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

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

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

#rockbox log for 2005-07-06

00:00:42amiconnBranches aren't too costly anyway. The coldfire doesn't depend as much on the pipeline as e.g. a Pentium 4
00:01:28amiconnTaken/not taken clock cycles: Forward: 3/1, Backward: 2/3
00:01:58stripwax_Ok, so not so bad then.
00:11:06stripwax_Oh, ffs. I've just found out that my iriver resetting is nothing to do with my local code changes - I'm getting that behaviour with a compile of cvs head code too. humph.
00:11:30thegeekhmm, does anyone know what PO# is ? (context: visa)
00:12:50stripwax_thegeek - visa as in credit card or visa as in international travel/employment?
00:13:15thegeekcredit card;)
00:13:16bill20r3"purchase order" ?
00:13:36bill20r3internal numbers companies use to track thier spending
00:13:37bill20r3make one up.
00:13:39thegeeksame from another # ;)
00:14:47stripwax_post office number i.e. zipcode?
00:15:26amiconnstripwax_: Do official builds work without resetting your iriver?
00:15:29thegeekI just left it blank
00:15:57stripwax_amiconn - yep. i've just noticed some crufty versions in my local dirs so i'm checking out everything from cvs again
00:16:22 Quit t0mas ("good night")
00:18:51stripwax_wow, yep. must have had some shit code somewhere that wasn't mine ;-) ok, back to work
00:27:16 Quit hardeep ("BitchX: it does a body good")
00:29:41CassandraRight - I gave up on multiple %a tags per line. If anyone else wants a pop, be my guest.
00:33:20 Quit Aison ("( :: NoNameScript 3.72 :: )")
00:34:58amiconnCassandra: I guess this will be almost impossible with the current approach
00:35:30amiconnI'm thinking of the following system:
00:36:07amiconnInstead of basing everything on lines, position every wps element in a bounding box
00:37:15amiconnThe box can contain just about every single element, including scrolling text, which would scroll inside the box
00:37:45Bagdersounds like a very good plan
00:37:51 Quit OnkelJonas ("User pushed the X - because it's Xtra, baby")
00:38:03amiconnAlternating display would be provided externally, defining whether the box is shown or not
00:38:13CassandraI think this is a good idea, ami. Would require significant changes to the WPS file format though.
00:38:36amiconnYeps, and it requires significant changes to the scrolling code as well
00:39:19amiconnThere are elements which would have to adapt themselves to the box size (progress bar, peakmeter)
00:39:29Cassandraamiconn: I can think of ways to fudge it with the current approach, but the code would be hideous.
00:39:59amiconnThe bitmap display is the first that uses pixel positioning, independent of the line concept
00:40:03CassandraIt also paves the way for a spectrum analyzer element.
00:41:27CassandraRight. AFAIK the only other thing anyone wanted for 2.5 was Linus request for more solidity to SongDB support.
00:41:44stripwax_amiconn - hm, maybe we need that scroll_rect fn after all? ;-)
00:42:01amiconnNot really
00:42:26amiconnThere is no real difference between scrolling line based and box based
00:42:29CassandraActually one other thing would be the A-B repeat patch.
00:42:40amiconnJust a few more parameters per scrolling 'line'
00:42:47CassandraHowever, that messes with Archos keymaps.
00:42:57Cassandra(Pitch up/down)
00:43:11CassandraI knew there was a reason why it hadn't been committed.
00:43:13 Join TCK [0] (
00:44:00amiconnCassandra: As Linus said earlier today (or y'day for me). One of the biggest problem with patches is that they are often tailored for one platform, ignoring the others
00:44:22amiconnRockbox is multiplatform, grr!
00:44:35CassandraIt's not going to work anyway.
00:44:38 Join hardeep [0] (
00:44:43stripwax_amiconn - no, I guess not. I'm just still not comfortable with the idea of doing all that pointless glyph-framebuffer copying instead of framebuffer-framebuffer copying when scrolling text
00:44:50CassandraThe WPS context menu will break it.
00:45:18amiconnThat's a thing I have to carefully consider when chaning the scrolling etc - it has to continue to work on player, without too much code divergence
00:45:36CassandraNo, scratch that.
00:46:02CassandraWPS context menu won't interfere.
00:47:22 Join Sucka [0] (
00:47:40hardeepred build
00:47:49CassandraHe's messing with the pitch screen: See
00:48:14CassandraAlso, this patch won't actually work on iRiver, due to the SELECT key not being chordable.
00:48:38 Quit Coldtoast ("Peace and Protection 4.22")
00:49:03amiconnCassandra: What did Linus say? Multiplatform problem in wps alignment code - player broke
00:49:28CassandraI'll go fix it.
00:50:36 Join webguest61 [0] (
00:50:58 Quit niobos ("off to get some sleep")
01:00:09 Quit TCK- (Read error: 110 (Connection timed out))
01:18:23 Join webguest53 [0] (
01:19:14webguest53amiconn: think others have been thinking in similar ways
01:20:22 Quit webguest53 (Client Quit)
01:21:12stripwax_Woohoo! PCM peakmeters ! Kindof!
01:21:22 Quit TCK (Read error: 54 (Connection reset by peer))
01:22:37stripwax_although... it looks pretty random right now which isn't really the intention :-(
01:26:56 Quit Stryke` (Read error: 60 (Operation timed out))
01:28:50CassandraArgh! What the hell does IO3:AddrErr mean?
01:29:10amiconnThat means address error
01:29:18CassandraWell yes.
01:29:31amiconnArchos or iriver?
01:29:34CassandraAnd why does this code work on Archos and die on iRiver.
01:32:35CassandraAnd why is it now not working on iRiver when it's logically the same code as last time? know
01:34:13hardeepCassandra: reinstall the codecs as well
01:34:34hardeepthe header was changed
01:35:03hardeepwe should put some versioning there, similar to plugins
01:36:48CassandraThanks. That was it.
01:37:01CassandraGods I'm getting tired of catching the digital flow.
01:37:58Cassandraamicon: Commited. Can you verify that player build is still working, please?
01:38:12 Quit webguest61 ("CGI:IRC")
01:39:28hardeepCassandra: the player sim works pretty well for this type of verification
01:40:44 Join Stryke` [0] (
01:44:24CassandraI'm still getting the last few seconds of the final track clipped.
01:45:05CassandraAlso resume info is definitely not being reset at the end of the playlist. (The index needs setting to -1)
01:46:21hardeepCassandra: strange, it was working for me yesterday... maybe one of the changes today broke it again
01:47:33 Quit Sucka ("a bird in the bush is worth two in your house")
01:47:49CassandraRight. Builds are now clean. I'm done for the day.
01:48:44 Quit Cassandra (" Try HydraIRC -> <-")
01:49:31***Saving seen data "./dancer.seen"
01:55:03hardeepSlasheri: around?
02:13:49RoriI look forward to peak meters
02:13:50 Quit hicks ("Too lazy to change my quit message")
02:16:34 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
02:17:15stripwax_Rori .. mmm.. will be nice when I've debugged it but it's getting pretty late..
02:19:08Roriexcuses ;)
02:25:27 Join amiconn_ [0] (
02:30:53 Join Biple [0] (
02:31:01BipleHello there
02:31:27BipleHows the radio coming along
02:34:00 Quit amiconn (Read error: 60 (Operation timed out))
02:34:01 Nick amiconn_ is now known as amiconn (
02:41:55 Join courtc_ [0] (
02:42:44 Join markun_ [0] (
02:42:47 Quit markun (Read error: 113 (No route to host))
02:42:58 Quit courtc (Nick collision from services.)
02:43:02 Nick courtc_ is now known as courtc (
02:50:13 Nick markun_ is now known as markun (
02:54:34 Quit Biple ("CGI:IRC")
03:13:46 Quit markun ()
03:13:56 Quit Moos (" HydraIRC -> <- The dawn of a new IRC era")
03:18:06 Quit hardeep ("[BX] Tony the Tiger uses BitchX. Its Grrrrrrrrreat!")
03:49:35***Saving seen data "./dancer.seen"
04:05:33 Join QT_ [0] (as@area51.users.madwifi)
04:15:52 Quit QT (Read error: 113 (No route to host))
04:16:30 Quit cYmen ("zZz")
04:25:00Rorigood question
04:41:48 Join StrathAFK [0] (
04:55:40 Join ashridah [0] (
05:00:22 Quit Strath (Read error: 110 (Connection timed out))
05:07:08 Join lostlogic [0] (
05:25:19 Nick StrathAFK is now known as Strath (
05:49:22 Join CheeseBurgerMan [0] (~Dan@
05:49:36***Saving seen data "./dancer.seen"
06:09:28 Join thegeek_ [0] (
06:12:18 Quit thegeek (Read error: 104 (Connection reset by peer))
06:13:00 Join Gfreaky [0] (
06:13:11 Nick Gfreaky is now known as Gamefreak (
06:13:20 Nick Gamefreak is now known as Gfreaky (
06:13:24Gfreakyhey yall
06:13:26Gfreakyanyone awake?
06:13:36CheeseBurgerManI'm awake.
06:19:43Gfreakyi was wondering if you could help me with rockbox? the main menu is really small and i was wondering how to make it larger?
06:20:09CheeseBurgerManI have no idea. I've never used Rockbox.
06:20:12CheeseBurgerManI own the H300
06:20:22Gfreakyahhh, i have a H120
06:20:23HClchange fonts, probably
06:20:27*HCl goes back to sleep
06:20:50Gfreakyor, what was called an ihp-120 when i bought it lol
06:20:57Gfreakyi love the thing
06:21:05CheeseBurgerManGood for you. ;)
06:21:13Gfreakynow i just gotta convert my wmas and i'll be good to go with only rockbox :)
06:24:13ashridahGfreaky: well, potentially, if someones' willing to rewrite ffmpeg to work in integer-only code, wma could be supported. i'm just not sure there's enough interest as it stands
06:24:24ashridah(interest from capable parties, that is)
06:24:35Gfreakyi hate wmas anyways
06:24:52Gfreakyit would be nice to have them supported but its just as good for me to get rid of the suckers
06:24:58ashridahheh, true, but reencoding it will end up potentially adding more loss, unless you've got wma's of your own cds
06:25:17Gfreakyi dont think the loss will be that bad
06:25:20Gfreakyand its only one or two cds
06:25:42ashridahwell, it depends on what you go to. FLAC wouldn't add any MORE loss, in theory.
06:26:33CheeseBurgerMan"and its only one or two cds" <=== does that mean you have the CDs?
06:27:06Gfreakyquick question: the docs say i need to "run" this file from my player
06:27:14Gfreakybut when i click ont he rockbox dir it bring up and option menu
06:27:22Gfreakyam i doing something wrong?
06:27:58ashridahGfreaky: which file are we talking about?
06:28:01 Join Naked6 [0] (
06:28:06Gfreakylike a custom font
06:28:11Naked6Hey FELLAS
06:28:13Gfreakyand a custom graphic
06:28:18Gfreakywps or something like that
06:28:35ashridahoh. uh, they go in .rockbox/fonts and ./rockbox respectively.
06:28:39Naked6anyone here know alot about FM waves
06:29:00Gfreakyyeah well
06:29:06Gfreakythen the doc says i need to "play" them
06:29:08Gfreakyhow do i do that?
06:29:36ashridahthen use a-b to get a menu up, then go to display, and change the font from there. that's the only way i know how to use them. holding down the control stick may bring up a context menu that lets you use them as well
06:29:48ashridahthe WPS can be chosen in the same way as the font
06:30:05ashridaha normal install of rockbox on a player will have included a fair chunk of fonts tho
06:30:12ashridahso adding a custom one may not have been necessary
06:30:17Naked6you guys know how nice cars like land rovers can pick up text from radio stations.... what if it were possible for our irivers to do the same? maybe it is not possible but if anyone knows I want to request this as a feature of rockbox.
06:30:32Gfreakyaaahhhhh! Thanks!
06:31:30Gfreakyquick other question
06:31:36Gfreakybit irrelevant
06:31:54Gfreakydo most boomboxes support, say, recording to a cassette from a line that the h120 has?
06:32:12ashridahGfreaky: if they've got RCA in plugs, probably.
06:32:17ashridahi find not many do these days
06:32:19Naked6some old boom boxes had mics
06:32:21ashridahcheaper ones at any rate
06:32:48ashridahyou don't really want the extra gain a mic-in will give you tho
06:33:29Naked6i would just go to goodwill and buy a tape recorder with aux inputs
06:34:04Gfreakycause like my car only has a cassette player
06:34:09*ashridah doesn't even own anything that'll play tapes anymore
06:34:12Gfreakyand im interested in hearing what it will sound like on tape
06:34:16ashridahwell, there is mum's car, that's about it :)
06:34:32CheeseBurgerManThen all you need is a cassete adaptor.
06:34:34Gfreakyi want a bit of a warmer analog sound
06:34:43Gfreakygot one :)
06:34:45Naked6DUDE. NO
06:34:52*ashridah shudders
06:34:53Naked6get a stereo
06:34:57Naked6for your car
06:35:06Naked6your raping your life away
06:35:34Naked6you need a CD/mp3 player
06:35:35Gfreakywas that referenced to the tape adapter or wanting to listen to stuff on a tape?
06:35:38Naked6FOR YOUR CAR
06:35:56Naked6ebay has to have them for a decent price
06:36:03CheeseBurgerManThe tape adaptor would let you listen to your iHP-120 in the car.
06:36:10Gfreakyi have one
06:38:52CheeseBurgerManSo then you (rather obviously ;)) plug the 3.5mm end into your iHP-120 and put the cassette in your cassette player.
06:39:05Gfreakyi know :)
06:39:20Gfreakyi'm curious to hear what my music would sound like as a tape :)
06:39:31CheeseBurgerManIf you have one.
06:39:34CheeseBurgerManTest it. :P
06:39:43Gfreakywell the thing is
06:39:43CheeseBurgerManSeems like the obvious thing to do.
06:39:51Gfreakyi want to compare the same music in each format
06:39:57Gfreakyand to do that i need to convert one to the other
06:40:14CheeseBurgerManBut then you'll lost quality, and it won't be a fair test.
06:40:32CheeseBurgerManTo do that, I'd test it how you'll normally listen to it.
06:41:06Gfreakyi just wanted to put wav/flacs on a cassette
06:41:14Gfreakyand see what they sound like compared to the digital files
06:41:19CheeseBurgerManOn the cassette.
06:41:23CheeseBurgerManThat's different.
06:41:28Gfreakyyep :)
06:41:42CheeseBurgerManThen you don't need the cassette adaptor. ;)
06:41:48Gfreakybut the thing is
06:41:54Gfreakyi need to find a way to convert it in the first place
06:42:32Gfreakyrockbox doesn't play flacs very well does it?
06:42:36CheeseBurgerManI don't know.
06:42:50Gfreakyits quite "skippy"
06:42:53CheeseBurgerManWAV and FLAC will sound the same.
06:42:58CheeseBurgerManSQ wise.
06:43:03CheeseBurgerMansince their both lossless.
06:43:17CheeseBurgerManand FLAC has a smaller file size.
06:44:43Naked6you guys know how nice cars like land rovers can pick up text from radio stations.... what if it were possible for our irivers to do the same? maybe it is not possible but if anyone knows I want to request this as a feature of rockbox.
06:46:01CheeseBurgerManAnd WavPack has a smaller file size than either FLAC or WAV.
06:46:06CheeseBurgerManand is still lossless.
06:46:41CheeseBurgerManBut Naked6, I have no idea.
06:47:12Naked6just think how sweet it would be if we had text while on FM
06:47:22Naked6we could start making recording programs
06:47:29CheeseBurgerManThe amount use FM, it wouldn't be all that important.
06:47:30Naked6to include the id tags
06:47:34CheeseBurgerManBut yes, it would be cool. :D
06:47:39CheeseBurgerManIF it's possible.
06:48:17Naked6it might be on a spectrum the fm channel does not normally see
06:49:03Gfreakywow this gapless is rather spiffy
06:49:07Gfreakyit works like a charm
06:49:12Gfreakytakes a while to load flac files but
06:49:17Gfreakyits very nice
06:49:20CheeseBurgerMan"well, WavPack -h is always smaller than FLAC -8"
06:49:21CheeseBurgerManIf that means anything to you, Gfreaky. ;)
06:49:25CheeseBurgerManMeans about nothing to me...
06:49:42CheeseBurgerManQuoted from DreamTactix291.
06:49:45Gfreakywavpack high compress and flac 8 compression quality i *think*
06:50:40CheeseBurgerManYeah, I figured something like that.
06:54:26 Quit Naked6 ()
06:54:31Gfreakythats what it is that i have in wma
06:54:33Gfreakythe wall :-/
06:54:43Gfreakyyeah, i need to borrow that my from my friend again
07:02:46ashridahFLAC tends to be fairly large, yeah, so it won't compress down to a meg or two per minute like mp3, ogg and wma can
07:03:21CheeseBurgerManYes, but that's typical of any lossless format.
07:03:57CheeseBurgerManJust one of those things that lossless is.
07:05:32ashridahindeed, which means a much higher bitrate, and thus, longer load times, and more of them
07:06:06CheeseBurgerManBut higher sound quality.
07:06:16CheeseBurgerManWhich is really the whole point of lossless.
07:06:29ashridahnot higher. highest :)
07:07:26 Nick CheeseBurgerMan is now known as CBM-away (~Dan@
07:09:29 Join LinusN [0] (
07:10:37 Nick QT_ is now known as QT (as@area51.users.madwifi)
07:49:14*LinusN is thinking of a new interface for the wake-up alarm
07:49:39***Saving seen data "./dancer.seen"
07:49:48LinusNa "Play later..." option in the browser context menu
07:50:18LinusNor "Timed playback" or something
07:53:45Bgermorning ;)
08:04:09LinusNBger: i removed it
08:04:16Bgerwhy ?
08:04:28LinusNamiconn: morn
08:04:54Bgermorning, amiconn
08:05:51LinusNBger: hmm, maybe i misinterpreted it...
08:06:42Bgerfor me i don't see any reason to remove it
08:07:29LinusNi put it back now
08:08:15*amiconn now has a 4-grey rockbox logo on his iriver shown at boot and info->version
08:08:37amiconn...converted with a heavily changed bmp2rb
08:09:02Bgeri pasted it just because i was surprised by the reaction of the iriver's support representative
08:09:33amiconnThe 'box' part in the logo is filled with light grey
08:09:43LinusNyeah, i missed the part where he said he told exactly what he had done
08:09:57LinusNamiconn: keeeeewl!
08:10:23amiconnI dropped the 'show ascii' mode for now. Is that a problem?
08:10:33amiconn(from bmp2rb)
08:12:27LinusNi kind of liked that option
08:12:42LinusNbut you can leave it for now
08:12:43amiconnTrciky to support with > 1 bit depth...
08:12:54LinusNamiconn: true
08:12:56amiconnbmp2rb now supports 3 target formats:
08:14:03amiconn(1) Archos recorder, Ondio, Gmini 120/SP, Iriver H1x0 monochrome bitmaps (2) Player gfx library format (3) Iriver H1x0 4-grey mode
08:14:49LinusNi assume ascii still works for (1) ?
08:15:08amiconnNot atm; I disabled ascii mode
08:15:34LinusNit's just a nice way of seeing that it interpreted the bitmap correctly
08:15:38amiconnIn fact I *could* support ascii mode for all depths, as I changed the processing quite a bit
08:15:40LinusNnot a big deal
08:16:17LinusNlet the one who really needs it do the work of fixing it
08:16:22amiconnread_bmp_file() reads the bitmap and puts the data in an 1 byte/pixel array (like the comment said but unlike it was implemented)
08:16:50amiconnThe byte conatins the true brightness of the pixel, computed from the palette
08:17:14amiconnNo more lame guess which colour is foreground and which one is background
08:17:47amiconnThen there is an additional function, transform_bitmap(), which creates the target format from that
08:18:35amiconnASCII could be generated from the 1 byte/pixel bitmap without even transforming
08:18:54amiconnJust ' ' for brightness >= 128 and '*' for <= 127
08:31:33 Join B4gder [0] (
08:36:03Bgermorning, B4
08:36:20B4gderhowdy ho
08:36:52*B4gder started his morning with over on hour without internet access!
08:44:19*amiconn wonders about the inconsequent fix for wps alignment on the player
08:44:43amiconnI don't see why the player shouldn't allow alignment (character based).
08:45:23amiconnAnd - in case this is undesired the alignment tag processing should be disabled as well.
08:45:32amiconnNo need to process data, then discard it
08:46:03B4gderI agree
08:46:30B4gderI guess the alignment is less useful on the charcell display due to the limited number of characters
08:49:59amiconnYes it's less useful, but first reading alignment info then not using it is nonsense
08:50:27amiconnLinusN: I enabled ascii output
09:10:26 Join Delazon [0] (
09:10:39DelazonHello ya all!
09:11:36*B4gder holds up his cup of coffe in a welcoming gesture
09:14:25DelazonI already had my morning coffe... byt any way SAULTE
09:14:37DelazonSalute it should be :D
09:16:26Delazonim new here and way i jumped on the train was that i wanted to find out anything about the release date for the Rockbox H3XX...
09:16:38LinusNthere is no date
09:16:41DelazonI couldnt find it on the site or in any FAQ....
09:16:54Delazonthanx for your answer then.
09:17:11DelazonBy the way, are you guy´s swedish?
09:17:20LinusNi'm swamped with work and Real Life(tm)
09:17:21B4gdersome of us are
09:17:22LinusNyes i am
09:17:49LinusNDelazon: you should change your character encoding from utf8 to iso-8859-15
09:17:51Delazonok, i guessed that by your names.... good work you are doing.
09:18:18Delazonhmm... im using Chatzilla in Firefox :(
09:18:37DelazonIm not used to IRC in general.
09:19:11DelazonIs it better now?
09:19:12B4gderwe'll try not to tease you ;-)
09:19:18 Quit pike (Read error: 110 (Connection timed out))
09:31:29HClgah. i hate spam
09:32:56ashridahi don;t like green eggs and spam, i don't like them, sam i am
09:35:17*HCl should probably change his usage of the word spam since in the original meaning, it doesn't make much sense..
09:35:25HCl"gah, i hate sending people annoying mail"
09:35:45HClpsmam maybe
09:35:50HCli hate people sending me annoying mail
09:36:01amiconnI'll add high-/truecolour bmp support to bmp2rb. Seems simple
09:36:26LinusNway to go
09:37:34amiconnThat's only as input format for now, and it discards colour infomation
09:37:59amiconn(computes a almost exact visual brightness from it)
09:38:33amiconnbrightness = (3 * red + 6 * green + blue) / 10
09:39:53HClis there any way to make rockbox display the recycled and system volume information folders?
09:40:17HCli need to delete them..
09:40:32B4gderthey aren't shown?
09:40:48B4gderI thought they were
09:40:53LinusNme too
09:40:53HClnot on my iriver, at least
09:40:57amiconnshow files == all
09:41:09HCli thought i had that option on.. but ok
09:41:22stripwax_Really, really rubbish PCM vu meter patch at
09:41:32HCli just managed to do it in the windows command prompt anywho
09:41:52stripwax_If anyone would like to point me in the right direction that would be great... like, is the approach even valid? I'm just guessing, really
09:42:38HClno clue, sorry :/
09:43:39LinusNstripwax_: interesting approach, i must say
09:44:46Maxime`Mrnyou guys never sleep? o_O lol
09:44:51stripwax_LinusN yeah.. I think it's kind-of in the right direction, reading directly from the pcm buffer indexed on the current value of BCR0. But I also make some crap assumptions
09:45:38stripwax_and of course it's not even slightly optimised. And I reread SAR0 before and after BCR0 to check SAR0 hasnt' been updated by the DMA0 interrupt in the meantime.
09:45:53DelazonAre the H3xx your main priority at the moment?
09:46:02LinusNDelazon: far from it
09:46:02B4gderDelazon: no
09:46:55stripwax_LinusN - by the way, what's the normal way to create patch files like this? I cvs diff -u file > patch.txt cvs diff -u file2 >> patch.txt etc ... ! :-s
09:47:00Delazoni just digging for some light in my H3xx tunnel :D Im a little tired of my player and want some thing new to play around with ;)
09:47:10LinusNstripwax_: i think those calculations should be made by the pcm dsp code
09:47:23stripwax_LinusN ?
09:47:37HClstripwax_: you can just diff directories..
09:47:38B4gderstripwax_: yeps, that's a fine way to make a patch
09:47:43LinusNthe peak info can be precalculated
09:48:02stripwax_LinusN - I agree, it can, but then you have to take into account latency, no?
09:48:12LinusNat the same time it does the mixing and resampling
09:48:28stripwax_LinusN - hmm, where is that code?
09:48:32LinusNthen the peak info can be stored in a separate buffer
09:48:39stripwax_(still finding my way around rbx)
09:49:08LinusNstripwax_: i'm as lost as you are when it comes to the iriver playback code :-)
09:49:40***Saving seen data "./dancer.seen"
09:50:08HClhm, lag.
09:50:37stripwax_HCl ?!
09:50:40HClmy improvement to the hashing algorhythm yesterday was at least enough to give my two copies of sunday bloody sunday a duplicate hash :)
09:50:53stripwax_HCl hehe
09:51:04B4gderHCl: the hash is only calculated by the songdb tool, is that right?
09:51:06HClwhich was one of the things that was annoying me
09:51:10stripwax_LinusN - ah, apps/dsp.c and/or apps/playback.c I'll take a look tonight maybe
09:51:14stripwax_gotta go.. t'ra!
09:51:15LinusNHCl: psst, it's "algorithm"
09:51:19HClB4gder: yes, but we might calculate it on target in the future
09:51:25HClLinusN: thanks :p
09:51:31HClare you sure? *checks dict*
09:51:44stripwax_LinusN - maybe it's algorhythm if it's got a good beat?
09:51:46HCl :p
09:51:49HClokay, i'll keep that in mind
09:52:16LinusNi remember "The Algorhytms" being the band that played at the wedding in Ghostbusters
09:52:39B4gderthere was a wedding in ghostbusters? :-)
09:52:45 Part stripwax_
09:52:45LinusNwas it ghostbusters or honey i shrunk the kids?
09:53:03*HCl imprints the proper spelling of algorithm in his mind
09:53:09DelazonNo wedding in shrunk the kids ;)
09:53:51ashridahghostbusters 2 might have had one
09:54:26LinusNit's probably neither of those
09:54:46LinusNmy memory isn't that good
09:55:55B4gdergoogle shows loads of references to the name
09:55:56ashridahthere was a sort of weirdo wedding near the end of ghostbusters 1 with the vigo wizard dude
09:56:05LinusNnot that one
09:58:18ashridahi think they implied a wedding in between ghostbusters 1 and 2 tho
09:58:42B4gderthat must be it
10:09:07 Quit Delazon ("Chatzilla [Firefox 1.0.4/20050511]")
10:09:54 Quit CBM-away (Read error: 104 (Connection reset by peer))
10:13:47 Nick Lynx_awy is now known as Lynx_ (
10:19:05HClwhats this crash
10:19:07HCli keep getting it
10:19:14HClI40:PDIR1FULL at 40102000
10:19:54LinusNhave you tried recording?
10:20:14HCljust playback
10:20:18LinusNwhen do you get it?
10:20:23HClwhen trying to stop my song
10:20:26HCli had it yesterday too
10:21:17LinusNthat interrupt shouldn't even be enabled
10:21:24LinusNlooks like a major crash
10:22:07HClgotta go.. i'll try to look at it later
10:29:52LinusNaah, i'm an idiot
10:31:04LinusNjust wanted to tell you all
10:32:40ashridahwe knew :)
10:32:56B4gderyour wife already told us
10:33:24ashridahwho else but an idiot would commit to a lifetime of rockbox development
10:38:56 Join Aison [0] (
10:48:13LinusNnew issue: how can we handle the case when the ata driver hangs because there isn't enough power to spin up the drive?
10:48:33Rickbarf all over the floor?! :D
10:48:51LinusNthere is an endless retry loop in ata_read_sectors()
10:48:55ashridahwatchdog timer interrupt?
10:49:24ashridahassuming one can even set it up across all platforms
10:49:54amiconnHCl: For your "0% Add a "Filesystem" link in the root when browsing the id3 database, since at the moment it completely blocks users from using the normal filesystem" todo, I may have a better suggestion
10:49:55LinusNthe retry loop is there to handle r/w errors due to shaking the device
10:50:05ashridahthe big problem i find is that there's no way to do this when USB is active is there?
10:50:22 Join Zagor [0] (
10:50:33LinusNashridah: true
10:50:45ashridahcan you forcibly disable the usb chip?
10:51:01ashridahbecause i'd really prefer the OS at least treated it as a problem *now* not when it tries to write to the disk
10:51:24ashridaheven if what may have been in flux between buffers is already lost
10:51:48ashridahi've accidentally let it run down a few times, and it's invariably been when it's plugged in in usbmode
10:51:58LinusNuse the charger
10:52:19amiconnHcl: Instead of having a separate tagbrowse mode, we could show the tag database as a virtual sub-dir in the root like we show the additional volumes with multivolume support now (Ondio)
10:52:21ashridahtrue, i didn't try plugging that in, infact, i yanked the usb plug :)
10:52:45LinusNashridah: i mean. don't do extensive usb sessions without the charger attached
10:52:47ashridahi doubt the os would have been overly happy regardless tho
10:52:58ashridahLinusN: yeah, well, i forgot. it's not something i do often
10:53:45ashridahbut my point is i'd rather the operating system had to deal with it as a 'wtf, the usb device vanished?!' than 'wtf, the drive won't respond' or however the usb-storage chipset handles the situation
10:54:12LinusNi get it
10:54:28ashridahof course, that's not the same thing as your original problem.
10:55:30LinusNno, our problem is that 1) the boot loader hangs if it can't load the firmware because the disk won't spin up and 2) rockbox hangs when it tries to save the settings
10:56:10LinusNwe can fix it by checking the battery voltage, but i don't like that solution at all
10:56:40LinusNsince it prevents utilizing the full battery capacity
10:56:42amiconnLinusN: Iirc the retry loop in ata.c has a timeout
10:57:16LinusNamiconn: not that i can see
10:57:29ashridahlike freaking out and hitting the brakes and then getting out and pushing when a car's fuel gauge hits 'E', it might still have enough fuel to get to a petrol station :)
10:57:48LinusNah, TIME_BEFORE(current_tick, timeout)
10:58:29ashridahbut yeah, i'd do the quick-n-dirty method in the first case, toss in a timer, attempt to spin it up, hopefully the timer won't get stuck by the ATA bus being jammed, and say 'Are we out of power' for a few seconds on the lcd, then shut down
10:59:03amiconnLinusN: yup
10:59:50amiconnIt still doesn't help sometimes, probably because of up-level code not checking return values, and retrying...
11:00:53 Join pike [0] (
11:06:21 Join markun [0] (
11:17:25 Join webguest11 [0] (
11:27:51amiconnLinusN: RLD froze the unit forever until forced poweroff, not only for a minute
11:28:07 Join Cassandra [0] (
11:42:35 Join yyz [0] (
11:44:18 Quit Maxime`Mrn ()
11:44:19markunamiconn: I'm implementing replaygain using the mixer. Any idea where I can set the volume so it will coincide with the track change?
11:44:59LinusNamiconn: ah, yes
11:47:15amiconnmarkun: My best guess is that it should be done by the playback code. It knows best when the track changes, and should have the replaygain value in the associated structure
11:47:31amiconnOf course the pcm playback latency makes it a bit tricky...
11:47:40markunyes, that's the problem
11:48:00LinusNimho, the pcm playback code should know where the track changes are in the pcm buffer
11:48:34amiconnThe pcm playback code needs to know where the DMA is as well, for the peakmeter
11:48:41amiconnAn associated problem...
11:49:15amiconnHow large are the DMA chunks?
11:49:42***Saving seen data "./dancer.seen"
11:49:58 Join Moos [0] (
11:50:25MoosHi guys :)
11:51:32markunmorning Moos
11:51:44Mooshi markun
11:53:23amiconnmarkun: You also need to take care how you set the mixer. The mixer value is non-constant, used as a prescaler when boosting bass / treble
11:56:21LinusNamiconn: the chunk size seems to be 32kbytes
12:00:05markunamiconn: I took care of that.
12:01:08amiconnLinusN: Not too large, less than 0.2 sec. For the peakmeter it might be sufficient to just calculate the peak per chunk
12:19:02 Join einhirn [0] (
12:24:45 Quit yyz ("later")
12:30:56 Join webguest69 [0] (
12:31:37webguest69I've just installed the latest version of iRiver Rockbox and when i turn the iRiver on i get this message "Error accessing playlist control file"
12:34:04webguest69any ideas?
12:34:56webguest69also when you crank the bass control up, the maximum volume dimishes
12:36:07 Join cYmen [0] (
12:38:21LinusNwebguest11: the "Error accessing..." message is usually because the .rockbox directory is missing
12:38:34LinusNor that there is a filesystem problem
12:38:35webguest69oh right
12:38:48webguest69Also the volume level decreases as you turn the bass up
12:39:01LinusNyes, that's intentional to prevent clipping
12:39:20webguest69oh right
12:39:38LinusNyou can't have max volume with cranked bass, so either the bass has to be decreased or the volume
12:39:56 Join Febs [0] (
12:39:57LinusNrockbox lowers the volume while the original firmware lowers the bass
12:40:08webguest69oh right. I believe iRiver had an option so you could turn "bass management" on or off on their 300 series
12:40:12LinusNi'm not sure which is best
12:40:44LinusNpersonally, i think iriver made the better choice
12:40:49webguest69well i suppose you could have a switchable option such as "decrease bass, decrease volume, or none (warning, may cause clipping)
12:41:05FebsLinus, I agree. Fletcher-Munson effect...
12:41:18LinusNwebguest11: of course, but rockbox already suffers badly from "optionitis"
12:41:21webguest69well what would u think of a switchable option as said above?
12:41:36FebsBass is perceived to be louder at higher volume levels, so the decrease in bass is not as noticable as the volume is increased anyway.
12:41:54LinusNFebs: yup
12:42:02webguest69It would be brilliant if the radio tuner gets supported as well
12:42:16LinusNwebguest69: of course, that's just a matter of time
12:42:33webguest69yea, and being able to record from it as well
12:43:03webguest69With the backlight fading, is it possible to keep it "fixed" at a dimmer or brighter level?
12:43:12LinusNwebguest69: no
12:43:38webguest69and finally, working peak meters in the wps would look good :-)
12:43:47LinusNof course
12:44:51webguest69but at least iriver rockbox is coming on nicely
12:45:24webguest69Does iriver rockbox automatically turn the machine off to prevent the battery getting excessively low like on the iRiver firmware?
12:47:14Slasheriwebguest69: in theory, it's possible to keep the dimming level fixed but that _will_ eat battery
12:47:32webguest69just wondered because i believe li-ion batteries get damaged if allowed to excessively discharge
12:48:12LinusNdoes the original firmware turn off with "battery low" if you turn it on?
12:48:37webguest69it shows "battery low" then switches off
12:50:46LinusNdoes it do that even when started via the rockbox boot loader?
12:56:01thegeek_it does
12:56:08thegeek_as long as the battery is not _too_ low
12:59:34 Join ]RowaN[ [0] (a2b0y@
12:59:48]RowaN[nowhere seems to have stock of 30gig 5mm drives for my iriver =[
13:02:22amiconnLinusN: For the bass vs. volume, I think rockbox made the better choice... and even way earlier, on the player
13:02:59LinusNamiconn: why do you think so?
13:04:13amiconnRockbox never sacrifices the desired treble/bass boost. If you crank up the volume, you'll notice it stops increasing
13:04:29amiconnYou might not notice the bass & treble decreasing from your desired settings
13:04:48 Quit webguest11 ("CGI:IRC")
13:05:26 Quit webguest69 ("CGI:IRC (EOF)")
13:07:08FebsAmiconn, isn't that exactly why the iriver choice is better?
13:07:28FebsIf you don't notice the change, then you haven't really departed from your "desired settings."
13:09:10FebsAnd because of the Fletcher-Munson effect, the bass boost is less necessary at higher volumes than at lower volumes anyway.
13:10:42 Quit thegeek_ (Read error: 145 (Connection timed out))
13:11:34 Join thegeek_ [0] (
13:14:45FebsThinking about this more, if you want the amount of bass *perceived* by the listener to remain constant as volume increases, you *have* to scale back the bass boost. That's exactly what the Fletcher-Munson equal loudness curves represent.
13:17:34 Join hicks [0] (
13:18:44Moosniobos began the tagdb conversion
13:19:48amiconnFebs: That would be a loudness function; something that isn't present in iriver rockbox yet (and also not for archos player)
13:20:17amiconnSo far only archos recorder and Ondio have it, because the MAS chip used there does support loudness directly
13:20:28markun]RowaN[: Do you live in Europe?
13:21:25FebsBasically, yes, it is a loudness function. But if your reason for limiting the volume instead of scaling bass as volume increases is because you want to keep the user's "desired" settings, the point still applies.
13:21:35amiconnFebs: The problem with loudness on iriver is that it doesn't fit nicely with the treble/bass ranges possible in the UDA chip
13:22:12FebsI guess the question is what do you mean when you say that you want to keep the user's "desired" settings.
13:22:22amiconnEspecially treble has a much too low range, and it isn't possible to do bass/treble attenuation
13:22:23markun]RowaN[: You might try alternate:
13:22:47amiconnFebs: Basically I would like it to work as on archos recorder and Ondio
13:23:02FebsWhich unforunately I am not familiar with.
13:23:15amiconnI.e. have a (configurable) loudness option, and allow treble & bass manipulation on top of that
13:23:23LinusNamiconn: you mean without compensation?
13:23:41LinusNthe recorder allows clipping
13:24:00amiconnIt's probably necessary to do digital filtering on iriver to achieve this
13:24:34LinusNmy personal issue is that i want to crank the volume higher, and i don't mind the bass being decreased
13:24:37amiconnLinusN: I never heard the recorder clipping, provided I stay below the true 0 dB limit (92% volume)
13:24:47]RowaN[markun: yes im in London England
13:24:58LinusNamiconn: that's the point
13:25:07]RowaN[ahah, wonder if they have stock
13:25:22LinusNthe recorder allows setting a volume higher than the limit of potential clipping
13:25:30LinusNbut not the iriver
13:25:33amiconnYes, but that's because the MAS allows to set volume > 0 dB
13:25:37amiconnThe UDA does not
13:25:41markun]RowaN[: I'm colorblind so I can't check if it's in stock :)
13:26:09markun]RowaN[: Ah yes, within 24 hours it can be shipped
13:26:46markun]RowaN[: Did you sell your 20GB yet?
13:28:16Febsamiconn, are the bass and treble boost on the UDA a "shelf" type EQ?
13:28:41amiconnLinusN: There's another difference. The MAS3587/MAS3539 clipping happens in the analog domain if you set the volume too high.
13:28:59LinusNamiconn: nope
13:28:59amiconnThe MAS3509 and UDA1380 clipping is digital; it sounds much more harsh
13:29:26]RowaN[sell my 20gig? its fucked coz i dropped it ='[
13:29:43amiconnFebs: "shelf" type??
13:30:00markunyou can have mine, than I can buy a 30gb :)
13:30:06Febsi.e., all frequencies above/below a certain frequency boosted equally.
13:30:29LinusNmaybe i have amnesia, but i thought the clipping when (max(bass,treble) + loudness + volume) > 0db was digital
13:31:14amiconnLinusN: There must be some difference; when I set volume to 92% on archos recorder I can crank up bass quite a bit before it will start clipping slightly
13:32:02amiconnOn iriver, without prescaling, it's clipping like crazy if I set the volume to 100% == 0 dB and then push the bass only a little (4 dB)
13:33:02amiconn2 dB is still bearable because the music doesn't use the full range that often
13:33:45amiconnFebs: I don't remember; you could check the datasheet
13:34:09FebsWill do. Not a big deal, just something I've been curious about.
13:34:58amiconnLinusN: I could compare that to archos player with prescaling disabled
13:40:58 Nick Aison is now known as Microsoft`in`die (
13:41:04 Nick Microsoft`in`die is now known as M (
13:42:19 Nick M is now known as Ms`in`die`Tonne` (
13:44:30 Join t0mas [0] (
13:49:45***Saving seen data "./dancer.seen"
13:50:49 Join austriancoder [0] (~austrianc@
13:51:20austriancoderZagor: are you here?
13:52:44Moospfffffffffffff London it will be a city for olympique plays of 2012 :(
13:52:49Zagoraustriancoder: yes
13:52:50MoosParis lose
13:55:42 Join bipak_ [0] (
13:57:08 Nick Ms`in`die`Tonne` is now known as Aison (
13:59:16 Quit Febs ("Chatzilla [Firefox 1.0.4/20050511]")
14:02:08*B4gder wrestles the gcc on arm demons
14:07:12 Join RotAtoR [0] (
14:11:10 Quit bipak (Read error: 110 (Connection timed out))
14:12:29zethats ok, by 2012 consciousness will reach its final stage of evolution and we'll cooperatively create reality at will
14:12:54zeya know, either that or planet-X'll swing back by and cause another mass extinction
14:14:28LinusNze: eeeh?
14:15:30zewell ya know the mayans had some interesting calanders
14:15:50zeand apparently an account of the last 14 billion years or so of the universe's history and evolution
14:15:59zeand their calander ends in 2012
14:16:42zeso supposedly everything'll be coming to some kind of pinnacle at that point
14:16:53 Quit Rick (Read error: 104 (Connection reset by peer))
14:17:08B4gdernooo, the world ended at y2k
14:18:17zey2k was just a round number in an arbitrary number base from an arbitrary epoch
14:18:29zethe mayan calander's supposed to be based on fundamental cycles of existance
14:18:44B4gderoh yes, maya is much less arbitrary
14:18:58B4gderbecuase they are older
14:19:20zebecause they supposedly figured out stuff we still haven't
14:19:35B4gderso said the y2k people
14:19:48zenot really
14:19:48B4gderand other end-of-the-worlders
14:20:11B4gderthey all have something figured out that the rest of us haven't
14:20:29]RowaN[the sooner this shit is over the better i say
14:20:36zepersonally i'm pretty impressed if an ancient civilization actually concieved of there being such a thing as 14 billion years in the past
14:20:48zethats more than you can say for the y2k people
14:20:55B4gderI don't believe that ze
14:21:19zeyeah i dunno anything for sure
14:21:22zeits all just stuff i heard
14:22:15 Join Rick [0] (
14:23:20]RowaN[the end of the world isnt as bad as it sounds.. the media duplication company i work at earned a heap of money duplicating y2k bug testing cds =]
14:23:49zewell according to the interpretation i saw of the mayan calander
14:23:58zethe universe has been undergoing an accelerating evolution
14:24:08zewhich culminates in the highest possible state of consciousness around 2012
14:24:32]RowaN[my morgage wont be paid off by 2012 =]
14:24:33]RowaN[all hail the mayans, um-ba-chu, um-ba-cha
14:25:15 Join DangerousDan [0] (~Miranda@
14:26:15]RowaN[bah when i send an email to a hotmail address, hotmail server responds by saying its "queued the message for delivery" ... 15mins later and its still not been delivered
14:47:17 Part rwlogix ("...")
14:52:21amiconnLinusN: Hmm? You changed the RLD description, and then changed it back?
14:53:23LinusNdamn browser cache
14:56:52 Join ghostiger [0] (~ghostiger@41bb5f12f9ce5a96.session.tor)
14:58:06 Join webguest14 [0] (
14:58:54webguest14in wps, how do i clear a line before scrolling the next line of data
15:00:16webguest14%s%It;%s%Ia I would like to clear the track line before displaying the artist info
15:00:40B4gderno can do
15:00:45B4gderI think
15:02:47 Part webguest14
15:06:20 Join Febs [0] (
15:17:52 Quit DangerousDan (Read error: 131 (Connection reset by peer))
15:20:19]RowaN[olympic games 2012 r gonna be held in london, whoopie
15:23:35 Part Nibbler ("Leaving")
15:26:51Chamoisparis was the best
15:28:56 Join yyz [0] (
15:33:36 Quit silencer1 (Read error: 101 (Network is unreachable))
15:37:25 Quit dionoea (Read error: 101 (Network is unreachable))
15:37:48 Quit yyz ("Leaving")
15:39:48 Join dionoea [0] (
15:40:06 Join silencer [0] (
15:41:44 Join yyz [0] (
15:49:48***Saving seen data "./dancer.seen"
15:53:46 Join Nibbler [0] (
15:53:47 Quit Nibbler (Client Quit)
15:56:24 Join [1]Chamois [0] (
15:56:50 Quit Bger (Read error: 110 (Connection timed out))
15:56:50*t0mas runs cvs update on his local copy... after more than a week...
15:58:32B4gderquite a lot of updates you will see
15:58:47B4gderspeak like yoda we will
15:59:36*t0mas moves B4gder to #ipodlinux
15:59:44t0masthey do that some times :P
16:00:03*LinusN spots a half-ready tag database creator in apps/
16:00:27CassandraIn C, thanks be to &deity;
16:00:40t0masyes, what is that tag database thing?
16:00:49t0masis it like the one HCl's working on?
16:01:25LinusNit creates a database in the same format
16:01:51t0masbut then on target?
16:02:00LinusNnot yet
16:02:27t0masbut it should do that in some time?
16:02:27LinusNmaybe one day
16:02:58B4gderNiels is our man behind that
16:03:15B4gderfeel free to start improving
16:03:34B4gderhe said he'll be back to work on it in 3 week or so
16:03:52LinusNour tag parsers should be moved to a separate library
16:04:34B4gderthen this C version could be used both for host and target
16:04:42B4gderusing the Rockbox tag parsers
16:05:31amiconn*use the rockbox force*
16:08:51t0masthis line:
16:08:52t0mas%t4%?Fn<Next F: %Fn|%?Ia<Next A: %Ia>;%?It<Next T: %It>>;%t2%?bt<Battery: %bl %bt>;%t2%?fc<Codec: %fc %fbkbps %ffKhz>
16:08:57t0masis that valid? :P
16:09:30B4gderI noticed we only support 800 bytes WPS
16:09:37LinusNt0mas: doesn't it work?
16:11:18t0masI don't know
16:11:19 Quit Chamois (Read error: 110 (Connection timed out))
16:11:19 Nick [1]Chamois is now known as Chamois (
16:11:31t0masit doesn't look like it's going to work :)
16:11:32LinusNt0mas: then why the question?
16:11:41t0masbecause I don't have my iriver here
16:11:53FebsSpeaking of WPS, I noticed an issue with today's bleeding edge build that I don't think was present on prior builds ...
16:12:18 Join Bger [0] (~Bager@
16:12:27FebsNew lines on the WPS overwrite the prior line without erasing the prior line. At least in alternating sublines.
16:12:34LinusNso you'd rather make us spend time evaluating your wps line, rather than test it yourself later?
16:12:58t0masno, I just hoped you could tell me it's wrong in 1 look
16:12:58LinusNFebs: yes i know
16:13:20t0masit's my sisters line... and she's complaining that it doesn't do what she wants
16:13:23amiconnPerhaps the alignment patch broke it...
16:13:30LinusNamiconn: probably
16:13:38LinusNbut i haven't yet seen how
16:13:51amiconnBad patch, that ;)
16:13:59CassandraI am never touching the WPS code again.
16:14:34amiconnNote the ;)
16:15:11amiconnHowever, I don't like your fix for the player. It parses alignment info, the discards it
16:15:58CassandraFeel free to fix it. ;)
16:16:22 Part LinusN
16:16:38CassandraI didn't think the alignment stuff was terribly useful for the player.
16:17:08CassandraAnd at that point, I was wanting to bugger off to bed.
16:17:39amiconnIt might be that it isn't useful, but then the info shouldn't be paresed either
16:18:04amiconnEither parse and use, or don't even parse
16:20:06*t0mas is still looking at his wps line
16:20:11t0maswondering what she wanted it todo...
16:20:55 Join deadlands [0] (
16:21:01amiconnI think I know why this overwriting happens. It does happen only with non-scrolling lines, correct?
16:22:10 Quit ashridah ("sleep")
16:22:58amiconnFebs: Does it only happen if the new line is too short to need scrolling?
16:24:26deadlandsI've discovered a little quirk with the H1X0 mp3 codec, low bitrate files produce crackling on playback.
16:24:35FebsNot sure how I could tell. Lines long enough to need scrolling overwrite the old text anyway. But, I guess the answer to your question is yes, I've noticed it on lines that do not need scrolling.
16:25:28amiconnThen the case is clear. That happens if someone replaces one function with another without knowing the exact differences (lcd_puts_scroll() -> lcd_putsxy() in this case)
16:26:00amiconnlcd_puts_scroll(), like lcd_puts(), clears the whole rest of the line, while lcd_putsxy() doesn't
16:26:18amiconn...since lcd_putsxy() has no line concept
16:26:23FebsThe problem also happens on lines with dynamic content, like the elapsed time display. E.g., When 3:20 changes to 3:21, the right edge of the 0 remains visible.
16:26:39amiconn...and the positioning code uses lcd_putsxy() unless it needs to scroll
16:26:44t0masthat's the putsxy() thing amiconn said
16:27:45t0mashm.. how do we clear a line?
16:27:50t0maswrite all spaces to it?
16:28:14amiconnWe could draw an inversevid rectangle across the line
16:28:25amiconnAnyway, this wps aligment looks quite hackish to me
16:29:39t0massssst, Cassandra starts feeling bad ;-)
16:31:16Cassandra'S OK. Not *my* hack. ;)
16:33:32t0masno, but we are the 2 persons who read it ;)
16:41:45 Quit deadlands ("CGI:IRC")
16:52:13CassandraHmm. How far are we from getting voice UI on iRiver?
16:56:04B4gderI guess we need someone to grab the issue and make it happen
16:56:25CassandraI really don't know enough about audio/buffering to be able to do it.
16:56:34amiconnB4gder: I'd very much like your per-target localisation and voicing to happen soon
16:57:43amiconnIt's a really nice idea, and it might save some space for the sake of the archos even though the code will prbably get a little larger
16:58:06amiconnWe're getting closer and closer to the final size barrier on fm/v2...
16:58:08 Join Nibbler [0] (
17:06:15 Quit yyz ("ciao")
17:15:42]RowaN[20gig iriver H320 is going for 179inc delivery on if anyone cares
17:16:23 Quit Chamois (" HydraIRC -> <-")
17:17:59 Join Chamois [0] (
17:19:02 Part Nibbler ("Leaving")
17:26:57*bill20r3 waits for a port
17:35:28t0mas]RowaN[: tnx, I care :D
17:35:33t0masI want an H3xx
17:35:45crwlh3xx's are big and ugly :P
17:37:01bill20r3like a brick
17:37:03bill20r3I have a 340
17:40:29t0masI don't want it because it's looking good :P
17:40:39t0masI want it because my H120 starts to fall apart
17:40:48t0masand my sister wants to have the H120
17:40:57]RowaN[h3xx dont have optical out, thats what i dont like about em
17:41:11crwli'd rather buy iAudio X5 if i couldn't find h120 anywhere, but it's probably quite expensive
17:42:40bill20r3I'm likeing my 340
17:42:50bill20r3I got a 120 remote for it, and flashed on the korean firmware
17:45:47Zagorwho is niobos?
17:46:00B4gdertagdb in C author
17:46:10Zagortagdb what?
17:46:30Zagorah, creating the db?
17:46:42B4gderin pure C code
17:47:23Zagorany reason that's not a plugin?
17:47:36B4gderwe want the code to work for both host and target
17:47:51B4gderso it probably will become some sort of lib and then a plugin will use that
17:48:30B4gderbtw, it doesn't fully work yet afaiu
17:49:52***Saving seen data "./dancer.seen"
17:51:23B4gderit wasn't really easy to think of where in the source tree the code would go
17:51:40elinenbeZagor, B4gder: the tagdb commit will become a plugin on the target that creates a browsable id3lib?
17:52:00B4gderonce it works, yes
17:54:07Zagoroff-topic: that vote was effectively a draw: no side got their changes through
17:55:23elinenbeB4gder: so, then it does the same thing as the perl/java apps in the wiki/cvs?
17:55:31elinenbeonce it is working properly...
17:55:36elinenbeoh −− great!
17:55:39t0masZagor: they're going for another round?
17:55:47B4gderof course it will probably be quite a lot slower
17:56:44t0masB4gder? do you mind reading a patch for me?
17:57:00B4gdercan't do it right now, am about to go home
17:57:02t0masI have the feeling I hacked at it to much... and last time you read some WPS stuff you had nice comments
17:57:13B4gderbut mail me and I
17:57:16t0masno problem, I'm not in a hurry
17:57:17B4gder'll look later
17:57:19t0masok, thanks
17:57:44Zagort0mas: the commission has said they will not write a new proposal, so we're back to square one. (i.e. EPO can continue to issue illegal software patents)
18:01:15 Quit B4gder ("go go go")
18:06:34t0masConditional images for the WPS are working :)
18:06:49t0masnow waiting for Daniel to read it... and tell me what should be done better... haha
18:07:39t0masanybody has comments for the new tags?
18:07:49t0mas%xd = display, and %xl = load
18:07:59t0mas(most of the code is written by kenshin)
18:14:36 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:24:59 Quit ]RowaN[ ("bbl")
18:26:08 Join solex_ [0] (
18:26:49t0masis there some way to output a tone with the iriver audio system?
18:30:05Slasherit0mas: Not yet but i think i could do a function to do that
18:31:07t0masok, and what if I have a few samples... is there some way to just drop them in the audio system?
18:31:48Slashericurrently no, i have to design a low level architecture that can mix from several sources
18:32:58HClamiconn: sure.. sounds fine...
18:33:15Slasherii think i will do some redesign to the pcm playback system..
18:33:21t0masSlasheri: so there is no way to play some sound?
18:34:01Slasherit0mas: if you really want to try, you could use for example pcm_insert_buffer -function
18:34:17Slasheribut don't use it when codec playback system is active
18:34:28markunSlasheri: I'm working on replaygain, but don't know where to set the replaygain volume. Any ideas?
18:34:31 Join tvelocity [0] (
18:34:36t0mashm... codec playback stops when I call audio_stop() right?
18:34:54Slasherimarkun: Hmm, maybe you could do that in dsp.c
18:35:03Slasheri(software volume scaling)
18:35:11Slasherit0mas: yes
18:35:22markunSlasheri: yes, could do that but now I'm just using the mixer.
18:35:57Slasherimarkun: ah, then you can still set the mixer values in
18:36:09 Join bagawk [0] (1000@bagawk.user)
18:36:17 Nick Lynx_ is now known as Lynx_awy (
18:36:21Slasherijust add DSP_SET_REPLAYGAIN or something like that settings
18:36:31markunSlasheri: Won't it be too early?
18:36:41t0maspcm_insert_buffer(char *buf, long length)
18:36:49t0masthat buf... that's 16bit pcm sample?
18:37:02Slasherimarkun: yes it would because pcm buffer has the latency..
18:37:22Slasherit0mas: yep, 16 interleaved stereo
18:37:37t0masoh shit...
18:37:38markunSlasheri: So is there something else that get's called when the track get's played? The wps switches at the right time..
18:37:40t0masI have some mono samples
18:38:01t0masmarkun: there is some event... HCl uses it for his runtimedb
18:38:12Slasherit0mas: Then you need to resample.. or use the dsp somehow
18:38:28*t0mas just changes his data...
18:38:40t0masinterleaved is normal PCM wave data right?
18:38:57Slasherimarkun: yes, audio_track_changed in playback.c
18:39:12Slasherijust do a search "track_changed" and you will find it
18:39:38Slasherit0mas: yep
18:39:47Slasheriraw pcm should be ok
18:40:22t0masas long as it's stereo :)
18:40:53t0masbut I can just copy it like this right: 1 2 3 4 -> 1 2 1 2 3 4 3 4
18:41:00t0masif 1, 2, 3 and 4 are bytes?
18:41:12Gfreakyis there a guide or something on the site on how to use the simulator?
18:41:18HClby the way, i had a crash with a build of this morning, was that known?
18:41:21bagawkShould be
18:41:23Gfreakyi wanna test some wps and i dont want to have to constantly reload it
18:41:27HCllike, immediately on bootup
18:41:29Gfreakydo you know where?
18:41:37t0masGfreaky: simulator doesn't work with WPS atm
18:41:42Gfreakyoh ok then
18:43:08bagawk shows some keys at the obttom
18:43:29Gfreakyi really must admit
18:43:37Gfreakyi am not a programmer and have no past experience
18:43:41Gfreakybut writing a wps is very easy
18:43:56Gfreakythat's the point i guess :)
18:44:13 Join Nibbler [0] (
18:49:38markunLooks like replaygain is working :)
18:52:11 Join [1]Moos [0] (
18:56:44Gfreakyhow accurate is the estimated time left for battery?
18:58:24t0masnot really accurate...
18:58:30t0masSlasheri changed it a bit last week
18:58:45t0masbut I don't know if it got better...
18:58:57Zagorit is quite accurate for archos devices. no idea about iriver.
19:02:00t0masfor iriver it isn't really good... but it's getting better :)
19:03:55Gfreakyfor wps
19:04:08Gfreakyis there a command to dynamically center text?
19:04:18t0masin the latest cvs version
19:04:21bagawkThere is?
19:04:25Zagorhas anyone done a battery discharge curve for the iriver?
19:04:27t0masit's new
19:04:31bagawkWow rockbox has changed to much
19:04:37Gfreakycool, that helps a lot :)
19:04:38t0masZagor: yes, but it's depending on cpu speed
19:04:45Zagorah, quite so
19:04:54t0masGfreaky, bagawk: it's not really working good
19:05:06amiconnZagor: Estimated runtime on my recorder v1 is totally off
19:05:15amiconn(by a factor of 3 or so)
19:05:19markunZagor, t0mas: The curve looks more or less the same:
19:05:34Zagoramiconn: really? it's very accurate on mine. at least to within an hour.
19:05:38t0masmarkun: check the times
19:05:54t0masmarkun: they differ a lot...
19:06:17amiconnZagor: It says >17 h when fully charged, but I only get ~5:30
19:06:21Zagorthe curve is the same, but the time changes a lot :-)
19:06:30amiconnThe cells are new, 2500 mAh Sanyo cells
19:06:54Zagorwow, 5:30? i didn't get that little even with the original green ones!
19:07:14amiconnI suspect the improved charging algorith cuts off too early for larger capacity cells
19:07:17markunt0mas: Yes, but the curve looks the same, just scaled.
19:07:45t0maswhat is that rockbox cerified
19:07:46amiconnZagor: I got ~4:30 before with 2200 mAh cells (Ansmann brand), and suspected the cells...
19:07:55Zagori must admit i'm basing my judgement on pasts experience. i haven't followed the development very closely the last few months.
19:08:08HClt0mas: they've been out in korea for a while.. oh wait, that O.o;
19:08:19HClno idea. what *is* that rockbox certified
19:08:26Gfreakyjust ooc: is the estimated time generally on the good side or bad side?
19:08:38Gfreakyie if it says 7 do i have 6 or do i have 8?
19:08:52ZagorHCl: there is no such thing as "rockbox certified"
19:08:59HClZagor: yet they advertise with it.
19:09:21Zagori guess they mean "rockbox has figured out how to work with it, so it will probably work once they port to h300"
19:09:25t0masso that's why I askewd
19:09:29markunI think Jeff is just excited about rockbox, that's all.
19:09:32 Quit Moos (Read error: 110 (Connection timed out))
19:09:32 Nick [1]Moos is now known as Moos (
19:09:50HCli wonder whether its any better than the h100 remote
19:10:03bill20r3has a little D-pad, that'd be nice
19:10:55HClyea but i'm getting the feeling its just a remapping of the buttons
19:12:18bagawkI forget, do floating poitn math (like sine/cosine) work on archoses?
19:12:34HClafaik we don't have any floating point math available
19:14:15HClwell, only emulation, and you don't want that..
19:15:09HClscary how misticaudio advertises the h140 as rockbox ready when its still in full development..
19:15:24Zagorwell the h140 is ready ;-)
19:15:58markunAnd for a lot of people at works just as good as the original firmware..
19:16:27HCli need to try a latest build
19:16:29markunSo it's is as ready for rockbox as for the original
19:16:33HCli had plenty of crashes yesterday
19:16:45HCldid linus find anything after the bugreport i gave?
19:16:49 Join Lear [0] (
19:17:20HCli guess thats a no..
19:18:11amiconnI had a crash some mins ago when I tried to play audio.
19:18:27HCli had several.. at the moment rockbox crashes when i try to boot it
19:18:31HCland that was a build of this morning
19:18:38amiconnMy guess is this happened because I didn't update the codecs, but simply crashing isn't nice
19:18:40SlasheriHCl: do make clean
19:18:47HClSlasheri: yea, i'll do that..
19:18:55amiconnThe codec api should be version-checked the same way as the plugin api
19:19:50 Join rjamorim [0] (~blah@
19:20:08markunSlasheri: replaygain works, but when I'm playing a song and switch to another song it takes about 1 second before the new volume is set.
19:20:48Learmarkun: what kind of replaygain? I was thinking of adding the same, you see... :)
19:21:20markunLear: howmany kinds are there?
19:21:32markunIt now just works for ogg
19:21:53Learmarkun: only one, but for different formats, and implemented in different ways...
19:22:08markunIf we add ape tag support for mp3 we can have replaygain there as well.
19:22:40Learmarkun: shouldn't require ape for that... mp3gain should write ID3V2 tags, IMHO.
19:23:31Learmarkun: and I was thinking of hacking Tremor to get "proper" replaygain as well. You can't really do it properly after clipping. ;)
19:23:34markunLear: mp3gain changes the mp3 itself. foobar2000 add ape tags and sets the gain there.
19:23:48markunLear: I was also thinking of that :)
19:24:16Learmarkun: mp3gain at least writes its undo info as ape tags, and I think you can tell it to only write tags and not modify the file.
19:24:32markunBut we should get the 24bit samples (sometimes 25) from Tremor and do the conversion to 16bit (or even 20) ourselves.
19:24:34LearBut supporting ape as well for mp3 is okay.
19:25:09LearThe internal format for Tremor is, AFAICT, fixed point with 9 fraction bits, which is easy to extract and process.
19:25:25LearBtw, how/where did you do the pow() calculation?
19:25:46markunI didn't the mixer works with db.
19:26:32markunLear: I don't change the samples, just adjust the volume.
19:27:36LearI guessed that much when you said the volume was set too late. But I didn't know the mixer worked with dB...
19:28:07markunLear: about clipping, when the peak is bigger than the maximum we can do a >> 1 and add 6 db to the volume.
19:28:24LearSo you parse the tag in metadata then, and just convert it to something the mixer groks?
19:29:05markunLear: Yes, I convert it to a value in tenth-of-db.
19:29:17markunShall I make you a patch?
19:31:10LearNah, I don't really need it (yet, at least). But will you look at the "proper" implementation later? If so, not much point in me doing it too. :)
19:32:09 Nick Lear is now known as Lear_away (
19:34:00 Quit Aison ("( :: NoNameScript 3.72 :: )")
19:35:34 Join LinusN [0] (
19:35:41LinusNhey folks
19:35:44bagawkrandom question: Where did and 192.168.x.x come from?
19:35:54bagawkhey Lear_away
19:36:20Zagorbagawk: you mean, why those specific numbers?
19:36:25bill20r3127.0.0.1 is squarely in the middle, 192 must have been arbitrary
19:36:28bagawkZagor, yes
19:36:32bill20r3unless it's something interesting in octal
19:36:38bill20r3or hex.
19:37:02amiconnit is
19:37:33Zagornote that 192.168 is not the only private net. 10.0 and 172.16 are too.
19:37:37bill20r3can I get 1/2 a point?
19:38:35amiconn10.0.0.0/8, and are all private
19:39:12LinusNSlasheri: why is the tag info so late in the wps?
19:41:55*t0mas is using 10.x.0.y/24
19:42:31t0masx = network number
19:42:50markunSlasheri: Sometimes after I switch songs I hear about 2 seconds of the new song, than a very short blib of the old song and then the new song continues..
19:42:59LinusNme too
19:43:19amiconnt0mas: Nothing wrong with that, you just subnetted the (only) private class a network
19:44:11Slasherimarkun: Hmm, i think that is a bug when crossfade is disabled
19:44:12*HCl goes to check whether his crash has disappeared
19:44:22HClLinusN: any idea about that crash earlier today?
19:44:36HCli'll make clean and try again..
19:44:47SlasheriLinusN: probably that late tag info update is also due to the disabled crossfade
19:44:56LinusNamiconn, Bagder : is this something we want?
19:44:58Slasheriit will be fixed
19:45:09LinusNSlasheri: great
19:46:06*LinusN is working on the a-b repeat patch
19:46:41 Join solex [0] (
19:46:45SlasheriLinusN: btw, i am just trying to remove the sleeps from pcm_playback.c. I think the problem is with pcm buffering so the sleeps are not necessary when i find the bug.
19:46:50markunSlasheri: and it would probably also fix the late replaygain
19:47:01Slasherimarkun: yes it should :)
19:49:36amiconnLinusN: I'm not sure. I didn't use triggered recording for real a single time
19:49:53***Saving seen data "./dancer.seen"
19:50:16amiconnA propos recording: I have the suspicion that the recording glitches some people still observe(d) have to do with rombox vs. normal rockbox
19:50:51amiconnThat can't be tested now, as rockbox grew too large for rombox even on the v1...
19:51:29t0masany swedish people here?
19:53:27 Join ghode|afk [0] (
19:54:26ghode|afkLinusN: the a-b repeat patch is it this one?
19:55:11bagawkghode|afk, that is a old one
19:55:30ghode|afki know
19:55:45rjamorimExcuse me...
19:55:46rjamorimHave you guys ever seen this?
19:56:18bagawkrjamorim, loks like corruption
19:56:32t0masrjamorim: have you unplugged the iriver without unmounting it?
19:56:39bagawkrjamorim, scandisk might fix it
19:56:42HClhow do i skip to the previous track...?
19:56:50t0masHCl: rwd 2 times
19:57:02rjamorimt0mas: not that I remember
19:57:06rjamorimFormatting fixed it
19:57:18rjamorimI was just wondering what could have cause it
19:57:24 Quit t0mas ("brb... reboot :(")
19:57:50ghode|afkthanks bagawk]
19:58:59 Quit ghode|afk ()
19:59:06HClt0mas, it wasn't working..
19:59:13 Quit solex_ (Read error: 110 (Connection timed out))
19:59:33bagawkGood bye
20:00:09 Quit bagawk ("Leaving")
20:01:29HClat least i haven't had a crash since i did make clean
20:01:34*HCl goes to rate his songs..
20:01:45 Quit tvelocity (Read error: 113 (No route to host))
20:02:45 Part rjamorim
20:03:05HClruntimedb works much better now that the wps loads info faster
20:06:30HClhas anyone noticed how next song id3 info doesn't work for songs without a tag?
20:06:37HClwhen it should fall back to filename
20:06:53Lear_awayIf you have the right WPS, yes...
20:06:59 Nick Lear_away is now known as Lear (
20:07:07HClwhat do you mean?
20:07:35LearSomething like this: %?It<%?In<%In - >%It|%Fn>
20:07:57LearThat is, track number and title if available, filename otherwise.
20:07:57HClhm right.....
20:08:03HClso its a glitch in the wps i use
20:08:11amiconnLinusN: Imho we should change the default date for the FAT32 driver from being fixed at 2003-08-01 to the build date.
20:08:25HCli second that
20:08:30amiconn...for units without rtc of course
20:10:11LinusNamiconn: sure, why not?
20:10:36CassandraWoohoo! EU throws out software patent law.
20:10:50HClwhile setting my song rating
20:11:16*HCl wonders why he gets all the crashes o.o
20:11:51 Join t0mas [0] (
20:14:06*HCl stares at rockbox.
20:14:18HClits not displaying anything for this song... no title, no artist, no album...
20:14:30HCli know its hair - let the sunshine in..
20:14:58CassandraJust lucky I guess, HCl.
20:16:02 Join cYmen_ [0] (
20:16:28 Quit cYmen (Read error: 104 (Connection reset by peer))
20:16:42t0masshouldn't we add some shutdown battery level?
20:16:55t0mas(just before the disk is unable to spinup... for example)
20:17:40HCli don't know about you people
20:17:45HClbut rockbox is really unstable here
20:18:04HClhm, well, it recovered this time..
20:18:20*HCl is mostly skipping songs really quickly, setting rating, then skipping
20:18:21LinusNHCl: you are the only one reporting those errors
20:18:35LinusNok, i'll try that
20:19:06t0mashe isn't the only one...
20:19:11HCljust did it again...
20:19:14t0massomeone else reported unstable cvs builds last night
20:19:15HClit just... doesn't skip
20:19:17HCland playback stops
20:19:26HClit hasn't crashed though
20:19:38HClbut it looks as if the codec thread crashed
20:19:51HClplayback is just frozen at 21 seconds, and its not paused
20:20:20HClwhen i press stop and then play again it works again
20:21:18SlasheriHCl: Hmm, do you know how to reproduce that bug? (i have also found it but haven't have time yet to look it)
20:21:43LinusNHCl: does it matter if you rate it or not?
20:21:52HClLinusN: i dunno
20:21:59LinusNfind out
20:22:09HClSlasheri: what i've been doing is just rating a song, skipping to the next one, rating it, skipping, etc.
20:22:50HClnot smooth
20:23:06HClwhen skipping repeatedly it kept playing the first second of a song over and over for a long time
20:23:25 Join webguest30 [0] (
20:24:36HCldoesn't seem to happen if i don't rate them
20:24:44HCli've been skipping forward for a while now without getting the bug
20:25:07 Join Coldtoast [0] (
20:28:01LinusNso it might be the database rating that is the problem
20:28:09HCli doubt it
20:28:21HCli'd say its the few seconds that you allow each song to play
20:29:34 Quit cYmen_ (Remote closed the connection)
20:29:59CassandraTry it without db rating to be sure?
20:33:53HClthere it goes again..
20:34:07HClkay, i'll let each song play for about 8 seconds then skip it
20:34:44t0mashave you enabled crossfading?
20:34:46HCli really need to rate my songs though :/
20:35:02t0masmaybe that doesn't like quick skipping
20:35:18LinusNHCl: is move-to-next-folder enabled?
20:35:28HClum, no idea
20:35:39HClprobably set to the default setting
20:37:30 Join cYmen [0] (
20:38:31HClmaybe it has to do with the wps context menu
20:38:35t0masexport PATH=/usr/local/m68k/bin/:$PATH && make && make zip
20:38:41t0maswrong window :)
20:38:46HClthe set rating code really doesn't do anything critical / anything that can crash
20:39:46HCloh, there it goes again
20:40:10HClSlasheri: well, to reproduce it, i recommend just putting the song rating in your wps and to start rating your songs, and it'll happen sooner or later
20:41:29HClalso, when skipping i just had a song start at 0:04, not 0:00
20:41:41HClbut that might be due to the crossfade that was probably so soft that i didn't hear it
20:42:02 Join _Mark [0] (
20:42:44LinusNSlasheri: is audio_current_track() guaranteed to return a valid pointer?
20:43:29LinusNno it doesn't
20:43:37HClthats it then
20:43:38LinusNHCl: you should check for a NULL pointer
20:43:50Slasheriyep, it might return a NULL pointer
20:43:59 Quit cYmen (Read error: 104 (Connection reset by peer))
20:44:06LinusN20.38.46 # <HCl> the set rating code really doesn't do anything critical / anything that can crash
20:44:25Slasherihehe :)
20:44:42HClwell i relied on it returning a valid pointer
20:45:17t0masthat centered lines things...
20:45:27t0masdo you think anybody cares if I try to fix it?
20:45:40t0mas(the way you said... clearing the full line)
20:46:42 Join hardeep [0] (
20:49:23CassandraI think people would be delighted if you fixed it.
20:49:30LinusNt0mas: go go go!
20:49:37 Join Aison [0] (
20:49:47CassandraWhat things are we waiting for before freeze, still, btw?
20:50:12LinusNa-b, perhaps
20:50:24HCli still don't think thats causing it though
20:50:39ChamoisCassandra: fm radio from austriancoder
20:50:50LinusNChamois: that't iriver stuff
20:51:07LinusNthe freeze is only for archos
20:51:30LinusNthe 2.5 release will not be for the iriver
20:51:59Coldtoastthink there'll be an iriver release this year?
20:52:10Coldtoastor is that the sort ofthing you just can't predict
20:52:12LinusNi definitely think so
20:52:16 Join cYmen [0] (
20:52:25SlasheriHmm, i really can't find what causes the glitch/pop when playing the first mp3 file. Playing ogg file first doesn't cause any problems and mp3s after that are also fine :/
20:53:12ColdtoastI was listening to an MP3 today and when I adjusted the volume, I coule hear faint ticking as it got louder
20:53:33Coldtoastneed to check tho. It was on some stuff I usually don't listen to. Never noticed it before
20:54:16t0masColdtoast: It wasn't 44.1 khz was it?
20:54:32ColdtoastI'll check right now actually
20:54:42t0masas I've heard some things too... and some other people reported it for 22 khz
20:54:43Coldtoastwas only with adjustign the volume
20:55:43ColdtoastI get constant tickign thru 22KHz mono tho
20:56:06Coldtoastthat's the resampling?
20:56:12CassandraIs someone looking at A-B? I had a brief poke, but it messes with keymappings in a non-good way.
20:56:20HCllinus is looking at it
20:56:38Coldtoastyeah. was 44KHz t0mas
20:57:06LinusNCassandra: yes, it conflicts somewhat with the pitch screen
20:57:13Coldtoasthmmm. but the stuff I usually listen to is also 44KHz
20:57:17Coldtoastjust looking now
20:57:41LinusNCassandra: i don't have a really good idea how to do it good
20:59:05t0masI've fixed the centered overwriting bug...
20:59:11t0masbut only for the first few seconds
20:59:13 Join einhirn [0] (
20:59:51CassandraLinus: I didn't either. I suspect that's why this patch has been sitting around for so long.
21:00:04LinusNit has other issues as well
21:00:23CassandraI did think of one possibility. Turning on A-B repeat disables pitch change. Turning it off re-enables it.
21:00:28CassandraNot sure I like that though.
21:00:38LinusNwas about to suggest that too :-)
21:00:57CassandraWhy don't we throw the issues back at the submitter.
21:01:34t0masolddrawmode = lcd_get_drawmode();
21:01:34t0maslcd_fillrect(0, ypos, LCD_WIDTH, strh);
21:01:37LinusNCassandra: we did
21:01:50t0masthen it inverts the whole screen
21:02:04t0mas(olddrawmode is an int)
21:02:27CassandraAnd he didn't reply? Then it's his own fault if it doesn't make it into 2.5.
21:02:34LinusNCassandra: in irc
21:02:34HClSlasheri: do i get track unbuffered events when skipping? i'm getting the impression i'm losing a lot of data while i've been setting ratings..
21:02:43t0masand it doesn't do that when you add "lcd_set_drawmode(STYLE_INVERT);" after the lcd_get_drawmode();
21:02:52t0masanybody has an idea why that's happening?
21:02:59LinusNCassandra: then he updated it with the current solution
21:04:08 Quit courtc (Read error: 113 (No route to host))
21:04:11CassandraI really don't like the current solution.
21:04:12SlasheriHCl: you should get the events, please use logf to verify
21:04:17t0masLinusN? have you got a moment to look at lcd-h100.c line 273?
21:04:29 Quit _Mark (Read error: 60 (Operation timed out))
21:04:37HClSlasheri: i have... i don't seem to get any rundb writes..
21:04:44LinusNt0mas: yes?
21:04:47CassandraPersonally I'd never use either feature though.
21:04:53t0massee my bug above...
21:04:57HClit does load them though
21:05:02t0masolddrawmode = lcd_get_drawmode();
21:05:05HClbut it doesn't seem to write them
21:05:07t0masthat makes a mess of my screen
21:05:13t0mas(and olddrawmode is an int)
21:05:34LinusNi don't see anything wrong on line 273
21:05:42HCli'm too tired to look at it tonight, i'll investigate properly tomorrow
21:06:01t0masLinusN: no, down from there
21:06:08t0masdrawmode = mode & (DRMODE_SOLID|DRMODE_INVERSEVID);
21:06:22t0masand the other one: return drawmode
21:06:38LinusNyes, he makes sure that the drawmode only contains valid bits
21:06:41t0masit fucks up when I get that mode... and then set it again
21:07:42t0masbut there are 5 drawmodes...
21:07:49t0mashow can it check with only 2?
21:08:31LinusN0,1,2,3,4 fits nicely in 3 bits
21:08:54t0masok, but still...
21:08:59LinusNbit 0-1 are the mode, bit 3 is the inverse modifier
21:09:05t0masit gets messy if I get the draw mode... paint something... then set it
21:09:10LinusNsolid|inverse == 7
21:09:38 Join tvelocity [0] (
21:10:08 Join courtc [0] (
21:10:52t0masthen why is my screen partly inversed after this:
21:10:52t0masolddrawmode = lcd_get_drawmode();
21:10:52t0maslcd_fillrect(0, ypos, LCD_WIDTH, strh);
21:10:52DBUGEnqueued KICK t0mas
21:11:17t0mas(there was a lcd_set_drawmode(DRMODE_INVERSEVID); statement in betwee, but I removed it for testing)
21:11:21LinusNhow come you don't set the drawmode?
21:11:36amiconnt0mas: There are *8* drawmodes
21:11:37t0masit was there... but I removed it for testing
21:11:42LinusNdrawmode is SOLID by default, so the fillrect will be solid
21:11:47amiconnIt's mentioned in lcd.h
21:12:00t0mas#define DRMODE_COMPLEMENT 0
21:12:00t0mas#define DRMODE_BG 1
21:12:00t0mas#define DRMODE_FG 2
21:12:00***Alert Mode level 1
21:12:00t0mas#define DRMODE_SOLID 3
21:12:02t0mas#define DRMODE_INVERSEVID 4 /* used as bit modifier for basic modes */
21:12:09LinusNt0mas: "bit modifier"
21:12:12amiconnDRMODE_INVERSEVID is a modifier that can be ORed to the other modes
21:12:19t0masah ok
21:12:28t0masso I should OR the inversed to the old mode?
21:12:30amiconnThe table in the wiki describes what each mode does
21:12:32t0masand then set back the old one
21:12:36LinusNt0mas: no
21:12:38 Quit cYmen (Read error: 104 (Connection reset by peer))
21:12:46LinusNgetmode, setmode, draw, setmode
21:13:11 Join cYmen [0] (
21:13:13LinusNif you want to clear it, you need to set the mode fo FG
21:13:17LinusNsorry, BG
21:13:17amiconnThe STYLE_INVERT is a completely different, much older thing that doesn't have to do with the new gfx core
21:13:45 Quit courtc (Remote closed the connection)
21:14:23CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:14:23*t0mas doesn't understand anymore
21:14:40amiconnIf you want to clear a rectangle, the drawmode must be set to DRMODE_SOLID|DRMODE_INVERSEVID (or DRMODE_BG|DRMODE_INVERSEVID)
21:14:41t0masamiconn: how should I clear a rectacle?
21:15:06amiconnThere's some quite logical concept behind the drawmodes
21:15:10t0masand then just set back to the old drawmode?
21:15:12 Join courtc [0] (
21:15:30*LinusN is confused too
21:15:52*t0mas just does wat amiconn said ;)
21:15:58LinusNi would accept DRMODE_FG|DRMODE_INVERSEVID
21:16:31t0masah thans amiconn
21:16:47t0masLinusN: FG + INVERSED = nothing
21:16:53amiconnFirst, the FG and BG is bound to the source element. Note that I distinguish 2 types of elements:
21:17:00LinusNhmm, i see now
21:17:25amiconn(1) Elements which are foreground only, like lines, (filled) rectangles etc
21:17:45amiconn(2) Monochrome bitmaps have both foreground and background
21:18:22 Quit cYmen (Read error: 104 (Connection reset by peer))
21:18:27amiconnDRMODE_INVERSEVID just logically swaps foreground and background _of the source_, so elements of type 1 then consist of background only
21:19:15amiconnDoing it this way is the only way to make DRMODE_COMPLEMENT|DRMODE_INVERSEVID doing something useful for bitmaps
21:19:45amiconnThen the other 3 modes (FG, BG and SOLID) are logical as well
21:19:59amiconnSOLID is just the bit combination of FG and BG
21:20:18amiconnOnly COMPLEMENT is a special reuse of the otherwise do-nothing case
21:20:28LearAny math wizards here?
21:20:56*LinusN looks away
21:21:08 Quit Coldtoast (Read error: 110 (Connection timed out))
21:21:48LinusNSlasheri: u there?
21:22:01***Alert Mode OFF
21:22:21amiconnt0mas: Since the wps display functions are only called from the GUI thread, you don't need to save the old mode
21:22:32amiconnYOu can safely assume the old mode is DRMODE_SOLID
21:22:47amiconnReset it to the same before returning
21:23:09t0masamiconn: but what if I do lcd_set_drawmode(DRMODE_SOLID|DRMODE_INVERSEVID);
21:23:11amiconnThis is documented in the wiki as well, below the drawmode table
21:23:14t0masand after that write some text?
21:23:53amiconnset DRMODE_SOLID|DRMODE_INVERSEVID, then your clearing fillrect(), then set DRMODE_SOLID, then draw your text
21:23:54t0masthen I should reset it to the old mode right?
21:24:22amiconnI have done this in numerous other places you can use for reference
21:25:33t0masyes, I've found one
21:25:43t0masgrep is a wonderfull tool ;)
21:26:06LinusNcool, VeryGoodSite is the #6 top wiki contributor this month :-)
21:26:18amiconnNote that the handling has to be different when the code might be called from other threads, or from an interrupt (beware!)
21:26:37amiconnThe scroll thread is one such example
21:29:07 Quit hicks (Remote closed the connection)
21:29:30t0mashm... Cassandra? have you looked at this one: ?
21:33:04Cassandrat0mas: I have now. Doesn't seem terribly popular.
21:33:44amiconnEr, I was just looking at my iriver. It is playing music for some time now
21:34:01t0masthe overwriting bug is fixed
21:34:11amiconnWhile it is still playing, the wps is frozen at track 6, time 0:00, and the backlight is on ....
21:34:27t0masI've heard that before
21:35:07SlasheriLinusN: yes :)
21:35:09Gfreakythis is really weird
21:35:21Gfreakyit says my player has 18.9 GB
21:35:26Gfreakyand all the stuff on it = 14.5 GB
21:35:35Gfreakyyet it says i only have 11 megs of free space
21:37:04LinusNSlasheri: i just wanted to know why the audio_next/prev/pause/resume functions are executed in the calling thread and not in the audio thread?
21:37:17LinusNi am about to move them
21:37:22amiconnMore strangeness: At the start of track 9, the wps decided to un-freeze itself ????
21:37:23SlasheriLinusN: ah, there is no reason for that
21:37:27Slasherigood :)
21:37:30LinusNwell, i have already, but not commited
21:37:40 Join Infirit [0] (
21:37:50Slasherii think it's safe to commit that change
21:38:02 Quit Infirit (Client Quit)
21:38:03LinusNthat makes the car adapter mode work again
21:38:10amiconnSlasheri: Any idea what might cause this weird behaviour
21:38:19Gfreakyany ideas whats up?
21:38:40Slasheriamiconn: Hmm, no ideas yet :/
21:38:45LinusNand i think that's the way to go on most cases, since it prevents weird thread races
21:38:46 Quit webguest30 ("CGI:IRC (EOF)")
21:39:01Slasheriah, that's good then :)
21:42:28 Join _Mark [0] (
21:44:21 Join webguest07 [0] (
21:45:36amiconnHmm, since nobody objected against lcd_mono_bitmap() / lcd_mono_bitmap_part(), I'll do that transition now.
21:45:45amiconnExpect a huge commit soon :)
21:46:48Slasherii will also commit some pcm buffering fixes soon if i manage to find one bug :)
21:47:12amiconnHopefully nothing using (or in the near of) lcd_bitmap() calls...
21:47:26Slasherii don't think so :)
21:48:04amiconnJust found another issue with current track info:
21:48:37amiconnIf you enabled 'follow playlist' you should end up at the currently playing track when switching to the browser.
21:48:56amiconnSame goes at the end of a playlist, where the browser should stay at the last played file
21:49:09amiconnHowever, it ended up at the second last file...
21:49:23Slasherihmm.. it doesn't follow the playlist correctly?
21:49:41amiconnIt seems that it does while playing, but not at the very end
21:49:55Slasheriah, thanks
21:49:56***Saving seen data "./dancer.seen"
21:50:03amiconnHmm, some testing may be due
21:50:18amiconnI hope to be able to do that after my commit
21:53:06LinusNSlasheri: could you explain the concept with the track change handling in the playback code?
21:53:33 Join Coldtoast [0] (
21:56:22 Join cYmen [0] (
21:56:42SlasheriHmm, there are three cases when track change could happen: user requests it, codec requests a new track or codec switch will happen. If track is changed by user and buffer is not filling, we will set ci.reload_codec = true that will cause the codec to request a new track. This will happen in function codec_request_next_track_callback. It handles the situations that the next track is already buffered (track change on fly), unbuffered (load from disk) or new ...
21:56:48Slasheri... codec when codec switch will be initiated
21:56:59Slasheriirssi pasted that twice?
21:59:13LinusNSlasheri: and the track change event?
21:59:22SlasheriLinusN: and when track really changed on audio output, pcm buffering system will send an event to the playback code
21:59:37LinusNvia a callback?
21:59:49LinusNthat sends a msg to the audio thread?
21:59:50Slasherithat event is added to the pcm data stream when playback system advances to the next track
21:59:54Slasheriyep, it's callback
22:00:17Slasherithere is codec_track_changed in playback.c
22:00:26Slasherithat is the callback function pcm buffering will call
22:00:31LinusNdo you think we can use the same technique for A-B marker events?
22:01:12SlasheriHmm, should be possible because we use audiobuffer_add_event(fptr) to add the callback to the pcm stream
22:01:23Slasheriso we could add any callback there
22:03:49*LinusN is in a-b patch mode
22:06:15 Quit cYmen (Connection reset by peer)
22:06:16 Quit _Mark (Read error: 104 (Connection reset by peer))
22:06:29amiconnHow do you add a callback to a data stream?? :puzzled:
22:07:19LinusNeach pcm chunk struct has a callback pointer
22:07:38 Join cYmen [0] (
22:08:41amiconnWhat happens if I call audiobuffer_add_event(fptr) twice while playing the same chunk?
22:09:08Slasheriamiconn: only the last callback will be added
22:09:22amiconnAh, so the pointer is replaced
22:09:40LearAnnoying. I have a fixed-point exp10 now, but the precision isn't quite what I expected. Is it a bug? :/
22:09:47amiconn...and possibly a callback isn't executed that shhould be
22:10:15amiconnWe should at least keep that in mind
22:10:26Gfreakyanyone know why my player says only 11 megs are left when it says it has 18.9gb and all the files only take up 14.5?
22:10:27amiconnIt may be that the runtime db doesn't like that
22:10:27Slasheriamiconn: we can make a workaround to this if we really have any risk replacing existing pointers
22:11:00LinusNGfreaky: trashcan?
22:11:19Leargfreaky: or bad FAT tables. See if scandisk or somesuch complains...
22:11:20LinusNcalled "Recycled" in windows
22:11:31Gfreakywhat about it i mean?
22:11:36Gfreakyjust finished running scandisc and no complaints
22:11:46Learcould be stuff in there that takes space.
22:12:25amiconnRECYCLER or the beloved 'System volume information'
22:12:35Gfreakyok that frees up some
22:12:48Gfreakywhat about "system restore: obsolete data stores"?
22:12:50Gfreakydo i need that?
22:12:55t0maslinux and rebooting
22:13:02t0masit is needed...
22:13:09t0mashanging hald process in the do_ide state
22:13:16t0masand no way to kill it
22:13:53 Quit cYmen (Broken pipe)
22:13:54 Join cYmen [0] (
22:14:09 Quit RotAtoR ()
22:14:16Learshouldn't be system restore info there, I think...
22:15:10Gfreakywell there isnt any more lol
22:15:21Leart0mas: been there, done that... :) (or at least something similar; don't know exactly where it was, but it was disk related)
22:15:33t0masLear: it's my iriver here :(
22:15:51t0masafter mounting it > 10 times it just hangs hald... and won't hotplag anything after that
22:16:23LearNow I remember; it was when trying to manipulate files on a possibly flaky CF card in an USB card reader. :)
22:25:20t0massee the mailinglist
22:25:29t0massomeone with and issue you told me about last week
22:25:36t0masthe > 16 tracks problem on archos
22:26:08 Quit t0mas ("brb... rebooting")
22:28:03 Join t0mas [0] (
22:30:26LinusNt0mas: the "norwegian language course" bug
22:30:46 Join spiralout [0] (
22:31:05Bagdernorwegian language courses are dreadful ;-)
22:32:03Bagderhaving more than 16 tracks that fit in memory at once
22:32:18LinusNt0mas: the bug submitter was listening to a language course with short norwegian phrases, each in one mp3 file
22:33:02t0masis there a fast way to generate an I04 somewhere?
22:33:23LinusNto test the new panic?
22:33:36LinusNjmp 0
22:33:56t0masasm("jmp 0");
22:34:01t0maslike that in C?
22:34:10amiconnIsn't I04 IllInstr?
22:34:17LinusNamiconn: yes
22:34:18t0masyes it is
22:34:34amiconnYou colud the use the special opcode assigned to that
22:34:35LinusNplenty of illegal instructions in the vector table
22:34:46LinusNamiconn: ah, yes
22:34:58amiconn(no joking)
22:35:00t0masthat would be a good test :)
22:35:18FebsI've started work on updating the FAQs to reflect recent developments. (Linus, thanks for the suggestion on MR that this would be better done on the wiki that at MR.)
22:35:43FebsOne thing that became clear to me immediately is that the "General" FAQ page actually contains many FAQs specific to the Archos hardware.
22:36:07*LinusN has a nice idea about that
22:36:31FebsUnless anyone objects, I plan to move those to a page specific to Archos. I think this makes sense and is consistent with the fact that there is a separate FAQ page for iriver.
22:36:52LinusNdo so
22:37:34FebsIt also occurred to me that it would be easier for people to find the FAQ pages if there were a link directly to them on the navigation bar on the left side of the Rockbox home page.
22:37:51LinusNFebs: that applies to almost every page
22:38:32Bagderthe faqs are in the top of the docs page
22:38:32LinusN- mom, i cleaned my room
22:38:41LinusN- good, how did you do?
22:38:44FebsLinus: true.
22:38:48LinusN- i put it all in a huge pile
22:39:01LinusN- but how do you find the stuff?
22:39:09LinusN- i put everything on top
22:39:40 Join _Mark [0] (
22:40:50t0masmy rebooting on UIE works
22:40:58t0masand amiconn's crash test too
22:41:33FebsLOL. :) On the other hand, since FAQs are directed mostly to new users and those unfamiliar with the project, it seemed to me that the front page would be a good place for a direct link to the FAQ.
22:41:44amiconnHummhomm :/ The player sim doesn't like my lcd changes
22:42:01FebsWas just a thought. In any event, I'll continue working on organizing and updating the FAQs themselves.
22:42:13LinusNamiconn: hehe, surprise :-)
22:42:18webguest07Febs: are noobs so dumb they can't work out that a FAQ would logically be linked from Documents ?
22:42:22LinusNFebs: much appreciated
22:42:32LinusNwebguest07: yes!
22:42:32amiconnLinusN: It doesn't find it's framebuffer anymore
22:42:46webguest07that's just tragic!
22:42:53amiconnThere was definitely something wrong with the declaration
22:42:56FebsWebguest7: After spending some (too much) time at Mistic River, I'm afraid that I have to answer yes.
22:43:15webguest07true - the forums there aren't pools of enlightenment are they ?
22:43:19Cassandrawg7: I recently told someone on the forums how to create a config file for use with a WPS.
22:43:37CassandraTwo posts later, someone was saying they didn't know how to do this.
22:43:44CassandraI think that says it all, really.
22:43:53webguest07hehe - notepad is a tough app to get a handle on eh ? :)
22:44:29amiconnLinusN: That reminds me, every time I login to freenode and join #rockbox, this reminder is displayed:
22:44:31amiconn[22:31:22] -ChanServ- [#rockbox] "Be sure to check out the FAQ before you look for 'general' information!"
22:44:31FebsNo, they're not. It is a welcome change of pace to come here and read dialog by people with a better grasp of the English language than the folks at MR, many of whom speak English as their native tongue.
22:44:33webguest07it depresses me, the number of people out there who fail to read stuff
22:44:39amiconnThe link is outdated...
22:45:04webguest07Febs: You mean they *claim* to speak it as their native tongue. The evidence says otherwise :)
22:45:14FebsHa ha! True.
22:45:27CassandraThey speak American, which is a strange bastardised version of English.
22:45:53webguest07even american is preferable to "l33t"
22:46:35t0masI know some people speaking l33t english all time...
22:46:38t0masthat's just irritating
22:46:55t0masespecially if they claim to know linux... but don't even know what iptables is :X
22:47:05webguest07I'll forgive someone for not knowing when to choose between "its" and "it's" - but people who think they have to abbreviate everything to the minimum number of letters is just plain lazy
22:47:27t0mas"Oh man, I'm duch a l33t h4x0r, I gonna h4x0r your l4m0 b0x in 10 seconds."
22:47:39t0mas3 minutes later: "WTF? iptables? what's that?"
22:48:07t0mas5 minutes later: "You're just a nerd... stop talking about that shit."
22:48:11zewebguest07: i dunno bout tht, cuz b4 ya no it yer tpn 2much
22:48:13zewebguest07: :p
22:48:14webguest07"my skript k1dd1e programz c4n't h4ndle Iptables"
22:48:23dionoeat0mas: lol :)
22:48:44t0maswebguest07: I had someone hacking a few weeks ago... (was an EFFnet idiot)
22:48:48t0masasked for my IP
22:48:55webguest07that old one
22:48:56t0mas3 minutes later... ping timeout
22:49:07t0masdon't know what toold he had...
22:49:12t0masbut I saw someone doing that before :P
22:49:16webguest07you could have been more original, and given him or something
22:49:22webguest07they all loopback
22:49:37HCllooks a lot less like localhost
22:49:47webguest07to a twit anyway
22:49:57t0masok, I'll do that next time...
22:50:05t0masthere are more 13 y/o hacker wannabees ;-)
22:50:21 Quit west-acre ("I-n-v-i-s-i-o-n 2.0 Build 3515")
22:50:22webguest07I didn't think EFNET offered stealthed IPs yet anyway
22:50:25CassandraI found a great site to point idiots at.
22:50:33webguest07if you have to *ask* someone on IRC for their IP. Well. Really.
22:50:53webguest07Cassandra: nice. I'll book mark that :)
22:50:54t0masand I dare to claim my box is almost "unhackable"
22:51:05 Part LinusN ("Leaving")
22:51:14 Join LinusN [0] (
22:51:17HCli stopped caring for security a while ago :p
22:51:25LinusNamiconn: changed
22:51:25webguest07well there's always the chacne the moron will own a botnet
22:51:29Cassandrat0mas: I think they'd have trouble hacking my wireless router. :)
22:51:41webguest07I have an 8MB link at home, but I still don't want it flooded
22:51:54amiconnbash is weird, at least in cygwin
22:51:54t0maswebguest07: that's the only think I can't have here :)
22:51:58t0masjust 2 mb link here
22:52:01ColdtoastI'm sitting here on 1.5Mbit :/
22:52:13t0masyeah... well mine is effectively 1.5
22:52:16webguest07t0mas: heh - I'm only paying for 4MB as well :)
22:52:27 Quit _Mark (Read error: 54 (Connection reset by peer))
22:52:39amiconnIs it to be expected that if I open a bash window, the do startx & from there, and I close the bash window, x11 gets also shut down?
22:52:49webguest07when they turned it on the router was negotiating up around 6.5MB, but now my ISP is upgrading me to 8MB for free!
22:52:49Coldtoastwow! you have a 8MB line??
22:52:56t0masamiconn: that's normal
22:53:01Coldtoasthow's that possible?
22:53:04Bagderamiconn: it seems they have chosen that approach, yes
22:53:09webguest07amiconn: yep
22:53:23t0masColdtoast: it's possible here too...
22:53:24webguest07Coldtoast : Sorry - 8Mb
22:53:26t0mas8 Mbit
22:53:27t0masnot MB
22:53:29ColdtoastMB tho?
22:53:33amiconnWhy? I start it in the background with &, and I am able to do other stuff in the shell....
22:53:36webguest07my bad
22:53:36ColdtoastI'm beign persnickety
22:53:57*Cassandra didn't have broadband at all until 2 months ago.
22:54:03webguest07I held the shift key down too long. It's this web client - it's a bit slow :)
22:54:06t0maswhere do you live Cassandra?
22:54:08CassandraNow I have 512k, and I love it. ;)
22:54:17CassandraJust outside Oxford, England.
22:54:21ColdtoastI started out with 512/128 myself
22:54:40t0masI felt horrible when I had no adsl for 2 days... 6 months ago :)
22:54:45Coldtoasthopefully they'll install the equipment for ADSL2+ where I am
22:54:47CassandraCan't afford more than 512k, really.
22:54:52t0masfucking slow 128/128 ISDN link
22:54:56webguest07ADSL2 will be nice :)
22:55:01t0masyes :)
22:55:04webguest07I'm hoping my exchange gets upgraded with it too
22:55:11Coldtoastmy ISP started rolling ADSL2+ out this year but my state's not scheduled at all for the DSLAMs they need
22:55:19webguest0724Mb is the initial offering isn't it ?
22:55:22amiconnThat reminds me...
22:55:25webguest07in the UK anyway
22:55:28t0masit's already availible here :D
22:55:31Coldtoastyep. ADSL2+ is 24Mbit
22:55:33t0mas20 mbit though...
22:55:41t0mas40 euro's / month
22:55:47webguest07not bad
22:55:56Coldtoastonly righ tnext to the exchange tho. I'll prbably get aroudn 20Mbit where I am
22:55:59t0masincluding some live tv shows (sports)
22:56:00webguest07I think I pay 30 a month for the 8Mb
22:56:01Coldtoastnot too far from teh exchange
22:56:18ColdtoastI'm on 1.5Mbit and get the full 1.5Mbit
22:56:18HCli pay 10 euro a month for 100mb
22:56:25HCl :p
22:56:34*t0mas looked it up... it's about 430 m here
22:56:41webguest07who did you sleep with to get that rate HCl ?
22:56:52t0masno, he's living at the campus :P
22:57:05HCl :P
22:57:06*t0mas has 1 year left...
22:57:07webguest07ah - the life of a student. Remember it well. Didn't have the intermaweb then though.
22:57:15HCli'm on the university network administration team
22:57:17HCli guess that helps? :P
22:57:26webguest07you still have to pay tho :)
22:57:30*t0mas has to look into that too
22:57:32HCli didn't used to
22:57:36HCli used to get my money back
22:57:39HClfor being on the admin team
22:57:49HClbut now we just get free outings instead
22:57:52HCllike going bowling or bbq
22:57:54Coldtoastwebguest07: when you guys are filesharing, do you have your own wifi networks running on campus?
22:57:58HClwhich is more fun for the money anywho
22:58:03Coldtoastso you don't get busted filesharing?
22:58:07webguest07Coldtoast - I'm not on a campus.
22:58:11HClwe have our own wifi network, yes
22:58:12Coldtoastoh. ok
22:58:36ColdtoastI took my x50v for a walk around campus here (it's 2mins from my back fence) and there were about 3 running
22:58:37t0masgetting busted? HCl and me live in the Netherlands :P
22:58:49HClour university almost advertises
22:58:52HClwith its network
22:59:14Coldtoasthasn't Sweden just passed laws on filesharing tho? Does that eventually filter over to you guys?
22:59:27t0maswell... It might...
22:59:33t0masbut not in < 20 years
22:59:43Bagderour new laws are pretty lame and won't affect much imho
22:59:43Coldtoastcool. I should move there
22:59:47t0masand then they'll forbid kazaa...
22:59:49ColdtoastI'd move there if it was cold
22:59:55t0masas nobody uses that anymore
22:59:57Coldtoastkazaa suck arse anyway
23:00:15Coldtoastthey'll probably bad the original Napster first :)
23:00:18t0masanother 30 years before they'll find out about bittorrent
23:00:41t0masour gouvernment is always running after the technology...
23:00:46t0masthey just don't get it :P
23:00:51HCli'm gonna demonstrate if they try to pass a law
23:01:05HClstupid brein already limited us to 15gb upload to the rest of the internet
23:01:17HClper week
23:01:24t0masHCl: buy the new Pc-Active magazine...
23:01:26Coldtoastper week?? nice!
23:01:29HCland wifi is no limit
23:01:36t0masand read my article about extra cost added to DVD's
23:01:53t0mas(heffingen in dutch)
23:01:55ColdtoastI dumped all the wires in the house here a few months back
23:02:01*amiconn is still mass-rebuilding simulators
23:02:03ColdtoastSOOO much nicer having no cables
23:02:23*t0mas has wires here... build into the wall when we moved here
23:02:31Coldtoastmight invest in some 7.7dbi replacement antennaes tho
23:02:34t0mas1 gbit is way faster than 54 mbit
23:02:38Coldtoastyeah. that'd be nice too
23:02:42t0masthe only wireless thing here is my laptop...
23:02:49t0masoh and my mothers old laptop
23:02:52Coldtoastwell, mine's 108mbit but still
23:03:01Coldtoasteffectively probably getting around 12mbit
23:03:07HClwe only have 11mbit..
23:03:09amiconnI never experienced such a thing as a stable wireless connection
23:03:09HCland since a week or 2
23:03:16HClthere's been a jerk pulling the entire bandwidth
23:03:22*t0mas has a stable wireless connection
23:03:32t0masasus accesspoint
23:03:34amiconnThey're always more or less flaky
23:03:38t0masand build in intel client
23:04:18ColdtoastI'm the only one here with gbit LAN
23:04:26t0masamiconn: My AP is dumped somewhere in a closet... close to my switches and router... and get 90% signal strenght in the garden
23:04:35Coldtoasteverybody else has 100Mbit and my switches and routers are all 100Mbit
23:04:38HCli bet i'm the only one here with root access to a computer with gbit internet o.o
23:04:52t0masHCl: I guess you are...
23:04:53Coldtoastwhat router are you using t0mas?
23:05:02t0masColdtoast: linux router
23:05:04HClnot that i can do anything with it
23:05:05amiconnIt's not the distance and signal strength. Often windows says signal strength 'execellent' yet the connection is still flaky
23:05:11t0masjust a simple P2 @ 233 mhz
23:05:13Coldtoastor you built a box?
23:05:24ColdtoastI'm using a Linksys WRT54GS
23:05:26amiconnIt seems to depend on the combination of wifi card, ap and OS
23:05:30Bagderhonestly, windows is less good at doing wireless
23:05:32Coldtoastquite a famous little wifi router that
23:05:36t0masand my modem is a simple E-tech shit thing... I got it from my ISP for free
23:05:36HClmy dad has a linksys wrt54g
23:05:49amiconnAt home I prefer trusty wires
23:06:02Coldtoastall we use the network for is ADSL
23:06:07Coldtoastno file transfers at all
23:06:13t0masamiconn: asus AP + pcmcia card works great... asus AP + intel centrino works great too... asus AP + alcatel doesn't work that good
23:06:15Coldtoastso wifi is AWESOME
23:06:42amiconnWell, my laptop has 11b intel builtin, and that works well with most ap's.
23:06:51t0masHCl: is the mirror at your university?
23:06:52amiconnWitth the T-Sinus 1054 it is flaky
23:06:58HClthats our university server, yes.
23:07:07Coldtoastyeah. I also need wifi for friends' laptops and my PDA
23:07:13amiconn(Which is in fact a Samsung iirc)
23:07:18t0masHCl: you can't use that to DoS someone off the internet ;-)
23:07:41HCltechnically i could
23:07:41amiconn54 MBit/s is even more flaky with the T-Sinus (Level One cards, both PCI and PCMCIA)
23:07:52HClbut i'd get my arse kicked by the other admins in no time at all
23:07:52t0masHCl: yes, but that's the same with me...
23:07:53amiconn...even in the same room
23:07:53HCl :p
23:08:01t0masI technically can use company servers...
23:08:08CassandraWhy has my cat taken up residence under my desk?
23:08:15t0masbut in real life I can't.. I'll get fired
23:08:20HClsame thing here
23:08:25Coldtoastamiconn: 802.11N should be interesting
23:08:26HClthough i have an university server
23:08:29HClthats titania
23:08:40HClwith 2gb of ram and 4 xeons of 550mhz each
23:08:49t0masah your ftp?
23:09:00HClamong other things
23:09:04amiconnColdtoast: I don't think so
23:09:10t0masmy home server is a simple dual P2
23:09:14t0mas2x 400 mhz
23:09:19t0maswith 512 mb ram...
23:09:23Coldtoastit's full duples tho, for one thing, is it not?
23:09:31t0masfast enough to do everything it has to do here
23:09:41amiconnIf even the current setup doesn't work reliably, moving to a new one probably won't help
23:10:02Coldtoastbut it's not just a faster version of what we have now tho, afaik
23:10:05*t0mas copies his new audio book to his iriver...
23:10:14t0masand goes to sit downstairs
23:11:01 Join webguest37 [0] (
23:11:18Coldtoastaaah! nice build
23:12:32 Join chester [0] (
23:13:59chesterhello - do any of use dir /s /b to generate playlists?
23:16:32t0masI don't
23:16:51t0masamiconn / Cassandra? where's the resume data stored?
23:16:59t0masa file in .rockbox ?
23:17:23LinusNt0mas: in sector 61 iirc
23:17:39t0masow... so no way to reset it?
23:17:44amiconn/.rockbox/.playlist_control stores the current playlist
23:17:54t0maswhen rockbox isn't booting... because resume is set to a WMA file?
23:17:58amiconnThe resume index is stored in the config sector
23:18:22amiconn...which is start_of_first_partition - 2, if that is > 0
23:18:47amiconnUsually sector 61 on harddisks, and sector 30 on flash memory (Ondio)
23:18:58t0masok, and is there some way to reset it?
23:19:02t0maswhen rockbox isn't booting?
23:19:11t0mas(because it can't load the WMA file it tried to play)
23:19:29amiconnYou can try if deleting .playlist_control helps
23:19:40Bagderdoes it really hang if you try to play a wma file?
23:19:44amiconnOtherwise use dd to overwrite sector 61
23:19:48amiconn(hard way)
23:19:49Bagderor rename the file
23:19:51 Part chester
23:20:07t0masso there is no simple way to help a noob having this problem? :)
23:20:20Bagderrenaming isn't _that_ hard
23:20:26t0masbecause I'm not going to let someone use dd if they don't know how it works :)
23:20:45Bagderc'mon, it could be fun! ;-P
23:20:46amiconnBagder: It might be that renaming doesn't help. Isn't the internal playlist system index based?
23:21:14Bagderah, righ if he played a dir
23:21:23Bagderthen we would need to move it to another dir
23:21:27Bagderhe would
23:21:36amiconnOne could still delete that dir, or at least the problematic file
23:21:50Bagderor rename it so that the file order is modified ;-)
23:21:51amiconnBut I would try deleting .playlist_control first
23:22:02amiconnAll that gets lost is the current playlist
23:22:43t0maswell... wish him good luck :)
23:22:52amiconnHmm. Why does the x11 player sim link, but the win32 sim doesn't :/ ?
23:23:07*amiconn goes studying .map files
23:23:48hardeepremember to take a compass with you
23:27:19hardeepdo we have a feature-freeze yet?
23:28:06t0masoh about that
23:28:16t0masBagder? isn't it better to do that another way?
23:28:26Bagderdo what?
23:28:34hardeepOkay, I wanted to add another feature that people have been requesting: "Insert from" which inserts all tracks starting from the current
23:28:34t0maslike letting Zagor press the button... and then use patch to put in bugfixes
23:28:45t0masso we don't have to stop committing features for iriver?
23:28:58Bagderyes I think so
23:29:05Bagderwe should make a branch for the release
23:29:51t0masthat would be ok too
23:29:54BagderI just don't currently feel I have enough time to grab this task myself
23:30:03t0maswho manages cvs?
23:30:16Bagderanyone can make a branch
23:30:22Bagderbut Linus, me and Zagor are admins
23:30:40HClyea, a branch for the release is definately a good idea
23:30:51*t0mas will read it when he's back :)
23:30:57HClthat way development can continue and copies are made of important tools like songdb
23:31:13 Join CheeseBurgerMan [0] (~Dan@
23:31:15Bagderyeps, that way it won't slow down any development
23:31:21amiconnAhaahh! The playersim doesn't break because it doesn't use the framebuffer
23:38:12 Join hicks [0] (
23:40:09 Quit Zagor ("Client exiting")
23:42:50*amiconn ponders rewriting lcd-win32.c some more
23:45:32 Quit ghostiger (Remote closed the connection)
23:46:57HClpff. it takes 15 minutes alone just to read all my ddr songs..
23:49:01 Quit spiralout (Read error: 104 (Connection reset by peer))
23:50:00***Saving seen data "./dancer.seen"
23:50:20CoCoLUSwouldn't a seperate release branch be obsolete long before it's bug-free with the current rate of new feature additions?
23:50:49BagderCoCoLUS: it wouldn't get any new features
23:51:04 Join preglow [0] (
23:51:08HClthe branch would only get the bugfixes
23:51:12CoCoLUSyeah i know
23:51:16HClwhere the main branch gets both bugfixes and new features
23:51:25CoCoLUSbut the main line would get all the new features
23:51:38preglowhi all
23:51:42HClthats the whole idea of a feature freeze
23:51:43Bagderhey preglow
23:51:46HClhey preg
23:51:52preglowanything fun going on?
23:51:58HClpreparing to play ddr..
23:51:59amiconnhi mr. codec :)
23:52:40HClmight it be possible to add a logtype to logf, and a filter on what type of log messages to store/print?
23:52:45preglow28k8 life is not good life, internet wise :/
23:53:31HClor would that be bloated and should i just comment out the messages i don't want
23:53:55CoCoLUSwhat i want to say is, when the release branch is ready, would there still be someone interested in it? it might look fairly outdated in comparison to the main tree, with all it's new features, albeit not stable
23:54:01HClpreglow: especially when you're used to 100mbit..
23:54:29preglowno shit, i left that connection lying battered and bruised in an empty flat some four days ago
23:54:35HClCoCoLUS: thats always the case in most open source projects, and yes, most of the time people will look at the stable release first
23:54:38BagderCoCoLUS: all those who use 2.4 today, yes
23:54:40CassandraLots of people stick to the stable branch of Rockbox.
23:54:59CassandraI probably would if I didn't do development work.
23:55:05CassandraLet someone else work the bugs out.
23:55:13CoCoLUS2.4 was before the iriver development started, right?
23:55:50 Quit Febs ("Chatzilla [Firefox 1.0.4/20050511]")
23:56:36amiconn2.4 was 20-something Dec 2004
23:56:45Bagderyes, and iriver work started earlier
23:56:51amiconnIriver started earlier
23:57:05Bagderin september 04 in fact
23:57:20amiconnAround the same time as Ondio dev started...
23:58:17*amiconn just saved a whole conversion layer in the win32 player sim

Previous day | Next day