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 2005-09-18

00:00:16rasherUpdate 13 Sep 2005: Please upgrade to TWikiRelease03Sep2004 or apply patch for SecurityAlertExecuteCommandsWithRev. −− PeterThoeny - 13 Sep 2005
00:00:17amiconn ,
00:00:27 Quit sneakums ("taster, you're the automatic saint")
00:01:19rasherMan, every release were affected
00:01:38rasherah, no
00:02:37Bagderthe rev thing was used against us
00:02:51solex_um, where's the publication date of these alerts?
00:03:28Bagderthey are wiki pages, you can check their history
00:03:43Bagderthe rev page is from sep 9 2005
00:04:09rasher Researcher discloses issue 2005-09-09
00:04:25solex_thx, found it, too
00:04:35amiconnThat's ... odd
00:04:55Bagderand they seemed to have announced it widely on sep 13
00:05:15linuxstbWas Rockbox included in " 5. Compile e-mail list of administrators of public TWiki sites" ?
00:05:28solex_and published it on sep 16
00:05:38solex_not exactly 0-day, but close
00:06:27Bagderthe actual delete was made 23:17 CET on the 16th
00:06:38amiconnWhy do they publish an article about such a serious security hole one year after fixing it???
00:07:04solex_amiconn: there have been two vulnerabilities
00:07:07Bagderthis was not fixed
00:07:10Bagderuntil just then
00:07:28amiconnBagder: My first thought was that you or Linus took it offline because of someone mentioning this flaw here in the channel...
00:07:29Bagderaccording to that page at least
00:07:48solex_the first one (from 2004) was used to attack a wiki from the german chaos computer club
00:07:50Bagderamiconn: I noticed that mention and I mailed both Linus and Zagor
00:07:57Bagderbut it wasn't good enough
00:07:59solex_(europe's largest hacker org)
00:08:25amiconnBagder: I know that there were two vulnerabilities, the Rev thing being the second
00:08:28Bagderin retrospect, of course I should've shutdown the twiki parts
00:08:47solex_Bagder: well, you learn from these things
00:09:17Bagderwe've been hit before :-/
00:09:23solex_after reading #rockbox today, i immediately switched off clamav on my machine since there's a vulnerability, too-
00:09:31amiconnHowever says that releases up to TWikiRelease02Sep2004 are vulnerable, and TWikiRelease03Sep2004 fixes it
00:09:34amiconn2004 (!)
00:10:21rasher NOTE: If you are running an unmodified TWikiRelease02Sep2004, simply copy the patched lib/TWiki/,
00:10:32amiconnAh, an additional patch is needed, but that's from Feb 2005...
00:11:02BagderI think it is just their naming that is confusing
00:11:06***Saving seen data "./dancer.seen"
00:11:09Bagderthey seem to have made the fix on sep 11 2005
00:11:38Bagdernot that it matters now
00:12:21amiconnHmm, indeed a confusing scheme....
00:12:46 Quit gromit` (Remote closed the connection)
00:18:51 Join gromit` [0] (
00:29:58amiconnAnother archos firmware test recording hung after 10 h 47 min
00:31:23 Quit t0mas (Read error: 104 (Connection reset by peer))
00:31:37Bagder36 different IP addresses have used the twiki flaw
00:33:41Bagderseems to be a popular backdoor attempt
00:34:13rasherooh, they've been trying to backdoor you?
00:34:28Bagderlots of those guys tried hard
00:34:57*amiconn 's machine is lagging like hell atm
00:35:18amiconnLoading a 700 MB file into the hex editor...
00:35:47rasherWhy is your hex editor caching the entire file?
00:35:56amiconnFast searching...
00:36:05rasherSounds silly for such a large file
00:36:15amiconnNow it's running fine.
00:36:22amiconnMy machine has 1 GB of RAM
00:37:09rasherThat leaves 300mb.. I guess XP is breathing heavily now :(
00:37:25amiconnNo, everything running smooth again
00:37:43amiconn(with like a dozen programs opened in parallel)
00:40:09 Join webguest67 [0] (
00:41:33webguest67hi everybody
00:42:04webguest67hi rasher
00:43:27webguest67i just started getting into rockbox and when i load a wps with a bitmap it doesn't show up, any ideas of how to fix this?
00:43:50rasherUse the correct syntax :)
00:43:55rasherI'd point you at the wiki, but..
00:43:56rasherhang on">
00:44:31 Quit webguest67 (Client Quit)
00:57:29Bagdersomeone did run processes on the server
00:58:25Bagderas the http user
00:58:46rasherwould need an apache exploit for that to be really useful
00:58:58Bagderor any other local exploit
00:59:10rasheroh, true that
00:59:40rasherwell at least there's a pretty well-documented point-of-entry
01:00:31fuzzieit's a lot more worrying when you can't tell where they got in
01:00:52Bagderbut still, seing a 'bash' running that ain't us...
01:01:34Bagderbut I really don't want to shut it down completely
01:04:31 Join webguest67 [0] (
01:07:48BagderI've kept the scripts for later analysis, now I need sleep
01:09:19webguest67that worked!
01:09:36webguest67thanks rasher
01:10:00rasherYou're welcome.
01:13:48 Quit Moos ("Parti")
01:21:05 Join davidd [0] (i=davidd@
01:21:30 Quit matsl (Remote closed the connection)
01:26:05 Join Paul_The_Nerd [0] (
01:31:11 Join ashridah [0] (
01:31:30 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
01:35:29 Quit davidd ()
02:04:06 Part Shapeshifter
02:04:09 Quit Domonoky (Read error: 104 (Connection reset by peer))
02:06:12 Join CheeseBurgerMan [0] (
02:07:59 Quit dpassen1 ()
02:11:10***Saving seen data "./dancer.seen"
02:19:14 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.6/20050716]")
02:29:50 Quit noC|andY`fRa (Read error: 104 (Connection reset by peer))
02:41:08 Join DJ_Dooms_Day [0] (
02:53:43 Join DarkShadow [0] (
02:57:29 Join noiz [0] (
02:57:55 Join bitshift [0] (
02:59:37DarkShadowI'm having a problem with creating a song database for my H120.
02:59:38bitshiftwhats happening
02:59:48bitshiftany info for rockbox on iaudio x5?
03:00:00DarkShadowWhen I run the .bat to make the database, a dos prompt just flashes onto my screen, then closes, and it doesn't do anything.
03:00:41rasherbitshift: not really.. haven't heard from the guy working on it for a while
03:01:21bitshiftwhat involves developing such software for the device?
03:02:12rasherFirst making a way to run your own code on the device, then writing drivers for Rockbox, I guess
03:02:37rasherDarkShadow: what error message do you get?
03:03:09DarkShadowI don't get an error message.
03:03:21rasherwhat does it say in the dos prompt?
03:03:33DarkShadowThe dos prompt just pops up too fast for me to read, then closes.
03:03:55rasherrun the .bat from within a dos prompt
03:04:17DarkShadowHow do I get to F/.rockbox though?
03:04:29DarkShadowWhen I tried it, the dos prompt didn't like the period before the folder.
03:04:53rasherhaha, silly windows
03:04:56rasherno idea
03:05:36fuzziecd ".rockbox" works here
03:05:42DarkShadowI think there's a code thing you can put on the end /(something) when running it to make it stop...not sure what it is though.
03:05:58DarkShadowWith the parenthesis?
03:06:24fuzzieno, with the double-quotes.
03:07:00fuzzieunless i'm mistaken about the meaning, dictionary isn't very helpful
03:07:21DarkShadowIt says Invalid Directory.
03:07:34fuzzieand this is win2000/XP?
03:07:35DarkShadowIt even shows it when I do 'dir'... it has it with the period before too.
03:07:43DarkShadowYeah, XP.
03:08:00fuzziethat's strange
03:08:07rasherAnd I guess you're on the right drive if you see it
03:08:11fuzzieif you use properties on the batch file, can you not configure it so it doesn't close the window?
03:08:28DarkShadowI'll check.
03:09:15DarkShadowIt doesn't really have anything.
03:11:59DarkShadowI renamed rockbox without the period, and ran it in command, and now it's working.
03:12:06DarkShadowIt had an error, I was missing a file.
03:12:40 Join DJ_Dooms_Day|AF [0] (
03:12:41DarkShadowDo both of you use the DB?
03:14:06 Quit DJ_Dooms_Day (Read error: 104 (Connection reset by peer))
03:14:26 Join Script [0] (
03:20:17DarkShadowMy H120 says 'Codec Failure' now...and I can't turn it off.
03:20:44rasherDid you move .rockbox back?
03:21:25rasherreinstall rockbox, and make sure you overwrite all files
03:21:35DarkShadowWell... I couldn't rename it I copied it back over.
03:21:50DarkShadowThat's what I was planning on doing, but I can't get this to do anything.
03:21:57DarkShadowIt's like frozen.
03:22:09DarkShadowPlugging it into USB doesn't do anything either.
03:22:18rasherReset it then
03:22:45DarkShadowAh...there's the button. ;)
03:26:10DarkShadowHow do I restart everything?
03:26:33DarkShadowIt says "error accessing playlist control (file?)'
03:27:19rasherplug in usb and reinstall
03:27:46rasherWhat do you mean how?
03:27:52rasherYou've done this before, surely
03:28:07DarkShadowUhh, is it called flashing it?
03:28:27rasherno... just unzip unto the drive
03:28:49DarkShadowI did...and it still has that error.
03:29:26rashercan't you just ignore it
03:30:15DarkShadowThe boot thing comes up, then the Rockbox logo, then it pops up infront of the Rockbox picture... "Error accessing playlist control file"
03:31:05rashercan't you just make it go away?
03:31:11rasherpress stop or select or something
03:32:16DarkShadowah... I got it now.
03:32:25DarkShadowHow do I get that playlist thing working though?
03:32:27 Quit Riddy (Read error: 110 (Connection timed out))
03:32:55rashershould be fixed next time you create a playlist I think
03:34:28DarkShadowAh...brilliant, thanks.
03:36:41DarkShadowAnother thing... I was reading through the forums the other day how Rockbox had the problem of letting the mp3 player run out of battery, and that polymer batteries get damaged when they run out of batteries.
03:36:53DarkShadowSo...just the other day, I didn't know it was on, and it totally ran out of battery.
03:37:44rasherI don't think Rockbox can run them far enough down as to damage them
03:37:52rashernot sure though
03:39:13DarkShadowOkay...then what would the worry be?
03:39:33rasherthat Rockbox failed to actually shut down
03:39:49rasherbecause it tried to spin up the harddisk, which it couldn't
03:39:50DarkShadowSo only that it wouldn't save the playlist or something?
03:40:19rasherwell that, and it'd sit there trying to spin up the harddisk
03:40:43DarkShadowWell that's a relief.
03:40:46DarkShadowThanks much! :)
03:41:07DarkShadowDo you use the database?
03:41:31rashernot really
03:41:49DarkShadowJust file tree?
03:48:30 Quit CheeseBurgerMan (Read error: 104 (Connection reset by peer))
03:50:39DarkShadowIs there a limit to the speed that scrolling increases two when going down a big list?
03:50:52DarkShadowIt takes me like a minute and a half to scroll through the list of all my songs.
03:51:31rasherno idea
03:51:32 Quit rasher ("Ex-Chat")
03:52:48 Quit ender` (Read error: 110 (Connection timed out))
03:55:36DarkShadowAnyone here now?
03:56:44 Join CheeseBurgerMan [0] (
03:57:19DarkShadowHow do you record with rockbox?
03:57:30DarkShadowMessing around with it for a while now, I haven't found it. :-P
03:58:12 Join hd [0] (
03:58:48DJ_Dooms_Day|AFi dont think you can yet
03:59:03DJ_Dooms_Day|AFtheir still adding a lot of thins
03:59:19DarkShadowI guess that would explain it. :-P
04:03:19DarkShadowHow do you install .patch files? (I'm trying to have a different keyboard−−keyboard.patch)
04:03:56fuzziei think there's a record option in the debug menu?
04:04:00fuzziebut .. it breaks stuff
04:05:42DarkShadowAlright, I'll avoid it then.
04:05:45 Join QT [0] (i=as@madwifi/users/area51)
04:11:14***Saving seen data "./dancer.seen"
04:14:15 Quit goa (Connection timed out)
04:14:16 Nick hd is now known as goa (
04:15:13DarkShadowWell I think I'll head off for now.
04:15:19DarkShadowThanks so much for all your help, Fuzzie.
04:16:57 Quit DarkShadow ("CGI:IRC (EOF)")
04:17:26 Quit QT_ (Connection timed out)
04:40:32 Quit CheeseBurgerMan (Read error: 104 (Connection reset by peer))
04:47:33 Quit cYmen ("zZz")
04:59:39 Join pengo [0] (n=xtofu@
04:59:49 Quit DJ_Dooms_Day|AF ()
05:15:32 Part noiz ("Leaving")
06:11:18***Saving seen data "./dancer.seen"
06:21:08 Quit webguest67 ("CGI:IRC")
06:22:34 Quit bitshift ("fractal2 mirc script (ver 1.0betar2) ·")
07:50:04 Quit ashridah (Read error: 110 (Connection timed out))
07:50:41 Join ashridah [0] (
08:11:20***Saving seen data "./dancer.seen"
08:42:47 Quit Febs (Read error: 110 (Connection timed out))
09:13:06*HCl yawns
09:20:24 Join ender` [0] (i=ychat@
09:23:57SlasheriWhat was the maximum file size rockbox supports? Just thinking if i could record 4h continuously with iriver..
09:25:02Slasheriif it's around 2GB (2^31), that is not possible then because there is no currently file splitting
09:50:46amiconnYou shouldn't try more, as the recording code doesn't check for that iirc
09:51:31amiconn(on iriver)
09:52:11Slasheriok, then i will try manually stopping and restarting the recording after two hours :)
09:52:44Slasheribtw, does original firmware support longer recordings to wav files?
09:54:33amiconnI have no idea
09:54:42amiconnNever recorded with the original firmware
09:54:55Slasheriah, almost same here
09:55:21SlasheriHmm, or maybe i have enough time to code the splitting feature before recording.. =)
09:59:24 Quit ender` (Read error: 104 (Connection reset by peer))
10:02:11pengoSlasheri, ihp120 has recording limits under 60 mins.. definately wouldn't record longer than 74min.. there were different limits for wav and mp3 recording.. can't remember exactly
10:03:48 Join ender` [0] (i=ychat@
10:04:20Slasherihuh.. that's not good
10:04:33pengo(with original firmware)
10:04:41Slasheriat least over 3h to wav shouldn't be a problem
10:04:57Slasheribut there must be something weirdness in the code then
10:05:01pengothat's what's stated in the manual at least
10:05:12pengoi'm not sure i've ever tried recording that long
10:11:22***Saving seen data "./dancer.seen"
10:38:08 Join amiconn_ [0] (
10:39:33 Quit amiconn (Nick collision from services.)
10:39:33 Nick amiconn_ is now known as amiconn (
10:45:11 Join webguest41 [0] (
11:06:27SlasheriHmm, the wav splitting seems to work well :) Now listening the recorded files (splitted to 1 MB per file) with iriver and playback is continuous
11:06:57pengoi hope you make the default larger tho :)
11:07:04Slasherihehe, sure :D
11:07:04pengoand make it auto generate a playlist file
11:12:29Slasherinow it will name the files following: rec_0001_00.wav, rec_0001_01.wav, ... Maybe two digits is enough if we have something like 650MB - 2 GB file size (should be configurable)
11:13:24pengoshouldn't have an extra ending if there's only one file
11:13:29amiconnWhy double numbering?
11:14:07Slasheripengo: Hmm, so the numbering should go: rec_0001.wav, rec_0001_02.wav, rec_0001_03.wav or something like that?
11:14:15pengoSlasheri, yeah i reckon
11:14:24Slasheriok, that sounds better :)
11:14:31pengotho it's still confusing looking..
11:14:40Slasheritrue.. but files will sort right
11:15:14amiconnThe numbering should go rec_0001.wav, rec_0002.wav etc
11:15:27Slasheriamiconn: no, if we are splitting the current file?
11:15:40pengoamiconn might be right.
11:15:56Slasherimaybe that's better..
11:16:00amiconnThis is how the already existing recording screen handles it.
11:16:33pengoSlasheri, you could have a rec_0001.wav, rec_0002.wav, and rec_0001.m3u (or whatever playlist format)
11:16:39amiconnOn archos, you can set automatic time split, can do a manual split by a button press, and finally it auto-splits when approaching the 2GB barrier
11:17:20Slasheriamiconn: btw, will the recording screen in debug menu be the future recording menu on iriver or will it change? If not, i could implement menu system there etc. to set the recording parameters
11:17:51amiconnWhy diversify code???
11:18:13amiconnThere already is a recording screen. Imho we should use as much common code as possible
11:18:39Slasheriah, so then we should port that screen to work
11:21:37amiconnAway now
11:23:05pengois there any widget code in rockbox?
11:23:17pengolike buttons and menus and stuff
11:26:03HClunfortunately not really
11:32:37 Join linuxstb_ [0] (
11:34:13linuxstb_There is a menu "widget". And maybe one or two more - e.g. the progress bar used in the WPS screen.
11:44:06 Join Shapeshifter [0] (
11:47:15 Join Lear [0] (
11:50:11 Join Yokalosh [0] (
11:50:18 Part Yokalosh
12:01:56Shapeshifter may I ask a question about a wps?
12:03:28linuxstb_Go ahead.
12:03:35zeyes, but now you have used your single question daily allotment.
12:03:46Shapeshifter lol ok...
12:05:44Shapeshifter Since I upgraded to the last rockbox version, none of the images used in the wps are beeing shown. e.g. in snapwise:
12:05:44Shapeshifter#Snapwise v1.0, made by Aru (
12:05:44DBUGEnqueued KICK Shapeshifter
12:05:47Shapeshifter%ac%s%?it<%?in<%in - |>%it|%fn>
12:06:01Shapeshifter%ac%s%?It<%?In<%In - |>%It|%Fn>
12:06:05Shapeshifter %pv %bl (%bt)
12:06:09ShapeshifterI have to change the first line to:
12:06:13Shapeshifterand make a empty line after it, so evereything is aligned correctly. Before of the last version everything was fine without these manual changes
12:06:24Shapeshifter why is that
12:06:48Shapeshifter this is how it has to look like
12:07:30Shapeshifter and if I dont make the changes manually it just shows x0|snapwisebg.bmp|0|0| in the first line insteaqd of the image
12:08:57Slasherithe format has changed, take a look at the newer wps files found at misticriver
12:09:30linuxstb_It's documented here: custom wps&hl=en&client=firefox-a
12:09:51Shapeshifter allright thank you
12:10:10Shapeshifter yeah I know that page I didnt notice the change
12:11:25***Saving seen data "./dancer.seen"
12:13:15linuxstb_How does Rockbox deal with a partitioned hard disk? Does it just use the first partition?
12:13:58 Join Moos [0] (
12:23:10 Part Shapeshifter
12:29:41 Quit Lear (Remote closed the connection)
12:39:36 Join cYmen [0] (
13:07:58 Quit Sucka ("a bird in the bush is worth two in your house")
13:08:01 Join linuxstb__ [0] (
13:09:36 Join ze__ [0] (
13:10:04 Quit ze (Nick collision from services.)
13:10:09 Nick ze__ is now known as ze (
13:13:15 Quit linuxstb_ ("CGI:IRC (Ping timeout)")
13:23:08 Join Febs [0] (
13:24:44 Quit linuxstb (Read error: 110 (Connection timed out))
13:25:41 Join Sucka [0] (
13:39:25 Quit Maxime ()
13:41:14 Join Maxime [0] (
13:51:04 Quit pengo ("qubero - open binary")
13:52:25 Join Lear [0] (
13:54:36 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
13:56:11CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
13:56:11DBUGEnqueued KICK Script
13:56:11*Script Winamp 5.08 - The Police - The bedīs too big without you (playing) - 44kHz 128kbps stereo - 4m 24s long |>>>>>>>>>>> (85%)|
13:57:37LearStrange... __builtin_alloca isn't available when I build the simulator...
14:10:24 Quit edx ()
14:11:26***Saving seen data "./dancer.seen"
14:22:03LearSlasheri: lately Rockbox has been "forgetting" to buffer file data (i.e., I start playback, and when it is time for the first "full buffering", nothing happens, it seems, so playback stops)...
14:22:43LearHave you heard of that problem before? Could be related to me listening to oggs mainly now, or possibly some changes I've made in metadata.c... :)
14:23:15SlasheriLear: interesting, at least i haven't had that problem yet
14:23:35Slasherican you reproduce it?
14:23:38MaximeI have a strange problem
14:23:52MaximeOne time I boot the player, no sound..
14:23:56MaximeStop it, restart it
14:24:00Maximethe sound come back
14:24:13LearI've seen similar things before; when first "full buffer" is about, it just buffers some data (rest of the track?), but the next buffering after that works as expected.
14:24:19SlasheriMaxime: that must be bacause of some hardware differences between the units :/
14:24:37Slasheriand that issue will be hard to fix without having such a unit
14:24:39LearBut no, I can't easily reproduce it. It just happens sometimes...
14:25:11LearAt first I suspected too low anti-skip buffer, but it can even happen with a 30s buffer.
14:25:24LearAnd most times I've noticed it, the player has been on a desk...
14:25:27SlasheriLear: hmm, that might be when we skip to the next folder
14:25:34Slasherion that situation, the buffer always runs empty
14:26:22LearNope, folder skipping isn't involved (it's off anyway). Always in the middle of a playlist.
14:26:34SlasheriHmm, weird..
14:26:45LearAnd it runs out of buffer in the middle of a track.
14:27:11Slasherihave you tried a fresh daily build without any own patches to code?
14:27:38LearNo, not yet. I guess I should at least remove the metadata changes... :)
14:29:08LearBtw, regarding metadata.c, when should get_metadata return false?
14:30:07SlasheriHmm, when the file is badly corrupted / unrecognized
14:30:19Slasherireturning false will prevent starting of the playback
14:32:00LearWhat about IO error reading e.g. vorbis comments?
14:33:13Slasherii think it's safe to skip the file then
14:33:31LearI.e., return false...
14:36:18Slasheribtw, i have just enabled the "real" recording screen on iriver. But it's not quite usable yet, there is lot to do to support all of the functions
14:39:05 Quit paugh (Excess Flood)
14:40:22 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
14:41:11 Quit paugh (Client Quit)
14:41:40 Join AliasCoffee [0] (
14:43:34linuxstb__Slasheri: I don't think playback.c always reacts correctly when get_metadata() returns false.
14:43:56linuxstb__e.g. the crashes reported when playing FLAC files with id3v2 tags - get_metadata() returns false for such files.
14:44:13linuxstb__There's a couple of places where the return value is ignored.
14:45:30Slasherilinuxstb__: Hmm, that's true
14:46:20linuxstb__If it doesn't affect playback, is it possible to always call get_metadata() before you initialise the codec? This would allow get_metadata() to change the codec type after analysing the file.
14:46:25SlasheriBut if get_metadata sets taginfo_ready to false if the reading failed, that should be not an issue (and we could do that also)
14:46:59SlasheriNo, that would spin up the disk
14:47:25linuxstb__Don't you need to spin up the disk to load the codec?
14:47:35Slasherino, if it's buffered
14:47:39linuxstb__get_metadata() also spins up the disk.
14:47:59Slasheriyes, it always spins it up
14:48:40linuxstb__Is get_metadata() called when you start buffering a new file, or elsewhere?
14:48:57crwlcdcd /cygdric
14:49:21Slasheriit's called always before buffering that file. At the same moment the metadata (including codec type) is save to the trackinfo structure
14:49:45Slashericrwl :D
14:49:54crwlthank you cygwin or someone
14:49:58linuxstb__Sorry, I have to run now. back later.
14:50:46crwli managed to do a great windows xp trick - the window that was on top didn't have the keyboard focus
14:53:02 Join edx [0] (
14:53:29 Join webguest72 [0] (
15:12:44 Quit Lear ("Chatzilla [Firefox 1.4/undefined]")
15:23:11 Quit linuxstb__ (Read error: 104 (Connection reset by peer))
15:23:13 Join linuxstb_ [0] (
15:35:44Slasherihehe, now recording peak meters work :)
15:40:15 Quit t0mas (Read error: 104 (Connection reset by peer))
15:45:35 Join muesli- [0] (
15:52:15 Join t0mas [0] (n=Tomas@unaffiliated/t0mas)
16:00:14SlasheriHmm, in recording.c why is peak meter always set to playback mode?
16:00:27 Join muesli__ [0] (
16:02:56amiconnSlasheri: That's because the MAS has two registers for reading the quasi-peak value,
16:03:12amiconnone connected to the ADC and one connected to the DAC
16:03:24Slasheriamiconn: yes, but i tought it would be logical to have it in "recording mode" while recording :)
16:03:35amiconnThe DAC reading is always valid because recording monitor is always enabled
16:03:43Slasherianyway that doesn't matter, i will still implement that for software codecs also
16:03:55Slasherihmm, ok :)
16:04:05amiconn...but the ADC register only works when recording from line in or mic, not from s/pdif
16:05:14amiconnI think that shoould be changed for swcodec platforms, or perhaps there is no specific mode setting necessary
16:05:41amiconnPeak calculation is always done across a bunch of PCM samples with swcodec
16:05:55Slasherii changed it so on swcodec platforms it will be on "recording mode" while recording, so the peak_meter_peek calls the right function to fetch the peak data from
16:06:17amiconnWhy not have just one function?
16:06:57SlasheriHmm, that is also possible.. Then we have to put the function to a different file
16:07:12amiconnHmm, probably
16:07:17Slashericurrently it's on pcm_playback.c for playback (and firmware layer should not access to the apps/)..
16:07:27amiconn...or we need to unify pcm_playback and pcm_record
16:07:56amiconnI'm not sure whether this would be a good idea
16:08:24Slasherime either :) we still might have units that would not be capable for recording
16:08:26amiconnI thought a bit about the general architecture, concerning recording, playback, and buffering
16:09:04amiconnI think that the current separation between buffer/ track handling and pcm playback is the right way
16:09:36amiconn...and the firmware layer should allow pcm playback and recording at the same time.
16:09:41amiconn...if possible
16:09:51Slasherihmm, interesting
16:10:21amiconnIn the high-level code, allowing both playback and recording at the same time doesn't make sense
16:10:34amiconn...because then we would need to split the buffer
16:10:53Slasheriyes. But for example playing voice samples during recording might be a good idea
16:11:01amiconnThe low-level full-duplex operation opens the door for interesting sound effect plugins
16:11:14Slasheriah, true also :)
16:11:28***Saving seen data "./dancer.seen"
16:11:55amiconnIn short: PCM playback & recording should be allowed full duplex on the units that support it
16:12:24amiconnFile playback and recording should be exclusive (but voice UI still possible since low level supports full duplex)
16:12:36Slasheriyep. I don't think that would be a big problem with the current architecture
16:13:01amiconnThe voice file buffer is set aside anyway, shouldn't make a big difference between recording & playback
16:15:31amiconnlinuxstb_: Rockbox uses the first primary FAT32 partition it can find.
16:16:01amiconnLogical drives are not (yet) supported, but I intend to change that
16:16:39linuxstb_amiconn: Thanks. Just gathering info for the iPod.
16:17:12amiconnIt uses a very simple but effective approach - it simply tries to mount every non-empty partition entry as FAT32. If that fails, it tries the next etc.
16:17:25amiconnSo the partition type field doesn't matter at all
16:18:23amiconnRockbox for Ondio is almost the same, only two differences: (1) It handles 2 drives (internal flash + MMC). (2) It handles FAT16 in addition to FAT32
16:18:24Slasheribtw, does iriver have serial or parallel flash chip?
16:18:45amiconnDo you mean the firmware flash rom?
16:18:54Slasherii think it's parallel..
16:19:11amiconnIt's a standard flash eeprom. They're always parallel
16:20:10Slasheriyes, unfortunately. If it has been serial then it could be easy to build a mod to recover the flash contents (possible useful for developers)
16:21:01amiconnYou would have to unsolder it anyway, regardless of interface
16:21:25 Quit ashridah ("Leaving")
16:22:32 Quit Script ("Bambalam .gx.")
16:24:37 Quit muesli- (Read error: 110 (Connection timed out))
16:30:51 Join tucoz [0] (
16:31:56tucozGood thing that most of the wiki were taken care of by google & co
16:33:23tucozI read that some person at ubuntu had a recent attack from a .ru. Just that he didn't refer to the attackers as gentlemen, but scriptkiddies :)
16:35:08Slasheriamiconn: Hmm, why? Isn't the cpu tri-stated while it's powered off?
16:37:41 Join zenx [0] (
16:38:06zenxwhat's up with the website?
16:38:14amiconnSlasheri: How would you power the flash rom without also powering the CPU when it's soldered in?
16:38:29linuxstb_zenx: Read the news item at the top of the front page.
16:39:09Slasheriamiconn: true, that would be hard.. :) but never mind, there is probably no easy and cheap way for everyone to do that
16:39:40zenxoh, that's fucked... are there backups?
16:40:31linuxstb_zenx: Yes.
16:41:53tucozI simply can't comprehend what the point of doing that is.
16:42:04zenxany optical gurus here?
16:45:14zenxumm, don't know if this has been discussed before, but: has anybody considered connecting an ADAT lightpipe signal to the h120? the adat lightpipe uses the same toslink plastic cable as the s/p-dif, but it's a different protocol. it transfers 8 channels at once (in a serial order)
16:46:18SlasheriHmm, i don't think that would work.. Checking the cpu datasheet
16:47:36zenxhere's a brief description:
16:48:59Slasheri"Optical interface components and optical fibres are identical to the cost effective components as applied with the optical S/PDIF (TOSLINK) interface."
16:49:02Slasheriso it may work
16:49:45zenxand this is information about the data transfer:
16:50:00Slasheribut i really doubt that, the protocol is different
16:50:18crwlbut if the pulse speed and other features (what there are) differ enough, the S/PDIF chip wouldn't understand anything about the signal anyway, right?
16:51:01crwlsome kind of weird data transfer mechanism via the optical connectors would be funny, though
16:51:18zenxyeah, well this is what I don't have info about
16:51:31zenxwether the pulse speed differs and suc
16:52:25crwli have no idea if there's such thing as pulse speed, nor do i really know anything about toslink. but anyway :)
16:53:40zenxsince the rockbox site is down, can anybody send me some datasheets or specs of the s/pdif chip and the optical receivers of the h120 ?
16:53:51Slasherithe cpu is compliant with IEC958-3 Ed2 and Tech 3250-E Standards only
16:55:47crwlis it currently possible to monitor the toslink input signal with headphones?
16:56:02zenxthat's s/pdif and aes/ebu
16:56:23Slashericrwl: with few lines of code it should be :P
16:56:38crwlthe headphone out in h100 series should be of quite good quality, but my soundcard's one might not be
16:56:46crwlso it could be used as a kind of sound card.. :P
16:56:53crwland i have optical out on the sound card
16:56:57Slasheriwithout any resampling it should be easy to reroute the digital input to the dac
16:57:07zenx IEC958-3 Ed2 : this is the cpu, and IEC958-3 Ed2 is aes/ebu ?
16:57:09crwli could actually use this if it was possible :)
16:57:46SlasheriHmm :) maybe i will try that :P
16:58:08crwli'd just need a toslink -> 3,5 mm (optical miniplug, what's its name?) adapter
16:59:43Slasheriyes, i really would need one too. Currently i have used normal toslink cable and put it to the connector (some stuff over the cable to keep it on place)
17:00:05crwlyes, that's not really handy :)
17:00:12Slasherii have no idea where i could find those cables
17:00:19Slasheriyep :D
17:01:13zenx<Slasheri> the cpu is compliant with IEC958-3 Ed2 and Tech 3250-E Standards only
17:01:17zenxwhere is this?
17:01:28Slasherizenx: on the cpu datasheet
17:01:50zenx this?
17:02:03zenxok thx
17:02:07crwlit just feels stupid to order some component worth 1 EUR from the net
17:02:12Slasherithe cpu has builtin drivers for the optical interface
17:03:28 Quit tucoz ("Leaving")
17:16:45 Join matsl [0] (
17:23:45 Quit webguest41 ("CGI:IRC (EOF)")
17:32:17 Quit Febs (Read error: 110 (Connection timed out))
17:32:59 Join bagawk [0] (i=1000@unaffiliated/bagawk)
17:34:53zenxfound the Adat chip:,10,0,0,1,0
17:37:26 Quit muesli__ (Read error: 104 (Connection reset by peer))
17:44:49 Join Lear [0] (
17:48:04zenxwell, spdif has a data rate of 2.8224 Mbit/s and Adat has a data rate of 12.288 Mbps, so it's unlikely that it could be used :-)
17:52:38 Join markun [0] (
17:55:05 Join Dma-Sc [0] (
17:55:52 Quit zenx ("CGI:IRC (EOF)")
17:56:52Dma-Sci was wondering if i could get a complete listing of available wps tags somewhere? though i suppose it was on the bombed wiki :/">
17:57:59fuzzienot entirely up to date..
17:58:36bagawkI had no idea rockbox was hacked
17:58:46bagawkNo wonder my inbox seemed smaller...
17:58:50Dma-Scah yes didn't tried google cache ;)
17:59:19Dma-Scthought this page is a handmade cache of google cache of rockbox wiki ;D
17:59:26Dma-Scthought > though
17:59:46fuzzieyes, if you look at the directory it's in, you can see how much work rasher put into it
18:00:02fuzzieso as soon as the wiki backup is restored it'll be no problem to put most of the changes back :)
18:00:26Dma-Scyes indeed a cool move :)
18:07:15Dma-Scwps formating options are quite complete, and with images support it gives a blast compared to ajb-player wps i used before ;)
18:11:16 Join wacky [0] (
18:11:30***Saving seen data "./dancer.seen"
18:14:28 Join noC|andY`fRa [0] (
18:22:18wackyhey guys, something totally offtopic here..
18:23:01wackyI'm about to release a software instrument tuner, and I'd like to know what are the steps towards a free software release .. and I know you guys know about it
18:26:00 Join linuxstb [0] (
18:27:25linuxstbwacky: Step 1 - Choose a license based on how you want others to be able to use your code.
18:29:07wackyhey :) GPL is my friend
18:29:34Slasherilinuxstb: btw, i got the iriver recorder screen basic functions working :) peak meter etc.
18:30:11linuxstbSlasheri: Nice. Does that mean it's no longer in the debug menu?
18:30:34Slasheriit's still in debug menu but there is now the "real menu" also :)
18:30:41Slasherii think i will commit after the feature freeze
18:31:00linuxstbAre you using the existing (Archos) recording code?
18:31:14Slasheriit's very usable
18:32:09linuxstbApart from a recording codec system, what's left to do?
18:33:19Slasherihmm, i think there is not so much left to do to support the full pcm recording functions the recording screen offers
18:34:01Slasheriof course the codec support will be the hardest one (but still not very hard)
18:34:31 Join dropandho [0] (
18:34:38dropandhohey all
18:34:58dropandhoyou happen to notice that the wiki link on site is broken?
18:35:26Slasheriamiconn, linuxstb, do you think would it be better to use the full buffer space while recording and spinning the disk up every 2-3 minutes or keeping the disk continously running?
18:35:51Slasheridropandho: the wiki is deleted by crackers
18:35:54linuxstbWhy would you want to keep it running continuously?
18:36:03 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
18:36:09wackyspinning the disk makes noise in the recording
18:36:13Slasherilinuxstb: i was just thinking what would be a better option
18:36:29Slasheriwacky: hmm, yes.. when using the internal mic
18:36:35wackyI though we could even try to eliminate these high frequencies with some algorithmes..
18:36:40dropandhoslasheri- man...did we loose anything?
18:36:42AliasCoffeeif you spinup/down the disk, does that mean that you'll end up dropping samples as per the iriver firmware?
18:36:47linuxstbWhy is recording any different to playback?
18:37:06Slasheridropandho: there is a backups about a month old and much of the contents have been recovered using the google cache..
18:37:12wackyyes, and even if your microphone is too close
18:37:15Slasheribut there is still much work to do to put the wiki back online
18:37:42dropandhoslasheri- great...this is pretty sad
18:37:45SlasheriAliasCoffee: no
18:37:52Slasheriwe wont have that glitch bug
18:38:13wackywas there anyone working on remote support ?!
18:38:28dropandhoslasheri- dont u think someone should put a headline on the homepage calling for anyhelp with this....maybe they have it cached or saved
18:38:29Slasherii don't think anyone is working with it currently
18:38:31wackyI got my LCD broken, and bought an iAudio (since H3xx is discontinued!)
18:38:39wackybut I just *want* to run the Rockbox somewhere, anywhere!
18:38:57dropandhowacky- there was one guy working on a patch...xavier or never got committed tho
18:39:06Slasheridropandho: hmm, i don't maybe. But i think any help is greatly appreciated when the wiki is up again (restoring the pages etc.)
18:39:26wackyyeah I know.. that's the version I'm running.. but it's nothing complete and up to date with the latest features
18:40:19 Join Febs [0] (
18:40:45amiconnSlasheri: Spinning all the time is bad. It literally eats battery power
18:41:07amiconn2-3 minutes is definitely worth it
18:41:10Slasheriamiconn: good, i thought that also. So it's then good to spin it down
18:41:31amiconnI did some measurements with video playback on archos, which needs to rebuffer every ~25 seconds
18:41:47amiconnEven then it's still better to spin down in between
18:42:03Slasherihehe, nice
18:42:09 Join wacky_ [0] (
18:42:14 Quit wacky ("ahuh")
18:42:29dropandhoslasheri- i think it is important to let folks know this happened on the front page...and solicite else are people suppose to know?
18:42:40amiconnIt'll probably not worth to spin down on archos with upcoming wav playback, at least at full 44.1/16/stereo
18:42:58amiconnThis will need to rebuffer every ~10 seconds...
18:43:12amiconnSame for wav recording of course
18:43:51Slasherihmm, that's quite often
18:45:09amiconnwacky_: I'll probably look into making a remote-only build for iriver, i.e. using the remote lcd *instead* the main lcd
18:45:22amiconn...hoping that this will help H3x0 development
18:45:36dropandhoamiconn- nice..thanks for the remote work....ive been waitin for that!
18:45:49dropandhoamiconn- it is hard to make the 2 work at the same time?
18:46:32amiconndropandho: This won't be the "real thing", but a hack to ease H3x0 development. It allows to make a bootloader for H3x0 without having an LCD driver for the colour lcd
18:46:45wacky_oh yeah ?!
18:46:50amiconnIt'll probably be easier to work from there...
18:46:57wacky_is that something possible right now ?
18:47:13 Nick wacky_ is now known as wacky (
18:47:32dropandhogot it- thanks
18:51:49 Quit bagawk ("Leaving")
18:53:42 Join Dan [0] (
18:56:46wackyamiconn - is it something really involving ? or is the code disposed to that quickly ?
18:58:43dropandhoquestion tho...i know you guys are all about being why would folks jump to develop h3x0 if h1x0 isnt complete
19:01:41Slasherihehe, rockbox will be never complete :)
19:01:58Slasherithere is always something to develop
19:02:29dropandhohehe- i understand that
19:02:48dropandhobut there hasnt even been an iriver why jump to the next model?
19:03:03Slasheriyes, that's true
19:03:35dropandhoyou guys are in a feature make sure the other models are cool
19:03:42 Join DangerousDan [0] (
19:04:17amiconndropandho: It's not "jumping to the next model", but extending the user and developer base by another model
19:04:55amiconnThe H3x0 is similar to the H1x0 hardware-wise, so it should actually push forward iriver development
19:05:53dropandhook- sold!
19:06:07dropandhoi guess im greedy...and i wanna see my guy running nice
19:06:09dropandhothanks all!
19:15:36 Join muesli- [0] (
19:19:05 Quit dropandho ()
19:35:37wackyamiconn - I was just wondering if "compiling a version with native remote support" was something I could put some time into
19:35:44wackyor if it was something very involving
19:40:31bluebrother^can someone help me with a perl problem?
19:41:07 Quit Seed (Read error: 104 (Connection reset by peer))
19:41:11 Join Seed [0] (
19:41:22bluebrother^I want to install a module locally. How can I use this in a perl script?
19:42:01 Join webguest79 [0] (
19:42:42 Quit Dma-Sc (Read error: 110 (Connection timed out))
19:42:42webguest79bluebrother: does export PERLLIB=/home/user/lib:$PERLLIB work for you?
19:43:34webguest79and use the module like you normally would. Like use module in perl.
19:43:46webguest79i.e in the script
19:44:21CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
19:44:21*bluebrother^ tries
19:45:26linuxstbinteresting comments about the speed of Ogg:
19:45:42linuxstbHas anyone considered putting codec code in IRAM?
19:46:02webguest79isn't that what is done at the moment? at least parts of it
19:46:32amiconnlinuxstb: It's way better to use IRAM for data than for code on coldfire 5249
19:46:32linuxstbOnly data is in IRAM (AFAIK)
19:46:42webguest79oh, sorry
19:46:52amiconnSome code is in iram, but no codec code
19:47:06linuxstbamiconn: Why is is it better to use IRAM for data?
19:47:16Learamiconn: actually, a few Tremor functions are in iram...
19:47:35amiconnThere are two reasons. (1) The 5249 has an instruction cache, but no data cache
19:48:48amiconn(2) Uncached instructions are loaded from dram using burst mode.
19:49:04linuxstbSo basically, the speed-up of putting code in IRAM is minimal?
19:49:31amiconnUsing burst mode for data is only possible with movem, and only if the data is line aligned
19:50:25 Quit webguest79 ("CGI:IRC (EOF)")
19:51:20LearI'm wondering about a comment about a benchmark for Coldfire FFT code: "Without invalidation of all cache locations program requires even more cycles, than in the case when cache is disabled."
19:52:30 Quit Febs (Read error: 104 (Connection reset by peer))
19:55:27amiconnLear: Some core code is in iram too (at least memset, memcpy and lcd_update irrc)
19:56:11LearAnd the Tremor functions put there are the ones that are called _very_ often...
19:57:08amiconnYes, it should pay off to put functions in iram that are called often, and from various places
19:57:48 Join cybernetseraph [0] (n=cybernet@tor/session/x-2a1dc73876c1b5c3)
19:58:14amiconnThe coldfire icache is direct mapped, so depending on the alignment code will sometimes get flushed from the cache due to cache aliasing
19:58:14cybernetseraphHey, just thought i'd mention that the wiki on appears to be down.
19:58:51fuzzieif you look at the front page, it clearly states that the wiki was deleted
19:58:55fuzzieit is being worked on :)
19:59:15cybernetseraphfuzzie: oh, sorry. Must have missed that.
19:59:29cybernetseraphmust be one of those days.
19:59:50bluebrother^I'd say that information is a bit easy to over-read.
20:00:02cybernetseraphoh, that's it. i hadn't reloaded the site since i loaded it, and that wasn't there when i first surfed over.
20:00:04fuzzieif you need info from it, rasher has put the google cached versions at"> for now, until the backup is restored
20:00:18 Join Dma-Sc [0] (
20:00:24fuzzieor, the google cached versions of anything recently modified.
20:00:30cybernetseraphthanks. I was just checking up on how the iriver support was coming along
20:01:22Dma-Scfor wps, it would be cool to be able to display several values one after the other (with timer stuff) in a condition case
20:03:43cybernetseraphWell thanks for the info. I'm no good at audio/embedded development, so good luck guys, i'll keep checking back. Then again, do you have some donation interface? I get my student loan and i'd happily chuck some change your way.
20:05:23amiconncybernetseraph: There's a link on the front page...
20:06:11cybernetseraphgod, i really am dim witted today. Thanks for the lack of flaming stupid questions. i think i need more coffee. Goodbye all.
20:07:03 Part cybernetseraph (""shutdown -h now"")
20:07:10 Quit edx (Read error: 110 (Connection timed out))
20:09:40Dandid noone backup the wik?
20:10:02fuzziethe last backup is a month old, hopefully it'll be restored today sometime
20:10:35fuzziebut that page of rasher's has almost all the changes up to the point of deletion
20:10:39Danany ip logs?
20:10:41fuzzieso it'll just be a little work to get it back up to date
20:10:47fuzzieand, yes, it was an russian ip
20:11:15fuzziejust executed 'rm -rf /' through the exploit, no local exploits or anything, so logs/system/etc are intact.
20:11:31***Saving seen data "./dancer.seen"
20:11:32Danhave you checked for a proxy on it?
20:12:58amiconnIt was a dialup ip iirc
20:14:00fuzzie85.140.9.74, which resolves to something with 'pppoe' in it, which seems more likely to be broadband
20:14:21RickThat sucks
20:14:34Rickpppoe = adsl
20:14:58Danill scan it
20:15:41Rickdid we lose all of the irc logs too?
20:15:43Rickthat'd suck
20:16:07fuzziethey're fine
20:16:12fuzzieonly stuff owned by the apache user was lost, afaik
20:18:05Danport 21 is open on eit
20:18:13Danbut its not an ftpd i dont think
20:18:36fuzziei'd expect it's a dynamic ip anyway
20:18:52Rickyes, pppoe is usually dynamic too
20:18:58Danmine isnt
20:19:40Danhas anyone emailed abuse@ ?
20:20:40Rickthere have been a number of * visitors to #rockbox
20:21:11Danhow many?
20:21:42Rickrecent logs? three different people
20:22:29*Rick goes back to idling like he usually is.
20:23:03Danheres what id to
20:23:22Danscan the ips for vulns
20:23:39Rickwhat good would that do?
20:23:45Danif they can be rooted easily then he was running proxies off a botnet
20:24:44Danthen you could root them and sniff his net out if that was the case
20:24:52Danthen take it from there
20:25:23Danemail abuse@ first though
20:26:18 Quit muesli- (Read error: 104 (Connection reset by peer))
20:27:06 Join Musicmad [0] (
20:28:01Musicmadhi guys - does somebody have an idea what is going on here?
20:31:52 Join muesli- [0] (
20:31:52 Join webguest64 [0] (
20:32:31webguest64Musicmad, sounds almost like the oppisite of what I was experiencing a while ago. I got a high souding noise on boot sometimes.
20:32:47Musicmadwebguest72: ok - never had that.
20:32:51webguest64Which was fixed in the v5 bootloader
20:33:14webguest64where the uda is reset early in the boot process
20:33:29Musicmadok - I have tried with a recent build - still no sound in the morning :)
20:34:30webguest64annoying, so do you think it has to do with a build, or bootloader?
20:34:42webguest64Like, was this introduced lately?
20:35:11Musicmadno idea actually - I _think_ it has been like this for 3-4 weeks.
20:35:22webguest64But I believe there are minor hw differencies in the different h120/h140's
20:35:30Musicmadmine is a 120
20:35:32 Join solex [0] (
20:35:50webguest64As I and a couple of others had this noise on boot (sometimes), where others didn't
20:35:59Musicmadanyway - a reboot is pretty fast, but in the long run it would be nice to avoid it.
20:36:08webguest64Musicmad, :)
20:37:11webguest64And it's hard to find out what causes it, as it is occasionally. But, we were able to fix the issue I was having. It just takes some time
20:37:32webguest64I mean, find and fix
20:37:56Musicmadyeah - I don't know if any info gathered from the debug menu will help?
20:39:04webguest64If you are able to found out as much as possible when this happens, and maybe suggest something to the devs later on.
20:39:16webguest64you will be more likely to see a fix.
20:39:47MusicmadI know - trying to establish a pattern.
20:39:54webguest64Like, what do you think causes it. (good)
20:40:17webguest64If there is something that can trigger this.
20:40:30Musicmadyeah I know about good bug reports thanks :)
20:41:03webguest64I had my issue maybe twice a week. It was the usb-bootloader that narrowed it down, as the noise were present at that stage.
20:41:28webguest64Which meant that it couldn't be the firmware, as this wasn't loaded at that time.
20:42:16webguest64Musicmad, hope you'll find this out. gtg.
20:42:21 Part webguest64
20:47:01 Quit solex_ (Read error: 110 (Connection timed out))
20:50:40 Quit muesli- (Read error: 113 (No route to host))
20:55:16 Quit matsl (Read error: 104 (Connection reset by peer))
20:55:30 Join stripwax_ [0] (
20:56:42 Join matsl [0] (
21:06:15 Quit matsl (Read error: 104 (Connection reset by peer))
21:07:42 Join matsl [0] (
21:09:04 Quit wacky ("Leaving")
21:09:22 Join Shapeshifter [0] (
21:09:30 Join edx [0] (
21:09:56 Join Jefe [0] (
21:12:42Jefedoes anyone know where I can get a rfvMux filter?
21:14:26 Quit Jefe ()
21:16:51 Quit Musicmad ("Trillian (")
21:17:10Dma-Scdoesn't rockbox handle dir based playing like playlists, i mean with tracknumber and total tracks info for the current dir? i thought it did on my abj.
21:19:10 Part Shapeshifter
21:24:11 Join Zagor [0] (
21:28:44Dma-Scehh, yes it does... i was just using the wrong info (id3tag tracknumber instead of playlist tracknumber) X_X
21:32:03stripwax_anyone know any technical reasons why 'id3 info' (or really 'tag info') is not an option on the context-sensitive file menu in the browser?
21:32:10stripwax_would be v. useful!
21:32:19 Join muesli- [0] (
21:35:20 Quit muesli- (Client Quit)
21:52:32 Join billy1ear [0] (
21:52:55billy1earis something wrong with web page
21:53:14Zagoryes, see front page
21:54:33billy1earguess i should learn to read. anyone have the latest manual, or know somewhere else to download it?
21:56:03 Join preglow [0] (
21:57:19preglowwiki still dead? :/
21:58:13Zagoryes, we'll get the backup restored tomorrow
21:58:43preglowahh, you've got backups
21:58:44preglowthat's great
21:59:06preglowapache logs intact?
21:59:08Zagorunfortunately not completely up-to-date (since Murphy is always watching)
21:59:16Zagoryes, we know who did it
21:59:29Dandont be so sure
21:59:35preglowjust let me know if he's situated in oslo
21:59:50preglowi'll blow up his mailbox or something
21:59:58Zagorwell that's hard to say. naturally we only know the IP, which was a russian dialup.
22:00:00Danthat ip DID have a proxy on it
22:00:01amiconnhi preglow
22:00:05preglowamiconn: hi
22:00:19preglowZagor: you can pretty much forget anything happening at all, then
22:00:41 Join rasher [0] (
22:00:47Dani can see 2 options
22:01:03ZagorDan: correct. same same.
22:01:07Danrequest logs from the proxy owner, which i doubt will work
22:01:19amiconnpreglow: How costly would it be to derive the peak level from an mp2/mp3 frame?
22:01:19Danor root the proxy
22:01:31Danthatl get you an ip
22:01:39amiconnNo full decode, just enough to obtain an approximately correct level
22:01:40Danif not then email abuse@
22:02:31preglowamiconn: hmm, lemme think
22:03:06preglowyou'd haveto do a full huffman decode, which will cost some cpu and quite a lot of memory
22:03:26preglowyou're thinking archos here?
22:04:09amiconnMeanwhile I suspect the peakmeter i2c to cause the recording glitches, but reading less often is going to be imprecise
22:04:59amiconn...and since I'm thinking about scanning the frame structure anyway, I had the idea that it might be possible to obtain the peak level as well
22:05:02preglowyou've actually tried recording without the peak meter, or... ?
22:05:22amiconnI'm currently trying with a much lower peak rate
22:05:37amiconn*peek rate
22:05:51preglowto be quite honest, i don't think you've got a chance
22:05:56preglowespecially not with mp3
22:06:27preglowlayer 3, i mean
22:06:39amiconnPeeking once per refresh (20 Hz) instead of a busy-yield cycle (roughly 600 Hz)
22:06:45 Quit DangerousDan (Read error: 110 (Connection timed out))
22:08:02preglowas far as i know, there isn't much you can do apart from actually decoding the waveform to see where the peaks are
22:08:13Dma-Scallowing display timers in conditions cases for wps would allow nice play, pause, ff, etc... graphic animations ;)
22:08:34preglowi can control my enthusiasm...
22:10:14 Quit Lear ("Chatzilla [Firefox 1.4/undefined]")
22:10:31Dma-Sci mean sublines
22:10:55preglowamiconn: but this will most certainly add bloat, just the huffman decoding alone requires quite a few lookup tables
22:10:56Dma-Scfor use with timers ;)
22:11:34***Saving seen data "./dancer.seen"
22:15:42preglowamiconn: the mas only records in layer 3, yes?
22:16:51amiconnVersion 1 and version 2 layer 3. No version 2.5
22:17:12amiconnMax. frame bitrate is 192kbps
22:17:39preglowwell, you MIGHT be able to get some approximate peak info from the scale factors
22:19:05preglowbut even if it would work, it would probably be too inaccurate to be usable as a recording peak meter
22:19:24amiconnI wonder which part of decoding mp3 is the most computationally intensive
22:19:32amiconnHuffman decode? imdct?
22:19:52preglowa split between imdct and filter bank resynthesis, i would think
22:20:48preglowgod, i wish i could fix that ugly imdct hack of mine soon
22:21:26amiconnHmm. How much precision could one expect from using the scalefactors?
22:21:59amiconnIt might still be better than these "quasi peak" readings from the mas
22:22:53amiconnPerhaps I should perform some experiments to determine how this "quasi peak" behaves for various input signals
22:22:54preglowto be quite honest, i have no idea
22:22:58preglowi'm not even certain it would work
22:23:28preglowi don't know how well the scale factors correlate with the signal
22:23:50preglowthey're used before quantising to make sure all signals have a pretty uniform level
22:23:53preglowto make the most of the quantiser
22:24:26preglowand you have to combine those over all the frequency bands
22:24:40preglowand you don't really know if some of them might cancel out to make a lower peak than you would expect
22:26:01preglowif you're serious about finding out, hacking libmad should be easy
22:26:43preglowi might even do it myself tonight
22:26:55preglowsince i got vaguely interested, heh
22:36:08 Join Arishon [0] (
22:36:36 Part Arishon
22:47:37 Quit matsl (Remote closed the connection)
22:50:38 Join DangerousDan [0] (
22:53:24Dma-Scimplementing something like "?%id!=%Id<bla|blu>" in wps would also be great :)
22:57:41 Quit dpassen1 (Read error: 110 (Connection timed out))
22:59:40 Join linuxstb__ [0] (
23:02:45 Quit Dan ()
23:03:47 Quit billy1ear ()
23:09:18 Quit linuxstb ("CGI:IRC (Ping timeout)")
23:12:41 Join BoD[] [0] (
23:12:46BoD[]Hello worldz0rz
23:12:47 Quit rasher ("Ex-Chat")
23:14:09 Quit linuxstb_ (Read error: 110 (Connection timed out))
23:14:36BoD[]so... what's up ? :)
23:16:26BoD[]I was wondering if I should buy a comodore e-vic 20
23:16:45BoD[]maybe it may be "rockboxable" :)
23:27:13 Join webguest13 [0] (
23:32:14 Quit Dma-Sc ("")
23:35:41 Join coob [0] (
23:42:12 Join ]RowaN[ [0] (
23:42:27]RowaN[quick question.. rockbox for h10 .. is that something that might happen?
23:42:59]RowaN[iriver h10 i mean
23:43:06 Join linuxstb [0] (
23:43:08coobh10 is closer to the ipod :)
23:43:24fuzziemaybe, someday, but it's completely different from the existing iriver ports
23:43:34coobso if linuxstb gets his ipod port going, h10 would be possible
23:43:40linuxstbcoob: Do you know if the guy trying to port uclinux to the H10 got anywhere?
23:43:56coobstuck in bootloader alley i think
23:43:58preglowlinuxstb: you've got some ipod porting experiment going on?
23:44:20linuxstbpreglow: I'm going to start when my ipod arrives - which will hopefully be tomorrow.
23:44:39linuxstbUnfortunately, I seem to have ordered an ipod with an unsupported LCD.
23:44:58coobyeah no luck on that yet
23:45:09preglowlinuxstb: excellent!
23:45:15coobmaybe you'll be lucky and get an older 60gb...
23:45:32preglowit's just a matter of disassembling anyway, no? :P
23:46:15preglowlinuxstb: well, if you get anywhere, chances are good i'll join in on the fun when i get some cashflow again
23:46:25coobonce you manage to decrypt the fw :0
23:46:43preglowcoob: well, haven't you guys cracked that already?
23:47:03coobthe dmca doesn't allow me to reveal that information
23:47:18coobwell, it probably does, but i'm not getting shot for it.
23:47:29coobor th eu equivalent of the dmca
23:48:01coobpreglow: we have been known to read the flash by beeping it out...
23:48:06preglowi don't believe the dmca can give you any trouble on these grounds
23:48:22coobyes, but you're not a lawyer :)
23:48:51preglowi can pretend i'm one
23:48:56linuxstbDo we know the make/model of the LCDs in the ipod?
23:49:05cooblinuxstb: there are many.
23:49:08t0masyou here?
23:49:30linuxstbIs there a list anywhere? I'm having trouble finding detailed technical docs on the ipod.
23:49:48coobhaha of course you are, there aren't any! welcome to apple :)
23:50:13linuxstbI thought the ipodlinux project would have documented things though.
23:50:15coobno real lcd info there
23:50:38linuxstbEven the source code doesn't really refer to the names of many components.
23:50:43coobcoding is more fun than documenting.
23:50:59linuxstbBut it makes it hard for new devs to join in the project.
23:52:03coob <- think thats the datasheet for the mono lcd
23:52:32coobbear in mind nearly all RE work is done without opening iPods
23:52:44coobi think bern's gf told him he couldn't open it :)
23:52:51coob(his original 2g)
23:53:56cooband no hardware other than what came in the box
23:56:08linuxstbIs all the other important hardware on the 4G units understood now?
23:56:16linuxstbi.e. enough to get Rockbox working?
23:56:38HClas far as i understood they haven't even been able to get both cpu cores working
23:56:40BoD[]i would buy an ipod if rb would run on it
23:56:45coobHCl: yes we have
23:56:50HCloh, nice.
23:56:53coobjust don't really utilise them
23:56:54HClsince when?
23:57:01]RowaN[i told my m8 to buy a h1xx, and he bought a h10 hehe, stupid twat
23:57:04HCli thought only one cpu core worked..
23:57:13HCland only since recently you could set the clock
23:57:18coobjust had a lot of cache issues
23:57:29coobbeen able to set the clock for months
23:57:36HClyea, recently
23:57:37HCl :p
23:57:48coob * The LCD uses the HD66753 controller from Hitachi (now owned by Renesas).
23:58:02coobthats the mono one, i think the photo stuff was all re'd from the bootloader.

Previous day | Next day