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

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

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

#rockbox log for 2005-01-17

00:02:56jyp_The patch has been submitted
00:06:42 Join Cassandra [0] (~Christi@
00:08:49***Saving seen data "./dancer.seen"
00:17:32Zagor"const long zerocluster = 2;" looks a little unnecessary :)
00:18:19Bagder"use const, not #define"
00:18:52Zagoryeah, but i was mainly thinking about the type. it's not likely to exceed 16 bits...
00:20:51crash_Zagor: Bagder i was fooled too, but at a closer look his emailadress let me think it might be a fake
00:21:00crash_but would habe been nice ;)
00:22:59jyp_Zagor: Might be unnecessary but hurts not...
00:23:15jyp_Consitent with cluster => long
00:23:29Zagoryeah, but it's always 2...
00:24:23Zagoralso you only changed one of the declarations of it
00:24:38jyp_The other is in fat16 iirc
00:24:56jyp_and I thought fat16 => short
00:25:01Zagoryes. but we can't have only half the code 16-bit safe. it would be a mess.
00:25:41Zagorhmm, you're right. damn ;)
00:26:32jyp_heh ;)
00:26:42jyp_I wasn't certain ;)
00:26:57gromit`peter parker :))))))
00:32:03Zagori'm off to bed. see you tomorrow.
00:32:04 Quit Zagor ("Client exiting")
00:33:52amiconnBagder: Do you know how the iRiver button assignments in rockbox correspond to the real iRiver buttons?
00:34:16amiconnI'm not sure about the MODE, SELECT, ON and OFF buttons.
00:35:27amiconnI guess SELECT is pushing the joystick. The 3 right buttons have symbols that look like Play/Pause, Stop and A/B to me.
00:36:09Bagderthe left one says 'rec'
00:36:15amiconnI can't image how this corresponds to ON, OFF and MODE
00:36:31Bagderwe'll just have to settle on something
00:36:37amiconnYes, REC is obvious, as are the direction keys.
00:36:50amiconnI guess Linus did this assignment?
00:36:58 Join quelsaruk [0] (~kvirc@
00:37:01amiconnI need that for the player pic background
00:37:08amiconnhi quelsaruk
00:37:09Bagderamiconn: I think Linus did it, yes
00:37:18Bagderpossibly based on Zagor's wiki page
00:37:33*amiconn checks
00:39:09amiconnBagder: Hmm, the wiki page (ButtonAssignments) does also mention the symbolic names only :(
00:39:28Bagderyes, I see...
00:42:38Bagderwe'll need to confront Linus with it
00:43:52 Join Christi-S [0] (~Christi@
00:45:19Christi-SSo, anyone have any idea when my recording screen patch will get looked at?
00:45:57BagderI don't even have an archos atm
00:46:22Christi-SBummer. That must be slowing things down.
00:46:46Bagderactually, I've been rather slow in Rockbox for months now
00:47:18Christi-SAny particular reason, or just life getting hectic?
00:47:42Bagderjust life and time, the usual stuff
00:48:42*Christi-S nods.
00:48:54BagderI'm on paternery leave for the moment
00:49:00Bagderit leaves less time for hacking
00:49:09Christi-SI think I'm about to hit a busy patch - I'm worried the documentation may suffer. :(
00:49:18Bagderwhile working I could do more during work hours ;-)
00:49:29Christi-SCongratulations, and commiserations in whatever proportions you deem appropriate. ;)
00:49:50Bagderits generally only goodness
00:50:00Bagderjust less time for computer-related stuff
00:50:15*Christi-S nods. Priorities change.
00:50:38*quelsaruk grins
00:52:49amiconnquelsaruk: How's the spanish voice file coming? ;-)
00:53:07quelsarukit's in my computer...
00:53:44quelsarukbut my screen continues broken, and i'm a bit lazy to try to make it again with the laptop
00:53:56quelsaruki haven't even tested
00:54:27quelsarukso i don't know if the voice speaks nice spanish or is somehow like a foreigner speaking spanish
00:54:49jyp_Like, mexican ? ;P
00:56:15Christi-SI was looking for a nice British voice the other day. I came across one vendor who had clearly never heard a British person actually speak.
00:56:20*jyp_ should remember to stfu
00:56:49Christi-SI'd be tempted to buy AT&T natural voices except that they want to tax me at least $30 for not being American.
00:57:59jyp_Better not take any risk with could-be terrorist.
00:58:17*jyp_ just fumbled again.
00:59:16quelsarukjyp_: like dwihno when he speaks spanish :D
00:59:32quelsarukit's the same that if you hear me speaking english..
00:59:42quelsarukyou notice i'm not english...
00:59:59Christi-SNow you see, if everyone just spoke English, it'd all be fine.
01:00:07jyp_I for one probably cannot speak either correctly.
01:00:12Christi-SProper English, that is. *British* English.
01:00:25quelsarukChristi-S: everyone should speak spanish
01:00:55quelsarukwe had our chance to conquer the world, but failed...
01:01:04jyp_I can only speak euro-english & utter "Me voy a la playa".
01:01:06Christi-SWell, everyone can speak Spanish as their second language if you like.
01:04:17jyp_time to sleep
01:04:20*jyp_ waves
01:04:23 Quit jyp_ ("Leaving")
01:04:36preglowi've got a guy yelling in spanish outside my window right now
01:04:44preglowi'd prefer he stopped it
01:05:06Christi-SShhh. Don't say it too loud or everyone will want one.
01:05:40preglowthe fact that it's 1am doesn't seem to bother him
01:05:58amiconnBagder: What is the proper way to correct a (still pending) "cvs add" when I accidentally forgot to specify "-kb" (binary)? Simply reissue "cvs add -kb" ?
01:06:07quelsarukwhere is that preglow?
01:06:13preglowquelsaruk: norway :P
01:06:19 Quit Cassandra (Read error: 110 (Connection timed out))
01:06:31Bagderamiconn: I think I would try cvs rm first, then add it again
01:06:34preglowi'm living in a student village, lot's of exchange students
01:06:38preglowlots, even
01:06:49quelsarukand he is yelling in spanish??... these spanish guys... you should kill all spanish people :P
01:07:07preglowquelsaruk: i'll tell him you gave me spains permission
01:07:15quelsarukof course
01:07:41preglowconsider him dead, then
01:09:18amiconnBagder: Hmm. If I leave the files there, "cvs rm" tells me that I have to remove them first. If I take them out, "cvs rm" then says "know nothing about <file>". Trying "cvs add" again says it's already added...
01:10:08Christi-SIt's like Schroedinger's CVS committal.
01:10:33amiconnAh, found it. I need to name each file individually on remove.
01:13:38Bagdertime to sleeeeep
01:15:15preglownow, what the hell would make a method with no 'return's in it exit before the end of the method?
01:15:24preglowthis is the queerest bug i've had happen to me in a long time
01:16:16Christi-SStack corruption?
01:16:39Christi-SUsually the thing behind really weird bugs.
01:17:00preglowyes, i think i've covered that already
01:17:11preglowsince that usually what my conclusion is as well
01:17:17preglowand it's almost exclusively correct ;)
01:18:19preglowi've even bloody disassembled the code to make sure gcc hasn't been on crack while optimizing again
01:19:22quelsaruki think is sleep time even here in spain... cu other day
01:19:47quelsarukamiconn: when i get my screen back i'll show you my spanish voice
01:20:19 Quit quelsaruk ("i seek a life of honor free from all false pride...")
01:28:18preglowgreat, three hours of debugging and i now find out it's the host application's fault
01:28:43 Quit MooMaunder (Remote closed the connection)
01:28:48preglowi'll just go cry myself to sleep, then
01:39:06 Quit preglow ("gah")
01:44:20 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
02:00:10 Quit amiconn (" nite")
02:08:52***Saving seen data "./dancer.seen"
02:39:37 Join stevenm [0] (
02:39:46stevenmHi. Is anyone working on the Calculator plugin ?
04:08:53***No seen item changed, no save performed.
04:20:44stevenmI have a little piece of code here that can quickly (and accurately) find sin(x), cos(x) and e^x (and potentially other stuff too) that can be integrated into the Calculator plugin for Rockbox
06:08:54***No seen item changed, no save performed.
07:02:29 Quit einhirn (Read error: 104 (Connection reset by peer))
07:11:16 Quit stevenm ("Leaving")
07:30:13 Join einhirn [0] (
08:08:58***Saving seen data "./dancer.seen"
08:59:21 Join LinusN [0] (
09:06:13 Join Zagor [242] (
09:08:17crash_oink oink
09:13:20dwihnomeow meow
09:14:29Zagorthe rockbox zoo
09:15:08LinusNdoes anyone know what the problem is with the cyrillic support?
09:19:50 Join midk [0] (
09:19:50 Join Zagor_ [242] (
09:19:50 Quit LinusN ("disconnecting from stoned server.")
09:19:50 Join LinusN [0] (
09:19:50Zagorlast i hear we were using the iso character set, while most russian people use the windows charset
09:19:50Zagorbut that was long ago. is there a new report?
09:19:50 Quit Zagor ("disconnecting from stoned server.")
09:19:50 Nick Zagor_ is now known as Zagor (
09:19:50LinusNin the forum
09:19:58LinusNso we garbe up the file names?
09:21:08Zagorhe's using a studio. afaik we have never supported cyrillic for those.
09:21:50 Quit midk (Remote closed the connection)
09:22:05LinusNah, didn't see that
09:26:25LinusNpisses me off
09:33:04LinusN...and now those poor gmini users will suffer too :-)
09:35:41Zagori seriously doubt there a country where a manufacturer is allowed to unilaterally and without notice alter the warranty
09:37:02LinusNi think it might be worth checking
09:37:10Zagorhowever in most countries it will cost more than $68 to prove it...
09:37:32LinusNnot necessarily
09:37:55LinusNjust call Sverker (swedish joke)
09:45:21 Quit Nibbler (Connection timed out)
09:49:48 Join Lynx_ [0] (HydraIRC@
09:51:05dwihnoLinusN: He'll throw the archos' in his red trashcan :(
09:53:23Christi-SOooh, I didn't realise I was logged in here.
09:53:56Christi-SLinus - Any idea when you might get around to looking at my recording on start patch?
10:03:23LinusNthat will cost you :-)
10:03:46*dwihno sends LinusN to Sverker. "Är det så? Är det så?"
10:04:06LinusN"skall det vara på det viset, tycker du?"
10:05:35*Bagder looks at LinusN an dwihno
10:05:43dwihno"Nu pratar du teknisk mumbo-jumbo, som konsumenten inte förstår"
10:09:02***Saving seen data "./dancer.seen"
10:18:59 Join MooMaunder [0] (~me@
10:19:19 Quit MooMaunder (Client Quit)
10:19:19 Join MooMaunder [0] (~me@
10:22:56Christi-Serm, i have a half eaten slice of marmalade on toast. Will that do, Linus?
10:26:04LinusNcertainly, fax it right away
10:30:39Christi-SUnfortunately I finished eating it before you replied. Life is harsh sometimes. I could fax you my stomach contents, I suppose.
10:31:08LinusNjust puke on the fax
10:31:30LinusNor fax an xray shot
10:32:11Christi-SYour wish is my command.
10:32:28*LinusN unplugs his fax
10:36:53 Join Nibbler [0] (
10:49:52 Join [IDC]Dragon [0] (
11:13:57 Join quelsaruk [0] (~kvirc@
11:14:03quelsarukhi, good morning
11:14:29LinusNmorn quelsaruk
11:15:05quelsarukmur da bra LinusN? (or mar du bra... y always forget)
11:17:05LinusNjag mår bra
11:19:53[IDC]DragonI only know Smörebröd
11:19:56quelsarukso then it's "mår du bra?".. isn't it?
11:20:22LinusN[IDC]Dragon: ...which isn't a swedish word :-)
11:20:33quelsaruki wonder you you write å
11:20:53LinusNwell, i use the å key on my swedish keyboard :-)
11:21:25quelsarukhmm... that's not fair
11:21:55 Quit Ka (Nick collision from services.)
11:23:08Lynx_Hows the tag database coming along?
11:23:21Zagorvery well. i'll commit the first version soon.
11:23:44 Join Ka_ [0] (
11:23:46Lynx_Zagor: cool :)
11:24:22[IDC]Dragonthen I really need to tidy up ID3 tags :-{
11:24:47Zagor[IDC]Dragon: no rush, you can't play songs from the id3 browser yet anyway :-)
11:25:41Zagori'm committing an early version to let interested people have a peek. it's far from completed.
11:25:50Lynx_[IDC]Dragon: since MusicBrainz and 'godfather' tidying up the tags is not too much effort any more...
11:26:00Nibbler <−− great tool for renaming/id3tagging mp3s
11:26:01[IDC]Dragonon a wider scope, what did you do to tree.c
11:26:13Zagora lot :)
11:26:38[IDC]DragonI'm waiting for that, to perhaps add dynamic mounting
11:27:12Zagorall file-specific code (nearly) is moved to filetree.c
11:28:01Zagortree.c: 416 insertions(+), 801 deletions(-)
11:28:19Lynx_Nibbler: that sounds also good, does it find stuff with no info on the files at all? like song1.mp3 ?
11:28:44Zagorthere's a new "tree context" structure introduced, for passing information between tree.c and dbtree.c/filetree.c
11:29:12NibblerLynx_: yes, it uses - but you MUST have the mp3s with index numbers b4 the title (can be after the artist) so that if you order them alphabetically you get them in the right order
11:30:03Lynx_Nibbler: ah, ok. musicbrainz may have a little advantage then, it also finds single songs
11:30:29Nibblerwow nice1. but is there a linux commandline client for it?
11:31:58Lynx_Nibbler: not an official one, it seems. some of the kde players/taggers support it's database, though
11:32:36 Quit [IDC]Dragon ("CGI:IRC")
11:32:43Nibbleroh well, i can check whatever is leftover by mp3cddb from within windows
11:39:59 Join amiconn [0] (
11:41:40crash_LinusN: are you planing to implement a statistic function?
11:41:52crash_like: songx was playe 10 times
11:42:01crash_or song y was played 20 times
11:42:11 Join [IDC]Dragon [0] (
11:42:13crash_so you could find out your favorite songs ;)
11:43:18[IDC]Dragonoh, Jens is here meanwhile :-)
11:44:07amiconnLinusN: I found a problem within lcd-recorder.c for iRiver. Now I'm looking for a special preprocessor hack:
11:46:42 Quit quelsaruk (Read error: 110 (Connection timed out))
11:53:50amiconnThere are initialized arrays, which need to be variable size (compile time) depending on the display height. How do I do this?
11:54:11amiconnI.e. I want something like this:
11:55:16LinusNfyi: lcd-recorder.c for iRiver is named lcd-h100.c
11:55:58amiconnNot for the sim...
11:57:46amiconnif I #define ARRAY_LEN 5, the preprocessor should resolve the array definition to char array[5] = {0xFF, 0xFF, 0xFF, 0xFF, 0xFF};
11:58:04amiconnif I #define ARRAY_LEN 3, the preprocessor should resolve the array definition to char array[3] = {0xFF, 0xFF, 0xFF};
11:58:43[IDC]Dragonso make the end values conditional
11:59:53[IDC]Dragonamiconn, different subject: we need a _norom firmware for the player, did you see that?
12:00:26amiconnHow? This can get very longish when done for larger numbers. Currently I have a kludge (only 2 different sizes actually occur), but I would like to know if this is possible to solve with the cpp for the general case
12:00:47amiconnYes, saw that.
12:01:01[IDC]DragonI meant for 2 or few sizes
12:01:25[IDC]Dragonno preprocessor loop ;-)
12:01:33crash_LinusN: did you see my question? :)
12:01:53LinusNstatistic function? for the db?
12:02:57crash_yes :)
12:03:09crash_like amarok at the pc
12:03:36crash_i think this would be a great feature, to see which songs you really use ...
12:03:43LinusNsomeday someone will implement that
12:04:01LinusNthere is already a patch that implements a simple log file
12:04:04crash_ok, juste wanted to know if this is something you have in mind while coding the db
12:04:10crash_maybe i'll have a look at it
12:04:13LinusNthen you can use an offline tool to analyze it
12:04:16crash_when your basics are done
12:04:31LinusNthe db is simply a static id3 database, mainly for browsing
12:04:54crash_like the iriver one i know
12:05:07crash_i think i mixed some features here :/
12:05:15crash_will have a look at the patch
12:05:20LinusNa statistic function doesn't necessarily need an id3 database
12:05:29crash_yeah i know
12:05:36crash_file name is more relevant
12:06:40Zagorwe have explicitly decided against mixing statistics and user marks into the id3 database. the id3 database is read-only.
12:07:17*LinusN goes to lunch
12:09:04***Saving seen data "./dancer.seen"
12:12:04amiconn[IDC]Dragon: (player_norom) It would be a good idea to test this before publishing. I wonder how to do this...
12:13:44Lynx_Zagor: how long does it take to build the index, and does it have to be rebuilt completely if one song is to be added?
12:15:05 Join lImbus [0] (
12:15:05ZagorLynx_: currently there's only a "create" function. but an "update new files" is not unthinkable.
12:15:19Zagorspeed naturally depends on disk and computer speed
12:15:37Lynx_Zagor: ah, so it's never built on the archos...
12:16:09Lynx_then i guess it does not take too long anyway
12:17:03 Join quelsaruk [0] (~kvirc@
12:20:42 Join methangas [0] (
12:45:38*Zagor braces for red builds...
12:47:56Zagorow, several sims broke
12:48:59Zagorhmm, i tested recsim
12:49:19Zagorgah, forgot to commit sim makefiles
12:53:42 Join ashridah [0] (
13:04:15 Quit Christi-S (Read error: 110 (Connection timed out))
13:05:34quelsaruki'm making one test, can anyone tell me if he/she can see these japonese writing and which irc client is he/she using? Thanks :)
13:06:46[IDC]Dragonamiconn: I once tested _norom on my FM torso
13:07:00 Join ripnetuk [0] (
13:07:18[IDC]Dragonbut not for recorders, that was a "blind shot"
13:07:21ripnetukIve been playing with the iRiver sim, which has made me even more impatient :) any gnews on the bootloader???
13:07:31Lynx_quelsaruk: only question marks visible...
13:07:48ripnetukyou can see SO many songs at once on the larget screen
13:07:58quelsarukthx Lynx_ (using mirc, isn't it? )
13:08:14Lynx_quelsaruk: no, hydrairc
13:11:39[IDC]DragonZagor: did you use an older version of tree.c for your filetree.c?
13:11:40amiconn[IDC]Dragon: Would you say, since _norom works for both recorders and fm, it will work for players for sure? There's no security net without boot rom...
13:12:12[IDC]Dragonamiconn: the added security is minimon in the bootloader
13:13:11[IDC]Dragonit should be rather safe, since the _norom part is nothing which got modified while adapting the Player
13:14:01amiconnHmm, then perhaps I should simply make a _norom version, and add another big fat warning...
13:14:23[IDC]Dragonperhaps let somebody volunteer
13:14:34[IDC]DragonI did the same with the recorder
13:15:34[IDC]Dragondid you test the minimon in the bootloader on a player?
13:18:06amiconnI did; it worked.
13:18:40amiconnOnly uart_boot doesn't allow to switch the rom-based minimon to the higher bitrate
13:18:57[IDC]DragonI know
13:19:06amiconnI still wonder why you chose 14400 bps as default when 38400 bps are possible
13:19:10Zagor[IDC]Dragon: are you missing something?
13:19:33[IDC]DragonZagor: I added a few lines for multivolume
13:20:00[IDC]Dragonin the quicksort compare function
13:21:19amiconnZagor, [IDC]Dragon: Shouldn't cvs warn if you try to commit and the central version has changed meanwhile, and no allow committing?
13:21:26Zagorok. i'll dig it up and reinsert it
13:21:30Zagoramiconn: this is a new file
13:21:46[IDC]Dragonthere's more cvs traffic after my stuff
13:21:51amiconnArgh! Of course then this doesn't happen.
13:22:25[IDC]DragonZagor: do you know when you've branched?
13:23:11[IDC]DragonJens did 2 other fixes to tree.c after multivolume
13:23:12Zagorstill, cvs should complain when it can't update a file so this shouldn't happen without notice.
13:23:27Zagori'll check them all
13:24:39[IDC]Dragonmaybe other files are affected, too, depends how long you've worked "in isolation"
13:24:57Zagori haven't worked in isolation. i've ran cvs update for all commits.
13:25:59[IDC]Dragonfor the new files, I mean, which contain code moved from the existing
13:26:15Zagoronly your multivolume patch was lost
13:26:34[IDC]DragonJens' fixes are in?
13:26:58[IDC]Dragoncvs didn't like me...
13:27:31Zagorjens patches were on code that is still in tree.c
13:28:54[IDC]Dragoninteresting scenario, from a configuration management point of view
13:29:18Zagorindeed. i still think there must have been a complaint from cvs that I have missed
13:29:53[IDC]DragonI don't see how cvs could tetect moved code
13:30:03Zagorit fails to apply the patch
13:31:03[IDC]Dragonbut the diff would just say "remove this big block of file A, add another big block in file B"
13:31:37Zagorno, your patch says "insert this code in file A, between lines X and Y". and when it can't find X and Y in file A, it will give up and complain to the user
13:31:44[IDC]Dragonno cross reference that it's the same or a similar block
13:36:44[IDC]DragonI see, so it won't delete the block in file A if the block in no literal match
13:37:22Zagorexactly. it will fail if it can't apply the patch cleanly.
13:37:47ashridahcvs can fudge things a little if the patch applies cleanly a few lines down
13:38:02ashridahat least, '
13:38:04ashridahpatch' can
13:38:10ashridahcvs might be a bit more strict
13:38:55amiconncvs does the same as patch for checkout, but most likely not for checkin
13:40:06Zagorashridah: yes
13:40:25Zagoramiconn: checkin never patches. only update does.
13:40:50ashridahamiconn: checkin will just assume that any off-by-ones are part of the patch to be checked in
13:42:33amiconnYes of course, but it refuses to check in when the central copy changed meanwhile
13:43:56amiconn..instead of moving the lines to apply them cleanly to the central copy
13:44:22Zagorexactly. you need to run update first.
13:45:00amiconnZagor: You broke the Win32 sims...
13:45:32Zagoryeah, i noticed. haven't looked at it yet.
13:46:00[IDC]DragonZagor: don't worry about our nagging. Thank your for the ID3 browsing!
13:46:36Zagori never worry ;)
13:46:58Zagortried it yet?
13:47:43Zagorbtw, the database file goes in the .rockbox dir
13:47:57[IDC]Dragonhere I can't try
13:48:06[IDC]Dragonno cvs
13:48:22amiconnZagor: I could try it now, if the win32 sims would build... ;-)
13:48:34Zagorhehe. i'll take a look.
13:49:02amiconnI'll check im parallel. Some really odd errors..
13:49:44[IDC]DragonI'd need a feature like a bleeding edge source tarball
13:50:56Zagoramiconn: aha, i've broken oldC compatbility by declaring variables below code
13:51:54amiconnCode police! ;-)
13:52:00[IDC]DragonI could at least run the perl script on my box, in preparation
13:52:11amiconnBut I wonder why; don't you use gcc to build the win32 sim?
13:52:59Zagori do, but version >3 doesn't complain about that. only <3 does. and only the win32sim gcc is <3
13:56:11amiconnI'm building with gcc 3.3 (on cygwin). No warnings/errors here.
13:57:06Zagoryeah, i'm using an old 2.96 mingw version since I need it to cross-compile
13:57:22[IDC]Dragonfrom : use MP3::Info;
13:57:33[IDC]Dragonwhat's that?
13:58:13[IDC]Dragona reason for failure here
13:58:31Zagorit's the id3 library
13:58:49Zagorwe'll need to extract the .pm file and distribute it alongside for it to work on win32
13:59:01[IDC]Dragonsome bunch of perl I don't have?
13:59:04amiconnZagor: How am I supposed to build such a database? Is there some documentation?
14:00:43Zagoramiconn: documentation? what's that? ;)
14:00:48Zagori'll add it to the wiki
14:01:12[IDC]DragonI can't reach that :-(
14:01:24Zagorme fix
14:01:25[IDC]Dragonah, now, finally
14:03:39[IDC]Dragonwhere should I place it?
14:03:55Zagorwait a bit, i'll adjust
14:04:44[IDC]DragonMP3 subdir, ok
14:06:30[IDC]Dragonit's running...
14:06:53[IDC]Dragonfor how long on 60 GB?
14:07:22Zagora few minutes
14:08:00[IDC]Dragonshould I see a file growing somewhere? or does it write all at once?
14:08:23Zagorall at once, since it needs to grab and cross-index all data
14:09:08***Saving seen data "./dancer.seen"
14:10:10[IDC]Dragonah, finished
14:11:28[IDC]Dragonit prints some max length, so your field sizes are tailored to the worst case?
14:14:59amiconnWhat are the maximum lengths? I have some songs with very long names....
14:15:31Zagorthere is no maximum
14:16:41Zagorother than it probably wont work with >2GB database :)
14:18:33amiconnDoesn't work here :(
14:18:41Zagorwhat happens?
14:18:43amiconn(win32 sim)
14:18:55amiconnI get the following output in the debug window:
14:18:56 Quit ashridah ("sleep")
14:19:11amiconndb_load(0, 0)
14:19:12amiconn0 read(603979776) returned -1
14:19:12amiconnNo filesystem found. Have you forgotten to create it?
14:20:47LinusNwell, have you? :-)
14:21:28Lynx_can i get the perl script somewhere, too?
14:21:41amiconnLynx_: It's in cvs
14:22:30Lynx_hmm, don't really know how to use cvs...
14:23:11[IDC]Dragonyou can use the web frontend for just getting files
14:23:35[IDC]Dragonthe script is here:
14:23:58Lynx_ok, thx
14:24:20[IDC]Dragonor just click on the file from front page cvs activity
14:25:16Zagornow MP3::Info is included in
14:27:53[IDC]DragonI grabbed the files one by one and tried to compile
14:28:02[IDC]Dragonbut must have missed one
14:28:23[IDC]Dragonwhere e.g. is db_load() ?
14:28:34Zagorin dbtree.c
14:31:15Zagorgosh, "print STDERR" is not displayed on win32 machines! /me change
14:31:57[IDC]Dragonah, SOURCES got a .txt from the download
14:36:33[IDC]Dragonhmm, it hags when I select view database from F2 menu
14:37:22Zagortold it wasn't complete ;)
14:37:28Zagortold you
14:37:42Zagoron the win32 sim?
14:37:47[IDC]Dragonwhen rebooted, it still hangs :-(
14:37:56[IDC]Dragonno, RecV1
14:38:14[IDC]Dragonhe red LED is flashing rapidly
14:38:18Lynx_Zagor: probably not important and win-specific: opendir d:\/System Volume Information: Invalid argument at line 62
14:39:11[IDC]Dragonguess I have to drive home with no music today
14:39:23Lynx_Zagor: and the stuff printed to stderr is printed on my win...
14:39:24Zagor[IDC]Dragon: just clear the RTC ram and it will fallback to file browser
14:39:48[IDC]Dragonwe still lack that button combo
14:40:05Zagoryeah, but it's easy on the v1s
14:40:07[IDC]DragonI know, I can use 2.4
14:40:18[IDC]Dragonor pull the cell
14:41:19Zagorwhat is the max numbers for your database?
14:41:40ZagorLynx_: is that some hidden/restricted directory, perhaps?
14:41:47[IDC]Dragon599 artists, 901 albums, 9183 songs
14:42:25[IDC]Dragonmax artist:48, album 68, song 116, songs/album 64
14:43:04quelsaruk[IDC]Dragon: do you say that red led is slashing after you reboot?
14:43:14Lynx_Zagor: i'm not sure what that was again, i think it's windows recovery info in there. It'll just happen to every windows user who runs the script on a drive root, i.e. d:/, maybe some people have an extra partition for their music...
14:43:17Zagorlooks about like my numbers
14:43:52ZagorLynx_: yeah, i'll make it stop dying on that error
14:43:53quelsaruki got that error, checked and was HDD error, affecting the .rockbox folder... (that's how i found out i had bad sectors..)
14:44:08[IDC]Dragonquelsaruk: yes, it shows disk activity when booting, then goes straight int flashin
14:44:24[IDC]Dragoninto flashing
14:44:54quelsaruk[IDC]Dragon: do you have archos firmware also in flash? can you boot using the secure boot option??
14:45:13[IDC]Dragondon't worry
14:45:24amiconnLynx_, Zagor: "System volume information" is protected from access even for administrators. Since it is also hidden, should not try to enter it anyway though...
14:45:47Lynx_Zagor: actually it's important to change that, because windows will put the system volume info folder onto the box also when connected via usb, and running it on the root of the box will be the default for most users
14:46:13Zagoryes, fix coming up
14:46:14quelsarukdoes anyone worry about you [IDC]Dragon? ;) amazing....
14:46:18amiconn[IDC]Dragon: Btw, with my long song names, I once ran into the talk queue length limit with file name spelling
14:46:48[IDC]Dragonany harm?
14:47:05 Join R3nTiL [0] (~zorroz@
14:47:37amiconnLynx_: (svi folder) (1) You can prevent windows from putting this folder, by disabling system restore for the drive. (2) You won't run into that error anyway, because there are no access restrictions on FAT32.
14:49:28Lynx_amiconn: right, didn't think of that it's no problem on fat32...
14:51:30[IDC]Dragonthe flashing LED error even happens before having a chance to enter USB mode, how that?
14:52:09Zagor[IDC]Dragon: sounds like a build failure. try the bleeding edge.
14:52:50[IDC]Dragonbefore I switched to ID browsing it behaved ok
14:53:43Zagori've been switching back and forth all day without problems... :)
14:53:55[IDC]DragonI'll try again later
14:54:24amiconnZagor: I forgot After the Win32 sim prints those messages, it does nothing, i.e. hangs too.
14:54:43amiconnIt worked before I switched to ID3 browsing, then restarted it.
14:54:50Zagoramiconn: do you have a database?
14:55:07amiconn(Small one, containing only 2 albums)
14:55:09Zagorthe code is cursed :)
14:56:02amiconnWhen I remove the database, it starts normally (but doesn't display any file of course)
14:59:31Zagorsounds like i should add more checks in the initial loader
15:00:58[IDC]Dragonsame on the box: when I rename the database, it starts but is blank
15:02:04[IDC]Dragondo you load it before USB mode?
15:02:54Zagorno, it's loaded in tree.c
15:06:22ZagorLynx_: see the TagDatabase for instructions now
15:09:41amiconnZagor: When I start the sim when it's set to file browsing and the database in place, and then switch to id3 browsing, it displays some errors in the debug window, then crashes (closes)
15:10:30Zagorwhat error messages?
15:11:45Zagorif you remove sector3D.bin and start up again, you should see a listing of the basic database numbers. are those consistent with your database?
15:14:43amiconnThey're totally off, much too large
15:15:08amiconnVersion: RDB82
15:15:08amiconnNumber of songs: 419430400
15:15:08amiconnSongstart: 64060000
15:15:08DBUGEnqueued KICK amiconn
15:15:08amiconnSonglen: 805306368
15:15:09amiconnNumber of albums: 67108864
15:15:09***Alert Mode level 1
15:15:09amiconnAlbumstart: 400c0000
15:15:10***Alert Mode level 2
15:15:10amiconnAlbumlen: 536870912
15:15:12amiconnNumber of artists: 50331648
15:15:14amiconnArtiststart: c00d0000
15:15:16amiconnArtistlen: 603979776
15:15:41amiconnFor real there are 2 Artists, 2 Albums, and 25 Songs
15:16:06Zagorok thats it then. hm...
15:16:23[IDC]Dragonis endianess taken care of?
15:17:04Zagoryes, if LITTLE_ENDIAN is set. usually gcc takes care of that, guess not here.
15:17:05Lynx_Zagor: the .bat file does not work, the script does not take / as a path in windows (or \ ), it exits with "no songs found..."
15:17:44ZagorLynx_: ok
15:18:26[IDC]DragonLynx_: I did "perl −−path e:\"
15:18:42Lynx_[IDC]Dragon: yes, that works
15:18:51Zagor[IDC]Dragon: yes, but i'd like to avoid the drive letter if possible. otherwise everybody has to edit the bat file.
15:19:07Zagormaybe that's unavoidable?
15:19:32Bagderlooooads of tabs in that code
15:19:52[IDC]Dragoncode police!!
15:20:12ZagorBagder: yeah, that's the MP3::Info. i didn't change more than I had to in that code.
15:20:26BagderI think it should be re-intended at least
15:20:31Zagorfeel free
15:24:15Bagderthis is a good comment from the script:
15:24:17Bagder"# whoaaaaaa nellllllyyyyyy!"
15:24:38LinusNNelly Furtado
15:25:11***Alert Mode OFF
15:25:50Zagoramiconn: could you try changing that #ifdef LITTLE_ENDIAN to #ifdef __LITTLE_ENDIAN__ and see if it makes a difference?
15:26:15LinusNthe perl script creates the binary database, right?
15:26:37LinusNand it generates big endian integers
15:26:48 Quit R3nTiL ()
15:27:49Bagderindented now
15:28:07Bagderthere's no perl-intent-line-or-region
15:28:15 Join edx [0] (
15:30:42Lynx_Zagor: $0 =~ m/(^.)/; my $drive_letter = $1;
15:31:54amiconnZagor: Doesn't change anything :(
15:32:27Lynx_Zagor: well, that works for windows, don't know how useful it really is, this will get / in linux i guess
15:34:23ZagorLynx_: best idea so far anyway.
15:34:43 Join webguest43 [0] (
15:35:04webguest43hi ! do you guys have any thoughts on the linux on ipod project ?
15:35:08Zagoramiconn: ok, we'll need to find which defines to use
15:35:20Zagorwebguest43: plenty. which do you want? ;)
15:35:22Bagderwebguest43: I think cool, but pointless ;-)
15:35:42Bagdernot very useful
15:36:36Bagderand it can't even play mp3s properly
15:36:39webguest43just would like to know if you guys think it will fly
15:36:59Bagder"fly" meaning what?
15:36:59Zagorwebguest43: it flies already, doesn't it?
15:37:16LinusNwebguest43: i don't think it will be a replacement for the original firmware
15:37:23webguest43well, i dont know much about this, but once you got a nice linux kernel running applications should be simple to develop ?
15:37:25BagderZagor: not on all models
15:37:36Bagderwebguest43: they are
15:37:38ZagorBagder: true
15:38:11Bagderwebguest43: but to make it play mp3 files properly, you need to make a real effort
15:39:27webguest43webclient verrry slow...
15:39:38Bagderuse a real IRC client! ;-)
15:40:23webguest43is nono at work ;)
15:43:59Zagorhmm, it appears BYTES_BIG_ENDIAN is the correct define to check against
15:45:02Zagorhowever it's not set for sh-elf-gcc. sigh...
15:45:33Bagderjust add a new Rockbox one in the config-*.h files
15:45:46Zagorthat doesn't help the simulator case
15:45:58Zagorsimulators can be on both
15:46:18Bagdercan yes
15:46:28Bagderbut it hardly happens
15:46:29Lynx_Zagor: you could add a −−windowsCurrRoot flag, check $^O for MSWin32, and then get the drive letter as above...
15:47:55webguest43so you guy dont think it will ever be a proper replacement for the apple firmware, even considering the vast amount of linux programming skills out there ?
15:48:43Bagderwebguest43: since it has been around for quite some time now, and they still don't do better, I doubt that
15:49:01Zagorwebguest43: haven't you heard? the ipod firmware is perfect. there is nothing that could be improved. just ask ipod owners.
15:50:28[IDC]Dragonask the folks at #ipodlinux
15:50:55Bagderthey will have a different opinion, I'm sure ;-)
15:51:42LinusNwebguest43: my opinion is this: the fun was to make linux run on the ipod, not to create a replacement firmware
15:54:59webguest43ok, so i can sell the ipod.
15:55:13Bagderyeah, buy one that can run Rockbox ;-)
15:55:31webguest43another one which might be stupid: is it possible to do usb on the go on the iriver ?
15:55:43Zagoronly the h300 series
15:56:37crash_webguest43: i made the experience that it doenst work with all kind of devices
15:56:50crash_my h320 doesnt recognize my casio exilim
15:56:58crash_although it's a massstorage device
15:57:13crash_my archos jukebox wasn recognized as well :/
15:57:24crash_another iriver and a gmini didnt make probs though
15:57:43crash_hope rockbox can do something about this *hint* ;)
15:57:56 Quit methangas (" WOW! This IRC Client ownz! HydraIRC -> <-")
15:58:09webguest43so its a hardware problem in the usb part ?
15:58:46Zagorcan't say. sounds more like a firmware issue.
15:59:56 Nick Lynx_ is now known as Lynx_awy (HydraIRC@
16:01:52amiconnZagor: Maybe endianess is not the problem here. system.h uses #ifdef LITTLE_ENDIAN to conditionall define SWAB32() etc. This does work.
16:03:59webguest43i mean doing usb to go on h140 ?
16:04:10Zagoramiconn: are those ever used by the simulators?
16:08:04Zagoramiconn: can you check if the endianess in the db header looks ok?
16:09:12***Saving seen data "./dancer.seen"
16:11:34webguest43do you guys think it will be difficult to port rockbox to ihp340 once it runs on 140 ?
16:11:54LinusNwebguest43: not very
16:14:57webguest43do you guys think the 340 is worth 150 euros more than the 140 ?
16:16:34Zagorwebguest43: where do you see that? I've found they cost almost the same nowadays
16:16:34 Part LinusN
16:22:53amiconnZagor: How is the header supposed to look like?
16:23:31crash_webguest43: i have the posibility to compare them by each other
16:23:48crash_so i would say the h3xx have a very cool display with color
16:24:12crash_in comparison this is the bigest adantage of the h3xx series to the h1xx ones
16:24:40crash_usb-on-the-go, doesnt work that good for me, ad it wasnt one of the reasons to buy it
16:25:02crash_i think u have to know for yourself if you want to pay for de device with color display ;)
16:25:28crash_and by the way there is no possibility to port the usb-on-the-go to the h1xx cause this is a hardware issue
16:25:58amiconnZagor: The numbers in the db header look ok, big endian.
16:29:14amiconnYou can find it here, for reference:"> (4 KB)
16:32:37Zagorworks here: Marie Fredriksson, Äntligen Bästa.
16:37:34 Join mecraw_ [0] (~lmarlow@
16:44:31amiconnZagor: Which platform?
16:45:05webguest43thanks crash, i like the idea of the dock for 320, but find it weird that it comes without remote ?
16:49:14 Join grivell [0] (
16:51:07Zagoramiconn: linux, x86
16:55:27 Part grivell
16:58:36amiconnZagor: The X11 sim also does not like id3db: It says:
16:58:37amiconnWe open the real file 'archos/.rockbox/rockbox.id3db'
16:58:37amiconnUnsupported database version 2288752, aborting.
16:58:37BagderCopyright (C) 2002 might not be entirely correct in dbtree.c
16:59:00 Join oxygen77 [0] (
16:59:06 Quit webguest43 ("CGI:IRC (Ping timeout)")
16:59:14Zagoramiconn: i'm running the x11 sim without problem
16:59:20amiconn(that's also on cygwin)
16:59:42BagderBE32() shouldn't be used on a single byte, should it?
16:59:46Zagori think cygwin is the key here
16:59:51amiconnLITTLE_ENDIAN is *not* defined (I added "#error Little endian" within)
16:59:51ZagorBagder: it's not, it's a word
16:59:59Bagder16 bits it is
17:00:06Zagoramiconn: that's what i thought all along
17:00:07Bagderon some platforms
17:00:24Zagorright well those platforms weren't an issue when this was written :)
17:00:37 Join webguest23 [0] (
17:06:01 Join webguest24 [0] (
17:07:15 Quit webguest23 ("CGI:IRC")
17:12:36 Join elinenbe [0] (
17:13:06elinenbeI have a question. When I run the makedb.bat file I get this error "Can't locate in @INC (@INC contains: .) at line 566.
17:13:07elinenbeBEGIN failed−−compilation aborted at line 566.
17:15:27 Quit webguest24 ("CGI:IRC (EOF)")
17:16:03 Join webguest72 [0] (
17:16:05Zagorelinenbe: fixed, download it again
17:16:25Zagorhowever Lynx said the bat file is incorrect, so you will need to run the script manually
17:16:36elinenbeZagor: wow, that was quick.
17:16:41elinenbeI'll let you know how it goes.
17:16:55elinenbeI do think there should be a much quicker way to browse by ID3
17:17:05Zagori'm afraid i have to go. see you all later.
17:17:09 Part Zagor
17:25:15 Quit webguest72 ("CGI:IRC (EOF)")
17:28:12 Join cline [0] (
17:32:57 Quit cline (Client Quit)
17:35:02elinenbeit works now.
17:36:32 Join webguest83 [0] (
17:37:00 Quit webguest83 (Client Quit)
17:39:05 Quit ripnetuk (Remote closed the connection)
17:39:24 Join clne [0] (
17:40:58 Join Spida_ [0] (
17:46:59 Quit clne ("CGI:IRC (EOF)")
17:48:32quelsarukleaving, cu!
17:49:53 Part quelsaruk ("Ooh, look, a shiny object...")
17:52:27 Quit Spida (Read error: 110 (Connection timed out))
17:53:26 Join clne [0] (
17:54:00 Quit clne (Client Quit)
17:59:11 Join webguest01 [0] (
18:00:17webguest01tu es là
18:09:16***Saving seen data "./dancer.seen"
18:09:45[IDC]Dragonamiconn: idn't you have a file stress test plugin or so?
18:11:07[IDC]Dragonthinking about Pieter's problems with Ondio
18:11:39[IDC]Dragonhe sent me a corrupted dump
18:11:49[IDC]Dragonbut it may not tell much
18:13:20[IDC]Dragonperhaps it's a good idea to make the file system test code and your plugin identical
18:13:43[IDC]Dragonthen we can do the same test in the sim and on the hardware
18:14:19[IDC]Dragonand check if they run apart
18:18:14 Join webguest39 [0] (
18:20:44 Nick Lynx_awy is now known as Lynx_ (HydraIRC@
18:26:05amiconn[IDC]Dragon: The fs test plugin writes a file with pseudo-random data, using the mp3 buffer. It fills the buffer with the data, then saves in 2 chunks.
18:26:17amiconnThe split point between the chunks is also random.
18:26:56amiconnThis action is performed in a loop, so the resulting file size will be a multiple of the mp3 buffer size
18:27:14amiconnThe loop count is configurable at compile time
18:27:22 Quit webguest01 ("CGI:IRC (Ping timeout)")
18:27:42[IDC]Dragonsounds good
18:27:50amiconnThen it reads back the same chunks, and checks with the prng if the read data is what it should be
18:29:47amiconnThis mechanism uses the deterministic behaviour of the prng, so this allows a full comparison of the data (not only crc), without needing to store the data to compare with elsewhere.
18:30:25amiconnIt is relatively slow though; the file can get very large
18:30:25[IDC]Dragonis it your quick generator?
18:30:50amiconnInitially I wrote this to check whether ata could be the cause for the recording problems.
18:31:03amiconnNo, it's rockbox random()
18:31:45[IDC]Dragonhowever, the content shoudn't matter
18:31:59[IDC]Dragonif different on the test code
18:32:03amiconnWriting and comparing an 1.7 GB file took 10 hours on the recorder
18:33:00amiconn(1.7 GB == 1000x mp3_buffer_size)
18:34:33[IDC]Dragonit doesn't have to be so big for Ondio ;-)
18:36:08amiconnWell, it even can't ;) I ran this once on the Ondio (filesize ~80 MB), without problems
18:37:40[IDC]Dragonwe'd probably need some mixed actions, like creating dirs files within, deleting some inbetween, etc.
18:38:53[IDC]Dragonperhaps best on the MMC, so it can be completely blank at start
18:39:39amiconnDid you see already?
18:40:05[IDC]Dragonno, but the email
18:40:36[IDC]Dragondo you want a copy?
18:41:46amiconnAh, that's the dump. Anything abnormal with it?
18:42:07[IDC]Dragonhaven't downloaded it yet
18:42:41amiconnMay be it is useful, please send a copy if possible
18:42:47amiconn(off now)
18:57:24[IDC]Dragon(off, too)
18:57:28 Quit [IDC]Dragon ("CGI:IRC")
19:01:30 Quit webguest39 ("CGI:IRC")
19:14:13 Join einhirn_ [0] (
19:14:34 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
19:14:52 Nick einhirn_ is now known as einhirn (
19:20:36 Join Pieter_ [0] (
19:20:48 Nick Pieter_ is now known as Pieter__ (
19:22:17 Quit Pieter__ (Client Quit)
19:25:16 Join R3nTiL [0] (~zorroz@
19:43:43 Quit R3nTiL ()
20:00:35amiconnBagder: r u there?
20:09:07 Quit oxygen77 (Read error: 60 (Operation timed out))
20:09:18***Saving seen data "./dancer.seen"
20:38:46Bagderhere now
20:48:06 Join Stryke` [0] (
20:49:50 Join onox [0] (
20:52:00onoxis there anyone here who has knowledge about the z80?
20:55:11amiconnBagder: I checked the endianess problems with cygwin gcc a bit more. Neither LITTLE_ENDIAN nor __LITTLE_ENDIAN__ are defined :(
20:55:47amiconnHowever, we may "misuse" _X86_ , which is defined.
20:56:58amiconnonox: Maybe I can help...
20:57:37onoxI don't have a problem with progrmming it, but rather using it in circuits
20:57:49onoxwhat's the min. current a pin of the z80 requires?
20:58:55amiconnSorry I don't know the hw specs.
20:59:38amiconnBagder: And for a different topic: I wanted to ask whether the simulator build system will be changed to use the SOURCES files in the near future.
21:00:01amiconnThis would make solving the lcd fillrect() issue for the iRiver sim much easier...
21:00:06onoxwell, the manual says: power supply current max. 30 mA
21:00:18 Join jyp [0] (
21:00:51onoxbut I don't know if any electronics work with low current
21:04:00onoxamiconn: do you have experience with electronics?
21:12:42amiconnonox: Yes, I think so.
21:16:54 Quit onox ()
21:17:02Bagderamiconn: I don't know about "near future" but I would like it to be done, yes
21:18:09amiconnI found a way to hack this dependency in. Is there something like "elseif" for makefile, or do I need to nest the if ?
21:18:44Bagderlet me check...
21:19:03Bagderno, there's only 'else'
21:44:04 Join elinenbe_ [0] (
21:44:04 Quit elinenbe (Read error: 104 (Connection reset by peer))
21:44:08 Nick elinenbe_ is now known as elinenbe (
22:00:57 Join onox [0] (
22:09:19***Saving seen data "./dancer.seen"
22:22:38 Quit einhirn (Read error: 104 (Connection reset by peer))
22:25:53 Join [IDC]Dragon [0] (
22:27:22[IDC]Dragonhi again
22:27:32amiconnhi Jörg
22:27:50[IDC]DragonPieter missed me/us
22:28:21amiconnDid you already download his image? It's 500 MB... (I did it)
22:28:33[IDC]Dragonstarting to do so
22:28:38[IDC]Dragon500 MB?
22:28:49[IDC]Dragonhow many images is that?
22:29:26amiconnyup. Zipped, but only a very little compression. It is exactly one MMC image.
22:29:48[IDC]Dragon512 MB MMC?
22:30:09amiconnSeems so, yes.
22:34:15[IDC]Dragonhmm, my rate for this file is rather low
22:34:26[IDC]Dragon~40 KB/s
22:34:45 Quit onox ()
22:34:46amiconnI got 60 KB/s (starting with ~100)
22:34:54amiconnThis will take a while ;-)
22:35:08[IDC]Dragon3:35 ETC
22:37:06[IDC]Dragondidn't he brag abot his 100Mbit ? ;-)
22:37:35 Join einhirn [0] (
22:48:19 Join Zagor [242] (
22:48:30[IDC]Dragonhi Zagor
22:48:50amiconnhi again
22:49:22[IDC]DragonI have no luck with the ID browser. Now I get a CPU Adr. Error @ 0xFFFFFFFF
22:49:26amiconnZagor: Did you read the log (my remarks concerning endianess & cygwin)?
22:49:31Zagornot yet
22:51:13amiconnI found no suitable symbol other than _X86_
23:06:53Zagori find it very strange. why does cygwin gcc differ so much from "regular"?
23:07:13amiconnI don't know.
23:07:16jypWhat's the difference ?
23:07:32jypI mean, which is annoying ?
23:07:34Zagorjyp: cygwin doesn't define LITTLE_ENDIAN
23:07:36amiconnZagor: Do you want a list of the internally defined symbols?
23:08:02Zagoramiconn: thanks, but i think i will try to find some gcc & cygwin developers and ask them
23:08:08jypI think maybe it is not compiled natively but cross compiled
23:08:49amiconnIn contrast, sh-elf-gcc on cygwin *does* define BIG_ENDIAN as it should
23:08:57Zagor[IDC]Dragon: did you grab the latest code? i added some checks which should avoid fail more gracefully
23:09:06jypok ;p
23:09:08[IDC]DragonI thought I did
23:09:30[IDC]Dragonbut now with cvs I can do that more "controlled"
23:10:02amiconnZagor: The db loading clearly slows down boot :(
23:10:25[IDC]Dragonif enabled, I hope
23:10:31Zagori'd say the half-second sleep slows it down more...
23:10:46amiconnIt does this also when not enabled.
23:11:00Zagorbut yeah, it will me made on-demand later
23:11:16amiconnTried latest cvs on recorder -> I09:CPUAdrErr at FFFFFFFF
23:11:29[IDC]Dragonsame as I get
23:12:33Zagoryeah, i can't run cygwin here so I can't test those problems right now. grab the bleeding edge if you want to try it.
23:13:13[IDC]Dragonyou think it's a cygwin compile problem?
23:13:19amiconnThis happens immediately after leaving the menu from enabling ID3 browse. Red led is lit, setting is not saved. ON for RoLo does hang
23:13:58[IDC]DragonI did try a bleeding edge before...
23:13:59amiconn[IDC]Dragon: You still use sh-elf-gcc 3.3.1?
23:14:33[IDC]Dragonon this machine, I think so
23:14:44amiconnThat lowers the chance that it's caused by cygwin. I use sh-elf-gcc 3.3.5 meanwhile.
23:14:53[IDC]Dragonon the notbook, I installed Eric's latest
23:16:08[IDC]Dragoncompile done. yes, it's 3.3.1
23:16:43amiconnZagor: Same effect with bleeding edge from
23:17:42[IDC]Dragonan adress exception is a misaligned access?
23:17:50amiconn(only that the led is flashing, not simply lit)
23:17:57amiconn[IDC]Dragon: yup.
23:18:04Zagorbleeding edge works perfectly for me. this is nuts.
23:18:21[IDC]Dragonthen it tried to access -1
23:18:50[IDC]Dragonwhich is probably a special value, instead of a pointer
23:18:51amiconnZagor: You probably have a different database?
23:19:06[IDC]Dragonvery likely ;-)
23:19:14amiconn(As I now indexed my whole box, it's 787 KB)
23:19:25[IDC]Dragonunless you guys have an identical taste
23:19:34Zagormy database is 2.4 MB
23:19:44[IDC]Dragonmine about 2.2
23:20:46[IDC]Dragonsome bug in the perl creation, perhaps?
23:21:09Zagora very interesting thing is that is appears my disks cache is valid even when the disk is spun down. i can browse back and forth, and the disk led flashes briefly but the disk doesn't spin up
23:21:28[IDC]Dragonthat's nice
23:21:34Zagorsend me your databases for testing
23:22:01[IDC]Dragonany particular upload location?
23:22:16[IDC]Dragonor just email?
23:22:17amiconnZagor: dcc send?
23:22:21Zagorwait, i'll recreate mine first. maybe i broke something in the last few changes to it.
23:23:18BagderNumber of songs : 5873
23:23:32Bagderfind /data/mp3/ -type f -name "*mp3" | wc -l
23:23:46Zagori don't index songs without tags
23:23:58amiconnAh, I just wanted to ask that :)
23:24:07BagderI don't think I have any songs without tags
23:24:07Zagorthe "<no artist>__<no album>" album becomes way too big if we do
23:24:10Bagderbut I'll check
23:24:21Bagderdoes the verbose output tell?
23:24:36Zagorno it doesn't. it should.
23:25:27amiconn"214 untagged files skipped"
23:25:38Zagoryeah. perhaps even a list.
23:28:31CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:28:31*Bagder fires up easytag
23:29:42amiconnThe list only when in verbose mode, please.
23:30:35Zagormy recreated list works without problems
23:31:00amiconnSo we should send ours?
23:32:49amiconnHuh? Did you get the file? It says "failed: connection closed", but transferred all data !=
23:32:55Zagori got it
23:34:31Zagorit is corrupt. something in the perl script is not windows safe.
23:35:00amiconnHmm. What perl is that in the wiki, as I used this one now?
23:35:02[IDC]Dragonso you don't need mine, I guess
23:35:12Zagor[IDC]Dragon: no
23:35:20[IDC]DragonI use ActivePerl
23:35:28Zagoramiconn: it's a link to the head, i.e. always latest version
23:36:07amiconnHmm. I'll try cygwin perl now.
23:36:49Zagoroh, "what perl". i think it's an official win32 port. not activestate or cygwin.
23:37:10Zagorwe use it at work a lot
23:38:11amiconnIt added drive letter & friends to the db...
23:38:54Zagoryeah it will add complete path unless you use −−strip
23:40:16Zagorinteresting. database version is ok, song table pointer is ok, number of songs is ok, but length of song name field is corrupt.
23:41:41amiconndb generated with cygwin perl doesn't cause a crash.
23:41:53amiconnHowever, I get "Dir buffer full" now
23:42:21Zagoryes, that is normal. i haven't added a moving window to it yet. the artist table is too long to fit in the dir buffer.
23:42:40amiconnHow do I switch browse mode?
23:42:56ZagorF2 quickscreen or menu
23:43:41amiconnNo, I mean within ID3 browse, switching between by artist, by song name tec
23:44:12Zagoryou can only browse it as a tree yet, i.e. going into the artists lists his albums. going into an album lists the songs.
23:45:20amiconnThat's useless... It's the default hierarchy of the fs I use anyway (except that I have one additional level above that - a (rough) genre)
23:45:37Zagoruh, it's not ready...
23:46:32amiconnNow the box froze after switching back to file browsing (having increased the max-files-in-dir-buffer setting in between, but not yet restarted)
23:47:46Zagori wonder why the perl versions produce different output
23:48:47amiconnScrolling the id3 db is sloo-ow
23:51:00amiconnThe id3db browser should not crash with a corrupt db file, imho
23:51:36Zagorof course not. fix committed.
23:52:11amiconn(different topic) With the current implementation, the leave-all-menus-if-language-changed workaround is no longer needed, I checked this.
23:54:04amiconnBug in id3 browser integration: If switched on, I cannot browse .wps, .cfg, .lng, .rock ... because I also get the id3 browser when I try...
23:54:21Zagoryup's even not leavable, I'm stuck...

Previous day | Next day