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 2002-09-23

00:16:37 Nick TotMacher is now known as tot|n8 (
00:34:59***Saving seen data "./dancer.seen"
00:46:51 Join Jet8810 [0] (
00:50:55 Join langhaarrocker [0] (
02:24:37 Join Foxinou [0] (
02:25:04Foxinouhi guys
02:26:25 Part Foxinou
02:29:02 Quit Jet8810 ("Client Exiting")
02:35:01***Saving seen data "./dancer.seen"
02:36:35 Quit xtac ("ircN 7.27 + 7.0 for mIRC (2002/01/10 00.00)")
02:43:36 Quit langhaarrocker (Read error: 110 (Connection timed out))
02:50:14 Nick seb-away is now known as _seb_ (
04:17:48 Join PsycoXul_ [0] (
04:17:48 Quit PsycoXul (Read error: 104 (Connection reset by peer))
04:17:56 Nick PsycoXul_ is now known as PsycoXul (
04:27:12 Join elinenbe [0] (
04:35:04***Saving seen data "./dancer.seen"
05:06:12 Nick _seb_ is now known as seb-sleeo (
05:06:21 Nick seb-sleeo is now known as seb-sleep (
06:09:45 Quit Hes (
06:09:45 Quit Hadaka (
06:09:45 Quit tot|n8 (
06:09:45 Quit seb-sleep (
06:09:45 Quit datazone (
06:09:45 Quit elinenbe (
06:09:45 Quit PsycoXul (
06:09:45 Quit mecraw (
06:09:45 Quit Synthe (
06:09:45 Quit mem (
06:09:45 Quit webmind (
06:09:45 Quit Schnueff (
06:09:45 Quit DownKaos (
06:12:46NJoinelinenbe [0] (
06:12:46NJoinPsycoXul [0] (
06:12:46NJointot|n8 [0] (
06:12:46NJoinDownKaos [0] (
06:12:46NJoinHes [0] (
06:12:46NJoinseb-sleep [0] (
06:12:46NJoinmecraw [0] (~mecraw@
06:12:46NJoinSynthe [0] (
06:12:46NJoindatazone [0] ([B5hF+FdPv@
06:12:46NJoinmem [0] (
06:12:46NJoinHadaka [0] (
06:12:46NJoinwebmind [0] (
06:12:46NJoinSchnueff [0] (
06:13:32 Quit Schnueff (
06:13:32 Quit webmind (
06:13:32 Quit mem (
06:13:32 Quit Synthe (
06:13:32 Quit mecraw (
06:13:32 Quit elinenbe (
06:13:32 Quit PsycoXul (
06:13:45NJoinelinenbe [0] (
06:13:45NJoinPsycoXul [0] (
06:13:45NJoinmecraw [0] (~mecraw@
06:13:45NJoinSynthe [0] (
06:13:45NJoinmem [0] (
06:13:45NJoinwebmind [0] (
06:13:45NJoinSchnueff [0] (
06:21:11 Quit datazone (Read error: 104 (Connection reset by peer))
06:21:18 Join datazone [0] ([GvyNP2SjP@
06:35:05***Saving seen data "./dancer.seen"
06:38:12 Quit Schnueff (
06:38:12 Quit webmind (
06:38:12 Quit mem (
06:38:12 Quit mecraw (
06:38:12 Quit elinenbe (
06:38:12 Quit Synthe (
06:38:12 Quit PsycoXul (
06:38:12 Quit Hes (
06:38:12 Quit Hadaka (
06:38:12 Quit datazone (
06:38:12 Quit tot|n8 (
06:38:12 Quit seb-sleep (
06:38:12 Quit DownKaos (
06:38:47NJoindatazone [0] ([GvyNP2SjP@
06:38:47NJoinSchnueff [0] (
06:38:47NJoinwebmind [0] (
06:38:47NJoinmem [0] (
06:38:47NJoinSynthe [0] (
06:38:47NJoinmecraw [0] (~mecraw@
06:38:47NJoinPsycoXul [0] (
06:38:47NJoinelinenbe [0] (
06:38:47NJointot|n8 [0] (
06:38:47NJoinDownKaos [0] (
06:38:47NJoinHes [0] (
06:38:47NJoinseb-sleep [0] (
06:38:47NJoinHadaka [0] (
06:44:13 Quit elinenbe (
06:44:13 Quit mecraw (
06:44:13 Quit Synthe (
06:44:13 Quit mem (
06:44:13 Quit webmind (
06:44:13 Quit Schnueff (
06:44:13 Quit PsycoXul (
06:44:13 Quit Hes (
06:44:13 Quit Hadaka (
06:44:13 Quit tot|n8 (
06:44:13 Quit datazone (
06:44:13 Quit seb-sleep (
06:44:13 Quit DownKaos (
06:44:42NJoindatazone [0] ([GvyNP2SjP@
06:44:42NJoinSchnueff [0] (
06:44:42NJoinwebmind [0] (
06:44:42NJoinmem [0] (
06:44:42NJoinSynthe [0] (
06:44:42NJoinmecraw [0] (~mecraw@
06:44:42NJoinPsycoXul [0] (
06:44:42NJoinelinenbe [0] (
06:44:42NJointot|n8 [0] (
06:44:42NJoinDownKaos [0] (
06:44:42NJoinHes [0] (
06:44:42NJoinseb-sleep [0] (
06:44:42NJoinHadaka [0] (
06:46:19 Join adiamas [0] (
06:46:34 Nick adiamas is now known as adi|home (
06:48:31 Quit adi|home (
06:48:31 Quit elinenbe (
06:48:31 Quit mecraw (
06:48:31 Quit Synthe (
06:48:31 Quit mem (
06:48:31 Quit webmind (
06:48:31 Quit Schnueff (
06:48:31 Quit PsycoXul (
06:48:31 Quit Hes (
06:48:31 Quit Hadaka (
06:48:31 Quit tot|n8 (
06:48:31 Quit datazone (
06:48:31 Quit seb-sleep (
06:48:31 Quit DownKaos (
06:48:38NJoinadi|home [0] (
06:48:38NJoindatazone [0] ([GvyNP2SjP@
06:48:38NJoinSchnueff [0] (
06:48:38NJoinwebmind [0] (
06:48:38NJoinmem [0] (
06:48:38NJoinSynthe [0] (
06:48:38NJoinmecraw [0] (~mecraw@
06:48:38NJoinPsycoXul [0] (
06:48:38NJoinelinenbe [0] (
06:48:38NJointot|n8 [0] (
06:48:38NJoinDownKaos [0] (
06:48:38NJoinHes [0] (
06:48:38NJoinseb-sleep [0] (
06:48:38NJoinHadaka [0] (
06:58:08 Quit adi|home (Remote closed the connection)
06:58:13 Join adiamas [0] (
06:58:30 Nick adiamas is now known as adi|home (
07:00:48 Join Zagor [0] (
07:01:46 Join datazone_ [0] ([NdgZxBeWG@
07:16:42 Quit datazone (Read error: 110 (Connection timed out))
07:16:49 Nick datazone_ is now known as datazone ([NdgZxBeWG@
07:18:38 Join LinusN [0] (
07:24:21LinusNor mecraw12?
07:24:38Zagormorning linus
07:24:47LinusNmorning Zagor
07:25:06 Join rwood [0] (
07:25:11LinusNi'm looking into the poweroff patch
07:25:30LinusNi don't want it to run on interrupt
07:25:41LinusNi want it in the power thread
07:25:45Zagoruh, definitely not
07:25:50Zagori haven't read it yet
07:26:13LinusNBTW, the ATA thread knows about user activity, doesn't it?
07:26:22Zagoryes. i was thinking about that too.
07:27:01LinusNah, last_user_activity
07:27:35LinusNbut that isn't really user activity, only disk-related user activity
07:27:47Zagorah, right
07:28:40LinusNok, i was thinking of letting the backlight thread record the user activity, but it seems so "secret", if you know what i mean
07:29:02Zagoryes, i agree.
07:29:10LinusNmaybe a call from button_irq is still good
07:29:24LinusNbut i want the timekeeping to be done in power_thread
07:29:40rwoodon a make after loading the current daily build, Updating dependencies for the apps dir fails because lang.h hasn't been built yet - or is this only on cygwin?
07:30:02LinusNrwood: happens all the time
07:30:17rwoodLinusN: ignore it?
07:30:37 Join merwin [0] (~none@
07:30:59LinusNyo d00d!
07:31:34*merwin hasn't been very involved lately
07:31:53rwoodLinusN: it will probably be another week on the player i2c stuff - the city electrical planner has screwed up our move - we are in the new building, but have no power for production or labs
07:32:07LinusNdon't worry
07:32:29rwoodLinusN: is there anything on i2c for recording that would be interesting?
07:32:34LinusNdo you still have your recorder on the operating table?
07:32:41merwinLinusN: is recording being dev'd yet?
07:32:51rwoodLinusN: yes - for the forseeable future
07:33:12LinusNrwood: but no way of measuring things right now?
07:33:43LinusNmerwin: yes
07:34:04rwoodLinusN: measuring what?
07:37:08LinusNi would like to measure EOD, PR and RTW when recording, maybe along with PA15, PB14 and PB15
07:37:30LinusNthe timing info in the data sheets doesn't satisfy me
07:39:29rwoodLinusN: i can probably add the signals, same problem with getting it done - probably a week or so
07:40:26*merwin forsees languages being a pain in the ass for upkeep :-)
07:40:52LinusNrwood: ok, no need then. thanks anyway
07:41:10LinusNmerwin: we are addressing that problem
07:42:16merwinLinusN: how so?
07:44:23LinusNwe will have scripts to keep track of the translation files, much like the coloured build table
07:45:05LinusNa red square for every untranslated string and so on
07:45:38LinusNand we will hopefully have maintainers for the languages
07:45:42merwinLinusN: ahh... now that is a good idea :-)
07:46:05LinusNit's not reality yet, though
07:48:26*merwin dropped his Ericsson T68i phone ($300+) in a keg... was underwater (beer rather) for about 20 seconds... it's still in perfect condition!
07:49:24LinusNa friend of mine dropped his cell phone from his shirt pocket into his own coffee cup :-)
07:49:51LinusNinstant phone death
07:50:29LinusNberr-safe phones, eh?
07:50:57LinusNnever saw that in the advertising
07:51:13merwinyeah, I was doing keg stands (standing upside down drinking out of the keg) and it dropped out of my shirt pocket :-)
07:51:19merwinthis thing is like a rock
07:51:30merwinI have a newfound respect for ericsson
07:52:04LinusNevery phone should be designed for keg stands
07:52:10LinusNcommon daily use
07:52:27rwoodmerwin: send them the story - next thing you know, you'll be a TV star
07:52:51merwinrwood: hehe
07:52:55merwinLinusN: damn straight :-)
07:53:24LinusNstraight? rather upside down :-)
07:54:25merwinLinusN: well, damn upside down then! ... This morning the LCD was blotchy and looked kind of bad, so I spent $30 and ordered a replacement LCD screen, but 12 hours later it was almost 100% clear again
08:16:30 Join Bagder [0] (
08:16:42Bagderhey ho
08:21:39Bagderonly 300 mails to read ;-)
08:24:11rwoodZagor: turning the led on in ata_flush before the call to write_sectors and moving the led(false) to after wait_for_end_of_transfer turns the led on until the disk has spun up - i don't know if the transfer is actually complete, but probably close enough that the power off timer will cover the write
08:24:42rwoodZagor: read turns the led off after wait_for_end_of_transfer write does it before - any reason?
08:25:10Zagorno, looks like an error
08:25:30rwoodZagor: do you want me to send a patch file?
08:25:39Zagornah, i'll fix it here
08:26:43rwoodZagor: is there a key combination that could be used for logical power off on the player - pause, ata_flush and power off in sequence?
08:27:43merwinrwood: good idea
08:27:54Zagorhow do you mean?
08:28:16LinusNaha, a shutdown()
08:28:38LinusNi like it
08:29:05Zagorwell, we are desperately short on keys on the players...
08:29:10rwoodi'd trade mute for shutdown
08:29:32rwoodbut then i never use mute
08:29:47merwinrwood: I wouldn't turn menu+off into poweroff :)
08:29:59merwinhow about holding menu+on for 2 seconds?
08:30:26rwoodmaybe 1 second
08:30:56rwoodi'm usually in a hurry getting out of the car when i shutdown
08:30:57merwinrwood: we want to be sure that you actually want to turn it off though... 1 second could just mean trying to slowly get into the id3 info screen :)
08:31:07merwinrwood: configurable!
08:31:31rwoodi'll take 2 seconds - another config item would be too much
08:31:37LinusNrwood: agreed
08:31:54merwinrwood: 2 seconds is shorter than you think
08:32:13merwinrwood: just think... buy a turbo car and have to wait 60 seconds just to turn it off :-)
08:33:08Zagorrwood: do you leave the archos in your car or take it with you?
08:33:41rwoodZagor: leave it in the car - heat probably isn't good for batteries
08:34:59rwoodmerwin: i went to a consumer focus group to evaluate car options today - one of the options was a finger print reader to replace the key - i'll be that that would take a while before you could start the car
08:35:06***Saving seen data "./dancer.seen"
08:35:34merwinhehe... a finger point reader? how do you have someone else start your car?
08:36:02rwoodrecognizes 4 finger prints & key still works
08:36:31Zagorrwood: sounds like what you really want is this:
08:37:14merwinrwood: ahh... well, that would probably promote body mutilation :-) "are you the owner of this car?" "yes, why?" "can I shake your hand" (hand reaches out and finger is chopped off) hehehe
08:37:48rwoodanother option was MP3 player built into audio system with cd-rip in car or module loaded on PC
08:37:55 Join linuxstb [0] (
08:38:06PsycoXuli dunno
08:38:15Bagderuh, rockbox feaure mail spammed :-)
08:38:15ZagorAudio A8 has a radio transmitter in the key. So you only need to have it in your pocket to unlock and start the car
08:38:16PsycoXulif they're not willing to beat/shoot you for your key's
08:38:22PsycoXulthey're probably not willing to start chopping
08:38:28ZagorBagder: yeah, I updated a whole bunch
08:38:55merwinZagor: wow, I don't know if I like that either :)
08:39:56rwoodanother option - wireless internet access to your car - lock/unlock, check fuel, turn lights on/off, download MP3 to audio system, etc
08:39:56HesGood morning.
08:40:15Bagdermorning hes
08:40:19HesHow's world?
08:40:31*merwin is going to bed
08:40:46 Quit merwin ()
08:41:14LinusNHes: world is fine
08:42:53Hesplayed some music using rockbox at a wedding on saturday night... after our band's gig
08:43:17rwoodZagor: i've been looking at the ata read retry - do you want to retry the whole transfer n times and/or read individual sectors after an error?
08:43:27Hesrecorded the gig on minidisc though 8-)
08:43:31LinusNHes: about your charging code, it takes 40s to sample battery voltage levels, and then it sleeps for an additional 60s. So it's 1:40 for each loop?
08:43:44HesOh, it does?
08:43:58HesThat's a bug. It should sleep less.
08:44:13Zagorthe whole transfer, I think. and i think we should retry n seconds, not n times, since the most likely cause of errrors is physical jarring
08:44:21LinusNPOWER_AVG_SLEEP is 10
08:44:36HesOk, that is correct
08:44:49Hes /* sleep for roughly a minute */
08:44:49Hes sleep(HZ*(60 - POWER_AVG_N * POWER_AVG_SLEEP));
08:45:01LinusNah, ok. I am blind
08:45:07HesIt sleeps for 20 seconds in the end. The comment is old.
08:45:14HesSorry for my bad comments again 8-)
08:45:34HesLooking at doing improvements there?
08:45:36LinusNsorry for my bad eyesight :-)
08:45:52LinusNHes: i am looking into putting the auto-poweroff there
08:46:00HesThat'd make sense.
08:46:29HesProbably enough to run the auto-poweroff once per minute.
08:46:48HesAnd only if charger is not plugged in.
08:47:08ZagorBagder: i'm splitting the lcd.c into lcd-player and lcd-recorder. this, of course, makes the simulators very upset. have you given this any thought?
08:47:10rwoodZagor: good idea - i'll start working on it
08:47:19LinusNmecraw has an idea of a sleep timer that shuts off regardless, after a longer timeout
08:47:27BagderZagor: yes, the player sim would need both files
08:48:04BagderZagor: or possibly only the recorder one, with some stubs for the player functions redirected to the recorder versions
08:48:56HesLinusN: and restart charging with the archos charger then?
08:49:12rwoodZagor: 1 more question - do you want a tight loop on the retries, or sleep a fraction of a second between retries?
08:49:27LinusNHes: hehe, i didn't think of that
08:49:32Zagorrwood: a small delay is probably good
08:49:50rwoodZagor: i agree - let others have a turn
08:50:17LinusNHes: i guess it should never shut off when the charger is attached... :-)
08:50:29HesLinusN: nope, since it will auto-poweron anyway.
08:51:09LinusNand we don't want the users to use the steenkin' Archos charger, do we?
08:51:10HesThe little beast powers on if you plug anything in the DC input even if there's no electricity involved. 8-)
08:51:19Hesan unpowered charger is fine
08:51:30HesLinusN: Nope, if they don't actually want to
08:51:38Hesit's a Good Thing to be able to choose
08:57:31 Join langhaarrocker [0] (
08:59:57linuxstbGood morning. Has anyone looked at my cuefile patch I sent to the mailing list last night?
09:00:18HesI've been looking forward to the poweroff timer, it's really easy to power the thing on accidentally, especially with spare batteries in the front pocket of the pouch
09:00:45BagderI vote for us starting to use the "patch tracker" on sf
09:00:52LinusNi agree
09:02:30langhaarrockerDoes that mean patchers need a sf login?
09:03:09langhaarrockermoin. Must go unfortunately. Have some stupid lessons at some stupid school.
09:03:55linuxstbThe controversial part of my patch is that it is in firmware - my view is that it is very closely linked to the id3 information, so it belongs in the same place as id3.c
09:04:41Bagderlanghaarrocker: no, it would mainly mean that patches would be posted to a better system/queue
09:05:24Zagorlinuxstb: i still disagree. id3 is an integral part of mp3 playing. cue files are not. how many mp3 players do you know with built-in cue file support?
09:06:06Zagori consider playlists more closely linked than cue files. and playlist handling is in apps.
09:06:19linuxstbBut cuefiles provide the same information as that contained read by id3.c - track titles and running times.
09:07:00linuxstbIt just seems a lot cleaner to keep it all together - i.e. whenever a an id3 header is read, we know we also need to read the cuefile.
09:07:12 Quit rwood ()
09:07:33Schnueffa .cue for this purpose is more like a .m3u
09:07:41Zagorafaik cue files provide track names, not track titles?
09:07:51ZagorSchnueff: my point exactly
09:07:54Schnueffso people can click on the .cue later to start playing the associated playlist
09:08:06Schnueff"associated playlist" wrong term
09:08:39linuxstbSchneuff: no - my choice was to make cue support "invisible" - i.e. it is read whenever the user plays an mp3 file containing a cue file.
09:08:59Zagoran mp3 file does not contain a cue file. a cue file references an mp3 file.
09:09:17linuxstbZagor: that's true.
09:09:40linuxstbSo should the cue file be ignored if the user just plays the mp3 file?
09:10:03Zagor.cue should be treated as .m3u
09:10:04Schnueffi think so too
09:10:09linuxstbAnd should cue files be allowed inside .m3u files?
09:10:21LinusNi don't get it
09:10:30LinusNwhat feature does cuefiles add?
09:10:33Schnueffone could make some new filetype for that
09:10:47SchnueffLinusN: u can play whole-cd-rips trackwise
09:10:59LinusNand a playlist can not?
09:11:22LinusNwhat gaps?
09:11:26Zagorbetween tracks
09:11:28linuxstbI use it for a different reason - I make a long radio recording, and use a cue file to make marks in the recording - like DVD chapters.
09:11:33Zagorbetween files, i mean
09:11:41LinusNaha, an entire mp3 file wor the whole disk?
09:12:22Schnueffif u have it as one single mp3, a playlist is insufficient
09:12:22Schnueff(unless we extend it with time offsets for each track)
09:12:22Schnueff(time offset + length)
09:12:55*LinusN fills his coffee cup
09:14:09Schnueffdoing this with .cue files is a hack, in my opinion, but winamp has a plugin to support this
09:14:29Schnueff(although i dont use winamp, so i could not try :)
09:14:45Zagori think cue file support is cool, but I want it handled like any other playlist
09:15:13PsycoXulcan playlists reference playlists?
09:15:33Zagor<Code Police>structs should not be called *_t. only types are _t </Code Police>
09:15:39linuxstbBut I may want to play three or four 30 minute mp3 files in succession, each with a cue file - how would this work.
09:15:39ZagorPsycoXul: no
09:15:43PsycoXulwhy not?
09:15:47linuxstbZagor: OK. I'll change that.
09:16:08linuxstbZagor: I am aware I waste RAM - I'm also going to work on that.
09:16:12Zagorlinuxstb: it doesn't work right now, just like we can't have playlist-in-a-playlist. when that works, cue files in playlist works too
09:16:16Schnueffi guess one can start with .cue file and then make something better
09:16:25Schnueffcombined .cue files / playlists
09:17:18linuxstbI've thought about this a lot, and I just think it is more natural for the user to "play" the mp3 file, and for Rockbox to automatically find and use the cue file.
09:17:24PsycoXulideas arise, are implimented, and their seperateness of implimentation slumps into the rest with time and evolves to become a smaller, cleaner part of the whole, and it is good
09:18:20Zagorlinuxstb: that means we require all mp3 with cue files to have id3v2 'cue' tags
09:18:47PsycoXulthere's id3v2 'cue' tags?
09:18:51linuxstbNo - the current patch just looks for a ".cue" file with the same name as the ".mp3". I don't like id3v2 tags.
09:18:59Zagorlinuxstb: ok
09:19:09Schnueffthere's something like marking specific times in the track
09:19:29Schnueffbut i guess it's not meant for trakc times
09:19:47LinusNcan you guys reach
09:19:59Schnueffno, but i got it in my wwwoffle
09:20:02LinusNor has it moved?
09:20:11ZagorLinusN: i can't reach it
09:20:12Schnueffdunno exactly, had problem too in the past weeks
09:20:32LinusNRIAA has shut it down :-)
09:21:15webmindfuck riaa
09:21:16Zagorlinuxstb: also, you can't cut down MAX_ID3_TAGS to four. it's too low.
09:22:04linuxstbZagor: It's fine for me :-) But this was because of my "stupid" memory management. I'll fix this soon and Cue-file support should only take about 8K of buffer space.
09:22:45Zagorlinuxstb: optimise more. keep a list of positions instead of the whole file in ram
09:22:46linuxstbLinusN: Is that good or bad?
09:23:10linuxstbBut it needs to know that track and title names of up to 99 tracks.
09:23:14LinusNwhat us taking 8K?
09:23:30LinusN99 tracks?
09:23:30Schnueffif one got it merged with playlist, it should take less mem
09:23:31Zagorlinuxstb: that can be read when needed. we don't need it in ram all the time.
09:23:33linuxstbI'll repeat that. Track title and track performer for 99 tracks.
09:24:10LinusNhow common is 99 tracks?
09:24:12Zagoralso cuefile doesn't need to be MAX_PATH. we require it to be in the same dir as the mp3 file.
09:24:32linuxstbOK. I agree that the memory usage can go down a lot. This will be my next priority.
09:24:38LinusNZagor: technically it still needs to be MAX_PATH
09:24:56LinusNit *may* reside in the root, and be 260 bytes long
09:24:57ZagorLinusN: doh, right
09:25:06Heslinuxstb: check what the playlist code does - the playlist is not in memory, just offsets to the file where the actual mp3 filenames are
09:25:44Hesyou could do the same, store offsets to the track information in the cuefile and read it when you need it
09:25:46LinusNthe behaviour could ne the same
09:25:51linuxstbI will need to buffer MAX_ID3_TAGS number of track titles in RAM from my cue file.
09:26:15linuxstbHes: That is what my current implementation does - but it buffers the whole cuefile in RAM.
09:26:51LinusNso using the playlist approach could be the solution?
09:27:13linuxstbI have one problem with merging it with playlists - I want the progress bar to show the entire track, with "tick marks" to indicate the individual tracks.
09:27:14Hescould support Huge cue files.
09:27:17LinusNand if done cool enough, you can use the ordinary shuffle as well! :-)
09:27:49linuxstbBut maybe I am on my own with that.
09:27:58Zagorlinuxstb: have you tested having a cue file in a playlist?
09:28:01 Quit langhaarrocker (Read error: 110 (Connection timed out))
09:28:12Zagorlinuxstb: i think that's a cool thing
09:28:49linuxstbZagor: No, only in a directory, and then next/prev works between files with and without cuefiles.
09:29:41Zagorok. the only thing i'm against, really, is the introduction of application data in the id3 struct
09:30:08Zagorsince cue files are handled in wps.c, it *is* application data...
09:30:38linuxstbDo any files in the firmware directory use the id3 data?
09:32:10linuxstbAnd cuefiles contain the same information as id3 tags...
09:32:54Zagorin a sense, yes. but look at how mp3 files are handled. mpeg.c gets a filename, plays it, end of story
09:32:59Schnueffcue files are for makeing the toc on a cd.
09:33:07Schnueffthey won't contain all id3 files
09:33:11Schnueffeh tags
09:33:14Zagorwith cue files, it's so much more complicated. we need to simplify it.
09:33:29Schnueffwhat about supporting time tags on each mp3
09:33:50Schnuefflike file.mp3[:30.12-2:30.0] or such thing
09:34:09linuxstbI think my cuefile patch is very simple - tiny change to id3.c and then small changes to wps.c and wps-display.c.
09:34:10Zagornah, lets not invent new "standards" :-)
09:34:11Schnueff(cdparanoia-like syntax)
09:34:23linuxstbAll the complicated work will be the memory management done by cuefile.c
09:34:32Zagorlinuxstb: yes, the code is simple. but the design gets muddled.
09:34:33Schnueffbut one could load a .cue file than like an extended playlist
09:34:40*Hes ignores the freezing weather and drives to work
09:34:56Schnueffand one could play a track from one big-mp3 and then from another big mp3 by using extended playlists
09:35:59linuxstbSchneuff - I personally don't use cue files, but they seem to be the only available "standard" for the job.
09:36:06Schnueffi agree, new standard are not desired. but cue files are not really meant for such thing
09:36:13Schnuefflinuxstb: yes, i am aware of that
09:36:30linuxstbBut people use them (e.g. mp3cue for winamp).
09:36:45Schnueffyes, it would be good to support them
09:36:59Schnueffbut it could be easier, by allow playlist entries to have a start time and a length
09:37:17linuxstbThere is no harm in allowing that.
09:37:28linuxstb... but how would you name the track?
09:37:35Schnueffand then u could load a .cue by constructing such a playlist
09:38:02Schnuefflinuxstb: hm yes good point
09:38:03ZagorSchnueff: people don't want to change their data. they simply want us to support it.
09:38:12Schnueffalthough extended m3u support such things
09:38:30Schnueff(for example for playing a http mp3 stream, u can't tell the name of the stream from the url)
09:38:43Schnueff(but u can still have it in an m3u)
09:38:47Zagorlinuxstb: you could check for (and read) cue file in wps.c:441
09:38:53linuxstbIn my mind this is also linked to the eventual "editting of mp3 recordings" that people will be asking for.
09:39:13linuxstbi.e. 1) Create a cuefile on the rockbox 2) (optionally) split the mp3 file according to cue file.
09:39:40Schnuefflinuxstb: how do the .cue files support having 'tags' in them exactly?
09:40:04linuxstbLook at the comments at the start of the cuefile.c that I posted to the mailing list...
09:40:07LinusNit would be cool to be able to insert queue points while recordding, just press a key for "add queue point"
09:40:18Zagorhmm, no that would mean spinning the disk again..
09:40:27Zagor(my wps.c:441 idea)
09:40:29LinusNZagor: why?
09:40:46Schnuefflinuxstb: will have a look
09:41:45Zagorwhat i'm getting at is that this is an id3 extension. it won't be the last. people will have lyrics files, album cover images, videos, you name it. we can't keep adding that kind of application data to struct mp3info every time.
09:42:06Schnueffhm, but such things are supported by id3v2.4
09:42:26linuxstbMy view is that id3.c should be moved to apps.
09:42:32Zagorthat's not the issue. id3.h is the issue.
09:42:48Zagorlinuxstb: maybe. but then we get sticky issues with mpeg.c instead
09:43:16linuxstbI think I will get sticky issues if I move cuefile.c to apps (but I'll have to think more about that).
09:44:30Schnuefflinuxstb: i guess these tags are the things the cd-text support (the track-name-thing for cd players)
09:45:09linuxstbmpeg.c calls the mp3info function, and it is at this time that the cuefile information is also needed.
09:45:22Schnueffso if u burn an audio cd with such a cue and then play it in a cd-text compatible hardware, the names appear
09:47:33linuxstbZagor: Re: line 441 of wps - I think that may work.
09:47:37Zagorhow about this: we add an apps callback, "read_app_data()" or something, that mp3info calls for each new file.
09:47:50Zagorlinuxstb: no, that would spin up the disk on every track change
09:47:57Zagoreven if mpeg data is already cached
09:48:07linuxstbTrue. I'll catch up soon....
09:48:11BagderI think a callback sounds fine
09:48:38Zagorbut how will the app know when the data is no longer needed?
09:48:56Zagorfree_app_data() :-)
09:49:04mecrawmorning, all
09:49:09Zagorhi mecraw
09:49:49LinusNmecraw: hi! I am applying your patch, with modifications
09:50:12Schnueffmaybe the app should handle at least two data sets
09:50:12mecrawdid i see some auto-poweroff comments earlier?
09:50:12Schnueffso it can be called for the current and next song, when disk is spinning
09:50:23Zagorre cuefile path: we already have mp3entry->path. it's just the last 3 chars that differ. that can be handled in a stack var to save 260 bytes/cue file
09:50:27LinusNi'm letting the power management thread handle the timeout
09:50:31mecrawLinusN: cool
09:50:49LinusNi can't see why the sleep timer is in global_settings?
09:51:15mecrawLinusN: i wasn't exactly sure where to put it, i figured i'd let the Swedish triumvirate decide :)
09:51:43LinusNi'm leaving the sleep timer out for now
09:52:20LinusNi'm not sure how we want to handle that in the best way
09:52:51LinusNso i'll apply the simple auto-poweroff and let the sleep timer be discussed further
09:54:44 Quit elinenbe (
09:54:44 Quit PsycoXul (
09:54:44 Quit mecraw (
09:54:44 Quit mem (
09:54:44 Quit webmind (
09:54:44 Quit Schnueff (
09:54:44 Quit Synthe (
09:54:56NJoinelinenbe [0] (
09:54:56NJoinPsycoXul [0] (
09:54:56NJoinmecraw [0] (~mecraw@
09:54:56NJoinSynthe [0] (
09:54:56NJoinmem [0] (
09:54:56NJoinwebmind [0] (
09:54:56NJoinSchnueff [0] (
09:57:37LinusNmecraw: who would want an auto-poweroff less than 1 minute?
09:57:56Bagder"202 messages saved and deleted" the weekend's share of rockbox mail ;-)
09:58:37BagderLinusN: I could very well imagine having it set to less
09:59:08Bagderwhy would I leave it doing nothing for a hole minute?
09:59:25 Quit Schnueff (
09:59:25 Quit webmind (
09:59:25 Quit mem (
09:59:25 Quit mecraw (
09:59:25 Quit elinenbe (
09:59:25 Quit Synthe (
09:59:25 Quit PsycoXul (
09:59:25LinusNthe power management thread has a resolution of 1 minute :-(
09:59:29Bagderhow long is the stock archos timeout?
09:59:45NJoinelinenbe [0] (
09:59:45NJoinPsycoXul [0] (
09:59:45NJoinmecraw [0] (~mecraw@
09:59:45NJoinSynthe [0] (
09:59:45NJoinmem [0] (
09:59:45NJoinwebmind [0] (
09:59:45NJoinSchnueff [0] (
09:59:46Zagori think 1 minute resolution is ok
09:59:52BagderLinusN: ... I would *never* set it less than one minute, didn't I tell you that? B-]
10:00:13LinusNok, so 1 minute resolution can be OK?
10:00:22*Bagder erases the log too
10:00:50Bagderlet's use that for now and wait for the crowds to complain
10:00:54ZagorLinusN: yes
10:01:49LinusNis a plain integer ok for timeout setting, or should we use a 1,2,3,5,10,30,45,60,120 minute setting approach?
10:02:01mecrawsleep timeout of less than 5 minutes is probably useless (except for my testing) :)
10:02:01Zagorplain integer is ok
10:02:18Schnueff0 for never
10:02:35LinusNmecraw: how would a user want the sleep timeout to work?
10:02:42HesAnd we can change the power thread a little to allow for a shorter timeout.
10:02:50LinusNHes: true
10:02:56linuxstbTo summarise our cuefile discussions - I'll keep it in firmware for now (so mpeg.c can call it at the right time), but improve the memory handling. Is that OK?
10:03:14mecrawLinusN: "shut off in 10 minutes no matter what"
10:03:18HesI have been thinking about modularizing the thread a bit (moving things from the thread function to other functions it'd call)
10:03:30Zagorlinuxstb: it's ok, but I won't commit it. i want to solve the design issue first.
10:03:39LinusNwould anyone want a longer autopoweroff than, lets say 10 minutes?
10:03:50ZagorLinusN: probably :-)
10:03:53BagderLinusN: I doubt it, then they disable it
10:04:09LinusNso a 4-bit field would ne enough?
10:04:13linuxstbZagor: Do you still want cuefiles to act like playlists? i.e. the user plays the cue file?
10:04:26BagderLinusN: I think so, yes
10:04:32Zagorlinuxstb: no, i agree with you that auto-finding the cue file is a good feature
10:05:15mecrawi just put a bunch of times in just to appease the masses, even though someone will want it to poweroff after 3 hours and 12 minutes.... can't please everyone
10:05:19linuxstbSo is "apps or firmware" the only outstanding issue (apart from RAM) ?
10:06:25mecrawLinusN: people will want sleep poweroff times of longer than 10 minutes... maybe not for idle poweroff though
10:06:50Zagorlinuxstb: it's i little more than just where the files are located. it's about which code does what work.
10:07:33 Quit datazone (Remote closed the connection)
10:08:07LinusNmecraw: is it intentional to not save the sleep poweroff timer?
10:08:40linuxstbZagor: Sorry. but I'm not sure I see the difference.
10:09:06mecrawLinusN: the sleep poweroff should only be used for one use
10:09:30 Part LinusN
10:09:51Zagorlinuxstb: I don't like id3.c doing cuefile work. i much prefer a generic callback.
10:11:14linuxstbZagor: I understand that point - that's what I meant about putting cuefile stuff in "apps".
10:12:16Zagorlinuxstb: good. then id3.c should not call read_cuefile() but rather read_app_data(), which will in turn call read_cuefile and store that data locally (not in mp3entry struct)
10:13:23 Join LinusN [0] (
10:13:34Schnueffcue file support with vbr is not very accurate, right?
10:14:01Schnueffwrong? hm
10:14:01ZagorSchnueff: not 100%, but shouldn't be too bad either
10:14:03LinusNonly in 1% steps
10:14:32Schnueffiirc, the bitrate was averaged or such thing. but with long rips, the bitrate should vary a great deal
10:14:37LinusNso in a 100 minute MP3 it would be in 1 minute resolution
10:14:57LinusNif we use the standard VBR headers
10:15:09ZagorLinusN: no, it would just be 100% correct in 1 minute resolution. you can still make a decent average estimate for smaller resolution
10:15:19Schnueffhm, for ripping whole cds with cue file (which is the standard application) thats way to bad
10:15:25Zagorlike we do for ffwd/rew
10:16:32linuxstbWork is calling me - I'll have to leave that problem to you for now. Another problem is that even in CBR files, the frame length can vary by one byte (the padding byte).
10:16:55LinusNZagor: i stand corrected
10:16:56Zagorbpt includes that byte, doesn't it?
10:17:02Zagorbpf, i mean
10:17:10linuxstbSorry, what's bpf?
10:17:16LinusNbytes per frame
10:17:46linuxstbThat's what I mean - bytes per frame can be either (for example) 576 or 577 bytes, depending on the padding bit.
10:18:16LinusNvery few files use the padding bit, what i know
10:18:39LinusNmaybe internet radion streams
10:18:57linuxstbYes - but I've found some :-(. I've written a very simple "mp2cut" utility that cuts layer 2 recordings. Some of these have padding bytes.
10:18:57LinusNbut i doubt it
10:19:05linuxstbMy examples are digital satellite broadcasts.
10:19:25linuxstbBut I agree - it is very rare.
10:19:32mecrawLinusN: did you see my response to your sleep timer question?
10:20:04linuxstbSee you later.
10:20:05 Quit linuxstb ("using sirc version 2.211+KSIRC/1.0")
10:20:14LinusNmecraw: no
10:20:46mecrawsleep is only used for one use, so there is no need to save it
10:21:12LinusNZagor: FYI, the padding bit can't be included in bpf, since it is different for each frame
10:21:21LinusNmecraw: i know
10:21:44Schnueffmecraw: people could get tired of adjusting it to their needs (and then they sleep without timer :)
10:21:50Zagorfun fun fun on the autobahn...
10:21:51LinusNbut the user is expecting the settings to be persistent
10:22:13ZagorLinusN, mecraw: you are discussing different things. idle poweroff and sleep are two separate functions.
10:22:14Bagderyes, they might like to use the same sleep timer every time
10:22:20LinusNi'd rather have a sleep setting, and a "go to sleep" button of some kind
10:22:57ZagorLinusN: most devices activate the sleep function anew every time. toggling off, 15, 30, 60, 90 minutes or some such.
10:23:22BagderZagor: it could still default to the previous selected time
10:23:26mecrawthat's how my tv works, i have to set it every time
10:24:03ZagorBagder: it could, but I don't quite see the point. i say we follow how other devices work.
10:24:12mecrawthe default would be cool, but i didn't want to take up more space in rtc than was necessary
10:24:13Bagdersure, that's not a big thing
10:24:27Zagorespecially if we only have like 4 different times
10:24:51 Join pyvasene [0] (
10:24:56mecrawso, the 22 times are too much :)
10:25:20Zagorhehe, yeah i'd say so :-)
10:25:54LinusNi'm only implementing the auto-poweroff for now
10:26:08mecrawfor now the really long timeouts will be useful since we can't turn repeat off
10:29:18Zagormy 5800 file playlist takes a while to complete too :-)
10:29:56 Nick tot|n8 is now known as TotMacher (
10:30:14BagderZagor: increase the pitch! ;-)
10:30:15LinusNis anybody working on the repeat option?
10:30:16mecrawLinusN: are you going to implement the idle poweroff so it powers off when paused?
10:30:28LinusNdo we want that?
10:30:45Zagori think we do
10:30:46mecrawLinusN: quelsaruk is looking into the repeat stuff
10:30:57mecrawi think so too
10:31:50LinusNi'm setting it to 0..15 minutes, 0 being OFF
10:32:14LinusNor is 10 enough?
10:32:14Zagordo we need to limit it so much? waste a byte, to avoid complaints
10:32:30mecrawsometimes i try to turn off rockbox ~200 songs into a 2200 song playlist, but don't hold stop long enough, so i have to start all over... now i can just pause it with idle poweroff set to 30 seconds and just leave it alone
10:33:06Zagormecraw: min idle off is not 1 minute :-)
10:33:16LinusNi need to implement a set_int with a special case for the 0 value...
10:33:54ZagorLinusN: use set_option then
10:34:08LinusNthat's why i want to limit the max
10:34:24LinusNor it would be a HUGE array of options
10:35:04Zagorwell with an option we can quantify it a bit. off,1,2,3,4,5,10,15,20,30
10:35:10***Saving seen data "./dancer.seen"
10:35:25LinusNhence my question before
10:35:38mecrawwhy not let it be less than 1 minute?
10:35:44Zagori know. i wasn't thinking
10:35:48LinusNis a plain integer ok for timeout setting, or should we use a 1,2,3,5,10,30,45,60,120 minute setting approach?
10:35:55LinusNthat was my question
10:36:41Zagormecraw: because the power thread only wakes up once/minute
10:36:54LinusNin the current solution, it can be changed
10:37:15Zagori don't quite see the need for <1min shutoff, though
10:37:30LinusNme neither
10:37:30Zagormecraw will simply have to learn to shut off properly :-)
10:37:42mecrawLinusN: why not do the sleep part too? isn't it a heavily requested feature?
10:37:50LinusNthe "clean" shutdoen idea is good, btw
10:37:52mecrawZagor: :p
10:38:09LinusNi'm not sure i like the UI approach
10:38:16Bagderit is very frequently requested indeed
10:38:30ZagorLinusN: how is the approach?
10:38:32LinusNis it ok to have that in the settings, non-persistent?
10:38:42Zagorsleep? yes.
10:38:48LinusNok, then
10:38:53mecrawZagor: i won't have to worry about shutting down properly once you have disk writing working and we can store the settings for each playlist :_
10:39:18Zagorhmm, wait. do we have anything else non-persistent in settings?
10:39:23BagderI think the sleep-timer doesn't have to be considered a setting
10:39:32Bagderit could be a "function" in a menu
10:39:45ZagorBagder: exactly what I was thinking :)
10:40:10LinusNthat's why i wanted to wait with the implementation and discuss it first
10:41:30Zagorcommit idle-off first
10:42:03 Nick TotMacher is now known as tot|away (
10:42:52Bagder8 languages in cvs now, pretty cool
10:45:38LinusNis that OK for the poweroff?
10:46:57mecrawi want 83 minutes!
10:46:59*mecraw hides
10:47:22*Bagder swings
10:48:04*mecraw falls over... due to intoxication and the wind from Bagder's miss
10:53:54mecrawgood night from the Rockies... can't wait to build from CVS in the morning!
10:54:09Bagdernight mecraw
11:03:17 Quit pyvasene ("Leaving...")
11:03:20 Join TotMacher [0] (
11:04:19Zagorany opinions which bugs we should fix for 1.4?
11:04:55LinusNi'm a little puzzled by the priorities
11:06:09LinusNrandom regularly repeats, is that a confirmed bug?
11:06:25Bagderregarding 612509, did we fix that fat_mount after the 1.3 release?
11:06:47ZagorBagder: let me check
11:07:00ZagorLinusN: half confirmed :-) i still can't repeat it.
11:07:27LinusNthe discussion has turned into a feature request
11:07:41ZagorBagder: that was fixed in 1.3. his problem is a truly odd fat partition, IMHO
11:07:59Bagderprobably yes, but then we could set that to fixed I guess
11:08:45ZagorBagder: well, it's still something of a problem since it works with archos firmware
11:09:15Bagderah, sorry, I misread your comment, I thought it was fixed _after_ 1.3... ok, then it is a bug
11:10:06ZagorLinusN: i interpret that request as a workaround for the bug
11:12:54 Join pyvasene [0] (
11:22:27LinusNZagor: i don't
11:22:55ZagorLinusN: his request does not explain the bug
11:23:01LinusNi think he wants resume to resume in the same playlist, but doesn't want the random seed to be the same
11:23:22LinusNbecause he wants "true" random
11:23:29Zagorsure. but that still doesn't explain the behaviour he gets.
11:24:17LinusNstopping the player without pausing explains it perfectly
11:25:25Zagorah, right. could be that.
11:25:51LinusNas i explained in my long comment
11:26:34LinusNi just hate the word wrapping in SF's bug tracker
11:27:24Zagori agree
12:25:32ZagorBagder: are you there?=
12:35:14***Saving seen data "./dancer.seen"
12:51:16 Join Jet8810 [0] (
12:52:57 Join Schnueff_ [0] (
12:53:02 Part LinusN
12:56:59 Quit Jet8810 ("Client Exiting")
13:13:05 Quit Schnueff (Read error: 110 (Connection timed out))
13:16:41 Join kargatron [0] (
13:17:26kargatronin request tracker, is high priority # correspond to high or low priority? seen it used both ways
13:18:14Zagorhigh # == high prio
13:18:50kargatronah, cool (since Zagor bumped up wps tag request :-)
13:19:17kargatroni wish request tracker could sort by last modifed, to see latest comments easily. is that configurable?
13:20:11Zagorno. sourceforge gives us no last-modified date
13:25:00 Join LinusN [0] (
13:25:30 Quit tot|away (Read error: 110 (Connection timed out))
13:25:45 Join tot|away [0] (
13:27:32kargatronwhen completely refreshing/replacing jukebox contents, is it recommended to reformat the drive? to get rid of fragmentation, etc?
13:29:18Zagorcan't hurt
13:33:01*Bagder is back
13:33:53 Nick Schnueff_ is now known as Schnueff (
13:38:07ZagorBagder: can you look at the playsim? my lcd split broke it in some strange ways
13:38:22Zagorthe charset break is the least strange :-)
13:42:09Bagderwell, it almost works for me
13:42:18Bagderjust some graphical errors
13:42:46Bagderand the cursor moves wrong
13:43:19Bagderhow are we gonna solve the charset issue for the player sim? any ideas?
13:45:36Zagori want it simulated closely to the real thing. with lcd_define_pattern and all
13:45:51Zagormaybe we can simply blow it up 2x to get the right "feel"
13:45:59Bagdertoo bad no one really cares about the player sim
13:47:48Zagorif someone else fixes the charset(s), can you write the code?
13:48:25BagderI could
13:54:31 Quit Schnueff (Read error: 110 (Connection timed out))
14:05:44 Quit mem ("reset")
14:11:14 Join langhaarrocker [0] (
14:13:57langhaarrockerBagder: And the winner is ... ?
14:14:34Bagderoh, right
14:18:08ZagorLinusN: red build
14:20:13LinusNthose f*ing simulators
14:20:43*Bagder watches LinusN's mouth with soap ;-)
14:20:55*Bagder goes and learns to type
14:24:22langhaarrockerBagder: you should wash the keys u, c and k on LinusN's keyboard because it's theam that don't seem to work.
14:25:13 Quit langhaarrocker ("Trillian (")
14:26:26 Join langhaarrocker [0] (
14:28:32pyvaseneluc : Vu que tes congés au Japon se sont transformés en congés sans solde... Tu vas meme peut etre devoir de l'argent a Alcove...
14:29:02 Join Schnueff [0] (
14:34:22 Nick Zagor is now known as Zagor|away (
14:35:15***Saving seen data "./dancer.seen"
14:42:25 Join _Snorlax_ [0] (
14:47:41 Nick seb-sleep is now known as seb-school (
15:12:40kargatronjust finished applying mp3gain to my 70GB of mp3s - really does make a postive difference in reducing having to volume adjust a lot (i'd previously peak-normalized).
15:12:52Schnueffhm converting those charsets codes_{old,new}.png with a perl script would spoil the joy of the non-programmers?
15:13:56kargatronAlthough I'm only halfway through my current jukebox, think i might refresh anew, since i'm constantly fooling with the jb vol.
15:14:01Bagderplease spoilt that :-)
15:14:02kargatronyeah, was thinking the same, re: perl script.
15:14:12Schnueffi got one
15:14:33kargatron(though my charset understanding is nil - you don't learn that stuff when you get a physics degree :)
15:15:14Bagderkargatron: hey, we'll be happy to feed lots of info to you until you get it ;-)
15:16:11LinusNkargatron: but mp3gain won't help when playing with Rockbox, will it?
15:16:31kargatronoh, is the header that's adjusted by mp3gain not read by rockbox?
15:16:52kargatronoh, thought that it was adjusting a std header
15:16:55kargatronoh well
15:16:57BagderI guess we could make it
15:17:10kargatronmakes a difference while listening at my computer, so still worth it
15:17:23LinusNit's adding an ID3V2 tag, or a field in the Lame INFO header
15:17:57kargatronYA feature request :)
15:18:12LinusNi've been thinking of adding support for that, but i haven't come around to that
15:19:03kargatronregardless, if you listen at your computer, the effort's worthwhile, imo. i notice it.
15:19:41Schnueffrow is high-order nibble in those charset?
15:20:30Bagderyes... or is it LSB first?
15:20:42*Bagder checks
15:21:54BagderLSB first
15:22:07Schnueffso 0x14 is A
15:22:19Schnueff0x54 is a
15:22:30Schnueff(in old charset)
15:22:56Bagderuh, you're talking about that png
15:24:02Bagdersorry, 45 is A in the old
15:24:15Bagder41 in the new
15:24:19Schnueffok better, i have it that way atm
15:30:07 Join bobTHC [0] (
15:30:14bobTHChi all!
15:32:11 Join YeAhx [0] (~aarond@
15:32:17 Quit langhaarrocker (Read error: 110 (Connection timed out))
15:32:21bobTHCi recieve a mail with "eol formats were problematic", what the deal with *.lang end of line ?
15:32:48BagderbobTHC: what about them?
15:32:52LinusNthere were problems if the last line didn't have EOL
15:33:10LinusNfixed now
15:33:11YeAhxnew builds are released even with no changes? is it automated?
15:33:15Bagderin the bdf font, yes
15:33:35BagderYeAhx: the daily builds are fully automated, yes
15:34:08LinusNwe rebuild every 20 minutes, if a file has changed
15:34:44YeAhxIm loving the lack of delay between tracks
15:35:11LinusNYeAhx: took some effotr to remove it
15:35:56YeAhxmy old rio wasnt bad but I dont think I notice any at all on my archos with rockbox
15:36:10kargatronthat only affects tracks that were contiguous but separated?
15:36:21LinusNor Lame encoded
15:36:35kargatronor does it shorten gaps between random mp3s?
15:36:52LinusNkargatron: it removed all non-mp3 info
15:36:53kargatron(say if lame encoded...)
15:37:33YeAhxoh its better with LAME encoded?
15:37:55YeAhxno wonder, thats my main encoder how
15:38:20LinusNi read somewhere that LAME does a better job somehow regarding the gap between tracks
15:38:40Schnuefflame man page documents a −−nogap option
15:38:47Schnueffbut i think that has to be specially enabled
15:39:08YeAhxso does ogg
15:39:09Schnueff(and it worx on a sets of contiguous .wave files, i presume)
15:40:00LinusNi guess that it encodes the two files as they were one WAV file, keeping the bit reservior filled in between
15:40:04YeAhxogg vorbis files are the only type I have got to play with no gaps on my computer
15:40:20Schnueffyes, and if takes sample from the next file to fill mp3 frames, i guess
15:43:23 Join pimlottc [0] (
15:45:04YeAhxstill, I dont think they playback as well on my computer as they do on the archos with rockbox
15:45:39LinusNyou don't need to suck up to us. we'll be nice to you anyway :-)
15:46:34pimlottcany recommendations for a program to keep sync between hd and archos under linux?
15:47:23Bagderpimlottc: I wrote a little perl script that works for me
15:47:34Bagderupdates my archos according to my local dir changes
15:47:43Schnueffbidirectional is more difficult
15:47:47pimlottcBagder - mind sending a copy?
15:48:00pimlottcschnueff - actually unidirection I think is all that I need
15:48:11pimlottcBagder - danke.
15:48:27pimlottcI have been trying to use rsync but it seems to be copied a lot more than is necessary
15:48:35 Quit _Snorlax_ ("gittar ny!")
15:48:37kargatronstupid question - do people talking of 'syncing' their jb have collections that are <= 20GB?
15:48:51pimlottckargatron - myself, yes
15:49:11Schnueffrsync usually pays attention on the file date
15:49:30kargatronwas just trying to figure out what synching would mean with collection >> jb
15:49:33Schnueffmaybe one has to use −−modify-window option to account for the low time resolution on vfat
15:49:56pimlottcSchneuff - it still sems to copy too much with the −−size-only option
15:50:23Schnueffand u set −−whole-file too?
15:50:39pimlottcI've also been thinking about what could be done in rockbox to aid the syncing process
15:50:50pimlottc(i'm aware there is no control when usb is plugged in)
15:51:17Schnueffhm strange
15:53:12pimlottcalthough I admit I haven't come up with much
15:54:34pimlottcI didn't set −−whole-file, I thought that only affect the how it copied
15:54:52Schnueffits just to make sure that rsync doesnt try to compute checksums on archos
15:54:59Schnueff(for partially transmitted files and such)
15:55:31Schnueffyes, it only applies after size/time check
15:55:34 Quit YeAhx (Read error: 104 (Connection reset by peer))
15:56:08 Join YeAhx [0] (~aarond@
15:56:12pimlottcI have not set it to use checksums at all
15:56:21LinusNtime to go home, bye all!
15:56:21pimlottcsince it'd just have to slurp the whole file over USB
15:56:45 Part LinusN
15:56:57Schnueffpimlottc: i thought that what the -W (−−whole-file) switch was for
15:58:09pimlottcschneuff - for what, turning off checksum?
15:58:46pimlottcas far as I see it doesn't use checksusm unless you specify −−checksum
15:59:04pimlottcand the runtime seems to agree
15:59:12Schnueffhm default is to use checksums, i'm quite sure on that
15:59:26Schnueff−−checksum says: use checksum even if dates are the same and size matches
15:59:53Schnueffbut, as u use −−size-only, it shouldnt make much of a difference anyway
15:59:53 Join YeAhx1 [0] (~aarond@
16:00:05pimlottchmm good point looks like you're right
16:00:50pimlottcI just know when I do −−dry-run it gives results much faster than if it's checksumming everything
16:00:54Schnueff(but maybe u often do tag updates are such things? then it would surely make a difference)
16:02:07Bagder" The iPod provides no gap suppression, so that in between every track there is a noticeable gap of about 1/2 second (or up to five seconds if the hard disk decides to spin up at the same time.) "
16:02:51 Join datazone [0] ([p5zkfbdmr@
16:03:08pimlottcschnueff - not really, once a file is in the collection it is not likely to be changed (perhaps rarely). I have a staging area for incoming files where I fix all that :)
16:03:09Bagder" It is incapable of playing music while reading from the hard disk"
16:03:43Bagderipod surely seems a bit "limited" in its software
16:03:53Bagder"So if you have a file that is longer than 32mb, it will play the first 32mb, then pause for 3-5 seconds while reading the next 32mb chunk into memory. "
16:05:16Schnueffpimlottc: hm ok. maybe give it another try with the -W option, to make sure no checksums are used
16:05:54 Quit mecraw (Read error: 104 (Connection reset by peer))
16:10:26elinenberead about how one person uses .cue files to remove these gaps on the PJB (Personal Jukebox)
16:11:28Bagderisn't .cue files for when you have one big file and want it to work as if it was multiple ones?
16:13:05Schnueffits for burning the TOC when u copy an audio cd.
16:13:10Schnueffbut u can abuse it for that purpose
16:13:54Schnueffand there's a plugin for winamp which supports that (mp3cue)
16:14:29YeAhx1did he think I was sucking up? hehe
16:14:44YeAhx1I just love rockbox
16:14:52BagderI *think* it was a joke, but you never know with Linus ;-)
16:15:06*Bagder is always serious
16:15:23*Bagder always ignores mail too B)
16:15:32YeAhx1sorry I was disconnected and went away for coffee so this probably makes no sense
16:15:37 Quit YeAhx (Read error: 110 (Connection timed out))
16:16:09YeAhx1this is what Im reffering to
16:16:13YeAhx1YeAhx: still, I dont think they playback as well on my computer as they do on the archos with rockbox
16:16:14YeAhx1 LinusN: you don't need to suck up to us. we'll be nice to you anyway :-)
16:16:47Schnueffdont confuse us now:)
16:29:43 Quit YeAhx1 (Read error: 104 (Connection reset by peer))
16:30:32 Join YeAhx [0] (~aarond@
16:32:11YeAhxhow crapola
16:33:13YeAhxwhat is a TOC file anyway?
16:33:49YeAhxI think there is an invisible one on a CD I just bought
16:33:50Bagderyou mean toc header?
16:34:21*Bagder stands silent in a corner
16:34:23YeAhxone that I cant get iTunes to name for me so I cant encode it which is pissing me off
16:34:38 Join edx [0] (
16:34:43 Quit TotMacher ()
16:34:43YeAhxI mean I could encode it but not with my choice of encoder
16:35:17***Saving seen data "./dancer.seen"
16:40:12 Join YeAhx1 [0] (~aarond@
16:40:18 Part YeAhx1
16:40:22 Join YeAhx1 [0] (~aarond@
16:41:16YeAhx1sorry I guess I should sign off, this connection just has to drop every few minutes and I dont want it to get anoying
16:42:19 Nick edx is now known as edx|studying (
17:02:15bobTHCwho make the pitch code ?
17:03:07bobTHCu think it's possible like the ff/fd config menu to configure steps....?
17:04:05bobTHCbecause it's to bad to listen to much "click" noise
17:04:38Bagderof course its possible
17:05:58bobTHCit's would be great....
17:06:11 Quit YeAhx (Read error: 110 (Connection timed out))
17:06:36bobTHCin all case, thanx a lot to linus for this so good option!
17:06:54Bagdergotta run
17:06:56 Part Bagder
17:09:37YeAhx1there is pitch code?
17:12:45bobTHCwhat ?
17:12:46YeAhx1the font thing sounds pretty cool
17:12:53YeAhx1pitch mode
17:13:06bobTHC>>>biatchin ??
17:13:30YeAhx1hehe yeah totaly
17:13:40Schnueffwhats the meaning of that
17:17:17YeAhx1it just sounds like a handy feature to have
17:18:17Schnuefffunny even
17:20:47 Join YeAhx [0] (~aarond@
17:21:22YeAhxI was just laughing at your reactions to my use of that word
17:21:43Schnueffit looked like a typo to me
17:22:08YeAhxwell usualy I would say "bitchin" but I like to alter it every once in a while
17:22:42YeAhxand most people dont even know the meaning of "bitchin" I probably dont either, I just use it as an alternative to "thats cool!"
17:23:46bobTHCand what's roflmao ?
17:24:03YeAhxrolling on the floor laughing my ass off
17:24:30bobTHCok thx..
17:25:51YeAhxno prob
17:26:33pimlottcbiatch is normally only an alternate of the noun form
17:27:36YeAhxah ic
17:27:39YeAhxoh well
17:29:03 Part kargatron
17:30:04YeAhxoh finaly found lyrics to a song that didnt come from the same source after looking at several pages that had the same one with the same errors
17:30:19YeAhxand it has the ending yeaaah
17:30:42YeAhxJimi Hendrix - 1983...(A Merman I Should Turn To Be)
17:35:00YeAhxin case anyone was wondering
17:35:30YeAhxI guess I better leave before I keep babling cyas
17:35:49 Quit YeAhx ("yip")
17:39:18 Quit YeAhx1 (Read error: 110 (Connection timed out))
18:04:47 Join langhaarrocker [0] (
18:09:14 Nick dw|gone0r is now known as dwihno (dwihno@Bald067.Baldakinen.Umea.SE)
18:09:14DBUGEnqueued KICK dwihno
18:14:39 Quit pyvasene ("Leaving...")
18:32:20 Quit pimlottc (Read error: 110 (Connection timed out))
18:35:21***Saving seen data "./dancer.seen"
18:52:28 Part bobTHC
18:53:03 Quit langhaarrocker (Read error: 104 (Connection reset by peer))
19:00:12 Join Phantom [0] (
19:00:15PhantomHI !
19:00:25PhantomI ve got a big Bug
19:00:42dwihnobug bug?
19:01:38Phantomat 0F00082A
19:05:50 Quit Phantom ()
20:33:49 Join Phantom [0] (
20:34:27Phantomnobody ?
20:34:49PhantomI ve a big bug and nobody to talk about
20:35:24***Saving seen data "./dancer.seen"
20:35:28 Part elinenbe
20:38:35PhantomHEELP : big fatal error
20:45:27 Quit Phantom ()
21:12:19 Nick seb-school is now known as seb-away (
21:56:24 Quit DownKaos (Read error: 104 (Connection reset by peer))
22:05:59 Quit edx|studying ()
22:14:44 Nick dwihno is now known as dw|gone0r (dwihno@Bald067.Baldakinen.Umea.SE)
22:14:44DBUGEnqueued KICK dw|gone0r
22:22:13 Join Ennan [0] (~Doe@CC53164-B.GRONI1.GR.HOME.NL)
22:26:21Ennanpeople here
22:28:16 Quit Zagor|away ("Client Exiting")
22:29:29Ennanhmm guess not.
22:30:38 Nick seb-away is now known as _seb_ (
22:35:25***Saving seen data "./dancer.seen"
22:51:56 Join elinenbe [0] (
23:28:07 Nick _seb_ is now known as seb-away (
23:35:07 Quit Ennan ("Client Exiting")
23:53:52 Nick seb-away is now known as _seb_ (
23:56:10 Join DexterAYS [0] (
23:56:27DexterAYShi dudes, do you already have a german translation?

Previous day | Next day