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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2002-09-09

00:28:06 Join EnnaN [0] (
00:28:07***Saving seen data "./dancer.seen"
00:32:52 Join AaronD [0] (~Yepyep@
00:34:44AaronDI found the solution to my corupted file problem woohoo, might actually get to use my player now
00:37:05EnnaNcongrats :)
00:38:49EnnaNyou a rb hero? just installed rb on my 1 day old ajs ..
00:45:31 Quit AaronD (Read error: 104 (Connection reset by peer))
01:05:23 Quit TotMacher ()
01:09:39 Join langhaarrocker [0] (
01:17:55 Quit EnnaN ("Leaving")
01:18:23 Quit langhaarrocker ("Trillian (")
02:28:11***Saving seen data "./dancer.seen"
02:42:06 Join lohap [0] (
02:44:26 Join hardeep [0] (
02:45:01 Quit lohap (Client Quit)
03:36:50 Join elinenbe_ [0] (
03:36:50 Quit elinenbe (Read error: 104 (Connection reset by peer))
03:56:36 Join pist [0] (
03:56:52 Part pist
04:05:44 Join rwood [0] (
04:23:27 Part rwood
04:28:14***Saving seen data "./dancer.seen"
04:37:10 Quit elinenbe_ ()
05:04:57 Join elinenbe [0] (
05:48:50 Quit Hes (
05:49:44NJoinHes [0] (
06:22:45 Nick _seb_ is now known as seb-sleep (
06:24:49 Nick seb-sleep is now known as seb-away (
06:24:53 Nick seb-away is now known as seb-sleep (
06:28:15***Saving seen data "./dancer.seen"
06:46:50 Nick edx is now known as edx|school (
06:55:15 Join datazone [0] ([NyzPmBte4@
07:21:49 Join rwood [0] (
07:43:26 Quit Hes (
07:43:26 Quit fraggle (
07:43:26 Quit datazone (
07:43:26 Quit elinenbe (
07:43:26 Quit adi|home (
07:43:26 Quit sylvr (
07:43:26 Quit dwihno|gone (
07:43:26 Quit laotan (
07:43:26 Quit rwood (
07:43:26 Quit seb-sleep (
07:43:26 Quit edx|school (
07:43:26 Quit Synthe (
07:44:15NJoinrwood [0] (
07:44:15NJoindatazone [0] ([NyzPmBte4@
07:44:15NJoinHes [0] (
07:44:15NJoinelinenbe [0] (
07:44:15NJoinadi|home [0] (
07:44:15NJoinseb-sleep [0] (
07:44:15NJoinedx|school [0] (
07:44:15NJoinfraggle [0] (
07:44:15NJoinSynthe [0] (
07:44:15NJoinsylvr [0] (
07:44:15NJoindwihno|gone [0] (dwihno@Bald067.Baldakinen.Umea.SE)
07:44:15NJoinlaotan [0] (
07:46:31 Join Bagder [0] (
07:48:31rwoodfor what it's worth, I just posted RoLo with a faster descramble routine - the multiply and divides were taking all time time
07:55:22 Quit hardeep ("Leaving")
07:57:26PsycoXulso how about a new WPS menu that just gives /.rockbox/*.wps for options... could do the same for *.eq too
07:58:20Bagdersounds like a nice thing, yes
08:01:52adi|homeBagder ive got a nice little keyboard going :)
08:02:21adi|homethat was yesterdays version
08:02:25adi|homeive come a bit farther...
08:02:36adi|homegot a 256 char buff for the keyboard..
08:02:44adi|homescrolls nicely on one line of the screen
08:11:27 Join hardeep [0] (
08:12:39Bagderadi|home: nice
08:12:51adi|homegot a better screen shot in a min
08:25:57 Quit rwood ()
08:28:19***Saving seen data "./dancer.seen"
08:41:17 Join rwood [0] (
08:58:36 Quit pimlottc (Read error: 110 (Connection timed out))
09:06:00 Quit hardeep ("Leaving")
09:10:56HesGood morning
09:13:57Bagderhey hes
09:18:21HesAh, new battery specialist in sight, good
09:20:50adi|homethere ya go bagder
09:20:54adi|homeversion .85
09:21:25adi|homemind you the del, shift and save don't work yet
09:21:34adi|homeim going to position them on the function buttons
09:21:36Bagderare those f-keys?
09:21:38adi|homeoff to exit
09:21:46adi|homei couldn't think of an easier way to do it
09:21:54Bagderseems cool
09:21:56adi|homecursor position on the screen is just dirty code wise.
09:22:11adi|homeany other chars that seem nec off hand?
09:22:23adi|homei can fit 14 more easily :)
09:23:06Bagdercomma and colon I guess
09:23:40Bagderit might depend on what the input is fore
09:25:51PsycoXul% and | so people can edit their WPS configs
09:26:09Bagderah right
09:26:30PsycoXulmaybe newline too in that case heh
09:26:54PsycoXulor something
09:39:35adi|homeare in..
09:39:40adi|home%| we can wait on ;)
09:48:02 Join RipnetUK [0] (
09:58:33 Join Zagor [0] (
09:59:09adi|homehey zagor
09:59:38Bagdermorning zaggie
09:59:48Zagoradi|home: ooh, looking good!
09:59:48PsycoXuli've been getting an odd bug lately
10:00:13PsycoXulwhere playback will stop for no apparent reason, the WPS goes blank, and it sits there reading the HD and then spinning down over and over a few times
10:00:41PsycoXulif i press on it goes to the browser and then pressing on again does nothing
10:00:46PsycoXulbut then i can browse around and play another song just fine
10:01:13Zagorsleeping late is nice...
10:01:19CtcpIgnored 2 channel CTCP requests in 5 hours and 55 minutes at the last flood
10:01:19*Zagor goes to make breakfast
10:03:58adi|homeokay.. enough keyboard for now..
10:04:09adi|homeZagor we are going to have to discuss what happens when you go into the keyboard
10:04:18adi|homeas far as what to do with the data that is 'typed'
10:04:30adi|homeright now its all on one line that scrolls the screen as needed.
10:04:41adi|homewith no way to look back at what you typed.
10:04:58adi|homebut i figure we can scroll 'typed data' once you pause for a given time
10:05:29adi|homePsycoXul: no need for a CR, each line added to WPS can be done by entering into the keyboard again.
10:05:33adi|homewe could have an edit menu
10:05:38adi|home'new line' would be an option
10:06:10PsycoXuloh and # for custom EQ files too btw
10:07:22PsycoXulor those might better be made by setting it and then saving the settings
10:07:43PsycoXulso nm heh
10:09:04adi|homei figure keyboard will just be called by what ever needs it
10:09:14adi|homeand the 'data' gets returned in a global buffer
10:09:18adi|homeright now its 256 big
10:09:25adi|homewhich should be more then enough for nearly anything
10:09:55adi|homeisn't there a quick way to go from lower case to upper case?
10:10:02adi|homeby adding or sub some value?
10:10:26Bagderyes it is
10:10:42adi|homeadding that does lower to upper yes?
10:10:44Bagder0x61 lower case
10:10:49Bagder0x41 upper case A
10:10:58adi|homesubtract that.. got ya..
10:20:20adi|homedo we have an ischar function?
10:22:11adi|homeheheh shift works now :)
10:22:13adi|homethat was nice and easy
10:22:47Bagdercool, does all the letters change on screen when you press shift?
10:23:38adi|homeall the a-z
10:23:45adi|home# and punc don't ;)
10:28:21***Saving seen data "./dancer.seen"
10:36:29adi|homeheheh that was a fun bug to fix
10:36:40adi|homeokay.. so now i can shift case w/o losing cursor position
10:36:48adi|homeman... ppl aren't going to like this code..
10:36:54adi|homecode police are going to want to play :)
10:37:18adi|homeokay.. i DEL going tomorrow..
10:37:22adi|homeand its as good as ready
10:37:31adi|homeuntil we decide what SAVE is going to do..
10:37:39adi|homegive that some thought folks.. im off to sleep
10:38:02Zagorsave should just return the string from keyboard();
10:40:58adi|homeright, but we if aren't using malloc'd memory..
10:41:04adi|homehow does that string not get lost?
10:41:20adi|homeor should keyboard() recieve a buffer and buffer size to store to?
10:46:46adi|homebeta for testing will be ready tomorrow sometime
10:46:49 Join LinusN [0] (
10:46:59Bagderhey ho linus
10:47:41LinusNI ran a very interesting test this weekend
10:54:19ZagorLinusN: tell us
10:59:35LinusNI compiled two versions of the Recorder rockbox
10:59:48LinusNone with ATA_STANDBY, and one with ATA:SLEEP
11:00:26LinusNATA_STANDBY: 10.5 hours
11:00:32LinusNATA_SLEEP: 6.5 hours
11:00:44Zagoryou have a seriously wacky unit...
11:00:52LinusNare you sure, Zagor
11:01:02Zagoryes. since i've tested mine.
11:01:11LinusNdo you think my unit is the only "wacky" one?
11:01:12Zagorbut if you have it, others probably have it too
11:01:41Zagorso how the heck can we ever detect which sleep to use?
11:01:53LinusNdon't know.....
11:02:15LinusNi guess we have to read the ROM version number
11:02:33Zagor?? we have the same version, don't we?
11:02:42Zagori have 1.24
11:03:20ZagorBagder, what's your rom version?
11:03:33Zagorwe need more measurements
11:04:10adi|homehows that?
11:04:10Zagorit's really weird how many hardware bugs there is in these things...
11:04:22Zagoradi|home: huh? you gf? ;)
11:04:47adi|homeheheh i wish
11:04:54adi|homemy ex was like that at one point
11:05:12adi|homebut more like 36-20-30
11:05:20adi|homeheads off to bed for real
11:08:25LinusNI noticed that the scroll thread sometimes outputs too short lines when using proportional fonts
11:08:58LinusNand that the cursor highlight in time/date is totally screwed
11:09:02BagderI have the same Linus has
11:09:12Zagorall font bugs await loadable-fonts before fixing
11:09:15BagderI gotta go, see ya later
11:09:18 Part Bagder
11:12:13LinusNi compiled a cgg-3.2 with binutils-2.13 this weekend, and fixed the linker script
11:12:44LinusNso now we can use the latest gcc if we want
11:12:55LinusNbut i'm not sure it will work all that well
11:13:20LinusNgot a lot of warnings when compiling the Rtx code
11:13:29Hadakaso no more illegal instructions from gcc?
11:14:04 Join TotMacher [0] (
11:14:10Hadakagcc-3.0 had broken size optimizations btw, dunno how it works for sh1 but gcc-3.2 should be able to produce smaller code if wanted
11:14:46Hadakabut good to know - so now I can upgrade my toolchain-src
11:15:25LinusN3.2 produces a smaller mod file by default
11:15:31LinusNabout 1k smaller
11:16:04LinusNbut i haven't had any problems with 3.0.4
11:16:37LinusNbut i haven't optimized for size
11:17:39Hadakaoh I mean broken as in not optimizing - not broken as in producing broken code
11:18:16LinusNyou talked about illegal instructions
11:18:41Hadakaoh that was just a reference to gcc-3.1 - producing sh4 instructions on sh1
11:30:43LinusNZagor: do you know what that guy is talking about in feature request #605773?
11:32:01LinusNi find it silly, masking a potential bug report in a feature request
11:32:19Zagorah, yes. he wants to be able to resume if he accidentally presses STOP in a playlist
11:32:33LinusNfunny, mostly it is the other way around, filing feature requests i bug reports
11:32:53LinusNZagor: no, his box is crashing
11:33:00Zagorah, read too fast :-)
11:33:29LinusNthe last paragraph is so confusing...
11:34:17LinusNas far as i can see, his problem is a crashng rockbox on a specific file, not the resume feature itself
11:34:36LinusNwe can't do more that ask if you want to resume
11:37:41LinusNand why didn't he file a bug report "crashing when playing a file"?
11:38:01Zagorpeople are strange..
11:39:05LinusNi wish we could add a column on the front page that sais "fixed" on the fixed bug reports
11:39:34Zagoryeah. i just don't know how to extract that info.
11:39:55LinusNme neither, hence "wish" :-)
11:41:41TotMacherwhere is the recent cvs activity gone ?
11:42:14LinusNis it gone?
11:42:26Zagorit is sometimes missing for a few minutes. it's a glitch.
11:42:27TotMacherafter reload
11:42:33TotMacherits there
11:46:40LinusNZagor: are you working on the FAT, or did you just say so? :-)
11:46:45Zagori am
11:51:40 Quit Hes (
11:51:40 Quit fraggle (
11:51:40 Quit dwihno|gone (
11:51:40 Quit sylvr (
11:51:40 Quit elinenbe (
11:51:40 Quit datazone (
11:51:40 Quit laotan (
11:51:40 Quit adi|home (
11:51:40 Quit rwood (
11:51:40 Quit Synthe (
11:51:40 Quit seb-sleep (
11:51:40 Quit edx|school (
11:57:46NJoinrwood [0] (
11:57:46NJoindatazone [0] ([NyzPmBte4@
11:57:46NJoinHes [0] (
11:57:46NJoinelinenbe [0] (
11:57:46NJoinadi|home [0] (
11:57:46NJoinseb-sleep [0] (
11:57:46NJoinedx|school [0] (
11:57:46NJoinfraggle [0] (
11:57:46NJoinSynthe [0] (
11:57:46NJoinsylvr [0] (
11:57:46NJoindwihno|gone [0] (dwihno@Bald067.Baldakinen.Umea.SE)
11:57:46NJoinlaotan [0] (
12:10:29LinusNhehe, now we have a 400-file limit discussion on Yahoo as well :-)
12:10:49LinusNand Björn is getting all the credit for Rockbox :-)
12:10:57Zagoryeah. but it's not as obnoxious as the one on our mailing list
12:11:08Zagori am?
12:11:35LinusN"Björn, your firmware is excellent!! Thanks for a very good job. If you have a
12:11:36LinusNfavorite charity, let
12:11:36LinusNme know and I will make a donation to them in your name."
12:11:54TotMacheri have 500 albums on one directory !
12:12:00TotMacherso i cannot read it
12:12:06Zagorah, right. although "your" doesn't necessarily mean mine. it can mean "ours" too.
12:12:29Zagorhe's just speaking to me
12:13:01Zagori'm always careful talking about "we have decided" and "we did" or "we feel" so people should not get the feeling it's a one-man project
12:13:27LinusNZagor: i'm not blaming you, i just though it was funny
12:14:08LinusNhehe, i often find myself talking about "the Rockbox" instead of just "Rockbox" :-)
12:14:40LinusNTotMacher: is this a problem for you?
12:15:01TotMacheryes i have personal complexes because of this
12:15:12LinusNTotMacher: how long are the dir names?
12:15:22TotMacheri always have long things :)
12:15:27LinusNon average
12:15:34TotMacherjust kidding
12:15:39TotMacheri dont have so much subdirs
12:15:49TotMacherbut i have more that 400 files in one :)
12:16:37TotMacherit is my "too much heard" directory
12:16:46TotMacherbut i want to conserve those mp3
12:18:00Zagorjust split it up and make a playlist
12:18:23TotMacherbut then i cannot directly access those files
12:18:48TotMacherand i am using
12:18:54TotMacherjukebox synchronizer
12:19:04TotMacherso my hdd and jukebox are alway synced
12:19:16Zagorcannot access?
12:19:33TotMacheri want to play just one file
12:21:00Zagorthen do that. navigating through 400 files is much slower than going into one dir and then search among 100
12:21:39TotMacherif i have to search in 6 direactorys a 100 files
12:21:44TotMacherit is very abusing
12:21:59Zagorso scrolling through 600 is *better* ???
12:22:00TotMacherthen its better in one dir
12:22:19TotMacheryou have to read just one time the directory
12:22:23TotMachergo ind
12:22:25TotMacherand go out
12:22:34TotMacherin avarage
12:22:37TotMacherit is more search time
12:23:47Zagorreading a second dir takes, what, 200ms. then you have 1/6 the files to go through. i'm not buying you arguments. but i don't have to, you are free to change this in your own firmware.
12:25:01LinusNZagor: maybe tools/configure could do this?
12:25:05TotMacherno !
12:25:10TotMacheri have to go one dir up
12:25:11ZagorLinusN: no
12:25:14TotMachertakes user intervetion
12:25:26TotMacherthat bothers the user
12:25:40ZagorLinusN: the people with this problem don't compile their own firmware
12:25:58ZagorTotMacher: yeah, and searching 600 files takes no user intervention. right.
12:25:58LinusNZagor: you just told Totmacher to do that
12:26:37TotMacherhow much more ram does one file take ?
12:26:59LinusNnumber of characters in the file name + 9
12:27:15LinusNnumber of characters in the file name + 7
12:27:15Zagorthe point is, TotMacher, that no matter how large we set the limit people will still always complain
12:27:27Zagor400 is more than 99% ever come close to. so that's what we have.
12:28:25***Saving seen data "./dancer.seen"
12:28:26Zagorthose with extreme needs are free to customize their firmware to suit them
12:29:50TotMacheri think a firmware which is used on a device which easily could contain more than >3000 files in a dir should support at least 1000
12:30:22Zagorfat add dir support in 1983. you should try it.
12:31:00TotMacherbut i dont have a criteria to sort my files in those directorys
12:31:10Zagorhow do you find them, then?
12:34:04TotMacherit is not rational to have more
12:34:09TotMacherthan 400 files in a directory
12:34:36TotMacherthe directory i use with those files, is just a garbage diretory for all mp3 i cannot longer hear
12:34:47TotMacherall the discussion about it
12:35:01TotMacheris not good for the popularity
12:35:10LinusNTotMacher: i agree
12:35:15TotMachermost of the things
12:35:19TotMacherare not rational
12:35:34TotMacherand people mostly
12:35:38TotMacheract not rational
12:35:44Zagoryou are wrong
12:35:46Zagormost people do
12:35:56Zagormost people are perfectly happy with this limit
12:35:58TotMacherand if they hear about a limitation
12:36:01TotMacherof the firmware
12:36:02Zagormost people didn't even know it's there
12:36:37LinusN"i was thinking about rockbox, but i read a lot about limits, so i didn't try it"
12:36:39TotMacherthey have a bad feeling about it
12:36:48TotMacherand wont even try it out
12:38:30TotMacherand i am concerned, that 99% not act rational :)
12:38:37TotMachermost of the time
12:39:16TotMachernot my problem anymore
12:39:21TotMacheri will split up my directory
12:39:25TotMacherwith a bad feeling :)
12:40:47 Quit Hes (
12:40:47 Quit fraggle (
12:40:47 Quit dwihno|gone (
12:40:47 Quit sylvr (
12:40:47 Quit elinenbe (
12:40:47 Quit datazone (
12:40:47 Quit laotan (
12:40:47 Quit adi|home (
12:40:47 Quit Synthe (
12:40:47 Quit seb-sleep (
12:40:47 Quit rwood (
12:40:47 Quit edx|school (
12:41:32NJoinrwood [0] (
12:41:32NJoindatazone [0] ([NyzPmBte4@
12:41:32NJoinHes [0] (
12:41:32NJoinelinenbe [0] (
12:41:32NJoinadi|home [0] (
12:41:32NJoinseb-sleep [0] (
12:41:32NJoinedx|school [0] (
12:41:32NJoinfraggle [0] (
12:41:32NJoinSynthe [0] (
12:41:32NJoinsylvr [0] (
12:41:32NJoindwihno|gone [0] (dwihno@Bald067.Baldakinen.Umea.SE)
12:41:32NJoinlaotan [0] (
12:46:20Zagori'm running a software development project here, not a presidential election
12:46:21Zagoras the faq says, people used big dirs because playlists sucked. they don't suck any more, so people can stop using big dirs.
12:46:30Zagorit's very simple to change for those who want it different
12:46:34Zagorthe HUGE majority don't need it changed, and should not suffer for the demands of a tiny minority
12:46:48Zagoram I wrong?
12:46:58TotMacherhat is DISCRIMINATION !
12:47:02TotMacheryou nazi
12:47:28TotMacherits ok
12:47:59LinusNthe main question is "how much is enough"
12:48:31LinusNok, how many users do you thing have 1000 file/dirs?
12:48:50TotMacherat least !
12:49:00Zagorfirst, estimate how many people have tried rockbox so far. i'd say 300-500.
12:49:15TotMacheronly ?
12:49:16Zagorhow many of those have complained about the dir limit? 3.
12:49:33TotMacherbut they complain hard
12:49:33Zagori'm being conservative. maybe it's several thousand who has tried it.
12:49:36LinusNZagor: then maybe there are lots of people with 600-file dirs just waiting to try rockbox :-)
12:49:48TotMacherYES !
12:50:02 Join Bagder [0] (
12:50:09 Nick TotMacher is now known as tot|klausur (
12:50:11Zagorcan you see my point now?
12:50:20LinusNso, how may people will we shut up if we raise to 600 files?
12:50:26 Nick tot|klausur is now known as TotMacher (
12:50:26LinusN1, maybe
12:50:27Zagor1 or 2
12:50:33TotMacheri always saw your point
12:50:38TotMacherbut i have to go now
12:50:53TotMacheryou are too rational :)
12:51:03LinusNthose people don't complain about the 400-file limit
12:51:16LinusNthey complain about that there is a limit at all
12:51:23TotMacherbad for the rockbox marketing
12:51:25TotMacherthis limit
12:51:34BagderI disagree
12:51:38LinusNrockbox isn't marketing
12:51:41Bagderits good marketing
12:51:45LinusNit's software development
12:51:54Bagderwe prefer big mp3 buffer
12:51:57Bagderthat's goodness
12:52:28LinusNbut if we had a dynamic handling somehow, we would have the cake and eat it
12:53:03Bagderlet's do that in version 1.8 B-]
12:53:12LinusNi say let the one who complains write the code
12:53:31LinusNSHOW ME THE CODE!
12:54:23LinusNi wonder what happened to the dynamic memory discussion
12:54:28LinusNit just died
12:54:37BagderI don't miss it
12:54:45LinusNi miss it
12:54:49 Join lohap [0] (
12:55:04LinusNi want to see how mr Wood will solve it
12:58:31 Quit lohap (Client Quit)
12:59:20LinusNby the way, i like this:
12:59:24LinusNtext = (char*)realloc(text, strlen(text)+1);
13:00:27LinusNefficient, dynamic memory usage
13:01:12Bagderwell, most of the times the same pointer will be returned ;-)
13:07:31ZagorBagder: if you feel like some extra web hacking, we can extract the whole SF tracker contents here:
13:07:42Zagorwould be good to mark bugs as "fixed" etc.
13:07:43BagderI know
13:08:07BagderI made an automatic script that can download it
13:08:15Bagderas you need to login as admin to do it
13:08:34Bagderbut I'm not quite sure what to do with it ;-)
13:09:14Zagori'd like a "Resolution" row in the bug list, for starters
13:09:40LinusNme too
13:09:49Zagorand we could have a much better list of feature requests
13:10:15Bagdercatgorized only if we introduce good categories
13:10:27Bagderbut we'd need a hell of a xml parser
13:10:40Zagorjust having all on one page would be an improvement
13:12:09BagderZagor: can you get the helvR08.bdf font showing?
13:12:16Bagderit occurred to me I never actually used that font
13:12:24Zagorcan't remember. i'll try
13:12:26Bagderthe makefile overrides it with that other one
13:12:45Bagderand when I carefully made sure I got the helv-one, it doesn't work
13:13:20LinusNBagder: how far have you come with the font stuff?
13:13:39BagderI have no working prop font
13:13:57BagderI wanted to get the ajf work first with prop
13:14:24Bagderas the .bdf parser will need some "fgets()" function or similar
13:15:37BagderI thought about doing a full fopen(), fgets(), fclose() layer
13:15:58Bagderbut I'm not sure if anything else will need it
13:17:31Zagortext file handling usually gets a lot easier with an fgets(), so it might be a good idea. i'm not sure if we should make a full f* layer or simply a gets() though.
13:18:06Bagderwell, fopen() and fclose() would make the fgets() nicer
13:18:17Bagderbut those three would probably be enough
13:18:47LinusNi think fputs() will be nice later on
13:19:04LinusNfor config files and stuff
13:19:12LinusNjust handy, not necessary
13:19:18Bagderbut writing is pretty easy with write() anyway
13:19:23LinusNit is
13:19:51LinusNit's just easier without caring about LF
13:24:58LinusNZagor: do you have permission to publish the ISD-200 specs?
13:27:57Zagornot explicitly. but cypress doesn't even sell the isd200 or isd300 any more, so I figured I might just aswell publish it and see if they complain
13:28:08LinusNand the ata specs?
13:28:30Zagorthat's downloaded from the web, so that should be no problem. it's a draft.
13:29:01LinusNyes, but i think they may dislike that we redistribute working drafts
13:29:16Zagor"Any commercial or for-profit replication or republication is prohibited." we are neither.
13:29:42Zagorthey might. if they complain, i will remove it
13:29:51LinusNPermission is granted to members of NCITS, its technical committees, and their associated task groups to
13:29:51LinusNreproduce this document for the purposes of NCITS standardization activities
13:29:58LinusNwe are neither .-)
13:30:31LinusNhas anyone tried the digital out from the recorder?
13:30:39Zagornot me
13:31:02LinusNi don't even know how to connect it :-)
13:31:14Zagori think it's described in the yahoo faq
13:32:37LinusNwhere is that?
13:33:50Zagorthere wasn't much in it, though
13:44:43BagderZagor: tried the font?
13:44:51Zagoruh, no
13:46:28Zagoris there some way to stop gcc from complaining about differing const in function parameters? I can't compile the ata-sim because linux uses const* filename and we dont (and no, we don't want to have it)
13:47:58BagderI don't think so
13:48:45Bagder-Wno-conversion perhaps
13:48:48ZagorBagder: nope, that font is bugging bad for me
13:49:06Bagderyeah, for me too
13:49:24BagderI guess that's a problem in the bdf2ajf converter or something
13:49:46ZagorHes: you alive?
13:51:14Bagderso I guess I need to fix that first, so that we know we understand the bjf format properly
13:52:28ZagorI hate it when people hit "reply" when they post a new subject to the mailing list. messes up my threading.
13:55:02 Join Tsd [0] (Tsd@
13:56:58HesI'm here
13:57:21HesI've been quite busy lately, haven't had time to listen to the jukebox too much...
13:57:27Hesor update from cvs
13:57:51Zagorok. i made the sleep handling a bit more relaxed a few days ago. you should try it and see if it improves skipping.
13:58:17HesWilld o.
13:58:31HesI will try to cough up a build today.
14:01:43 Quit Tsd ()
14:03:28 Join pimlottc [0] (
14:10:14LinusNZagor: about the "poor man's ROR", why not use asm(" rotr %0" : "+r" (data));
14:10:43Zagorbecause it's taken directly from descramble.c, which is ran on many platforms
14:11:05Zagorfeel free to fix it in rolo.c
14:23:34LinusNi am silly
14:23:45LinusNthe rotr instruction is 32 bit only
14:23:50Zagorboo :-)
14:28:27***Saving seen data "./dancer.seen"
14:42:54 Nick seb-sleep is now known as seb-school (
14:47:23 Nick Zagor is now known as Zagor|lunch (
14:51:23webbie_sup ?
14:51:47LinusNnot much, working
14:52:28webbie_school here
14:53:07pimlottcspeaking about rolo shorten 'descrambling' to 'descramble'
14:53:12pimlottc(11 chars on player ya know)
14:53:48Hadakahey, have you guys gotten your hands on a multimedia jukebox yet?
14:56:47Hadakahm, I wonder how similar that is
14:57:02LinusNit is a totally different beast
14:57:16LinusNi think it is an ARM CPU
14:57:40HadakaI wonder if they'll keep the same firmware system for that though
14:59:01LinusNpimlottc: done
14:59:24LinusNHadaka: maybe, meybe not
14:59:48LinusNthey still load the firmware from disk, for what i know
15:00:16LinusNbut i don't know if they scramble the firmware
15:01:16Hadakahm, well loading the firmware from disk is a good thing - and if the scrambling isn't difficult to handle, then it's good
15:06:20 Join champi [0] (
15:06:40champiI can't get the simulator working
15:06:52LinusNwin or linux?
15:07:17champiX error in rockbox-1.3-recorder-sim:
15:07:17champiX Error of failed request: BadAccess (attempt to access private resource denied)
15:07:17DBUGEnqueued KICK champi
15:07:17champi Major opcode of failed request: 2 (X_ChangeWindowAttributes)
15:07:17champi Serial number of failed request: 34
15:07:18***Alert Mode level 1
15:07:18champi Current serial number in output stream: 38
15:07:45champiyeah :-)
15:07:47LinusNi've never seen that
15:07:49 Join wagonr [0] (
15:08:13champineither do I
15:08:29Bagderdid you build it yourself?
15:08:41champinut I just downloaded the executable
15:09:30champiI'm gonna try to build it myself
15:09:33BagderI wouldn't be surprised if there's some library somewhere that differ or something like that
15:09:40wagonrbuy an Archos to try it in real :p
15:09:40Bagdertry building it
15:09:55Bagderthe sim is still very useful for development
15:10:42champiwagonr: tu sux
15:15:10 Quit PsycoXul (Read error: 104 (Connection reset by peer))
15:17:19***Alert Mode OFF
15:18:48champiwow it's not so easy to build ;)
15:18:53champibut same error
15:18:59champiwith my own build
15:19:40Bagderwhy isn't it easy to build?
15:21:13champijust kidding : you have to create a directory and run de configure script that is in another one :)
15:21:27champibut it build without any error
15:22:00Bagderso then I guess you need to check which X-function that fails
15:23:00champiI gues it has something to do with : X_ChangeWindowAttributes
15:23:51BagderXChangeProperty possibly, yes
15:27:51 Quit wagonr (" -= Des tonnes d'antisèches, techniques de triche, fausses excuses, cour")
15:30:39 Join EnnaN [0] (
15:31:18EnnaNany1 here?
15:31:57EnnaN:) just wanted to say "this is great" to anyone associated with rockbox...
15:32:07CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:32:07*EnnaN happy
15:32:16 Nick Zagor|lunch is now known as Zagor (
15:32:34*Bagder bows
15:32:40EnnaNjust got a ajs 10 saturday
15:32:46EnnaNinstalled rockbox yesterday
15:32:49EnnaNsmiled today
15:33:05Bagderthat's the spirit!
15:33:33EnnaNcouldn't find it, but is the "queue in playlist" option in development?
15:33:41LinusNnot yet
15:33:47EnnaNtoo bad..
15:33:59LinusNhow would you want such an option?
15:34:13LinusNinsert the selected song in the list? as in "how could we code that for heavens sake?
15:34:21LinusNir append to the end of the playlist?
15:34:21EnnaNor 'what should it do..
15:34:26LinusNor both?
15:34:38EnnaNwell, what i would really like myself, is a "play next" option
15:34:47EnnaNi like to play just a really big playlist at random
15:34:57EnnaNbut sometimes you feel like 1 certain song
15:35:10EnnaNjust play song next, and then continue with random playlist
15:35:15LinusNok, and the continue the the shuffled list?
15:35:30LinusNdo you need to queue several songs?
15:35:55EnnaNcould be usefull, but i think coding that would be harder...
15:36:04LinusNdon't worry about the code
15:36:23LinusNwe need to sort out the wishlist first
15:36:25EnnaN:) well, if it is too hard, nobody is gonna implement it, cause other shit might be usefull...
15:36:49EnnaNwell, it would be great if you could just say play this next, and than this, etc..
15:36:50champiI'm not sure but I think there's an option like this one in the official firmware (but it's really really buggy)
15:37:00LinusNqueuing one song won't be difficult
15:37:03EnnaNthere is an "queue" option
15:37:15EnnaNbut i don't know if it plays it as next song
15:37:50EnnaNthe nice thing about an option like this is that you will never have a silence, and you still can handpick your songs....
15:37:51LinusNi wonder how we should do it button-wise...
15:38:15EnnaNwell, while playing, you can go to the browser...
15:38:24LinusNZagor: good idea
15:38:28EnnaNthen you have a lot of buttons that are nog used..
15:38:29Zagorand replace normal play with BUTTON_PLAY | BUTTON_RELEASE
15:38:44EnnaN(*wtf is button repeat?*)(
15:38:47LinusNwith no repeat in between
15:38:59LinusNEnnaN: hold the button and it repeats
15:39:00Zagorlastbutton strikes again :-)
15:39:08LinusNZagor: :-)
15:39:32LinusNEnnaN: we are talking about a tricky handling of the keys
15:39:39LinusNa famous bug
15:39:41LinusNin rockbox
15:40:00champiin the official firmware that's what you do, you can queue a song by pressing play, but I repeat : it sux ! It plays something than the song you have selected !
15:40:18EnnaNi know..
15:40:27LinusNchampi: rockbox would never do such a thing :-)
15:40:39EnnaNbut does it play that song NEXT? or just cat it to the playlist...
15:40:39champilol sure :)
15:40:59LinusNwe haven't had *any* bugs that would play the wrong file :-)
15:41:09champiit plays it next but does not return to the playlist after it
15:41:10*LinusN is looking in the other direction, whistling
15:41:21EnnaNexcept for the luser=1d10t error?
15:41:50EnnaNwould it be hard to implement it to play the playlist afterwards?
15:41:54champiin the officiaf firmware, it displays the name of the song you wanted, but it plays another one ! strange bug...
15:42:08LinusNEnnaN: no
15:42:14 Join langhaarrocker [0] (
15:42:25EnnaNbtw: i didn't actually come in here to add to the wishlist at first, but just to say tnx :)
15:42:31LinusNi would implement it as two playlists, with priority
15:42:43EnnaNshould i submit this anywhere?
15:42:51EnnaNor is mentionning it here enough...
15:42:55LinusNEnnaN: there is already a feature request on that
15:43:04EnnaNnice :)))
15:43:57LinusNZagor: how about implementing it as two playlists, one which is "consumed", and has priority?
15:44:15LinusNthe problem is of course memory
15:44:25Zagorah, a stack
15:44:30LinusNsort of
15:44:39LinusNrather, a queue
15:45:20LinusNmemory isn't a problem
15:45:21Bagderno, that queue would be a stack
15:45:31Zagorsounds ok to me. a problem is that even small numbers of songs eat up a lot of space when we have to store full path
15:45:34Bagdersince you could queue more files
15:45:38LinusNyou just malloc() a block for each file :-)
15:45:40EnnaNlinusN: so if the priority playlist is empty, the "standard' is played? and as soon as you insert one in the prio, it plays that next...thats your idea?
15:45:48LinusNEnnaN: yes
15:45:55LinusNBagder: a FIFO
15:46:02EnnaNthat would be EXACTLY what i want!!!
15:46:16BagderLinusN: lifo
15:46:29EnnaNi've been screaming for that in winamp for a long time, and now in #3 it is finaly there :)
15:46:30LinusNwhy, then you can't queue a file after the queued one
15:46:34ZagorLinusN: FIFO or LIFO should probably be configurable. otherwise we'll get bug report "queued file was not played next"
15:46:42LinusNZagor: i suppose
15:46:56LinusNi can see the setting before me:
15:46:58BagderZagor: possibly the hold-play pop-up could offer both ways
15:47:01LinusNQueue Stack:
15:47:11LinusNclear as mud
15:47:26Bagdernono, ?IFO:
15:47:29BagderL/F ?
15:47:48EnnaNLinusN: clear as water! at least, i'd want it that way.. :)
15:47:49LinusN%cIFO, type?"L":"F";
15:48:00 Nick elinenbe is now known as elinenbe|here (
15:48:11langhaarrockerJepp, that' readable!
15:48:20elinenbe|heremorning ... well, good afternoon all.
15:48:23EnnaNyou could do it fifo/lifo/filo/lilo (for the redundency ;)
15:48:49langhaarrockerin SIMD?
15:48:58LinusNcan we have a fixed length queue?
15:49:10LinusNlike, 5 songs?
15:49:16Zagoryes, I think we should
15:49:24LinusNat least for now
15:49:24Zagor4 songs is 1KB...
15:49:34champiEnnaN: there IS a lilo for the archos, it's called ROLO ;)
15:50:02EnnaNchampi: i know, but i do not think it loads "li".. :)
15:50:43EnnaNgeneral: what time would there people be around here? (and especially, in what timezone would that time be?)
15:51:14LinusNthere are people here 24/24
15:51:41langhaarrockerMostly when I have to work or sleep ... :(
15:51:53Bagderbut be aware, most of the people here have no life ;-P
15:52:18EnnaNbagder: well, i'm a CS student, so i'm not supposed to have a life either :)
15:52:35Bagderok, you may fit well here then ;-)
15:52:49langhaarrockerNot life but at least an instance.
15:53:31LinusNZagor: or maybe the same approach as the dir buffer?
15:53:47langhaarrockerBut that doesn't matter: we don't use dynamic memory -> no garbage collection -> the instance is forever.
15:53:53EnnaNlanghaarrocker: i bet you're from holland, seeing your name?
15:54:00ZagorLinusN: the dir buffer assumes a single dir path
15:54:02langhaarrockerNo, germany.
15:54:51langhaarrockerEnnaN: But you're not the first to guess wrong. And I in fact understand and speak a very little bit dutch.
15:55:02EnnaN? your name looks dutch....
15:55:08LinusNZagor: i meant the allocation of name data
15:55:19LinusNa name buffer and an index array
15:55:28ZagorLinusN: right. but it still means potentially max 4 files/KB
15:55:44LinusNyes, but few people have that long paths
15:55:58Zagorbut yeah, null termination is a better way than a static MAX_PATH array
15:56:01LinusNso a 16 file queue with a 1k buffer, or something like that
15:56:24LinusNi'm on the job
15:56:41EnnaNnother general question (if i'm not anoying you too much); how hard is the rockbox code? how long would a reasenable coder take to understand some.....?
15:56:51LinusNit's easy
15:56:58LinusNplain, simple C
15:57:25LinusNwe work a lot to keep it simple, it's one of our main objectives
15:57:26langhaarrockerThe hard thing is to find out wether the SYNC pin of the MAS is connected to any interrupt or not.
15:57:32LinusNlanghaarrocker: hahaha
15:57:34EnnaNthat i knew, but if it is too lengthy and/or random written...
15:57:40Bagdera reasonable coder grasps it pretty quick
15:57:49LinusNlots of people have
15:58:01BagderEnnaN: dive in and check for yourself!
15:58:05EnnaNthink i'm gonna take a look at the code tonight when i get home...
15:58:25langhaarrockerI admit that it was _very_ easy to do the first steps. Contrats to the source scouts!
16:00:23LinusNhow should prev/next work?
16:00:41elinenbe|hereI see you are talking about queuing songs.
16:00:41*LinusN sees complications
16:01:05elinenbe|hereI think a great way is to hold play on a song ofr a few seconds, then 2 options come up insert/appent
16:01:26LinusNelinenbe|here: append to the list or to the queue?
16:01:32elinenbe|hereinsert will insert right after the current song playing, and append will append to the current list of songs playing
16:01:57LinusNelinenbe|here: and if you insert twice?
16:02:09elinenbe|herelist goes like this 1,2,3,4,5,6
16:02:14champiI think to many options will lead to a too much comlex interface
16:02:19LinusNmy list is 3000 songs
16:02:22elinenbe|herewhile plying song 2 you instert song 7
16:02:31elinenbe|herenow it looks like this 1,2,7,3,4,5,6
16:02:42elinenbe|herenow before 3 ends you instert 8
16:02:53elinenbe|herethen it looks like 1,2,8,7,3,4,5,6
16:02:55LinusNelinenbe|here: and if you insert song 8 before song 2 has played?
16:03:11LinusNok, your a LIFO guy
16:03:14elinenbe|herewell, it is only a feature WHILE songs are playing
16:03:24elinenbe|hereLinusN: that is the insert option
16:03:39elinenbe|herethen there is the append option, where if you did the same thing it would be 1,2,3,4,5,6,7,8
16:03:43champiThe idea with two playlist and priorities would be much more logical and easier to understand for the user
16:03:56LinusNi want 8 to appear after 7 aven when inserting
16:04:00elinenbe|herethe empeg car player has this options and it works GREAT!
16:04:23EnnaNwell, if some1 is here in a couple of hours, ill check back.. have a meeting now..
16:04:27EnnaNlatorz :)
16:04:40 Quit EnnaN ("Leaving")
16:05:03LinusNlanghaarrocker: why do you want the sync?
16:05:29langhaarrockerI want to detect when the MAS plays which mp3 frame.
16:05:55langhaarrockerNeed that to know which volumie info belongs to which frame for a split editor.
16:06:16LinusNa split editor in realtime?
16:07:05elinenbe|herewhat do you guys think about Hardeep
16:07:16LinusNand how do you know that the volume info isn't an average over several frames?
16:07:19langhaarrockerNo, press pause, go into split editor -> it buffers some mp3 data and works on that. When playing back this buffer 1st time it collects volume info. Voila: a graphical split editor.
16:07:27elinenbe|herewhat do you guys think about Hardeep's patch (where the cursor switches to the current playing song)
16:07:40LinusNelinenbe|here: i can see a good use for it
16:07:50LinusNi want it optional though
16:08:14elinenbe|herewell, he made it as an option. I really like it here.
16:08:41elinenbe|hereI think someone should commit it.
16:08:52 Nick elinenbe|here is now known as elinenbe (
16:10:04langhaarrockerHow long might an mp3 frame be at least?
16:10:21LinusNat least 24ms
16:10:30Zagori think it's good, since it's optional. elinenbe, will you commit it?
16:10:32LinusNup to 72
16:10:45elinenbealso, the oscilloscope demo is pretty cool −− the second one posted is much more effective then the first
16:10:59elinenbeZagor: only with your permission
16:11:03langhaarrockerI used lcd hardware scroll for that.
16:11:14*Zagor hands elinenbe permission on a silver plate
16:11:25Bagderooo silver...
16:11:39elinenbeit would be cool to have a definiable oscilloscope box for the WPS −− something like the status bar :)
16:11:49langhaarrockerI've got an updated oszi demo version that's switchable from scrolling to rolling.
16:12:27langhaarrockerelinenbe: that's difficult because the entire lcd must be scrolled.
16:12:36langhaarrocker(due to hardware scrolling)
16:12:51elinenbelanghaarrocker: did you do the original oscilloscope?
16:13:16elinenbelanghaarrocker: it would be nice to just have the peak values instead of filling it in.
16:13:17langhaarrockerI had to extend lcd.* for that.
16:13:22elinenbeI saw that.
16:13:32elinenbeit runs nicely though... the 2nd version.
16:13:42langhaarrockerThat can be done easily.
16:14:04elinenbelanghaarrocker: could you try to do that?
16:14:37langhaarrockerOk: Ill do that and post a 3rd version with filled / hollow, scrollable / rollable, adjustable speed.
16:14:51elinenbeadjustable speed?
16:15:08elinenbeis that instead of scrolling by one pixel, it scrolls by x pixels?
16:15:09langhaarrockerI can make it slower.
16:15:22langhaarrockerno, that's no fun
16:15:24elinenbewell, slower does not make it go with the music.
16:15:37langhaarrockerelinenbe: less updates.
16:15:49langhaarrockerlike 1 sample per more time
16:17:44elinenbelanghaarrocker: well, then it will not go with the music as well, I think it will look worse −− real time updates would be the best
16:18:21langhaarrockerIt stays realtime - it just scrolls slower.
16:18:35elinenbeah... I see.
16:21:22langhaarrockerDo you realliy write "oscillosgraph"? Dear me, my faulty english!
16:22:21langhaarrockerOr are there any suggestions for a better name?
16:23:25langhaarrockergone coding.
16:23:31 Quit langhaarrocker ("Trillian (")
16:26:14LinusNwhat would we want prev/next to do when having added/inserted a file?
16:26:30LinusNact if the added files are part of the playlist?
16:26:47Zagoryes. except prev of course can't go to the previously queued
16:27:37Bagderthe helvR08.bdf font gets all wrong when ajf'ified
16:27:53Bagderbut is pretty useful to point out flaws in the format
16:28:09Bagderit is gonna be hard to support a "static" font height
16:28:28***Saving seen data "./dancer.seen"
16:28:46BagderI'll have to write something down to explain this mess
16:29:03Zagorhow do you mean, static font height?
16:29:09Bagderone set height
16:29:13elinenbeLinusN: lookie here:
16:29:22elinenberead the LAST section on that page.
16:29:37BagderZagor: this font is said to use a 12x16 sized "box"
16:29:49BagderZagor: but each single letter is a lot smaller, within that set box
16:29:51elinenbeor you can just do a search for: "Insert/append from playlists menu" without the quotes.
16:30:24Bagdereach letter has its own height
16:30:55elinenbeLinusN: people absoutly love this method.
16:31:23Zagorelinenbe: it's not like "people" have tried a lot of different methods, is it?
16:31:42Zagorthey probably just love the feature
16:32:11elinenbeZagor: I assume, but the way this is set up, it satisfies nearly everyone.
16:32:16ZagorBagder: right, so do we really need to care about the box?
16:32:27elinenbeZagor: people just praise it...
16:32:41Zagorelinenbe: this is what we are about to fix, too. but without "append" for starters
16:32:50BagderZagor: not really, but that's the only "static" height we know about. To figure out font height, we need to check all letters in a defined string.
16:33:03Zagorwhat we are discussing is what happens with the *next* item you enqueue
16:33:03Bagderwhich might make things tricky
16:33:30ZagorBagder: we alredy did that, until I hardcoded a height just a few days ago...
16:33:50BagderZagor: but this would theoreticly make different lines use different heights...
16:33:58ZagorBagder: yes
16:34:15Bagderand we need to modify the ajf format ;-)
16:34:24elinenbethis is what the page says about the next enqueue: Enqueue: The first time you select this option, it inserts the selected item immediately after the current song. If you do this multiple times, it will insert the results after your last insert. Sort of like queing up songs on a jukebox.
16:34:26Zagorso probably there should be a "max height" and "max width" entry remembered for the font
16:35:03BagderZagor: possibly, but the max won't be very interesting for most things
16:35:19Bagderah, you mean for a given string?
16:35:27Zagorelinenbe: right. and we want that configurable. some people want enqueued files to always be played next
16:35:47ZagorBagder: no, I mean for the font. to use for line height in the dir browser for instance
16:36:11BagderZagor: well, in this font most letters are 8 or less pixels high, but the max is 16...
16:36:22Zagorwhat is 16?
16:36:28LinusNelinenbe: then it isn't like you described it
16:36:28Bagderthe max size
16:36:40Zagoryeah, but is any glyph actually 16 pixels high?
16:36:42LinusNyou said that songe were always inserted
16:36:44BagderZagor: I don't know if any letter actually uses the full height, no
16:36:58Zagorwe should have the max used, not the max stated :-)
16:37:03elinenbeZagor: well, I have used quite a number of hardware mp3 players and this feature on the empeg is incredible. Having this option with these features would be incredible.
16:37:26elinenbeLinusN: yes, I did explain it worng earlier
16:37:38elinenbeincredible :)
16:38:02 Join bobTHC [0] (
16:38:18bobTHChi all
16:40:46 Join quelsaruk [0] (
16:43:51 Nick edx|school is now known as edx (
16:46:16elinenbehello edx
16:52:44edxZagor, don't forget to update the sh-win page ;)
16:56:50 Part elinenbe
16:56:57Zagoredx: ah, will do that now
16:57:33 Part Bagder
17:04:33 Join elinenbe [0] (
17:05:14elinenbeLinusN: are you working on enqueue and append lilo/filo/fifo/lifo ?
17:05:55 Join edx|notebook [0] (
17:06:32LinusNelinenbe: i stopped, when i got too much to do with other stuff
17:07:23elinenbeLinusN: oh... I am quite excited about it. I will patiently wait.
17:23:01 Quit edx (Read error: 110 (Connection timed out))
17:27:48LinusNi think i just beat the world record in fast feature-request additions
17:29:01elinenbeLinusN: what do you mean?
17:29:24 Join EnnaN [0] (
17:29:26LinusNthe mono/stereo option posted to the feature-request database 30 minutes ago
17:29:35EnnaNhi again
17:29:43LinusNelinenbe: i just implemented it
17:29:59LinusNtook 33 minutes
17:30:14elinenbeahh... that's nice.
17:30:20quelsarukwau.. a record, LinusN
17:30:25LinusNi have to catch my train now
17:30:29LinusNcu later guys
17:30:54 Part LinusN
17:31:48 Quit mecraw (Read error: 54 (Connection reset by peer))
17:33:17 Join hardeep [0] (
17:33:33EnnaNhmm, is it normal for my original headset to distord much more that a *normal* 1?
17:38:28hardeephmmmm, red build
17:38:38elinenbeZagor: how is the fat32 writing coming?
17:38:49elinenbehardeep: I submitted your patch
17:39:12hardeepelinenbe: thanks.
17:39:13elinenbehardeep: but there is a bug when you turn file resume on.
17:39:31elinenbehardeep: te cursor is always on the first file which gives some interesting results.
17:39:44elinenbetry it on a file that is in the last few of a directory
17:39:59elinenbethat has both more then and less then 8 files in it.
17:40:09elinenbethen try hitting up/down
17:40:29hardeepelinenbe: yeah, that's what magnus also reported.
17:40:37hardeepi'll fix it
17:40:57elinenbehardeep: I really like that patch :)
17:43:55hardeepme too. :) it's one of those minor things that annoyed me on rockbox
17:45:02elinenbehardeep: now when you reboot though, the cursor is not on the correct file unless you go to that setting in the menu again.
17:46:00elinenbehardeep: hit on twice to get back to the song, then hit on again, and you get back to the correct position in the menu
17:46:21elinenbeand this is when rebooting.
17:47:31hardeephmmm, i don't quite follow. 1. enable browse current, 2. hit on 3 times, 3. reboot jukebox ?
17:47:53elinenbehardeep: no. let me explain.
17:48:57hardeepah, i see it. when you resume a file after reboot
17:49:14elinenbehardeep: 1) enable browse current 2) play some song 3) turn off 4) turn on 5) press on to get to tree (on first file in root dir (wrong file!)) 6) press on to get back to song 7) press on again (now on right song!!)
17:50:20elinenbehardeep: see this behavior?
17:50:53hardeepyep, will fix this also. :) it's related to resume
17:51:31elinenbehardeep: awesome!
17:54:01elinenbeI know langharrocker is not around who did the oscilloscope code, but I was wondering something. Couldn't the max values that the oscilloscope puts out be read and then a dynamic normalization be done to adjust the volume on the fly? Like a software based auto volume?
17:55:18elinenbealso, there is a problem with the oscilloscope code. When I am playing music and just watching the oscilloscope every few songs, then music just stops. I can go back to the directory browser and start the song over, but this is reproducable and kind of annoying.
17:55:47 Nick edx|notebook is now known as edx (
18:13:15 Nick Zagor is now known as Zagor|out (
18:14:58 Join sesam [0] (
18:15:01 Quit sesam (Remote closed the connection)
18:18:31 Nick elinenbe is now known as elinenbe|out (
18:28:30***Saving seen data "./dancer.seen"
18:30:34 Join langhaarrocker [0] (
18:31:14langhaarrockerHm. Now that Im here again elinenbe is gone :(
18:34:11langhaarrockerWell, I've sent the desired osci code to the mailing list, that's all I can do by now. CU later I've got to visit the practise room.
18:34:27 Quit langhaarrocker (Client Quit)
18:36:23bobTHCquelsaruk : u always work on the multilangual ?
18:36:36 Join LinusN [0] (
18:37:20LinusNhehe, of course i broke the sim builds :-)
18:37:30bobTHCas usual... ;)
18:38:14edx`MAS_BANK_D1' undeclared (first use in this function)
18:38:30LinusNi'm on it
18:38:39edxk :)
18:39:13bobTHCoscilo function is in the last daily ?
18:40:40elinenbe|outbobTHC: nope. it has not been comitted yet
18:41:14quelsaruksorry bobTHC??
18:41:40quelsarukdid u ask if i'm working on the multilingual support?
18:43:39quelsaruknot now, as Bagder said we would use a tool to "translate" from .lang file to language.h, we must first know how will this .lang files be....
18:43:49quelsarukso i've found another interesting thing to do
18:43:56quelsaruknon-continuous play
18:44:19bobTHCok.... i
18:44:31quelsaruki wan't my jukebox to stop playing once all my files have been played :)
18:45:02bobTHClike me
18:45:10quelsarukLinusN: none has worked before on this, no?
18:45:36quelsarukjust to know if there is something like this...
18:47:53elinenbe|outquelsaruk: no one is working on repeat on/off yet
18:49:49bobTHCC U later ... bye
18:49:53 Part bobTHC
18:51:02quelsarukthx elinenbe|out
18:56:02hardeepi think mecraw was looking into it. re: repeat on/off
18:57:47quelsarukok, thx hardeep, i'll ask him,
19:03:20 Part LinusN
19:34:21hardeepcvs is sloooooowwww today
19:38:20 Quit champi ("Fermeture du client")
19:52:41 Quit RipnetUK ()
20:06:02 Quit hardeep ("Leaving")
20:28:34***Saving seen data "./dancer.seen"
20:39:23 Quit sylvr (Remote closed the connection)
20:54:31 Join hardeep [0] (
21:09:34 Join xam_ [0] (
21:09:54xam_you guys are sooooo crazy!
21:10:13xam_rockbox rules. It's so much better than the original firmware
21:11:19xam_I just miss the recording option ... that's all ... and if there was .ogg support it would be perfect
21:13:50edxwe are working on the recording stuff... there's a long way to go.. and there where discussions about .ogg support - it is out of reach as it seems
21:14:34xam_I know ... missing manuals
21:15:23xam_what about deleting/moving/renaming files and creating/deleting/moving/renaming directories?
21:15:37Hesthat'll come with filesystem writing, eventually
21:15:43edxwe do not yet support writing to the fat filesystem
21:15:48Hesfilesystem writing is being worked on (it's a prerequisite for recording)
21:16:25xam_mhh, itsn't it a hitachi zh-1 processor? doesn't linux run on these cpus? ;)
21:17:04xam_at least the zh-3
21:17:15Hessh-1... doesn't run on it
21:17:25Hesruns on the sh-3
21:17:56elinenbe|outhardeep: great fix!
21:18:17 Nick elinenbe|out is now known as elinenbe (
21:22:09xam_even the music sounds better!
21:22:10 Nick edx is now known as edx|sleep (
21:22:29elinenbexam_ : and looks better too!
21:22:37elinenbexam_ : do you have a recorder or player?
21:22:43xam_elinenbe recorder 20
21:24:46 Quit quelsaruk ("KVIrc 3.0.0-beta1 "Eve's Avatar"")
21:24:50elinenbeah... then try this as the firmware:
21:25:04xam_elinenbe what's the difference?
21:25:18elinenbeplay a song and then go back to the menu, then to Demos, then to oscilloscope.
21:26:02xam_elinenbe ;)
21:26:11xam_elinenbe got it ...
21:26:17elinenbepress the F2 button for different option
21:26:25elinenbeI did not program this, but it is very nice.
21:26:40elinenbeF1 button for different modes.
21:26:54elinenbeup/down for speedup/slowdown
21:27:31elinenbeit may crash −− it is not completly stable yet :) (but it will not do anything bad - just hold down the off button)
21:28:50xam_elinenbe the font is different aswell ...
21:29:49elinenbeyeah −− that is the proportional font. you can see much more on the screen :)
21:29:58xam_whoaaaaa - the oscillograph is coooool
21:31:36xam_elinenbe and if you press up/down you can slow the oscillograph down/up if you want
21:34:24xam_elinenbe where did you get this firmware? from the CVS?
21:36:17 Nick seb-school is now known as _seb_ (
21:36:56elinenbeyes, and from a patch on the mailing list.
21:37:41xam_elinenbe you know where to get the sh-elf gcc? I already have proper binutils
21:37:56xam_elinenbe sh-elf-gcc debian packages
21:38:31xam_elinenbe I don't feel like wanting to bootstrap a crosscompiler right now
21:39:41xam_there is something in the debian repository called gcc-h8300-hms, but I don't think it's the right .deb
21:40:46xam_does anybody know whether these changes are already in the CVS?
21:40:51 Nick Zagor|out is now known as Zagor (
21:40:54Zagorno they're not
21:42:40 Nick xam_ is now known as xam (
21:44:28xamZagor will they be integrated in 1.4?
21:46:58xammhh, finally found that there are instructions for building a crosscompiler on the rockbox webpage ... seems I have to do it anyway ...
21:55:41Zagorxam: possibly
21:56:15xamsorry for all the questions, but what does the option "General Settings -> File View -> Browse Current" do on the patched version?
21:59:53 Nick _seb_ is now known as seb-away (
22:00:14Zagorit toggles whether the browser should change dir to currently-playing song or stay still
22:02:36xamZagor don't get it (/me too stupid?). "Browse Current Song" is default "no", but when change it to "yes" I don't see a difference
22:02:52xamZagor what do you mean by stay still?
22:03:38xamZagor what is the default action? I mean if I select/play a song, it automatically changes the directory ...
22:03:49Zagorxam: you only notice a difference when you play OFF or ON (go to dir browser). with "Browse current song" active, the dir browser will be positioned at the last played file
22:04:48xamZagor I see, thanks alot !
22:13:46xamI see you have implemented you own powermanagement, and battery recharge is using delta U detection (voltage drop). Do you know how it was done before or whether the battery life is better with rockbox?
22:14:54Zagorbattery life is better with some versions. some versions of some models seem not to like our power saving tricks, however
22:18:45elinenbeZagor: how is file writing progressing? making any improvements yet?
22:19:22Zagorstill laying the groundwork. there's lots of stuff to write (time/date handling, to just name one)
22:28:36***Saving seen data "./dancer.seen"
22:29:30elinenbeZagor: good luck with it.
22:31:01Zagorluck is not a factor ;)
22:31:42elinenbeZagor: it may be when you are testing it :)
22:32:16Zagoryup. that's just an old movie quote I enjoy using. from some NASA/space movie.
22:36:24adi|homeheheheh i like the osilligraph
22:37:24Zagorit sounds cool. i haven't tested it yet.
22:39:26elinenbeZagor: did you see what I wrote about the oscillograph a while back?
22:41:21elinenbeZagor: we know the max of each bit of music. Why can't we take an average of the past X number of samples, and then normalize it on the fly or something like that?
22:42:21elinenbeZagor: adjust the volume acordingly.
22:43:14HesWe could, yes.
22:44:38elinenbeHes: I think that could be a neat idea.
22:44:45elinenbeHes: something worth investigating
22:48:27Zagorelinenbe: how does that differ from the AVC?
22:48:42hardeepplayer doesn't have AVC does it?
22:48:59Zagorright. good point :-)
22:49:49elinenbebut, can we read the max values of each sample on the player?
22:50:03hardeepactually, how many people actually use AVC? I personally don't like it
22:50:15hardeeptoo many actually's :)
22:50:22elinenbeI don't use it either
22:50:22HesI don't like AVC, but might like it if it had longer decay times
22:50:43elinenbesame here. I think it is too quick. You can "hear" the changes.
22:50:49Zagori use it sometimes when randomising my entire collection. it's good to avoid having to reach for my amp remote
22:51:15elinenbeWhat I would be looking for is somehting that can adjust for most of the song.
22:51:30xamhardeep I don't like AVC
22:52:46xammhh, a notebook/adressbook function and a textviewer (even graphicviewer?) would be nice aswell once we can write to the filesystem
22:53:00Zagorelinenbe: we still won't know the level until we've played it
22:53:12Zagorxam: i'm sure someone will write it :-)
22:53:34elinenbeZagor: hopefully we could adjust the level throughout the song... slowly −− like the battery charging −− using deltas
22:53:38xamzagor hehe openoffice on archos
22:54:13Zagorelinenbe: yes
22:54:18elinenbeas we aquire mor samples in the song, as it plays, we would know more about the volume of the song.
22:54:48Zagorthe problem with both oscillator and soft-avc is that it kills our power-save plans. we've have to keep busy and sleep much less.
22:55:23elinenbein the beginning of a song it would most likely be inaccurate so we do not adjust the volume there. Only as it gets a certain percentage into the song can we start modifying the volume, and even then it would only be by such a little bit in most cases.
22:55:28hardeepyou could save the data to disk and re-use it the next time the song is played...
22:55:43adi|homejesus.. this whole wrap issue is turning into the 400 file issue all over again..
22:55:53elinenbeZagor: that and the oscillator carshes my recorder after a song or 2. CPU:blahblah error
22:56:19Zagorah, "Luck is not a factor" is actually from The Abyss. (gotta love IMDB)
22:56:34Zagorelinenbe: yeah, but that's just a bug. bugs get fixed.
22:56:52xamZagor someone should tell this microsoft
22:57:11elinenbeZagor: if IMDB ever goes down that will be a sad day. They should sell their database. But at least it is backed by
22:57:35elinenbeadi: you get annoyed quite easily. You hvae to learn to ignore.
22:57:58hardeepi don't know if it's still the case but you used to be able to download the entire imdb database
22:58:24Zagori agree with you adi. this option would be much more confusing than helpful
22:59:02Zagorhardeep: i would be surprised if that's still the case. they have built a nice business around that database these days.
22:59:12elinenbeSince there already is a patch and the code is written, it could be an option.
22:59:53adi|homeelinenbe: im not annoyed.. I just find it said.
23:00:02adi|homethe project leader has said _no_ and they still fight
23:00:09Zagorelinenbe: just because code is written doesn't mean it should be added
23:00:15*adi|home nods
23:00:38elinenbeZagor: agreed. That is what the Rockbox Unofficial Fanpage is for. :)
23:00:58elinenbewhere people hack the hack
23:01:05 Join _Dino_ [0] (~dan@
23:01:48elinenbeeverything will be compiletime, and you can set all the option in a form on the web page, and then it will compile your binary.
23:02:04elinenbethen you download it and save it to your jukebox
23:02:22Zagornot a bad idea. start hacking perl scripts :-)
23:03:09_Dino_Nice work people, I was very impressed when I tried rockbox today... very very impressed
23:03:11Zagorstill. too much options, even compile time ones, makes the code messy
23:03:19Zagor_Dino_: thanks
23:04:06elinenbe_Dino_: please send money! :)
23:04:08_Dino_archos should ditch their own os and go with this one, I wont use archos' again, unless this one crashes my hd!
23:04:37_Dino_how much do you want?!
23:04:50elinenbe_Dino_: I am just kidding
23:04:51Zagor_Dino_: oh, then don't press STOP+MENU+PLAY... ;-P
23:04:56_Dino_set up a pay pal account or something
23:05:09Zagor_Dino_: paypal only works in the us
23:05:28_Dino_oh, damn you blew my excuse already!
23:06:23elinenbe_Dino_: on the recored if you press up,up,down,down,left,right,left,right,B,A,B,A,select,start then you can play Tetris, Sokoban, or Wormlet!
23:06:53_Dino_yeah right!Not om my Studio20...
23:06:57elinenbeZagor: have you seen the other nibbler game (eat the dots game) that is out there for rockbox? It is really nice
23:07:35_Dino_Got a background in games, might get stuck into the src tomorrow... it looked pretty tidy
23:07:54Zagornope, haven't seen it. I have 21 fricking patches in my queue, and a fat driver to write... :-)
23:08:18elinenbe21 patches in your queue?
23:08:38Zagoryup. from the mailing list.
23:12:27 Join idefx [0] (
23:13:14idefxHi, i'm living tomorow (for paris, for those who read my mail...), well, before i go, how does the playlist work in latest build ??
23:13:29elinenbeit works good.
23:13:44idefxelinenbe : and how do I use it ??
23:14:34idefxlast time i tried (2 hours ago), when I hited play on a playlist, it displayed me the number of files, and then it didn't do anything.
23:15:05idefxmaybe my playlist format wasn't good ??
23:16:25elinenbeI guess. That is the way it should work
23:18:38idefxelinenbe : i've checked, in fact, i used a xmms playlist, and the file format isn't the same as the one i used before ;)
23:22:07idefxwhahou!, it's thousand's time faster than stock-playlist gestion !
23:22:39 Join TotMach3r [0] (
23:23:20EnnaNhmm, seems that if i press menu-play i find a mute mode? didn't find that in any of the faqs or other files..any other surprises 2b found?
23:23:42ZagorEnnaN: MENU+STOP is keylock
23:23:57EnnaNyep, found that out also :)
23:24:05 Nick seb-away is now known as _seb_ (
23:24:16Zagorkeylock is in the manual, though
23:24:24EnnaNzagor: that's why i tried menu+play in the first time..
23:24:39EnnaN(**that was not really correct english though..**)
23:24:43Zagortry MENU+ON
23:25:00EnnaNlemme see
23:25:08Zagorthat's an id3 tag browser
23:25:37Zagorour manual is embarassingly behind
23:26:19EnnaNbut i like the 2-line id3 tag thing better...artist first line, trackname 2nd..
23:26:34Zagorroll your own in /.rockbox/default.wps
23:26:34EnnaNonly thing, is it normal the artist doesnt scroll?
23:26:47_Dino_yeah, no artist scroll :(
23:27:08EnnaNzagor: could xplain that default.wps or point me to a file with info?
23:27:39_Dino_did archos have a key lock?
23:27:49Zagor_Dino_: yes they do
23:28:07_Dino_oh, well I never read the manual anyway...
23:28:18EnnaNlike -that- is in the manual?
23:29:56_Dino_can I feel the hd spin when it is charging, or am I going mad, always seems too loud when it's charging.. do you have any control over this?
23:30:36Zagor_Dino_: the archos firmware spins up the disk when charging is complete, to try to minimize battery damage. the players can't turn off the charger...
23:31:09*EnnaN still in love with rockbox for being ULTRA fast with playlists loading
23:31:10 Quit xam ("ChatZilla 0.8.7 [Mozilla rv:1.0.0/1]")
23:31:44EnnaNhow is this playlist speed done? first day: more than a minute to load 999 file-playlist...
23:31:48Zagorit's not just fast. we support 10x larger lists too
23:31:48EnnaNnow, sec or so....
23:31:57EnnaN:) i know and like
23:32:20Zagorthe question is not "how can we load lists so fast". the question is "wtf was archos doing"
23:33:13idefxwell, before going to bed, nothing yet to create/modify playlist on device ??
23:33:15Zagoractually, I know what they are doing. they are looking up the position of each single file while loading the list.
23:33:29Zagoridefx: not yet. i'm working on the necessary file writing code
23:33:40_Dino_Zagor : so that's a good thing then?!
23:34:04Zagor_Dino_: the spin-up, yes. but you should probably disconnect the charger when it happens.
23:34:20_Dino_Zagor: Will do..
23:34:38EnnaNzagor: why should you want to do that? to see if a file is there, and if not, remove it?
23:35:06_Dino_When will I be able to use some crazy piece of kit to flash the os in my player so it boots faster?
23:35:15_Dino_or is it rom?
23:35:42ZagorEnnaN: I don't know why they do it this way. obviously, we didn't think it was the best method... :-)
23:35:55idefxZagor : i forgot, nothing yet to write files ;) Ok, thanks, the playlist support may be sufficient to make me happy during my 'trip' ;). After that, i will add a good database and an efficient search function, for the next time i'll get an internet connection ;)
23:36:05Zagor_Dino_: we're not planning that. it's a bit too risky.
23:36:08EnnaNzagor: vvery big DUH
23:36:25 Quit idefx ("(long life to rockbox)")
23:36:34_Dino_Zagor: Doable though?
23:37:03_Dino_Sorry i should read some stuff before I ask these questions...
23:37:12Zagor_Dino_: everything is possible... you'd have to desolder the ROM chip (surface mounted) and flash it outside the device.
23:37:29_Dino_Zagor: Funk that!
23:37:31EnnaNi was in here earlier, talking about enqueueing numbers as "next" in playlist ... i know that it is in the request-list, but if i should help/think about it.....
23:37:33Zagormajor hassle for just a few seconds faster boot
23:39:45ZagorEnnaN: we've pretty much decided how that will be implemented
23:40:07EnnaNzagor: you did?? yez! how, when, feed me info! :)
23:40:35 Quit TotMacher (Read error: 110 (Connection timed out))
23:41:39Zagorwe'll create a queue playlist. enqueued files go into that, and each next-track first checks the queue list before the regular list
23:42:13Zagorthe queue playlist will work just like the ram (dir) playlist, only smaller. max 16 tracks or 1KB paths
23:44:18_Dino_does the ram have any distinction between code and data space? Does a smaller executable mean a bigger buffer?
23:45:14Zagor(to the last question)
23:45:17EnnaNzagor: any prognosses when someone is being a hero and implement this?
23:45:42_Dino_Zagor: cheers
23:45:49ZagorEnnaN: linus expressed interest, but had other things with higher prio right now.
23:46:29Zagor_Dino_: the executable is only a small portion though. the ram used by the program for other things than mp3 buffering is more important
23:46:56Zagorwe currently spend almost 250 KB on code and non-mp3 data
23:47:07Zagorwhile the executable is just 90 KB or so
23:47:22_Dino_Zagor: that's no good
23:47:42_Dino_Zagor: get rid of malloc, it's rubbish and a pain in the arse to debug
23:47:49Zagorit's still a lot better than archos :-) (who spend more than 512KB)
23:47:56Zagorwe have no malloc
23:48:18_Dino_Zagor: swear I saw it in the src, sorry
23:48:20Zagoror, we have it in cvs but we don't link it
23:48:28_Dino_Zagor: good
23:49:16 Quit Synthe ("Connection Lost - Excess Blood")
23:49:21_Dino_Zagor: gcc3.1's made some savings for me lately..(on another processor) have you given it a go yet
23:49:55Zagorsomeone tested it and gained a few hundred bytes
23:50:19Zagormost of the ram is used for data structures, such as the much-debated dir buffer
23:54:01 Join PsycoXul [0] (
23:55:16 Join Synthe [0] (
23:58:13_Dino_Zagor: Dissapointing saving... still as long as the situation is better than archos'

Previous day | Next day