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 2005-04-23

00:03:12*t0mas is going to bed :)
00:03:21 Quit t0mas ("c ya tomorrow")
00:03:34*amiconn is going to commit the fat.c fixes
00:13:00 Join DMJC [0] (
00:16:20 Join webguest09 [0] (
00:16:25 Quit webguest09 (Client Quit)
00:17:01 Join Leperkawn [0] (
00:17:20*Leperkawn dances
00:18:22ehntoowhat are you doing in here
00:18:28LeperkawnHola senor
00:18:35LeperkawnThis is a very random meeting.
00:18:47ehntoomy channel! I was here first!
00:18:51*ehntoo kicks Leperkawn
00:18:53*Leperkawn slaps ehntoo
00:19:10LeperkawnYour op status doesnt help you here1
00:19:13Leperkawn1 = !
00:19:30 Join lostlogic [0] (
00:19:47 Nick Leperkawn is now known as Lep_idle (
00:20:49thegeekdid anyone manage to compile rockbox with gcc 4.0 ?
00:21:10rasherHCl did for iRiver.. doesn't run though
00:21:42Bagdersee the wiki
00:21:49thegeekthe great wiki
00:21:57thegeekdid not think anyone would bother adding it there;)
00:22:04***Saving seen data "./dancer.seen"
00:22:16*ehntoo bows down to the greater powers that be residing in the wiki
00:32:54preglowthere are problems with my code for gcc4 as well
00:32:59preglowhcl didn't compile that
00:33:04preglowand i'm not about to fix it now
00:33:27rasherreal problems?
00:33:35rasherhaha, RoboCop!
00:33:53preglowno, just seems that gcc4 is a lot more fond of letting go of the registers i want
00:34:01preglowlot LESS fond
00:39:27 Quit matsl (Remote closed the connection)
00:41:08amiconncvs compile is lagging....
00:42:33 Quit Harpy (Read error: 60 (Operation timed out))
00:43:02 Quit Sucka ("a bird in the bush is worth two in your house")
00:44:38micoois Vladimir Pantelic in here?
00:50:17rasherwrong window :)
01:00:18 Quit DMJC ("Leaving")
01:08:52MoosCamarogood night all
01:09:12 Part MoosCamaro
01:17:24 Quit ehntoo (Read error: 60 (Operation timed out))
01:27:55 Join Welli [0] (
01:28:31 Quit Welli (Remote closed the connection)
01:46:59 Join KingFrog [0] (
01:47:47KingFrogany rockboxer here ?
01:48:11rasherdepends on your definition of rockboxer I guess
01:48:53KingFroguhmmm... 'someone who knows stuff well enough' (?)
01:48:53thegeekI rockbox, therefore I am
01:48:58oddi've been punching stones for days... 9 of my 10 fingers are broken in multiple places
01:49:53KingFroganyway, i was just wondering, is there a way to replace the graphics in a f/w? (i.e. the icons and stuff)
01:50:21rasherthe rockbox graphics?
01:50:53 Quit Lep_idle ("Chatzilla 0.9.67 [Firefox 1.0.3/20050414]")
01:51:08KingFrognope, the original ones
01:52:14 Join einhirn [0] (
01:52:40rashernot really a rockbox question then :) .. try looking n the forums on
01:53:24KingFrogI was prefering to archos players, not iRiver...
01:53:34rashersorry :)
01:54:13KingFrog(damn, my english is getting bad...)
01:54:49KingFrogso, no hope?
01:54:58rasherno idea, sorry
02:02:29 Join biptoria [0] (
02:02:52biptoriaowt worth testing today my little loverleys ?
02:09:04 Quit biptoria ("CGI:IRC")
02:20:06 Quit idlewild (Read error: 104 (Connection reset by peer))
02:22:08***Saving seen data "./dancer.seen"
02:25:43 Join idlewild [0] (
02:30:21 Join ehntoo [0] (
02:50:57 Quit idlewild ()
02:55:43 Join noclue2 [0] (
03:12:34 Join botaxxs [0] (~BoTaXxS@
03:12:39 Quit preglow ("hear me now!")
03:12:40 Quit ehntoo (Read error: 110 (Connection timed out))
03:18:08 Join courtc_ [0] (
03:23:30 Quit courtc (Read error: 110 (Connection timed out))
03:29:01 Nick courtc_ is now known as courtc (
03:36:15 Join ehntoo [0] (
03:36:47 Join ashridah [0] (
03:52:23 Join _n2_ [0] (
03:52:25 Quit noclue2 (Read error: 110 (Connection timed out))
03:56:23 Join noclue2 [0] (
03:59:43 Quit ehntoo (Read error: 60 (Operation timed out))
04:02:09 Join ehntoo [0] (
04:05:48 Join QT_ [0] (as@area51.users.madwifi)
04:09:19 Quit noclue2 (Read error: 145 (Connection timed out))
04:12:20 Quit _n2_ (Read error: 110 (Connection timed out))
04:13:04 Join Naked [0] (
04:13:24 Nick Naked is now known as Hadaka (
04:15:32 Join noclue2 [0] (
04:17:09 Quit QT (Read error: 110 (Connection timed out))
04:22:09***Saving seen data "./dancer.seen"
04:32:14 Quit ehntoo (Read error: 110 (Connection timed out))
04:33:42 Join Twingtwing [0] (~BoTaXxS@
04:33:43 Quit botaxxs (Read error: 104 (Connection reset by peer))
04:39:40 Quit noclue2 ("Leaving")
05:06:49 Join StrathAFK [0] (
05:28:44 Quit Strath (Read error: 110 (Connection timed out))
05:47:44 Quit Aison (Read error: 104 (Connection reset by peer))
06:17:27 Join stevenm [0] (
06:17:36stevenmHello people. What's all the business w/ GCC4 ?
06:18:00stevenmDo I need to update my devkit? Etc?
06:22:11***Saving seen data "./dancer.seen"
06:28:44 Join johntb [0] (
06:31:30ashridahstevenm: only if you like fixing bugs brought on by a new compiler
06:36:29stevenmashridah, nah, no thanks :)
06:36:42stevenmSo we're staying with 3 for the time being ?
06:42:37ashridahit would appear as much. SH-based stuff doesn't compile in 4 yet, and coldfire stuff compiles, but crashes immediately on boot or something.
06:43:09ashridahi'd be inclined to wait for the platform to stabilise and lesser architectures to get a good prodding before switching to gcc 4 myself.
06:47:21 Quit johntb ("Leaving")
07:19:41 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
07:31:27 Join XShocK [0] (
08:08:11 Nick StrathAFK is now known as Strath (
08:13:18 Quit stevenm ("Leaving")
08:22:14***Saving seen data "./dancer.seen"
08:44:03 Quit edx (Read error: 110 (Connection timed out))
09:04:17 Quit Hornet__ (Read error: 54 (Connection reset by peer))
09:05:01 Join lostlogic_ [0] (
09:06:06 Quit XShocK (Read error: 104 (Connection reset by peer))
09:07:46 Join Hornet [0] (
09:22:10 Join Harpy [0] (
09:23:33 Quit lostlogic (Read error: 110 (Connection timed out))
09:38:04 Join Lost-ash [0] (
09:38:06 Quit ashridah (Nick collision from services.)
09:38:16 Nick Lost-ash is now known as ashridah (
10:16:56 Join pfavr [0] (
10:22:18***Saving seen data "./dancer.seen"
10:25:30 Join edx [0] (
10:29:51 Quit pfavr ("ChatZilla 0.9.61 [Mozilla rv:1.7.7/20050420]")
10:39:50 Join matsl [0] (
10:50:24*HCl yawns.
10:50:46*HCl guesses he'll go and start on implementing the new database layout
10:55:34 Join t0mas [0] (
11:05:32*t0mas is away: breakfast
11:15:19*HCl wonders why the vars of the header of the tagdb are all global variables instead of properties of a struct, shivers, and changes :x
11:27:12 Join Tang_ [0] (
11:28:01Tang_"<KingFrog> anyway, i was just wondering, is there a way to replace the graphics in a f/w? (i.e. the icons and stuff)"
11:28:49Tang_the short answer is yes even in amodified with Rbx booloader HEX...
11:28:56rasherNo :)
11:29:02rasherHe meant in the Archos firmware
11:29:13Tang_hi rasher
11:29:19Tang_mistake sorry
11:29:19rashermade the same mistake myself :)
11:31:30Tang_is there so many Rokbox potential contributors using windows?*
11:32:06rasherno idea
11:32:40Tang_the crosscompiler is maybe motivated by something else?
11:36:20HClis cvs dead...?
11:36:41HClno. its me.
11:38:35 Join Patr3ck [0] (
11:47:10t0mas<Tang_> is there so many Rokbox potential contributors using windows?* <−− what do you mean?
11:47:24Tang_hi t0mas
11:47:25t0masyou want to know if it's possible to dev for rockbox on windows?
11:47:31Tang_in fact i wonder if
11:47:38t0masit is :)
11:47:58Tang_the crosscompiler is motivated by a "concrete" statement
11:48:13t0masI did a big part of the bmp thing on windows... as my laptop was on the office...
11:49:01Tang_as if several windows coders werent able to contribute since they were using windows OS
11:49:42t0masyou need cygwin or the rockbox Devkit
11:49:50t0masor access to a linux box for compiling...
11:50:21Tang_no that wasn't what i meant
11:50:33Tang_anyway not a problem
11:50:38t0masah... then I don't understand the question :)
11:50:44t0masI just used this for devving on windows:
11:50:46Tang_was just my curiousness
11:51:02t0masthe Devkit there :)
12:15:36 Quit KingFrog ("May the be with you.")
12:16:46HClso far my database adaptions are going okish
12:16:53HClMax genre length : 12 (Eurodance)
12:16:55HCl :)
12:17:38 Join MoosCamaro [0] (
12:17:46MoosCamarohi all
12:18:06MoosCamarohi Tang
12:21:07 Quit Twingtwing (Read error: 104 (Connection reset by peer))
12:22:19***Saving seen data "./dancer.seen"
12:26:17*HCl just thought of a fairly good way to keep it possible to make the songdb without needing the old songdb..
12:27:07HClwe can store the hash of the file in both the tagdb and the rundb, and just make all rundb offsets -1 (invalid)
12:27:13HClwhen you update the db
12:27:33HCland then you can synchronize on the player by comparing hashes. (between the two databases)
12:27:53HClwe'd still need a version capable of doing it off-target though :3
12:36:17 Quit Tang_ (Read error: 110 (Connection timed out))
12:44:57 Quit Patr3ck (Read error: 104 (Connection reset by peer))
13:16:55 Join idlewild [0] (
13:32:38*HCl slaps vim
13:44:32 Quit matsl (Remote closed the connection)
13:47:55 Join Aison [0] (
13:52:01 Nick QT_ is now known as QT (as@area51.users.madwifi)
13:54:07 Join Welli [0] (
13:54:21 Quit Welli (Client Quit)
14:10:31 Quit Strath (Read error: 110 (Connection timed out))
14:12:33 Join ehntoo [0] (
14:20:37 Join noclue2 [0] (
14:22:20***Saving seen data "./dancer.seen"
14:22:33MoosCamaroHCL: hey your works progress fine
14:27:29*HCl is working on the database v2
14:27:37HCli think i got the updated to the new format now
14:27:48HClaside from not making filehashes yet
14:35:36HCli dunno who called the perl script of death, but its not that bad o.o.
14:35:58rasherthe way it builds the database is fairly convoluted I'd say
14:37:16 Quit ehntoo (Read error: 110 (Connection timed out))
14:38:44 Join Sucka [0] (
14:40:47*HCl searches his dictionary for the word convoluted o.o.
14:41:16HCli guess o.o.
14:59:14*HCl thinks he got the new to work :)
14:59:18*HCl eats lunch
15:12:27 Join Shagnar [0] (
15:13:00HClhey shag
15:14:15Shagnarhello :)
15:18:52Shagnarsup? ?
15:19:02HClshort for "whats up?"
15:19:13HClyay soup
15:19:26HClanyways o.o.
15:22:33*Shagnar ate soupe few minutes ago ;-)
15:31:46 Part amiconn
15:32:14*HCl has never tried the tagdb before
15:33:46*HCl twitches.
15:33:52HCl"System Volume Information" !?!
15:33:59HClokay, i need to make sure to delete that one o.o.
15:35:47rasherI hate that
15:36:08Shagnarit's a windows-folder, but for what?
15:36:14rasherany way to turn it off?
15:36:15HCli saw restore points in it, i think.
15:36:22rasheryeah, it does
15:36:29HClit doesn't even show on my drive in windows >.<
15:36:37HCli need to find out how to make it visible
15:36:37rasherIt hides them
15:36:43HCli already have show system files
15:37:00HClhide protected operating system files.
15:37:04HCl*no* bad windows.
15:37:16ShagnarHCl perhaps try it over cmd
15:37:38HCltrying to disable hiding of operating system files throughout my laptop now
15:37:53HClnow its showing.
15:38:42HClheh o.o..
15:38:53HClMax songs per album: 712 (<no album tag>___/verrukkelijke715)
15:39:02HClthats gonna be painful..
15:39:10HCl1345 * 712*4
15:39:12HClfor every album.
15:39:20 Join Lost-ash [0] (
15:39:24HClum. o.o.;
15:39:34HCli gotta be wrong about that.
15:39:49*HCl nods, is.
15:39:55HClfor all my albums*
15:39:58HClnot every o.o
15:40:09HCl3.8mb of my tagdb is dedicated to album info, heh ;x
15:41:31 Quit Nibbler ("blubber")
15:46:35 Join cYmen [0] (
15:56:33 Quit ashridah (Read error: 110 (Connection timed out))
15:57:53 Quit Aison ("( :: NoNameScript 3.72 :: )")
15:59:00*HCl ish dumb.
16:10:49*Lost-ash hands HCl the cookie of self-analysis
16:18:49*HCl eats it.
16:22:23***Saving seen data "./dancer.seen"
16:34:52 Join botaxxs [0] (~BoTaXxS@
16:35:28 Join bagawk [0] (~Lee@bagawk.user)
17:03:52 Join preglow [0] (
17:10:44 Quit Lost-ash ("Leaving")
17:11:50HClhey preglow.
17:11:59*HCl 's songdb is broken with the album generation..
17:13:53HClwhats funny? o.o.
17:14:12preglowbroken things are funny
17:14:19HClwant to fix it for me then? :P
17:14:52preglowhaha, i said 'funny', not 'my area of experitise'
17:15:12preglowwhat are you using for db building thus far
17:15:13preglowthe perl thing?
17:15:50HCli'm not sure how, but the reference to albums from songs broke
17:16:32preglowbut no, i'm sorry, i'm a bit hard to motivate when i know i'll almost certainly never use the thing in question
17:16:47HClthen don't make fun of it either, please.
17:17:56preglowi didn't MAKE fun of it :P
17:18:25preglowbut no, no rockbox for me today anyway
17:21:19 Join mirak [0] (
17:22:06HCli found a bug in the current
17:22:14HClit wasn't even my fault
17:22:17HClit was already there.
17:22:23*HCl goes to commit fix.
17:26:39HClits still broken o.o.
17:26:45HClAlbum-> artist links are dead too.
17:27:09HClsomeone was really sloppy with the backwards links..
17:27:24HClthough we don't use them on rockbox, they need to be correct
17:27:48rasherI may have broken it recently
17:28:05rasherwhen trying to add support for albums with different artists
17:28:38rasherif that's not possible with the current suggestion, I'll cry :'(
17:29:26HClalbums with different artists?
17:29:47HClwell, this code here is definately wrong.. the one that maps albums to artists...
17:29:48rashermulti-artist albums
17:29:56MoosCamarocompilations for example
17:30:12HCl my $aoffset = $artistcount{$moo[0]} * $artistentrysize;
17:30:26HCl$moo[0] is the album name...
17:30:34rasherI'd be annoyed if I got 30 instances of "Matrix Reloaded Soundtrack"
17:31:00HClwhen $artistcount takes an artist name as argument
17:31:08HCland nah, it'll be possible.
17:31:08HClmaybe we should just get the album->artist link out?
17:31:11HClits not really needed anyways.
17:33:03rasherhaven't touched that iircah, so it was nmy fault :)
17:33:03HClhehe, ok.
17:33:03HClanyways. someone tried to lookup an artistname, but fed it the album name, that simply does not work :P
17:33:03HClgeeze, i'm glad i wrote a db tester :P
17:33:03HCllots of bugs in the current one
17:33:03 Join rashums [0] (
17:33:03rashumsHCl: the first "fix" is not needed afaics
17:33:07rasher- if($id3->{'ALBUM'}."___".$id3->{'ARTIST'} ne "<no album tag>___<no artist tag>") {
17:33:11rasher+ if($id3->{'ALBUM'}."___".$id3->{'DIR'} ne "<no album tag>___<no artist tag>") {
17:33:18rasherthat'll always be false now
17:33:22rasherer, true
17:33:30HClyea, well. okay.
17:33:32HClbut still
17:33:35rasherbut nevermind
17:33:36HClartist was wrong
17:33:40HClso i just fixed it everywhere.
17:33:52HClwould like..
17:33:59HClmapping the album name to the songs
17:34:04HClthen taking the artist of the first song
17:34:08HClbe acceptable for the artist of an album?
17:34:29rasherhow about a special case?
17:34:35HClwhat do you mean? oo.
17:34:43rasherlike "<multiple artists>"
17:34:45HCli'm not very good at perl programming, this is pretty much what i can still manage xD
17:34:46rasheror something like that
17:35:07HClwell, thats fine by me, only i don't know how to add that, heh.
17:35:19HCli can just take the artist of the first song for now
17:35:23HClat the moment its completely broken
17:35:27HCljust giving nonsense
17:36:12 Quit rashums (Client Quit)
17:37:14*HCl will commit.
17:40:16*HCl needs a better testcase than 5 random songs he happened to have on his ftp
17:40:50 Join webguest82 [0] (
17:41:04webguest82Hello, everybody
17:41:26 Join Psy^Dead [0] (
17:41:32Psy^Deadanybody home?
17:42:03webguest82I have some questions about xs 200 rockbox
17:42:24webguest82can anybody help me
17:42:33preglowjust ask
17:42:52Psy^Deadhello webguest
17:42:55webguest82is it possible to use rockbox on xs 200
17:43:03Psy^Deadnot yet
17:43:06Psy^Deadi think
17:43:12Psy^Deadwhat is an xs200?
17:43:26webguest82and if not is there a possibility that rockbox wil support it in the future
17:43:34webguest82yes gmini xs 200
17:43:35rasherThere's a port underway
17:43:39rasherbut it's stalled right now
17:43:53webguest82i really need it
17:44:08rasherYou're more than welcome to help out
17:44:23Psy^Deadwhy do you need it?
17:44:42webguest82because the original one sucks some how
17:44:44preglowthere'll be a long time before rockbox is usable by ordinary people on gmin
17:44:52preglowunless more people start to pitch in
17:45:02webguest82I would like to help
17:45:13rasherwebguest82: read the page I linked
17:45:13Psy^Deadhow many people contributed to getting iriver rockbox to where it is?
17:45:14pregloware you a programmer?
17:45:15webguest82but i'm not like in programming that well
17:45:20preglowthen you can't help much
17:45:27preglowi'm afraid
17:46:39webguest82perhapse i know some people who could help
17:46:59webguest82thx for the link
17:46:59Psy^DeadHCl here?
17:47:59webguest82see ya
17:48:03 Part webguest82
17:48:05Psy^Deadany news on that tag database proposal?
17:48:11Psy^Deadi saw your wikipage
17:48:16Psy^Deadit looks interesting
17:48:24Psy^Deadanyone actually making it happen?
17:48:31HCli'm actually working on it, yea.
17:48:39HCli think most people finally agree at least somewhat.
17:48:51Psy^Deadi think it sounds good :)
17:49:09HClat the moment i'm mostly trying to get everything to work as it did with the new format
17:49:09Psy^Deadits like how i use winamp's media library
17:49:16 Join _CoCoLUS [0] (
17:49:19HClbut for now its bugfixing the old format
17:49:31rasheractually.. wait
17:49:34 Quit CoCoLUS (Read error: 131 (Connection reset by peer))
17:49:34HCli'm bumping into a lot of bugs with the database tester i wrote that aren't actually my bugs :P
17:49:41rasherdo we really need an album->artist link?
17:49:43HCli'm not gonna commit it yet rasher
17:49:43 Nick _CoCoLUS is now known as CoCoLUS (
17:49:49HClno, like i said earlier, we don't :P
17:49:54HClwhich is one of the reasons
17:49:57HClto why it could be broken
17:49:59HClwithout people noticing
17:50:10Psy^Deadwhy does hcl
17:50:15Psy^Deadwrite in haiku? :P
17:50:22HClhaiku? :x
17:50:31rasherSo there's no problem with multi-artist albums reall?
17:50:36HClyou mean not long sentences at once? o.o.
17:50:41HClrasher: nope, none at all.
17:50:43rasherif we adjust the format slightly
17:51:01HCli don't think we even need a format adjustment, i dunno.
17:51:09HCli'm not focusing on that at the moment..
17:51:16rasheralright, I wasn't sure.. thought there was an album->artist table
17:55:14*HCl shakes his head
17:55:16HClnot as far as i know
17:55:24HClthough i might be wrong.
17:55:45*HCl goes to check on what he's broke in his
17:56:16rasherI'm inclined to suggest starting from scratch, picking bits from
17:56:37HClyea, well. um.
17:56:43HCli'm not good enough at perl for that
17:56:47rasherheh :)
17:56:48HCland its not that bad, really.
17:57:02rashermay just be me
17:57:15rasheralso, I think I was looking at it like really late one night
17:58:37HCli found what i broke :)
17:58:41HClthats the last bug :)
17:59:00Psy^Deadhcl, still planning rockbeans?
17:59:06HClbut its long term.
17:59:17HCli'm definately gonna make it once i find some time
18:03:50 Quit idlewild ()
18:05:59HClworks :)
18:06:12*rasher ponders having a second look at "rockout"
18:06:32HClwould be nice, it would make a good base for rockbeans if you make it fairly abstract
18:06:50*HCl fixed his last bug in the
18:07:01rasherI'm basing it on pong :)
18:07:05HClohh wait.
18:07:06 Join johntb [0] (
18:07:07HClyou're right
18:07:10HCli'm confused XD
18:07:19HClnevermind :)
18:07:20rasherI'm right?
18:07:30HClyea, i thought you meant rockblox for a sec
18:07:35HCl :P
18:07:49*HCl ish happy that his works perfectly now :)
18:07:55*HCl goes to celebrate with chocolate milk
18:08:22HCli'd commit it, but its not very useful yet
18:08:34HClaside from working with the accompanying testdb.c
18:12:56 Join cYmen_ [0] (
18:13:31 Quit cYmen (Client Quit)
18:14:10 Join XShocK [0] (
18:17:09 Join LinusN [0] (
18:18:00LinusNHCl: the "quarrel" with bluechip goes way back, long before the license dispute
18:18:04 Join idlewild [0] (
18:19:25LinusNthe main issue being that he wants to be anonymous, and we don't want anonymous contributions
18:20:03LinusNthe has been numerous arguments about a lot of things
18:20:25LinusNso we decided that we can live without his code
18:21:00LinusNthe "audio" filter stuff is nothing but a pretty way of setting the registers in the MAS MP3 chip
18:21:20LinusNwe have most of that functionality in rockbox anyway
18:22:24***Saving seen data "./dancer.seen"
18:23:29*rasher is annoyed by the ahx codec thing
18:23:31preglowwants to be anonymous?
18:23:55HCli agree that we don't need it.
18:24:03preglowsounds paranoid
18:24:06HClhe does.
18:25:40HClany objections against adding testdb.c and to the cvs?
18:25:43rasherthere was something in the mailing list recently
18:25:49HClthey'll have do not use new database format etc printed when ran
18:26:11HCli'm thinking they might be nice to show what the new format is like if anyone's interested o.o.
18:26:19HClnah, i'll just put them on my ftp instead o.o.
18:26:55rasherI totally need to convert my mail to maildir :-\
18:27:10HClafter 7 years of C coding.
18:27:18HCli *finally* learned how to do the do while statement
18:27:38LinusNHCl: i don't mind it being in cvs, as long as it doesn't mess anything up
18:27:42HClLinusN: okay
18:27:44HCli'll commit it then
18:27:49HCland no, its not going to mess anything up.
18:27:53LinusNbut testdb.c is a bad name
18:28:01HCli'm open to any suggestions
18:28:10HCli renamed it to testdbv2.c at the moment, but go ahead and suggest something
18:28:28LinusNcvs isn't very good at renaming and moving stuff
18:28:36HClyea, i know
18:28:41HCli haven't committed it yet, so :)
18:28:56HClwhat do you want it to be named?
18:29:00preglowany plans to move to svn? :>
18:29:13LinusNpreglow: we've talked a lot about it
18:29:27LinusNHCl: i haven't followed the discussion
18:29:44LinusNis it a new database, or an extension of the existing?
18:30:51HCla little bit of both.
18:30:59HClpretty much a rearrangement of the existing
18:31:02HClwith extensions.
18:32:09LinusNok, let's wait with the commit for now
18:32:30LinusNi'll have a word with bagder tomorrow
18:32:35HCli'll just put it on my ftp then with links to it on the wiki
18:32:49HClyea, i'm not gonna commit any code that changes it till it works decently
18:33:20HClat least now that i got all the bugs out of, it shouldn't be too hard to adapt the code into being able to read it, since my test program can read it fine now
18:33:31 Quit bagawk ("Leaving")
18:33:35HClhmm, except that i need to test the fileentry structure, forgot that.
18:34:27preglowshouldn't 'moveq.l #0xb0, %d0' work? i can't see why it shouldn't work
18:35:18LinusNmoveq holds the immediate value in the opcode itself
18:35:21preglowbahh, it's probably meant to be signed data
18:35:28LinusNand there are only 3 bits for that
18:35:34preglowno, there are eight
18:35:41preglowaddq and co uses three bits
18:35:45preglowmoveq uses eight
18:36:14preglowmov3q uses three, but isn't supported in our coldfire
18:36:48preglowbut i think gas somehow expects signed data there
18:37:20LinusNnothing in the manual says it should be signed
18:37:35LinusNis this in an "asm" statement?
18:37:42 Quit lostlogic_ (Client Quit)
18:37:53preglowcoldfire reference says the data are sign extended, so might be some foolishness involved
18:37:56preglowit's an .S file
18:38:16LinusNso try using a signed expression then
18:38:54LinusNbut you might not want it sign extended...
18:40:25*LinusN has to reboot
18:40:28 Part LinusN
18:40:42 Quit Psy^Dead (Read error: 60 (Operation timed out))
18:42:08 Join LinusN [0] (
18:42:17HClgreat. fileentry works fine
18:42:34 Join Chamois [0] (
18:43:41preglowi might not want it sign extended indeed, i'll just use a immediate move, it's just a %macsr setup anyway ;)
18:44:21preglowand it was a sign thing, moveq.l #-100, %d0 worked fine
18:47:58 Part LinusN
18:55:40 Quit Ka_ (Read error: 110 (Connection timed out))
19:00:35 Join Ka_ [0] (
19:09:03 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
19:11:29 Join muesli- [0] (
19:12:14 Quit botaxxs ()
19:13:59 Quit edx (Read error: 104 (Connection reset by peer))
19:19:11muesli-so what?
19:19:11 Join edx [0] (
19:19:37HClhow goes muesli- ?
19:19:38*muesli- yawns to hcl
19:19:45muesli-just yawning ;)
19:19:49muesli-hi rasher
19:19:57*HCl gets thrown all the way to the back
19:20:02HClwatch where you're yawning
19:20:03HCl :p
19:20:11 Quit edx (Client Quit)
19:21:07muesli-no worries mate, will not fight for the top yawning position in this # ;)
19:21:41HCl :P
19:22:29preglowcatch the digital flow
19:22:53preglowwow, this is a new one
19:23:05preglowI0B:Line-F at 32F468FC
19:23:15HClold plugin?
19:23:23HClhave you recompiled cleanly?
19:23:27preglowshouldn't be
19:23:29HCland cvs updated
19:23:30preglowjust make cleaned
19:23:57preglowprobably faulty plugin api
19:23:59rasherthat looks fun
19:25:22preglowi thought plugin versioning was supposed to catch stuff like this
19:25:34 Join lolo-laptop [0] (~lostlogic@
19:26:18rasherdid you mess with the plugin api?
19:28:24preglowno, but someone did
19:31:28rasher17:18 <preglow> but no, no rockbox for me today anyway
19:31:37preglowyes, i know...
19:31:44rasherdon't stop! :)
19:32:00preglowi just granted myself another half hours attempt
19:32:08HCl :P
19:33:11 Join edx [0] (
19:38:46 Quit idlewild ()
19:56:32 Join Camilo [0] (
20:01:38preglowi wish this thing started working soon, i'm sick'n tired of looking at asm
20:03:09HClwhatcha working on?
20:05:02 Quit muesli- (Read error: 104 (Connection reset by peer))
20:06:52 Quit Camilo ("Chatzilla 0.9.67 [Mozilla rv:1.8b2/20050404]")
20:11:17 Join LinusN [0] (
20:14:46preglowHCl: imdct_l
20:14:48preglowas usual
20:16:46 Part LinusN
20:17:08preglowi'm not used to dealing with assembly stuff of this size, heh
20:17:21HClthats why we made programming languages ;x
20:17:38HClbecause asm code is sooo hard to keep clean and organized
20:18:22preglowi'd just LOVE to use c
20:18:29HCli know :/
20:18:39preglowi hate playing register allocator
20:18:40preglowi suck at it
20:21:44*preglow puts on cowboy bebop soundtrack
20:22:28***Saving seen data "./dancer.seen"
20:22:31 Join idlewild [0] (
20:22:36preglowi frigging love this music
20:23:02 Join Psy^Dead [0] (
20:23:11*Psy^Dead lurks
20:23:15HClhey Psy^Dead
20:23:33 Nick Psy^Dead is now known as nobby (
20:23:38HClif you're interested in the new database progress so far, there are two programs that can work it on my http
20:23:54nobbyi can wait til iriver progresses a bit
20:24:33nobbyhow much work is left to get from playing 5 seconds of mp3 to playing full tracks?
20:24:50rasherHCl: mind if I add more stuff to the status? Such as 0% save runtime info on play etc?
20:25:13 Quit idlewild (Client Quit)
20:25:18HClrasher: not at all.
20:25:30rasherjust to complete the list a bit
20:25:47preglownobby: only linus knows
20:25:58HClyea, tomorrow i'm most likely gonna fix it so it'll work with the code
20:26:02HCland then start on the runtime db
20:27:57rasherthere, added a few steps
20:28:08rashernot sure how well they fit
20:28:21HClits okay.
20:28:45 Join idlewild [0] (
20:45:35 Quit thegeek (Read error: 54 (Connection reset by peer))
20:46:10 Join thegeek [0] (
20:52:42preglowahhh, it appears i'm a fool
20:52:58rasheroh noes
20:53:09rasherdoes this mean that you found out what was wrong though?
20:53:53preglowwell, i certainly found out why it sounds like shit
20:55:40preglowand it also appears i have a significant coding job to do before i'll find out if it works properly
21:02:29preglowouch indeed
21:15:30 Quit mirak (Operation timed out)
21:15:51 Quit lolo-laptop ("Client exiting")
21:41:07 Quit idlewild ()
21:43:20 Join _n2_ [0] (
21:46:30preglowanyone know what @−−−− means in arm assembly?
21:46:34preglowliteral pool placement?
21:47:20rasherthat's an excellent symbol
21:48:49preglowgoogling it was not very easy, heh
21:49:10preglowmainly because google seemed to just ignore it
21:50:55 Quit noclue2 (Read error: 60 (Operation timed out))
22:20:33 Join stevenm [0] (
22:22:32***Saving seen data "./dancer.seen"
22:22:40 Nick _n2_ is now known as ehntoo (
22:30:44HClwhat would be faster? searching for a file hash, or binary searching for a filename?
22:30:51HCl(including hashing of the file)
22:31:25HClif we can make hashing a file fast, i might even be able to readjust the database to sort on hash instead so we can binary search on hash
22:32:12rasherhashing a file where?
22:32:16rasheron target?
22:32:21HClon target
22:32:27HClcompared to binary searching for the filename.
22:32:38rasherI could try
22:32:48HClwe definately need the hashes to be unique
22:32:50rasherwas there a crc routine somewhere?
22:32:58HClsomeone mentioned crc16
22:33:02HClits not likely to be good enough.
22:33:14HClwe'd need crc32 at the very least.
22:33:17HCli think.
22:33:18rashercrc32 is quickly good enough
22:33:28HClin theory, crc16 would be good enough
22:33:32rasher512bytes is easily enough
22:33:34HClbut it'd have to be a darn good hashing routine.
22:33:43HClwe need to test and find out.
22:33:46rasherI did
22:33:53HClits one of the issues i'll come across tomorrow
22:33:56HCldo tell
22:33:59rashertested crc of the first X and last X bytes
22:34:16rasheras in, read first X and last X, crc32 that
22:34:22rasherI'll find the results
22:34:39HClok, please do
22:34:49HCloh wait..
22:35:01 Quit Chamois (" Want to be different? HydraIRC -> <-")
22:35:04HCli'm forgetting stuff XD we want to allow files that are copies of eachother..
22:35:16HClthough they would have the same tag and runtime info though.
22:35:26HCli guess we could still look up files by hash.
22:35:28 Quit MoosCamaro (Read error: 104 (Connection reset by peer))
22:35:31HClits just a matter of which is faster.
22:35:37 Join MoosCamaro [0] (
22:35:39rasher"regarding my file-hash experiment: 512byte from start of file 512 kb from end of file produced no duplicates.. 256 256 produced one"
22:35:50rasheroh, kb :)
22:35:52rasherI was lying
22:35:57HClnot bytes?
22:36:02HCl512kb is a lot...
22:36:12rasherlet me check the script I used :)
22:36:13HCleither way, we're gonna need hashes.
22:36:18HClso we need them anyways.
22:36:35rasherno, bytes
22:36:35HCli just want to know whether hashing + binary search for hash is faster than binary search for filename
22:36:38rashernot kb
22:37:06Bagderif the hash is on the music, then I think you should search on the file name
22:37:28HClrasher: okay.
22:37:38rasherI think it'd be absolutely great to use hashes on the music
22:37:40Bagdersince you won't always have the music around, but you will have the file name
22:37:57HCli never considered people temporarily removing files.
22:38:11rashernot for this matter specifically, I've no idea which'll be better
22:38:20HClrasher: it would map identical files to the same tag/runtime info
22:38:34HCland we can sort the hashes to enable binary search
22:38:49HClcomparing a hash is much quicker than comparing every character of a string..
22:38:59HClits just the generating of the hash that can be slower
22:39:11rasherI'm not sure I can come up with a scenario where people would want to distinguish between several identical files
22:39:31HClno, we wouldn't really want to in the database anyways
22:39:35HClthe id3 tag would be the same
22:39:40HCland the runtime info would be the same
22:39:43HClwhich makes sense to me
22:39:45HClif they're the same file
22:39:54rasherthe id3 tag might be different
22:40:00HClnot if the hashes are the same it won't.
22:40:05rasherdepending on how we hash
22:40:13rasherwell my idea was to seek to the first mpeg frame
22:40:31HClokay, well... at the moment i'm assuming hashes will be unique..
22:40:43HClwe need to keep account of it if they won't be.
22:41:08rashereven assuming we hash as usual
22:41:15rasherthey might not be unique
22:41:26HCli dunno.
22:41:33rasherif they're the same file
22:41:33HClit simply depends on our hashing function.
22:42:05rasheroh, you mean hash the filename as well?
22:42:14HClno, not with the filename.
22:42:20HCljust a hash of the file data.
22:42:44rasher22:40 <HCl> okay, well... at the moment i'm assuming hashes will be unique..
22:42:50HCloh. sorry.
22:42:54rasherthen we can't make that assumption
22:42:55HClto files that aren't copies.
22:42:57HClof eachother.
22:43:52rasherI'll see if I can make a hashing function
22:44:05HCl :)
22:45:00rasherI expect crc32 to be available all over the net :)
22:52:50preglowmd5 :P
22:53:04rashercrc is supposed to be faster?
22:54:07rasherI think that's why bagder suggested it
22:54:29Bagdercrc32 is only 32bit while md5 has longer hashes
22:55:04BagderI'm fairly sure crc32 is faster
22:58:27rasherhow does one add something to viewers again?
22:59:20 Quit Harpy (Read error: 60 (Operation timed out))
22:59:38Bagderto the viewers.config file you mean?
22:59:52rasherI'm an idiot...
23:00:41rasherI was looking at one of the other lines, and wrote sort.rock instead :)
23:01:08rasheroh err..
23:01:12 Join Camilo [0] (
23:02:30rashernow I don't get it
23:02:47rasherthere's a line in viewers.config, the .rock is in viewers
23:02:55rasherbut it doesn't show in "open with"
23:03:00rasherwhat am I missing?
23:03:40Bagderyou playing with a new plugin?
23:04:23 Quit Ka_ (Read error: 110 (Connection timed out))
23:04:40rasheryes.. well, one to hash files
23:04:49rasherjust to test out how fast it could be
23:04:57Bagdercould it be the size of the buffer for the viewers?
23:05:18BagderI recall some problem with that before
23:05:31rashermaybe.. what do I do?
23:06:18 Quit Sucka ("a bird in the bush is worth two in your house")
23:08:12*Bagder checks
23:09:03 Quit Camilo ("Chatzilla 0.9.67 [Mozilla rv:1.8b2/20050404]")
23:11:28Bagderno, it should support 32
23:11:49 Join amiconn [0] (
23:12:10rasherI must be missing something then..
23:12:19Bagderhi amiconn, I have a very minor nit/question about your fat patch
23:12:33 Join Sucka [0] (
23:12:36amiconnrasher: Did you restart rockbox after correcting viewers.config?
23:12:46amiconnBagder: which is...?
23:12:59rasherI did
23:13:06rasherboth through rolo and regularly
23:13:24Bagderamiconn: there's a strncpy() of size 12 to a 12 byte buffer, which could leave the buffer without a terminating zero
23:14:23amiconnTalking 'bout line 1259?
23:14:50amiconnIn this special case, it can't
23:15:32amiconnThis part is only executed if 'dotdir' is true, and then name is always either '.' or '..'
23:17:39rasherGar.. it just won't show up!
23:18:27Bagderfinally a chance to run it in the sim and use gdb! ;-)
23:18:41 Join Ka_ [0] (
23:18:42rasherwhat, me?
23:20:20rasherwell.. how?
23:20:47Bagdergdb rockboxui
23:20:52Bagderb source.c:1234
23:21:41Bagderwhen it stops, use 'n' to single-step, 's' to step into functions and 'p' to display variables' contents
23:22:29rasherany good ideas to where to break?
23:23:10Bagderapps/filetypes.c has the viewers code
23:23:21Bagderread_config() could be a place to start
23:26:41rasherthis is going to take a while..
23:28:04 Join amiconn_ [0] (
23:28:13rasherages in fact
23:29:23rasherI'm still doing "viewer"
23:29:46Bagderset a bette break point and use 'c' (continue)
23:30:21Bagdercontinues until it hits a break point
23:32:47*rasher blinks at the code
23:35:35 Quit amiconn (Nick collision from services.)
23:35:35 Nick amiconn_ is now known as amiconn (
23:37:16 Join muesli- [0] (
23:37:27rasherit works fine if I assign it a fileextension and "play" one of those files
23:37:34rasherwhy wouldn't it show up in "open with"?
23:38:19Bagderrasher: apps/onplay.c the list_viewers() function I think
23:38:38rasherhm.. think I'll just leave it be for now
23:38:48rasheror not
23:38:52rasheroh, I will :)
23:39:11 Quit nobby ()
23:48:08HClwell, fileextension should work for now.
23:53:38rasherstill trying to find a description of crc32
23:55:29HCllet me just settle in my room and i'll try to help
23:55:56rashernoooo I *want* to do it by myself :)
23:56:43Bagdersearch on
23:56:49HClokay :)
23:57:59muesli-g'n8 folks
23:58:25 Quit muesli- ("ich will Khe!!!")

Previous day | Next day