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 2002-08-14

00:01:26RipnetUKam i right in saying that the screen buffer is 1 byte for the first 1x8 column on the top left, then another byte for the 1pix wide column directly to the right of that and so on?
00:02:27adi|homegod damnit
00:02:37adi|homeare ppl ever giong to stop the 200 files limit discusion
00:03:23*adi|home wasn't commenting at RipnetUK, he was commenting at the emails
00:06:02 Part aku ("Client Exiting")
00:08:52RipnetUK:) how fast is this thing? is it worthwhile shifting screen data around a byte at a time (in which case you wont be able to easily smoothly scroll in the y direction) or is a bit at a time gonna be OK?
00:09:45Zagorwe've got about 50fps frame rate, so we probably want to be able to go bit-for-bit
00:12:03RipnetUKsorry, are u saying we have enough processing cycles left over to do it bit-by-bit, as opposed to a 8 bit column in one hit?
00:12:13RipnetUKthat would be easier
00:12:16adi|homeRipnetUK: whatcha working on?
00:12:29ZagorRipnetUK: are you talking horizontally or vertically?
00:12:39Zagorstart with whatever is easiest
00:12:55RipnetUKadi - im working on an experimental scrollbox, ie a large window a bit of which can be displayed on the screen at a given time, to allow the entire 3/4 lines of id3 to scroll as a block
00:12:59***Saving seen data "./dancer.seen"
00:13:05RipnetUKzag - vertically, hoz wouldnt be a problem either way
00:13:33RipnetUKzag - you are right... i will start with bit by bit, and optimize if it works/is too slow
00:17:20matslZagor: Playlist browsing seems wanted. I have a simpler idea.
00:17:42Zagormatsl: tell me
00:18:13matslWhy not just display the playlist.index
00:18:46matslIf next song is hit again, just increase the index and display it.
00:19:04matslWhen no more activity. Start playing that song.
00:19:11Zagorumm, that's an int
00:19:17Zagori don't get it
00:19:59matslok. Let say your on song nr 2.
00:20:35matslPress + and 3 is displayed. If you continue holding + the numbers 4,5,6.. etc will fly by on the screen.
00:20:42matslRelease + when you are there.
00:20:57Zagorwell, people want to see the names of the files. not just the number.
00:21:42matslI know but it will be rather slow and some filenames isn't that descriptive either in the beginning.
00:22:39matslWhen using id3 display you will also be confused since you will be browsing the playlist which contains filnames.
00:23:04matslZagor: just an idea anyway.
00:23:25Zagoryes. i'm not sure it will silence the wolfs, though :-)
00:24:10matslZagor: Probably not but probably much easier to implement.
00:25:21matslZagor: besides. When the wolfs see how sliw it will be to browse the playlist they come whining for the fast browse ;-)
00:25:38matslslim/slow ;-)
00:48:19RipnetUKgot the basics working... (at least a seperate buffer, which can be displayed on the main screen, at a given position, and a given amount of it), will continue tommorow... gotta sleep, thanks for yer help cu later
01:44:42 Nick dwihno|party is now known as dwihn0r (dwihno@Bald067.Baldakinen.Umea.SE)
01:44:43 Quit RipnetUK (Read error: 104 (Connection reset by peer))
02:11:08 Join lithermon [0] (
02:13:03***Saving seen data "./dancer.seen"
02:17:08matslTime to go. Cu l8tr today!
02:17:20 Quit matsl ("Client Exiting")
02:24:38 Nick Zagor is now known as Zagor|sleep (
02:41:59 Quit lithermon ()
02:48:39 Join RipnetUK [0] (
03:07:33 Quit Hadaka (Ping timeout: 14400 seconds)
03:09:14 Join BlueOtto [0] (
03:11:36 Quit BlueOtto (Client Quit)
03:12:52 Quit mistax ()
03:58:25 Join TotMacher [0] (
03:58:26 Quit RipnetUK (Read error: 104 (Connection reset by peer))
03:59:51TotMacherwhats the deep discharge ?
04:13:08***Saving seen data "./dancer.seen"
04:17:29 Quit TotMacher ()
05:12:56 Join RipnetUK [0] (
06:13:12***Saving seen data "./dancer.seen"
06:28:32 Join telliott [0] (trilluser@
06:28:33 Quit RipnetUK (Read error: 104 (Connection reset by peer))
06:30:31telliottI must be missing something but how do I record with Rockbox (Archos Jukebox Recorder). I'm pretty sure I downloaded the right version.
06:32:03dwihn0rIt's not possible
06:32:21dwihn0rRockbox does only support playing music ... (at least, right now)
06:35:56telliottI'm still getting used to it. It does load playlists faster than the default software, and more than 999 tracks.
06:36:04dwihn0rYeah :)
06:36:06dwihn0rIt's rather kickass!
06:37:08telliottcan I create playlists "on the fly" while playing music?
06:37:31dwihn0rThat is not possible yet.
06:38:47telliottIs it possible for future firmware to improve record quality, using Lame or uncompressed WAV?
06:39:04dwihn0rwhat do you mean?
06:39:33telliottbetter recording quality than with the default software.
06:39:59PsycoXulif somebody can figure out how to program the MAS's DSP they can do something with it
06:40:30dwihn0rProgramming the MAS is the key.
06:40:40PsycoXulbut otherwise its stuck with the mp3 encoder/decoder chip's default functionality
06:40:42dwihn0rBut it might require a combination of ninja tricks + zen
06:42:21PsycoXultelliott: ChanServ should have msg'd you with a reference to information
06:42:53telliottThat would be cool. I'm moving from a Rio Volt SP-250. This dosen't seem to have all the playback options that I've seen but it does play WITHOUT skipping while walking.
06:43:23telliottI did read the FAQ
06:44:17PsycoXul"`You know they've reintroduced the death penalty for
06:44:17PsycoXulinsurance company directors?'
06:44:17PsycoXul`Really?' said Arthur. `No I didn't. For what offence?'
06:44:17DBUGEnqueued KICK PsycoXul
06:44:17PsycoXulTrillian frowned.
06:44:17PsycoXul`What do you mean, offence?'
06:44:19PsycoXul`I see.'"
06:47:31 Part telliott
06:48:21CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
06:48:21*adi|home screams
06:48:39adi|homei realllly wish ppl would RTFM before coming in and asking questions
06:48:51adi|homecourse i guess ppl having been dealing with that for ages
07:09:13 Join LinusN [0] (
07:09:43adi|homeAHHHHHHHHHHH its LinusN!!!
07:09:46*adi|home runs screamming
07:09:50Mode"#rockbox +o LinusN " by ChanServ (ChanServ@services.)
07:10:03*LinusN feels the power
07:10:35Topic"Don't let this happen to you: " I did read the FAQ <-- telliott (trilluser@ has left #rockbox"" by adi|home (
07:11:19LinusNadi|home: interesting topic...
07:11:24*adi|home nods
07:11:26adi|homei try
07:11:34dwihn0rhellö hellö
07:11:36elinenbetime for me to drink.
07:11:48dwihn0rgood morning Linus
07:11:57dwihn0rWelcome back to *drumroll* #ROCKBOX!
07:12:02adi|homei have a better one... wanna see?
07:12:05dwihn0rYour host today is ADI
07:12:23elinenbeöl öl öl öl öl öl öl
07:12:29elinenbebeer beer beer beer beer
07:12:30dwihn0rWe sure have got a great show up ahead today! Lots of upcoming CVS commits and a couple of red builds
07:12:39Topic""Asking a question answered in the FAQ? Thats a kickin. Asking to ask a question? Thats a kickin"" by adi|home (
07:12:42adi|homehows that?
07:12:59 Join RipnetUK [0] (
07:13:11Mode"#rockbox +o adi|work " by adi|home (
07:13:34adi|homeLinusN: does the logbot do anything other then log?
07:14:24*adi|home points at logbot
07:14:24dwihn0rI want all lines regarding me nuked! :O
07:14:27adi|homecheck the website
07:14:48adi|homesorry.. anything you say in this room is by definiton property of Rockbox
07:14:56adi|homedidnt you get your copy of the NDA?>
07:15:47dwihn0radi|home: Och med tomma burkar ska vi fylla våra säckar!
07:15:53LinusNadi|home: i think the logbot is on the alert for some bot attacks too
07:16:06adi|homedo you happen to have the code for it
07:16:12adi|homeid be interested in looking
07:16:12LinusNat least that is what it was made for a long time ago (Dancer)
07:19:21dwihn0rLinusN: you don't know when the pharmacy opens?
07:20:27>>>"SEEN" used by adi|home ( [snoop prevented]
07:20:40adi|home ./msg logbot SEEN LinusN
07:20:44LinusNthe pharmacy at the central station is open 24 hours
07:20:45adi|home-logbot- You need glasses badly
07:20:51>>>"SEEN" used by adi|home ( [snoop prevented]
07:21:04>>>"SEEN" used by adi|home ( [snoop prevented]
07:21:17>>>"SEEN" used by adi|home ( [snoop prevented]
07:21:55>>>"SEEN" used by adi|home ( [snoop prevented]
07:22:26>>>"top" by adi|home (
07:22:33dwihn0ryou are the almighty web ninja
07:22:35dwihn0rand MAS ninja
07:22:40>>>"TOP" by adi|home (
07:22:41dwihn0rMAS ninja (zen++)
07:23:25>>>"SPELL asleep" by adi|home (
07:24:11LinusNi certainly don't feel like a MAS ninja...(grumble..grumble..distortion...grumble)
07:24:31dwihn0rMisa think you do great MAS job!
07:24:42dwihn0rDistortion on the player models?
07:24:43>>>"calc 1+1" by adi|home (
07:24:50>>>"CALC 1+1" by adi|home (
07:25:47>>>"EXCHANGE LIST" by adi|home (
07:26:11adi|homeEXPLAIN foboar
07:26:17>>>"EXPLAIN foobar" by adi|home (
07:26:24>>>"EXPLAIN god" by adi|home (
07:26:29>>>"EXPLAIN sun" by adi|home (
07:26:43dwihn0rLinusN: so who gives a shit, player = bad, recorder = good :)
07:26:44>>>"hel" by adi|home (
07:26:46>>>"help" by adi|home (
07:27:01>>>"CMD" by adi|home (
07:27:11dwihn0rthe firmware for the player models should show a text
07:27:18dwihn0r"Buy a recorder instead"
07:27:20>>>"UPTIME" by adi|home (
07:27:34>>>"RULES" by adi|home (
07:51:53>>>"SEEN" used by elinenbe ( [snoop prevented]
07:52:04>>>"SEEN" used by elinenbe ( [snoop prevented]
07:59:13>>>"HELP" by elinenbe (
07:59:20>>>"CMD" by elinenbe (
07:59:39>>>"TELLME" by elinenbe (
07:59:48>>>"UPTIME" by elinenbe (
08:10:30 Nick Zagor|sleep is now known as Zagor (
08:10:30 Quit RipnetUK (Read error: 104 (Connection reset by peer))
08:10:33Zagorhey all
08:11:13ZagorLinusN: I've fixed the id3v2 seek, but get a very strange bus error(!) on one test mp3
08:11:22Zagorwant a patch, to have a look?
08:12:37LinusNZagor: make sure that all addresses are even
08:13:13***Saving seen data "./dancer.seen"
08:13:14LinusNmprbuf_read, mp3buf_write and mp3buf_swapwrite must all be even at all times
08:13:27Zagorbut, hmm, why does it get odd. odd...
08:13:50 Nick Zagor is now known as Zagor|breakfast (
08:17:28 Join Bagder [0] (
08:18:20dwihn0rHello, mr Bagder, welcome back to #rockbox! :D
08:18:30Bagderthank you sir
08:19:34dwihn0rYour host today is adi|home, and the topic is "My bisexual rockbox is having an affair with a D.A.P"
08:20:08Bagderwhen does the show start?
08:20:28dwihn0rDunno, we're still waiting for the audience to fill the room ;D
08:20:52*Bagder takes a seat in the front row
08:23:10*LinusN buys popcorn
08:24:57Bagdersomeone in Nigeria needs my help!
08:25:28dwihn0r*roflmao* :)
08:25:57dwihn0rNigerian mail (tm) :)
08:26:35 Join bobTHC [0] (
08:26:41bobTHChi all!
08:27:22Bagdermorning bobTHC
08:27:24dwihn0rcomment allez vouz?
08:27:42bobTHCtres bien...
08:29:50dwihn0rc'est parfait, non?
08:31:28 Join matsl [0] (
08:31:39Bagdermorning matsl
08:39:28 Nick Zagor|breakfast is now known as Zagor (
08:43:34Bagderfunny little thing
08:44:46bobTHCwithout any UI to replace ;)
08:46:07Bagderthe UI *is* replaaceable, just change cassette player! ;-)
08:49:18matslBagder: Organizational Q about the simulator icons. Did some more experiment with it yesterday and though I create the code under uisimulator/common ...
08:49:37Bagdersounds wise
08:49:53matsl... just to realize that this would create the first dependency from firmware to uisimulator!?
08:50:43Bagderseveral firmware functions are replaced with sim stuff when the sim is built
08:51:10matslI had to include a -I../common
08:51:34Bagderah, well the first include file to go into the common, yes
08:51:50Bagderbut that should not be made in the firmware makefile
08:52:03Bagderwe should pass it to it from the sim's makefile somehow
08:52:08Bagderpreferably at least
08:53:12matslI just added it in the x11/makefile
08:53:49Bagderaaaah, right, it never uses the firmware's makefile in the sim build
08:54:52matslBagder: OK. So I'm on the right track then!
08:55:22Bagderit sure sounds so
08:56:04Bagdermorning hes
09:16:57ZagorLinusN: i've been doing some thinking about the bit swapping
09:17:14Zagorif we drop the big list, we win 128KB
09:17:34Zagorthat's 8 seconds of 128kbit mp3 data
09:18:30Zagorso even if the slower bit swap means we have to start the disk a few seconds earlier, we actually gain spindown time
09:19:11 Join Hadaka [0] (
09:19:19LinusNZagor: probably, but we will also slow down the response time
09:19:34LinusNFast forward will be slower, for example
09:19:45LinusNtrack changes will be slower
09:20:06LinusNand the big table is 64k, not 128
09:20:15Zagorit's 64k shorts
09:20:41Bagderof course, during the summer we have shorts ;-)
09:20:42LinusNme silly
09:21:16LinusNZagor: but go ahead and test it
09:22:50Zagorwell, i'm a bit busy :-)
09:26:36matslcu l8ter
09:26:56 Quit matsl ("Client Exiting")
09:28:49 Join RipnetUK [0] (
09:29:10Bagdermorning RipnetUK
09:29:19Bagderhow's your C skills developing? ;-)
09:31:51*adi|home beats Bagder about the head and neck with a soft pliable LART(c)
09:32:07*Bagder falls to the floor
09:34:15*adi|home applauds wildly6
09:34:29*adi|home points to the sign that reads "The beatings will now commence."
09:35:02*Bagder returns to his seat in the front row
09:35:03*adi|home beats LinusN senseful with a soft pliable LART(c)
09:35:34*adi|home shoves a soft pliable LART(c) up Zagor's nose and pulls it back out
09:35:52*adi|home wipes his snot ridden LART(c) on datazone
09:36:05*adi|home says quietly "And that ends act one"
09:36:12*adi|home waits for the applause
09:36:22*adi|home hears the crickets
09:36:34Bagderadi has gone more nuts than usual ;-)
09:36:49*adi|home nods happily
09:36:55adi|homebut at least im entertaining :)
09:37:12 Nick dwihn0r is now known as dwihno|zombie (dwihno@Bald067.Baldakinen.Umea.SE)
09:37:53Mode"#rockbox +ooo webmind Zagor TummGBG " by adi|home (
09:37:53Mode"#rockbox +ooo Synthe mbr|gone PsycoXul " by adi|home (
09:37:53Mode"#rockbox +ooo fragglet Hadaka ironi " by adi|home (
09:37:53Mode"#rockbox +ooo Hes matp elinenbe " by adi|home (
09:37:53Mode"#rockbox +ooo dwihno|zombie datazone datazone-work " by adi|home (
09:37:54Mode"#rockbox +ooo Bagder bobTHC RipnetUK " by adi|home (
09:37:56Mode"#rockbox -ooo webmind Zagor TummGBG " by adi|home (
09:37:57Mode"#rockbox -ooo Synthe mbr|gone PsycoXul " by adi|home (
09:37:59Mode"#rockbox -ooo fragglet Hadaka ironi " by adi|home (
09:38:01Mode"#rockbox -ooo Hes matp elinenbe " by adi|home (
09:38:03Mode"#rockbox -ooo dwihno|zombie datazone datazone-work " by adi|home (
09:38:05Zagorbored, adi?
09:38:05Mode"#rockbox -ooo Bagder bobTHC RipnetUK " by adi|home (
09:38:09adi|homeokay.. enough of that fun :)
09:38:12*adi|home smirks
09:38:19adi|homei just got back from a call
09:38:21adi|homewinding down
09:38:30*adi|home crawls to the corner and twiches for a while
09:43:13*Bagder thinks that today's episode of the "adiamas show" was a mighty fine one ;-)
09:47:31 Join Synthe` [0] (
09:48:02RipnetUKmorning bagder - c skills all coming back to me nicely now thanks :) i used to know c quite well, but have done 4 years of Delphi, which as made me lazy :)
09:48:52RipnetUKgonna see how much I can sneeakly do today at work via VNC :)
09:49:13 Quit Synthe (Read error: 113 (No route to host))
09:49:17Bagderavoiding work is an art ;-)
09:49:48 Nick Synthe` is now known as Synthe (
09:54:26dwihno|zombieIn that case, the people working on this project are Picasso, Michelangelo, van Gogh etc. ;)
09:56:45ZagorProgram received signal SIGBUS, Bus error.
09:56:45Zagor0x0f0003ee in ata_read_sectors (start=0, count=1 '\001', buf=0x909104d)
09:56:45Zagor at drivers/ata.c:187
09:56:45DBUGEnqueued KICK Zagor
09:56:45Zagor187 for (j=0; j<SECTOR_SIZE/2; j++)
09:56:51RipnetUKare we all teenage mutant ninja turtles then :)
09:56:58Zagorbluech :-(
09:57:32BagderZagor: ugha
09:59:29Bagderrockbox is ranked 193 on sourceforge, based on activity
09:59:55Bagderout of 45000 hosted projects
10:00:55Zagorwe were 182 on the 11th :)
10:00:57RipnetUKim not suprised... there is a LOT of updates - i love it - each time I check the email, my mp3 player has a new feature
10:01:20RipnetUKim just hoping that not too much work is done on lcd.c / h during the next couple of days...
10:01:28 Join notch [0] (
10:01:30RipnetUKat least apart from by me
10:01:37BagderRipnetUK: it has been pretty stable lately so I don't expect that
10:01:46ZagorRipnetUK: there will be. i'm rewriting the font code, remember?
10:01:56CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
10:01:56*Bagder is silent now
10:01:57RipnetUKneway, gotta take my 'meatspace' to work now, catch ya later
10:02:03Bagdersee ya RipnetUK
10:02:05RipnetUKmy code is largely seperate new functions anyway
10:02:11Zagorok, good
10:02:23RipnetUKso hopefully shouldnt give diff TOO harder time...
10:02:25RipnetUKc u soon
10:07:58LinusNZagor: any progress with the address error?
10:08:23Zagornothing yet. it's not an odd address.
10:08:24 Nick mbr|gone is now known as mbr (
10:08:34Zagor0x0f0003fa in ata_read_sectors (start=0, count=1 '\001', buf=0x909104d)
10:08:34Zagor at drivers/ata.c:189
10:08:34Zagor189 for (j=0; j<SECTOR_SIZE/2; j++)
10:08:42Bagdermorning mbr
10:08:50Zagorthe start=0 is wrong, which could be gdb or it could be the problem
10:09:30BagderZagor: line 189 doesn't seem likely to be the actual address is it? Or is it a stack problem perhaps?
10:10:15Zagorthe stack trace is ok, and a DEBUGF confirms this is the function that stops.
10:10:29Zagorbut yes, the line is unlikely. probably the next line is the one:
10:10:37Zagor190 ((unsigned short*)buf)[j] = SWAB16(ATA_DATA);
10:10:50Bagderright, seems much more likely
10:10:55Zagorbut buf is ok, and j is 1
10:13:16***Saving seen data "./dancer.seen"
10:15:27LinusNZagor: is this when reading MP3 data or when parsing the tags?
10:15:55LinusNZagor: buf=909104d is a very odd address
10:20:00Zagordoh. it is...
10:21:50LinusNwhat have you done to mess with the mp3 buffer indices?
10:22:08Zagornothing. all I do is add lseek().
10:22:41LinusNsend me the code
10:23:14 Join pyvasene [0] (
10:23:53LinusNZagor: dcc isn't my cup of tea
10:24:00Zagorah, right. sending a mail.
10:26:34LinusNPsycoXul: are you there?
10:28:45LinusNZagor: it looks like an lseek/read bug
10:29:08LinusNPsycoXul: have you experienced the distortion bug lately?
10:29:47PsycoXulit still pretends to have something to do with the state of the headphone jack on start/shutdown
10:29:51PsycoXulbut yeah
10:30:22LinusNcan you install the latest bleeding edge build?
10:31:03LinusNPsycoXul: i think i know why the headphone jack matters
10:32:00LinusNthe ROM firmware may get an rs232 framing error in the startup when the headphones are inserted
10:32:21LinusNand that *may* trigger some code that messes with the MAS
10:33:43PsycoXulok current cvs still distorts
10:33:58LinusNok go to the debug menu
10:34:06LinusNand select "MAS info"
10:34:20LinusNkeep the music playing
10:34:57LinusNdesign code?
10:35:37LinusNise + to go to PIOData
10:36:27PsycoXulok it switches between E1010 and 61010
10:37:52LinusNnext, StartupConfig
10:38:53LinusNfirst plloffset48?
10:39:28PsycoXuljumps around a lot
10:40:27PsycoXuljumps around
10:40:34LinusNand next?
10:40:43LinusNwooo again
10:41:01LinusNok, now press PLAY
10:41:17PsycoXulwent higher pitch
10:41:23PsycoXuland went to CECEB
10:41:24LinusNand play again?
10:41:46PsycoXulback to how it was
10:42:02LinusNstill distortion?
10:42:20LinusNok, outputconf?
10:42:53LinusNll gain?
10:43:04 Nick pyvasene is now known as pyvasene|Kf (
10:43:15LinusNnext should be 0, 0, 8000
10:43:50PsycoXuland yeah
10:44:03PsycoXuland it occasionaly switches to something else and back real quick
10:44:14LinusNyeah, i have noticed that
10:44:42LinusNyour MAS is somewhat special
10:45:08PsycoXullucky me
10:45:14PsycoXuldoes it have super powers?
10:48:39*notch thinks the debug menu is a thing of beauty
10:48:49LinusNit doesn't set the PLL frequencies correct at init time
10:49:16PsycoXuli booted it up with the headphones unplugged
10:49:31PsycoXulwhich got rid of the distortion and the stable PLL values are different
10:50:19LinusNweirdo indeed!
10:50:41LinusNPsycoXul: what ROM firmware revision?
10:51:17PsycoXul32d is 5D9D0 or 2D0DE if i press play, and 32e is CECEB or A2319
10:52:51LinusNPsycoXul: what were the values before you pressed play? 5d9d0?
10:54:20 Join aakil [0] (
10:54:24LinusNok, it seems that the unit is in a weird state sometimes
10:58:56PsycoXulso i just went from
10:58:59PsycoXuldrwxr-xr-x 2 root root 16384 Dec 29 2001 Vespertine/
10:59:00PsycoXuldrwxr-xr-x 2 root root 16384 Dec 29 2001 Vespertine (bad rip)/
10:59:00PsycoXuldrwxr-xr-x 2 root root 16384 Mar 12 02:43 Vespertine [ogg2mp3]/
10:59:07PsycoXuldrwxr-xr-x 2 root root 16384 Dec 29 2001 Vespertine/
10:59:07PsycoXuldrwxr-xr-x 2 root root 16384 Dec 29 2001 Vespertine(bad rip)/
10:59:07PsycoXuldrwxr-xr-x 2 root root 16384 Mar 12 02:43 Vespertine[ogg2mp3]/
10:59:40PsycoXulwith the first one 'Vespertine' on rockbox showed the contents of 'Vespertine (bad rip)'
11:00:16PsycoXulthen after i renamed them rockboxs shows 'Vespertine' as the contents of 'Vespertine[ogg2mp3]'
11:02:26PsycoXuli just moved 'Vespertine' to 'Vespertine [Good]' and now all 3 display their proper contents..
11:03:04LinusNPsycoXul: try this:
11:05:17PsycoXulbefore it was either like
11:05:22PsycoXulloud with distortion, or quieter without distortion
11:05:28PsycoXulthat one is quieter with distortion
11:07:01PsycoXulbut it's 5D9D0 and CECEB
11:07:12LinusNthis is so weird!
11:08:12LinusNgo through the debug menu values again
11:08:50PsycoXuli just rebooted it again and it went away...
11:09:15PsycoXuland in the mas info thing
11:09:22PsycoXulthe first time press play it doesn't change anything
11:10:13Zagoruhhh, this odd address problem is not so easily solved :-(
11:10:54Zagoryou're right, linus. it's lseek's "fault"
11:12:25 Part bobTHC
11:12:51PsycoXulyeah all the values are the same as the first time except the PLL's which are what i said last
11:12:52Zagorthe problem is this: lseek(511); read(2000); *bang*
11:13:20Zagorthe read get's an odd address
11:13:27 Join bobTHC [0] (
11:13:42Zagorhow the heck do we fix that...
11:14:09aakilZagor: It works when you do an lseek(510) though right?
11:14:18LinusNZagor: why would the read get an odd address for that?
11:14:19Zagorthe only thing I can think of is to read into a local buffer and memcpy to the odd address buffer. but that ain't no fun for 1MB mp3 data...
11:14:22Zagoraakil: yes
11:15:03ZagorLinusN: because read() copies single left-over byte from the previous sector and then calls fat_read() to read the next, into the next byte
11:15:36PsycoXulso i don't suppose any of you knows how to setup so that a loaded effect into an sblive dsp effects what you hear from the mic input... heh
11:16:02LinusNZagor: ah, i see
11:16:06aakilZagor: Could this be related to the fact that you need to read info a word-aligned buffer
11:16:19Zagoraakil: that is precisely the problem
11:18:04LinusNZagor: the easy approach is to round off the seek amount in mpeg.c
11:18:21PsycoXulheh guess not.. oh well
11:18:38LinusNwe don't need to skip the *entire* id3 tag
11:18:41Zagoryes, but it will bites us sooner or later somewhere else. it's a file.c problem
11:18:51LinusNZagor: yes it is
11:19:10Zagordoes the mas search byte-for-byte for the header? so a single bad byte doesn't affect much?
11:19:30LinusNno, a single bad byte doesn't hurt it
11:19:55Bagderit is a quick-fix for the immediate problem at least
11:19:58LinusNit will, however, lose the sync
11:20:17LinusNbut it loses it anyway today, when playing ID3 tags
11:20:22Bagderlose? it can't have one before, can it?
11:20:45LinusNsure it can, from the previous song
11:21:04Bagderright, if the following has no id3 tags or oddities
11:21:27LinusNit will lose it anyway because of the V1 tag...
11:21:38Bagderit needn't
11:21:43Bagderyou could stop before that
11:21:45Zagorwe should skip that too
11:21:59LinusNand it *may* be a Good Thing to lose the sync...
11:22:10Bagderwhy so?
11:22:19LinusNotherwise we may hear strange sounds between songs
11:22:39LinusNbecause of the bit reservoir and stuff
11:22:49Zagorare you sure about that?
11:22:55BagderI think it would be better to skip the tags anyway and possibly go off sync on purpose
11:23:14 Nick pyvasene|Kf is now known as pyvasene (
11:23:18LinusNbut the noise would probably not be very audible
11:26:06Mode"#rockbox +o Bagder " by ChanServ (ChanServ@services.)
11:26:22Topic"rockbox - why make it harder?" by Bagder (
11:26:45ZagorI like "how hard can it be?"
11:26:55Topic"rockbox - how hard can it be?" by Bagder (
11:31:06ZagorLinusN: can we do an 8-bit access to ATA_DATA ?
11:32:09LinusNi don't think so
11:32:28LinusNand why would we want that?
11:34:35LinusNi'd rather have read() return an error when the buffer address is odd
11:34:54Bagderwe need to support lseek() to an odd index
11:36:18LinusNbut i'd rather not impede performance because of it
11:36:41Bagderwe need to detect the special occation somehow
11:38:45Zagorwe can read just the first and last byte of the sector 8-bit and the rest 16-bit. that wouldn't affect performance.
11:39:07Zagorbut i guess the ata interface doesn't work like that :-)
11:40:09LinusNi'm sure it doesn't
11:52:11Zagorhmm... we have the exact same problem in id3.c
11:52:46aakilHmmmm... it looks like the two bytes are being swapped due to endianness. Does the result buffer for swap.b also need to be word aligned?
11:54:20Zagorthat's not really relevant, since it's the storing of the 16-bit word in the odd address that crashes
11:55:07Zagorthe result buffer is a register :-)
11:55:58LinusNZagor: an odd-address aware ATA read loop will do it
11:56:16LinusNread 16 bits, write two bytes, read 16 bits...
11:56:35Zagoryes. will be slow, though :(
11:56:38LinusNslow, yes, but non-crashing
11:56:54LinusNand it will only choose to do it if the buffer address is odd
11:57:04Zagori'll test it
11:57:21LinusNso it will have one fast and one slow loop, and choose between them
11:58:02LinusNit can even be smarter than that, by shifting the register contents
11:59:40Zagorumm, you got some code for that? sounds too complex to be fast in my ears
12:00:29LinusNi don't. it would probably require some assembler code
12:04:26LinusNStill, the mpeg code must make sure that it doesn't seek() an odd number of bytes
12:04:39LinusNfor performance reasons
12:09:36Zagorthe dma is 16-bit, right?
12:10:07LinusNbut no
12:11:45Zagorcare to explain?
12:12:43LinusNi think it is 8-bit
12:13:20***Saving seen data "./dancer.seen"
12:13:33LinusNbut we give it addresses in 16-bit DRAM
12:13:46LinusNso it must be aligned
12:14:41LinusNstill, it only reads 8 bits at a time
12:14:42Zagori was just thinking how much chance we have at feeding only the exact mp3 data and no garbage
12:14:59LinusNi don't think the DMA cares
12:16:10LinusNi'd rather ditch that idea and go for performance
12:16:44Zagornow, yes. but it would be nice to only send mp3 data, to minimise the gap btwn tracks
12:16:59LinusNwe can do it half-hearted
12:17:21LinusNwe can load the entire file in memory, but advance the DMA pointer between tracks
12:17:45Zagordoesn't help if the address must be word aligned
12:17:46LinusNwe lose a few K of MP3 data, but it may be simpler
12:18:03LinusNi don't think the DMA address must be aligned
12:18:47Zagorok, nice then. we could just seek to the odd address and then read to mp3buf_write+1
12:20:22Zagorseek(511); read(odd_addr);
12:20:41Zagora little too much low-level knowledge to be really kosher, but it would work
12:21:15LinusNso the read() would copy the leftover byte to the odd address and then perform word reads for the rest?
12:21:40LinusNbut doable
12:22:05LinusNthen it has to store two pointers in the id3 info structure
12:22:16LinusNboth start and end of the track
12:25:18Zagorwhy? id3v1len tells us the end of track (adding that now)
12:25:56LinusNi am silly
12:26:26LinusNi still had the "reead-the-whole-shebang and advance the dma pointer" idea in my head
12:42:00RipnetUKare there any docs on how fotns are stored? are they store in native (column) format? are they padded to a fixed width? im talking about the normal fixed fonts here
12:42:27Zagorthat's being changed right now, so no docs
12:42:45Zagornot even the code is necessarily true :-)
12:43:13Zagorcopy lcd_putsxy() and you'll be fine :-)
12:43:45RipnetUKok, are we gonna dump the fixed width and prop. fonts, and go with only loadable fonts for simplicity
12:47:55 Quit adi|home (Read error: 110 (Connection timed out))
12:48:38LinusNPsycoXul: ready for another test?
12:48:47PsycoXulsure why not
12:53:47PsycoXulwell i can't get it to distort so far
12:54:59PsycoXul4th try still sounds good
12:55:03LinusNdon't worry, the distortion will return... :-(
12:56:27LinusNZagor: i thought you did the checksum thing with the Player settings saver?
12:56:44PsycoXul5th try
12:56:47PsycoXulstill no distortion
12:57:20PsycoXuli tried charging it for a minute
12:57:27PsycoXulbefore the last try
12:57:31ZagorLinusN: yeah, but I forgot to actually ever set last_checksum :-)
12:57:49Zagori'll merge his patch (or that one line of it, at least)
12:57:50LinusNwill you fix it?
12:59:53PsycoXulso i read i think on the mailing list at one point one of you said a vu meter would be possible
12:59:59 Quit aakil (Read error: 110 (Connection timed out))
13:00:05PsycoXulon players too?
13:00:18RipnetUKto whoever is re-writing the LCD code - can we have a generic set of LCD routines which take a pointer to a buffer (and a buffer width) and work on those, so it can be shared between the 'real' LCD buffer, and the virtual ones I am working on?
13:00:30RipnetUKat the moment, i am duplicating all the existing functions
13:03:48PsycoXulLinusN: well it's 10 for 10 distortion-free
13:04:51LinusNPsycoXul: good news, but i just can't understand how this fix could make the distortion go away
13:05:06LinusNcan you reproduce it with an older rockbox version?
13:07:29LinusNPsycoXul: the MAS3507D doesn't have any registers where we can read the current output energy (for a VU meter)
13:08:00PsycoXuland thats how you'd do it on the recorder? there's no other tricks?
13:09:01 Join Blaster_Master [0] (~noone@
13:11:17LinusNPsycoXul: no other tricks. There are two registers (Left and Right) for exactly that purpose on the MAS3587F
13:11:43Blaster_MasterA quick question : (on the player version)is it at all possible to reduce the textsize and incorporate some kind of progressbar (like on the player) ? or just reserve one of the lines of text to graphics(progressbar)
13:12:29LinusNBlaster_Master: what do you want to accomplish?
13:12:32bobTHChe use another dsp ... Micronas MAS 3587F and perhaps it's implemented in
13:13:11bobTHCsorry for the lag...
13:13:16ZagorBlaster_Master: there is no graphics on the player, only characters. although a few characters can be redefined to make very simple graphics
13:13:29PsycoXul8 chars at a time
13:13:35PsycoXuland they can be animated
13:13:35LinusN4 on the older LCD
13:14:39PsycoXuland they're like 5x8 but the bottom line only shows in double-height iirc
13:14:41Blaster_Mastereven a | or > like this >>> = 37,5%
13:15:13Blaster_Masterwould help, or just numbers
13:16:23PsycoXulwe could draw a line accross the bottom of the first 8 chars .. heh
13:16:34PsycoXulwith like a little mark on each end
13:16:55Blaster_Masteroh well it is just be bitching :) , if you dont implement it,,, so what ill just buy the player :oD
13:17:29LinusNYeah, an ASCII progress bar is always possible to do
13:17:37PsycoXulits the recorder's that already have a progress bar i think...?
13:18:09 Nick LinusN is now known as Linus|lunch (
13:18:28Linus|lunchPsycoXul: keep on testing the distortion
13:18:48Linus|lunchtry to get it back with an older rockbox version and then try the new again
13:20:12Blaster_Masterperhaps : >|−−−−-< ... >-|−−−−< ... >−−|−−-< ... >−−-|−−< and so on
13:21:23 Join mistax [0] (
13:22:13ZagorBlaster_Master: for a 4 minute song, it would only move every 20 seconds or so. not very exciting.
13:23:07Bagderif you redefine the char in the mean time...
13:23:15Bagderyou can move it more often
13:23:28Blaster_Mastercorrect but if it is possible to subdevice each - with - then / then | then \ and -
13:23:41Bagderhow man pixels wide are the player characters?
13:23:51Zagori still question the point. numbers say a lot more than a bar.
13:24:12Blaster_Masteryes but , still a bar is fancy
13:24:22PsycoXulwith 8 chars you can have 40 pixels wide
13:24:41Blaster_Masterperhaps an option between number or bars ...
13:26:12BagderI'd say that it would take an implementation for people to try in order to judge properly
13:29:49Blaster_Masterif a cahr is 8 pixels wide and can be changes, why not a one dot at a time it world make 8*8=64 point to move around with. (asuming the char can be altered)
13:30:04Bagderit is 5 pixels
13:30:09Bagder5*8 = 40
13:31:31Blaster_Master4 minutes = (4*60)/40=6 seconds for a update
13:31:55Blaster_MasterTHAT world ..... rock
13:32:07Bagderwell, feel free to submit a patch
13:32:51bobTHCdistorotion or not distortion ? ;)
13:33:17PsycoXulis there already a 5x8 or 5x7 font that can be put into defined-patterns?
13:37:09Bagderwe have such a font as .bdf sure
13:37:10Linus|lunchYou guys still forget that the old LCD has only 4 user-defined characters
13:37:22Bagderthis progress bar would need only one
13:37:27 Nick Linus|lunch is now known as LinusN (
13:37:40Bagderthe rest would be '-' or something
13:37:52Bagderand that single one would be reprogrammed when moved
13:38:07LinusNoh, i misunderstood. i thought you ditched the −−−−|−−- idea
13:38:13 Quit mistax ()
13:38:24Bagderthe −−−−|−−- idea needs only one reprogrammable letter
13:38:43Bagderto give single-pixel precision that is
13:38:51LinusNoh, i misunderstood. i thought you ditched the −−−−|−−- idea
13:39:54*Zagor gets a sense of Deja Vu
13:41:03Bagdertoday's subscriber count is.... 210
13:42:31PsycoXulwell if you're only using 1 custom char, you can push that pixel accross the whole 55 pixel width...
13:45:37 Join matsl [0] (
13:47:07 Join mistax [0] (
13:47:33PsycoXuli got the distortion back on the 2nd boot with my last cvs build
13:47:36*matsl back again with new bios and new USB hardware ...
13:48:06PsycoXulafter several boots again after that with your test build i still haven't gotten it back
13:48:26mistaxok, Linus I'm back, I've been scandisking and complaining a bit about the sound quality issue on the players
13:48:44mistaxand now I've updated to the newest mod
13:48:56LinusNmistax: try my latest hack:
13:49:22LinusNPsycoXul: this is great news!
13:49:34PsycoXulyeah lets see if it works for mistax
13:49:46mistaxold display, right?
13:49:46LinusNby the way, i mailed a bunch of questions to Micronas
13:50:02LinusNmistax: sorry, it's new. you want old?
13:50:18LinusNhang on...
13:51:12LinusNTo provide you appropriate support kindly provide us some information about your company and project (application, features, schedule, volume) as well as the version of MAS3507D you are using.
13:51:45LinusNmistax: try now
13:51:58BagderLinusN: fill in: Haxx, Rockbox, rewriting firmware, MAS3507D and MAS3587F ;-)
13:52:04LinusNand when telling about rockbox they mailed back again
13:52:23Bagderthey did? wow ;-)
13:52:28LinusNwith some application notes about how to use their evaluation board
13:52:40LinusNso i repeated the Q:s
13:52:49LinusNno answer yet...
13:53:06 Nick dwihno|zombie is now known as dwihno (dwihno@Bald067.Baldakinen.Umea.SE)
13:53:51LinusNFreiburg, Germany
13:55:01Blaster_Masterwhen my archos distorts, i turn the volume,bass and trebble down to 0 , reboots and turns them down again until the distortion is gone....
13:55:39Blaster_Masterthe distortion disappears then bass is turned below 3
13:55:46LinusNBlaster_Master: does it help for you to turn the treble up instead of down?
13:55:46Blaster_Masteror dampenes
13:56:12Blaster_Masteron werry little
13:56:30Blaster_Masterthe bass is the "main" thing on mine
13:56:38LinusNBlaster_Master: then you have another type of distortion
13:56:45mistaxyeah defenetely need some work on the bass
13:56:59LinusNBlaster_Master: can you hear the distortion through the Line-Out jack as well?
13:57:52Blaster_Masteryep, it is there also... and is not bieng affected by the bass sttting
13:58:02Zagorbah, my root directory starts at sector 0x25320... fat sure is fun!
13:58:11PsycoXulBlaster_Master: have you tried leaving the bass and volume at the defaults and turning the trebble all the way up?
13:58:37LinusNBlaster_Master: you should try my mod too
13:58:59Blaster_Masterhowever if i put the archos 5.07a software on it is still there!! but when i use the Flat option it all goes away. i the put on the rockbox software and the distortion "resumes"
13:59:14PsycoXulhmm weird
13:59:38LinusNBlaster_Master: then you have problems with your headphones/amplifier
14:00:37Blaster_Masterwerry, i thourght i had broken it, but i fixed as described klater, might it be that the new values (7dB) i added to the allready present value for amplification ?
14:01:06*Blaster_Master spells badly
14:01:44LinusNi don't understand what you mean? New values?
14:03:13mistaxwhat volume have you got, It works fine with 74% and starts distortin at 76&
14:03:17Blaster_Masterwhen rockbox starts up all sound values is set to 70%,7db and 7db , i asume the those values is written somewhere ?
14:03:44Blaster_Master)not now ofcourse, now it saves the values)
14:03:51LinusNyes of course. they are sent to the MP3 decoder and the DAC
14:04:37Blaster_Masterwhat is those values i ADDED instead of set to ....
14:05:07LinusNah, now i see what you mean. you mean that the values are added...
14:05:13LinusNno they are not
14:06:02LinusNthe reason that the distortion returns when you install the rockbox again is that rockbox does not have a Flat setting
14:08:01Zagorboo. fseek() only handles 2G files :(
14:08:13LinusNhow lame :-)
14:09:30Zagorwhat is ultramookie???
14:10:02Bagderseems to be mp3/pda related stuff
14:10:26Zagorseems to be hosted on a pda too, by the speed of it...
14:11:27LinusNthey say that our firmware rocks, is that true? :-)
14:11:51Bagderif you hold it in your hand and moves it back and forth it rocks ;-)
14:12:30LinusNBlaster_Master and mistax hear hard rock regardless of which song they play :-)
14:13:22***Saving seen data "./dancer.seen"
14:13:25Zagoryou can feel a slight twitch when disk spins up. maybe that's what they mean?
14:13:44LinusNZagor: only with the toshiba disk
14:14:04Bagderso it rocks only with toshiba? ;-P
14:14:09Zagoryeah! my extra boosted drag racing disk
14:14:16LinusNi love it
14:15:03Blaster_Masterplease make an option to make the disc shake
14:22:57Zagorbuh, I want to use both unix open(2) and rockbox open()...
14:23:48LinusNZagor: ?
14:24:54Zagormy fat test code runs on the unix box, accessing an image. but the ideal test image, of course, is my usb connected archos on /dev/scsi/host1/bus0/target0/lun0/part1.
14:25:53Zagorbut fseek() (which I use to read sectors from the image) can't handle that big a file, while lseek seems to be able to do that. too bad symbols lseek/open/close/read are already taken by the code I want to test...
14:26:20Zagori'll have to do sime trick like in the sims
14:26:21BagderZagor: #define-mania!! ;-)
14:27:10LinusNBlaster_Master and mistax, is there any change with my new mod?
14:27:35 Join TotMacher [0] (
14:27:38mistaxI just pressed in my litte pin that you plug in the charger on!!!!!!!!!!!!!!!!
14:28:01Zagormistax: bad move...
14:28:02LinusNhi, deadmaker
14:28:19bobTHChello !
14:28:21TotMacher what is deep discharge ?
14:28:44LinusNit waits with the charging of the batteries until they are down to 10%
14:28:48Bagderwe need to add that Q to the FAQ
14:28:55Bagderif the option is gonna remain like that
14:29:21LinusNI think we will do it a little different in the future
14:30:01ZagorBagder: we should at least put it in the manual... :-)
14:30:03bobTHCa real deep discharge whould be great !
14:30:12BagderZagor: hehe, good idea ;-)
14:30:38bobTHCwith screen blinking and led flashing ....
14:31:22bobTHCand in a safemode just to drain the last electron
14:32:01bobTHCwithout risking the hd or register error
14:32:45Blaster_Masterhang on ...
14:32:48ZagorbobTHC: why?
14:33:10bobTHCto preserve battery quality...
14:33:25Zagorthese are nimh, not nicd
14:35:12Blaster_Masterindeed there nis a change, now my display is NOT updateing, it is just showing Jukebox ver : 5.06, but im able to play music ...
14:35:21LinusNBagder: what keys are used on the sims these days?
14:35:34Bagderthe same as before
14:35:38Zagorsame as always, just a few new optional
14:35:39BagderI never remember them ;-)
14:35:46LinusNBlaster_Master: hang on...
14:36:07LinusNBlaster_Master: dosnload again
14:36:11LinusNBlaster_Master: download again
14:36:14Blaster_Mastersame url ?
14:36:42LinusNBagder: but i thought we changed them to avoid the numeric keypad
14:37:01BagderLinusN: we added some keys yes, so you can use the new ones too
14:37:15BagderI can't remember those either ;-)
14:38:30Blaster_Masteryes, now the display works, what am i looking for ? distortion fix ?
14:40:03bobTHCzagor : NiMH batteries handle the "memory effect" much better than NiCd
14:40:03bobTHCbatteries and require less exercise to prevent it, but requier too ...
14:40:29LinusNBagder: maybe we should update the manual
14:40:42mistaxfuckelifuck, now the little pin is hopping around in there!!
14:40:56BagderLinusN: "we" should
14:41:01LinusNmistax: open the box
14:41:06*Bagder smiles
14:41:10mistaxI guess I'm gonna do that
14:41:16*LinusN looks away
14:41:36BagderI think I've done a pretty good job just bringing the initial manual
14:41:49LinusNindeed, it is cool as hell
14:41:49BagderI refuse to become responsible for its maintenance
14:41:59LinusNBagder: hahahahahahaaaaaaa
14:42:23LinusNyou should have thought of that before you made the manual
14:42:29Bagderpeople who do UI work should update the manual too
14:42:35Bagderor I'll hit them ;-)
14:42:45*LinusN sticks to the drivers
14:43:07LinusNBagder: actually it's the download page
14:43:32Bagderwhat is?
14:43:46LinusNthat doesn't mention the non-numerickeypad keys
14:44:13Bagderah, ok
14:44:25Bagderthe uisim/README should too
14:44:56mistaxI'll open mi box tonignt with my little daddy(so he'll be resposibel...)
14:45:11Blaster_Masteryes, now the display works, what am i looking for ? distortion fix ?
14:45:57mistaxcan't test any more today cuz my batteries are empty, HOPEFULLY tomorow!!
14:46:08LinusNok, good luck fixing your unit
14:46:24mistaxI'm gonna need it
14:46:34LinusNand don't turn it on until you have fixed it
14:46:47LinusNthat little pin might shortcut something
14:46:54mistaxyeah I know
14:49:02LinusNBlaster_Master: yes, see if there is any change in the distortion
14:49:22 Nick Zagor is now known as Zagor|away (
14:49:22DBUGEnqueued KICK Zagor|away
14:53:04Blaster_Masterno distortion:) but i dident have any to begin with, i told you i fixed it . and i fannkly dont know how to reproduce the problem
14:54:23LinusNok, fine then
14:54:34LinusNi'll commit my changes and hope for the best
14:54:52LinusNthanks for you help
14:57:01Blaster_Masterthis might sound like bull, but i think the general sound quality improoved ....
14:57:31LinusNit may very well be so
14:57:49Blaster_Masterwhat did you do ?
14:58:01Blaster_Mastermagic ?
14:58:17LinusNi wrote 0 to a register in the DAC that is 0 by default at powerup :-)
14:58:57LinusNand i flipped the WSEN pin before initializing the MAS3507D, to reset it
14:59:46LinusNthe DAC write ensured that the sample data from the MAS is interpreted correctly and not being shifted one bit
14:59:55Blaster_Masterperhaps it gets odd readings at startup (tri-stated signals)
14:59:58LinusNthat might have been the disatortion problem
15:00:30LinusNthe MAS reset was because it seemed to ignore the PLL settings otherwise
15:00:42LinusNand played the music at the wrong pitch
15:00:45RipnetUKwoohoo :) got fonts rendering on my scrollboxes - ive had a thought, the actual buffer can be treated as a scrollbox, with its buffer* happening to be the real LCD buffer, and then we can use the same print/line drawing etc routines for both... what dya think?
15:02:30CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:02:30*Bagder doesn't get it
15:03:15RipnetUKbadger - you mean me?
15:03:23Bagderyes ;-)
15:03:29Mode"#rockbox -o adi|work " by ChanServ (ChanServ@services.)
15:05:44RipnetUKwhat I have done is defiend a struct for the scrollbox info, buffer address, size, current scroll pos, position on screen etc, so we can have mroe than 1 of them. I ahve copied and hacked the line drawing stuff, the SET_PIXEL and CLEAR_PIXEL macros, to take a new first param - a * to the struct, and they now work on those. If we have one where the buffer is the real lcd_buffer, and the width is LCD_WIDTH aetc, those generic routines i made/nicked will
15:06:02RipnetUKive also got a draw_scrollbox function, which renders it, and is called from lcd_update,
15:06:29RipnetUKall working nicely... im a bit worried about the ongoing changes to lcd.c... but im having fun, and my c is all coming back to me now
15:06:49Bagderright, so you can basicly set a pointer to refer to the frame buffer of your choice?
15:08:15RipnetUKyes, but its slightly mroe complex than that, because the drawing routines need to know the size (x,y) of the buffer to calculate the correct addresses for a given x,y
15:08:30Bagderright, I was simplifying
15:08:48BagderI think it is the right way to proceed
15:09:07LinusNRipnetUK: so lcd_update is a little more complex now?
15:09:18RipnetUKcool :) im now going to work on getting it scrolling, and then hook it into tree.c so we can have the entire ist scrolling back and forth...
15:09:59RipnetUKlinux - actually, lcd_update just calls my draw_all_scrollboxes, thats the only change. It copies all the scrollboxes (1 at the mo) into the real buffer, before it does its hardware magic to display it
15:10:17LinusNi'm worried about performance
15:10:20RipnetUKi think it would be too complex if we have multiple buffers to check each pixel to see which buffer it is in
15:10:20Bagderyes, lcd_update() still reads the same frame buffer, right?
15:10:43RipnetUKyeah, but we could find out which region (scrollbox) a given x,y is in as we render the pixels, if my way is too slow
15:11:02RipnetUKafter i get it scrolling left/right, im gonna try it on real hardware...
15:11:07BagderLinusN: it doesn't change performance unless you scroll a lot
15:11:43Bagdersince lcd_update() still works as today
15:12:00RipnetUKim copying bit-by-bit at the moment, so there is toom for optimization there by doing it byte-by-byte like lcd_bitmap does
15:12:29BagderRipnetUK: also, you're likely to want to use the lcd_update_rect() from the scroll thread
15:12:37LinusNbut lcd_update copies the scrollboxes to the framebuffer before updating, right?
15:12:47*Blaster_Master goes away
15:12:53LinusNby Blaster_Master
15:12:57 Quit Blaster_Master ()
15:13:06BagderLinusN: of course that takes time
15:13:07LinusNBagder: so then there is worse performance
15:13:22Bagderthere's no scroll today, so of course a scroll takes more time than none
15:13:25TotMacherwhat is the disortion problem ?
15:13:41LinusNTotMacher: some player units have crappy sound with rockbox
15:14:11Bagderhowever, if the scrollbox is a pre-written buffer the scrolling is only done with thay copy without lcd_puts() being used
15:14:17LinusNBagder: all i want is that lcd_update takes the same time as today when no scrolling is done
15:14:42BagderLinusN: yes, and I believe it will
15:15:03LinusNRipnetUK: copying bit-by-bit???
15:15:41Bagderbit by bit is gonna be very slow
15:16:19LinusNhow about using lcd_bitmap()?
15:16:44Bagderno, writing a custom one is much faster
15:16:48 Join Lowfiler [0] (H@
15:16:58Bagderand works on >16 pixels height ;-)
15:17:04LinusNhi Lowfiler
15:17:05Lowfilerwebmaster here? :)
15:17:10LinusNwe are here
15:17:13Lowfilerfound a bug on the page
15:17:18Bagderwe are all webmasters ;-)
15:17:33Lowfilerthe two download links
15:17:38Lowfileron the end of the site
15:17:43Lowfilerthey are both linked on
15:18:12Lowfilerone should linked on
15:18:45Bagderhehe, right
15:18:49Bagderme fix
15:19:34Lowfileri download gnu-sh right now...
15:20:08LinusNdammit, Bagder! you did it first! :-)
15:20:14LinusNmy commit failed
15:20:23Lowfilerah uhm
15:20:36Lowfilercongratulations to your brilliant software :)
15:20:50Lowfilerare more games planned for the future? :P
15:21:02*Bagder wants breakout
15:21:27*Lowfiler wants snake :
15:21:36Bagderwrite one!
15:21:48notchChaps, Is there a 'playing a song now' bit set globally somewhere ?
15:22:04Lowfileri dont have the right knowledge of programming :)
15:22:17LinusNnotch: mpeg_is_playing()
15:22:31BagderLowfiler: so, its the perfect time to learn! ;-)
15:22:40Lowfilersome links? :)
15:23:04*Bagder looks at his toes
15:24:00Lowfileris the complete rockbox team from sweden?
15:24:05Bagderoh no
15:24:07LinusNnot at all
15:24:09Lowfiler(which is named on the website)
15:24:11Bagderwe're from all over
15:24:17Lowfilerare germans there
15:24:28LinusNthe "core" team are from sweden
15:24:36LinusNthe initiators
15:24:40matslWe must have a game for the player too. Any ideas?
15:24:55Lowfilerya i understand
15:26:44LinusNmatsl: maybe mastermind?
15:27:07Bagdermatsl: "guess a number" ;-)
15:27:13RipnetUKit doesnt seem to slow up on real hardware, and thats without any optimizaion
15:27:28LinusNRipnetUK: good news
15:27:28Lowfilerya hangman :)
15:27:31BagderRipnetUK: that's neat
15:27:36mbrlibmpeg3-1 transcode video-dvdrip
15:27:50mbrsorry :)
15:28:03matslwith hangman I could find use for my text input function as well!
15:28:17RipnetUKgot my name scrolling in a window on top of the tree/wps, name is > width of screen, and graphics (a line) also scrolling correctly :)
15:28:38BagderRipnetUK: *cool*
15:28:57Lowfileris a time-off (sleep) mode for the player planned? - that would be perfect
15:29:15Lowfileri would build it by myself if i had the knowledge :)
15:29:19 Quit ironi (Remote closed the connection)
15:29:19BagderLowfiler: there's a feature-request posted for such a feature, yes
15:30:09 Join ironi [0] (~ironi@
15:34:50LinusNhow do we describe Balance as a number?
15:34:58LinusN0 - 50% - 100% ?
15:35:12Bagderperhaps -X .. 0 ... +X
15:35:15LinusNor -100% - 0 - +100% ?
15:35:53BagderI prefer the latter
15:35:53bobTHCL [−−−−|−−-] R
15:36:03Bagderbob wins
15:36:17LinusNnot much accuracy there
15:36:30LinusNon the player
15:36:32 Join adiamas [0] (~adiamas@
15:36:48 Nick adiamas is now known as adi|home (~adiamas@
15:38:50*Bagder listens to Totmacher ;-)
15:39:18Zagor|awayfor the swedes in the channel:
15:39:21 Nick Zagor|away is now known as Zagor (
15:39:21DBUGEnqueued KICK Zagor
15:42:04bobTHCtranslation mistakes ?
15:42:19LinusNbobTHC: really bad mistakes in general
15:42:54LinusN"This means that the hard drive - the heart of the computer - works fast as lightning"
15:43:23LinusN(explaining the GHz rating)
15:43:34bobTHCouch really big
15:45:58LinusNis percent a good unit for Balance?
15:46:08LinusN-100% - +100%
15:47:48bobTHC2 sec...
15:47:52ZagorLinusN: sure
15:49:40CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:49:40*Zagor loves when people suggest new features without checking if they are already added...
15:50:23Bagderis anyone looking at the rolo problems?
15:50:30LinusNnot me
15:51:24Zagorme neither
15:52:50matslIs balance already added?
15:53:04LinusNi am looking into it now
15:53:34matslI was confused by zagors note
15:53:49Zagori was referring to browsing with the remote
15:53:52LinusNhe was talking about the remote control
15:54:03matslSo I gather ...
15:54:33matslLinusN: yes -100% and 100% for balance is as good as 0-100% for volume. same thing isn't it.
15:54:52*Bagder agrees
15:54:57LinusNok, i'll add Balance this evening
15:55:25Bagderabout the rolo, I think it would be really cool to get that running
15:55:33LinusNyeah, i know
15:55:36Bagderwe could get more people into running rockbox by default
15:56:25LinusNi'll take a peek at it this evening too
15:56:37Bagderyou're a cliff ;-)
15:56:46LinusNi'd like it too, to run the gdb stub easier
15:56:51matslLinusN: An alternative to balance is to have two sliders. One for each channel. Squeeze that one into two lines on the Player!
15:57:20Zagorbalance is much simpler
15:57:24LinusNstill, it has to translate into one or two numbers for the internal processing
15:57:30LinusNthe UI stuff is secondary
15:57:31Bagderwell, with the 55-pixel −−−−|−−- bar it could work
15:57:37*matsl just joking zagor
15:58:20matslI know. I will probably never touch balance except to test it.
15:58:35LinusNtime to go home!
15:58:40Bagdersee ya LinusN
15:58:49 Part LinusN
16:01:22Bagderany windows users doing wincvs around?
16:01:30TotMacherbalance suxx :)
16:09:59 Join Blaster_Master [0] (
16:10:07*Blaster_Master back
16:10:56 Quit Blaster_Master (Client Quit)
16:13:26***Saving seen data "./dancer.seen"
16:28:47Zagorwohoo! things go a lot faster when reading many sectors per call
16:31:33matslZagor: HD related. Would it be possible to do playlist browsing using id3-tags? means reading id3-tags while moving in the playlist.
16:31:45Zagorpossible, but slow
16:32:03matslHow fast can you get it
16:32:08dwihnooch med tomma burkar ska vi fylla våra säckar!
16:32:15*dwihno is officially a P.E.T master
16:32:38Zagormatsl: hard to say. you have to call mp3info() for every single file. maybe 4 files per second
16:33:13matslok. That's to slow I guess.
16:33:27Zagorjust look at how much time xmms/winamp takes to do the same thing, and that's on a pc with fast everything
16:34:01matslZagor: Hmm... right.
16:34:26 Quit TotMacher ()
16:35:04matslZagor: I just dislike the idea of viewing id3-tags and the browse using filenames. But seems like a war we can't win!?
16:35:28Zagori think the answer is a pc-created database
16:35:39Zagorthat enables the much-wanted search feature too
16:36:23matslZagor: Right. We should fill the box with software that is needed in runtime.
16:57:52 Nick Lowfiler is now known as Low[a] (H@
17:01:38matslBagder: Tried to get a standalone version of bmp. Followed your description from the mail-archive. I'm getting compile errors. Is this the way I have to go?
17:02:55BagderI didn't use it standalone though
17:03:08BagderI made it get used from within the sim and printf()ed to stdout
17:03:10matslI see that now when I read again.
17:03:56matslBagder: Something for tools instead?
17:04:18Bagderyes it would be really good to make a standalone bmp => C bitmap tool
17:04:58matslAny comments before I try it.
17:05:35Bagdernot really, just copy n' paste the bmp code and it shouldn't be that tricky
17:05:50matslOK. I'll give it a spin.
17:06:46matslBagder: IRC releated. How come you all answer so promptly? Do you stare at the window all the time or might there be some thang that helps you se that you are wanted?
17:07:18BagderI get a different color on things when I am addressed
17:08:09matslI suspected that. I have tried to configure X-chat to get that but without success. any hint?
17:08:12Bagderbut I do have the window visible almost all the time
17:08:52BagderI have 'colored nicks' activated
17:09:15matslme too.
17:17:43 Quit pyvasene ("using sirc version 2.211+KSIRC/1.2.1")
17:19:23Bagdertime to go, see ya around
17:19:25 Part Bagder
17:21:06 Quit elinenbe ()
17:21:59 Join aakil [0] (
17:24:13bobTHCu think 4x6 fonts is useable and readable ?
17:25:24bobTHCfor the recorder
17:42:39notchbit small for us getting older.... the recorder screen is so small - 4x6 may be readable if it was a bigger size
17:43:38notchin fact how about 20x30 for use when driving?
17:43:39bobTHC5x7 ?
17:43:52bobTHCwhen displaying listing
17:44:26bobTHC20x30 ! it's a bit large
17:44:44bobTHConly 3 or 4 lines...
17:46:02notchbut you are probably about 1m away from the unit when driving :-) Dont want to have too many crashes whilst squinting at the screen :-)
17:46:33bobTHC2 lines max !
17:46:36notchI think you would have to test your font size and see how it looks.
17:46:47notchI feel a patch coming on...
17:50:47*notch notch leaves for the sunny weather here in England... oh.. its cloudy again...
17:50:58*notch damn.
17:51:07 Quit notch ()
17:53:54 Join ah_ [0] (
17:56:21bobTHCfrancais !!!
17:56:41bobTHCde boulogne.....
17:57:08bobTHCc kool! ca fait longtemps ke tu suis le projet ?
17:57:20ah_pv no ?
18:01:07 Quit aakil (Read error: 110 (Connection timed out))
18:02:54bobTHCbye all
18:13:28***Saving seen data "./dancer.seen"
18:15:26 Quit matsl ("Client Exiting")
18:18:28 Join zzuuy [0] (
18:21:33 Quit zzuuy ("ChatZilla 0.8.7 [Mozilla rv:1.0.0/20020530]")
19:01:02 Join lohap [0] (
19:02:23Zagortry the new fat code. file loading is significantly faster.
19:03:32lohaphi all i have created the menu icons for the jukebox and i was just wondering what is the mailing list so people can have a look
19:03:50Zagorlook here:
19:05:49 Quit lohap ()
19:19:20 Quit ah_ (Read error: 110 (Connection timed out))
20:13:32***Saving seen data "./dancer.seen"
20:18:59 Join elinenbe [0] (
20:23:18elinenbehow do I apply that patch that was just sent over email? I use "patch -p0 -i file" but it does not work?
20:23:31Zagorit's not against the latest cvs. i'm working on it.
20:23:45elinenbeah... that is the problem.
20:23:59Zagorlooks like a very good patch, though
20:24:08elinenbewell, Zagor, the rockbox gods wish you good luck
20:24:37dwihno Cool done, by Hardu
20:25:09elinenbeI wonder how hard it would be to make audible fast forward and rewind? :)
20:25:26Zagornot very hard, but this is a great middle-step that I didn't ever think of
20:25:48dwihnoaudible ffw is not required imho
20:25:49elinenbeI think when you hit stop that the audio should fade out quickly −−- just move the volume towards zero.
20:25:58Zagordwihno: of course it is, but this is a good step
20:26:10elinenbethen reset the volume once the audio has stopped. winamp does this and it sounds wonderful
20:26:17Zagorelinenbe: nice idea
20:26:19dwihnosome kind of ffw/rew acceleration would be neat and prove extreme useful for audio tapes
20:26:39 Nick elinenbe is now known as elinenbe|store (
20:26:44dwihnoI agree. nice idea elinenbe|store
20:27:13dwihnoZagor: that font editing software proved a trifle difficult to use... I've gotten quite far with the conversion though.
20:28:22Zagordwihno: ok. maybe there are better programs, i just sent you the first i found
20:28:54dwihnoZagor: It's ok! :) I just had to get an X server and now it works... Thanks for the link btw.
20:30:43dwihnoIs the font loading code ready for propfonts?
20:31:09dwihnoah, okay... it's no stress really...
20:31:09Zagori've been busy with other code
20:31:18dwihnoATA speedups etc. ?
20:31:29Zagoryup, and the id3 tag bug fix
20:31:55dwihnoid3 tag bug?
20:32:07dwihnoIs the speed boost from the ATA stuff big?
20:32:55Zagoryes, quite
20:33:25dwihnoCool! :D
20:33:38dwihnoSpeedups = pleasant surprise :)
20:33:45Zagornow buffer filling takes about two seconds excluding spinup
20:34:19Zagorand the player is probably even faster
20:34:33Zagor(only tested it in debug mode, which is a lot slower due to all debug output)
20:34:37dwihnoThe only thing I think about is the speed difference between the player and the recorder...
20:34:50Zagoryeah, there's something fishy about it
20:35:15Zagorah, there it is!
20:35:40dwihnoI found the fish!
20:35:47dwihnoIs there any DevCon planned?
20:36:13dwihnoIsn't it possible to do a hw mod on the recorder to use the parallel bus instead of the serial?
20:38:40RipnetUKwhere is the fonts array declared?
20:38:52ZagorRipnetUK: firmware/chartables.c
20:38:59RipnetUKcheers Zagor :)
20:39:08dwihnoZagor: it would be kickass
20:39:16Zagordwihno: why?
20:40:24RipnetUKsorry, i meant the array of available fonts, not the actual bitmaps...
20:41:01ZagorRipnetUK: in lcd.c, around line 570
20:43:28Zagorhardeeps is an excellent programmer
21:15:05mistaxis it hard to desolder/solder back the main board?!
21:15:33mistaxI mean is it easy to get a "kall lödning"
21:18:05Zagorit's quite easy
21:23:38RipnetUKta Zag - sorry took ages to respond
21:24:08Zagorno problem
21:27:00mistaxI just accomplished my first dissasembly of my AJB6k
21:27:34mistaxcuz my DC conntact loosened
21:28:15mistaxbut I didn't dare to desolder..
21:28:26mistaxhow many times have u desoldered?!
21:29:53mistaxand soldered back?!
21:30:36Zagoroh, I don't know. 5-6 maybe
21:30:46mistaxI'm gonna buy an external charger tomorrow
21:31:17Zagorif you have the money, go for a quick charger. they are lovely.
21:31:39mistaxthen I'm gonna wait till I've bought some new LEDs and porhaps some sound mod stuff and then desolder an UPGRADE!
21:32:23mistaxu've got any tip?!
21:32:38mistaxhave u done the sound or LED mod?
21:33:16ZagorI mostly use my recorder20, which already has much more powerful sound
21:34:41mistaxyou betrator(alltså förrädare)
21:34:53elinenbe|storerecorder6: the new patch makes me recorder go crazy −− skips to a new song every few seconds.
21:35:09Zagorelinenbe|store: really? wow, strange.
21:35:23Zagorthe ffwd patch?
21:35:42elinenbe|storeyeah. I am getting CRAZY behavior
21:36:01elinenbe|storeit says it is in the same song, but every 2 seconds it loads a new song up.
21:36:38Zagoreven if you never use the search feature?
21:38:21Zagorvery strange!
21:40:27Zagorit works totally flawless on my rec20 and player-old
21:41:41Zagordid you ever try the fat patch before I added the search code?
21:42:42mistaxdoes your backlight work any good on your player-old, zaGOR?!
21:42:54mistaxoops caps-lock sorry
21:43:02Zagoryup, works fine
21:43:12mistaxmine doesn't
21:43:30mistaxmy god how do I handle my player!
21:43:57mistaxthe DC contact is pressed in the backlight broke
21:48:26dwihnoZagor: optimizing the decoding code perhaps
21:48:37dwihnoZagor: but if a hw mod is required it might as well be skipped.
21:49:45dwihnoZagor: but I still think it would be quite neat with soime kind of toggle for the ffw function for let's say ... audio tapes
21:49:51RipnetUKsorry to be dozy, but would you lot prefer if I made my scrollbox stuff a #ifdef, if so, where is the correct place to declare it, config.h?
21:49:56Zagorwhoa! the player fills the buffer in less than two seconds.
21:50:11Zagordwihno: you can configure the stepping size
21:50:30dwihnoZagor: already? kickass! :)
21:50:32Zagorso set it to 30 seconds and it's be very quick
21:51:10ZagorRipnetUK: if could be good during development. just put it in your Makefile
21:51:17dwihnoZagor: is it easily configured, I mean, if a skip to the configuration menu is required in order to activate a faster searching it's a rather dull environment to work in :(
21:51:34Zagordwihno: the config is saved nowadays, you know?
21:51:37RipnetUKzag - thanks
21:51:43Zagorbut, yes, i'm working on repeat accelleration
21:51:58mistaxwhat is that rs232 converter mod goog for?!
21:52:15Zagormistax: target debugging with gdb
21:52:19Zagorit rocks
21:53:01dwihnoZagor: I know, but if I listen to a 5 minute track and then switch to an audio tape, I need to go to the configuration and modify the settings...
21:53:10mistaxand the line in 2 way serial port mod?!
21:53:36dwihnoZagor: ffw speed should be togglable with let's say a function button
21:56:08Zagordwihno: no. it's too special and underused to be on such an important function. how often do you toggle search speed on your cd player?
21:56:19Zagoraccelleration and step size config will be enough
21:56:41dwihnoZagor: how often do you have several audio books on CD? :)
21:56:43Zagormistax: that's what you use for gdb debugging
21:57:30Zagorgoing out a bit, back soon
21:57:33 Nick Zagor is now known as Zagor|out (
21:57:33DBUGEnqueued KICK Zagor|out
22:02:08 Join LinusN [0] (
22:13:34***No seen item changed, no save performed.
22:14:18mistaxLinus, have u done the sound or LED mod?!
22:14:53LinusNmistax: what sound or LED mod?
22:16:16LinusNah, no
22:16:31LinusNi remember we talked about this, yes
22:16:54mistaxwe did?
22:17:59LinusNyeah, you were sitting in your country house
22:18:22LinusNand we looked at leds on
22:18:38LinusNor am i mistaken?
22:19:13mistaxwell as you maybe recall I earlier today presssed in my DC contact on my JB6k
22:19:30mistaxand opened my jukebox
22:19:45mistaxbut didn't dare to desolder
22:20:07LinusNok, is it still open?
22:20:34mistaxcuz I think I'm first gonna get LEDs and porhaps sound mod stuff and fix my DC contact all in one smäll
22:20:47mistaxno, I closed it up, why?
22:21:08LinusNi wanted to see the markings on the DAC and MAS chips
22:21:37mistaxmaybe, MAYBE I'll open it tomorrow..
22:22:17LinusNno rush
22:22:50mistaxok, have u desoldered it any time?
22:23:03LinusNi've lost count
22:23:39mistaxso it's easy?1
22:24:20LinusNyes, with a godd, hot soldering iron
22:25:54mistaxhave u ever got a "kall lödning"
22:26:08LinusNand some soder-wick (don't know what it's called)
22:27:48LinusNkall-lödning? jodå, lots of times
22:28:29mistaxtennfläta what is that?
22:30:34LinusNit's a copper plait (fläta) that absorbs the melted solder
22:32:48mistaxit's like a type of "tenn sug"?!
22:33:31LinusNit replaces the sucker :-)
22:38:12mistaxright =)
22:56:17HesFast forward/reverse - excellent !!!
22:57:49mistaxwhere do u buy a good "tenn fläta"?!
22:58:23LinusNmistax: elfa, for example
23:03:00 Nick Zagor|out is now known as Zagor (
23:03:00DBUGEnqueued KICK Zagor
23:04:20Zagorelinenbe|store: still getting crazy behaviour? did you run the bleeding edge or compile yourself?
23:08:53LinusNZagor: did you reset the config block version to 0?
23:09:54Zagoryes, we never understood why you bumped it
23:10:04 Join Bagder [0] (
23:10:21Zagorand added two bytes for contrast
23:10:50LinusNthe reason was to make sure that the checksum check fails on older versions that has the contrast setting set to 0
23:11:06LinusNthe double contrast was a mistake
23:11:32LinusNto avoud that bug report we got anyway
23:11:54LinusNi just couldn't understand why he got contrast=0
23:12:03LinusNas i had taken countermeasures
23:12:05Zagorok, so we should set it back to 1?
23:12:37LinusNi'll do that
23:13:39mistaxwhich should I buy?!
23:16:10LinusNi can't reach right now
23:32:17 Part LinusN
23:42:37RipnetUKdoes sleep implicitly yield as well? i assume it is a kind of 'yield and I do not want to be called again for x seconds' is that correct?
23:43:02dwihnoWarcraft 3 sure is addictive
23:43:06ZagorRipnetUK: yes
23:43:10RipnetUKthanks Zagor

Previous day | Next day