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

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

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

#rockbox log for 2005-07-01

00:01:45CassandraWhat's the problem?
00:01:45RoriI want to have next track displayed scrolling with artists as in 'artist - trackname' and scroll it if it's too long
00:02:40RoriI have this bit working I think 'Next. %s%?Ia<%Ia|%?d2<%d2|[\root]>>'
00:02:40Roriwhich shows the artist
00:02:41Roribut I am not sure what comes after it for '- trackname'
00:02:41Coldtoasthere's what I use
00:02:41Coldtoast%sNext: %?Ia< %Ia| unknown artist> -%?It< %It| %Fn>
00:02:46Rorilet me try
00:03:00CassandraYou need a %s on the end of that.
00:03:13Coldtoastyou have to close %s?
00:03:13CassandraOh, sorry, it's at the front.
00:03:19CassandraNope. :)
00:04:01 Join Stryke` [0] (
00:04:09Rorithat don't work
00:04:19Coldtoastwhat do you mean it doesn't work?
00:04:25Roriwth is the unknown artist bit anyhow?
00:04:26Coldtoastit's what I'm using and it works
00:04:49Coldtoastheh. if the artis field in the ID3 isn't filled, it says "Unknown artist"
00:04:58Coldtoastif it's filled, it displays it
00:05:14Roriwhere should the scroller be? at the end?
00:05:25Roripaste again properly please :)
00:05:33Coldtoastand it's followed by track name, if the field is filled in teh ID3 and if not, uses the filename
00:05:38Coldtoast%sNext: %?Ia< %Ia| unknown artist> -%?It< %It| %Fn>
00:05:56Coldtoastor this
00:06:02Coldtoast%sNext: %?Ia< %Ia| unknown artist> - %?It<%It|%Fn>
00:06:07Rorisomething is wrong with that
00:06:21RoriI can get the artist with my previous statement but not the track title
00:06:54RoriThis gives me the artists not 'unknown artist' Next. %s%?Ia<%Ia|%?d2<%d2|[\root]>>
00:07:01Coldtoastand mine just says "unknown artist"?
00:07:11Coldtoastyeah. yours isn't giving the artist name at all
00:07:18Coldtoastit's giving the %d2
00:07:31Coldtoastwait a bit til it gets toward the end of the song
00:07:35Coldtoastor change mine thusly:
00:08:03RoriI just want 'Next: artist - title' that scrolls if too long
00:08:14Coldtoast%sNext: %?Ia<%Ia|%?d2<%d2|[\root]>> - %?It<%It|%Fn>
00:08:16Coldtoasttry that
00:08:19Roriand grabs the filename if it is not tagged
00:08:56Coldtoastyou'll have to wait til it gets toward the end of the track tho to get next track by ID# or next track filename
00:09:11 Join ashridah [0] (
00:09:31Coldtoastjust for the first track tho. Once it goes to the next track, it should work. Just the first track it kinda sorts everything out it seems
00:09:46Bagdernot the first track
00:09:50Bagderthe first 2mb buffer
00:10:01Coldtoastthere you go then
00:10:01RoriColdtoast that only gives me artist
00:10:06Rorinot next track title
00:10:12Coldtoastwell, first 2mb buffer will be the first track you play tho :)
00:10:22Coldtoastare you reading what I'm typing?
00:10:29Coldtoastplay a track
00:10:41Rorino was trying to wps on the iriver
00:10:44Coldtoastjust let it play. Eventually it'll fill the info you want
00:10:45RoriI am
00:10:48Rorisays BT -
00:10:53Rorino title :)
00:11:14Coldtoastsounds to me like BT is what the ID3 has as track title
00:11:24Coldtoastall I can say is mine works perfectly on my player
00:11:36Coldtoastyou can even try mine out if you like
00:11:39Roriwell title is working above
00:11:49RoriTitle. %s%?it<%it|%?fn<%d1|[\root]>>
00:11:56Roriworking above
00:12:00Coldtoastthat's title for CURRENT track tho
00:12:07Coldtoastall that is for current track
00:12:13ColdtoastI thought you wanted next track?
00:12:15Roriworks for next track too
00:12:45ColdtoastTitle. %s%?It<%It|%?Fn<%d1|[\root]>>
00:12:50ColdtoastTHAT's for next track
00:12:58RoriI want 'artist - title' of next track in one line that scrolls for next track
00:13:03RoriI want both on the same line
00:13:06Roritogether :)
00:13:31Coldtoastwell, all I can say is what I posted earlier gives you exactly that
00:13:35Coldtoaston my player, at least
00:13:49Roriit's working now
00:13:55Roriwasn't before.
00:14:03Coldtoastit's not buggy
00:14:06Roriwhen I went to the next track
00:14:12Coldtoastyou just have to know how it works
00:14:45Roriit takes ages to kick in
00:14:55Coldtoastthe very first time the buffer is filled, for the first 2MB you'll get no next track info
00:15:19Coldtoastit's not a bug tho. it's just how it works
00:15:33Coldtoastit's only the first time you fill the buffer tho
00:15:45Coldtoastafter that it all works nicely
00:16:01RoriOK I just selected another track and it's still not showing next track correctly
00:16:12Rorieven after filling buffer
00:16:15Coldtoasthehe. well...
00:16:34Coldtoastdunno if it's been sorted but the WPS doesn't update right away in the latest bleeding edge builds :)
00:16:37Roriit worked briefly
00:16:56Rorirebooting it
00:17:03ColdtoastI'll try mine
00:17:54 Join rw_ [0] (
00:18:04 Quit Sucka ("a bird in the bush is worth two in your house")
00:19:20RoriI can't get it working
00:19:28 Quit stripwax_ (Read error: 110 (Connection timed out))
00:19:30Roriit works briefly then nothing
00:19:41 Nick rw_ is now known as rwlogix (
00:19:49Rorigonna grab a bleeding edge to check
00:23:11Coldtoastyep. mne works
00:23:11Coldtoastupdates as the track changes
00:23:11Coldtoastthere's just the first 2MB thing but that's normal
00:23:11 Join stripwax_ [0] (
00:23:11 Part stripwax_
00:23:11 Join stripwax_ [0] (
00:23:12stripwax_ello again.
00:23:28Roribeats me. will post elsewhere and ask for suggestions
00:23:37rwlogixhmm...i'm not quiet sure if this is the perfect chan for this, but, has anyone thought about connecting an iriver via tcp/ip? ,)
00:24:00Bagderrwlogix: using what connection?
00:24:02stripwax_rwlogix - hm, how and why??
00:24:07rwlogix.oO(yes, you had to build some hardware for this *g*)
00:24:08stripwax_;-) junx
00:24:17rwlogixstripwax_: why? ;)
00:24:21Bagderrwlogix: then write a tcp/ip stack for Rockbox and run for it
00:24:25Coldtoastyou CAN do that tho
00:24:33CassandraTCP over line out. What a fabulously hacky idea.
00:24:46crwlover optical out (and in), 1337
00:24:47Coldtoastconnect it to a different machine and map the iriver as a network drive. hehe
00:24:50Bagdermake a true serial and do ppp
00:24:54rwlogixcrwl: yeah ;)
00:25:04rwlogixBagder: the stack is not the problem ,)
00:25:06stripwax_coldtoast, mmm, nice. or just buy an NSLU2 ;-)
00:25:09thegeekyou could write a sounder driver for windows
00:25:10Bagderrwlogix: so what is?
00:25:17rwlogixBagder: the hardware...
00:25:30Coldtoaststripwax_: well, it IS connecting to the iriver via TCP/IP tho :)
00:25:32rwlogixi only wanted to collect "ideas" - you know
00:25:33thegeekas long as you had bitperfect output it would work
00:25:38Bagderrwlogix: getting a serial port is not that hard I believe
00:25:40CassandraYou could run a webserver on your Jukebox. How l33t!
00:26:08rwlogixBagder: is there a driver for the serial port by now?
00:26:18CassandraRemote WPS. :)
00:26:26Bagderafaik, there is no serial port without hw mods
00:26:29Bagderso no
00:26:29 Join n0bby_ [0] (
00:26:39Bagderbut easily done
00:27:09rwlogixCassandra: i thought about using it as kind of streaming-client ,)
00:27:14CassandraEven cooler, remote login to your JB using the UIsim. :)
00:27:49Bagder... when we get linux sound support for the sim, we should run it on ipod ;-)
00:28:30 Part Stryke`
00:28:32stripwax_;-) heh heh
00:28:40 Join Stryke` [0] (
00:30:33CassandraThen we could implement Internet filesharing.
00:30:40Cassandra(Now I really am getting silly.)
00:31:10stripwax_really not sure i see the point ....!
00:31:38CassandraIt's one of those things where there's very little point other than to get Slashdotted. :)
00:31:47Coldtoastwhy isn't there linuxfor iriver? there's linux for ipod :)
00:32:08CassandraNo-one could be arsed.
00:32:10Bagderbecause we see no point
00:32:11thegeekbecause using linux for it is just not ver smart
00:32:16stripwax_because no-one's done it yet?
00:32:24CassandraPersonally I don't want Linux on my MP3 player.
00:32:27stripwax_but it should be an easy-ish port.
00:32:27CassandraI want Rockbox on it.
00:32:30Coldtoastit's good enough for the ipod guys tho :)
00:32:37stripwax_yeah but what do they know :-)
00:32:38Bagdernot really
00:32:43thegeeklook at rockbox progress
00:32:48thegeekthen at ipod linux progress
00:32:51thegeekand true
00:32:55thegeekipod has wierd chips
00:32:57thegeekbut still
00:32:57Coldtoastheh. the ipod uses dual 90MHz processors doesn't it?
00:33:03thegeekit also has a much larger usergroup
00:33:13CassandraRockbox is light years ahead of the iPod linux player app.
00:33:24stripwax_i thought the different gen ipods had different hardware anwyay
00:33:31Bagderthey do
00:33:40CassandraGrid computing using MP3 players!
00:33:54Bagderbut they're all portalplayer and they seem rather similar, if you read the data sheets
00:33:57CassandraWe should port Folding@Home to Rockbox.
00:33:58rwlogixyeah, an iriver-cluster
00:34:02 Join muesli- [0] (
00:34:02Coldtoastis there ANY hack that allows the ipod to be used as a mass storage device for transferring and playing music the way the iriver has it?
00:34:16stripwax_??? you can do that anyway can't you?
00:34:23Coldtoastor would you have to wait for rockbox (if it ever happens)?
00:34:29ColdtoastHELL no stripwax_
00:34:35stripwax_wow, ipods suck
00:34:36BagderI don't hold my breath
00:34:38Coldtoastsure, you can use it as a mass storage device
00:34:47rwlogixstripwax_: that's why we are here, nor? ,)
00:34:49Coldtoastbut the DAP part is entirely separate it seems
00:35:30rwlogixs/nor/right/ .oO(what the hack?!)
00:35:38Coldtoastafaik you can't play music transferred via mass storage and it's not the easiest of tasks to access the music stransferred via itunes using mass storage
00:35:42rwlogixi should go to bed
00:35:43CassandraHmm. I still can't get over how much quieter my PC is since I ripped out the GeForce 4Ti and replaced it with a 6200.
00:35:51muesli-can we already count up to 4colors?
00:36:07rwlogixred green blue yellow
00:36:09rwlogixi can...
00:36:11stripwax_muesli who you asking
00:36:23muesli-in generell
00:36:36rwlogixmuesli-: -l ,)
00:36:37n0bby_i can count to 8 colors
00:36:38muesli-in general
00:36:48muesli-on rbx dude :D
00:36:52CassandraCan I have a ticket back to the universe where #rockbox was in English, please?
00:36:59n0bby_red orange yellow green blue indigo violet octarine
00:37:15stripwax_muesli - not yet.. assuming you're talking iriver?
00:37:21rwlogixCassandra: *bliing* here you are
00:37:34stripwax_here's a bug:
00:37:35n0bby_theres a patch, but the author of it didnt finish it and then died or something
00:37:36CassandraThat's better. :)
00:37:37muesli-stripwax_ of course..what else ;)
00:37:52stripwax_if "hard drive power off" is enabled, automatic bookmarks-on-stop don't work
00:37:55stripwax_(as far as i can tell~)
00:38:07stripwax_n0bby- :-) markun iirc
00:38:08CassandraA patch for what?
00:38:16stripwax_markun you about??
00:38:28stripwax_Cassandra- 4-color greyscale
00:38:34stripwax_markun - 4-color greyscale ;-)
00:38:37n0bby_you did the greyscale patch markun?
00:38:40markunAh, that patch :)
00:38:42CassandraOh, right.
00:38:47n0bby_FINISH IT YOU BASTARD :P
00:38:54stripwax_now, now., ..
00:38:56markunYes, but now ami will finish it I think.
00:38:57CassandraIt'd be totally invalid now amicon's revamped the greyscale lib.
00:39:30markunBut I got optical out working today!
00:39:40Coldtoasthey. crossfading seems to work better since the fast track skip was implemented
00:39:41rwlogixthis weekend i'm at "life aid", maybe i look into all that stuff afterwards
00:39:51stripwax_life aid? live8?
00:39:52n0bby_why dont you have CVS access, markun? :P
00:39:56CassandraGo you. Should we be expecting a patch soon?
00:39:57markunI have..
00:40:22n0bby_i dont see optical in the changelog
00:40:26rwlogixstripwax_: yes, i think that's it ,)
00:40:28Coldtoastneeds far less buffered for it to work by the look of it
00:40:40markunI'm getting rid off a small bug and then you can have your patch
00:41:02n0bby_patch? noooooooooo, COMMIT :P
00:41:20*Bagder starts fiddling with also to make sound
00:41:25*n0bby_ commits himself to cvs
00:41:27CassandraYou see this is why we don't have anonymous cvs commit access. :_
00:42:35CassandraMost of the work in changing Rockbox is actually making sure you haven't actually broken something hideously before you commit.
00:42:57Bagderand then some more after the commit ;-)
00:43:09CassandraWell, if you're me, yeah.
00:43:37markunNext project will be to send the track names over spdif..
00:43:55rwlogixyeah, that's sounds like it's very simple to maintain and extend rockbox, perfect!
00:44:53Bagderwell, some people say there's tiiiiny lack of docs
00:45:01CassandraRockbox is incredibly feature rich, multiplatform code. You want something easy to maintain, try true.
00:45:04rwlogixas usual ,)
00:45:10CassandraOr easier still, false.
00:45:46rwlogixwell...false is not that simple :)
00:45:48RoriI give up on this wps
00:45:57 Join n0bby__ [0] (
00:46:04RoriI can't get it to display squat for next track now
00:46:17Coldtoastone other thing abou tRockbox too, one thing that really sets it apart
00:46:31Coldtoastis it' also being blind-friendly
00:46:53CassandraThat doesn't actually add much to the code complexity though.
00:47:10Coldtoastno, but it's a very important aspect I think
00:47:28Coldtoastnot to you or I, obviously
00:47:40CassandraIt's certainly the thing that got me from "wow, that's cool" to being an active contributor.
00:47:43Coldtoastit's like installing wheelchair ramps in schools
00:48:15Cassandra(I'm partially sighted - while I don't need it for myself, it's quite important to me on a personal level, having been born blind.
00:48:41markunIs there a way to substract a patch from another?
00:48:45Coldtoastdefinitely a factor for you then :)
00:48:54ColdtoastI wear glasses tho. does that count? :)
00:48:54Bagdermarkun: patch -R reverts a patch
00:49:12CassandraI guess that'd depend on you.
00:49:17Coldtoastdon't suppose any of you know the show Pure Pwnage?
00:49:29rwlogixColdtoast: if you have less than -4.00 ,)
00:49:43ColdtoastI don't
00:49:52ColdtoastI'm +, not -
00:49:56 Quit muesli- (Read error: 148 (No route to host))
00:50:15Coldtoastcan't remember by how much tho
00:50:15rwlogixi don't know if i'm + or -
00:50:40Coldtoastone of my friends makes my glasses. He knows
00:50:55Coldtoastgot these Silhouettes I'm wearing for $50
00:51:06rwlogix<- left, 0.75; right, 1.00 ,)
00:51:23Coldtoastthey're something like $550 to buy usually
00:52:15rwlogixhmm...mine are not that expencive
00:52:23rwlogix-c +s
00:52:30markunfor anyone who want's to try out the optical out:
00:52:51markunthe patch also includes a patch to set the noise shaper, but that shouldn't matter.
00:52:59Coldtoast'm talking $550 AUD tho rwlogix
00:53:24 Join ep0ch [0] (~ep0ch@
00:53:45ep0chis it just me or is gapless broken?
00:54:32stripwax_depends. what kind of files are you playing and what version of rbx?
00:54:51ep0chall file types, even vorbis
00:55:07ep0chlet me do more investingation...
00:55:11rwlogixColdtoast: that doesen't much... ,)
00:55:58stripwax_markun - hm, what does the noise shaper do and should that be a separate patch?
00:56:13ep0chand vorbis playback is always at full cpu boost ;s
00:56:17stripwax_ep0ch - which rockbox version? changes were made recently because it was changing track too early......
00:57:19ep0chhangon let me investigate, would help a bit if seeking worked on Vorbis...
00:58:10ep0chplayback is like really dodgy though with the latest build...
00:58:18ep0chsometimes it will play sometimes it wont
00:59:05Coldtoastreally? damn!
00:59:29rwlogixgood night boys and girls
00:59:49rwlogixsweet dreams...
01:00:41CassandraAFAIK I'm actually the *only* girl working on Rockbox. I find this slightly depressing.
01:00:51stripwax_markun - oh that's in hardware? cool
01:00:58ep0chCassandra: welcome to IT
01:01:09ep0chit's a blessing for you :p
01:01:52stripwax_markun - any reason that can't just be committed ?
01:01:52CassandraIf I were trying to get a date, maybe. :)
01:02:08ep0chdamn you got me sussed
01:02:32markunstripwax_: The optical led eats power, so it should be an option.
01:02:35rwlogixCassandra: that's normal (when i think about linux development)
01:02:48stripwax_markun - i meant the noise shaper..
01:03:00CassandraIt's also been the case in pretty much every company I've worked for.
01:03:06 Quit n0bby_ (Read error: 110 (Connection timed out))
01:03:11markunstripwax_: I forgot to commit :)
01:03:24rwlogixCassandra: depends, if you work in the office....
01:03:42ep0chi just finished my last day at work today :)
01:03:53ep0chgood old redundancy pay out
01:03:56CassandraYeah, but I was the only *techie* female.
01:04:24rwlogixCassandra: in germany we have something called "girls day"
01:04:28stripwax_~30% of the developers I work with are female. but in previous company there were just 3 vs 80 blokes
01:04:49rwlogixwhere girls can get in touch with technical jobs
01:05:00ep0chquick question... C or PHP?
01:05:08rwlogixep0ch: PERL
01:05:30rwlogixwell...bash ;)
01:05:43stripwax_cassandra - lol!
01:05:46rwlogixif you want low-level with a unix-shell, use /dev/port ;)
01:05:58stripwax_i love that language
01:06:15ep0chok, looking for another career in IT, which skills do you think are up and coming?
01:06:29stripwax_ep0ch is it all about skills, or what you use them for?
01:06:34CassandraDatabase programmer is always a safe bet.
01:06:42ep0ch(sorry for being off topic)
01:06:56stripwax_ep0ch no, I mean, what do you *want* to do?
01:07:03CassandraAnd sysadmin, but I wouldn't wish that on anyone, and the pay is shite compared to programming anyway.
01:07:12ep0ch*good question!*
01:07:13rwlogixep0ch: if you want a job? say you know XML and mssql, thats enough
01:07:23ep0choh no it aint
01:07:26stripwax_rwlogix not where I work
01:07:30 Quit webguest91 ("CGI:IRC")
01:07:39ep0chbanks want soooo much more...
01:07:44rwlogixstripwax_: maybe, but thats seldom - IMHO
01:07:46stripwax_ep0ch - kerching
01:08:17stripwax_ep0ch - which bank?
01:08:20CassandraI want someone to pay me a programming salary to work on Rockbox. Any offers?
01:08:35stripwax_s/all/any ? ;-)
01:08:59rwlogixstripwax_: $BOSS always wants to hear buzzwords
01:09:02CassandraYou don't work for Misys, do you epoch?
01:09:29stripwax_rwlogix none of my bosses have ever cared for buzzwords. I once had a manager who only cared about buzzword bullshit and I left because of it
01:10:38stripwax_now colleagues, that's different. i can buzzword till the cows come home with the people i work with.
01:10:58rwlogixstripwax_: at the moment i would do nearly any job - even if i had to *choke* M$
01:11:47stripwax_rwlogix one of the top employees apparently. and they have a kick ass research centre. not sure i'd want to work for them but there's much worse out there
01:12:00*rwlogix misses words at the moment ,)
01:12:09stripwax_rwlogix - for example you wouldn't want to work for either of my former employers..
01:13:35thegeekI would love to work for ms actually
01:13:45thegeekthey are not the big "evil" as some think they are
01:13:49thegeekjust another compnay
01:14:05thegeekyou could easily find much worse people to work for
01:14:13thegeekand it seems like ms devs really have it quite nice
01:15:39rwlogixmy dream would be a small company where i would have very different things to do
01:15:55stripwax_rwlogix - google's pretty small, relatively
01:16:05rwlogixi don't want to sit in my chair and prog C all day
01:16:19stripwax_rwlogix - why small company?
01:17:09 Quit n0bby__ (Read error: 110 (Connection timed out))
01:17:11rwlogixstripwax_: coz there you have job variation
01:17:31rwlogixi don't think you could have this the same way in a big company
01:17:34stripwax_rwlogix i have MASSIVE job variation and there are over 20,000 employees...
01:17:52rwlogixhmm...okay - but we noticed before that you're luky ,)
01:18:18rwlogix.oO(is "hmm" a german phrase?)
01:19:04rwlogixstripwax_: what do you do? ,)
01:19:22stripwax_rwlogix - work at a bank
01:20:05rwlogixoh, i don't wanted it that explicit ;)
01:21:04 Quit ghostiger2 (Remote closed the connection)
01:21:38stripwax_;-) application development, support, training, financial model, optimisation, troubleshooting, ...
01:21:44ep0chcya, slasheri: why is vorbis playback stuck at full boost?
01:21:50 Part ep0ch
01:22:33 Join kergoth [0] (
01:23:10stripwax_[getting a bit OT here right now]
01:23:10rwlogixstripwax_: sound great ,)
01:23:48rwlogixwell, i have to get up early - good night
01:24:21stripwax_me too
01:24:22 Part stripwax_
01:24:33 Part rwlogix ("*zonk*")
01:24:48 Part kergoth ("Leaving")
01:31:56RoriI love TPB
01:33:50 Quit markun ()
01:35:21MoosBagder if you here wiki spam :(
01:35:38Plugh_half the fun of tpb is downloading unreleased stuff and other items that otherwise cost more than they're worth
01:35:45Plugh_the other half is legal.php
01:46:58***Saving seen data "./dancer.seen"
01:48:08 Quit gromit` (Read error: 110 (Connection timed out))
01:54:10 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
01:59:26 Join ansivirus [0] (
02:01:39 Join gromit` [0] (~gromit`
02:10:11 Join courtc [0] (
02:15:30 Join ShockerEngr [0] (
02:18:17 Join amiconn_ [0] (
02:20:52CassandraHeya ami. Any idea why the new fonts aren't being built by the autobuider despite me putting them in the repository.
02:24:15 Part Moos
02:25:53CassandraOh, that commit didn't trigger a rebuild. Weird.
02:34:00 Quit amiconn (Read error: 110 (Connection timed out))
02:34:00 Nick amiconn_ is now known as amiconn (
02:40:13 Join jwalk [0] (~jwalk_edm@
02:40:42jwalkamiconn: want an adapter cable for your remote?
02:52:09 Quit jwalk (Read error: 104 (Connection reset by peer))
02:55:45 Join kenshin [0] (
03:05:01 Quit DomZ ("CGI:IRC")
03:06:47 Join elinenbe [0] (
03:08:32 Quit elinenbe_ (Read error: 131 (Connection reset by peer))
03:08:32 Nick elinenbe is now known as elinenbe_ (
03:13:52 Quit cYmen ("zZz")
03:20:41 Quit hardeep ("[BX] PARTYTIME! 5 seconds to the millenium! 4...3...2...1...EOF From client")
03:22:34 Join TCK [0] (
03:47:00***Saving seen data "./dancer.seen"
03:58:51 Join tiegs [0] (
04:01:59 Quit tiegs (Client Quit)
04:02:22 Join tiegs [0] (
04:05:19 Join QT_ [0] (as@area51.users.madwifi)
04:07:50tiegsanyone here?
04:09:00 Quit tiegs ("CGI:IRC (EOF)")
04:09:18 Quit QT (Read error: 148 (No route to host))
04:16:17 Quit kenshin (Read error: 110 (Connection timed out))
04:17:01 Quit ShockerEngr (Read error: 110 (Connection timed out))
04:25:19 Quit ashridah ("Leaving")
04:47:01 Quit dwihno ("fixa")
04:48:16 Join dwihno [0] (~dw@
05:05:04 Join kenshin [0] (
05:07:27 Quit pike (Read error: 54 (Connection reset by peer))
05:08:06 Join MO-Pantsu [0] (
05:09:15 Quit dwihno ("test")
05:10:28 Join dwihno [0] (~dw@
05:12:10 Join pike [0] (
05:25:24 Quit Rori (Read error: 110 (Connection timed out))
05:25:37 Join tiegs [0] (
05:42:54 Quit tiegs ("CGI:IRC (EOF)")
05:43:00 Join tiegs [0] (
05:43:47 Quit tiegs (Client Quit)
05:43:59 Quit RotAtoR ()
05:47:03***Saving seen data "./dancer.seen"
06:00:26 Quit kenshin (Read error: 110 (Connection timed out))
06:08:36 Join tiegs [0] (
06:14:43 Quit tiegs ("CGI:IRC (EOF)")
06:14:44 Join heidi [0] (
06:19:26 Join kenshin [0] (
06:20:12kenshinI'm pretty sure my setup is jacked. Specifically, I'm trying to name some files in Swedish but the names only look right to KDE.
06:20:26kenshinNothing else displays them right.
06:40:47 Join cs1 [0] (
06:42:30cs1something i just noticed with my H120... it boots rockbox just fine, but recently (like within the last couple of days) i've lost the ability to boot the original firmware.
06:42:57cs1yes, i know about either starting with "play" on the remote, or holding "record" on the main unit
06:46:37cs1i'm running H120 firmware v1.63US, patched with rockboot v2 and the 20050701 cvs snapshot
06:47:48cs1booting with either of the magic keys gets me as far as "Starting original firmware"
06:47:56cs1at which point it just hangs
06:48:59cs1i can reproduce this with at least the last 3 cvs snapshots (haven't been reflashing my firmware each time)
06:49:39 Join LinusN [0] (
06:51:37cs1as it has been for the last 3 minutes or so
06:56:33cs1i declare my iriver officially unable to boot the original firmware
06:56:49 Quit kenshin (Read error: 110 (Connection timed out))
07:00:59LinusNcs1: how's your battery?
07:01:37cs1pretty good... i'm plugged in to the wall right now
07:01:52LinusNhave you performed a scandisk?
07:02:09LinusNthe iriver fw is pretty picky about the filesystem status
07:02:29cs1i suppose i could do that. rockbox boots just fine, this isn't even giving me the iriver splash screen
07:03:39LinusNi'd say this situation would have been dangerous if you didn't have the v2 bootloader
07:04:08LinusNbecause if you weren't able to boot rockbox either, you would have a brick
07:04:16cs1it appears my battery was really low
07:04:27cs1rockbox was still running and booting and playing...
07:04:31LinusNclassic faq item
07:04:41cs1but i had to charge for quite a while
07:05:00cs1impressive just how much of a pig the original firmware is
07:05:35LinusNyou only think that now when you have seen the light :-)
07:06:34*cs1 edits the iriver faq right now
07:07:36cs1nevermind. i see it
07:09:21 Join ashridah [0] (
07:09:33 Quit courtc (Read error: 110 (Connection timed out))
07:09:52 Join courtc [0] (
07:27:50 Part cs1
07:47:06***Saving seen data "./dancer.seen"
07:52:20 Join Ancelot [0] (
07:52:51AncelotAnybody alive?
07:53:02LinusNmy heart is still beating
07:53:27Ancelottell me the truth, i dont wanna hate ya
07:53:37LinusNok, it's not beating then
07:53:41 Join Noel_sad_song [0] (
07:54:12Ancelotif i put the rockbox in my h340 right now, Will it work, or It'll become in a nice pocket warmer?
07:54:28LinusNnothing will happen at all
07:54:34LinusNsince there is no boot loader
07:54:38AncelotNot even porn?
07:54:46AncelotShame on you!!
07:55:17AncelotThough is good to know.
07:55:19*LinusN feels really ashamed
07:55:32AncelotI was thinking
07:55:35Ancelotyou should ask 1 dollar
07:55:42Ancelotat each misticriver member
07:56:12LinusN...if i had a penny each time... :-)
07:56:36AncelotIs kinda easy.. since you guys are doing this for a hobby
07:56:53LinusNwe have a paypal account for donations
07:57:01Ancelotya i know
07:57:07Ancelotbut donations are not obligations
07:57:13Ancelotfree will is a bitch
07:57:23LinusNof course
07:57:46Ancelotwasn't easier to crack each firmware and mod it?
07:58:15LinusNit would be 1) harder 2) more boring 3) illegal
07:59:18Ancelot1) Indeed 2) No personal satisfaction 3) Illegal? You would not be even asking money in exchange..
07:59:43LinusNredistributing the modded firmware would be illegal
07:59:53LinusNsince they have the copyright
08:00:16LinusNbut yes, we could distribute it as a patch to get around it
08:00:22LinusNlike we do today for the boot loader
08:00:46AncelotYa, rockbox is not quite a "firmware", but an addon, right?
08:01:04LinusNrockbox is 100% firmware
08:01:28LinusNnot a single byte is borrowed iriver
08:01:31 Join B4gder [0] (
08:01:36Ancelotno no no
08:01:39Ancelotthat is not what i meant
08:02:07LinusNB4gder: krorink
08:02:31B4gderI see another one without enough coffee :-)
08:02:38Ancelotsince you came up with an entire firmware, why having the original installed?
08:02:40Noel_sad_songsorry I interrupt.. LinusN.. yesturday I put the 29th build on my player and rockboy stoped working... did I do anything wrong?
08:03:06B4gderNoel_sad_song: upgrade again and it'll pop back alive
08:03:07LinusNAncelot: because people may want to run the original until rockbox is fully functional
08:03:37Noel_sad_songyou mean with todays biuld?
08:03:59LinusNNoel_sad_song: yes
08:04:34AncelotStill rockbox can't play doom3
08:04:36Noel_sad_songok, thanks a lot.... oh.. by the way.. great job.. I mean it!!
08:04:46Ancelotnor porn
08:04:56Ancelotshame on you twice
08:05:47LinusNwait for PornLoader v3
08:06:14AncelotI guess you will need p2p support
08:08:46AncelotI guess it was really frustrating not to be able to do things with your players cause of the firmware
08:09:02Ancelotthat's pretty much the idea of rockbox, am i right?
08:09:12B4gderand the fun of it
08:09:54B4gderyou're the first to say that
08:11:58Ancelot ~~~~~~:p
08:12:16B4gderthen again, my nick is not badger
08:12:40*Ancelot runs away in fear
08:15:19Noel_sad_songhey, what's wrong with wma codecs? is it a legas issue?
08:15:35B4gderits a closed format
08:15:52B4gderthat is a huge wrong
08:16:43 Join StrathAFK [0] (
08:16:51AncelotAnother of those great ideas brought to you by Gill Bates
08:18:23B4gderthe idea is not bad for ms
08:18:27B4gderit is just bad for users
08:19:08Noel_sad_songf*** i have 4k wma files...
08:19:17B4gderyou are doomed
08:19:26Ancelotwanna hear something funny¿?
08:19:46Ancelot3/4 of HIS files, are now in my h340... -.-
08:20:23Ancelotw/e rename *.wma *.mp3
08:21:26Noel_sad_songI heard ogg files use up the battery faster
08:21:30Noel_sad_songor do they?
08:21:44 Join ghostiger [0] (~ghostiger@bafb3613187f9248.session.tor)
08:23:23Ancelotyup 3851 wma files
08:26:35LinusNNoel_sad_song: high quality ogg requires more cpu power to decode, thus it eats up more battery
08:27:14AncelotWe do not recommend using the UK Firmware for the U.S. models. unless you have no need to purchase files online and use the DRM support. When you state that the UK players are "inferior" to the US players, we hope you understand that this is simply a personal opinion on your behalf. Millions of customers who use the DRM support would beg to differ.
08:27:45AncelotB4gder: Do u know of any good software to convert wma > mp3?
08:27:55B4gderno idea
08:28:00Noel_sad_songi see.. so what woud suggest? mp3?
08:28:02B4gderI've never even seen a wma file
08:28:15Noel_sad_songI use Easy CD/DA audio converter
08:28:37LinusNNoel_sad_song: i suggest you use the format you feel is best
08:28:44Noel_sad_songsorry bout the extra "audio"
08:28:45LinusNif you like ogg, use it
08:29:16AncelotLinusN still, if you want to use rockbox.. wma is out of the question
08:29:34LinusNAncelot: yup
08:29:59B4gderunless you make it happen of course
08:30:02 Join webguest36 [0] (
08:30:08B4gderthere is code
08:30:10LinusNthere is a free wma implementation, but it needs a lot of rewriting to work with rockbox
08:30:46LinusNso wma isn't impossible, it just requires a lot of work
08:30:50Anceloti could try it out.. but then again, it's easier for me to build a house..
08:30:53Noel_sad_songand you don't feel like writing right? :P
08:31:03LinusNlet me put it like this
08:31:20*Ancelot hides...
08:31:37LinusNto rewrite the free wma codec, you need to have a clue
08:31:56LinusNand if you do have a clue, you wouldn't be using wma in the first place
08:32:32Bgervery well said ;)
08:32:35Bgermorning btw
08:32:48AncelotLate night here -.-
08:33:04LinusNso the problem is finding a wma user with a clue
08:33:43LinusNthe same actually applies to most windows specific programming in this project
08:34:27 Quit Strath (Read error: 110 (Connection timed out))
08:34:57Noel_sad_songdo you guys have an idea why iriver ppl don't put playlist otg on their firmwares?
08:35:15B4gderthere's no money for them for doing that
08:35:29B4gderthey probably left these firmwares a long time ago
08:35:34B4gderthey've moved on to the new players
08:35:47AncelotBig mistake
08:36:00B4gderall manufactorers do like this
08:36:19B4gderthey sell hw and forget about sw upgrades as quick as possible
08:36:32AncelotI think
08:36:36B4gderthey rather have customers buy the new gadget instead
08:36:41Ancelotthe least they could do is open source
08:36:50B4gderthey want to sell the new one
08:37:03B4gderwhich people don't do if they can improve their sw inifintely
08:37:09AncelotThink about this for a second
08:37:23B4gderI've thought about this for many years
08:37:37B4gderand mp3/DAP is not special
08:37:38AncelotWould you buy a new DAP from a company whose last player had a crappy soft?
08:37:46B4gdermost consumer electronics work like this
08:37:50Noel_sad_songyup i understand.. but.. knowing nothing about marketing and sales techniques.. I can say I'll think it twice next time I buy a DAP
08:37:58B4gderAncelot: the market proves people do that
08:38:08B4gdergood SW doesn't sell
08:38:14B4gderfancy HW specs do
08:38:55B4gderthere must be a reason why all mp3 player firmwares suck
08:39:02LinusNthe misticriver community seems to be a living proof
08:39:17Noel_sad_songwhere are you from? would you hate me if I told northamericans can't stop buying crap?
08:39:18LinusNlots of them have "moved on" to the newer models
08:39:37Bgeranyway, i bought my player becouse of the opportunity some day to run rockbox ..
08:39:39B4gderyes, "I hope they fix it this time"
08:39:41LinusNi think that applies to the entire world
08:40:05B4gderit is not an american sympthom as I see it
08:40:14LinusNi believe rockbox has been an eye opener for many
08:40:27AncelotYou are the mesiah
08:40:31Noel_sad_songprobably.. actually Ancelot got his h340 cos of the colour screen :P
08:40:38LinusNbut the vast majority of dap users don't know about rockbox, and don't care
08:40:52Ancelotand why would they
08:40:55Ancelotif they have IPOD
08:41:03*Bger doesn't see any reason to buy other DAP after rockbox running on H3x0 and prays his "toy" to have long life...
08:41:23Ancelotif your toy dies, buy viagra
08:41:26LinusNi have had my archos recorder for years
08:42:08LinusNand it still plays music just as well as it did when i bought it
08:42:11dwihnowee for archos \o/
08:42:15LinusNno need to buy a new one
08:42:45B4gderyou guys read about the swedish "cassette fee" for portaple music players?
08:42:49AncelotUnless you really need a featuer (and let's get real, nobody needs none of this new stuff, just to show off)
08:42:58Bgerthe Hxxx irivers have one important advantage - all is done in sw
08:43:05Noel_sad_songyeh i see your point
08:43:29dwihnoB4gder: the what-what?
08:43:40B4gderdwihno: 0.40 öre/MB
08:43:48Noel_sad_songhaven't heard of it
08:43:53BgerAncelot there are many people buying something because it's new, fancy, etc...
08:44:17B4gdergonna add 240 SEK to a 60GB player
08:44:18dwihnoB4gder: one of the new laws of today?
08:44:57Bgerthis is ... <no comment>
08:45:00dwihnohow f*cked ain't that :(
08:45:04Ancelotya, i could use more disk space, but still is not a big deal. If I had less, I'd still be happy
08:45:09B4gderits so silly it hurts
08:45:25Ancelot240 SEK?
08:45:31 Quit pike (Read error: 110 (Connection timed out))
08:45:46Noel_sad_songwhat's SEK??
08:45:51B4gderswedish currency
08:46:25Ancelotu talkin bout taxes
08:46:37B4gderyes, a new grand tax in sweden
08:46:48Slasheriwe have had that casette fee a long time in finland
08:46:59Ancelotonly for players?
08:47:00B4gderSlasheri: on portables?
08:47:17dwihnothe swedish government will lagstifta sönder everything. the internet filtration will expand amd whooppee, farewell free speech.
08:47:18Ancelotquiito :O
08:47:26Slasheriand it's very frustating, i am buying for example all dvd-r media from foreign countries
08:47:34SlasheriB4gder: yep, portables and set-top-boxes
08:47:48AncelotI see your point guys..
08:47:50B4gderwell, now we have it too
08:48:07Ancelotand Slasheri.. I had 2 buy my player straight off from singapore...
08:48:23B4gderwhen I backup my digitial images, I'm sponsoring the record industry!
08:48:26Ancelotjust because here it costs around 800 bucks
08:48:42 Quit Seed (Read error: 54 (Connection reset by peer))
08:48:50*LinusN bought up a nice cdr/dvdr stash in time
08:49:00dwihnoB4gder: obviousley lobbying works.
08:50:33AncelotNext issue of Rockbox Chronicles: Rockbox goes Pr0n!
08:52:50 Join Seed [0] (
08:54:04 Nick Ancelot is now known as Ancelot`Zz (
08:54:09 Nick Ancelot`Zz is now known as Ancelot_Zz (
08:54:26 Part Noel_sad_song
08:54:56 Nick Ancelot_Zz is now known as Ancelot`Zz (
09:12:05LinusNsweden has had a cassette fee for ages, the law was only adjusted and the fees have become much higher
09:13:00LinusNiirc, it has been extended for digital media so it is based on memory capacity and not music playtime
09:13:23LinusNstriking a severe blow on hard disk players and dvd media
09:13:51 Join markun [0] (
09:14:09markunHm, the CPU stays boosted when I play ogg's now.
09:14:31markunGood morning btw :)
09:15:09markunSlasheri: optical out works fine now.
09:15:31Coldtoastis it in a bleeding edge yet markun?
09:19:26 Join potato_ [0] (
09:19:43potato_hello everyone
09:20:01markunpotato_: hu
09:20:09potato_hi markun
09:20:42potato_it seems that I am unable to rewind while playing an mp3 now, ff works ok though, is that a known problem?
09:21:27potato_latest bleeding edge
09:22:43markunno problems here
09:23:16B4gderit is a known issue
09:23:25B4gdersince yday or so
09:23:59markunWell, I can't go back to a previous song, but rewinding in the works.
09:24:12markun.. in the song ..
09:24:49potato_markun: for me, skipping back to the previous song works, however, rewinding to a previous position in a song doesn't work
09:24:57Coldtoastmarkun: there's one way you can skip back
09:25:16Coldtoastmake sure the HDD's spinning and it'll skip back to the prev song
09:25:38Coldtoastpotato_: let the HDD spin down and try
09:26:17potato_Coldtoast: aha
09:26:31Coldtoastdoes it still work?
09:26:43ColdtoastLinusN realised that one. heh
09:26:52potato_yes it works now, with the hd spinned down
09:27:02Coldtoasthmmm. interesting
09:27:10potato_Coldtoast: yep
09:27:14Coldtoastyou're using a build from teh last 2 days?
09:27:31potato_no, the latest cvs built
09:27:52Coldtoastisn't that a build fromt eh last 2 days?
09:28:18potato_I have no idea, I just cvs update and built it like an hour ago
09:28:18 Quit heidi ("CGI:IRC (EOF)")
09:28:24Coldtoastthat'd be the latest bleeding edge wouldn't it?
09:28:31potato_it should be
09:28:54Coldtoastok. skipping back doesn't work here unless the HDD is spinning. Linus found the same thing
09:29:28potato_hmmm rewinding while playing the song, with the HDD spun down does not work now
09:29:39potato_it seems like it works randomly
09:31:16potato_I see, with the HDD spun down, rewinding only works when all the songs aren't buffered yet
09:32:13 Join Chamois [0] (
09:33:40 Join stripwax_ [0] (
09:33:51potato_never mind that last statement, now with the HDD spun down, and all the songs buffered, rewinding works
09:34:08potato_apparently it works whenever it feels like it o_O
09:34:38markunstripwax_: morning
09:37:42LinusNmarkun: how about committing the s/pdif support?
09:38:30markunLinusN: It's always on now. Should I just commit and then later commit when it's configurable?
09:39:02LinusNalways on is acceptable
09:39:15LinusNi assume it's a very tiny hack
09:39:33 Join hardeep [0] (
09:40:42markunLinusN: yes, not very big:
09:40:57markun(also contains code to set the noise shape filter)
09:42:21LinusNi guess the only thing needed is a config option to control the s/pdif power
09:42:26LinusNi can do that
09:43:13stripwax_is spdif power automatically turned off if playback is stopped?
09:43:42markunLinusN: I turned spdif power on by default in power.c, shall I leave it like this until you commit the config option?
09:43:52LinusNdo that
09:44:20 Join Godeater [0] (
09:44:22LinusNi'll add the config option today
09:44:38markunLinusN: btw, when I last comitted the newline I typed in the comment was not visible. Should I use \n or something?
09:44:54LinusNi guess it should fit nicely in the playback settings
09:45:15 Part stripwax_
09:45:16LinusNmarkun: don't use -m
09:45:38CassandraOh yeah, settings.
09:45:41LinusNthen you will be able to edit the comment in vim
09:45:46markunLinusN: I didn't and used vim to type to comment..
09:45:58CassandraI am possessed of a desire to move "Manage Settings" onto the root settings menu.
09:46:07CassandraAny feelings either way about this?
09:46:12LinusNCassandra: i'm with you
09:46:24LinusNand i also want a separate "Save sound settings"
09:46:26B4gderI got an idea the other day
09:46:32B4gder"save resume point as bookmark"
09:46:39LinusNthat saves a .cfg file containing only sound related settings
09:47:07***Saving seen data "./dancer.seen"
09:47:12*LinusN runs to the lab
09:52:33 Quit potato_ (".")
09:56:12markunLinusN: committed
09:58:59CassandraHow much of a difference does the noise shape filter make, btw?
10:01:57markunCassandra: I couldn't notice it. Just followed the datasheet. Maybe someone with very good ears can?
10:04:55 Quit Bger ("BitchX: the ONLY bug-free client")
10:06:20 Join Bger [0] (~Bager@
10:09:30CassandraWell, I'll check it out later. I'll wait for Linus to commit the configuration option first I think, since I have no use for optical out.
10:17:07 Join [1]Chamois [0] (
10:25:01 Quit Godeater ("CGI:IRC (EOF)")
10:25:33 Quit Chamois (Read error: 145 (Connection timed out))
10:25:33 Nick [1]Chamois is now known as Chamois (
10:37:05Slasheriyes, that rewinding is still experimental and not fully functional (same with seeking)
10:37:40SlasheriI think the next thing i will do is to make these working better
10:38:31 Quit Seed (Nick collision from services.)
10:38:39 Join Seed [0] (
10:48:01 Join courtc_ [0] (
10:49:25 Join [1]Chamois [0] (
10:52:53 Quit courtc (Read error: 110 (Connection timed out))
10:53:48 Join preglow [0] (
10:56:14 Join Harpy [0] (
10:56:21*B4gder tries to get interrupts in his arm driver
10:56:29 Quit Chamois (Read error: 145 (Connection timed out))
10:56:29 Nick [1]Chamois is now known as Chamois (
10:56:35CassandraSounds painful.
10:56:53 Nick QT_ is now known as QT (as@area51.users.madwifi)
10:57:04markunpreglow: word?
10:58:06preglowin four hours i'll have delivered this bloody thesis, and will be found running screaming naked around trondheim
10:58:13 Join webguest12 [0] (
10:58:19preglowwatch for the headlines
10:58:55markunpreglow: last hours as a student?
10:59:00preglowthankfully, yes
10:59:21preglowi'll be missing all the spare time, though :/
11:05:33Coldtoastyou going straight into the workforce?
11:07:17 Quit Coldtoast (Read error: 104 (Connection reset by peer))
11:08:28 Join tvelocity [0] (
11:10:29 Quit ansivirus (Read error: 145 (Connection timed out))
11:16:51preglowmore or less, yeah
11:17:08preglowbut first i'll be moving
11:17:10preglowmoving is always fun
11:17:46preglowpacking everything down, relocating packed stuff, packing everything out, ahh, that's life
11:19:34 Join ansivirus [0] (
11:20:12amiconnLinusN: I have an idea how to solve the poweroff-while-textviewer-running-doesn't save-position problem
11:20:36amiconnIn fact we already have most of the solution
11:20:46LinusNyeah, a broadcast
11:21:41amiconnIt's just that SYS_POWEROFF has to be reacted on in the plugin
11:22:09amiconnAnd then the idle poweroff has to send this event too
11:22:39amiconnCurrently it is only send from the button driver for software poweroff platforms iirc
11:23:08LinusNi believe that's the way to go
11:23:36LinusNsend sys_poweroff from the idle timeout and intercept it in the plugins that want to handle it
11:24:53amiconnThe latter should be very simple
11:25:09amiconn...and it would even catch a button poweroff in the plugin
11:25:23amiconn(on sw poweroff platforms)
11:26:40LinusNit could be as easy as replacing shutdown_hw() with queue_broadcast() in powermgmt.c
11:27:11amiconnSame as the button driver currently does
11:27:19LinusNworth a try
11:27:33LinusNdefinitely a more elegant solution
11:27:50amiconnOf course then the code handling that event has to be included on all platforms
11:27:58LinusNisn't it already?
11:28:13amiconn...and the poweroff (click-click on rec v1, and from menu on player)
11:28:20amiconnsimply have to send that event too
11:28:32LinusNah, they call shutdown_hw() today?
11:28:40amiconnI guess so
11:29:23amiconnBtw, line 130 in button.c should be changed to #ifndef HAVE_POWEROFF_WHILE_CHARGING
11:30:05amiconn(for upcoming H3x0 port)
11:30:20 Nick Lynx_awy is now known as Lynx_ (
11:30:37amiconnChecking for model usually is evil, as we put it
11:30:53 Join [1]Chamois [0] (
11:32:29amiconnThe sleep timer should of course also send SYS_POWEROFF
11:33:44amiconnHmm, and plugins should have a new return value
11:33:55amiconn...for returning because of SYS_POWEROFF
11:34:23B4gderand they should be able to send it too
11:34:48LinusNa sys_poweroff() function
11:35:19amiconnOkay, then the return value isn't needed
11:35:31B4gderwell it depends
11:35:39 Quit Chamois (Read error: 145 (Connection timed out))
11:35:39 Nick [1]Chamois is now known as Chamois (
11:35:42amiconnHmm, I think it isn't necessary at all
11:35:45LinusNi can't make my mind up if "shutdown" or "poweroff" is the best term :-)
11:36:06hardeepbest of both worlds
11:36:09amiconnThe plugin should call the default event handler after doing it's own processing
11:36:19LinusNamiconn: agreed
11:36:23B4gderpowerdown? :-)
11:36:29amiconnThat will do the housekeeping and then shutdown
11:36:44LinusNlike today
11:37:02amiconnSYS_SHUTDOWN seems to reflect it better
11:39:44hardeep audio_stop();
11:39:51amiconnThe default_event_handler_ex callback function could handle shutdown in the viewer plugin. It can take a parameter which is unused today.
11:40:04 Join Zagor [0] (
11:40:40amiconnThis parameter could then tell the SYS_* event to the callback, which will react accordingly
11:41:00amiconn...not even needed :)
11:41:23amiconnviewer_exit(() will already do what's needed, both for USB and shutdown
11:43:07LinusNok, so who does this?
11:43:19LinusNamiconn: i take it you're busy with lcd stuff?
11:43:29amiconnI am
11:43:37LinusNthen i'll do this
11:43:59LinusNbtw, optical config option is committed
11:44:16amiconnclean_shutdown() is already defined for all platforms. Nice.
11:44:34 Join west-acre [0] (
11:44:44amiconn..only a #ifdef check for model again :(
11:44:47west-acrehey, optical out, how is it activated?
11:45:13west-acrewait. the light is on now :S
11:45:37LinusNwest-acre: the current bleeding edge build has optical always on
11:45:54amiconnLinusN: There's more to LCD stuff than I first thought. (1) I have to fix the simulators too (2) I have to modify bmp2rb.
11:45:54LinusNthe next will have an option in the playback settings
11:46:06LinusNamiconn: how fun :-)
11:46:07west-acrerite. kewl. presumably the power consumption is very low.
11:46:49LinusNi wonder if the idle timeout code should be moved to apps/?
11:47:10***Saving seen data "./dancer.seen"
11:48:09amiconnLinusN: Shutdown from the menu on player and the click-click shutdown from the browser on rec v1 need to be changed
11:49:17B4gdernow the ADC and LCD is about the main things lacking in the h300 code
11:50:34B4gdersince it pretends to be a h100 lcd now, and with a little adc patch it can now be made to build and link
11:50:54amiconnI'd rather not do this
11:50:54B4gderbut the ADC patch makes no sense so I won't commit that
11:51:00preglowLinusN: perhaps we should switch resampling rates on the uda when the optical out isn't active
11:51:07LinusNbtw, does resampling work with all formats now?
11:51:22LinusNpreglow: why?
11:51:31preglowLinusN: well, it saves us from having to resample
11:51:40LinusNpreglow: i don't get it
11:51:49preglowLinusN: if optical out isn't enabled, the point with always using 44.1khz vanishes
11:52:21LinusNthe 44.1kHz limit is in the cpu
11:52:31LinusNnot the uda
11:52:38preglowwell, yeah, but we can switch to 22050hz, yes?
11:52:44LinusNah, yes
11:52:55amiconnWe *could* switch 44.1/22.05/11.025, but always using 44.1 has advantages
11:52:57preglowthat's what i mean, i might have expressed myself a bit poorly as usual
11:53:01preglowamiconn: what?
11:53:03LinusNthat's true
11:53:16LinusNamiconn: crossfade
11:53:24B4gderI guess transitions will be harder otherwise
11:53:25LinusNpreglow: crossfade
11:53:35amiconn(1) With proper upsampling it will sound better (2) Way easier mixing (for crossfade, voice UI...)
11:53:56B4gderright, the voicing may go over song changes
11:54:01preglowi'm pretty certain the uda does upsampling better than we ever can
11:54:10preglowsince it doesn't actually have to do it in the digital domain
11:54:12markunspdif doesn't support < 32kHz I think..
11:54:15amiconnOkay, (1) Isn't valid for source sample rates of 22.05 and 11.025
11:54:18preglowmarkun: corrent
11:54:34preglowcorrect, even
11:54:48LinusNbtw, does resampling work with all formats now?
11:54:51amiconnThe uda does no upsampling
11:54:57preglowlinusn: doubt it
11:55:04preglowlinusn: didn't work with vorbis last time i checked
11:55:10LinusNhow come?
11:55:13preglowi haven't used rockbox for days so i don't know :/
11:55:25markunAnd cpu is always boosted now :(
11:55:30LinusNit is?
11:55:32markunwhen playing vorbis
11:56:08LinusNyes, i see that
11:56:43preglowmarkun: did you ever get your ov_read mod to work?
11:56:53markunpreglow: no
11:56:59preglowany idea why?
11:57:28markunI just passed the internal buffer with the 24-bit samples, but I didn't get any output.
11:58:10markunIs it cleared when I leave ov_read?
11:59:24 Join webguest90 [0] (
12:01:52 Join cYmen [0] (
12:02:58amiconnBagder: markun isn't resolved to a realname in the CVS activity list
12:03:48markunchange it to Marcoen if you like, that's my real name.
12:04:38markunthanks :)
12:05:12amiconnmarkun: I know, but this is done by a script only accessible for the server admins
12:05:51markunI thought everyone else used their real name to login to cvs :)
12:06:47amiconnI don't
12:11:27LinusNi wonder if we should broadcast SYS_POWEROFF or not?
12:11:38LinusNtoday it goes to the button queue only
12:12:38LinusNbroadcast seems better, but we don't have a synchronization mechanism for poweroff
12:12:45LinusNonly usb
12:16:18amiconnDo we have other threads that need to perform housekeeping before shutdown?
12:16:28LinusNnot really
12:16:39LinusNi'll go for button queue only, to begin with
12:17:17LinusNshutdown_hw() takes care of the stuff that could be done by the other threads
12:17:32LinusNlike stopping playback, flushing ata etc
12:22:09markunLinusN: When I try to set the CPU freq to 11MHz it goes back to 48MHz, is that a feature?
12:22:25 Join [1]Chamois [0] (
12:23:10LinusNmarkun: are you playing music?
12:23:57preglowas for the cpu freq switch
12:23:58LinusNlooks like the audio thread does something
12:24:02preglowwavpack actually decodes without boosting
12:24:25preglowi wonder if always switching to 128mhz when we do heavy work is beneficial
12:24:29preglow120, even
12:24:37preglowi kinda like the current approach
12:24:48LinusNhmmm, is the backlight handler boosting?
12:25:00markunah, yes.
12:25:10markunbecause of the fading..
12:25:13preglowit boosts to be flicker free
12:25:23LinusNok, that's the explanation
12:25:36preglowit used to flicker horribly when the cpu freq changed before
12:25:40preglowgave it a really broken look
12:25:57LinusNi see
12:27:07markunany word on wavpack encoding?
12:28:33 Join Moos [0] (
12:28:42MoosHi guys
12:28:53markunmorning Moos!
12:28:57Bgeris there any word in the datasheets about power consumption of CF5249VF at different freqs?
12:29:07Mooshi Markun, how goes?
12:29:29markunMoos: fine, optical out is committed
12:29:40LinusNthe "brag sheets" from motorola claim that it's linear to the frequency
12:29:49Moosmarkun: congratulations :))
12:30:39 Quit Chamois (Read error: 145 (Connection timed out))
12:30:39 Nick [1]Chamois is now known as Chamois (
12:30:58BgerLinusN any figures ?
12:31:13LinusNhmmm, don't remember
12:32:01LinusN1.3 mW/MHz
12:32:50Bgerwow:) lower than what i expected
12:33:45*Bger found a bug in iriver firmware :) surprise:)
12:44:59hardeephaha, it works
12:45:18hardeepany complaints if I add a feature to move to next folder once the current one completes?
12:46:14LinusNgo ahead
12:46:51 Join Coldtoast [0] (
12:46:56LinusNi'd like an option for it though
12:47:23LinusNyou will be the #1 misticriver hero
12:47:34Coldtoastdo any mono MP3s give L+R output at all?
12:47:39hardeepwhat i've always wanted to be
12:47:51Coldtoastlistening to the only other set of mono MP3s I listen to and they're left out only
12:48:50Coldtoastand they have a constant clicking noise thru them
12:50:17 Quit preglow ("CGI:IRC")
12:50:21Bger <= this is the bug if someone is interested in it
12:50:35 Join spiralout [0] (
12:50:49Coldtoasthmmm. all my mono MP3s have constant clicking and no right channel
12:51:37 Quit Chamois (" Want to be different? HydraIRC -> <-")
12:51:46 Quit west-acre ("—I-n-v-i-s-i-o-n— 2.0 Build 3515")
13:02:12LinusNSlasheri: can you explain the problems with seeking backwards?
13:03:56 Quit edx (Read error: 145 (Connection timed out))
13:04:34*HCl got new headphones :)
13:04:44markunHCl: Which?
13:05:13ColdtoastI want to buy ER6s I think
13:05:38ColdtoastI'd buy E3Cs but they're quite expensive
13:06:14ZagorColdtoast: have you listened to them?
13:06:32Coldtoastno. I;d be buying blind
13:06:40Coldtoastnobody here stocks them
13:06:54Zagori did that and was hugely disappointed.
13:07:31Coldtoastwhat's that? ER6 or E3C?
13:07:45Coldtoasthmmm. ok
13:08:09ColdtoastI just need a good set of not-too-expensive in-ear phones
13:08:32Coldtoastmaybe e2c's then?
13:09:35Zagori haven't listened to any of the shure headphones, unfortunately. i use sonys ex-70 myself.
13:10:14Coldtoastare they in-ear?
13:10:27Coldtoasthmmm. ok
13:10:46Zagori think they're called ex-71 nowadays
13:10:53hardeepi'm not a fan of the ex-70 personally... the bass is overpowering
13:11:02hardeepmuddies the sound
13:11:02 Quit Cassandra (Read error: 145 (Connection timed out))
13:11:32Zagoryeah, i'm not claiming they are better than shure. just better than etymotics. (which you can't even force to reproduce bass tones)
13:11:34Coldtoasthow about those?
13:11:53ZagorColdtoast: yeah those are the ones
13:12:00ColdtoastI'll buy them
13:12:06Coldtoastthey're hella cheap
13:12:20Zagornot risking much money :-)
13:12:40LinusNi really like my ex-70's
13:12:41Coldtoastgrr! no PayPal tho
13:12:59ColdtoastI refuse to buy internationally without Paypal
13:12:59ZagorColdtoast: you should be able to find them pretty much everywhere though
13:13:21Coldtoastthat cheap tho?
13:13:28Coldtoastanyway. must get ready for work
13:15:21HClmarkun: sennheiser px 200
13:15:42crwlpx 100 are better ;)
13:16:45HClthey had those too
13:18:21crwli hear px 200's have problems with bass response if you can't position them perfectly on your head
13:18:45crwlOTOH, they are somewhat insulated, which might be a good thing
13:18:54crwlbut other than that, they're same than px 100's
13:19:19HClthey're closed
13:19:34HClthey seal pretty well
13:20:33Zagori find it interesting how much opinions differ when it comes to headphones. i couldn't find a single bad review of the etymotics before I bought them. and some people swear by Koss' "The Plug", which sounds like total crap in my ears.
13:23:02SlasheriLinusN: sure, currently the buffering implementation ignores to seek backwards if that position is not buffered
13:23:05markunMy The Plug doesn't sound very good, but good enough to take with me on the train so I don't hear all the screaming kids.
13:24:19LinusNSlasheri: i think it ignores it even if it is buffered
13:25:01Slasherihmm, interesting. I will look that when i get to home :)
13:27:17HClcan't we keep the entire current song buffered till its done?
13:27:37B4gderwhat if it is 150MB
13:27:40LinusNwhen i do a short left-click, it gives me this:
13:27:41HClgood point
13:28:08LinusNrewind: 0
13:28:29 Join oxygen77 [0] (
13:28:36 Join Sucka [0] (
13:28:41LinusNseek failed (reload s
13:29:07SlasheriLinusN: Hmm, are you sure the whole trak is buffered?
13:29:28LinusNyes, i play it from the start, wait a few seconds and the left-click
13:29:44Slasheriah, so there is no buffer refill between that click
13:29:56Slasheriok, i will fix that soon :)
13:31:04LinusNnormal rewind seems to work ok
13:31:17HClwhats normal rewind?
13:31:25LinusNhold left
13:31:43LinusNit's ff_rew(0) that fails
13:32:37amiconnI wonder why this (skip back/forward, seek backward/forward) is so hard to get working properly. I haven't checked it myself, but imho the necessary logic is identical to what we use on archos
13:33:20LinusNyes, but the code is 100% new, unfortunately
13:33:24amiconn...and of course it should work in all cases, independent whether the part skipped or seeked to is buffered or not
13:33:32 Quit ashridah (Read error: 110 (Connection timed out))
13:33:55LinusNaudio_ff_rewind() does this:
13:33:57LinusN ci.seek_time = newpos+1;
13:34:03LinusNwhy +1?
13:34:06amiconnA track might always be larger than the available RAM, especially with lossless codecs
13:34:58LinusNaha, in mpa.c:
13:35:00LinusN if (ci->seek_time) {
13:36:09 Join ashridah [0] (
13:37:40B4gderwhoa, what a hardeep commit!
13:47:14***Saving seen data "./dancer.seen"
13:47:45amiconnLinusN: Imho the codec api is a bit strange. It uses some fields directly for parameter passing.
13:48:33amiconnThat's a bit unclean I think, and it *will* break if we ever support more than one codec running at a time
13:51:12amiconnWell, there's only one api struct. The core code couldn't distinguish which codec wrote a value there, and the codecs could not decide whether a value is directed to them, or another codec
13:52:01B4gderI agree with amiconn
13:52:16B4gderwe might want to have two codecs at once
13:52:43ashridahie, voice menus
13:52:45amiconnEven if a second codec might be built in, it should also use the api
13:52:54LinusNi thought each codec would have its own instance of the struct
13:53:07 Join pike [0] (
13:53:19amiconnLinusN: You would duplicate the whole struct just for the parameter fields ???
13:54:59LinusNif understand correctly, the codec api contains several callbacks
13:55:25LinusNwhich will probably be different for the voice playback and the music playback
13:55:34B4gderah, yes
13:56:10 Join bipak_ [0] (
13:56:30B4gderci.request_buffer = codec_request_buffer_callback;
13:57:20B4gderor am I on drugs?
13:57:40B4gderthose are just normal api functions
13:58:23LinusNthose are the ones
13:58:52B4gderthey're the same for all codecs
13:58:58LinusNi *assume* some of them will be different for the voice playback
13:59:36LinusNso they can't use the same instance of the codec api struct
14:00:58 Quit B4gder ("go go go")
14:02:25amiconnLinusN: Hmm. The codec api does also contain pointers to core functions like memcpy() etc. These would then be duplicated as well
14:02:28 Quit bipak (Read error: 145 (Connection timed out))
14:02:39LinusNamiconn: yes
14:02:58amiconnHmm, wasting space?
14:03:24amiconnMaybe we should leave it as-is for now
14:03:46amiconnIs it possible to declare individual struct members as 'const'?
14:03:49LinusNyes, we can reduce the 1Mbyte PCM buffer with 16Kbyte if we run into memory problems
14:05:23 Quit webguest12 ("CGI:IRC")
14:05:32LinusNSlasheri: u there?
14:06:44*HCl ports rashers crc routine..
14:07:00amiconnLinusN: I mean, if the struct itself is not const, we should declare those pointers to core functions as const, to protect from accidentally changing them in the code
14:07:18HClwhy not have a sub structure with the parameters and duplicate that
14:07:53*HCl really should read more of the conversation before slinging suggestions around.. *goes back to crc*
14:08:06 Join RotAtoR [0] (
14:09:28LinusNamiconn: i see, i'm not sure if that's possible
14:13:11amiconnIiuc these should be constant pointers to constant data, but I'm not sure whether this will work in a struct definition
14:13:26amiconnint (*ata_sleep)(void);
14:13:30amiconnwould become
14:13:38amiconnint (const * const ata_sleep)(void);
14:15:51LinusNi don't think that works
14:17:23 Quit hardeep ("BitchX: the ONLY bug-free client")
14:22:05ashridahahaha. bug free. bitchx.that's funny/
14:26:34markunbitchx is always funny :)
14:26:45HClhorrid client
14:26:50HCli don't understand why people don't use irssi
14:26:57ashridahit's also buggy
14:27:09HClno problems here
14:27:12HClscreen is more buggy than irssi
14:28:44HClwow, a classcastexception
14:29:08HClwell duh
14:29:08HCl :)
14:29:08LinusNSlasheri: i think i have found the bug
14:29:47LinusNstart_pos is advanced by the first initialize_buffer_fill
14:30:13HClrasher: hashing the first 512 bytes is definately not enough, heh.
14:30:19LinusNso you can't seek back to before the first call to audio_check_buffer()
14:31:32HClfirst 32 kb
14:31:35HClseems to be much argh.
14:31:41HCloh wait.
14:31:48HClnope, 32 kb seems to be much more reliable
14:31:49crashdirssi pwns :]
14:31:53HClthese are indeed duplicate files
14:32:38 Quit oxygen77 ("CGI:IRC")
14:33:18 Join Godeater [0] (
14:39:22BgerHCl what about md5 ?
14:39:34HClmd5 is slow.
14:39:43HClplus i'm getting good results with crc32 on the first 32kb
14:40:36Bgertime will tell...
14:41:04HClthe db generator reports duplicates at the moment
14:41:14HClusers should be able to tell whether it does it correctly or not.
14:41:25Bgerthat's what i mean
14:41:43Bgerif there are no complains, then it's ok
14:44:08 Join oxygen77 [0] (
14:44:39*HCl wonders how to print an integer as hexadecimal in java
14:45:09HCli got a lot of duplicates
14:45:37 Join rasher [0] (
14:46:13rasherHCl: I believe my suggestion was to hash the first *and* last 512 bytes
14:46:21 Quit ashridah ("Leaving")
14:46:25rasherwhich did in fact not produce any duplicates on my entire collection
14:46:27 Join ashridah [0] (
14:46:31HClwell, i hash the first 32kb at the moment..
14:46:35HCldunno whats faster
14:46:48HCleither way, i think that with the current design we might not even have to hash on target
14:46:58rasherI do know which is more memory intensive
14:47:04HCl :P
14:47:05rasherBut.. the seek may be expensive
14:47:20HClfor now we won't need to hash on target
14:47:28rasheralso, doing crc on 32kb is probably a good deal slower than on 1kb
14:47:44rasherWell.. what if people rename their files?
14:47:48rasheron target
14:48:06rasherIt'd be pretty cool to have it still recognise it in that situation
14:48:42amiconnBoth hashing the first 32KB and hashing first+last 512 bytes does not protect from duplicates case the file is re-tagged
14:49:03rasherIt doesn't need to... as long as noone sees any duplicates
14:49:24rasherAlso.. if we're only hashing on the pc - it'd be nice to seek to the first mpeg frame for mp3s
14:49:29rasherand skip vorbistags for oggs
14:49:46HCli'll look at that later, the important bit is i have a decent hash for testing :P
14:50:06rasherIndeed, the hashing can always be tweaked
14:50:11HCli might even get away with a self-repairing runtime database
14:50:11rasherwith no ill effects
14:50:42rasherThe release notes are getting pretty extensive..
14:51:24rasherWould be cool to get the runtimedb in, and then do a release?
14:51:55 Join edx [0] (
14:52:27amiconnThe v2.5 was already taken a week ago or so
14:52:36amiconn*v2.5 snapshot
14:52:58rasherDidn't know that..
14:53:04rasherMust've missed it
14:53:09 Join fogcat [0] (
14:53:12rasherwellthen.. shouldn't have been updating the release notes then
14:54:17 Join xen` [0] (
14:54:33rasherwhat's happening to it now, then?
14:54:52 Quit markun ()
14:56:46fogcatI'm looking at listing the ID3 tag info as a scrolling screen instead of item by item
14:56:54fogcatWas looking at the scource to see how scrolling is done
14:57:04fogcatLooks like the easiest way would be to write the ID3 info to a temp file and call the viewer on that file.
14:57:10fogcatwould that work? Are temp files frowned upon?
14:57:28 Join Cassandra [0] (
14:57:39rasherWhat, like a context menu - "View all tags"?
14:57:58fogcatwas thinking as a replacement for the current view ID3 info
14:58:07rasherDoesn't seem too bad, but it clashes wonderfully with the edit id3 patch I have lying around
14:58:33SlasheriLinusN: oh, i think that is the problem because i changed that check_buffer code recently :)
14:59:16fogcatwould only be leaving the file around while info is viewed
14:59:18rasherfogcat: The patch removes the "View id3 info" menu choice, and adds viewing/editing id3 info instead
14:59:31HClsounds good
14:59:41rasherIt does. It also doesn't work.
14:59:50HCl :/
14:59:52rasherI'll try creating a patch that isn't outdated
15:01:29SlasheriLinusN: ah, yes. The idea why i did that change was to make sure that codec buffer gets fully filled. But unfortunately that throws away old buffered data from beginning of the track
15:01:53 Join elinenbe [0] (~elinenbe_@
15:02:13fogcatahh np... just looking for something really really simple as a spur to actually do some development
15:02:27 Quit elinenbe_ (" HydraIRC -> <- IRC has never been so good")
15:02:44rasherWell, go right ahead
15:02:49SlasheriDo you think it's enough to fix the backward seeking so the track will be rebuffered if necessary or should the buffering code try to keep always whole track buffered?
15:03:15fogcatfirst of all gotta sort out cygwin then dev envinonment etc etc
15:03:20LinusNi think trying to buffer entire tracks is a bad idea
15:05:57Slasherithat's true.. maybe i will just fix the seeking
15:06:39 Join webguest99 [0] (
15:07:11LinusNSlasheri: this check almost always fails:
15:07:12LinusN if (ci.curpos - difference < cur_ti->start_pos) {
15:07:38*amiconn just got a whole bunch of ideas how to improve the menu and settings look & feel
15:08:00LinusNstart_pos seems to be the position in the buffer, but curpos is the relative position in the song
15:08:25elinenbewhat would be the negative of buffering an entire track?
15:08:47LinusNelinenbe: waste of space, and complications then the track is larger than the memory
15:09:37elinenbedo you currently fill the memory buffer with as much as possible right now?
15:09:42fogcatamiconn : sounds interesting
15:09:48SlasheriHmm, start_pos should be the first file position that is buffered and curpos the current file position
15:10:29 Quit fogcat ("CGI:IRC")
15:10:56Slasheriboth are the indexes from beginning of the file
15:11:25LinusNi'm currently playing the third buffered song
15:11:41LinusNwhen i click left, i get this:
15:11:48LinusNnewpos = 65535
15:12:06LinusNci.curpos = 6657169
15:12:17HCli got a panic when plugging in usb
15:12:30LinusNdifference = -6591634
15:12:40SlasheriHmm.. :/
15:12:54LinusNcur_ti->start_pos = 2478497
15:13:17LinusNci.curpos - difference = 65535
15:13:48LinusNso it will be: if(65535 < 2478497)
15:14:15LinusNsomething is really wrong there
15:14:45Slasheriyes, i will investigate that
15:15:54HClwith a bit of luck
15:16:06HCli'll have the first working runtime db version committed soon :)
15:17:21SlasheriLinusN: Hmm, if the start_pos is correct, that seems to be ok
15:18:05Slasherii will have to fix that if where is the logf("Seek failed (reload song)");
15:20:51rasherLinusN: I'm trying to update the edit-id3 patch.. I did this a few days ago, but my tree got out of sync.. There are problems I don't know how to fix though, but I'll update the patch
15:20:51 Quit webguest90 ("CGI:IRC (EOF)")
15:21:01 Quit webguest36 ("CGI:IRC (EOF)")
15:22:16SlasheriLinusN: in fact fixing that shouldn't be hard, i will try :)
15:23:29HClHash: 0xf569592f
15:23:29HClRating: 0
15:23:29HClVoladj: 0x0
15:23:29HClPlaycount: 0x1
15:23:29HClLastplayed: 0
15:23:31HClweeeee :)
15:24:02*HCl goes to work on the searchengine a bit
15:24:04elinenbeHCl: awesome!
15:24:32elinenbeI think that the database viewer/seach needs to be "more integrated" into rockbox...
15:25:19rasherOkay.. really.. how do I create a patch where new files are included using cvs diff?
15:25:31rashercvs diff -Nu doesn't seem to cut it
15:26:09rasherHCl: as an added bonus - the edit id3 patch adds a method for calling plugins from the interface
15:30:04amiconnrasher: cvs diff only includes new files if you 'cvs add' them to your local working copy
15:30:25amiconnTo go without that, you need to use ordinary diff
15:30:51rasherI thought cvs add told the server about the file
15:30:56rasherwhich is why I wouldn't use it
15:31:00rasherall is good then
15:31:41rasherso I can safely cvs add whatever I want?
15:32:04amiconncvs add adds the file to your local copy. It would then be sent to the server when doing a cvs commit
15:33:12LinusNSlasheri: i can't see how the calculation of "difference" can be correct
15:35:21LinusNi assume "newpos" is an offset from the beginning of the file
15:37:20rasheredit-id3 patch updated.. it's broken though
15:37:45rasherDoes apply, however
15:42:11 Quit rasher ("CGI:IRC 0.5.4 (2004/01/29)")
15:43:39*HCl wonders whether things are working properly
15:43:53SlasheriLinusN: Hmm.. difference = newpos - curpos. Then if difference < 0 we are seeking backwards so negate the difference. Now we know how much to divide from curpos to get to the newpos
15:44:01HClanyone volunteering to add hashing to the perl version of
15:44:11HClits gonna be required if people will want runtimedatabase
15:44:28Slasheriand if that newpos < start_pos, we will have to reaload the track
15:44:37Slasherii don't see any problems with that :)
15:47:15***Saving seen data "./dancer.seen"
15:48:32CassandraI'd like to see songdb as a plugin.
15:48:37CassandraNot enough to write it though.
15:49:04HClwhat do you mean not enough to write it?
15:49:44HClincidentally, what happens when a plugin uses the audiobuffer for temporary storage?
15:49:57HCl*stares at searchengine*
15:50:04Bgervery good things
15:52:06 Join preglow [0] (
15:52:20Slasherihi preglow :)
15:52:24HClhey preg
15:52:27preglow'tis a lovely day
15:52:44HClhow so?
15:52:47HCltwas raining here
15:52:55preglowlovely here, for once
15:52:58preglowand i've just handed in my thesis
15:53:01preglowand i can leave this hell hole
15:53:05HClget to work
15:53:09HCl :P
15:53:11preglowgotta move first
15:53:42preglowfor now, such delightful tasks as washing the entire kitchen, livingroom, dishwashing, washing clothes, and packing remains
15:53:46Bger"wanting to raise the incomings, microsoft announced starting of selling advertisement areas on error messages"
15:54:56HCli think i managed to make the rundbdirty flag in the tagdatabase header redundant
15:55:18HClits all self-repairing now based on the hashes
15:55:38 Join ]RowaN[ [0] (a2b0y@
15:56:05Slasherigreat. and i have only three weeks left before i have to begin the military service (in civil) :/ so i will be then at least one month off from rockbox development .(
15:56:08]RowaN[guys do you think its worth me spending £100 on a new 30gig hd for my iriver, or is there a better player i can get for around that price?
15:56:22HClhm, bug..
15:56:26]RowaN[i would want something that has optical out
15:56:34HClwhy not 60?
15:56:40]RowaN[60 is too expensive
15:56:45]RowaN[well, so is 30 heh
15:56:47HClit seems to me that after buying 30, you'd have to upgrade at a later point
15:57:08]RowaN[yeah i know, but i really cant afford 60giger
15:57:15CassandraCan you get a thin form factor 60 these days?
15:57:20]RowaN[space aint much of an issue for me anyway
15:57:46]RowaN[if 1.8" is thin then yes @
15:58:09]RowaN[why did i have to drop my iriver!! *cries*
15:58:22preglowSlasheri: persuade them to let you serve as a rockbox coder
15:58:33HClCassandra: yea
15:58:40]RowaN[they should make 1.8" hd's with those sensors that stop it spinning when its falling (mercury tilt switch??)
15:58:42HClCassandra: 80gb is also out, just not publically for sale yet
15:58:58 Join ep0ch [0] (~ep0ch@
15:59:38amiconn]RowaN[: I also dropped my iriver :( Even though the disk was *not* spinning (the iriver was completely off) the HD suffered from it
15:59:40HCli just encountered something odd, as if memset was failing.. *investigates..*
16:00:00amiconnIt still works, but it is quite noisy while it spins. Definitely not good :(
16:00:07]RowaN[i read it suffers 3x more damage if it is spinning.. would have given me more hope
16:00:17Slasheripreglow: hehe :D that might be possible when i can begin working at the states work place i have selected (unix system administration). But there is still one month training period and that time i don't have access to computers
16:01:02]RowaN[so guys, there isnt a smaller/better/neater jukebox i should consider?
16:01:02amiconnThe iriver is too slippery. I never managed to drop my archoses. They are double-protected by the bumpers
16:01:04Slasherii hate that mandatory mil service we have in finland :)
16:02:30 Part webguest99
16:03:40*HCl has a leather protective thing around his iriver..
16:03:41CassandraThe iSkin is good for making an iRiver non-slippy.
16:03:56HClit came with my h140
16:04:06amiconnHCl: The leather case that comes with the iriver has 2 serious deficiencies for me
16:04:23amiconnThere are no holes for (1) reset and (2) line out
16:04:29HCli know i should probably get an iskin
16:05:54amiconnI probably need to get a new HD some time soon
16:06:22 Join PaulJ [0] (
16:06:34Slasherii have ordered a new battery from dapstore.. still waiting to get it
16:06:37Harpyi made a hole for line out.. should make one for reset too now that i'm testing rockbox
16:07:02HClhow expensive are iskins?
16:07:14HCli'm betting they sell them at outrageous prices when the cost to make them is probably like 3 bucks
16:07:55preglowSlasheri: oh, we have it as well, but it seems i'm too old for them now
16:08:13preglowand there was much rejoicing
16:08:55Slasherihehe, nice ;)
16:08:56]RowaN[3 bucks? more like 20 cents
16:09:29LinusNSlasheri: can startpos be bigger than curpos?
16:10:27preglowbut i gotta go buy a psu
16:10:29 Quit preglow ("leaving")
16:10:35LinusNare those offsets from the beginning of the codec buffer?
16:10:54SlasheriLinusN: no, it can't but indeed there are something wrong with seeking
16:11:14Slasheribeginning of the real file
16:11:22Slasherifile that is buffered
16:12:05amiconn60GB Toshiba MK6006GAH: EUR 175,68
16:12:21amiconnThat should fit into a H140, right?
16:12:47Moosyes it can
16:13:03]RowaN[where is it that price?
16:13:25MO-PantsuI'm experiencing some weird bugs in the bleeding edge
16:13:28LinusNSlasheri: so in the beginning of a track, start_pos == curpos?
16:13:47Slasheriand both are zero at the point
16:14:20LinusNstartpos is often much bigger than curpos
16:14:44*HCl stares blankly at an odd bug
16:14:46MO-Pantsuclicks and pops when I play or FF, stuttering when playing back an Ogg sometimes and sometimes when I FF and let go it stops playing and the hd led flashes really fast.
16:15:08amiconnThe 60GB model is much more shock resistant than the 40GB according to the specs
16:15:28SlasheriLinusN: hmm, that is certainly a bug
16:15:31Harpywhat would it need to get a support for displaying japanese tag info? is this only a font issue?
16:15:32MO-Pantsuanyone else noticed those probs in the bleeding edge? hoping it's not just me
16:15:54MO-PantsuHarpy it requires unicode support
16:15:56MO-PantsuI think
16:16:14Harpythe code is not using unicode?
16:16:21MO-Pantsushift-js or whatever it's called
16:16:34MO-PantsuI dunno I get confused with that stuff :)
16:16:41*HCl catches the bug and puts it in a jar for display
16:16:45*HCl makes holes in the lid
16:16:55MO-Pantsuso nobody else getting weird playback bugs in the latest build?
16:17:05Harpyi know there is more than one encoding for japanese text
16:17:54MO-Pantsuyeah well there is kanji and a number of other character sets in the Japanese lingo
16:18:11MO-Pantsucomplicated alphabet soup :)
16:18:55Harpyi mean like shift-js, euc-jp etc. and utf-8 works too, of course
16:20:26Harpyiso-2022-jp1/2/3, euc-jisx0213, sjisx0213.. sic
16:22:58 Part LinusN
16:23:47CassandraAre there full unicode BDF files out there?
16:23:52MO-PantsuSlasheri seems to have gone
16:24:15MO-PantsuI wanted to point out the playback problems I noticed
16:24:28SlasheriMO-Pantsu: no, i am still here :)
16:24:42MO-Pantsuwell did you catch what I said?
16:24:56MO-Pantsubtw mp3 seeking is accurate now but some other bugs cropped up
16:25:04ep0chis Vorbis playback still stuck in boost mode?
16:25:27]RowaN[hmm, german website.. why dont i trust germans
16:25:28SlasheriMO-Pantsu: Hmm, what kind of bugs do you mean?
16:25:33ep0cham scared to use Vorbis now...
16:25:57SlasheriI know the seeking and previous song button is broken at the moment
16:26:10MO-Pantsuwell now I get clicks and pops when I hit playback and when I seek and let go I get clicks and pops before it continues....
16:26:30Slasheriah, interesting
16:26:31MO-Pantsualso seeking sometimes when you let go it plays a bit for a second then stops, then plays again
16:26:40MO-Pantsuand also...
16:26:56MO-Pantsuwhen you FF sometimes it gets stuck for ages after you let go with the led flashing
16:27:05MO-Pantsuthere's 3 ;)
16:27:09]RowaN[40gig for 118 euros?
16:27:16BgerCassandra <= like this ?
16:27:20]RowaN[will that work in iriver?
16:27:45MO-Pantsuthe click and pops are nasty on headphones...hurts my ears
16:28:07SlasheriMO-Pantsu: thanks for the report, i will investigate what causes those pops
16:28:19SlasheriMO-Pantsu: btw, are you using the headphone connector?
16:28:31HCl]RowaN[: thats link to a 60gb one?
16:28:34HClthats a*
16:28:38MO-PantsuI tried to play an Ogg on first putting the new build on and it stuttered like crazy. seems OK now. But I had not rebooted so maybe it needed that
16:29:41HClshould work, do you have an h140 or h120?
16:29:46MO-Pantsuseems like hdd access needs working on to work well with when you do things while it's playing
16:30:08MO-Pantsuespecially when it buffers up
16:30:15]RowaN[h120 here
16:30:22MO-Pantsubuffering up sends things a bit flakey
16:30:51]RowaN[i dont understand why they are selling 40gig for cheaper than 30gig.. also they have 2 toshiba 30gig drives at different prices (all 1.8" size)
16:31:10MO-PantsuI want a 60gb drive but they cost too much
16:31:13HClyou'd have to open it up and see if you can fit 8mm in a h120
16:31:24HClmy guess is that it wouldn't
16:31:34Bger]RowaN[ maybe this 40gig is 2 platter (lower dentisty)
16:31:47ep0chwhat tall is the existing h120 drive?
16:31:55Bger1 platter
16:32:22MO-Pantsuis there a new graphics api on this new build? looks different
16:32:45]RowaN[hcl: 8mm , what do you mean?
16:32:51HCl]RowaN[: the drive is 8mm thick
16:32:58HCli doubt you can fit that in your h120 casing
16:33:04HClthe 30gb one is 5mm thick
16:33:13]RowaN[sounds like a challenge
16:33:36ep0chprobably why the 30gb is more expensive then
16:33:52Bgerep0ch that's what i tryed to explain
16:34:01Bgerit's newer
16:34:15amiconn]RowaN[: 1.8" disks come in 2 different heights, 5mm and 8mm
16:34:30amiconnThe 8mm drives don't fit into a H120
16:35:13amiconnThere should be a 5mm 40GB disk, MK4006GAH
16:35:31ep0chlets say i were to get a 30gb disk and fit it in, what would the procedure be to get it up and running? go into iriver firmware and format the disk?
16:35:43Bgeramiconn : in fact, you can put 8mm drive, but you must remove the soft "protectors"
16:36:00Bgerand this is bad
16:36:26Bgerep0ch something like this will be enough
16:36:38ep0chsoft protectors? you mean that rubber thing that i thought had no purpose?
16:36:45amiconnYes I can imagine. Even with these protectors the disk isn't protected too well...
16:36:48Bgerep0ch yes, this
16:36:55]RowaN[the prices on this german site are ex vat?
16:36:59MO-Pantsugapless playback actually works rather well for non-gapless non-lame mp3's too
16:37:15MO-PantsuGreat job Slasheri and Preglow etc
16:37:19amiconn]RowaN[: No they are incl german VAT, but excl. shipment
16:37:56MO-Pantsulistening to a dj mix that is non-lame mp3 and it hardly misses a beat
16:38:10*Bger will wait till 80GB drives are at the market and aren't so much exspensive ..
16:38:29Bgerat least i hope that i won't have other reason for buying new HDD....
16:39:01ep0cham suprised how quickly it took me to fill up 20gb :(
16:39:13Bgerthe same with 40GB ..
16:39:16HCldon't you hate it when things were working earlier but suddenly don't anymore..
16:39:19*MO-Pantsu is falling in love with his H140 all over again :)
16:39:56BgerHCl check the changes u've made ...
16:40:03ep0chi dare you to use iriver fw for a day!
16:40:22Bgerep0ch i'm still using it every day ...
16:40:44ep0chreally? i can't stand it now
16:40:49tvelocityiriver fw is not THAT bad
16:40:49 Quit oxygen77 ("CGI:IRC")
16:40:54MO-Pantsuit looks pretty
16:41:59]RowaN[i bet the hd would get taxed to hell when it arrives here in the UK from germany ='/
16:42:11ep0chget it marked as "gift"?
16:42:34MO-Pantsubut customs does not apply across europe?
16:42:39Bgeranyone knows who was working on Unicode support
16:42:51MO-PantsuI thought customs only applied outside the european union
16:43:00MO-Pantsuduty that is
16:43:14ep0chi think UK still have customs?
16:43:25tvelocityi don't think so
16:43:40MO-Pantsunot sure
16:43:54MO-PantsuI know the limit is a value of £18 from the US etc
16:44:08MO-Pantsubefore you get taxed
16:44:28MO-Pantsuand it can be potluck sometimes
16:44:35tvelocityyeah but the US is not in the EU :P
16:44:41ep0chahh that explains the price of the headphones on my parcel yesterday...
16:44:42MO-PantsuI have ordered in DVD boxsets over that and not got charged
16:45:18MO-Pantsuother times you get wacked not only with tax but a handling fee
16:45:28MO-Pantsuthe handling fee is a ripoff
16:45:40]RowaN[on this mindfactory all the 1.8" hds seem to have a listed height of "7.8 cm"
16:45:49tvelocityi'm glad I only order books from the US...
16:46:45MO-Pantsuwhat's wrong with the new Rockbox bleeding edge?
16:46:50MO-PantsuOgg playback is screwed
16:47:03ep0chcpu boost is stuck at 100%?
16:47:05MO-PantsuIt's stuttering
16:47:16]RowaN[ah now i see, tech description is fooked, but the title says the height
16:47:35MO-Pantsuis the stuttering caused by the cpu boost bug?
16:48:20MO-PantsuOK now I click on play an Ogg file again (same file) and it won't play it
16:48:28MO-Pantsuit stopped with the wps
16:48:54ep0chi thought maybe vorbis codec just got really really slow... so cpu boost is stuck high to try and cope
16:49:17MO-PantsuI can't play any Ogg files
16:49:41ep0cham sure slasheri knows... but wont say :p
16:49:49MO-Pantsunope...they won't play
16:50:04Slasheriep0ch: Hmm, i haven't looked at that yet
16:50:06MO-Pantsulets try the last daily
16:50:08SlasheriBut vorbis is slow
16:50:20ep0chrecently it's got really really slow
16:50:33SlasheriHmm, i will try out
16:50:33 Nick Ancelot`Zz is now known as Ancelot (
16:50:37]RowaN[ok, that site has 2 different 5mm toshiba 30gig hds.. one is 105, the other is 113 euros .. ?
16:51:36Ancelotcheck the rpm
16:51:58]RowaN[TOSHIBA 30GB HDD 4200rpm IDE 1,8Zoll 2MB embedded 5mm
16:51:58]RowaN[30GB Toshiba MK-3006GAL HDD 4200rpm IDE 1,8Zoll 2MB embedded 5mm
16:51:59ep0chprobably different models
16:52:09Slasheriep0ch: oh, there is something wrong with vorbis
16:52:16Slasheriit won't even play q10 files
16:52:34ep0chand has a hard time playing anything
16:52:40ep0chalways boosting
16:52:41]RowaN[both are the MK3006GAL it seems.. if the product description is to be believed
16:52:50*]RowaN[ will order the cheaper one =]
16:53:15MO-PantsuOgg is broken in daily too
16:53:28 Nick Ancelot is now known as Ancelot`Kain_Def (
16:53:33ep0chSlasheri: has Tremor codec been changed in the last 24-36 hrs?
16:53:42Slasheriep0ch: yep, really weird
16:53:48SlasheriThere must be some change somewhere..
16:53:54Slasheribecause last time i tested it worked
16:54:19MO-Pantsuok back to an even earlier build
16:54:23ep0chstarted happening to me at about midnight
16:54:33ep0chso its something before then
16:55:40amiconn]RowaN[: (1) The specs are correct, 'Tiefe 7,8 cm' means depth, not height
16:55:55MO-Pantsuyou broke my rockbox! lol
16:56:01amiconnheight is 'Höhe' which is correctly given as 5mm or 8mm
16:56:36amiconnThe duplicate article seems to be a bug in the shop...
16:57:28MO-PantsuOK build bfore today seems to work fine
16:57:44MO-PantsuOgg is broken in todays daily and bleeding edge
16:58:14MO-Pantsuwho broke'd it? ;)
16:59:59]RowaN[bah that site dont ship to England!!!!
17:01:09MO-PantsuI have 1.8gb left on my H140. I need to make better use of it
17:01:15amiconn]RowaN[: :(
17:01:26MO-Pantsuget rid of what I don't listen to and arrange it better
17:01:31amiconnI wouldn't have expected that
17:02:50MO-PantsuSlasheri are you looking to see what changed that might have caused ogg to stop working?
17:04:02godzirraWhats a good Air cd?
17:04:19ep0chermmm moon safari of course
17:06:27ep0chnothing has changed in tremor for 9 days...
17:07:09ashridahep0ch: lots has changed in the buffering code tho
17:08:27ep0chwouldn't you expect mp3 playback to fall over too if it's an issue with the buffering code?
17:09:17*HCl almost forgot what he was doing, but remembers
17:09:27*HCl has to listen to his air album in order to test his code :P
17:09:38HClself-repairing the runtime database
17:09:50ep0chwhich air album?
17:09:53HClmoon safari :)
17:10:00HCli don't like the others much
17:10:07ep0chme neither
17:10:20HClmoon safari is great though
17:10:25HClnow that i remembered what i was doing
17:10:32HCli might as well continue watching my flatmate play san andreas
17:10:40HClsince i have to wait till this album plays multiple times
17:10:41HCl :P
17:10:55ep0chhaven't got into SA
17:11:37ep0chtime i got up i think
17:14:22 Quit Sucka ("a bird in the bush is worth two in your house")
17:14:32*ashridah motherfucking overwrote his 98.93% completed saved game :(
17:14:40ashridahnot happy
17:15:59Bgervery bad, but not unrepairable
17:16:14SlasheriMO-Pantsu: yes, i am trying to find the cause
17:20:38Slasherii found it, fixing soon
17:20:53Bgerwhat was it
17:21:06Slasherismall change in pcm buffering
17:22:32Moosand about your FF/RW fixing?
17:22:55Slasherithat will come in same commit soon :)
17:37:49 Quit Ancelot`Kain_Def ()
17:40:10 Quit ep0ch (Read error: 131 (Connection reset by peer))
17:47:19***Saving seen data "./dancer.seen"
17:47:49 Join hardeep [0] (
17:48:30 Join ep0ch [0] (~ep0ch@
17:50:32]RowaN[ah, just got an email.. they DO ship to the UK, for 29 euros
17:50:51elinenbehardeep: the master of queuing, insterting, playlisting, and shuffling...
17:54:55*Bger reads and realises that *full* Unicode support is not so simple
17:55:31 Quit Godeater ("CGI:IRC")
18:01:10 Quit tvelocity ("Leaving")
18:05:44amiconnBger: The hardest part of full unicode support is the rtl/ltr handling
18:08:42Bgeri just got the fact that there are symbols such as "change to rtl" or "change to ltr"
18:08:46amiconnAnd of course we'd need font caching
18:09:47Bgerbut not full
18:10:00Bgerthis idea is not so bad
18:10:47Bgergotta go, nice weekend
18:11:33amiconnWe still need caching, because we can't load e.g. a chinese font completely least not on archos, but it would be a waste on iriver as well
18:13:33elinenbenot for someone who is chinese...
18:13:45elinenbelet's say they are reading a chinese text file?
18:18:27 Join preglow [0] (
18:24:22MO-Pantsuthx Slasheri ;)
18:24:59MO-Pantsu]RowaN[ how much in total for the 60gb?
18:28:10]RowaN[err, whatever it costs + 29 euros (to ship to the UK, godknows by what method)
18:28:29 Quit ]RowaN[ ()
18:35:47 Quit hardeep ("[BX] Does your mommy know you use BitchX?")
18:38:07 Nick Lynx_ is now known as Lynx_awy (
18:38:58HClgood news, my runtime database repaired itself
18:39:09HClwith a bit of luck all that needs to be added to the is file hashing
18:39:45 Quit ashridah ("sleep")
18:39:59godzirraDid rockboy break in the latest builds?
18:40:21godzirraI can't seem to exit it :)
18:42:48OnkelJonashah... my h120 just survived its second fall!
18:42:53OnkelJonasthat thing is pretty sturdy
18:43:37CassandraStop dropping it. :)
18:44:22*HCl pulls his runtime database off his iriver to analyze it with the tool..
18:46:11HClseems to have worked fine, apparently i played most of the songs twice
18:49:05 Join west-acre [0] (
18:53:38HClhrm.. bug in the searchengine..
18:55:05west-acrehey, the new feature to move onto next directory, is that always on ?
18:55:11west-acreor can it be changed ?
18:55:40CassandraI believe there's a config option.
18:56:01 Join mr_qno [0] (
18:56:35mr_qnoHi everyone
18:57:03Cassandra hello
18:59:25mr_qnoSlasheri : thank you for the bug fixing about the gapless. Now it works perfect on Vorbis. With Lame Mp3, it seems that the gapless is "version sensitive". I mean the gapless works perfect with Lame V3.96, but i have Mp3 encoded with Lame V3.92, and i can listen to a little "click" between songs.
19:00:52preglowmr_qno: earlier lame versions didnt encode the gapless info
19:01:15 Nick StrathAFK is now known as Strath (
19:01:30spiralouthow to find out the lame version? of a mp3 file?
19:01:31preglowi'm kinda surprised the gapless code works
19:03:21ep0chspiralout: mr questionman can help determine the version:
19:04:42spiraloutmhh i just looked into cdex and i guess all my mp3´s are 3.92 :-/
19:04:59spiraloutbut thank you
19:05:14preglowyou should upgrade
19:05:20preglowlame's gotten quite a lot better since then
19:05:24SlasheriHmm, i _think_ i fixed the vorbis playback issue. Please test
19:05:27HClblah, why doesn't my searchengine work
19:05:41ep0chSlasheri: cheers, will do
19:05:50mr_qnospiralout : i use cdex too, and my Lame version was 9.92 too :-/ ... That's why i will use foobar now.
19:06:07preglowSlasheri: vorbis playback problems?
19:06:18 Join Chamois [0] (
19:06:36*HCl stares blankly at his computer then slaps it.
19:06:39Slasheripreglow: vorbis was totally broken in the bleeding edge
19:06:46preglowdamn, btw
19:06:55spiraloutmhh EAC would be better but not so easy to use i think
19:06:58preglowi just remembered i do all my rockbox development on my server
19:06:59Slasherinot vorbis itself actually but the pcm buffering which vorbis uses
19:07:05preglowwhich i wont be having any more :/
19:07:14mr_qnopreglow : ok, that's why it doesn't "totally" work on eearlier versions
19:07:37preglowmr_qno: earlier lame versions just make perfetly normal non-gapless mp3s
19:07:41preglowso yes
19:07:44preglowthat's why
19:08:16HClwho's the wise guy
19:08:20mr_qnopreglow : So, now i have to re-encode ;-)
19:08:20HClwho messed with my spelling length
19:08:22HCl >>,
19:08:41HClokay. next time you change that, keep in mind that you have to change that in two locations.
19:08:44preglowtoken length or something?
19:08:53HClalso, it needs to be 32bit aligned, not 16bit
19:08:57preglowit was something about tokens, i believe
19:09:08HClit was to allow strings up to 255 chars
19:09:08 Join hicks [0] (
19:09:26HClits fine to change it to 100, but it has to be done properly
19:09:33HCleg, aligned to 103
19:09:37HCland changed in both the search engine
19:09:39HCland databox
19:09:57preglowthen you should document that
19:10:02preglowcomment or something, it's not apparent
19:10:08mr_qnoi have to go, bye
19:10:13 Quit mr_qno ("CGI:IRC")
19:10:13preglowmr_qno: have fun
19:10:51HCli know
19:11:05HClits because what i really need is a shared header file between two plugins
19:11:09HClbut there's not really a nice way to do that
19:11:31HClno wonder my searchengine choked on it
19:13:15*HCl patpats his kitty
19:15:25preglowi need to make crosscompiler gcc work in my 64 bit linux now
19:15:29preglowthat should be fun
19:15:44preglowbut now i desperately need something to eat
19:16:07HCloh by the way.
19:16:18HClthere's some REALLY deceptive translation when deleting files
19:16:23HClit says press PLAY to delete
19:16:24HClon iriver
19:16:30HClbut it actually wants you to press select
19:18:34HClsearch for songs with a playcount > 0
19:18:35HClhits 10
19:18:43HCland its my moon safari album i've been listening to :3
19:19:28*HCl will test the autorating then commit
19:21:29HClworks :)
19:23:57preglowgotta go do some laundry THEN get something to eat
19:29:51HClhuge commit
19:29:59HClwe have a runtime database now
19:30:12HClprovided you use the java tool to generate yourself a database
19:30:21HClthe perl one works, but since it lacks hashing, the runtime database won't work
19:30:33*HCl goes to watch what he broke
19:30:56HClit shouldn't be too hard to add a hash to the perl version
19:31:09HCli marked the location where the hash should go
19:31:12HClin the source
19:33:41CassandraI do not understand hashing or the tagdb well enough to do this.
19:33:52HCli'll try to add it
19:34:06HClits just a simple crc32 of the first 32kb of each file
19:34:12HClthat should be put in the designated spot
19:35:46ep0chSlasheri: Looks like Vorbis playback is back to normal now :)
19:35:53Slasheriep0ch: good :)
19:36:04HClunfortunately i don't understand perl very well o.o
19:36:06Slasherii am now fixing the seeking
19:36:17ep0chin Vorbis?
19:36:22 Join Yokalosh [0] (
19:36:23Slasheriwith mp3 ;)
19:36:34ep0chdamn! if only ;)
19:37:04Yokaloshdoes anyone know anything about debian linux?
19:37:12ep0chhow difficult would it be to add CUE support for rockbox?
19:37:19Yokaloshthe guys over in #linux and #debian are useless
19:37:33HClYokalosh: i guess..
19:37:33ep0chas it's very different, would it require lots of changes to rockbox?
19:39:03*HCl isn't sure what cue is
19:39:08CassandraCRC32 = sum mod 2^32 of all the bytes, right?
19:39:31HClCassandra: no idea :/ rasher researched the crc thing for me, i just adapted his work
19:39:32amiconnCassandra: nope">
19:39:46 Join bagawk [0] (1000@bagawk.user)
19:39:56HCli just ported that to java and increased the 512bytes to 32kb
19:41:24amiconnI'm pretty sure there is some neat perl module for CRC
19:41:32HClthere is
19:41:38HClbut i thought we didn't use modules
19:41:44CassandraModules how I hate thee.
19:41:53HCl$crc = Archive::Zip::computeCRC32($buffer, $crc);
19:42:01CassandraSomeone already introduced a vorbis parsing module.
19:42:05amiconnThen just paste the module code
19:42:17MO-PantsuI still can't get the flippin next track in wps to work how I want it
19:42:20HClyea, it totally broke the not needing an extensive perl
19:42:46ep0chHCl: basically a CUE file is paired with a very large mp3 file, the CUE file tells the MP3 player where each track starts in the large mp3 file. Used to get around the gap problems with splitting tracks seperatly with mp3
19:42:54HClep0ch: aha.
19:42:57amiconn(afk again)
19:43:09HCldoesn't sound like a hard thing to add, but i wouldn't know.. *looks at Slasheri*
19:43:10thegeekafk cue files are used with a lot of different type of files
19:43:17thegeekI've seen ape's with cue's too
19:43:28thegeekand I think flac's
19:43:36ep0chso basically rockbox would need to parse the CUE file first, then do a directory listing from that and do some seeking etc
19:43:57thegeekimho cue files is a bad solution
19:44:04HCli agree with that.
19:44:11Slasherithat cue thing would require some changing to the wps at least
19:44:14thegeekbut it's a bit like qwerty
19:44:15thegeekit's bad
19:44:19thegeekbut it's still used
19:44:25ep0chit was the only solution for proper gapless mp3 at the time
19:44:30MO-Pantsuwell if you have a lot of single MP3 files etc like I do
19:44:36CassandraWhy the hell is that 10000 times loop around the hashing code?
19:45:07HClCassandra: speed test
19:45:14CassandraOh, right.
19:45:23CassandraIt did seem a bit loony.
19:45:35HClyou should only look at the calccrc32 function
19:45:42HClthe rest isn't very important
19:46:09*HCl broke half of the builds
19:46:11*HCl goes to fix
19:46:19CassandraShouldn't be too hard.
19:46:32HClsomething with the track changed event only being available on iriver :P
19:46:42HClwell i can't help that
19:46:52CassandraBit bashing is a little harder in perl, but...
19:46:59HCli'll just make the runtime database work on iriver only till someone ports that event interface to archos..
19:47:20***Saving seen data "./dancer.seen"
19:47:37CassandraSurely it shouldn't be platform dependent.
19:48:52Slasheribackwards seeking fixed :) now forward seeking..
19:49:21 Quit Yokalosh ("using sirc version 2.211+KSIRC/1.3.11")
19:49:35ep0chand "previous"?
19:49:44Slasheriyes, that works fine
19:49:51HClCassandra: i know, but i don't know how to implement that event on archos, so this is a quick fix to fix the red builds
19:50:06ep0chSlasheri: when did it work?
19:50:19Slasheriep0ch: with my latest code, not yet in cvs :)
19:50:28ep0chok i'll let you off :D
19:50:37 Quit ghostiger (Remote closed the connection)
19:50:56thegeekHCl: how easy would it be to use mp3 tags to generate playcount data for the runtime db?
19:51:06HCli don't understand that question o.o..
19:51:31thegeekI use PLAY_COUNTER in foobar2000
19:51:32HClwhat now? o.o.
19:51:40HClwell..... hrm.
19:51:41thegeekit increases each time I play a song
19:51:50thegeekwould be nice to "import" that into the runtime db
19:51:52HCli just commented out my runtime database generator bit in the java generator
19:52:07HCland it would take quite a bit of code to get it to work
19:52:10HClbut technically, it could
19:52:28thegeekI'm in no hurry
19:52:48thegeekjust as long as it is possible to do it one day
19:52:52HClits a little too much work to do that all for one person who has play_counter :/
19:52:53thegeekI might even do it myself;)
19:53:08HCli can take a quick peek at what changes would be required
19:53:10HClin the java thing
19:53:10thegeekI know;)
19:53:16thegeekdont bother
19:53:17HClit might even be little
19:53:22MO-Pantsudoes anyone know why my wps doesn't work? next track does not show anything on that line
19:53:35MO-Pantsuit just shows an empty line
19:53:51HClokay well actually
19:54:00MO-PantsuI used this %s%?Ia<%Ia - %It|%Fn>
19:54:01HClif you uncomment all the runtime database code in the java generator
19:54:05MO-Pantsubut nothing displays
19:54:23MO-Pantsujust empty screen space
19:54:31HCland simply call getEntry(FileEntry file) on each file entry on which you have a playcount tag
19:54:46HCland then set the playcount in the RundbEntry that you get from that function
19:54:49HClit should work fine
19:54:58CassandraYou know we really shouldn't be using Perl modules.
19:55:05HClCassandra: i agree on that
19:55:10CassandraI don't think their license is GPL compatible.
19:55:22HCli dislike how i can't run the anymore
19:55:24HClon the other hand
19:55:41HClthe's tag parser is way inferior to the java version's tag parser
19:55:54HClcreating odd yPu entries and crap
19:56:38HClthegeek: its actually not *that* much work, the most work would be in reading your playcount tag from the file
19:57:39MO-Pantsumy wps is broken
19:58:20MO-Pantsuanyone know anything about wps?
19:59:03MO-PantsuI created a test.wps and loaded it with 'just' this in it %s%?Ia<%Ia - %It|%Fn>
19:59:12MO-Pantsuand it still will not display next track info
19:59:16MO-Pantsuon anything
20:00:21MO-Pantsuearth calling orson
20:00:25MO-Pantsuearth calling orson
20:00:29MO-Pantsucome in orson
20:00:54thegeekI've seen that too MO-Pantsu
20:01:09thegeekI think it is because rockbox does not know what is next in the "playlist" ?
20:01:16thegeekbecause if I make a playlist
20:01:17thegeekit works
20:01:26MO-Pantsulet me try a playlist
20:02:24MO-Pantsunothing whatsoever on that screen space
20:02:35HClthegeek: oh, one thing, it wouldn't support reading in an old database, so it would delete all the runtime info already present
20:02:51HClunless you add support for reading in an old database, which is a fair amount of work
20:03:24thegeekoh well
20:03:28HClnot hard or impossible, just a lot of work
20:03:28thegeekit's not that important to me;)
20:03:32HClk good
20:03:32HCl :P
20:03:36thegeekjust an idea;)
20:03:41thegeekI'll let it slumber for now
20:04:31CassandraErm, so each time you update the tagdb you lose your current rundb?
20:04:31HCloh poo.
20:04:36*HCl bites the database.c
20:04:39HClCassandra: nah
20:04:47HClCassandra: only with thegeek's feature of importing those tags
20:04:58HClwhich is merely theoretical at this point
20:06:03CassandraI wish someone'd port the Java tool to a real programming language.
20:07:07HCl :P
20:07:36bagawkHCl, every few minutes I am seeing new email... lol
20:07:40HCl :P
20:07:51HClthats what you get for subscribing to the cvs mail
20:09:31CassandraGood work on getting the dynamic database done btw, HCl.
20:09:39CassandraIgnore my bitching, I'm in a mood. :)
20:09:48HClits not yet done, but the basic layout is there and working
20:10:00CassandraWhat's still to come?
20:10:04HClthe thing i like most is that using the hashes, i was able to make it self repairing
20:10:11HClmaking it infinitely easier to maintain
20:10:18 Join belgarath [0] (
20:10:35HClmmm, better playcount handling
20:10:44HCladding options for volume adjustment and rating, and actually using them
20:11:01HClwriting a facility to keep track of time somehow in order to be able to keep track of how long it has been since a song was played
20:11:22HCland ofcourse extending the searchengine to use all that information, and adapting the searchengine to write directly to the internal playlist rather than /search.m3u
20:12:19HCland the other searchengine extensions that are needed to provide rio karma like functionality
20:14:19HClSlasheri: out of sheer curiousity, did you add that event to archos models as well? i just noticed that the red builds were only the simulators, not the actual targets
20:14:32belgarathwhat are the differences between lame 3.92 and 3.96?
20:14:40SlasheriHCl: no i didn't do that
20:14:43HClSlasheri: k
20:14:47 Join andrew [0] (
20:14:49HClthen my disable thing is okay for now
20:14:55HCltill someone adds it to archos
20:15:46HClcats are cute
20:16:26HClthey are!
20:16:47andrewi agree
20:16:48andrewi have 2
20:16:59*HCl has 1
20:17:05HClshared with 3 other people
20:18:21belgarathhmm.. is the new 3.97 beta lame stable?
20:18:33HCli dunno
20:18:36HClask in #lame?
20:18:49HClguess they don't have their channel on freenode
20:19:31ep0chits #mp3encoder?
20:19:38belgarathoh I'll check it out anyway
20:22:14Slasheriseek fix committed
20:22:23CassandraMust be years since LAME had a new version, surely.
20:22:53preglowthey do betas, i think
20:23:58Coldtoasthey. few updaes while I've been at work I see
20:24:08HClruntime database support
20:24:09HCl :)
20:24:11Coldtoastcan you skip to prev track properly now?
20:26:17ep0chyes apparantly
20:28:49 Join bangser [0] (
20:28:52bangserhi evb
20:29:43bangseris there a snoop way to aply a patch to an existing rockbox version?
20:30:04HClnot really, you need to get the source and recompile it
20:30:52bangserhmm, thats what I'm afraid of :)
20:31:46CassandraWhy not just download a new build from the daily builds page?
20:32:00dwihnoif the patch is safe and clean, then there's nothing to worry about.
20:32:29bangserare alle patches included in the daliy builds?
20:32:48CassandraDepends what you mean by "all".
20:33:05 Join Infirit [0] (
20:33:21bagawkbangser, no, all the "open" patches are not in the dailys
20:33:22Coldtoastall patches implemented at the time of compiling are included in the daily build
20:33:54bangseraha, perfect, so actually there is a snoob way to aply a patch :)))
20:33:57ep0chhmm ok maybe previous is still broke with vorbis?
20:34:13Coldtoasthow about mp3?
20:34:21bagawkbangser, If you want to use the patches, you ust compile yourself or beat someone with a stick until they do it :)
20:34:26Coldtoastwait for HDD activity to cease before doing it
20:34:27ep0chmp3 works differently
20:35:20ep0chone tap left to restart the current track, two taps is previous
20:35:32Coldtoastbut does it work, is the question
20:35:38bagawktime to go
20:35:42Coldtoasthas it been fixed since 7hrs ago?
20:35:46 Quit bagawk ("Leaving")
20:36:01ep0chgot "fixed" about 15 mins ago
20:36:06Coldtoastah ood good
20:36:13ColdtoastI'll dig my player out in a bit
20:36:22ep0chand Slasheri fixed vorbis playback too
20:36:30ColdtoastI saw vorbis listed
20:36:36Coldtoastbut not the track back fix
20:36:51Coldtoastoh hey. heh
20:36:56Coldtoastthe last bleedign edge fixed it
20:37:15Coldtoastoh. that's seeking
20:37:35Coldtoastis skip track included in the Rockbox definition of "seeking"?
20:38:40Slasheriyes, that backward skipping requires seeking
20:39:21Coldtoastdamnit! I want to buy these earphones from a guy in Germany on ebay but no paypal
20:39:24Coldtoastthat sucks
20:39:30bill20r3Mmm earphones
20:39:38ep0chwhich ones?
20:40:13Coldtoastsee, my theory is if you're going to list something as "Worldwide" you have to use a service like Paypal or Bidpay. not "Benk Deposit Express"
20:40:23ColdtoastSony EX-70
20:41:33ep0chheh i've gone through so many EX-70/71s...
20:41:39bill20r3I got some er6i's when they went on sale, I love them
20:41:48Coldtoastare they good tho ep0ch?
20:41:59ep0chi just got some ER6i's yesterday to replace the EX-71s :)
20:42:12Coldtoasthow do they compare?
20:42:18ep0chyeah i like em, i just only manage to make em last about 6 months :s
20:42:43ep0chEX-71s have more bass
20:42:49ColdtoastI like bass
20:43:04ColdtoastI currently use these Philips HD250s
20:43:13Coldtoastnot in-ear tho so I don't much like em
20:44:35 Quit andrew ("using sirc version 2.211+KSIRC/1.3.11")
20:44:36 Quit Infirit ("Leaving")
20:44:38ep0chbill20r3: you using the foam tips with your ER6is?
20:44:43 Join Infirit [0] (
20:44:52bill20r3no, the flanged rubber tips
20:45:12bill20r3my only complaint is that they can be a pain to get seated right
20:45:23ep0chme too, might try the foam ones... but won't they just get covered in wax after a while?
20:45:41ep0ch(says alot about my hygiene ;) )
20:46:02bill20r3I bought a box of alcohol swabs that I keep in my desk
20:46:08CassandraWow - other people with ER6i's. ;)
20:46:12bill20r3before I did that they would get pretty nasty
20:46:21bangserpercet: patch 1110332 is really in the daily and it works: now is can charge my jb via usb, while listenig, perfect,
20:46:26ep0chyou have them too Cassandra?
20:46:27CassandraI don't seem to get that much wax buildup.
20:46:53ep0chwhat's your opinion? personally i don't they're worth the RRP
20:47:07CassandraNoise reduction is better with the foam too.
20:47:30CassandraWeak bass response, but can't be beat for purity of sound.
20:47:46CassandraGuitar sounds totally amazing.
20:47:59Cassandra(Got mine of eBay, so didn't pay the RRP. :) )
20:48:05ep0chyeha me too :)
20:48:18CassandraI *heart* my ER6i's.
20:48:32ep0chi want to compare with some Shures now....
20:48:41Cassandra(I'd not tried the Shure E2c's which are really the only competition..)
20:49:01Coldtoastor the E3Cs
20:49:13ColdtoastI was thinking of going for E3c's
20:49:15CassandraThey new?
20:49:22Coldtoast~$200 AUD tho
20:49:25ep0chER6i can be a little hard to get out of my ears, maybe i'm just pushing them too far in?
20:49:37ColdtoastI think they've been around for a little while Cassandra
20:49:56CassandraThey don't need to go right in. Just until the actual earpiece is submerged.
20:50:34ep0chtake a photo of your ear and show me ;)
20:51:23CassandraI just realised I got confused.
20:51:38CassandraI have ER-4p's which are the higher specced model.
20:51:41 Join ShockerEngr [0] (
20:52:05Coldtoastphew! now THOSE are expensive!
20:52:24Coldtoast$330US for earphones!
20:52:24Cassandra*nods* I bought them to cheer myself up after my dad died.
20:52:38CassandraThey cost me 165 UK pounds, I think.
20:52:42preglowso there are actually earbuds that sound good?
20:52:46ep0ch$330 on ebay
20:52:55Coldtoastfound em for $233 ep0ch
20:53:06Cassandrapreglow: Oh yes.
20:53:09ColdtoastER-4 tho, not ER-4p
20:53:12ep0chstill, that's a lot of money!
20:53:29CassandraShure e5c's or Etymotic ER-4Ps are considered the best.
20:53:39CassandraThey are costly though.
20:53:44preglowi've more or less always sworn to earphones, but i'd love something a bit smaller
20:53:54preglowwell, good things often are
20:54:03Coldtoastthey'd BETTER come with warranty that covers wires being pulled out of the buds
20:54:10preglowoh yes
20:54:19preglowcan't count the times i've managed that
20:54:37Coldtoastyeah. that's one reason I only buy these Philips HD250s
20:54:42CassandraI suspect Etymotic'd repair them.
20:54:49CassandraBut I've never had a problem.
20:54:51ep0chand warranty that covers headphones being run over by a chair :(
20:55:00Coldtoastman. for that kind of cash, I'd want full replacement
20:55:15CassandraThey are sturdier than standard headphones.
20:55:25ColdtoastI really really have wired earphones
20:55:25Cassandra(although not much - there's a limit to what you can do.)
20:55:35ColdtoastI CONSTANTLY pull them out of my ears
20:55:42Coldtoastkeep snagging them on things
20:55:49*HCl looks around for a victim
20:55:57Coldtoastand it just pisses me off to no end!
20:55:58HClwhere's ashridah when you need him
20:56:04*preglow runs to the laundry room to hide
20:56:36HCli want someone who can use the java songdb and an iriver to give the the runtime database a test run..
20:56:44HCland has*
20:56:52ep0chtime for food
20:57:31*HCl prods amiconn
20:57:34CassandraThe Etymotics clip to you, so that's not so much of a problem.
20:57:39HClamiconn: you know a lot about archos, right?
20:57:51CassandraAlso means you can take them out to talk to people without dangling them all over the place.
20:58:18CassandraHCl - hold on, I'll try to give it a go.
20:58:39 Quit belgarath ("CGI:IRC")
20:58:46HCli'm not sure if i'm asking much, but i need the track change event thing implemented on archos platforms..
20:58:50 Join TCK- [0] (
20:59:09 Quit TCK (Read error: 131 (Connection reset by peer))
21:00:09 Join Febs [0] (
21:00:26 Join Ancelot`Kain_Def [0] (
21:01:03 Join belgarath [0] (
21:01:45belgarathassuming that the database file is created, do you need to be in view id3 database for the playcounts to be remembered?
21:02:04belgarathor is playing by the file tree fine
21:03:31CassandraBah, installing to the Jukebox under Linux is such a pain.
21:03:41FebsColdtoast: where are you located?
21:03:56 Quit belgarath (Client Quit)
21:03:58HClbelgarath: either will work
21:04:07HClhe shouldn't have left, heh.
21:04:18 Nick Ancelot`Kain_Def is now known as Ancelot (
21:05:15CassandraGreat. Now it's refusing to run java as root. See why I hate java?
21:05:27Ancelotwe all do
21:05:35Cassandra(Not for security reasons. Because it won't work.)
21:05:35HClsorry o.o
21:05:48Ancelotwe know
21:05:49HCljava is mostly nice for the programmer :p
21:06:02CassandraNever mind. I'll set up the automounting for the Jukebox properly and try again.
21:06:09Ancelota awfull for the processor
21:06:32HClthe good news is that the self repairing feature means you won't need the java version as long as hashing is added to the perl version
21:06:36HClgod i hate lag x.x;;;;
21:06:40HCli'll be back when the lag ends :/
21:06:42HCloh wait, thats now o.o
21:06:49*HCl bites the lagmonster
21:07:33HClhashes are required to stay the same though, and preferably using the same hashing routine so they're compatible
21:08:04*HCl goes to watch his flatmate play san andreas..
21:09:15preglowsan andreas
21:09:16preglowforgot about that
21:10:11CassandraArgh - what's the syntax to generate the db from Java?
21:10:20 Join tiegs [0] (
21:10:51 Quit tiegs (Client Quit)
21:10:54preglownow for the dish washing
21:12:08HClCassandra: java -jar Songdb.jar
21:13:15Cassandra[christi@slappy mp3]$ java -jar /home/christi/work/rockbox-devel/bin/SongDB.jar -path /mnt/mp3 -strip /mnt/mp3
21:13:15Cassandrayou can't specify more than one path!
21:13:15Cassandra[christi@slappy mp3]$
21:13:27HCl-path option is obsolete
21:13:29HCljust leave it out
21:13:44HClrun it without options to see its usage
21:15:03CassandraOK. This is going to take a while. Linux box is USB 1.1 only.
21:15:16HClyea, especially the hashing requires quite some bandwidth :/
21:15:29Ancelotgo linux!
21:15:39HClthe −−showduplicates option shows duplicate hashes, if you get the feeling its messing up
21:16:30HClso far all duplicates i got were real duplicates
21:17:22Cassandra*boggles slightly at the weird permissions on the tagdb.
21:17:57CassandraMust've screwed up my umask.
21:18:25CassandraHmm. My tabdb is 68 bytes long. That is not promising.
21:19:05HClthats just the header
21:19:51 Part Febs
21:19:55HClcould you maybe redirect output to a log and send it to me?
21:20:19CassandraSure. Just want to make sure it's not the permissions problem.
21:21:23HClit sounds as if it failed to open / hash the files
21:21:46HClsometimes linux mounts it with odd permissions
21:21:57HClalso java can have this thing about checking the mountpoint only and not the actual mounted disk
21:22:00HCli think
21:22:07HClcheck the permissions on the mountpoint as well..
21:22:17CassandraF*** bloody automounter has screwed itself and now I can't unmount the drive.
21:22:28HCli know someone ran it in linux successfully... don't remember who though.
21:23:45 Part bangser
21:26:00 Quit ep0ch ("Trillian (")
21:26:19CassandraRight. Running again.
21:27:28 Quit RotAtoR ()
21:27:40 Join ep0ch [0] (~ep0ch@
21:28:11CassandraSeems to be doing something a bit more sensible now.
21:29:44ghode|afkhmm do i run the songdb command from .roxkbox or root of my iriver?
21:30:26ghode|afkah nm
21:32:53CassandraSo what do I do with this once it's finished. Just play some tunes, or do I need to turn something on?
21:33:20HClmmm, toss the database files it generated into .rockbox
21:33:38HClalso, its a good idea to check the longest filename bit it spits out
21:33:46HClto see whether the filename was generated correctly
21:33:52HClcould you maybe paste it to me?
21:34:13HClfilenames are very important, and i should probably add more checks to see whether they're correct..
21:34:21HClthey should start with a \ and contain the absolute path to the file
21:34:40HClaside from that, it should work
21:35:36CassandraOh damn. Just ran it in the root directory, where it fell over at the end.
21:35:49HClit should be fine if you run it in the root directory
21:35:54HClas long as it generated proper filenames
21:36:11CassandraLinux root.
21:36:15CassandraIt doesn't have permission to write files there.
21:36:29ghode|afkHCl: are all the memory/time related karma features impossible on the h1x0?
21:36:35 Join einhirn [0] (
21:36:47HClghode|afk: not all of them...
21:36:55HClghode|afk: i'm planning some relative timer/counter
21:36:58ghode|afk4676 songs
21:37:14HClthat keeps track of the total amount of seconds during which the player was turned on
21:37:19HClgiving at least some notion of passage of time
21:37:39ghode|afkwhen do you think you will get ratings implemented?
21:38:46 Quit Chamois (Read error: 110 (Connection timed out))
21:38:57ghode|afkok so longest file name was 72chars, will that affect anything?
21:40:20HClits mostly statistics
21:40:25HClratings should be really easy to add
21:40:33HCljust a matter of adding it to the wps menu
21:40:42HCli just have no knowledge of the wps menu yet
21:41:31ghode|afkok i've got the db in the .rockbox folder, what do i do now? where do i go to see if its working?
21:43:09 Quit Infirit (Remote closed the connection)
21:43:23HClbest is to boot the player, check if id3 browsing works
21:43:25HClonce that works
21:43:29HClyou can try playing some tracks
21:43:45HCland entering the search query "playcount > 0" in databox, then "playing" the searchfile you generate with that
21:43:53HCli know, the interface will be improved later.
21:44:41 Join Infirit [0] (
21:44:52CassandraDoes it count complete tracks or partials as well?
21:46:02HClpartials, everything, in fact
21:46:05HClit needs to be improved
21:46:11HCllike after playing 75% of the song it should be counted
21:46:19HClat the moment it even counts it if you skip it immediately
21:46:30HClits still very crude
21:46:52 Join Yokalosh [0] (
21:47:15Yokaloshcan anyone help with debian?
21:47:24***Saving seen data "./dancer.seen"
21:47:25CassandraI get 2 "<no album tag>" entries.
21:47:50HClthats possible, you can use the −−dirisalbumname to replace them with the directory name instead
21:48:14HCli don't quite remember who, but there was someone who rather had <no album tag> than the directory name
21:48:32HClmaybe i should make it default and disableable
21:49:00CassandraSounds like a plan to me.
21:49:20CassandraHmm. Tried to play a track and the disk is in spin city.
21:49:40HClhm, what does the remote say? :/
21:49:45Coldtoastsomebody's offered to buy my xbox
21:50:00HCli wouldn't sell my xbox
21:50:05HCli use it daily
21:50:12Coldtoastit's as modded as you can get (almost)
21:50:12HClto watch futurama and pretty much everything
21:50:20HClwhat are its specs? :P
21:50:20ColdtoastI rarely use mine
21:50:39ColdtoastXenium modchip, little 60GB HDD
21:50:51HCli have a limited crystal edition, tsop modded xbox with an 250gb harddisk and a tweaked dvdrom :)
21:50:58Coldtoastbut I've hardware modded it so the RF receiver is integrated
21:51:03HClwith an usb->xbox converter
21:51:04HClnice :)
21:51:27HCli looked into that
21:51:35Coldtoastthere's my xbox
21:51:37HClbut it required permanently claiming an xbox slot
21:51:49Coldtoastnah. you can use a relay
21:51:49HCl :p
21:51:55HClyea, true
21:52:00Coldtoastbut I didn't bother
21:52:04HCllet me try to find a pic of mine
21:52:18Coldtoastactually, I'd better let the guy know
21:52:24CassandraOK, there's a bug in playback from the "All Tracks" selection.
21:52:25Coldtoasthe'd only have 2 controller ports
21:52:49preglowcrap game :/
21:52:57Coldtoastfunky old TV
21:53:00HCl :P
21:53:05HClyea :P
21:53:08HClstudent flat
21:53:17HCli'm still looking around for a tft screen
21:53:20Coldtoastheh. looks like it
21:53:44HClwe have our stereo set plugged into it with that plug at the bottom right :p
21:54:11CassandraHCl - can you use the song has been played criterion from
21:54:34CassandraThat'd make something else I've wanted easier in the long run.
21:55:31HClCassandra: could you elaborate a bit more how it works...?
21:56:09HCli think my english is degrading O.o. i need to work on that x.x
21:58:34 Quit Yokalosh ("using sirc version 2.211+KSIRC/1.3.11")
21:59:03CassandraThe search playlist is supposed to appear in the root, right?
21:59:55CassandraAudioscrobbler, or the "yes, a track has really been played" algorithm.
22:01:02CassandraEach song should be posted to the server when it is 50% or 240 seconds complete, whichever comes first.
22:01:08CassandraThat's the important bit.
22:01:27HCloh that
22:01:32HClyes, i'm planning that.
22:01:45HClbut at the moment we're still debating how to do it properly
22:01:54HClso i just went with the crude approach for testing
22:02:14CassandraBug in the search plugin: The root isn't refreshed if you run it from the root directory.
22:03:38 Join skel_ [0] (
22:03:57CassandraBut the rtdb seems to work.
22:04:02HCl :)
22:04:10Cassandra(although I don't remember playing "A Smile and a Whisper"
22:04:24HClit might be a hash duplicate
22:04:49HClyou could check by adding −−showduplicates while running songdb, and redirecting the output to a file
22:05:05HClif we have a hash duplicate, it means i have to increase the amount of bytes used to hash...
22:05:40HClhash duplicates will in general give properties to songs, while those properties actually belong to a different song
22:07:11CassandraYou've an off-by-one error somewhere.
22:07:31CassandraThe playlist is lagging one track behind the tracks I'm playing.
22:07:51HClwhat do you mean..?
22:07:54CassandraSo I play a,b,c do search and get a, b.
22:08:08CassandraPlay d next, do search, get a b c
22:08:10HClthe playcount of the previous file
22:08:15HClgets increased when the next track starts
22:08:39CassandraSurely should get increased when that track stops.
22:08:43HClnot at the beginning.. since its supposed to happen after a song has been played sufficiently
22:08:51HClits part of the crudeness at the moment
22:09:05HClits mostly proof of concept and foundation-laying code :)
22:09:13HClits still pretty crude
22:10:28*HCl goes to make a todolist..
22:10:31CassandraHere's another thought - how about using the Musicbrainz tag of a track to ID it rather than CRC32?
22:10:44HCla what?
22:11:00HCli really prefer crc over tags since crc is much much more independant
22:11:17HClbut its not hard to adapt the songdb to use a different hash by option, i guess.
22:12:22CassandraThing about MB tags is they're robust over filename changes and encoding differences.
22:12:54CassandraAlthough to be honest you really want a unique identifier.
22:13:01CassandraWould MD5sums work better?
22:13:16HCli guess the encoding difference is something that crc does not cover
22:13:21HClmd5 sums are slower
22:13:25HClthe only reason we get duplicates
22:13:30HClis because we only hash the first 32kb
22:13:32HClfor speed
22:13:47HCli could take rasher's suggestion and hash the first 16kb, and the last 16kb
22:13:52HClsee if thats a better way to hash quickly
22:14:07HClif we want no duplicates, we should just hash the entire files with crc32
22:14:23CassandraThat might be the best idea.
22:14:24HClbut that would be significantly slower
22:14:30HCland i mean really slow
22:14:36HClreading all songs in completely
22:14:41HClbefore being able to spit out a database
22:14:43CassandraWell, maybe you need not to have the db regenerated from scratch each time.
22:14:51HClat the moment, we do
22:14:57HClhashing takes most of the time anyways
22:15:20CassandraI think clashes are really bad.
22:15:24HClthey are
22:15:36HCli'll try to implement rashers suggestion
22:15:53HClwill you be able to test whether it will still create clashes when i do?
22:15:53CassandraIt should probably be a design goal to avoid them. I'd say a more important one than speed.
22:16:10HClthe idea was that by hashing only part of the file, we would be able to hash on target
22:16:17ZagorHCl: if you take separate chunks, don't use the last (often silen) chunk. take one in the middle.
22:16:22 Quit Moos (Read error: 131 (Connection reset by peer))
22:16:23HClwith the current design we don't need to hash on target unless the file is not in the tagdatabase
22:16:25preglowSlasheri: does resampling work for other formats?
22:16:31HClZagor: hrm.
22:16:52HClwell, its all not even meant to be production quality yet, just proof-of-concept quality
22:16:57CassandraLooks like the code is working well though.
22:17:40HCli really need to take a break soon
22:19:31CassandraBreak. Break. This is not worth killing yourself over. :)
22:20:16HCl :)
22:20:31 Join Moos [0] (
22:22:50 Join webguest44 [0] (
22:23:34*HCl goes to play with his cat and watch some tv, bbl
22:25:08 Quit Moos (Read error: 131 (Connection reset by peer))
22:25:22 Part webguest44
22:28:42 Join Moos [0] (
22:33:48MoosHi all: anyone who use db.jar can help please, i'm not a java user?
22:34:48Moosi just downlaod the 1.4.2 version
22:35:11Moosand when i want to open db.jar file, don't work
22:35:15Cassandrajava -jar SongDB.jar −−strip /path/to/rockbox/root /path/to/rockbox/root
22:35:26Mooshi cassandra
22:37:08 Quit Zagor ("Client exiting")
22:39:34 Join pill [0] (
22:40:03pillhuge props for rockbox on iriver hardware
22:40:07 Quit ep0ch ("Trillian (")
22:40:25pilli so much wish to contribute
22:41:33Bagdergo ahead!
22:42:21pillbut i cant code that good
22:42:58pilland i could try and correct the french translation, but i think Tang would hate me then
22:43:21Bagderwhy would he?
22:43:32CassandraIf you made it better, I'm sure he'd be pleased.
22:43:49Bagderwe all correct each others work all the time
22:44:02CassandraAnyone know what kind of a cable I need to connect the optical out on my H140 to a SB Extigy.
22:44:14HClhmmm, markun might know that
22:44:25Cassandra(which has those strange square plugs in the optical in socket.)
22:44:40CassandraOi. HCl, you're supposed to be resting. :)
22:44:44HCli know :/
22:44:52HClbut Moos needed help
22:45:18HCli'm probably gonna go sleep once this tv show finishes
22:45:22 Join Lmcmann [0] (
22:45:41LmcmannAfternoon all, I have a quick question.
22:45:57LmcmannDoes anyone know what happened to the RomBox builds?
22:46:23Bagderwhat about them?
22:46:26CassandraRockbox got too big to fit in ROM on the V2/FM.
22:46:55LmcmannThat's a bummer
22:47:17LmcmannHave to switch it back to normal to keep up with the latest changes.
22:47:23Bagdermy v1 one is too big too...
22:47:32Bagder"Output is 1848 bytes larger than max (192496)"
22:47:34LmcmannThanks for the info. Anyone know the last day one was built?
22:47:51CassandraNo idea, sorry.
22:48:12CassandraJust stick with Rockbox.ucl. There's not that much performance difference.
22:49:34 Join Stryke` [0] (
22:50:10LmcmannNo worries. Thanks for the help.
22:51:43LmcmannAmazing work from everyone. Thanks for the effort.
22:51:56LmcmannAnd starting on the 300, I'm eagerly awaiting that one <g>
22:52:12 Quit Lmcmann ("CGI:IRC")
22:52:53ShockerEngrdoes the rtdb update when running in file mode? or is just while running in db mode
22:52:53MO-Pantsudoes anyone know why my wps next track info only works occasionally? I left an album running for over an hour now and I saw it only display the next track a couple of times. The rest of the time the area of the screen that is supposed to show it is blank.
22:53:46ShockerEngrMO-Pantsu: it only shows the next track if it's buffered in memory −− at least from what i've seen
22:54:40 Quit TCK- (Read error: 110 (Connection timed out))
22:54:59MO-Pantsuthen why is it not buffering it into memory?
22:55:36Bagderif you want to know, the code is there to read
22:56:06BagderI guess the problem is that you have a scrolling line and that updates don't work in scrolling lines
22:56:14ShockerEngrthat ones beyond what i know, but you can check the audio buffer (it's under the debug menu) and see if it is buffer all the tracks
22:56:33HClShockerEngr: it updates as long as you have a tagdatabase generated
22:57:06*HCl stops commenting and goes to relax again
22:57:14MO-PantsuOK I tried someone elses wps and it works but I don't want to do it like they do it. I want it to show 'artist - title' and scroll it
22:57:33MO-Pantsubut the command does not seem to work properly
22:57:38Bagderthen read what I said
22:58:00MO-PantsuI have no idea about code. who is it who did the wps code?
22:58:11BagderI wrote something more
22:58:17BagderI guess the problem is that you have a scrolling line and that updates don't work in scrolling lines
22:58:28Bagderthis is a known flaw
22:58:34Bagdersince ages
22:58:54MO-PantsuOK let me test without the scroll
22:58:56BagderLOADS of people have written the wps code
22:58:57 Quit Bger (Read error: 113 (No route to host))
22:59:55crwldoes the updating work with sublines?
23:00:15crwlok, thought so too...
23:00:18preglowgotta goes
23:00:18 Quit preglow ("leaving")
23:00:35MO-Pantsuyep that is it
23:00:45MO-Pantsuscroller on next track does not work correctly
23:00:57Bagderscoller on _anything_ that updates don't work
23:01:00HCloh. i thought my tv show was finished, but apparently it hasn't o.o.
23:08:31ColdtoastMO-Pantsu: what's your code fo rthat line?
23:16:12amiconnHCl: What should the track change event contain, and what data might be needed apart from the event itself?
23:16:31HClamiconn: i prefer to get the same data that slasheri passes me
23:16:41MO-Pantsu%s%?Ia<%Ia - %It|%Fn>
23:16:45HClbut whats most important at the moment is the full path of the new song
23:16:48MO-Pantsuworks without the scroller
23:16:56HClin the future, i might want elapsed and total playtime
23:17:10MO-Pantsudoes not work with scroller
23:17:20HClwhat i mostly use is the id3 bit of the structure that slasheri passes me
23:17:27HCli can do fine with just that
23:17:31HCli hope thats available on archos
23:17:43amiconnIs that structure identical to the archos one?
23:17:47HCli don't know..
23:17:51amiconnid3 data is available on archos
23:17:55Bagderimo, it should be
23:17:57amiconn(in the mpeg thread)
23:18:04crwlhmm, i get "error accessing playlist control file" on boot with the current bleeding edge
23:18:07Bagderat least a subset
23:18:20Coldtoastlet me just alter my own WPS and see what happens
23:18:31Coldtoast%s%?Ia<%Ia|unknown artist> - %?It<%It|%Fn> is my line
23:18:41amiconnHCl: It seems your commit does not include this code. I believe it was added earlier?
23:18:57HClwhich code?
23:19:17MO-PantsuI always get unknown artist with that line for everything
23:19:24amiconnThe code that sends the event, and provides the structure
23:19:36HClyes, it was added earlier
23:19:38HClby slasheri
23:20:19*HCl doesn't have much energy left, so can't dig into code till tomorrow..
23:20:44HClits times like these when i wish i had a tv above my bed
23:21:28*amiconn could do almost completely without tv
23:22:14CassandraComputers play DVDs. What would you need a TV for?
23:23:14Slasheripops fixed
23:23:17HClcause i like a tv screen
23:23:27HCli'm planning to buy an tft tv
23:23:34HClto hook my computer, tv, and xbox up to
23:23:50ColdtoastMO-Pantsu: I just dunno what to tell you dude. mine works perfectly
23:23:50amiconnI think at least some of the tagdb/runtimedb error messages should be localised
23:24:25HClmost of the tagdb code in database.c is not used yet
23:25:44amiconnDo I understand correctly that the playback code calls rundb_track_changed(), and that you need ti->id3.path only?
23:25:52ColdtoastMO-Pantsu: can you send me your .wps?
23:27:19 Join tiegs [0] (
23:27:27HClamiconn: at the moment, yes. but i'd rather have the entire id3 structure
23:27:41HCli don't know what else the ti structure contains, and i'm too tired to check
23:28:21 Quit Cassandra ("If I were actually witty, this quitline would be funny.")
23:28:43amiconnYes. I think the id3 structure is the same as the one on the archos, only that the super-structure is different
23:29:26HClit should be fine if you just pass that, and remove the ti->
23:29:30amiconnapps/playback.h defines struct track_info which contains a struct mp3entry id3 ...
23:30:01crwlhey, is the PCM recording in the debug menu considered usable?
23:30:11amiconnand mpeg.c defines struct trackdata which also contains struct mp3entry id3
23:30:34Slasheribut mpeg.c is not used on software codec platforms at all
23:30:48Slasherithere has to be the id3 structure defined
23:30:50amiconnI have one problem with it though: This means a callback from firmware layer to application layer, which is bad
23:31:19amiconnSlasheri: struct mp3entry is defined in id3.h, hence available on all platforms
23:31:47amiconnHCl only needs that sub-structure, so it should be easy to adapt for archos
23:31:59amiconn(if we want to live with that ugly callback)
23:32:00Slasheriamiconn: yep, but we will cache the metadata.. So we need 10x those structures
23:32:18Bagderso does the archos
23:32:23amiconnWhat do you mean?
23:32:29Bagderit keeps N structs in memory
23:32:39amiconnThe archos stores up to 16 track info structures
23:32:42Slasheriamiconn: That track_info is allocated MAX_TRACK times
23:33:04amiconnYes, and so is struct trackdata
23:33:08amiconnon archos
23:33:15Coldtoastdo any of you have mono MP3s?
23:33:23amiconnstatic struct trackdata trackdata[MAX_TRACK_ENTRIES];
23:33:27SlasheriColdtoast: mono is not yet supported by dsp
23:33:39Slasheriamiconn: but that is defined in mpeg.c?
23:33:39Coldtoastthat's why I only get 1 channel?
23:33:40 Quit tiegs ("CGI:IRC (EOF)")
23:33:47amiconnSlasheri: yup.
23:33:49SlasheriColdtoast: yes
23:33:53Slasheriamiconn: ok, so we can't use it
23:33:59amiconnBut that shouldn't be a problem
23:34:04Coldtoastok. would that also be why the one channel I do get continually clicks?
23:34:11amiconnHCl only needs the pointer to the sub-structure
23:34:36Slasherii think i will take a look to the voice ui next
23:34:51Slashericodec/track buffering is now almost fully working
23:35:12SlasheriColdtoast: yep
23:35:27Slasherii will fix that soon (simple change to the dsp)
23:35:31amiconnFor the voice UI, you should talk to [IDC]Dragon if he pops in
23:35:38Slasheriok :)
23:35:44amiconnHe wrote the current implementation
23:35:48amiconn(most of it)
23:35:56 Quit Stryke` (Read error: 145 (Connection timed out))
23:35:59MO-Pantsuhold on
23:36:40MO-Pantsubung your own line in where mine it before the %pm
23:36:57MO-Pantsusee if it works
23:37:10amiconnBagder: What do you think about that callback?
23:37:15Slashericu tomorrow, nights :) ->
23:37:50BagderI agree about the badness in firmware => apps callbacks
23:38:24Bagderbut I don't have any good alternative right now
23:38:55amiconnHmm. At least the callback is set by an 'enabler' function, it isn't fixed
23:39:42Bagderaha, well then its not that bad
23:40:18amiconnSo basically mpeg.c has to implement that function, and actually call the callback function
23:40:45amiconn...and the parameter should be changed from a struct track_info * to a struct mp3entry *
23:42:12MO-Pantsustill showing 'unknown artist -' here Coldtoast
23:43:12ColdtoastMO-Pantsu: well
23:43:26Coldtoastum... don't want to complicate things here...
23:43:31Coldtoastbut it works on mine
23:43:35Coldtoastwhat build do you have?
23:43:45Coldtoastbleedign edge?
23:44:16amiconnBagder: Sounds like it shouldn't be hard, especially as there already *is* an event MPEG_TRACK_CHANGE in the mpeg thread
23:44:32amiconnThe benefits of modular programming... :)
23:44:54Bagderhow dows the wps detect track change?
23:45:09MO-PantsuI am going to rename my rockbox folder and copy over and try wps again
23:46:14BagderI don't see why this db stuff needs this more than the wps does
23:46:29amiconnBagder: In the mpeg thread there is a function track_change() which is called from some other places too, not only on the track change event if you skip a track etc
23:46:54Coldtoastyou're getting Unknown ARtist with my line I take it?
23:47:02Coldtoastthe wps you sent me works
23:47:27***Saving seen data "./dancer.seen"
23:47:37Coldtoastupdates properly on track change etc
23:47:43amiconnSlasheri: In what cases does playback.c call the track change callback?
23:48:53Coldtoastwith the next track info, does that give the info for the next track that's buffered?
23:49:04MO-Pantsuwell even after renaming rockbox and copying it over again and reloading that wps zip
23:49:34MO-Pantsusend me back the wps with what you added that you say works
23:49:43amiconnColdtoast: Yes, so the info is only available when the next track is already buffered
23:50:09Coldtoastamiconn: why not just read the next entry int he playlist tho? why even have anythign to do with the buffer?
23:50:14Coldtoastjust out of interest
23:50:27Bagderit is a complicated issue
23:50:31amiconnThe playlist does only contain the filename and path
23:50:41Bagderand in fact, not even that in memory
23:51:08Coldtoastok. I see
23:51:22amiconnArtist, title, album etc are read from the tag(s), for which the file must be read
23:51:37amiconn...which we don't want to happen in advance
23:51:40MO-Pantsumy tags are fine
23:51:54amiconn...because that (1) needs more power and (2) needs more RAM
23:52:01MO-Pantsubut your line does not seem to work
23:52:07MO-Pantsuon mine at least
23:53:06amiconnBagder: I know. Otherwise rockbox wouldn't be able playlists that large, and also not with such high speed
23:53:22MO-PantsuI just get 'unknown artist -' for everything no mattyer how long I leave it playing. if it is working for you then I can't figure out why
23:53:29 Quit west-acre (Read error: 110 (Connection timed out))
23:53:45 Quit ze (Read error: 104 (Connection reset by peer))
23:54:02ColdtoastI reall ymade no changes MO-Pantsu
23:54:23MO-Pantsuoh but I thought you added your line
23:54:41MO-Pantsu%s%?Ia<%Ia|unknown artist> - %?It<%It|%Fn>
23:54:41Coldtoastnah. wanted to test yours
23:54:48MO-Pantsuthat does not work on mine
23:54:58BagderMO-Pantsu: you are playing mp3, right?
23:55:03Coldtoasteven if you let the track play for a while?
23:55:20MO-Pantsuyes I leave it for ages
23:55:27MO-Pantsualways get unknown artist
23:55:28Coldtoasthow long's ages?
23:55:35MO-Pantsumy non scroller works of course
23:55:47MO-Pantsuseveral minutes
23:55:54MO-Pantsucan leave it for an hour and it won't show
23:56:03amiconnHmm. The runtimedb code adds ~2.5KB to the archos builds, even though a commit says it's disabled (??)
23:56:28Bagderrombox is no more
23:56:38Bagderon recv1
23:56:57amiconnArgh!!!! :((
23:57:29HClonly the enabling code was disabled
23:57:32Bagder"Output is 1848 bytes larger than max (192496)"
23:57:40HClthe supporting functions are still compiled
23:57:57amiconnBagder: For me it says "Output is 1744 bytes larger than max (192496)"
23:58:05amiconnDifferent GCC?
23:58:27MO-PantsuColdtoast what player have you got?
23:58:34Bagder3.3.1 here
23:58:38amiconnBagder: This is 3.3.5

Previous day | Next day