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 2005-04-26

00:02:10XShocKwhen you move the cursor it clears a part of scrolling line, but not all of it... looks bad
00:03:27 Join Strath [0] (
00:03:36HClthat happens when you select all songs of an artist
00:04:03HCli happen to have a good testcase
00:04:07HCli have two of those odd things
00:04:12HClin my album nevermind of nirvana
00:04:15HCli'll check it out when i have time
00:04:44HCli suspect it to be caused by bad id3 tags
00:05:12MoosCamaroHCl: yes it do probably
00:05:44MoosCamaroin the old version of iriver fw there was it
00:05:55MoosCamarobefore bugs fixing
00:06:19MoosCamarofw 1.60 i think
00:07:46XShocKthis is what i exactly mean. :)
00:08:15HClodd, my nirvana album doesn't seem to have anything odd when i look at it with my tag tool
00:08:42HClperl bug? o.o bug in our id3 parser?
00:08:50XShocKwhen i look at all other thing i have it seems all right.
00:08:57XShocKonly those strange folders...
00:09:20HClXShocK: is that in all songs mode or something else?
00:09:36XShocKin artist, albums, and songs...
00:09:41MoosCamaroin the end of list no?
00:10:11XShocKin artist and album there is one folder, which contain another folder and <All songs>. when you go there you have that bug.
00:10:48XShocKyes. in the end of the list
00:10:49HClyou'll need to load it in the simulator before i'll be able to tell whats going on
00:10:56HCli added debug messages to it..
00:11:13XShocK"ypK" with y letter having two dots on it
00:12:26HClthe two songs with corrupted tags in my nirvana album seem to be drain you and lounge act
00:12:49HCli'd almost say its a perl bug, heh.
00:12:52MoosCamarothe ogg files did the sames caracters in fw iriver 1.60
00:13:08HClyes, well, this isn't iriver 1.60, and i don't know what their problem was.
00:13:12HClso thats not gonna help, unfortunately :/
00:13:19MoosCamaroor older i don't remenber
00:13:26 Part Querty ("Leaving")
00:13:57MoosCamaroXShock: it's for MP3 files
00:14:12HClthe tagdatabase doesn't even support ogg
00:14:31preglowthink i'll bloody buy that
00:14:33XShocKi have all music in mp3. but i also have some video, programs.... but music is only mp3
00:14:51HClgeeze thats crappy.
00:14:59HCli crashed my iriver by trying to get a context menu on an album
00:15:09HClthats a bug that must've been present in the old database as well
00:15:17*HCl goes to reset, then sleep
00:17:14HCli have a lot of those bad things in my nirvana albums, for unknown reasons
00:18:22MoosCamaroit don't like Nirvana:-)
00:18:38HCli haven't listened to them in a while
00:18:55HClbut i have plenty of space and no reason to delete them
00:20:25HClXShocK: i think i might know what causes the cursor thing
00:20:35HClXShocK: its always when selecting "all songs", right?
00:21:27HClit happens cause the original count is higher than the eventual results
00:21:42HClwhen the results don't fill an entire screen
00:22:25*HCl still has no clue about the odd tag thing though..
00:23:49***Saving seen data "./dancer.seen"
00:25:06HCli say its something in the id3 tag parser library..
00:25:08 Quit ashridah ("Leaving")
00:25:15HClit specifically happens on my nirvana folder
00:26:42*HCl goes to run his db checker on it
00:28:59MoosCamarogood night all
00:29:11HClits literally stored in the database like that
00:29:14MoosCamaroHCl: good luck
00:29:17HClso its a bug in
00:29:33 Part MoosCamaro
00:29:54 Quit Sucka ("a bird in the bush is worth two in your house")
00:30:08ehntoointeresting problem, it sounds like.
00:30:16ehntooI'd volunteer to help
00:30:25ehntoobut I've forgotten nearly all the perl I used to know... =\
00:32:22ehntoooh, god, I love how fast rockbox boots.
00:32:42HCl :P
00:32:52ehntooI use it all the time.
00:33:45ehntoohmm... just wondering... what software audio player do most of you use?
00:34:42*preglow hugs foobar2000
00:35:07*ehntoo also hugs foobar2000
00:35:38amiconn(if at all)
00:36:28HClXShocK: fixed.
00:37:40HCli dunno who imported the id3 tag reading bit in the script, i'd appreciate it if that person would help nailing that odd 3 character album/artist/title tag thing.
00:37:52HCli'm gonna go sleep now, fixed that graphic glitch
00:38:11HClehntoo: winamp
00:39:48ehntooI just wish somebody would make a player with something equivalent to fb2k's tagger.
00:39:59HClwe're opensource, we can do anything.
00:40:07ehntooyeah, but nobody's done it
00:40:08HCli dunno whats fb2k's tagger is though.
00:40:13HClgo ahead :P
00:40:31ehntooit fetches album info from freedb, and tags files automatically
00:40:42HCloh right.
00:40:49HClwell iriver can't connect to freedb
00:40:51HClso cratch that.
00:41:13ehntooclosest I've ever seen was "Quod Libet"
00:41:22ehntoonifty little python player for linux
00:41:35ehntoosmallest footprint I've ever seen
00:42:36*HCl goes to sleep.
00:53:54 Join silencer_ [0] (~silencer@
00:56:56ehntoonight, HCl
00:58:48XShocKi was not here, sorry
01:00:29 Quit silencer (Read error: 60 (Operation timed out))
01:15:27HClgheh, i think i might be able to put my compiler class to use when parsing search requests on the new database o.o *goes back to sleep*
01:16:28 Join silencer [0] (
01:32:13 Quit silencer_ (Read error: 110 (Connection timed out))
01:32:49preglowi love the remote logo
01:37:51XShocKyep, looks cool
01:39:06 Quit andY`fRa (Read error: 54 (Connection reset by peer))
01:48:33 Join austriancoder [0] (~austrianc@
01:48:37austriancoderhi all
01:49:00austriancoderhave just commited some remote lcd stuff
01:51:11preglowanything excessively cool? ;)
01:51:50austriancodertext writing to remote lcd works fine
01:52:36pregloweyh, great work
01:53:16austriancodernext step.. get filebrowser working correctly on remote lcd
01:53:24preglowgood luck on that one
01:53:27preglowwill take a bit of work
01:54:34austriancoderit will be hard to make everybody happy
01:55:34austriancoderi want it, that the remote lcd shows all but on file playing it should use a own wps
01:57:06 Quit Camilo (Read error: 60 (Operation timed out))
02:00:39preglowbut i'm off to bed
02:00:40preglowlater, all
02:00:41 Quit preglow ("fop")
02:11:13austriancoderneed some sleep too
02:11:19 Quit austriancoder ("using sirc version 2.211+KSIRC/1.3.12")
02:14:53ehntoowhere did my remote get to?
02:14:57ehntooI want to test this.
02:20:35thegeekwhen are the daily builds built?
02:20:38thegeekare 00.00 ?
02:23:52***Saving seen data "./dancer.seen"
02:24:21ehntoothere are "bleeding edge" builds
02:35:50 Join XavierGr [0] (
02:36:40 Quit XavierGr (Client Quit)
02:52:13thegeekwhen I tried to "open with - mpa2wav" using the tag database
02:52:25thegeekmpa2wav just made a 44 byte large file
03:16:22 Quit Aison` (Read error: 104 (Connection reset by peer))
03:34:08 Quit Seed (Nick collision from services.)
03:34:15 Join Seed [0] (
03:35:09 Join Aison [0] (
03:56:15 Quit cYmen ("zZz")
04:03:32 Quit Aison ("( :: NoNameScript 3.72 :: )")
04:06:01 Join QT_ [0] (as@area51.users.madwifi)
04:07:37 Join Aison [0] (
04:08:56 Quit QT (Read error: 60 (Operation timed out))
04:22:56 Quit Aison ("( :: NoNameScript 3.72 :: )")
04:23:55***Saving seen data "./dancer.seen"
05:12:10 Quit ehntoo ("Leaving")
05:13:10 Quit XShocK (" HydraIRC -> <- The professional IRC Client")
06:09:36 Join matsl [0] (
06:22:22 Join Vicious [0] (~irc@a41587bbe8f5a504.node.tor)
06:23:58***Saving seen data "./dancer.seen"
07:10:34 Join bobdbob [0] (
08:07:47 Join LinusN [0] (
08:18:15 Quit matsl (Remote closed the connection)
08:23:56*LinusN just discover that his entire mp3 collection is corrupt, probably due to a disk failure :-(
08:24:02***Saving seen data "./dancer.seen"
08:24:33*HCl dreamt something like that happened to him :/
08:25:03*HCl goes back to sleep :/
08:26:00LinusNhopefully i can rescue most of it from my old hard drive
08:26:15LinusNi copied the entire collection when i upgraded my pc
08:32:56amiconnLinusN: Don't you make backups?
08:33:53LinusNwell, in fact i have the collection in several places. it's still intact on my archos, but i added stuff on the pc without synching to my archos
08:34:30LinusNand i have it on my old hard drive, and possibly intact on my lacie usb drive as well
08:34:33amiconnAh, yes, that may happen to me too, but most of the time in reverse
08:34:47amiconn(adding it to the archos, but not the pc)
08:34:49LinusNyou mean corrupt on the archos?
08:35:14amiconnOn sunday I corrupted my Ondio's internal flash
08:35:27LinusNbacking up such large quantities of data is a tedious task, and takes lots of space
08:35:53LinusNamiconn: oh, can it be rescued?
08:36:15amiconnYes, I already did. I mean, I corrupted the file system
08:36:19dwihnothe first bacup takes a lot of time... the incremental are a lot faster
08:36:41 Quit DangerousDan (Read error: 110 (Connection timed out))
08:37:01amiconnLinusN: The corruption was most likely caused by playing around with the archos firmware and the 2 GB card.
08:37:32amiconnWhen I rebooted rockbox, I found rolo'ing was no longer possible; rockbox told me that it couldn't read the file
08:37:49amiconnchkdsk moaned about cross-linked files, lost cluster chains etc
08:38:31amiconnI checked the beginning of the disk with a disk editor. A bunch of sectors got overwritten with a "nice" pattern
08:39:09amiconnEvery 32 bytes there was a zero byte...
08:41:25LinusNmy first investigation of my corrupt hard drive seems to indicate a repeated pattern of corrupt sectors
08:41:40LinusNlike if the LBA scheme was changed
08:43:50LinusN...and it isn't just my mp3 collection... :-(
08:43:50amiconnBut imho that means the disk isn't physically corrupt?
08:45:27 Quit edx (Read error: 110 (Connection timed out))
08:45:30LinusNit's a brand new sata disk
08:54:31 Quit Strath ("Client closed")
09:10:01 Join B4gder [0] (
09:32:06 Join Nibbler [0] (
09:36:16 Nick Lynx_awy is now known as Lynx_ (HydraIRC@
09:41:45 Join Harpy [0] (
10:05:37 Join Patr3ck [0] (
10:14:09 Join ashridah [0] (
10:19:38 Join webguest54 [0] (
10:24:03***Saving seen data "./dancer.seen"
10:26:27 Join edx [0] (
10:35:09B4gdershould I offer a fwpatcher that should work?
10:36:17 Join Camilo [0] (
10:36:33 Join Vicious_ [0] (~irc@556cdec9640869ca.node.tor)
10:37:10 Quit Vicious (Remote closed the connection)
10:38:11 Join bobTHC [0] (
10:39:38bobTHCmornin' mates !
10:39:56B4gdergood morning
10:40:23 Quit Camilo (Client Quit)
10:40:39 Join austriancoder [0] (~austrianc@
10:41:22rashermorning, good work on the remote
10:43:06austriancoderi am thinking about a wrapper for the lcd's. So i can set with 1 call e.g. the font of both lcd's.. is this is a good idea?
10:43:49B4gderI think you should start in the other end instead
10:43:57B4gdermake code that uses the lcd
10:44:03B4gderremote lcd I mean
10:44:10B4gderthen you
10:44:21B4gder'll figure out what more functions that are needed
10:44:54B4gderadding functions just because "they might be useful" is not a Rockbox philosophy :-)
10:45:02austriancoder+g+ ok
10:45:23*austriancoder hacks now menu.c
10:47:19B4gderanyone with a fancy demo written for qt? ;-)
10:47:42rasher"fancy demo"?
10:47:56rasherI'd say kde is one big fancy qt demo :)
10:48:05B4gderI've written a framebuffer driver, and I'll now run qt/E on it
10:48:16B4gderI'd like something that demos how it works/speed
10:48:23B4gderI have no mouse and no keyboard
10:52:13B4gderand limited ram of course :-)
10:58:46austriancoderi can see the settings menu on my remote lcd :)
10:59:21*B4gder cheers
10:59:37rasherthat was fast
10:59:48austriancoderits not that hard, as i thought
11:00:00B4gderwe should hack the simulator to simulate the remote as well
11:00:12amiconnaustriancoder: How do you intend to deal with the differing line count on main & remote?
11:01:32austriancoderamiconn: i dont now that yet.. at first i only want to get something nice on the remote...
11:07:38rashermore wikispam
11:10:00oddif i may, i'd just like to give you guys a huge cheer and thank you for what you're doing... i'm an owner of an iRiver H-140 and I can't wait to use rockbox on it :D
11:10:08oddso <3
11:15:51rasher :(
11:17:40amiconnrasher: You shouldn't promote it.... that's what they want..... now the link is in the irc logs
11:18:15rasherwon't help them much when it's moved though?
11:18:48rasherplus, links within a website doesn't crank up google-juice do they?
11:22:09HClthey do
11:22:34rasherThat's crazy
11:22:37rasherI think..
11:22:42HClthe amount of pages linking to another webpage
11:22:51HCladds to the credibility of that webpage being actually useful
11:23:06rasherI would've thought "the amount of external pages"
11:23:11HClyes o.o.
11:23:16HClwhat did you mean then?
11:23:23oddyou can use that new google html tag thing to prevent it
11:23:27rasherPages on the same site
11:23:44rasherlike, the irc log linking to it doesn't matter one bit
11:23:52bobTHCperhaps a antibot sign in procedure to avoid wikitrick ?
11:23:52HClhaving a cold sucks
11:24:20oddjust use the tag
11:24:25odddont need antibot signin
11:24:59B4gderthe tag doesn't prevent people from trying
11:25:03odddesigned for blogs but there's no reason it wouldn't work
11:25:13oddB4gder: the bots look for it and don't both
11:25:20oddand 99% of it is bots
11:25:23B4gderI very much doubt that
11:25:33oddif they don't now, they will soon
11:25:39B4gderthey'd rather just take a shot and see
11:25:41oddit couldn't hurt is all i'm saying
11:26:14oddyou could also add a little notice at the top or bottom of the page saying "links on this website are formulated not to produce Google PageRank"
11:26:20rashereven if it wouldn't stop it, at least we'd be doing the Right Thing
11:26:21oddwith a link to that article
11:26:46rasherWe don't want that :-O
11:26:58oddwhy not?
11:26:58B4gderwell, I also have no idea how to make twiki add it
11:27:11rasherthis would make all links in the rockbox wiki useless
11:27:13oddyou could do it only for outgoing links
11:27:17rashernot useless
11:27:26rasherwell, it would make real information harder to find via google
11:27:37rasherthe wiki has a lot of good links
11:27:57oddi was just tryin ta help ;x
11:28:00rasherthat deserves the rank.. well, I guess
11:28:06 Join Aison [0] (
11:28:22rasherFirst I thought about only enabling it on userpages, but that would be even worse
11:28:41rasherthey'd start spamming the entire wiki then
11:28:46rasherright.. :X
11:30:26*HCl yawns
11:30:47HCli need an instant-beep mode on my alarmclock..
11:32:57bobTHCso the best way imho is a antibot trick, because rockbox have a very good google rank and spammer, target this kind of wiki to increase bullshit sites
11:34:24oddi guess that'd be easy enough to implement... just a little distorted random-char image on post/edit
11:35:03rasherplease no
11:35:06rasherjust on sign-up
11:41:16bobTHCi'm agree, just on sign-up, but random char image requiere php things, perhaps a simple custom mental calculation form...
11:41:45bobTHCwho trap bots
11:51:12bobTHCsome info here :
11:52:20 Nick QT_ is now known as QT (as@area51.users.madwifi)
11:54:16*HCl goes to make coffee.
11:56:06austriancoderwhere can i find the code for the filebrowser?
11:59:33amiconnFile browser and db browser are rather integrated. filetree.c, dbtree.c, and tree.c
12:00:04 Quit Vicious_ (Remote closed the connection)
12:00:23austriancoderamiconn: merci
12:00:40HClabout that.
12:00:55HCli was planning on adding a "fileview" to the database browser
12:01:30*HCl switches to wireless..
12:01:40rasherhow about a root menu where you can choose between db browser and file browser
12:01:44rasheror something
12:01:54rasherI know there's an easy way to switch on archos, but.. :(
12:02:45amiconnI really prefer it as it is now (the easy switch is only there on the archos recorders, btw)
12:03:13amiconnI see the db browser as a nice addition that's not really needed
12:03:23rasherI find myself wanting to switch back and forth
12:03:44rasherdo it all the time with the irvier db as well
12:04:42 Join MoosCamaro [0] (
12:05:50MoosCamaroHi all
12:06:12amiconnThe only thing I used the db for so far is search for tracks... which I need really really rarely
12:06:34amiconn...and which even can be done without the db
12:06:51amiconnJust create a root.m3u and use the search plugin
12:07:05 Join cYmen [0] (
12:07:14rasherHm.. my music isn't organised like the db browser is.. and both ways mean something to me
12:07:25*HCl goes to make a little todo for the database on the wiki
12:10:17 Join Vicious [0] (~irc@e9079288032ac90b.node.tor)
12:12:41*HCl yawns.
12:12:53HClwhat is .tor...?
12:15:28oddi'd guess a file for use with the program Tor?
12:19:22MoosCamaroaustriancoder: congrates for your remote works
12:21:40HCli meant that persons tld
12:21:44HCl12:07 -!- Vicious [~irc@e9079288032ac90b.node.tor] has joined #rockbox
12:21:48HCl .tor ?
12:22:39HClspoofed, apparently.
12:22:39oddNo country name found for .TOR
12:22:59oddhmm actually maybe it's an irc connection through tor
12:23:05oddi know tor uses "nodes"
12:23:11oddi have no idea how it'd make a mask like that, though
12:24:05***Saving seen data "./dancer.seen"
12:25:02rasherprobably freenode doing that
12:25:17*HCl has one hour till class.
12:25:20HClshould i check my notes.
12:25:27HClor should i spend some time on rockbox.
12:25:57austriancoderMoosCamaro: thanks... today i will do an other commit, to use the remote lcd much more
12:26:21MoosCamarogood news
12:26:45MoosCamaroaustriancoder: bonne chance et merci
12:40:56*austriancoder has added a splash_remote function
12:44:10B4gderI don't like how the lcd code has most of the functions duplicated from the ordinary h100 lcd
12:44:21B4gder(which in turn has too much duplicate from the lcd-recorder)
12:44:45austriancoderi know.. i dont like ot too.. but there isn't an other way
12:44:51B4gdersure is
12:45:23B4gderthere are several ways to do it
12:45:33austriancoderfor instance
12:46:42austriancoderi think, now is the right moment for the new graphics api
12:46:44HClsplah(LCD_REMOTE, 2, "blah");
12:47:04B4gdera new graphics api doesn't solve this
12:47:10B4gderit just provides a new api
12:47:15austriancoderHCl: will do it this way now
12:47:37B4gderadding a numeric on all functions is a pain too for those that don't have more than one
12:47:42HClits just a suggestion, it requires a lot of stuff..
12:47:56HCllike checking whether thats actually possible on the target lcd, whether you're not going out of its x,y range, etc.
12:47:58B4gderI was more looking for making low-level functions generic with specific lcd / lcd_remote wrappers
12:48:05austriancoderB4gder: i have looked at other lcd drivers.. and all of them have duplicated functions
12:48:19B4gderI said that too
12:48:26B4gderrepeating the same mistake doesn't make it better
12:48:40austriancoderi told you some time ago, that i want some wrappers, and you told me, not to do it
12:48:49B4gderI did not
12:49:09B4gderyou were talking about different wrappers then
12:49:36B4gderI'm talking about lcd_drawline() and lcd_remote_drawline() should use the SAME underlying function
12:49:50B4gderand so should the lcd_drawline for recorder
12:49:51austriancodermaybe.. but lets think about to rewritte the drivers with much more sharing code
12:50:15austriancodersound good
12:50:59*austriancoder thinks about a good solution
12:51:40B4gderI think we should have a struct that sets up the basic data for a specific lcd
12:51:58B4gderthen you set a pointer to that struct and call the low-level function
12:52:21austriancoderi thought about the same
12:52:50austriancoderlet me try to define a struct
12:52:56B4gderplease do
12:54:24B4gderwe're doing this kind of thing for the sound api
12:55:03B4gder"we" might be a bit too much said, but Linus is ;-)
13:01:43austriancoderincomplet struct:
13:01:47*ashridah puts on his 'groupie' t-shirt.
13:02:28B4gderseems like a fine start
13:02:55austriancoderyeah.. we can extand it later will recoding
13:03:03B4gderbut getstringsize doesn't have to be there
13:03:32B4gderbut I agree, we'll fix it as we go along
13:04:08austriancoderso i will start to recode it - thats ok for you?
13:04:18B4gderoh yes
13:05:50austriancoderwill add a generic_lcd.h/c - there will be the code, which will be shared
13:06:43B4gderbitmap_lcd perhaps
13:11:31austriancodersounds good
13:13:07amiconnB4gder: You *can't* use the same underlying code for the different lcd routines
13:13:09B4gderor even lcd-bitmap to follow the existing naming
13:13:19B4gdersure you can
13:13:26B4gderit just needs some adjusting
13:13:49amiconnNo, the code will be substantially different as soon as we support greyscale
13:14:18B4gdereven more reason to have a struct like this
13:14:39B4gderthe remote lcd is not greyscale, is it?
13:15:56amiconnAll routines that deal with more than single pixels will differ significantly for different lcd data formats
13:16:57 Quit HCl ("Lost terminal")
13:16:57LinusNso there will be several functions, and some can be shared
13:17:37B4gdereach lcd-specific struct sets up a common or a specific function
13:17:48 Join hcl [0] (
13:19:56*hcl smacks screen
13:19:58 Nick hcl is now known as HCl (
13:19:59*LinusN wanders off to pick up his new car
13:20:24amiconnB4gder: Next problem: It will be necessary that a higher-level function (which might be shared) needs to call a lcd-specific low-level function. How could that be handled?
13:20:53 Part LinusN
13:21:07amiconn(Btw: I do have ideas how to split higher and lower levels, similar to how the grayscale lib does it)
13:21:45B4gderthe higher level ones wouldn't be shared, they'd simply be wrappers that set the pointer and call the shared function
13:22:41amiconnThe basic idea is to have 2 kinds of low-level functions, one dealing with single pixels, the other dealing with blocks of pixels (block of pixel == all pixels that share bytes in the lcd buffer somehow)
13:23:43amiconnThere are 4 implementations for each group, one for each drawing mode (the lowest level really only consists of 2 functions), which are callable via a function pointer array (idex == draw mode)
13:24:41amiconnThe higher levels all use these functions, but at least fillrect() and bitmap() do also need lowlevel knowledge
13:25:05B4gderbut this doesn't contradict what I'm saying
13:25:12austriancoderatm it is looking like this:
13:26:07amiconnB4gder: Maybe, but implementing shared functions before changing the api (and the way of internal handling) which severely complicate things
13:26:42*amiconn really wants an iriver. now! :(
13:26:58B4gderso the question is then, how close is the new api?
13:26:59austriancoderso it is now the best point for the new graphics api
13:28:01amiconnB4gder: As soon as my mmc driver modifications are done (and probably Jörg's usb-before-mount patch), I'll start working on that for the recorder core graphics
13:29:02amiconnShouldn't take too long; most things are already "implemented" in my head...
13:30:07austriancoderi will work on the bitmap_lcd functions... and see how we can connect them
13:34:37amiconnThe remote lcd bitmap drawing will probably be very similar to the archos internal lcd
13:35:17amiconnIn fact, all routines will
13:35:27amiconnThe h1xx main lcd will be different
13:36:17 Join preglow [0] (
13:36:19*preglow enters
13:36:28austriancoderlooks very nice:
13:37:26amiconnThat's one of the worst gfx routines ever
13:37:40amiconn...doing the invert pixel by pixel.
13:37:50amiconnI know that this is the current core implementation
13:38:33austriancoderi only take the current used core functions, and try to share some code over the drivers
13:38:52austriancoderas the new api is not ready, i cant do anything other
13:40:22amiconnYes... but this way, implementing the new api will get significantly harder (imho)
13:41:54austriancoderand why this?
13:42:03austriancoderwe will later also share some code over the drivers?!
13:44:34austriancoderi could do some optimations with using the framebuffer
13:45:01austriancodernet to set every pixel to a value.. instead set a hole block of the framebuffer to a value
13:46:53austriancoderamiconn: when will you start with the new graphics api.. in 2 weeks or a month?
13:50:35austriancoderamiconn: huhu +g+
13:58:37austriancoderi dont know what i should do now :(
14:10:22 Join Shagnar [0] (
14:12:02elinenbehas anyone here had the problem where the joystick button acts as if it is always pressed down (on their iriver)?
14:14:46ashridahhrm. no, i don't think i have. although i suspect my joystick isn't centering perfectly anymore, so it's aligned slightly to one side.
14:14:52ashridahit isn't triggering the button tho
14:15:16ashridahelinenbe: crack it open and see if the cups/springs/whatever are jammed/broken?
14:15:19amiconnaustriancoder: As I said, I'll start as soon as I find the time. Whether that will be in 1 or 4 weeks I can't say (yet)
14:17:19MoosCamaroelinenbe: me, joystick left side is very sensitif
14:17:25austriancoderamiconn: ok.. so i dont will change anything at the driver - no sharing functions
14:21:02 Join andY`fRa [0] (
14:21:34 Quit andY`fRa (Client Quit)
14:21:38 Join andY`fRa [0] (
14:21:38 Join Patr3ck_ [0] (
14:24:07***Saving seen data "./dancer.seen"
14:24:56 Quit austriancoder ("using sirc version 2.211+KSIRC/1.3.12")
14:25:28 Quit lostlogic ("Going to the moon")
14:39:20 Quit Patr3ck (Read error: 110 (Connection timed out))
15:02:22 Join lolo-laptop [0] (~lostlogic@
15:02:49 Quit MoosCamaro ()
15:04:01 Join MoosCamaro [0] (
15:15:25 Quit webguest54 ("CGI:IRC")
15:23:20 Join Sucka [0] (
15:45:25 Join DangerousDan [0] (
15:48:08 Quit gromit` (Read error: 54 (Connection reset by peer))
15:48:09 Join gromit`` [0] (~gromit`
16:00:41 Quit B4gder ("CGI:IRC")
16:01:26 Quit Sucka ("a bird in the bush is worth two in your house")
16:02:35 Join Sucka [0] (
16:06:24CoCoLUSi encountered some strange bug today
16:06:54CoCoLUSi planed on moving a few files from my computer to the iriver, and windows (when connected to rockbox) reported only 1 mb free space
16:07:06CoCoLUSafter booting the original firmware, windows displayed the correct 444 mb
16:07:08CoCoLUSany clues?
16:07:28 Join t0mas [0] (
16:17:59ashridahbroken FAT?
16:18:12ashridahwait. which version of windows?
16:18:20ashridahcould system restore have been enabled on the drive?
16:18:21CoCoLUSnever had that problem before with rockbox
16:18:26ashridahbecause that'll do fun things like that.
16:19:01CoCoLUSi have disabled system restore systemwide
16:19:01ashridahnote that rockbox has nothing to do with windows using the iriver as a drive, since the usb<->hdd bridge chip inside the iriver does everything, you just turn it on and go to sleep.
16:19:30CoCoLUSwell i can only describe what i've experienced
16:20:28ashridahwell, true, but as i say, rockbox has no interaction with the drive when it's plugged into a pc via a usb cable.
16:23:04MoosCamaroHCl: you're here?
16:24:10***Saving seen data "./dancer.seen"
16:32:26preglowapparently not
16:32:45MoosCamarohi preglow
16:34:19MoosCamarogood comit yesterday
16:34:28preglowoh, the patcher
16:37:53CoCoLUSi'm so sick of europe getting second class treatment compared to the us
16:38:15CoCoLUSthe sony psp... released date 1/9 !
16:38:28CoCoLUSand price: 249 € compared to 249 $
16:39:28preglowdifferent markets, different prices
16:39:54amiconnHCl ?
16:46:13 Quit Sucka (Read error: 104 (Connection reset by peer))
16:58:28 Quit ashridah ("Leaving")
17:00:14 Join Renko [0] (
17:01:15 Join tedboer [0] (
17:03:19tedboeri was just wondering, could portable players be used (iriver in my case) for realtime effects? in other words, is sound i/o fullduplex? if possible, what kind of latency would that have? i'd say processing power is more than sufficient for some nice effects...
17:03:30HClamiconn: sup?
17:03:35HClMoosCamaro: mm?
17:04:02amiconnHCl: Item #5 on your database todo should be easy. I'll look at it
17:04:42amiconnThere's even a small bug. The db browser should close the database file before rockbox enters usb mode. It doesn't
17:06:01HCltedboer: yes, it theory, it could.
17:07:02preglowtedboer: i'll be working on this
17:07:04t0mastedboer: you want echo and things like that?
17:07:16t0maspreglow: cool, have a plan?
17:07:17preglowtedboer: i'm planning a vocoder among other things
17:07:48preglowno plan, no, i don't even know if the audio chip supports full duplex operation, but i sure hope so
17:08:00t0masiriver does...
17:08:11t0masyou can listen to something you record in original firmware
17:08:26preglowt0mas: that could be a passthrought switch in the audio chip
17:08:37t0mashm... the volume control works ;)
17:08:49preglowthe volume is adjusted in the audio chip
17:08:57t0mashm... then I don't know
17:09:01preglownor do i
17:09:13preglowand i won't have time to work seriously on rockbox for several months
17:09:32preglowat least, i seriously doubt so
17:10:12t0masthe UDA1380 should support full duplex for the effect thing right?
17:11:35preglowtedboer: as for latency, i guess there is very little of it in the hardware, at least
17:13:14 Join Sucka [0] (
17:13:49t0masnever knew the UDA1380 supported 100 khz
17:14:21preglowit may do, but we're severly limited in what sfreqs we can use
17:14:25preglowno 48khz, for instance
17:15:34t0maswhat's the limiting component there?
17:17:17t0masA feature: "Digital mixer for mixing ADC output signal and digital
17:17:17t0masserial input signal, if they run at the same sampling
17:19:24MoosCamaroHCl: i 've got a error message in boot: "unsuported db vesion 1"
17:20:05MoosCamaroHCl: i can't see my db with your new version
17:20:10t0masMoosCamaro: you have to build the database with the new script
17:20:39MoosCamarowith the new script, it don't build
17:20:50MoosCamaroi don't know why
17:20:51t0masgiving an error?
17:20:51 Quit Sucka ("a bird in the bush is worth two in your house")
17:21:08preglowt0mas: cpu frequency
17:21:25t0maspreglow: can't it get to 48khz on max speed?
17:21:38MoosCamaroHCl: i cameback v1 it build, but i can't see it
17:21:41preglowthe clocks for the different components have to match up
17:21:50t0masah ok
17:22:11preglowwe can do 22.05, 44.1, 88.1khz, etc
17:22:16preglowbut not 24khz, 48khz, 96khz
17:22:49t0mashm... weird...
17:22:52preglownot really
17:22:54preglowjust bad design
17:23:00t0mashasn't iriver invented some trick for it?
17:23:06preglowthey resample
17:23:08preglowto 44.1khz
17:23:17pregloweven optical out is 44.1khz for 48khz files
17:23:21t0masstupid design then yes...
17:23:33preglowbut yeah
17:23:38preglowi'm really hoping we can do full duplex
17:23:50t0masyeah, I'm searching for some info on that
17:23:59preglowmaking this little player to a little effect box is part of the reason i joined the project in the first place ;)
17:24:27HClMoosCamaro: what?
17:24:38MoosCamarohi HCl
17:24:39t0maspreglow: I speak dutch... and philips is dutch...
17:24:47t0masso if I can find some telephone number...
17:25:55MoosCamaroHCl: with new cvs update, error message in boot:"unsuported db version 1"
17:26:12 Join hile [0] (
17:26:35t0masMoosCamaro: you said you had a problem creating a database with the second script version?
17:26:59t0maswhat's the error message?
17:27:02MoosCamarocan't create db with v2
17:27:33t0masyes, but why can't you?
17:27:35MoosCamaro"unsuported database version 1"
17:27:48t0masno, that's the rockbox error
17:27:59t0masyou have to create the database from your pc
17:28:00preglowt0mas: someone probably has this information already
17:28:05t0maspreglow: ok
17:28:06preglowlike xshock or linus
17:28:31MoosCamaroi can't see my db with this current cvs version
17:28:44t0masMoosCamaro: no, you can't
17:28:48t0masyou have to rebuild a new db
17:28:57MoosCamaroi do it
17:29:19MoosCamarobut not building wiyh new script
17:29:35MoosCamaroi attempt again
17:30:50MoosCamaroin the dir .rock, necessary the 2 scripts or just v2?
17:32:01t0masjust 2
17:32:36MoosCamarodon't build
17:33:06MoosCamaroi use make db file but buld't
17:34:41MoosCamarodon't build
17:35:22MoosCamaroi use songdbv2 but don't build
17:36:06t0masdid you read the wikipage?
17:36:30MoosCamaroyesterday my big db sounds good
17:36:46MoosCamarobut with today cvs not
17:36:57t0masyes, the db format changed
17:37:08t0masso have you rebuild the db with the new script?
17:37:18t0mas(not the one from the wiki, the one from cvs)
17:37:25preglowi see nothing that indicates it doesn't support full duplex
17:37:55t0maspreglow: they weren't sure... but philips told me it should be possible to record and output at the same time
17:38:04t0masas the AD and DA convertors are separate
17:38:23MoosCamaroi used new script, but when i clik in db build, perl don't open
17:39:12MoosCamaroin my .rock i've got songdbv2 and perl files
17:39:20MoosCamarobut don't sounds good
17:39:32t0maswhat's the filename of songdbv2 ?
17:39:33MoosCamaroi don't understand
17:39:58t0masrename it to
17:40:00MoosCamaropl file yea
17:40:01t0masand try again
17:40:18preglowt0mas: you actually called them?
17:40:37t0masthey're dutch... so for me it isn't a problem to call them ;)
17:40:39preglowbut yeah, like you're saying
17:40:46preglowthey're separate
17:40:51t0masbut they didn't know for sure... it was a helpdesk
17:40:52preglowenough rockbox
17:40:54*preglow vanishes
17:40:59t0masI gave them my email address
17:41:08t0massome tech guy will look at the question and email me :)
17:41:31t0masMoosCamaro: is it working now?
17:41:58MoosCamaroit build
17:42:16MoosCamaroi wait for see if error msg in boot
17:42:36MoosCamaroi've got big db
17:49:28MoosCamarorocks good
17:49:44MoosCamaroi'm so stupid guy
17:52:12t0mashm... someone should update the plugin api docs sometime...
17:52:34preglowgo ahead :P
17:52:43*t0mas was more thingking of doxygen
17:53:01preglowdoxygen is ok, if done properly
17:53:22t0mashm... maybe it's a good idea to add doxygen comments to some of the rockbox files
17:53:33preglowbut sure, it's a decent enough idea, ask the big boys what they think of it
17:54:15t0masI guess linus is still reading logs :)
17:54:59*t0mas prods Zagor
17:55:05t0masas he's the only one not away...
17:55:14t0mashmm... "idle 24:37:05, signon: Mon Apr 25 08:49:27"
17:57:08preglowlinus is more than busy driving his new car around, bless him
17:58:08bobTHCwhich car he bought ?
17:58:25t0masI guess we will see a "Feltzing Iriver extra's company" selling carkits soon then ;)
17:58:50 Quit Patr3ck_ ()
17:59:08MoosCamarohi bobTHC
17:59:47MoosCamarola forme?
18:00:16bobTHCyes ;)
18:00:53MoosCamaroi smoke one for you
18:02:33bobTHChehe, i cant do the same for u, i'm at work but dont worry when quit i'll do
18:03:07MoosCamaroa ta santé
18:03:34bobTHCcheeeeeeers ;)
18:04:19 Join Sucka [0] (
18:05:32preglowhe got a volvo v70, i believe
18:05:49preglowand he did mention modding it to be able to interface with an mp3 player, heh
18:07:09bobTHC:))) i'm a psychic, i found the car manufacturer without your help
18:07:41preglowsomeone cart this man off to the research lab, he can read my mind
18:09:20*t0mas swallows a remark on french people :P
18:10:59bobTHCfor me volvo V70 is a good choice if he take the 2.5L R 300HP
18:11:38t0masthat's not really a cheap one ;)
18:12:14bobTHCgood things are seldom cheaps
18:13:28preglowany car would do for me about now
18:14:36bobTHCa fiat 500 is ok ;) ?
18:14:46preglowas long as i can drive it from a -> b
18:14:52*t0mas isn't even allowed to drive a car ;)
18:15:35bobTHCwhy ?
18:15:46t0masI'm 17 :)
18:15:59t0mashave to wait a few months to get my license
18:16:25bobTHCthat change the life
18:17:23preglowit changes your life if you get a car
18:17:26preglowwhich i don't have
18:17:32preglowsince i can't afford one :D
18:18:38MoosCamarohere in Paris, car is not good all the day
18:18:59preglowif you live in a big city, you really don't need a car
18:19:07bobTHCyes of course in bigcity is the opposite
18:19:40bobTHCbut when you want to go on the countryside of the world ;) hard to do without ;)
18:20:02t0masbobTHC: paris has an excellent subway system...
18:20:28MoosCamaroyes it's true
18:20:29t0masand those thalys trains are usefull too...
18:20:38bobTHCthe subway is good, buses are not really usefull
18:21:20MoosCamaroBob: which city you live in France?
18:21:51bobTHCmarseilles like says english
18:22:21MoosCamarobeautiful city
18:22:41bobTHCand here the subway is totally unusable
18:23:13bobTHC2 lines, a dozen of station and close @ 19h
18:23:25MoosCamaroi'am fan of the O.M club of football of Marseille
18:24:13***Saving seen data "./dancer.seen"
18:24:21bobTHCme too, necessarily
18:29:20preglowi'm a fan of abolishing football from the planet
18:29:32preglowon the grounds of being a boring sport
18:30:19*preglow sneaks out before being killed
18:31:12MoosCamaroit's culture question
18:31:27bobTHChere it's not a question of which sport, however which footballclub
18:32:06bobTHCin norway u have the mre spectacular sport ever : curling
18:32:17preglowno, football is what's popular here as well
18:32:22preglowbut i love curling :P
18:33:30MoosCamaroi'am passioned by all sports
18:33:50MoosCamaroand pratic tennis and foot
18:34:38 Quit Sucka ("a bird in the bush is worth two in your house")
18:34:43bobTHCand the best of all sport never invent is rockboxing ;)
18:35:06MoosCamarohehe :)
18:35:08preglowi can see that becoming a viable sport as soon as i code my vocoder
18:36:47bobTHC:) just vocoder or the total FX package with chorus, tremolo..... ? ;)
18:38:03preglowi can see it becoming a sport of seeing who has the most kickass effects on their player
18:38:20preglowa shouting match until one of the players wet their pants
18:41:20bobTHCwhy not a speech synthetis engine when u have the wet hands ?
18:43:00preglowmuch harder
18:43:03preglowbut i've thought of it
18:43:29bobTHC120Mhz ... dont forget
18:43:38preglowwith crap ram
18:45:46preglowmy, i love vococers
18:45:46bobTHCbut 32MB...
18:46:11bobTHCto make funny phone jokes ? ;)
18:46:41preglowthey just sound plain bad
18:46:48preglowespecially in electro music
18:47:04rasherplug it between mic and computer while playing $FPS_OF_THE_MONTH
18:47:11rasherand scare your teammates
18:47:52preglowor just play an audio book or something
18:48:25preglowsounds like a bloody robot is reading literary classics
18:49:04bobTHCwith the darth vader voice
18:49:38preglowno, darth vader isn't vocoder
18:49:43preglowit's just flanger or something
18:51:00bobTHCimho u can be near with vocoder too
18:51:21bobTHCbut it's not a typically vocoder sound
18:52:18bobTHCorange vocoder do great things for me this days ;)
18:52:34pregloworange vocoder is pretty neat
18:53:53bobTHCyes, prosoniq release good stuffs
18:55:45bobTHCVoxCiter is one of this ;)
18:57:16HClMoosCamaro: you need to redownload
18:57:29HClthe makedb.bat relies on, and is in cvs now
18:58:29MoosCamarore HCl
18:58:37MoosCamaroyes i see that
18:59:34MoosCamaromy db works fine now(thanx t0mas)
18:59:50HCli suggest you update the script from cvs though
19:00:00HClnot use the one thats in the database folder on my server
19:00:05preglowthere's proper vocoder in use
19:00:16preglowsorry about mono, i hate audacity
19:00:36MoosCamaroyes i download it
19:00:59MoosCamarobut i forget rename .pl file
19:02:31HClyes, i'm saying you shouldn't download
19:02:34HClbut the one from cvs.
19:02:38HClthen you don't have to rename it either.
19:03:25MoosCamarohave you got in your plan, the kind filter?
19:03:57MoosCamaroBob: filtrage par genre
19:04:12HClgenre's are already stored. eventually the info will be combined into a playlist generator
19:04:39MoosCamaroin the browser
19:05:00bobTHCso yes MoosCamaro u will can generate playlist by genre ;)
19:05:02*HCl was more thinking in the form of a plugin or as an option in the menu.
19:05:20MoosCamarook thanks
19:05:34*HCl annoys his flatmates by watching anime\
19:05:55bobTHChehe subfans ?
19:05:59HClyes, i swear.
19:06:03HClyellow subtitles are the worst.
19:06:48HClthey made the subtitle colors in the colors of the characters o.o
19:08:33MoosCamaroHCl: have you got any news about ˙ bugs?
19:08:49HClthey're caused by a bug in the id3 parser part of the
19:08:53HClits not high on my priority list
19:10:05HCli'm kind of wanting mp3 support before starting on the runtimedb..
19:10:06bobTHC˙ is a real strange letter, even linux have trouble to handle it...
19:10:36bobTHCat least ftpd ;)
19:11:38 Quit bobTHC ("Smoke Weed Every Day !")
19:14:16MoosCamaroHCl/t0mas: PSV Eindoven play today for European Football Cup
19:14:36t0masdon't even know if they're good...
19:14:57MoosCamarothey played good
19:14:59*HCl generally doesn't care much for sports, but it can be fun to watch sometimes
19:15:56MoosCamaroit' good for Netherlands image in europe
19:16:17HClfor people who think soccer is a big deal...
19:16:18HCli guess..
19:17:48 Join TheVoid [0] (
19:18:16TheVoidhardware troubles... anyone feeling adventurous?
19:18:26*HCl is a software person
19:18:44*TheVoid remembers HCl
19:18:55*HCl also needs someone who can explain what code gets ran when a song is played/loaded.
19:19:15*TheVoid is a dumb end-user
19:19:27preglowTheVoid: still no disk?
19:19:39TheVoidya. here's the whole story:
19:21:07thegeekhow much work is it to extend the db with support for more tags?
19:21:24HCldepends on what you mean with more tags
19:21:36thegeekI mean
19:21:37TheVoidthe recharge plugin on the player snapped off into the player the other night (my reaction: NOOOOO!!!)... so i took the player apart and soldered it back on. after that, i've been getting ATA -11 (you know... hard disk wont start up)... so i was reading around, and it turned out two of the parts i unsoldered are used to power the HDD... so feeling relieved, i soldered those two connections back together, and.. well... that wasnt it
19:21:50thegeekcustom tags
19:21:58thegeekI use some custom tags for my mp3's
19:22:11HClcustom tags? you have your own tagging format? o.o.
19:22:13thegeekDate added to collection
19:22:30thegeekfoobar2000 can set whatever tags you want
19:22:45HClwell, you'd have to find a perl reader for those kind of tags
19:22:55HClthe code in is fairly adaptable
19:23:02HClbut the code in rockbox
19:23:05HClreally isn't very nice
19:23:10thegeekI see
19:23:15HCli tried to convert it to more maintainable a bit
19:23:15thegeekoh well
19:23:17HClwith my v2 update
19:23:19preglowTheVoid: well, i have no experience with archos hardware, so can't help you
19:23:21thegeekgood work btw:)
19:23:35HClyou'd have to browse through the entire code for it
19:23:40HCland see what regions would be affected
19:24:16HClwe can't just add any tags to the database
19:24:21HClsince the database format is static
19:24:23thegeekI understand that
19:24:47HClmostly cause writing a parser for xml and stuff like that is too slow for archos
19:24:49thegeekwhat about some more standard tags?
19:24:54thegeekis it implemented?
19:25:10HClwell, adding not, but last time played should be implemented in the runtime db
19:25:17thegeekI meant
19:25:17HCldate added isn't in the database.
19:25:21thegeekas in
19:25:24thegeekalbum date
19:25:34thegeekoh well
19:25:37HClwe only support id3v1 / v2, i think
19:25:39thegeekit works great as it is now
19:25:42HCli never touched the tag reader bit
19:25:49thegeekI'm spoiled
19:25:50thegeekfoobar2000 ;)
19:25:57TheVoidpreglow: i posted the problem in the forums... we'll see what happens
19:27:02HCli think the is pretty portable
19:27:07HClyou'd just have to alter the tag reader bit
19:27:10HClto support more fileformats
19:27:18HCland export more tags to the database part
19:27:22HCldatabase building*
19:28:35HClas for the actual database suppporting certain tags
19:28:48HClthey'd have to be generic ones, not custom ones
19:28:54HClfor them to get added
19:29:04 Quit TheVoid ()
19:49:02 Quit hile ("jotain muuta (nelonen)")
20:03:10HClany rockbox core developers awake..
20:03:32HClif i were to add some functions to operate the runtimedatabase, where would i put them...
20:03:36HClrasher: how're things with the hash?
20:07:24preglowi'll refrain from making a bad joke
20:08:14HCli really need the stuff soon :P
20:08:22*HCl prods rasher
20:08:41preglowwhat kind of hash do you need?
20:08:46HCl :3
20:08:52HClthat runs fast on embedded devices
20:09:07preglowin that case i can't help
20:09:11preglowi've no idea what runs fast
20:09:19preglowi do know that md5 isn't very fast, heh
20:09:19HClrasher looked at it...
20:09:30HCli pretty much need a plugin that scans all files on the target
20:09:35HCland reports for duplicate hashes.
20:09:46preglowcrc32 isn't too bad
20:10:00*HCl hiccups
20:10:09HClfinally finished something.
20:10:11HCland now i'm bored
20:10:12HCl :?
20:10:12HCl :/
20:10:15preglowhold your breath
20:10:22HCli can't, for long
20:10:57preglowhalf a minute does the trick for me
20:11:07HClpretty much.
20:14:29HCli need to know where i can add functions that'll be pretty much global..
20:14:38HClbut first i need hashing to work.
20:14:41HClfor which i need rasher
20:14:47*HCl violently shakes rasher back and forth
20:14:54HClwake uppppp x.x
20:15:03preglowwhat kind of functions?
20:15:14HClruntime database stuff, and general database stuff.
20:15:24preglowprobably add some new file
20:15:37HClget tagdata for file, store runtime data for file
20:15:43HCladd file
20:15:45HClremove file
20:15:46HClrename file
20:15:59HClsynchronize runtimedatabase
20:18:20HClwhere though, apps?
20:18:33preglowi'd guess
20:24:17***Saving seen data "./dancer.seen"
20:29:10*HCl is gonna love the "rating" option of the new database ^.^
20:30:11 Quit crash_ (Remote closed the connection)
20:30:21 Join crash_ [0] (
20:30:37HClcan anyone tell me whether i'm able to 16bit align shorts
20:30:42HClor whether they'll have to be 32 bit aligned
20:30:46HClfor archos
20:31:12*HCl guesses we'll find out.
20:31:21HClit'll be less tedious to fix it for the runtime db anyways
20:31:51preglowi should think the types need to be aligned on a boundary equalling their sizse
20:32:20preglowbut no guarantee
20:32:22preglowcheck the manual
20:32:22HClthen i'll nicely put two shorts next to eachother
20:33:01amiconnpreglow: you're right
20:34:23*amiconn compiles with a reload-db-after-usb patch...
20:34:41HClyou might have to look at a check that i think i added to db_init..
20:35:25 Quit crash_ (Remote closed the connection)
20:35:31 Join crash_ [0] (
20:42:17*HCl works on structurizing the database code.
20:45:24 Join Elmoe_ [0] (
20:45:31*Elmoe_ pokes pabs
20:46:19t0masamiconn: maybe do it like rolo does? check if the file is modified?
20:47:53amiconnt0mas: The reload itself is really simple. I added a db_shutdown() function, which is called before rockbox hands over disk control to usb and does what the current code 'forgets' - close the db file.
20:48:14amiconnAfter returning from usb, I added a call to db_init()
20:48:18t0masah ok
20:48:23t0masthen checking it isn't needed
20:49:14amiconnOnly one thing isn't working properly yet - if the current browse mode is 'id3db', it should return to the db root as it does with the filesystem
20:50:12amiconnI need to find out where the dir filter modes are defined...
20:54:05amiconnHmm, that's not the problem...
20:55:35*t0mas is still searching for an opensource dictionary...
20:55:56CoCoLUS ?
20:56:07t0maswith word explaination...
20:56:56t0masas it's useless to have just a wordlist
20:57:43CoCoLUSwell define "open source"
20:57:52CoCoLUSfree to use?
20:58:37t0masen to download
20:58:39t0masen = and
20:58:49t0masI need wordlists...
20:59:16t0masand downloading should be free... better would be GPL-ed wordlists... but that isn't really needed
21:00:34HClthe moby project?
21:00:46HClisn't websters dictionary free?
21:01:27*HCl restructurizes the database stuff into database.h and database.c
21:01:45t0masisn't downloadable
21:04:05amiconnDB reload is working now....
21:06:33amiconnThere's still a small glitch
21:07:08t0masCoCoLUS: no suggestions for my search? :)
21:07:18CoCoLUSnope :)
21:07:32amiconnGoing 'valid db' ==> 'no db or invalid db' and vice versa via USB mode causes display probs
21:08:06amiconn...if id3db browse mode is enabled when entering usb mode
21:11:59HClit should go back to file mode
21:12:04HClif the id3 db becomes invalid
21:12:16Rob-gcide is a free dictionary.
21:12:43rasherHCl: haven't looked at it.. I just have the function to hash files.. no recursive lookup thingy
21:12:45amiconnHCl: It doesn't do that even if you enable id3db mode and there is no db in the first place
21:13:30t0masRob-: tnx
21:14:27HClamiconn: yea, now you mention it. o.o. it should.
21:14:55rasherthat'd confuse people
21:14:59rasherunless there's a message
21:15:17HClit gives a message o.o.
21:15:44HCli don't think we shouldn't add nice features because it would confuse dumb people :/
21:16:07rasherthat's not the issue
21:16:15rasherthe issue is non-obviousness
21:16:41rasherdoing something other than the expected should be accompanied by a message
21:16:47rasherexplaining why
21:17:27HClit is!
21:18:33 Join Stryke` [0] (~Chairman8@resnet-241-86.resnet.UMBC.EDU)
21:18:40rasherI just reacted to your second statement
21:18:50rasherwhich seems sortof arrogant to me
21:19:50amiconnHCl: The invalid->valid id3db transition within usb mode is handled now. I'll leave the issue with the other transition for now
21:19:53 Quit Stryke` (Client Quit)
21:27:00amiconnMy id3 database on the recorder is 1.2 MB
21:27:18HClnot too bad.
21:27:21HClmine is 5.8
21:27:28HClmostly due to the fact that i have an album with 715 songs.
21:27:35amiconnMonster tags, or way more tracks?
21:27:37HClwhich causes all other albums to have space for 715 songs
21:27:38amiconnAh, yes
21:28:22amiconnI'll build a db for my player as well
21:29:10 Join Tang_ [0] (
21:29:25amiconnThe db of my 2 GB MMC (~66% full) card is really tiny though - 42 KB
21:29:47HClhows mmc?
21:30:17amiconnStill unfinished, but I'll resume working on it in a few
21:31:09MoosCamarohi Tang
21:33:01HCldoes C support whatdoyou call it.. functions with the same names but different arguments?
21:34:36CoCoLUSoverloaded functions.
21:35:08CoCoLUSc++ does
21:35:20CoCoLUSi don't think c does, though, could be completely wrong of course :)
21:35:50amiconnbah, oo stuff
21:36:00CoCoLUSi wouldn't call it that
21:36:47preglowc does not support overloaded functions, no
21:37:12CoCoLUSnot that it's hard to do without them :)
21:37:19preglownah, not really
21:37:22preglowthey're great sometimes, though
21:37:22HClcan i assume pointers are 32bits?
21:37:31preglowdo you need to assume that?
21:37:38HClyes o.o.
21:37:50HClcause i have to skip a pointer in my structure o.o.
21:38:00preglowthen skip sizeof(pointer)
21:38:01amiconnsizeof is your friend
21:38:04CoCoLUSover-riding- functions are nice
21:38:05HClgood point.
21:38:36preglowCoCoLUS: all if it can be nice if used properly
21:38:53preglowand plain irritating if used improperly
21:39:17CoCoLUSof course, but isn't that statement true for everything in life? :)
21:39:56preglowindeed ;)
21:40:50preglowbut no, operator overloading isn't oo, heh
21:41:14preglowjava doesn't have them, and java is oo ;)
21:41:53Tang_Hi preglow
21:41:59Tang_i'm back at home
21:42:19Tang_now i can work on the fornoobs section translation
21:42:42 Join muesli- [0] (
21:42:52 Join muesli_- [0] (
21:43:30 Quit muesli- (Client Quit)
21:43:34Tang_hi muesli_-
21:43:43muesli_-high tang
21:43:59*HCl wonders why stricmp is called strcasecmp in rockbox.
21:44:31Tang_anyway i wonder
21:44:44Tang_if the two nooby section
21:44:49preglowi didn't know there was anything called stricmp
21:45:00Tang_(status and general) shouldnt be merged
21:45:01preglowstrcasecmp i know of
21:45:47HClyou're right, i guess its a windows thing maybe? o.o.
21:45:49preglowmy man pages do not know of stricmp either
21:46:12MoosCamarohi Tang
21:46:19Tang_hi Moos :)
21:46:34 Part HCl
21:46:37 Join HCl [0] (
21:47:14MoosCamaroTang: holidays are finishe?
21:47:25Tang_yes back to Brest
21:47:42Tang_still in holiday thought
21:48:16MoosCamaroscuse I come back
21:48:48Tang_so about the Wiki any idea?
21:54:26HClwait, i just implemented a binary search for filename, but i shouldn't till we figure out whether we need to search for hash or filename
21:54:27 Quit muesli_- (Read error: 104 (Connection reset by peer))
21:54:30*HCl prods rasher
21:58:34*HCl guesses we'll need to be able to search for filename either way
22:01:37 Join Biptoria [0] (
22:02:25 Quit Biptoria (Client Quit)
22:05:00 Quit courtc (Read error: 110 (Connection timed out))
22:05:22 Join courtc [0] (
22:08:53 Join uski [0] (
22:10:05 Quit amiconn (Read error: 145 (Connection timed out))
22:14:12 Quit Shagnar ("( :: NoNameScript 3.81 :: )")
22:16:52HClmeh, i should probably implement a non recursive binary search..
22:24:18***Saving seen data "./dancer.seen"
22:25:52MoosCamarore Tang
22:26:34MoosCamaroare you here?
22:36:04HCli need more developers to be around >.<
22:38:16BagderI don't see how you can avoid searching for the file name
22:38:37HClwell, we have to search for file name either way when deleting a file
22:38:45HClbut otherwise, it can possibly be faster to search for hash
22:38:49Bagderand when loading the db entries
22:39:05HClhashing the file, then searching for the hash
22:39:12HClit kind of depends on which'll be faster.
22:39:26Bagderno, it depends on when you'll load the rundb entries
22:39:39BagderI suppose you load N entries at once
22:39:45Bagderwhen the songs are loaded into ram
22:39:48HClwe don't need to load any entries at once for now.
22:39:56HClwe load them for a file when needed o.o
22:39:59Bagderso when do you load them?
22:40:06HClwhen a new file gets loaded
22:40:16Bagderthat kills batteries
22:40:33HClyou suggest that all fileentries will be in ram?
22:41:05Bagderall the entries for the loaded songs should be
22:41:12BagderI'd even say must be
22:41:13HClwhats "loaded songs" ?
22:41:23Bagderwe have some 28MB ram for songs
22:41:35Bagderit means we'll load perhaps 10 songs at once
22:41:46HClthen you still load them one at a time, but 10 times o.o
22:42:00Bagderthey'll all be loaded at once
22:42:09Bagderthen the disk will spin and power down
22:42:11HClexactly how do you search for 10 fileentries at once?
22:42:17Bagderuntil 9.5 songs have been played
22:42:35HClthen you still load them one by one.....
22:42:41HClyou can't do a search if you don't have all the data
22:42:46Bagderyes, but you need to load them in advance
22:43:02Bagderand hashing on the file contents then is a very slow and costly operation
22:43:12HCli dunno about that
22:43:23HClrasher was trying only hashing the first 1024bytes
22:43:27HCland the last 1024bytes
22:43:35HClwe're not gonna hash the entire file.
22:43:40Bagderthe last might not be in there
22:44:09HCli don't really care either way..
22:44:20HCli just don't know which is faster, hash or filename, and we'll need filename either way.
22:44:33HCli wrote a binary search for filenames, and a sequential search for hash
22:44:43HClwe're gonna need to be able to search for both
22:44:43 Join amiconn [0] (
22:44:49amiconnmerf :(
22:45:03HClyou stole my merf expression! :x
22:45:13amiconnWindows decided to crash, and now Firefox killed all bookmarks :(
22:45:16HCl :/
22:45:25amiconn_ALL_ bookmarks
22:46:40BagderHCl: when we load a song, we don't know how much of it we load
22:46:50Bagderwe could end up just loading a tiny bit of the beginning
22:47:16HCllike i said, i don't really care, its just a question of which is faster
22:47:27Bagderyes, and I'm providing info
22:50:45HClokay, so filename is faster?
22:50:58Bagderand probably the only possible way
22:51:27HClhow do i truncate a filedescriptor?
22:51:35HClsay i want to remove x bytes at the end
22:52:26HClwe have a filelength function right, what is it?
22:56:22preglowamiconn: how did firefox manage that?
22:56:33amiconnI dunno
22:56:44preglowamiconn: are you certain the bookmarks file is not lying around still?
22:56:45amiconnBookmarks are empty, and all settings reset to default
22:56:56preglowamiconn: close firefox and restart
22:57:02amiconnI did...
22:57:12amiconnbookmarks.html is 431 bytes now...
22:57:23preglowcertain you've found all of them?
22:57:47preglowfirefox sometimes made a new profile for no reason
22:58:44MoosCamaroTang: private me
22:59:07HClBagder: is there a big memoryspace on iriver that i'm able to use as a buffer when deleting files from the player from rockbox?
22:59:09t0masAnybody intrested in a dictionary on rockbox?
22:59:40HCltoo much work to type in a word :
22:59:41HCl :/
22:59:52t0masjust type first cahrs
22:59:53amiconnpreglow: There is only one profile dir
22:59:56t0maschars :)
23:00:01t0masand scroll throug list..
23:00:26BagderHCl: why would you need a buffer to delete files?
23:00:56 Quit andY`fRa (Read error: 104 (Connection reset by peer))
23:03:28HClBagder: when i delete a fileentry, i need to shift all of the entries above it.
23:03:51Bagderah, that
23:04:00Bagderno you won't have much space for that
23:04:06Bagderwe want working on archos too
23:04:11Bagderwant it
23:04:14*Bagder can't type
23:04:15HClthen i'll do it with my limited 1k buffer
23:04:49 Quit Renko ("Leaving")
23:11:16Bagderoh how I like rockbox more than iriver's firmware
23:13:21 Nick Lynx_ is now known as Lynx_awy (HydraIRC@
23:13:23preglowonce it starts playing things, i'll even start using it for stuff
23:13:42preglowdamn, that'll be fun
23:13:57BagderI'm using it on my Archos now. I used my iriver on my way home from work earlier
23:15:09*t0mas senses a mad cow... quick! close the borders!
23:15:33*preglow itches to finish libmad opts
23:16:31 Quit Elmoe_ (Read error: 110 (Connection timed out))
23:20:51amiconnBagder: More wiki spam :(
23:21:13*HCl checks whether things still compile...
23:21:14amiconnDid you use the ordinary way to delete (=move to Trash web) for your last deletion?
23:21:21HClamiconn: did you commit your database after usb thing?
23:21:25Bagderamiconn: no
23:21:41amiconnHCl: yup
23:22:24*HCl prepares to fix conflicts :P
23:22:37*HCl made quite some rearrangements in the database code
23:22:58HClmoving database generic things to database.h and database.c, and putting only dbtree specific stuff in dbtree
23:25:41 Quit Harpy (Read error: 60 (Operation timed out))
23:29:26 Join amiconn_ [0] (
23:31:04t0masyou're good at perl right?
23:31:36preglowbut be fast, i've got people coming over
23:31:40t0mascan you help me with some string magic?
23:31:51t0masok, I'll drop some things in your privmsg
23:33:03 Join Vicious_ [0] (~irc@3549f791554ca0f0.node.tor)
23:33:04HClif this works, i'll commit
23:33:22HCland then call it a day :)
23:35:25 Quit Tang_ (Read error: 131 (Connection reset by peer))
23:37:59 Quit amiconn (Read error: 60 (Operation timed out))
23:37:59 Nick amiconn_ is now known as amiconn (
23:38:13HClnow to watch whether anything breaks :X
23:44:32 Quit Vicious (Remote closed the connection)
23:46:02 Join _ferenczy [0] (
23:54:50HClthats not good.
23:56:10*HCl fixes, commits
23:56:24HClnow why didn't the iriver sim break o.o.
23:56:52HClnm.. only the win32 sims broke..
23:57:07HClthat reminds me, the LITTLE_ENDIAN define is broken
23:57:09HClin x11 builds
23:57:14HClwhich is why the x11 sims didn't break
23:57:19HClwhen they should've
23:57:22 Quit lolo-laptop ("Client exiting")
23:57:38Bagderwe should set our own defines at a single central place, imo
23:58:00HCla config.h seems like a good idea to me.
23:58:05HClgenerated by the configure script
23:58:30HClthats how most configure scripts do it
23:58:35Bagderconfigure knows
23:58:46Bagderor could know at least

Previous day | Next day