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 2005-09-02

00:01:41 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.6/20050716]")
00:02:38***Saving seen data "./dancer.seen"
00:19:25 Quit wacky_ ("ooooohh yeah")
00:19:48 Quit Zagor ("Client exiting")
00:23:24 Quit Moos (" HydraIRC -> http://www.hydrairc.com <- Leading Edge IRC")
00:29:59 Join ashridah [0] (i=ashridah@220-253-121-127.VIC.netspace.net.au)
00:41:05 Quit ender` (Read error: 113 (No route to host))
00:45:07 Join TCK [0] (i=TCK@85-210-35-84.dsl.pipex.com)
00:59:42 Quit webguest61 ("CGI:IRC (EOF)")
01:00
01:00:54 Quit matsl (Remote closed the connection)
01:02:28 Quit Domonoky ("Trillian (http://www.ceruleanstudios.com")
01:14:07 Quit webguest05 ("CGI:IRC (EOF)")
01:35:40 Join solex_aw1y [0] (n=jrschulz@c147039.adsl.hansenet.de)
01:40:20 Quit Sucka ("a bird in the bush is worth two in your house")
01:48:01 Quit solex_away (Read error: 110 (Connection timed out))
01:50:39Wettarg, I can't compile anything... Yet another problem, someone help me ? :)
01:52:27ashridahgot any error messages?
01:52:38Wettyes, plenty of
01:53:05Wettit seems that I can't use libmad from a plugin...
01:53:38ashridahaah, you're using the codec plugins from another plugin. waay out of my depth :)
01:54:38Wett:p ok, i'll have to find on my own
02:00
02:02:39***Saving seen data "./dancer.seen"
02:17:03 Part cheriff_AWAY
02:33:53Wettcompiled ! but it doesn't work xD
02:41:24 Quit hicks (Remote closed the connection)
02:45:11ashridahhuzzah!
02:54:27Wetti'll see that tomorrow, good night !
02:54:36 Quit Wett ("Visitez www.coder-studio.com !")
03:00
03:09:59 Join Febs [0] (n=Febs@207-172-122-81.c3-0.rdl-ubr4.trpr-rdl.pa.cable.rcn.com)
04:00
04:02:41***Saving seen data "./dancer.seen"
04:15:57 Join coob [0] (i=pen0r@host-84-9-63-253.bulldogdsl.com)
04:25:56 Join webguest00 [0] (n=8ddfd339@labb.contactor.se)
04:29:41 Quit preglow ("leaving")
04:30:19 Join adiamas [0] (n=adiamas@ool-435559f8.dyn.optonline.net)
04:30:53adiamasBagder!
04:30:56adiamaslong time no see :)
04:43:01 Quit webguest00 ("CGI:IRC")
05:00
05:02:16 Join Omega11 [0] (i=Omega11@69-173-185-233.sbtnvt.adelphia.net)
05:08:59 Quit Slasher (herbert.freenode.net irc.freenode.net)
05:08:59NSplitherbert.freenode.net irc.freenode.net
05:15:38 Quit Omega11 ()
05:39:52NHeal(timeout) herbert.freenode.net irc.freenode.net
05:49:11 Join random_man [0] (n=405b5447@labb.contactor.se)
05:49:51random_manhello. i need some help with the newest rockbox cvs version for rockbox. is there anybody that thinks they could help me?
05:50:38random_manfor iriver **
06:00
06:00:07ashridahrandom_man: what;s the prob?
06:02:45***Saving seen data "./dancer.seen"
06:10:16random_mansorry i took so long
06:10:45random_mani cannot view bmps in my wps ashridah
06:12:32ashridahhrm. not something i've tried to do myself
06:13:11random_manwell i jsut updated the version on my iriver and then it wouldnt show the bmps last version worked fine
06:13:38 Join LinusN [0] (n=linus@labb.contactor.se)
06:13:50ashridahask linus
06:15:25random_manLinusN i am not able to see my bmps in my wps. would you know a reason for this? i just updated to bleeding edge build and this problem happened
06:15:47LinusNthe wps image tag syntax has changed
06:16:16random_manis it in the wiki?
06:16:26LinusNhttp://www.rockbox.org/twiki/bin/view/Main/CustomWPS
06:17:00random_manhaha just got there myself
06:22:01random_mancrap mp3 player froze
06:26:48random_manhmmm i still cant seem to get it.
06:27:20random_mani got this %xde|Boxea1.bmp|1|8|
06:29:00random_mani dont have a-d pics does that matter?
06:29:42LinusNshould be %x|e|Boxea1.bmp|1|8|
06:30:49random_mandid i read the wrong thing?
06:31:06random_manthe wiki says . Image display tag:= %xdn
06:31:14random_manthats the one i want right?
06:31:38LinusNbetter read further down the page where this is explained
06:32:35random_manok
06:33:07random_manwoo hoo it works no ill read
06:34:07random_manahhh i understand now
06:36:53random_manthank you LinusN
06:37:00LinusNyou're welcome
06:39:49random_mando you know how i would do the x and y measurments or do i have to guess?
06:43:19LinusNguess or calculate, your call
06:43:52LinusNone way is to take a screenshot and edit it
06:45:20random_mangood idea
07:00
07:30:55 Quit random_man ("CGI:IRC (EOF)")
07:37:44 Quit ashridah ("Leaving")
08:00
08:00:05amiconnmorning
08:00:20B4gderguten morgen
08:00:42LinusNmoo
08:01:04LinusNamiconn: nice work on the filename creation
08:01:14CtcpIgnored 3 channel CTCP requests in 14 minutes and 11 seconds at the last flood
08:01:14*amiconn didn't manage to actually reduce code size with his last change :/
08:01:42LinusNbut it certainly is better source code wise
08:02:05LinusNmoving the credits to a plugin saved quite a lot
08:02:19LinusN(damn, sourceforge is slow today)
08:02:24B4gderprobably the biggest saver for a while
08:02:28amiconnAnyway, at least the filename creation is more modular now, and the config files now use the same logic as all other numbered files used on non-rtc targets
08:02:46***Saving seen data "./dancer.seen"
08:03:13LinusNin fact, i'd like an option to have numbered file creation even on rtc units
08:03:28LinusN(aaaaaah, more options!!!!!)
08:03:34B4gder:-)
08:03:45amiconnoptionitis :/
08:03:52LinusNthat's a feature many have asked for
08:03:57amiconnReally?
08:04:02LinusNin the past
08:04:19LinusNthey hated our date names from the beginning
08:04:45amiconnAnother thing I came across is that the filename patterns for numbered and date+time names are different
08:04:46B4gder"what is that number for"
08:04:47B4gderhehe
08:04:56B4gder"it is a date"
08:04:57LinusNmostly since they didn't understand that they were dates
08:04:59B4gder"it is?"
08:04:59amiconnI left it that way, but perhaps it should be unified?
08:05:00B4gder;-)
08:05:20LinusNamiconn: i think it could
08:05:29amiconnI mean "R<datetime>" vs. "rec_<number>" etc
08:06:09amiconnIt's clearly visible in the source now that there are these functions
08:06:37amiconnHmm, is there a special way to call the credits plugin, or is it just an ordinary plugin?
08:07:34B4gderplugin_load("/.rockbox/rocks/credits.rock",NULL);
08:07:36amiconnAhh, there's a direct link
08:08:10LinusNyeah, we've changed our minds regarding direct links
08:08:47LinusNwhat's your opinion on this: http://sourceforge.net/tracker/index.php?func=detail&atid=439120&group_id=44306&aid=1179046
08:09:32B4gdersounds fair
08:09:42B4gderbut I don't use such things so I really can't tell
08:10:02LinusNit's mainly a convenience for people using the triggered recording
08:10:52amiconnHmm, the credits patch makes the plugin api incompotible w/o reason
08:11:16amiconnIt would have been possible to retain compatibility by adding the functions at the end
08:11:40amiconnJust a minor thing...
08:12:20LinusNi figured that we won't change it again before release
08:12:33LinusNso i took the opportunity to sort it
08:12:40amiconnAh
08:12:56LinusNdoesn't really matter
08:15:19LinusNamiconn: in what way is the file size display wrong when prerecording?
08:16:21 Quit einhirn ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
08:16:21amiconnIt uses num_rec_bytes, and that always counts up
08:16:33amiconnFor prerecording, we would need a correction value
08:17:26LinusNit definitely tried to correct
08:17:29LinusNtries
08:17:52amiconnIt does? where?
08:18:15LinusNah, no
08:18:34LinusNnum_rec_bytes is not adjusted when starting the recording...
08:18:47amiconnYes, and we can't simply adjust it
08:18:52LinusNwhy not?
08:19:18amiconnIt's written to in an isr, and this addition is non-atomic
08:20:01amiconnHmm, it doesn't matter that the addition in the isr is non-atomic, but a corrective subtraction would be non-atomic too
08:20:15amiconnAnyway, we don't have a correction value
08:20:31LinusNwe don't need a correction value
08:20:35LinusNwe just set it
08:20:48LinusNnum_rec_bytes = ...
08:21:53LinusNwe know startpos and audiobuf_write
08:22:15LinusNcalculate the size and set num_rec_bytes
08:22:30LinusNjust like we do in mpeg_get_num_recorded_bytes()
08:23:04amiconnHmm, we can do that after line 1775
08:23:25amiconnOh, btw, there's a wrap correction missing:
08:23:25LinusNyes
08:23:43amiconnLine 1772: audiobuf_read = startpos + offset;
08:23:53amiconnThere should be a wrap correction after this
08:24:09LinusNouch
08:24:35LinusNbtw, do you think some of the problems people have had with file splits could be because of the peak meter hogging?
08:25:14amiconnNot if they used a reasonably recent build
08:25:27LinusN"reasonably"
08:25:29LinusN?
08:26:19amiconnThe hogging is fixed since 2005-06-04
08:27:01amiconnSince then it was the other sort of wrong - the peakmeter peeked so little that it was imprecise
08:27:04LinusNsf is wacko atm
08:27:49amiconnHmm, the credits move saved ~3000 bytes
08:27:55LinusNwow
08:29:34amiconnIs there a reason why credits.raw now ends up in the build root?
08:29:49LinusNsounds like a mistake
08:30:09amiconnIt used to end up in apps/ before, and should perhaps be in apps/plugins/ now
08:30:14LinusNyes
08:31:02 Join _reiner [0] (n=reiner@mailout.imageware.de)
08:34:29 Part _reiner ("Konversation terminated!")
08:34:47 Join arkascha [0] (n=arkascha@mailout.imageware.de)
08:36:25LinusNamiconn: http://linus.haxx.se/recbytes.patch
08:40:48amiconnHmm, looks good to me
08:41:04amiconnBtw, you could cleanup some *.exe.stackdump files
08:41:27LinusNhehe
08:41:49B4gderI've grown to always make my patches with cvs diff | grep -v "^?"
08:42:19B4gderI wonder if there is a cvs option to hide them
08:42:25LinusNi bet there is
08:42:35B4gder-q doesn't help at least
08:44:16amiconnLinusN: I would put the num_rec_bytes calculation below the DEBUGF() call telling the new position
08:44:41LinusNsure, why?
08:45:20amiconnI just think this DEBUGF() logically belongs to the audiobug_read adjustment above
08:45:26LinusNof course
08:45:30LinusNdone
08:45:32amiconn*audiobuf_read
08:45:39amiconnaudiobug is funny
08:45:42LinusNfreudian slip
08:46:56amiconnThe header stuff below will probably be replaced with a prepend_header() function by me
08:47:48amiconnThat would be used for file split as well, and create_xing_header would resort to using the time when no framecount is available
08:48:28LinusNok
08:49:09LinusNi wonder if adding xing headers to split files is a good idea or not
08:49:30amiconnI think it is, we've discussed this already
08:49:30LinusNmakes it harder to join them
08:49:49LinusNstill, why split them in the first place
08:49:50LinusN?
08:49:52amiconn(1) Why would you split them in the first place if you want to join them?
08:49:56LinusN:-)
08:50:36amiconn(2) With no header the files are not detected as vbr by a number of players
08:50:49LinusNincluding rockbox
08:53:25amiconnWell, it could be done.
08:54:02amiconnAfaik some players read a couple of frames from the start. If they find frames with different bitrates, it's vbr
08:54:43amiconn(a couple >~ 100 or so)
08:56:06adiamaslinus!
08:57:43LinusNamiconn: that's always a method, but we have so far refrained from that because it would take time
08:57:58LinusNand the bitswapping makes it a little difficult
08:58:02LinusNadiamas: yo
09:00
09:00:58 Join DangerousDan [0] (n=Miranda@newtpulsifer.campus.luth.se)
09:01:23 Join ender` [0] (n=ender@tm.213.143.74.124.dc.telemach.net)
09:04:48 Join tvelocity [0] (n=tony@ipa175.1.tellas.gr)
09:05:25LinusNamiconn: this certainly looks like it would have been fixed by the peak meter hogging fix: http://sourceforge.net/tracker/index.php?func=detail&group_id=44306&atid=439118&aid=937100
09:12:35amiconnYes it should, however, there are still situations where the mas delivers corrupt frames
09:12:53amiconnOften they're not really corrupt, but not byte aligned
09:13:10LinusNthat's really weird
09:13:30amiconnmpeg audio is a bitstream so it shouldn't matter, however, many (most?) players expect the headers to be byte aligned
09:14:14amiconnI'm not sure whether we can do anything about it, afair this also happens with the archos firmware and long recordings
09:14:29LinusNthat sucks immensely
09:14:55amiconnI'm out of ideas what could be wrong with our mas handling.
09:15:08LinusNif there is anything wrong
09:16:15LinusNimho, the triggered recording should be off by default
09:16:23LinusNnow the default is "once"
09:16:54amiconnThe thing is that if a file starts with corrupt (or shifted) frames, this file isn't playable in many players, but if the corrupt/shifted frames occur later in the stream it may or may not play correctly
09:17:34LinusNok, so a "frame shift" plugin is in order :-)
09:17:51LinusNwould only take a few days to run on a large recording :-)
09:17:54amiconnI had some recording test files that were stuttering on playback
09:18:48amiconnBtw, it helps a lot to set independent frames to on if you plan to analyse the file in a hex editor...
09:19:01LinusNhehe
09:19:21amiconnAnother helpful mas feature is that it can write the frame count into the frames
09:19:29amiconn(aka timecode)
09:19:50LinusNnot that it matters, but the cfg file names for the trigger modes "once" and "repeat" are "no rearm" and "rearm"
09:19:51 Quit arkascha (Remote closed the connection)
09:20:14amiconnThis is how I found the wrap problem with the peakmeter hogging
09:20:26LinusNamiconn: yes, but it suffers from the counter overflow problem
09:20:36amiconnI needed a way to identify which snippet of mp3 overwrote the beginning...
09:21:32 Join ashridah [0] (i=ashridah@220-253-122-84.VIC.netspace.net.au)
09:21:46LinusNi'll change the option names to "once" and "repeat", plus change the default to "off"
09:22:47amiconnAgreed
09:23:58LinusNi also want to change the behaviour when you play a new file in the playlist browser
09:24:26LinusNit should quit and show the wps instead of staying in the playlist viewer
09:24:32LinusNimho
09:24:42amiconnHmm, depends
09:24:43*ashridah looks through his email
09:25:01amiconn7me whispers: another option... ;)
09:25:08LinusN:-P
09:25:09ashridahyeah. i'd definently put in a vote for bugzilla. it does a reasonable job. not that i've played with many of the other ones
09:25:36ashridahbiggest problem is that it requires mysql iirc.
09:25:37LinusNisn't bugzilla a monster (no pun intended)?
09:25:49ashridah(it's not db agnostic, by the looks of the dependencies)
09:27:31ashridahpossibly, but who cares about the size if it damned well works? bugzilla's been easy enough to use the many times i've dealt with it (mostly on icculus.org)
09:28:08LinusNi was not referring to the size
09:29:08 Quit adiamas (Read error: 110 (Connection timed out))
09:29:58LinusNbut we should probably try it
09:30:24ashridahcan't hurt to try all of them
09:30:55ashridahmany of them seem to require email-based bug submissions in a particular format (debbugs and gnats for eg)
09:31:34ashridahflyspray might be a reasonable alternative to bugzilla, since it claims "Flyspray is an easy to use BTS for those who don't require all the complexities of something like Bugzilla."
09:31:41ashridahuses php tho
09:32:37amiconnLinusN: Recording file size fix is working, tried it on my recorder
09:32:53LinusNgreat
09:33:35LinusNi wonder about the issue kurzhaarrocker brought up yesterday, about the "start in rec screen" not working if resume==yes
09:33:56LinusNi think it should have priority over the resume
09:34:32amiconnHmm, not sure
09:34:44amiconnThat may lead to unexpected behaviour.
09:35:02amiconn...like resuming playback after leaving this (initial) recording screen
09:35:21amiconnI think you're right, but then it should not resume when leaving the recording screen
09:35:45amiconn(which would be correct anyway from the option name. It's called "resume on startup")
09:37:12LinusNso it should set start_wps to false
09:38:24 Join kurzhaarrocker [0] (n=Phil@p50908055.dip0.t-ipconnect.de)
09:40:45kurzhaarrockerI just updated a patch for autoselection of the recording source: http://sourceforge.net/tracker/index.php?func=detail&atid=439120&group_id=44306&aid=1179046
09:41:25kurzhaarrockerI think that can be considered as bug fixing -> It's worth to consider wether it should be appliedt before V2.5
09:42:53LinusNi think so too
09:45:12kurzhaarrocker-recording source + trigger source
09:45:21LinusNkurzhaarrocker: nitpicking: use "bool" for the recording_toggled flag
09:46:28kurzhaarrockerThat's not nitpicking, that's help to improve my c :)
09:50:24kurzhaarrockerWas bool allowed in rockbox ever since? Or was there a change in code police rules somewhen?
09:51:33LinusNwe changed that when we went c99
09:51:42 Join banan_ [0] (i=banan@dalink.campus.luth.se)
09:51:43LinusNa long time ago
09:51:47kurzhaarrockerah. ok.
09:51:55LinusNwe still aren't c99 all over
09:52:10 Quit merbanan (Read error: 104 (Connection reset by peer))
09:52:11LinusNwe still frown upon variable declarations in the middle of the code
09:52:26LinusNbecause we're old and grumpy
09:52:34kurzhaarrockerwould c99 allow c++ comments too?
09:53:05LinusNunfortunately yes
09:53:14LinusNbut we frown upon that too :-)
09:53:19kurzhaarrocker:)
09:53:28kurzhaarrockerOnly dinosaurs know what a c64 was ..
09:54:31LinusN:-)
09:55:03kurzhaarrockerstill have 1 1/2 of those
09:55:07kurzhaarrockerI like c++ comments as that makes it easier to comment out code with c comments. But I prefer to stick to given design rules.
09:55:13LinusNcan someone think of a reason not to have "rec on startup" have priority over resume?
09:55:34zehey i'm not a dinosaur
09:55:48zei just happened to only be 3 when i figured out how to load programs from floppy on a c64 :p
09:56:02LinusNc++ comments are often evil, like when adding comments inside #define's
09:56:37LinusNwhen i was 3, there were no home computers ;-)
09:57:12kurzhaarrockerI think its confusing if you toggle "rec on startup" and nothing changes because of a side effect based on "resume on start".
09:57:19LinusNme too
09:59:15kurzhaarrockerOnce upon a time - when we had that "ask" option for "resume on start" it might have made sense though.
09:59:50kurzhaarrocker(never tried it then)
09:59:59LinusNi changed it just now
10:00
10:00:16LinusNamiconn: are you working on the Xing generation?
10:01:54LinusNkurzhaarrocker: when you create a patch, always be in the rockbox source root, not in apps/
10:02:39*kurzhaarrocker tries to remember that
10:02:50***Saving seen data "./dancer.seen"
10:07:49 Join arkascha [0] (n=arkascha@mailout.imageware.de)
10:07:51kurzhaarrockerze: floppies were for posh snobs. Adjusting the tape head of the datasette so that you could read the files of your friend to was the real art.
10:08:27zekurzhaarrocker: hey what do you expect out of a 3yo
10:08:46kurzhaarrocker:)
10:08:56zegetting up in the middle of the night and knowing to type LOAD followed by whats printed on the floppy then ,8 or ,8,1
10:09:04 Join cYmen [0] (n=cymen@nat-ph3-wh.rz.uni-karlsruhe.de)
10:09:13zein fact i dunno how i even managed that :p
10:09:20LinusNkurzhaarrocker: the peak meter doesn't work when auto-sourcing
10:09:39kurzhaarrockerYou have a source attached I assume, LinusN?
10:09:48kurzhaarrocker(or mic as source enabled)
10:09:58LinusNyes
10:10:11LinusNand it hung when i entered it from the recording screen :-(
10:10:30kurzhaarrockerhrgh. Why doesn't my recorder do that?
10:10:47*kurzhaarrocker checks
10:10:51LinusNi started by entering it directly after boot
10:10:57LinusNno peak meter
10:11:08LinusNthen i entered the recording screen and tried again from there
10:11:26LinusNand it froze
10:14:28LinusNsame thing happens even if i don't autosource
10:14:33LinusNit freezes
10:16:28kurzhaarrockerI can reproduce that bug with the current version. But with a version based on cvs of 31 Aug 05 and the auto source patch applied it works fine.
10:16:50LinusNcould be the new peak meter code
10:17:10 Quit cYmen ("back")
10:17:38kurzhaarrockerI can enter the trigger setup from stop mode and the peak meter works nicely (current cvs + autosource)
10:17:49LinusNstrange
10:17:59kurzhaarrockerBut not from recording screen
10:18:15 Join cYmen [0] (n=cymen@nat-ph3-wh.rz.uni-karlsruhe.de)
10:18:55LinusNhmmm, works for me too now, but not the very first time, must be my mistake
10:22:41 Quit DangerousDan (Read error: 104 (Connection reset by peer))
10:23:21LinusNit reinits the recording...
10:24:03LinusNthat's because audio_status() returns 0
10:24:46LinusNthat happens when prerecording is off
10:25:48kurzhaarrockermaybe I never tested that
10:26:02LinusNaudio_status() doesn't tell you if you are in recording mode or playback mode
10:29:45kurzhaarrockerLooks as if I have to access init_recording_done in mpeg.c instead.
10:31:06LinusNor add a new function, for example audio_mode()
10:31:37 Quit cYmen ("n forth")
10:31:52kurzhaarrockerok
10:32:31LinusNthere is a global variable called "mpeg_mode"
10:32:38 Join cYmen [0] (n=cymen@nat-ph3-wh.rz.uni-karlsruhe.de)
10:33:57LinusNwe could make the enum public, and add an audio_mode() function that returns the current mode
10:41:57amiconnkurzhaarrocker: regarding bool - if you checked my peakmeter changes, you might have noticed some int->bool conversions
10:43:02kurzhaarrockerI noticed that, amiconn, and to be honest It surprised me then, too.
10:46:26amiconnkurzhaarrocker: I only made it consistent. The function parameter resp. return value were int, but the internal variable of the peakmeter was bool.
10:46:38amiconnNot very logical...
10:46:57kurzhaarrockerdon't ask me about logic :)
10:47:59kurzhaarrockerThere's write access to mpeg_mode in mp3_playback.c and in mpeg.c. mpeg_mode is defined in mp3_playback. But it feels cheesy to make something like audio_mode() part of mp3_playback. Should mpeg_mode move?
10:48:02*kurzhaarrocker is confused
10:50:07kurzhaarrockeralso there seems to be no way to set mpeg_mode to MPEG_DECODER for devices with CONFIG_CODEC != MAS3587F
10:51:01*kurzhaarrocker smells a can labelled "worms"
10:53:39amiconnkurzhaarrocker: For MAS3507D and MAS3539F the mode is always decoder
10:54:00amiconn...and there can't be recording auto source as there is no recording
10:54:14kurzhaarrockerok, I see that a new function audio_mode() should be #ifdefed MAS3587F
10:54:24 Join preglow [0] (n=thomjoha@hekta.edt.aft.hist.no)
10:57:46kurzhaarrockerAnd still I am not convinced that the declaration of an enum MPEG_MODE should reside mp3_playback. What do you think?
10:58:19LinusNin fact, the entire audio code needs a rework
10:58:29LinusNbut that will be post-2.5
10:58:50LinusNso we could add a temp fix for this now, and clean it up later
11:00
11:00:25amiconnLinusN: You forgot to mark another issue as fixed in the release todo
11:00:34LinusNoh?
11:00:34kurzhaarrockerThen I'll declare the enum MPEG_MODE and audio_mode() in mpeg, but leave the mpeg_mode in mp3_playback.
11:00:45LinusNdo so
11:01:28 Join webguest68 [0] (n=d5ee48b7@labb.contactor.se)
11:01:40amiconnenum mpeg_mode should be declared in one of the .h files
11:01:51amiconnLinusN: Yes, the code size thing
11:01:52 Join einhirn [0] (i=Miranda@bsod.rz.tu-clausthal.de)
11:01:57LinusNamiconn: ah yes
11:02:01 Quit Strath (Read error: 104 (Connection reset by peer))
11:02:44 Join Strath [0] (n=mike@dpc674681214.direcpc.com)
11:03:34preglowhow much was shaved off in moving credits to a plugin?
11:03:41LinusN~3k
11:03:51preglowwoot
11:04:09preglowi'll see about optimising some codecs today, btw
11:04:15LinusNnice
11:04:29preglowdoes it coutns as bugfixing? :)
11:04:43LinusNpreglow: the playback/codec code has issues with the file/buffer positioning
11:04:54preglowLinusN: how?
11:04:58LinusNit confuses buffer and file positions
11:05:20LinusNso seeking doesn't work as expected
11:05:42preglowoh
11:06:06LinusNi made a fix in mpa.c, but it turned out to be a general issue
11:06:23amiconnpreglow: I'm awaiting your codec magic. ;)
11:06:45LinusNamiconn: i'm worried about the battery-low issue on the iriver
11:06:57*preglow summons Slasher
11:07:01LinusNwe should find a good way of solving this
11:07:06amiconnpreglow: Btw, do you intend to adjust the emac inits to the non-volatile behaviour
11:07:14amiconn?
11:07:31preglowamiconn: well, not much to adjust, really
11:08:04preglowcan probably remove a couple of moves from some asm blocks
11:08:57preglowthe biggest benefit of doing the macsr change is that i don't have to worry about the emac state when doing things like asmifying commonly used functions
11:08:59amiconnHmm, I can imagine there are some calls that are unnecessary now
11:09:01preglowlike mad_f_mul
11:09:52amiconnEach codec should only need to init the emac once, at startup
11:09:59amiconnSame goes for the dsp
11:10:27preglowshould, yes
11:10:52preglowi'll see about it
11:14:50 Part kurzhaarrocker
11:23:21amiconnLinusN: I hope you don't mind me putting the emac before the standard registers in the context. It made saving the registers easier
11:36:00LinusNamiconn: i don't mind at all
11:40:48 Join Sucka [0] (n=NNSCRIPT@host81-156-159-120.range81-156.btcentralplus.com)
11:45:42 Join Slasher [0] (i=miipekk@ihme.org)
11:51:07preglowSlasher: read twenty lines back
11:54:31 Quit cYmen ("oops")
11:55:52amiconnLinusN: What's the thing you worry about with the battery on iriver?
11:57:13LinusNthat it hangs when it can't spin up the hard drive
11:57:23LinusNand that you can't turn it off because of it
11:57:26amiconnThat's an issue on all platforms
11:57:41LinusNno
11:57:41amiconn...only that you can turn off the archos by hardware force
11:57:47LinusNexactly
11:58:08LinusNthe ata driver should be able to fail in this case
11:58:35amiconn...but it would hang for quite some time if it tries to shutdown and the box is unattended
11:58:59LinusNyes
11:59:16LinusNi wonder if we could make the ata driver detect this situation
11:59:22amiconnThe only platform where this is a non-issue is the ondio
11:59:39amiconnEither it does the shutdown, or power fails completely
12:00
12:00:14amiconnI would suggest a simple solution, as there is a second situation that leads to being unable to shutdown the iriver
12:00:50amiconn...and that is if the shutdown tries to kick in within an event loop not using the default handler
12:01:01amiconnI ran into this several times
12:01:03LinusNyes, but that would be a bug
12:01:25amiconnThere are many of these
12:01:27 Quit einhirn ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
12:01:32LinusNthen we should fix them
12:02:13amiconnI would simply add a timeout to the shutdown event, and shut down by force if not all threads have reacted to the event after that
12:02:22amiconn15 seconds should be enough
12:02:53***Saving seen data "./dancer.seen"
12:05:07LinusNthat's a good fallback mechanism, but i don't think that's the solution to the low battery issue
12:07:49 Join cYmen [0] (n=cymen@nat-ph3-wh.rz.uni-karlsruhe.de)
12:10:43 Join Moos [0] (i=DrMoos@m29.net81-66-158.noos.fr)
12:10:58MoosHello folks ! :)
12:18:30FebsAt the risk of demonstrating my naivete, would it not be possible to set a battery level threshold slightly above the point where the battery no longer is able to spin up the hard drive, and shut down automatically when that threshold is reached?
12:22:48 Quit Nibbler ("stromausfall")
12:27:40preglowwoops, gotta go
12:30:17coobFebs: we already do that...
12:30:34cooboh, whups, this isn't #ipodlinux :/
12:33:59 Join hicks [0] (n=hicks@zeus.mups.co.uk)
12:47:58 Quit amiconn (Read error: 110 (Connection timed out))
13:00
13:07:25 Join amiconn [0] (n=3e088e42@labb.contactor.se)
13:08:21amiconnbleh.
13:23:33MoosHi Linus, quick question: why there are the credits in the new plugin and still in infos/version ?
13:25:22LinusNbecause that's where we want it
13:25:57Moos2nd little question: when I go to PCM recording menu and leave it the playback don't work, need to restart for playback rework, is it a known bug or new?
13:26:01LinusNthe menu option simply calls the plugin
13:26:13Moosah ok i'm stupid scuse :)
13:26:14LinusNMoos: old, known bug
13:26:26Moosok thanks
13:26:36LinusNthat's one of the reasons that it's still in the debug menu
13:26:52Mooshehe true :)
13:28:18LinusNFebs: the problem with that approach is that we will have to set the threshold with a sufficient margin, and then we won't use the maximum capacity of the battery
13:28:53LinusNbut sure, if that's what it takes, we'll have to do it
13:29:41MoosI'm looking at the new Hitachi HD :P
13:30:19Moostravelstar C4K60, it will be good for our players :)
13:33:02MoosLinus: ZIF connector isn't good for irivers?
13:34:55Moosthey have better price then toshiba's :)
13:35:47LinusNzif? yuck!
13:37:25 Join einhirn [0] (i=Miranda@carlsberg.heim2.tu-clausthal.de)
13:38:27 Join Wett [0] (i=Wett@AMontpellier-251-1-52-119.w81-251.abo.wanadoo.fr)
13:40:31 Quit TCK (Read error: 110 (Connection timed out))
13:41:11FebsLinusN, you might find this interesting if you haven't seen it already: http://www.misticriver.net/boards/showthread.php?t=28157
13:41:31LinusNFebs: seen it :-)
13:42:11Febs:)
13:46:33LinusNWett: u there?
13:49:30 Nick mbr is now known as mbr_ (n=mb@stz-softwaretechnik.de)
13:54:55 Nick mbr_ is now known as mbr (n=mb@stz-softwaretechnik.de)
13:58:59Moosoh "id3 tag editor" plugin updated in the tracker
13:59:58MoosIt would be very cool if one day one core dev. add it to Rockbox, it's a very good feature :)
14:00
14:00:14 Quit Febs (" Try HydraIRC -> http://www.hydrairc.com <-")
14:02:55***Saving seen data "./dancer.seen"
14:02:55Mooshttp://sourceforge.net/tracker/index.php?func=detail&atid=439120&group_id=44306&aid=697085
14:04:40Moosmaybe you Linus, one day when you'll have a bit of time :D (a dream :) )
14:05:26WettLinusN : yes, sorry
14:06:13 Join Nibbler [0] (n=sven@port-212-202-77-14.dynamic.qsc.de)
14:06:58Wettbut mirc doesn't flash when my nick is said x_x
14:07:37LinusNmirc is evil
14:07:45Wettkinda
14:07:56LinusNi use xchat
14:08:19Wettxchat for windows is not freeware... neither mirc anyway
14:08:35Mooshydra IRC :)
14:08:35LinusNanyway, i've seen you asking lots of questions about using libmad in a plugin?
14:09:02Wetthm yes I had problems to compile
14:09:25WettI had to copy/paste a few functions from xxx2wav
14:09:59Wettfor maloc or memcpy
14:10:11LinusNmay i ask why you want this?
14:10:47Wettbecause I'm trying to port the video plugin which use mp3 sound, so I have to convert it to pcm :)
14:11:01LinusNaha
14:11:28Wettoh there also is a little problem
14:11:39 Quit edx (Read error: 110 (Connection timed out))
14:12:06Wettplugin memory is very short... I had to use memory over my allocated space, but I guess it's not very safe
14:13:01Wettin order to avoid dd to spinup continuously
14:13:10LinusNit would be a lot better if we changed the existing codecs to allow for memory playback
14:13:54Wett... Yes but I don't really see how I could do that
14:15:09LinusNbecause...?
14:15:16WettI could write something to copy mp3_playback style
14:16:35Wettbecause I don't understand all of your codec archi, but after all I could give a try, yes.
14:18:46Wettwhat do you think about mp3_playback functions ? May I try to write something plugins could use in the same way ? In fact, that's close to what I'm doing
14:21:37LinusNthat might work
14:23:42Wettand what about the memory space ? It works but...
14:25:05LinusNplugin_get_audio_buffer()
14:26:00Wettoh ! Ok, I still have al lot of things to learn :p
14:26:44amiconnWett:: The archos version also does this
14:27:00amiconn... plugin_get_audio_buffer() I mean
14:27:24Wettyes but I changed it, in fact I don't even remember why....
14:36:52t0mashm
14:36:57t0maselectronics question...
14:37:02t0mashow to use a relay?
14:37:06t0masit has 8 pins
14:37:14t0masI know wich 2 are for the coil
14:37:22t0masbut I have no idea how to connect the others
14:41:51 Join edx [0] (n=A@p54A85586.dip.t-dialin.net)
14:50:28 Join koniu_ [0] (n=koniu@host-ip194-145.crowley.pl)
14:51:48 Join henrikb [0] (n=c354a115@labb.contactor.se)
14:55:07henrikbLinnusN,amiconn I have looked inte the problem when .info files get asociated to .nfo files and i have found the problem, but....
14:56:58henrikbthe cleanest (codewise) way to solve the problem is to add a "." to all extensions in viewers.config and a small code change.
14:58:54henrikband I'm i little bit hesitant to change the viewers.config file without some kind of approval
14:59:08amiconnhenrikb: I think it could be solved in a similar way as the iriver filetype code does it
14:59:41amiconnIt doesn't try to match the extension from the end, but instead searches the last dot (.) first, then matches from there 1
15:00
15:00:48amiconnLook at probe_file_format() in apps/metadata.c
15:01:12henrikbThat was the small code change i had in mind, but it requires the "." in viewers.config
15:01:25amiconnNo, it doesn't
15:02:47 Quit ashridah ("Leaving")
15:05:09amiconnThe "." is implicit
15:05:15 Join paugh [0] (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
15:06:30henrikbWell, you are right but it would not hide another mistake i did in filetypes.c, I'll try to solve both problems instead.
15:07:14amiconnWhat's the other mistake?
15:08:54henrikbThe static extensions, ajz,m3u are saved with a "." dot in the filtypes lookup table, and the plugin extensions are saved without a dot.
15:11:22amiconnI'd say that changing the static ones to not include the dot would be the better way
15:15:02amiconnThis will even save some bytes, perhaps more than you might think
15:15:19 Join linuxstb [0] (n=d556da1b@labb.contactor.se)
15:15:32amiconnThe SH1 compiler aligns strings at 4 byte boundaries, and all strings have the trailing zero byte
15:15:59amiconnSo ".mp3" is 5 bytes which gets rounded to 8 by alignment
15:16:08amiconn"mp3" only takes 4 bytes
15:16:40amiconnAnd of course it allows to keep viewers-config as-is
15:17:19linuxstbOn a similar topic, I think there's still a bug where the file "1.wv" is not recognised as a .wv file. I remember having a quick look a while ago, but couldn't think of a clean way to fix it.
15:18:21amiconnlinuxstb: Hmm. Afaics probe_file_format shouldn't have such a problem...
15:18:58linuxstbI think the problem is elsewhere. It's not recognised as a supported Rockbox file in the file browser
15:20:17amiconnAh, I remember...
15:20:38linuxstbSomeone reported this on the mailing list a few months ago.
15:22:25 Join _Wett [0] (i=Wett@AMontpellier-251-1-35-239.w83-113.abo.wanadoo.fr)
15:25:43LinusNamiconn, linuxstb: line 243 in filetree.c:
15:25:45LinusN if ( !(dptr->attr & ATTR_DIRECTORY) && (len > 4) )
15:26:06LinusN"1.wv" is 4 chars
15:27:55B4gderhaha
15:28:08B4gderlovely
15:29:30linuxstb...and then the subsequent code assumes len > 4 (IIRC)
15:30:12amiconnlinuxstb: I think the solution is simple. the len > 4 check can go away as soon as the filetype check is fixed
15:31:35amiconnI think this check was introduced because of the backward matching of the string, to "ensure" (haha!) that the matching doesn't start before the start of the string
15:36:47 Quit Moos (Read error: 104 (Connection reset by peer))
15:37:02 Join Moos [0] (i=DrMoos@m29.net81-66-158.noos.fr)
15:37:47 Quit Wett (Read error: 110 (Connection timed out))
15:38:29linuxstbYes - ".flac" is longer than "1.wv". Would the solution be to modify filetype_get_attr to locate the extension of the filename being tested (strrchr?), and then do the comparison on that pointer?
15:40:40henrikblinuxstb, yes but you need a few more changes also, I'm on the case right now
15:42:05amiconnlinuxstb: That's what I'm trying to say all the time
15:42:44amiconnfiletype_get_attr() should work the same way as probe_file_format()
15:52:57pillhi
15:53:32pilli dont understand how %?fc works now that it's conditional
15:53:58LinusNit works like it always has
15:54:18LinusNuse %fc if you want the usual "mp3" string
15:54:26pillwell before i had %?fc<%fc|track>
15:54:34pillnow, that gives me ???
15:54:47linuxstbamiconn: Yes, sorry. I hadn't looked at probe_file_format.
15:55:40LinusNpill: what type of file are you playing?
15:55:45pilland i tried to paste your example, and it seems to work, but not with %?Fc
15:55:51pillmostly .mp3 and .ogg
15:56:12LinusNok, so %fc works but not %Fc?
15:56:18pillyes
15:56:27pill%Fc shows random types
15:56:56LinusNrandom?
15:57:14pillwell, it showed the coming file as MP2, and then as MP1
15:57:20pilland it was a vorbis
15:57:33pilltried on several files
15:57:50LinusNah, i see
15:58:53pillcouldn't we add a %fe for file extension?
15:59:03pillthat would save some of us the hassle
15:59:17LinusNhassle of what?
16:00
16:00:03pillwell, for most people i think that the fileext is correctly refering to the codec
16:00:18pilli understand that conditional %fc is better
16:00:37pillbut i would be happy with just the fileext
16:00:43pill(but that may be just me)
16:02:58***Saving seen data "./dancer.seen"
16:03:40amiconnThe file extension uniquely identifies the format *atm*, but this will very likely change in the future
16:04:14*pill shrugs
16:04:27linuxstbThe future will be next week (?), when I commit my ALAC decoder, it won't make much sense - ALAC will be displayed as "m4a".
16:04:41LinusNpill: so just having %Fc gives random codec info?
16:05:04LinusNor do you have to have it in a conditional?
16:05:15pill%?Fc<MP1|MP2|MP3|WAV|OGG|FLAC|MPC|A52|WAVPACK|track>
16:05:21pillthis gives randomness
16:05:36amiconnlinuxstb: With alac it's just a slight oddity. It'll get really interesting if we want to support ogg/vorbis vs. ogg/flac vs. ogg/speex
16:05:40preglowwill the freeze end even though there are quite a few bugs left?
16:05:49amiconn...or mp1/2/3 in a .wav container...
16:05:56pillhold, i'm trying constant %Fc
16:06:36linuxstbamiconn: I agree, but the problem with AAC/ALAC is identical.
16:06:40LinusNpreglow: we'll see
16:06:45linuxstb(identical to Ogg).
16:07:47linuxstbBut I think the solution is simply to ensure that we always call the get_metadata() function before the codec is loaded. The get_metadata() function will then have the responsibilty of confirming the filetype.
16:10:32pillLinusN: %?Fc<%Fc|track> shows ??? then nothing
16:11:05linuxstbI think the problem is that playback.c is setting the codectype to zero - to indicate the codec has not yet been loaded.
16:11:33linuxstbSo these seem to be bugs in playback.c - but I'm not worried, because I want to change how playback.c is working in that area anyway.
16:12:00linuxstb(to implement the get_metadata() call I mention above).
16:12:24LinusNpill: try this %?Fc<%?Fc<MP1|MP2|MP3|WAV|OGG|FLAC|MPC|A52|WAVPACK>|track>
16:12:34LinusN(i see the bug)
16:12:34pill%?Fc<MP1|MP2|MP3|WAV|OGG|FLAC|MPC|A52|WAVPACK|track> exactly shows MP2 for 2 seconds, then MP1, and then after more buffering is done, shows the correct file type.
16:12:42pillk
16:12:44pillhold
16:18:02pilldamn my usb connection is not working
16:20:20LinusNi gotta go
16:20:22LinusNcu later
16:20:23pillk
16:20:26pilli'll tell you
16:20:33pillonce i have that usb thing fixed
16:20:35pillcya
16:20:36LinusNi see the bug in the wps code
16:20:38 Part LinusN
16:28:34preglowhmmm
16:28:42preglowi wonder why the flaming hell layer12 is so slow
16:29:32 Join ripnetuk [0] (n=george@82-70-100-230.dsl.in-addr.zen.co.uk)
16:29:40 Part ripnetuk
16:30:55linuxstbpreglow: I liked your idea of splitting libmad into two codecs, each optimised for either L3 or L1/2 . Are you still thinking about doing that?
16:31:11preglowkind of
16:31:22preglowit's not somethign i'm going to do right now
16:31:24preglowbut it is possible
16:31:42preglowthe entire layer12 code and data would probably fit in iram
16:31:55preglowproblem is it wouldn't benefit layer3 too much, but anyway
16:33:30preglowit's not THAT much of a problem
16:33:37preglowbut
16:33:49preglowit will require a better codec decision mechanism
16:33:55preglowa lot of layer2 files are called .mp3
16:35:16preglowmeasuring codec efficiency with cpu boost factor was not very productive...
16:40:46Slashernice, now there seems to be enough space in iram to put even voice codec there
16:44:26amiconnSlasher: You mean the voice codec stack?
16:44:32Slasherah, yes
16:44:42amiconnI think that might be possible...
16:44:53Slasherthat should give a major performance boost (testing now)
16:44:56preglowhmm?
16:45:01preglowis there any more spare space?
16:45:08amiconnI reduced the main stack by 24KB, but the core IRAM by only 16KB
16:45:20amiconnThat gives 8KB of extra usable IRAM in the core
16:45:32preglowa good coded stack requires almost 32kb
16:45:35preglowcodec
16:45:47Slasherhmm.. in fact the iram will get quite full after that change..
16:46:19Slashercurrently DEFAULT_STACK_SIZE + 0x2500 bytes are allocated for one codec in iram
16:47:04preglowcalculating it that way is a bit hazardous
16:47:10amiconnYes. That's a little above 10KB
16:47:10preglowDEFAULT_STACK_SIZE just changed to 8kb, no?
16:47:17preglowexactly
16:47:21preglowlibmad needs more than that
16:47:32amiconnNo, DEFAULT_STACK_SIZE is and was 0x400
16:47:42amiconn(1 KB)
16:48:11preglowahh
16:48:20amiconnThe codec that eats most is libmad, and the debug menu says the codec thread uses 80% of its stack
16:48:31amiconnThat means a little above 8KB
16:48:43preglowwow
16:48:47preglowi thought it needed heaps more than that
16:49:15amiconnI only changed the main thread's stack size
16:49:54preglowyes, know
16:50:04amiconnbtw, 0x400 0x2500 is an odd value...
16:50:14SlasherHmm, allocating the voice codec stack from iram seems to increase performance much
16:50:19amiconn0x2900 == 10496
16:50:49amiconnBleh 0x400 _plus_ 0x2500
16:51:05amiconnDamn CGI::IRC
16:51:24preglowhahaah
16:51:30*preglow huggles irssi
16:51:47amiconnMy home network connection vanished for some reason
16:54:53 Join webguest77 [0] (n=3e4f4094@labb.contactor.se)
16:55:39webguest77i think i found abug in playback on iriver . if you play a file and stop halfway through then delete the file and press play to resume it will act strange
16:55:56preglowyes, the iriver playback code handles missing files badly
16:56:15 Join Sanitarium [0] (n=Test@213-202-138-155.bas502.dsl.esat.net)
16:56:20Sanitariumhey guys
16:56:29webguest77it just chnages the statusbar icon to play and nothing else
16:56:49 Quit webguest77 (Client Quit)
16:59:56 Join Sanit [0] (n=Test@213-202-138-155.bas502.dsl.esat.net)
17:00
17:00:52Sanithey huys, sorry about that
17:01:19Sanityou guys need any help with anything?
17:03:01 Join LameD [0] (n=55406634@labb.contactor.se)
17:03:32LameDI've been trying to build both on today's and yesterday's build. what am I doing wrong?
17:03:48 Join drfeelgood [0] (n=44256200@labb.contactor.se)
17:03:54B4gderLameD: what happens?
17:03:55LameDCC words.c
17:03:55LameDCC wputils.c
17:03:55LameDCC coldfire.S
17:03:55DBUGEnqueued KICK LameD
17:03:55LameDAR /home/Liberman/rb01092005/build/libwavpack.a
17:03:55LameDMAKE in dumb
17:03:56***Alert Mode level 1
17:03:56LameD(dumb) CC src/core/atexit.c
17:03:58LameD(dumb) CC src/core/duhlen.c
17:04:00LameD(dumb) CC src/core/dumbfile.c
17:04:02LameD(dumb) CC src/core/loadduh.c
17:04:04LameD(dumb) CC src/core/makeduh.c
17:04:06LameD(dumb) CC src/core/rawsig.c
17:04:08LameD(dumb) CC src/core/readduh.c
17:04:10LameD(dumb) CC src/core/register.c
17:04:11 Quit LameD (Excess Flood)
17:04:24 Join LameD [0] (n=55406634@labb.contactor.se)
17:04:31LameDsorry for flooding
17:04:38linuxstbLameD: Which version of gcc are you using?
17:04:45LameDlemme C
17:04:51preglowgcc 3.3.x, i'll wager
17:05:06LameD(the newest, just installed cygwin yesturday)
17:05:10LameD(but checking)
17:05:12B4gdeririver target?
17:05:14LameDyeah
17:06:05linuxstbType "m68k-elf-gcc −−version"
17:06:20LameD3.3.5.
17:06:31linuxstbThat's the problem - get 3.4.4
17:06:47LameDthanks!
17:06:57LameDwas thet that SED problem?
17:07:05LameDe=>a
17:07:06B4gderno
17:07:14B4gderconfigure detects the sed problem
17:07:38LameDso what's wrong with the gcc?
17:08:45B4gderI don't know
17:09:12LameDno wait.. that's strange, cygwin installer reports "gcc-core: C compiler 3.4.4-1" is installed. I'm redownloading
17:09:32preglowyou sure that's the cross compilder?
17:09:39preglowm68k-elf-gcc −−version
17:10:06henrikbTime to go home,
17:10:49henrikbI'll clean up any eventual messes i caused with my commit when I get home
17:10:53linuxstbcygwin will be reporting the native gcc, not the cross-compiler that you built yourself.
17:10:55henrikbBye
17:11:09 Quit henrikb ("CGI:IRC")
17:11:11LameDI'm sure cygwin reports in bash, version 3.3.5, but when I entered Setup.exe to dl the older version I saw there v. 3.4.4
17:12:29linuxstbYou will have two versions of gcc - the native version "gcc", and the Coldfire version - "m68k-elf-gcc".
17:12:57LameDafter reinstalling, bash still reports version 3.3.5 is installed
17:12:57 Quit drfeelgood ("CGI:IRC (EOF)")
17:13:30linuxstbThat's perfect for "gcc". For "m68k-elf-gcc", you need 3.4.4. You use your "gcc" to compile "m68k-elf-gcc".
17:13:56LameDyeah, i've got it... here's a copyNpaste
17:13:57***Alert Mode OFF
17:14:12LameDLiberman@libermanim ~
17:14:12LameD$ m68k-elf-gcc −−version
17:14:12LameDm68k-elf-gcc (GCC) 3.3.5
17:14:12DBUGEnqueued KICK LameD
17:14:12LameDCopyright (C) 2003 Free Software Foundation, Inc.
17:14:12LameDThis is free software; see the source for copying conditions. There is NO
17:14:13***Alert Mode level 1
17:14:13LameDwarranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
17:14:19LameDLiberman@libermanim ~
17:14:21LameD$
17:14:40B4gderthen update it
17:14:44linuxstbThat's the problem. You need to delete that installation, and compile 3.4.4
17:15:12LameDwait you're right i've dl'ed the gcc core :/
17:16:56 Quit Sanitarium (Read error: 110 (Connection timed out))
17:17:23LameD( I even remember looking at the version numbers in the document and seeing this is a newer version) but thanks i'm downloading
17:18:21 Quit amiconn ("CGI:IRC")
17:18:37 Join B4gd3r [0] (n=daniel@static-213-115-255-230.sme.bredbandsbolaget.se)
17:19:41 Quit webguest68 ("CGI:IRC (EOF)")
17:23:28 Quit B4gder (Read error: 60 (Operation timed out))
17:24:14***Alert Mode OFF
17:24:20preglowdoes codecs too link with plugin.lds?
17:24:38B4gd3ryes
17:24:50preglowwhere does codeclink.lds come rom?
17:24:52preglowfrom
17:25:26B4gd3rit is the output of the preprocess the build makes
17:26:13B4gd3rplugin.lds is preprocessed and then the codeclink.lds is used in the actual link command line
17:26:23preglowgood, that's what i thought
17:26:48B4gd3rthe plugins and codecs all use plugin.lds, just with some different defines
17:26:58preglowi'm just gonna add a new segment anyway
17:27:12preglowsection, that is
17:29:52preglowhmm
17:29:56preglowwhat's the KEEP macro good for?
17:30:51B4gd3rI only know what the lds file says ;-)
17:31:35 Join xen` [0] (n=xen@stg25-1-82-238-117-1.fbx.proxad.net)
17:32:20B4gd3rgoing home
17:32:21 Quit B4gd3r ("Lämnar")
17:36:30 Quit paugh ("Leaving")
17:37:20LameDwhy is it reported on the plugins page that chip8 emulator runs over iriver?
17:37:56LameDhttp://www.rockbox.org/twiki/bin/view/Main/PluginIndex
17:38:11 Join webguest45 [0] (n=d5ee48b7@labb.contactor.se)
17:47:27preglowwell, doesn't it?
17:52:44 Quit ender` (Read error: 113 (No route to host))
17:56:41 Quit arkascha (Remote closed the connection)
17:57:07 Join ender` [0] (i=ychat@tm.213.143.74.124.dc.telemach.net)
17:59:53 Nick Mark__`away is now known as Mark__ (n=Mark@cpc1-bele3-3-1-cust167.belf.cable.ntl.com)
17:59:53DBUGEnqueued KICK Mark__
18:00
18:02:40 Join Lear [0] (n=chatzill@h179n2c1o285.bredband.skanova.com)
18:02:59***Saving seen data "./dancer.seen"
18:09:19 Join amiconn [0] (n=jens@p54BD4D07.dip.t-dialin.net)
18:10:08LameDPreglow: Nope. It's not
18:12:24LameDand also the metronome.
18:12:55Mooschip8 working in irivers
18:13:14Mooschek the wiki page for dowloads games...
18:13:20Moos*check
18:13:25LameDworks for you? i'll go dl some other games
18:13:52LameDbut there's no metronome... you can't argue with that can't you ?
18:14:02Moosyes it work, since few weeks, Frederic works
18:14:16Moosmetronome, not yet
18:18:08LameDyou're right, I had broken games all that time on my player. thanks
18:18:22Moosnp :)
18:19:47 Join henrikb [0] (n=d9d77cc4@labb.contactor.se)
18:20:32LameDi've got to say, it's kinda sucks asses
18:22:38henrikbAnother space saver would be removing scan_plugins from filetype.c, it asociates viewers to extension based on a naming scheme and I don't think anybody uses it
18:23:26 Join us3r [0] (n=18d79b85@labb.contactor.se)
18:23:33LameDyes but I still believe chip8 games would still suck badly.
18:23:53 Join DangerousDan [0] (n=Miranda@newtpulsifer.campus.luth.se)
18:23:58us3rcredits.pl seems to be missinf from the source code file....
18:24:16MoosLameD: feel free to improve it ;)
18:24:56henrikbme fix
18:26:25LameDI've tryed banging my player on the floor. now it's much better, not moving too fast that's for sure!
18:28:58henrikbFixed, it will appear in the next daily tarball
18:29:15us3r:) ty
18:30:58 Nick Mark__ is now known as Mark__`afk (n=Mark@cpc1-bele3-3-1-cust167.belf.cable.ntl.com)
18:30:58DBUGEnqueued KICK Mark__`afk
18:31:18 Quit us3r ("CGI:IRC")
18:31:33 Join bagawk [0] (n=lee@67-42-194-6.eugn.qwest.net)
18:40:54 Quit linuxstb ("CGI:IRC")
18:47:17 Quit LameD ("CGI:IRC (EOF)")
18:51:55 Join QT [0] (i=as@madwifi/users/area51)
19:00
19:01:30 Quit xen` (Read error: 110 (Connection timed out))
19:01:51 Join QT_ [0] (i=as@madwifi/users/area51)
19:04:26 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
19:04:30 Quit henrikb ("CGI:IRC (Ping timeout)")
19:11:18 Join paugh [0] (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
19:11:54 Quit QT (Read error: 104 (Connection reset by peer))
19:16:29 Nick QT_ is now known as QT (i=as@madwifi/users/area51)
19:20:47Sanitanyhelp needed for the RB H300 project
19:27:22 Join Paul_The_Nerd [0] (n=chatzill@cpe-66-68-93-2.austin.res.rr.com)
19:29:59 Quit Paul_The_Nerd (Client Quit)
19:33:54 Quit Nibbler ("life is like a rental car, you fuck it up, and give it back.")
19:34:01 Join Nibbler [0] (n=sven@port-212-202-77-14.dynamic.qsc.de)
19:48:33 Quit Lear ("Chatzilla 0.9.68.5 [Firefox 1.0+/undefined]")
19:54:53 Nick Mxm`PasLa is now known as Maxime (n=flemmard@fbx.flemmard.net)
20:00
20:03:01***Saving seen data "./dancer.seen"
20:03:21solex_aw1yDoes anybody know details about the new conditionals in the wps?
20:06:21solex_aw1y(and why tf am i not able to change my nick!?)
20:06:44 Quit koniu_ (Read error: 104 (Connection reset by peer))
20:11:16 Quit solex_aw1y ("leaving")
20:11:50 Join solex_ [0] (n=jrschulz@c147039.adsl.hansenet.de)
20:12:22solex_conditional image display doesn't work here
20:12:34solex_i took the example from the wiki page
20:12:48solex_%?mm<%xdb|%xdc|%xdd|%xde>
20:13:01solex_well, almost: %?mm<%xdb|%xdc|%xdd|%xdc>
20:13:54solex_hm, everybody's asleep, i suppose.
20:14:36Maximesure
20:14:36Maxime:x
20:14:42MaximeI can't help so..
20:18:51 Join [1]Moos [0] (i=DrMoos@m29.net81-66-158.noos.fr)
20:18:51 Quit Moos (Read error: 104 (Connection reset by peer))
20:18:55 Nick [1]Moos is now known as Moos (i=DrMoos@m29.net81-66-158.noos.fr)
20:20:50 Join [1]Moos [0] (i=DrMoos@m29.net81-66-158.noos.fr)
20:20:50 Quit Moos (Read error: 104 (Connection reset by peer))
20:20:54 Nick [1]Moos is now known as Moos (i=DrMoos@m29.net81-66-158.noos.fr)
20:22:55dpassen1are all the images pre-loaded?
20:23:17solex_yep. I did it exactly like in the wiki
20:23:51dpassen1the images are all 1-bit bmps?
20:24:30solex_i think so. wait...
20:25:36solex_fsck
20:25:47solex_24...
20:28:47 Quit bagawk ("Leaving")
20:36:53 Join koniu_ [0] (n=koniu@host-ip194-145.crowley.pl)
20:39:14 Join arkascha [0] (n=arkascha@xdsl-213-196-214-210.netcologne.de)
20:46:00 Nick paugh is now known as AliasCoffee (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)
20:46:32 Quit koniu_ (Read error: 104 (Connection reset by peer))
20:47:34 Quit arkascha (Remote closed the connection)
21:00
21:05:20 Quit QT (Read error: 104 (Connection reset by peer))
21:06:47 Quit Mark__`afk ("Leaving")
21:19:09 Join QT [0] (i=as@madwifi/users/area51)
21:23:42 Join pilot000 [0] (n=c31ce021@labb.contactor.se)
21:24:06pilot000solex: still here ?
21:32:03 Join XavierGr [0] (n=XavierGr@ppp10-adsl-43.ath.forthnet.gr)
21:32:55solex_um, no :)
21:34:24pilot000ohh i see
21:34:36pilot000still problems with wps ?
21:35:32solex_not really, i am just too dumb to convert my images.
21:35:38solex_but i am buy atm
21:35:42solex_*busy
21:36:39pilot000then see you later, i finish my week in 20 minutes
21:45:34 Quit pilot000 ("CGI:IRC")
22:00
22:03:04***Saving seen data "./dancer.seen"
22:07:16XavierGranyone here with an Archos recorder or Ondio?
22:13:44 Quit XavierGr ()
22:19:14Sanitanyhelp needed for the RB H300 project
22:26:39 Quit banan_ (Read error: 104 (Connection reset by peer))
22:34:28solex_fyi: after converting my images to 1 bit per pixel everything works fine.
22:34:40solex_i will update my wps in the gallery in the next few days
22:34:47solex_(me == JochenSchulz)
22:35:08solex_it will have a custom "status bar" with the original symbols
22:35:16solex_only a little bit higher
22:35:38dpassen1glad you got it working
22:36:16solex_thanks. it's not finished yet (only tested conditionals for repeat mode)
22:36:38solex_but the new conditionals are great
22:36:45solex_no more nested %?
22:37:27solex_does anybody have an idea why i can't change my nick with /nick?
22:37:39CtcpIgnored 4 channel CTCP requests in 1 hour and 27 minutes at the last flood
22:37:39*solex_ is still new to IRC
22:37:39dpassen1i'm fairly sure that nested conditions still work
22:37:54solex_dpassen1: sure, but they're ugly and error prone
22:38:11dpassen1what message appears in the server window when you try /nick
22:38:18solex_nothing at all
22:38:30dpassen1maybe eventually some ANDs or ORs will appear
22:38:33solex_client: irssi
22:38:38 Nick dpassen1 is now known as dpassen (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
22:38:40 Nick dpassen is now known as dpassen1 (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
22:39:02solex_usually i have some kind of status window on window 0
22:39:11solex_but it isn't there
22:39:32solex_but well, i can live with the underbar
22:46:33 Nick amiconn is now known as am1c0nn (n=jens@p54BD4D07.dip.t-dialin.net)
22:46:41 Nick am1c0nn is now known as amiconn (n=jens@p54BD4D07.dip.t-dialin.net)
22:47:33solex_ah, now i get a message: nick already in use
22:47:41solex_which is strange since i registered it
22:48:13amiconnThen you can regain it
22:48:20solex_funny, i get the /msg
22:48:31solex_amiconn: how do i do that?
22:49:22amiconn. /msg nickserv recover solex <yourpassword>
22:49:36amiconn(of course w/o the dot)
22:49:46solex_msg nickserv recover solex madsci
22:49:54solex_msg nickserv recover solex mad_sci
22:50:02Maximebut with the /
22:50:02Maxime:x
22:50:12*solex_ curses
22:50:20solex_how dumb can a single person be?
22:50:27*solex_ blushes
22:50:34amiconnsolex_: That also happened to me once
22:50:49Maximewe haven't seen anything :p
22:50:57amiconnMy slash wasn't there because dreaded numlock was on (laptop)
22:51:00solex_and nobody'sloggin, of course
22:51:24amiconnchange your pw afterwards
22:51:39amiconn. /msg nickserv help is very useful
22:51:45solex_thanks!
22:58:07amiconnhttp://www.rockbox.org/irc/rockbox-20041009.txt <== if you wanna see my mistake
22:58:23amiconnLook around 09:58
23:00
23:00:24solex_:)
23:00:38solex_either i forgot my real pw or someone else was faster
23:01:05*solex_ now knows why he uses unimportant passwords for things like this
23:01:47solex_but i didn't know there are public logs
23:04:28solex_fsck it
23:23:57 Join Aison [0] (n=hans@zux166-181.adsl.green.ch)
23:27:24 Nick AliasCoffee is now known as AliasAle (n=pete@2001:5c0:8fff:ffff:8000:0:3e03:6822)

Previous day | Next day