconnPlus, it doesn't look "squeezed"
22:39:53Bonkersya, indentation does too, but it looks more like a block to me being on a seperate line
22:40:18preglowi like stuff like if (blah) { \n woopwoop(); \n } else { \n behbeh(); \n }
22:40:40Bonkersexacatly how I do it too
22:40:54Bonkerserr, scratch that, I do \n else \n
22:41:19Bonkersif (blah) \n { \n woopwoop(); \n } \n else \n { \n behbeh(); \n }
22:41:28Bonkersof course xemacs does it all for me, so it's no effort
22:41:34preglowhaha
22:41:39preglowdon't start me on emacs :)
22:41:44HCl22:15 <XavierGr> ooh and keep up the good work, thanks to you I have an mp3
22:41:44HCl player plus a gameboy!!!!!
22:41:48HCloops
22:41:52HClaccidental paste
22:42:06preglowbut ok, should i include a rockbox logo in the patcher, or something?
22:42:10HClbrb.
22:42:13HClpreglow: sounds nice
22:47:09ashridahHCl: is there a way to choose which lines get dropped yet?
22:47:35HClashridah: no, i have something in cvs for it, but its malfunctioning for unknown reasons.
22:48:19HCli'm hoping amiconn can take a look at it :3
22:48:24HClor someone else :x
22:48:26LinusNi have increased the max cpu frequency to 120MHz in cvs
22:48:53HClnice
22:48:59HCllet me try that with rockboy..
22:49:03preglowLinusN: great
22:49:19*Bagder pats LinusN on his back. Good work!
22:49:23ashridahLinusN: any worries about overheating at that speed?
22:49:30ashridahor is it just 140 that things are going too far?
22:49:42LinusN120MHz has worked for me so far
22:49:45*amiconn spots an active Bagder
22:50:03Bagdersemi-active
22:50:07*LinusN sucks at playing mario bros
22:50:14 Quit lImbus (Remote closed the connection)
22:50:20amiconnBagder: It seems buildzip.pl needs an addition and a fix
22:50:30HClpabs: *prod*
22:50:48LinusNpreglow: would be nice to see how libmad performs in 120mhz
22:50:50Bagderamiconn: addition for the .ovl and... ?
22:51:00preglowLinusN: i'll check it out now
22:51:17amiconnIt needs to include rockboy.ovl in .rockbox/viewers for the recorder targets, and it seems that it drops one of the 2 lines from viewers.config that refer to rockboy
22:51:21 Join lImbus [0] (~manuel@kernel.cycos.net)
22:51:27HClLinusN: 120mhz with open or closed case?
22:51:52amiconnBagder: Another necessary fix: don't include the executable in the zip for the simulators
22:52:00[IDC]Dragonrockboy.ovl will duplicate the zip size, or so
22:52:09amiconnI don't think so
22:52:18amiconnThere is much bss spacve in it
22:52:30amiconnThe code itself is ~65 KB
22:52:33[IDC]Dragonwell, uncompressed it's big
22:52:38[IDC]Dragonok
22:52:42amiconncode + data + rodata, that is
22:53:03Bagderbtw, I really want to make rombox a build option in the configure script
22:53:08LinusNHCl: on my 140, never opened
22:53:51[IDC]Dragoncan somebody privately tell me an exhaustive source of ROMs?
22:54:21HClLinusN: ok
22:54:32amiconnBagder: Why that? This way, you could only build either rombox.ucl or rockbox.ucl. They are linked differently, and you'd need to link both ways anyway for a rombox distribution
22:54:47amiconn...because the supplied .ajz needs to be linked for ram execution
22:54:56Bagderamiconn: to better allow #ifndef ROMBOX etc
22:55:08Bagderto be able to keep doing rombox even when rockbox grows
22:55:26[IDC]DragonI'll do a loader, one day...
22:55:33preglowwhat the hell
22:55:42HCl?
22:55:47preglowmake zip just freezes
22:55:54[IDC]Dragonmini-rockbox as firmware 1, I mean
22:55:58HCli built one.. its on my ftp
22:56:02preglowyes, mpa2wav is 800 megs
22:56:08LinusNpreglow: your binary might be very large
22:56:11HClbut ok, it might not have your libmad stuff
22:56:20preglowwhat the hell did i just do to make it 800 megs big
22:56:26LinusNcheck the map file
22:56:40Bagderdoesn't rockboy build on ondios?
22:56:53amiconnBagder: I think this isn't necessary. Either rombox fits, or it doesn't. I'd vote for throwing out some stuff from the debug menu in general though
22:57:14amiconnBagder: No rockboy on Ondio - lack of buttons :(
22:57:21Bagderoh
22:57:38amiconnYes, too bad. It would be even a bit faster....
22:57:39Bagderthe exe in the sim zip was not fixed because I didn't really care enough
22:57:43Bagderit doesn't matter much imo
22:58:09[IDC]DragonOndio: we're said to have no disk symbol when deleting dirs
22:58:26[IDC]Dragonis it necessary to yield?
22:58:34LinusNthe disk symbol is visible on iriver too :-(
22:58:49[IDC]Dragonbecause it has no LED?
22:58:54LinusNexactly
22:59:03amiconnBagder: Hmm. Imho it does matter somewhat. (1) There is a second copy of the executable lying around. (2) If you accidentally start it from there, it will not work.
22:59:04[IDC]Dragongood then
22:59:19LinusNHAVE_LED is undefined, because it has no software controlled led
22:59:21preglow*default* 0x00000000 0xffffffff
22:59:23preglowis that it?
22:59:23[IDC]Dragonjust make a different icon
22:59:28preglowthe last is length
22:59:36[IDC]Dragondepending on HAVE_MMC or not
23:00
23:00:07LinusNpreglow: let me see the map file
23:01:00preglowLinusN: gimme a sec, did a reconfigure
23:01:16amiconn[IDC]Dragon: Of course there is no disk symbol when deleting dirs, because there is no status bar then...
23:01:18[IDC]Dragonok, the deletion screen has no status bar
23:01:22LinusNthis might happen if you misspell ".idata" or ".icode"
23:03:12preglowLinusN: thanks for the hint, case closed
23:03:30preglowi had section("idata")
23:03:40*preglow tries stressing down
23:03:42***Saving seen data "./dancer.seen"
23:04:10*LinusN eagerly awaits the 120MHz benchmark
23:04:50preglow102% realtime
23:04:51 Join webguest43 [0] (~d99b54b6@labb.contactor.se)
23:04:55LinusNninja!!!
23:05:02*Bagder waves
23:05:05preglowhaha
23:05:07*preglow gets beer
23:05:14Bagderchampagne!
23:05:18 Quit webguest43 (Client Quit)
23:05:20LinusNsleep!
23:05:23preglowno, beer :V
23:05:31 Join webguest43 [0] (~d99b54b6@labb.contactor.se)
23:05:53LinusNcommit!!!
23:06:26coobsleep-commit on a beer drip
23:06:27webguest43congrats guys !
23:06:35LinusNthx
23:07:12webguest43quick question before sleep: does anyone know how to replace a standard usb plug with a mini-usb one ?
23:07:31webguest43it seems to me that the standard usb has 4 pins, and the mini 5 ???
23:07:36preglowLinusN: me commit?
23:07:45LinusNpreglow: would be fun
23:08:05preglowLinusN: it looks like shit, but i guess that's not critical, i'm in the process of fixing it up
23:08:51LinusNfeels good to have it in cvs, so we can work on it in stages
23:09:33preglowyep
23:09:35LinusNand as a backup, in case you do "make clean" on your .S files again :-)
23:09:44preglowi'll have it commited by tomorrow morning
23:09:48 Join muesli- [0] (muesli_tv@c-180-220-49.cvx-h.dial.de.ignite.net)
23:09:52LinusNfine
23:10:04muesli-g'day mates
23:10:08*LinusN needs about 30 hours of sleep
23:10:11LinusNhi muesli-
23:10:14preglowgod, that sounds nice
23:10:26muesli-hi LinusN, preglow
23:10:33muesli-what sounds nice? ;)
23:10:34LinusNor 30 glasses of beer
23:10:37preglow23:10 * LinusN needs about 30 hours of sleep
23:11:17amiconnhrmf
23:11:22LinusN...which will probably cause 30 hours of sleep... :-)
23:11:25preglowhahah
23:11:27preglowparalysis
23:11:29lImbuslol
23:11:42muesli-beer toxication ;)
23:11:56LinusNwell, have to sleep now
23:12:00LinusNcu tomorrow
23:12:00preglowgood night
23:12:06 Part LinusN
23:13:49webguest43sweet dreams !
23:17:10Bagderand thou shall make thy zips with .ovl files
23:17:20preglowshalt!
23:17:33*Bagder stands in the corner
23:17:58Bagderhm, build time now at 18 minutes
23:18:07preglowyou're estimate is off, btw
23:18:10preglowyour
23:18:23Bagderit is the previous build's time
23:18:33Bagderat least that's what I meant it to be
23:18:35preglowthen the build time changed by four minutes :V
23:18:43preglowwhich is pretty drastic
23:19:59BagderI bet it changes due to different loads on the host
23:20:09Bagderbut four minutes sounds too much
23:20:26 Join ripnetUK [0] (~mirc@82-70-100-230.dsl.in-addr.zen.co.uk)
23:20:35Bagdernow it says 1206 secs for the last one
23:21:06Bagderuuuh
23:21:13 Join DMJC [0] (~James@220-245-162-47-sa-nt.tpgi.com.au)
23:21:27pabserr
23:21:27*preglow kicks back with beer
23:21:28preglowahahaha
23:21:33pabsdoes get_time() work on the iriver?
23:21:37preglowlibflac is 170% realtime at best
23:21:39ripnetUKhi, does anyone here know if a device (a gps receiver) which has a rs232 interface that works with signal voltage +3 to +5v will need a level converter (like the max ones) to connect to a pc?
23:22:34preglowripnetUK: well, rs232 on pc is -15 to 15 volts or something, afaik
23:22:54Bagdergrrr, I need to go down to builds every 30 mins
23:23:17ripnetUKcheers preglow
23:23:19preglowBagder: what kind of box is it on?
23:23:45BagderAMD Athlon(TM) MP 1900+
23:23:48frank_rs232 is -12 to 12 volts. afaik
23:23:53preglowripnetUK: you should check out tolerance levels, it might work
23:24:05ripnetUKpreglow - i checked it out the practical way. It doesnt :(
23:24:23preglowripnetUK: you can get level converter chips easily
23:24:24amiconnBagder: The auto dependency stuff seems to need a significant part of the total compile time. The simulators used to build much faster with the old build system....
23:24:26 Join markun [0] (~markun@bastards.student.utwente.nl)
23:24:30preglowripnetUK: just need a couple of caps and you go
23:24:38ripnetUKpreglow - ive built a converter before to connect a phone to the pc
23:24:57ripnetUKbut i forgot the levels i was converting... i might go through my junk box and try to dig it out
23:25:21ripnetUKnight
23:25:22 Quit ripnetUK (Client Quit)
23:26:30markunamiconn, you still here?
23:26:37amiconnyup
23:27:08markunShould I commit my grayscale stuff? I now have lcd_graymap and lcd_bitmap in lcd-h100.c
23:27:52markuniriver-gray.c (the graylib api) only implements a few functions, so you get a lot of warnings on unused arguments.
23:27:55amiconnI don't know whether this would be good. I have no iriver, so I can't comment on the slowdown.
23:28:38markunbitmaps are a lot slower, all other drawing stuff works fine.
23:28:39amiconnOf course code that gets committed shouldn't cause warnings, but then this is my opinion
23:29:01Bagderit is mine too
23:29:01 Join Renko [0] (~Renko@host81-154-33-155.range81-154.btcentralplus.com)
23:29:14DMJCdid I just read that flac is 170% realtime?
23:29:19amiconnWell, I can accept warnings, but only in very special cases.
23:29:26markunok, then I will not commit tonight. Cleanup a bit first.
23:30:02muesli-g'night fellows..
23:30:04amiconn...like the win32 simulator cross-compiles with oldish gcc. This problem simply does not exist with a 3.xx gcc
23:30:26markunmuesli-: gute Nacht
23:30:49preglowamiconn: only one thing to do, get an iriver
23:30:56amiconn(and it's only warnings about implicit declaration, most likely because of incomplete/ incorrect system headers
23:31:00 Join Camilo [0] (~chatzilla@userca029.dsl.pipex.com)
23:31:06 Join Bluechip [0] (~BlueChip@cpc3-colc1-3-0-cust61.colc.cable.ntl.com)
23:31:41amiconnpreglow: Yeah, I keep thinking about it. I'm still not convinced that it would be a good thing though. No 80 GB model...
23:31:53amiconn(and other concerns of course)
23:31:54preglowamiconn: disk can be modded, surely
23:32:14amiconnThere simply is no 80 GB disk in 1.8", right?
23:32:15preglowi'm thinking about replacing the battery on mine
23:32:19preglowamiconn: correct
23:32:30DMJCactually there is...
23:32:33[IDC]DragonApple gets the first
23:32:38DMJCtoshiba did one recently
23:32:47[IDC]Dragonexclusively or so
23:32:48Bagderthey said Q3 2005
23:32:54Bagderiirc
23:32:54*amiconn likes his 80 GB-upgraded archos recorder, modded with white backlight as well
23:33:08*Bagder has one too
23:33:12Bagder80GB white
23:33:15[IDC]Dragonmy 100 GB is on order
23:33:16DMJCwhat does archos use?
23:33:18Bagderwith *black* bumpers ;-)
23:33:23DMJCthe larger form factor laptop drives?
23:33:37Bagder2.5" yes
23:33:39amiconnThen there is the archos player, and the Ondio...
23:33:51[IDC]Dragonit seems to be already difficult to get iHP140
23:33:58DMJCso preglow... how long till we can play flac audio?
23:34:13[IDC]Dragonthe online stores are tried had one in common: no stock
23:34:20amiconnThe 80->100 GB step isn't that significant.
23:34:23DMJCiHP-140 is the worlds best mp3 player atm imho
23:34:27[IDC]Dragons/are/I've
23:34:52[IDC]Dragonamiconn: but for me it's 60->>100 GB
23:35:05amiconnAh, then it is significant ;)
23:35:22*amiconn waits for the first 120 GB laptop drive
23:35:23pabsHCl: you around?
23:35:36preglowhaha
23:35:41preglowi wish i had gone for the h140
23:35:45preglowbut 20 gig is a lot
23:36:11[IDC]DragonI'd like to have a 137 GB, to "saturate" the USB bridge
23:36:14Bagderhttp://forums.rockbox.org/index.php?topic=478.0
23:36:18Bagdergood thread
23:36:18 Join webguest57 [0] (~1877aba4@labb.contactor.se)
23:36:19Bagder:-P
23:36:26webguest57hey
23:36:32amiconn[IDC]Dragon: Rather 128 GB. I dislike the vendor's disk arithmetics....
23:36:39pabsanyone have a response for my get_time() question above?
23:36:46pabsit seems to be behaving strangly for me
23:36:51preglowBagder: see an excorcist
23:36:53HClslightly
23:36:53DMJCeeew.. emacs
23:37:03pabsHCl: k
23:37:04[IDC]Dragonamiconn: me too
23:37:10HCland no, since the iriver does not have an rtc, its problematic to implement it
23:37:12webguest57how di i read the source code of firmware?
23:37:15*HCl is gonna go sleep..
23:37:16pabsgotcha
23:37:18webguest57do i*
23:37:37pabsHCl: k, i'll send the next round of patches later htis evening (my time)
23:37:38 Quit mecraw ("Trillian (http://www.ceruleanstudios.com")
23:37:44pabsHCl: so i'll ust message you with a URL or something
23:37:45HClk :)
23:37:46HClnight
23:37:47HCl*nods*
23:37:50pabsHCl: and send hte patch tot he mailng list
23:37:50pabsk
23:38:20DMJCwhatever happenned to that lawsuit over drive sizes?
23:38:33 Quit webguest57 (Client Quit)
23:38:40DMJCit's the ONE thing in IT that really sucks
23:42:56 Quit markun ()
23:43:31 Quit lImbus ("bedtime. good night fellas")
23:46:58*DeadMan is in the process of selling a brand new iHP140 to someone
23:48:09amiconn[IDC]Dragon: http://www.amazon.de/exec/obidos/ASIN/B0000UJFU6/qid=1109889476/sr=1-16/ref=sr_1_2_16/302-5633493-9341640 http://www.amazon.de/exec/obidos/ASIN/B0001FAP7W/qid=1109889476/sr=1-20/ref=sr_1_2_20/302-5633493-9341640
23:48:41 Quit muesli- (Read error: 110 (Connection timed out))
23:49:44 Join Chamois [0] (~3e2339dc@labb.contactor.se)
23:50:11preglowso they're still available
23:50:24preglowok, i'll just fight off the urge to get a h140, then continue as usual
23:51:07ChamoisLinusN is not here ?
23:51:09[IDC]DragonEUR 458,80 urgh
23:51:24DMJCEUR?
23:51:26DMJCwoah
23:51:29preglowChamois: just left
23:51:35DMJCthat should be like $500.00 au
23:51:36 Join YouCeyE [0] (foobar@youceye.user)
23:51:49DMJCand au$ is worth ~$0.80 US
23:51:49Chamoisi have a file for him
23:51:54Chamoisabout battery
23:51:58preglowholy shit, they're expensive in norway
23:52:10[IDC]Dragongoodnight
23:52:12preglowChamois: he'll most probably be here tomorrow morning
23:52:17 Quit [IDC]Dragon ()
23:52:18Chamoisyes
23:52:25Chamoisi wait fo tommarow
23:52:33Chamois*tomorrow
23:52:40DeadManI sold this brand new in the box player for 200
23:53:01amiconnDMJC: In that case it's > $500 au for sure
23:53:32DeadManBut since I did not pay for the player (Long story) I am in pocket :)
23:53:38DMJCfor a 140, an outphased model?
23:53:40DMJChell NO
23:54:03DMJCall the color ones are at $690 for 40gb
23:54:05 Quit frank_ ("Leaving")
23:54:14DMJCthat's how much my black and white one cost when i got it
23:54:39DMJC$550 is what i've heard
23:54:44DMJCno higher than that
23:56:27Schee459 EUR = 770 AUD, btw
23:57:37 Quit jyp ("poof!")

Previous day | Next day