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

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

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

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

#rockbox log for 2005-09-24

00:00:00Zagor_everything_ in the GUI would need locks and state storage. we're totally free of that today.
00:00:53TiMiD[FD]state storage ?
00:01:16TiMiD[FD]the threads running can keep their own state
00:02:10TiMiD[FD]I don't know a lot about the rockbox threading arch, but normally local variables are not shared between threads
00:03:08Zagorright, but far from everything is handled with local variables
00:03:31Zagortake the directory buffer (browser file list) as an example
00:03:32TiMiD[FD]so that's the problem :/
00:04:14Zagorin limited memory systems, you cannot use mega-size stacks
00:04:36 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
00:05:47TiMiD[FD]would that take so much memory to duplicate such informations ?
00:06:17Zagoryes. the dir buffer (for example) can be quite large.
00:07:40TiMiD[FD]so if there is no way to separate the two displays, what is the problem with xavier(s patch '
00:07:58TiMiD[FD](I looked a little and he just duplicated the drawings)
00:07:59Zagorif was quick and ugly.
00:08:18Zagorthe displays are not the same size, so the remote was cropped
00:08:28TiMiD[FD](I didn't test it)
00:09:16Zagorit was fun as a test, but not the right way forward
00:09:39TiMiD[FD]so the need here is to do some generic code to handle properly the display independantly of the size (like widgets) ?
00:10:23Zagoryes. our code already supports varying screen size, but the size is currently compiled-in.
00:10:56rasherZagor: twiki edit timeout of 60 minutes is a bit much
00:11:10Zagorrasher: agreed. i'll change it to 20.
00:11:48TiMiD[FD]ok so for example a bunch of functions that could allow handling of a file tree not only in the tree viewer (or something like that)
00:12:00TiMiD[FD]that would imply to rewrite a lot of code
00:12:27TiMiD[FD](and I'm only talking about the tree viewer, since it's the one I'm mainly interrested in)
00:12:30ZagorTiMiD[FD]: rather I see it as modifying the file tree code to simply draw the screen twice: once on main screen and once on the remote.
00:13:25Zagorthere are some issues, of course. like the list needs to scroll at different points and so on.
00:14:04***Saving seen data "./dancer.seen"
00:14:21Zagornote though that this is just my current thinking. we have not fully decided on the design of this yet.
00:18:04TiMiD[FD]I will look more closely to the tree source code
00:18:19Zagorthe scroll issue should be pretty simple, now that I think about it. focus on dircursor and adjust dirstart if it's too far away.
00:18:57TiMiD[FD]I'm a little bit lost in this code, I'm more used to oo programming and c sounds qite messy to me
00:19:32TiMiD[FD](no, I didn't said dirty :p)
00:21:05TiMiD[FD]maybe implementing the file tree parsing in independant widgets would make the code cleaner
00:21:28TiMiD[FD]and plugins would be able to use it
00:21:32 Quit ender` (Read error: 113 (No route to host))
00:22:44TiMiD[FD](quite useless though since it's the tree that is launching the plugins in most cases)
00:24:41TiMiD[FD]I was also wondering about one thing (just an idea) : would it be possible to handle the file tree like menus ?
00:24:43Zagortread carefully. "abstraction" is a foul word in these parts :-)
00:25:06Zagorhow do you mean like menus?
00:25:24TiMiD[FD]don't worry, I noticed :p
00:26:16TiMiD[FD]I mean a file tree is in presentation at least and in function (to the user) very similar to the file tree
00:26:40TiMiD[FD]directory is like sub-menu
00:26:59TiMiD[FD]file is like an option you enter in
00:27:34TiMiD[FD](well I didn't looked at the menus code too close too, but the code could be regrouped)
00:28:19Zagoryes, but there are also a lot of differences
00:28:29TiMiD[FD]so menus would become an abstract layer for file tree (oops I said the forbidden word :p )
00:28:35Zagorthe menus are hard-coded, the file tree is dynamically read from disk
00:29:00Zagorthe menu is translated, the file tree is not
00:29:21Zagorthe file list is sorted, the menus are not
00:29:47Zagorthe file list uses icons, the menu does not
00:29:50TiMiD[FD]I read somewhere on the wiki that dynamic menus were wanted
00:30:06Zagorwanted: yes. close: no.
00:30:12TiMiD[FD]ok :p
00:30:46Zagorthere are people who want the whole world configurable. I want to stay a bit closer to the ground. :-)
00:30:49TiMiD[FD](wth I certainly don't intend to recode the entire rockbox interface :p)
00:32:07ZagorI hope you don't feel I am just putting you down. It's a good discussion.
00:32:20TiMiD[FD]no it's not that
00:32:35TiMiD[FD](I'm not fool enough to recode everything)
00:33:06 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
00:33:07TiMiD[FD]i'm too lazy
00:33:39rasherHaving the menu structure seperated of the code would possibly allow to easier adapt the ui for say icon-driven interface (I'm sure the h300 people will be screaming for this) though
00:33:45Zagorlazy is sometimes good. it means only the really desired things get written and the fancy stuff is left out. :-)
00:33:47TiMiD[FD]and well maybe this story will be forgotten after eight or nine beers !
00:34:20 Quit Dan ()
00:34:38TiMiD[FD]but this time I don't know why I'm motivated
00:34:51TiMiD[FD]it doesn't happen a lot in a year
00:36:49TiMiD[FD]my first idea was to make multi thread like described in the mailing list (to allow me for example to continue reading my texts while changing the tracks, like on the iriver)
00:37:14rasherDoesn't the iriver sync the two displays 100%?
00:37:19TiMiD[FD](on iriver fw, it's more like an alt-tab though)
00:37:42TiMiD[FD]when you are viewing a text, you can press for example record to return to the main screen
00:38:03TiMiD[FD]and another record press will bring you back to the text
00:38:26TiMiD[FD]it's quite basic, but when I'm reading my notes it's useful
00:38:50rasherexcept that the text viewer is horrible
00:39:05rasherThe rockbox viewer auto-bookmarks by the way
00:39:16rasherso stopping it is not much hassle
00:39:40TiMiD[FD]it's some more keypresses ...
00:40:14rashercan't be many
00:40:33TiMiD[FD]but there are soe plugins that don't bookmark, for example games XD
00:40:45rasherstop, play, [whatever you want to do], select, play
00:40:56rasherof course
00:41:07rasherI'm just being argumentative
00:41:32Zagorrasher: about the displays, yes both displays show the same thing. but obviously not the exact same thing.
00:41:54*rasher boots the iriver firmware
00:42:14TiMiD[FD](btw, I can't use the rb text viewer cause it doesn't display japanese :( )
00:42:48*rasher eagerly awaits the unicode patch
00:43:06TiMiD[FD]no need to say i'm impatiently waiting for unicode support :)
00:43:28Zagorrasher: they're taking an easy way out by having 6 rows of text on the main lcd and 3 lines on the remote
00:43:52Zagorso they just page-flip twice as often on the remote
00:43:54TiMiD[FD]so the remote also displays text on iriver fw
00:44:01TiMiD[FD]never noticed this :)
00:44:05rasherZagor: the menu is 6/4
00:44:13Zagorrasher: ah, ok
00:44:28ZagorTiMiD[FD]: this was the file browser. don't think I've tried the text viewer :-)
00:44:38*rasher gets out of the iriver fw as quick as possible
00:44:55*TiMiD[FD] has some pbl with english :/
00:45:28Zagorrasher: the menu is visibly very different code. it scrolls while the file browser page-flips
00:45:44ZagorI'd like rockbox to work like the iriver menu does.
00:45:49rasherZagor: ah, haven't noticed
00:46:03TiMiD[FD]uh ?
00:46:13TiMiD[FD]wath's the problem with scrolling ?
00:46:20TiMiD[FD]display is too slow ?
00:47:00ZagorTiMiD[FD]: Nothing, I want to scroll. I just mentioned the iriver file browser doesn't.
00:48:25rasherHrm, am I missing something, or doesn't Rockbox/iriver do FM scanning?
00:48:55TiMiD[FD]it can scan radio stations
00:49:09TiMiD[FD]I don't remember the key though
00:49:27TiMiD[FD]maybe short + long right
00:49:46rasherRockbox on iriver?
00:50:11rasherone of my presets just appeared in the main menu?
00:50:30TiMiD[FD]I was speaking of original fw :/
00:50:41rasheralso, entering/leaving the radio screen changes the volume
00:51:10 Join bg [0] (
00:51:24bgnoticed 2.5 is out, but no iriver release yet?
00:51:35rasherbg: No.
00:51:42rasher2.5 was for Archos only.
00:51:44Zagorrasher: it's in the debug menu for good reasons :-)
00:51:52rasherZagor: It isn't, though
00:52:00rasherIt's right there in the main menu!
00:52:02bgrasher: when should we be expecting an iriver release?
00:52:28Zagormy eyes are going downhilll...
00:52:29TiMiD[FD]bg: no official release until irivier becomes fully supported I think
00:52:31rasherbg: Noone knows - see for a rough overview of what needs to be done
00:52:57rasherZagor: does rockbox do FM scanning?
00:53:06Zagorbg: when we can release it without getting tons of complaints about it :-)
00:53:18bgalways makes life easier
00:53:28Zagorrasher: dunno actually. never used radio myself (my archos doesn't have it)
00:53:35rasherFair enough
00:53:52rasherJust wondering if it's not there, or just not enabled on iriver
00:53:55TiMiD[FD]you can download a daily build from sept 22, it should be the same as official 2.5
00:53:56rasherguess I should rtfm
00:54:02bgheres another question... noticed my iriver firmware cant tune in AM, is that a hardware limitation or a software limitation?
00:54:07Zagorrtfc rather :)
00:54:08rasherbg: hardware
00:54:10 Quit paugh ("bbiab")
00:54:14rasherZagor: c being?
00:54:29rasherAh, but the manual knows about the fm screen
00:55:22rasher"Change frequency in 0.1 MHz steps. For automatic station seek, hold LEFT/RIGHT for a little longer."
00:55:23bgill be fine waiting for irivers official release... the original firmware doesnt bother me too much, just a few nagging problems that i can live with for the time being
00:55:31rasherAh, so it does
00:55:36rasherguess I wasn't holding it long enough
00:56:13bganyway, time to go do the drugs. you guys have a good night.
00:56:16 Quit bg ("Chatzilla [Firefox 1.0.6/20050921]")
00:57:49TiMiD[FD]time to sleep for me
00:57:56TiMiD[FD]good night guys !
00:57:58Zagorgood night
00:58:59rasherHm, archos users should chip in on
00:59:42 Join bagawk [0] (n=lee@unaffiliated/bagawk)
01:04:32linuxstbrasher: You're right about entering/leaving the FM screen changing the volume. But I think it only happens when your sound settings (bass, treble etc) are not zero.
01:05:01fuzzielinuxstb: were you the person with the modified faad2? if so, could you put it up somewhere?
01:05:04linuxstbSomeone mentioned it a couple of weeks ago - amiconn had a quick look at the problem, but didn't fix it.
01:05:08rasherlinuxstb: yes indeed
01:05:32rasherrasher: I was going to tell you you were wrong, but found that I had in fact changed the bass setting
01:05:41rasherUsually use flat settings
01:07:15linuxstbYou will need to uncompress it in the apps/codecs directory, and then modify the apps/codecs/Makefile and tools/configure scripts to add libfaad to the build
01:07:32fuzziethanks :)
01:07:56linuxstbAll I've done is to get it to compile - please double-check my changes. It is the CVS version of libfaad from (I think) 30 Jan 2005.
01:08:17linuxstb(I'm looking for my unmodified copy now...)
01:11:44linuxstbfuzzie: This is the original file I used:
01:12:13linuxstbIt should be the final GPL'd version of faad2.
01:12:22linuxstb(before they broke the license)
01:12:46fuzzieheh, thanks a lot
01:13:41linuxstbAre you planning to take it further?
01:14:31fuzziei am, if i can .. i have no experience with rockbox development, but i've built the simulator, poked at the source a bit, and am intending to buy an H120 for it
01:14:56Zagorrandom thought: On color displays, we probably will want to ignore the first button pressed when the backlight is off, since the user has no way of knowing where he is and which button is "safe".
01:20:10 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
01:26:44linuxstbSpeaking of color displays, has anyone given any thought to the lcd_framebuffer structure needed for the H300, or are we waiting until we know more about the LCD?
01:28:20ZagorI have been thinking 1 byte/pixel, for simplicity, performance and memory conservation. However some displays will want more depth for best image quality (for jpgs and video etc)
01:29:41Zagorperhaps a model with different video modes: 1byte/px for normal screens and 4byte/px for image viewing
01:30:09Zagorotoh we'll probably want album covers in the wps....
01:30:36ZagorI'm just worried scrolling will be terribly slow with 4 bytes per pixel
01:30:48Zagor(or require a cpu boost!)
01:30:48 Join Febs [0] (
01:31:40 Join Paul_The_Nerd [0] (
01:36:47linuxstbI've got no strong views on it - I just asked because I'm looking to see what needs to be done to get the H300 X11 sim to compile.
01:36:57 Join webguest00 [0] (
01:37:09bagawkZagor, I think you are underestimating the speed :)
01:37:29webguest00evening - has the patch to fix the auto power off bug been applied yet?
01:37:40bagawkZagor, I have 16mhz 68k mac that can do scrolling in color well
01:38:04Zagorbagawk: without help? (i.e. hardware)
01:38:19bagawkNo idea then
01:38:28bagawkBut I still think it could od it speedy
01:39:08rasherI guess the middle of the night is not the best time to get a list of FM stations available here
01:39:27rasherI can get the same 3 stations at 9 different frequencies, roughly
01:39:42Zagorbagawk: there's no doubt it would be possible, but we also want to keep the idle frequence down as far as possible. so performance is still an issue.
01:40:19rasherSpeaking of idle frequency - why are we not running the radio screen at 11MHz?
01:40:51Zagorno idea
01:41:02Zagoris it running at full tilt?
01:41:06rasherI have that in my builds, with absolutely no ill effects
01:41:40rasherI don't know how much would be saved. IIRC, 48->11mhz is not much difference
01:42:08ZagorI don't remember the numbers either.
01:42:15rasherOh, hm.. 58->39mA
01:42:20rasherthat's.. somethiing
01:45:40 Part webguest00
01:47:57linuxstbI would like to add an option to Sudoku to allow the user to keep the backlight on permanently. Is there an easy way to implement that?
01:48:56Paul_The_NerdHow much power drain does the backlight add anyway?
01:48:57rasherdon't know about the option, but keeping the backlight on permanently is pretty easy
01:49:56rasherlinuxstb: apps/plugins/grayscale.c does it
01:50:02Paul_The_NerdI was just thinking, with most of the games I keep wishing the screen wouldn't fade while I think about a next move.
01:51:14linuxstbrasher: Thanks.
01:53:00*rasher looks at ipodport
01:53:25rasherwhy, that's almost a small devicechart you have there
01:53:31rasherDeviceChart, that is
01:54:07 Join fogcat [0] (
01:54:32fogcatAnyone here know about how the wiki works?
01:54:35Moosgood night @ all
01:54:52fogcatI'm trying to reset my Wiki password and get " Administration error"
01:54:52 Quit Moos ("Glory to Rockbox")
01:55:10Zagorfogcat: I'll take a look
01:55:23fogcatZagor: ta
01:55:48 Quit paugh ("Leaving")
01:57:25rasherHm, maybe ipodlinux' wiki should be added to TWiki.InterWikis
01:57:36rasherI'm guessing linuxstb might want to link to it a lot
01:57:42Zagorrasher: good idea
01:57:51rasher(I can
01:57:54rasher't edit the page)
01:59:05Zagori'll fix it
01:59:27rasherAdding the FAQ link to the top bar seems like a good idea
02:02:08Zagoryeah I played around a bit with the looks and links
02:03:41rasherHm, maybe the tables should follow the style used on the non-wiki pages? Maybe with a bit more contrast, so joined cells are more obvious
02:04:07Zagorfogcat: looks like I need to update the ResetPassword page for the new twiki version. hang on.
02:04:13rasherTo me the tables stick out as being somewhat out of place
02:04:22rasherAnyway, I'm just being nitpicky
02:04:50Zagoryes, the tables are also very ugly :-)
02:05:57rasherI understand amiconn's complaint the other day that you couldn't tell that cells were joined with the look just after you upgraded
02:06:06rasher(no vertical devider)
02:08:58rasherI like the edits to the top/bottom bars
02:09:28rasherIt's very close to the cvs activity/daily builds tables
02:09:52Zagoryeah, that was the idea
02:11:01rasherExcellent, the interwiki thing works like a charm
02:11:32fogcatZagor: OK - thanks - I'll do it tomorrow now.
02:11:52 Part fogcat
02:11:58Zagoraww, just got it working :-)
02:13:07*rasher nitpicks the ipodport page
02:13:53rasherDoes the new twiki merge edits in some strange ways?
02:14:07***Saving seen data "./dancer.seen"
02:14:16rasherI just edited, unlocked and edited IpodPort and it's still at r1.2
02:15:26ZagorI don't know
02:16:21rashermaybe I'm just going insane
02:17:06rasherNo, still at 1.2 after another edit.
02:17:12rasherThis can't be a coincidence
02:18:31rasherI wonder if itunes for windows will encode alac
02:18:35 Quit dpassen1 ()
02:18:41Zagorare your edits visible in the diff?
02:19:09rasherseems like it's just merging consecutive edits by the same person (and maybe within some time window)
02:19:46fuzziesounds nice, i hate spamming wiki histories with a bunch of consecutive edits
02:20:00rasherYeah, it's a good idea
02:20:17rasherJust had me confused for a while because I wasn't expecting it
02:20:39ZagorI'm not so sure. The versioning is done by rcs externally, and you can't make rcs shoehorn new changes in an old rev.
02:20:55 Join webguest33 [0] (
02:23:19 Quit webguest33 (Client Quit)
02:27:11 Part the_winch
02:28:03Zagori'm off to bed. night all.
02:55:07 Quit bagawk ("Leaving")
02:56:48 Join Sucka [0] (
03:08:34rasherlinuxstb: codec failure with alac files ripped using itune
03:08:37rasherlinuxstb: codec failure with alac files ripped using itunes
03:10:26rasherItunes 5.01/Win32 specifically, if that matters
03:11:14Paul_The_NerdHow do ALAC file sizes compare to .WV of FLAC?
03:11:38rasherSomewhere inbetween FLAC at best setting and FLAC at normal setting I think
03:11:56rasherisn't sizes of lossless codecs pretty much the same?
03:12:05Paul_The_NerdYeah, pretty darn close
03:12:26Paul_The_NerdBut when the files are 15-20mb anyway, "close" can still be 1mb.
03:12:42rasherBut that's still not much
03:12:55Paul_The_NerdI was just curious
03:13:00Paul_The_NerdI'd read somewhere it decodes faster.
03:13:45Paul_The_NerdAnd that worried me.
03:13:50rasherI don't think so
03:14:00rasherIt's slightly more complicated than FLAC, it seems
03:14:16rasherIt all comes down to optimizations
03:14:29rasherThe algorithm isn't faster
03:15:15rasher(according to the author of the ALAC decoder and a FLAC developer)
03:18:55rasherThe commandline decoder fails as well
03:21:52Paul_The_NerdSo, it's an iTunes issues?
03:21:52rasher$ alac -f o.wav 01\ -\ Beneath\ Another\ Sky.m4a
03:21:53rasher(trak) unknown chunk id: edts
03:21:53rasher(moov) unknown chunk id: elst
03:21:53DBUGEnqueued KICK rasher
03:21:53rasher(top) unknown chunk id:
03:21:53rasherfailed to load the QuickTime movie headers
03:22:27rasherWhy do you think that?
03:23:00fuzzierasher: get the file to me somehow?
03:23:04Paul_The_NerdWell, you said you encoded it with itunes, but two different decoders fail?
03:23:13rasherPaul_The_Nerd: it's the same decoder
03:28:05 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
03:49:16 Join ashridah [0] (
04:05:11 Join QT [0] (i=as@madwifi/users/area51)
04:13:45 Quit TiMiD[FD] (Read error: 104 (Connection reset by peer))
04:14:11***Saving seen data "./dancer.seen"
04:16:02 Quit QT_ (Read error: 110 (Connection timed out))
04:16:15 Join TiMiD[FD] [0] (n=TiMiD[
04:30:31 Quit Sucka ("a bird in the bush is worth two in your house")
04:58:35 Quit cYmen ("zZz")
05:00:13 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
05:43:36fuzziedon't suppose anyone except rasher is around?
06:05:36fuzzieanyway, I got libfaad2 integrated into my tree, fixed it a bit more and wrote the skeleton of a codec
06:06:17fuzziei sort of need more metadata than the ALAC parser provides, it seems like it'd be a good idea to just integrate the faad2 mpeg4-file-parsing code
06:06:31fuzziegiven it doesn't have FIXME all over the place. but i guess i'll poke people about this tomorrow.
06:14:14***Saving seen data "./dancer.seen"
06:54:29 Join amiconn_ [0] (
06:54:42 Quit amiconn (Nick collision from services.)
06:54:43 Nick amiconn_ is now known as amiconn (
07:32:48 Join Maxime` [0] (
08:14:18***Saving seen data "./dancer.seen"
08:39:05 Join hshah [0] (
08:39:35 Quit hshah (Client Quit)
09:02:01 Join Vladoman_ [0] (
09:02:23 Quit Vladoman (Read error: 110 (Connection timed out))
09:04:03rasherNice, more missing attachments coming in
09:15:46amiconnMorning rasher
09:18:34 Quit paugh ("bbl")
09:21:50 Join rasher_ [0] (
09:23:20 Quit rasher (Nick collision from services.)
09:23:25 Nick rasher_ is now known as rasher (
09:24:19rasherI had all these questions and they never went anywhere.
09:24:28rasherWorked out most anyway.
09:24:51rasherHowever, 32 fm presets is not enough for everyone, it seems.. is this an important limit?
09:25:01rasheror just chosen because "it should be enouogh"
09:25:14rasheroh and, morning amiconn
09:25:33amiconnI think it was chosen because it should be enough *almost* everywhere,
09:25:50rasherNot enough for Paris apparently
09:25:58amiconnand it's a bit more than the archos fmr/ Ondio FM firmware allows (30)
09:26:18rasherMeanwhile, I struggle to find 5 channels with decent reception
09:26:52amiconnI don't know about how the iriver firmware handles it. Iriver is so unintuitive I didn't even find out myself how to enter radio mode...
09:27:15rasherI did manage to go to radio mode, but I never really figured out presets
09:27:20rasher(hold down play)
09:27:46amiconnYes, radio reception on H1x0 is significantly worse than on (new) Ondio FM, although they use the very same tuner chip
09:28:08rasherWith the same headphones?
09:28:20rasherNow isn't that curious
09:28:41rasherOh well, I don't listen to much radio.
09:30:12ashridahi don't think the iriver's calibrated properly myself.
09:30:22ashridahis it possible to get finer tuning than .05?
09:30:31ashridahor is that a limitation of the chip?
09:31:55amiconnRockbox tunes in 0.1 steps
09:32:39amiconnI dunno about other countries, but it seems that no terrestrial stations use xxx.x5 frequencies
09:33:01amiconnCable is a different thing though, but our devices don't have an antenna socket...
09:33:10ashridahiriver's firmware seems to think so
09:33:16rasherwell reception can be odd though
09:33:21ashridahit just seems like the S/N ratio is way too high for it to be tuned properly
09:33:36rashersometimes it's better, off the right frequency
09:34:05ashridahway too low
09:34:05ashridahsorry, not high
09:35:36amiconnWell, it's a pll synthesizer tuner, so tuning should always be exact (within the quartz precision)
09:42:22ashridahthen something must be off
09:42:28ashridahor the antenna sucks balls
09:42:35ashridahi can't get any channel without static
09:42:55amiconnDid you compare with iriver firmware?
09:44:15amiconnI just checked the datasheet, the minimum tuning step size depends on the quartz used
09:44:42ashridahamiconn: both have the same problem
09:44:48rasherZagor: the EZWPS page has something broken with attachments
09:45:58amiconnWith a 13 MHz xtal, minimum tuning step size is (exactly) 0.05 MHz (50 kHz)
09:46:06amiconnThat's what the Ondio FM uses
09:47:02amiconnThe H1x0 uses a 32.768 kHz xtal, which allows a minimum tuning step size of 32.768 kHz (0.032768 MHz)
09:48:31amiconnThat means, the iriver can't be tuned exactly to most xxx.x0 or xxx.x5 positions; it's almost always off a bit
09:51:30ashridahso there's no way to get better
09:51:50ashridahor could one in theory replace the crystal
09:52:38amiconn...and our routine always rounds down, so it might be more off than necessary...
09:52:47amiconnSomething to fix...
09:53:20rasherWhy would anyone use that instead of what the ondio uses?
09:53:30ashridahrasher: probably cheaper?
09:53:58amiconn32.768 kHz xtals are quite cheap, as that's what praktically all quartz watches use
09:54:25amiconn...and the oscillator needs slightly less power because of the lower frequency
09:56:42 Join DangerousDan [0] (
09:57:21CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
09:57:21*rasher looks at Tang's wpsgallery backup
09:57:34rasherit appears to be exactly what we found in MSN search's cache
09:57:51rasherof course, his copy has the images, but those are just screenshots :-\
10:04:03 Join linuxstb_ [0] (
10:14:20***Saving seen data "./dancer.seen"
10:16:59 Join Pieter_ [0] (
10:17:14Pieter_there's no more USB power mode in rockbox 2.5?
10:18:21rasherthere should be
10:19:15Pieter_yes i kno
10:19:18Pieter_i wrote it
10:19:25Pieter_it just doesn't work
10:19:43 Quit linuxstb (Read error: 110 (Connection timed out))
10:20:07rasherWell, I don't know what happened there
10:20:17Pieter_no.. i'll check the source :P
10:20:34rasherYou do that
10:20:43amiconnPieter_: FMR or Ondio?
10:20:45rasherWow, another student.utwente
10:20:51Pieter_amiconn: ondio
10:21:02amiconnHmm, strange. Maybe I should try 2.5
10:21:41amiconnSo you were the one who started USB power mode...
10:21:59amiconnI've improved upon that (at least I think it's an improvement)
10:22:19rasherlinuxstb_: are you around?
10:23:01Pieter_yes, i was.. it should be an improvement, but i can't get it to work...
10:23:07Pieter_seems to be there still
10:24:27Pieter_perhaps the line if(button_status() == USBPOWER_BUTTON) should use & instead of ==, but that shouldn't be the problem :P
10:24:33amiconnWorks for me in 2.5
10:24:37amiconn(Ondio FM)
10:25:54Pieter_i have the same thing.. let's try again
10:26:21Pieter_it says 'please remove inserted MMC'
10:29:18Pieter_i'll try installing manually instead of with the installer, it seems to have done something wrong with the plugins as well
10:31:04amiconnI used the zip file...
10:31:22amiconnI suspect there's something wrong with the 2.5 windows installer
10:31:31amiconnMany people had problems using that
10:35:44Pieter_well all the plugins would say incompatible version
10:36:57Pieter_but i'm extracting the zip file
10:37:00Pieter_takes a while
10:38:27Pieter_eh, this one's even better
10:38:30Pieter_doesn't boot
10:38:39Pieter_hangs at the rockbox logo
10:42:01Pieter_(oh wait, that was the wrong version.. i seem to be not completely awake yet :P)
10:46:37Pieter_ok, must have been the installer
10:46:45Pieter_because just manually installing it works fine
10:50:20amiconnThe installer is indeed wrong... It installs a build from 2005-06-07, whatever that is
10:51:39rasherThat's crazy wrong
10:51:59rasher2.6.1 time?
10:53:49rasherAnyone with strong opinions on iriver radio button-assignment?
10:54:14rasherI'm in agreement with this post:
10:54:31amiconnI'm not
10:54:43rasherwhat's wrong with it?
10:54:50amiconnYou can't call the main menu from the radio screen, there
10:55:14amiconns/there/it would allow to produce a loop/
10:55:32amiconnmain menu->radio->main menu->radio->...
10:55:36rasherI think he just means "leave with radio playing"
10:55:43rasherjust like is possible now
10:55:44amiconnFurthermore, it isn't necessarily logical
10:56:03amiconnAB is "menu", not necessarily "main menu"
10:56:41rasheryeah, I'm with you on that
10:56:52rashermaybe select should leave
10:56:54rasherlike it does the wps
10:59:18rasherSELECT - Leave with radio on
10:59:18rasherSTOP - Leave with radio off
10:59:18rasherA-B - Presets
10:59:18DBUGEnqueued KICK rasher
10:59:18rasherLong SELECT - Context menu
10:59:34rasherand I guess the mute thing on play is sortof logical if you're looking for pause
10:59:59amiconnI don't see the need for mute at all
11:00:11amiconnJust stop the radio, re-enter to continue
11:00:18amiconnSame amount of keypresses
11:00:53rasherTrue, can't hurt though - it's a decent analogue to the wps
11:01:14rasherUnless we end up lacking keys, in which case it's totally expendable
11:01:22amiconnPlus, it may rise other issues when mute is added
11:01:42amiconnLike, what happens if a radio recording is running and you mute?
11:02:02rasherYou get eaten by a grue.
11:02:02amiconnHmm, that seems to be a situation where mute might actually be useful...
11:02:21rasheranyway, I was intentionally leaving that out from my list
11:02:36Pieter_2.5 starts even faster? :)
11:03:06amiconnrasher: In fact your list makes sense to me, plus PLAY = mute
11:03:29amiconnMute should only silence the output of course, leaving the radio (and an eventual recording) running
11:03:49 Join ender` [0] (i=ychat@
11:03:58rasherso what happens when you leave and mute is on?
11:04:06amiconnIt should unmute
11:04:16rasheralso if you leave with radio on?
11:04:37amiconnThou shalt not do that! ;)
11:05:18rasherWell, it probably should just stay muted and from there on it's now the user's problem
11:05:21amiconnI think leaving with mute & radio on should act like leaving with the radio off
11:05:38rashergood idea
11:05:40amiconnWe have no mute for file playback, and that's for a reason
11:05:56amiconnIt makes no sense, pause is always better
11:06:41amiconnBetter tuning for H1x0 committed
11:07:55amiconnBtw, the tunning error was smaller than I though, because the frequency is divided by 4. Anyway, now the maximum roundoff error is half of what it was before
11:08:13amiconn(but can be positive or negative, used to be always negative before)
11:08:28rasherAh, so shouldn't be much noticable?
11:09:30 Join hshah [0] (
11:10:05rasherI'm not quite sure if the context menu is the one that offers the "ADD" entry, or the presets
11:10:19 Part hshah ("Leaving")
11:10:46rasherNah, it is the one with add in it
11:13:31amiconnHmm, your list is logical, but not necessarily convenient.
11:14:10rasherI'm all ears
11:14:18amiconnPresets on Select might be better
11:14:27rasherfor easier access?
11:14:57rasherI had that Idea, but decided against it to let logic win over laziness
11:15:06amiconnAnd, it would combine the radio actions. Short SELECT = presets, long SELECT = context menu
11:15:55amiconnBtw, the volume error with non-flat tone controls happens when first entering the radio screen
11:16:59rasherI'm torn on this button thing
11:17:24rasherBut then.. it'd be a-b to leave, which goes to menu
11:17:30rasherso it's not that illogical
11:18:39rasherHrm, am I now back to the suggestion in that forum
11:18:54rasherhaha yes
11:18:55rasheryes I am
11:19:30amiconnHmm, indeed
11:20:00amiconnMaybe I just misinterpreted the "AB go to main menu" thing as actually calling the menu
11:20:20amiconnIf this would be "leave with radio on", it seems okay
11:20:37rasherI don't see the point in wanting anything else
11:20:49rasherit's not like there's any state to be saved that isn't saved if leaving with radio on
11:21:52amiconnLeaving the radio screen always stops an eventual recording, so mute makes sense
11:28:01rasherWhee, I think we have another case of 0 == undefined according to preprocessors
11:28:03rasherhang on
11:28:56rasheroh well
11:29:10rasherguess I just screwed up then
11:29:12amiconnI found the place that causes the radio volume error.
11:29:28amiconnFixing it needs some rework of uda1380.c
11:30:45amiconnPerhaps not...
11:31:21CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
11:31:21*amiconn looks up the uda1380 datasheet
11:40:27rasherI'm confused
11:40:41rashermy button changes don't seem to be reflecing what's actually happening
11:42:22 Join einhirn [0] (
11:43:08amiconnRadio volume issue fixed.
11:43:40rasherwell that was fast
11:43:51rashermeanwhile, I haven't figured out how to change the buttons
11:45:01rasherI give up.
11:45:18rasherThis doesn't make sense
11:45:45amiconnShould be simple (except mute, as that doesn't exists yet)
11:46:06rasherIt should, but I'm messing it up
11:46:08amiconnOnly one pitfall - you'll need to introduce FM_PRESET_PRE
11:46:57rasherI can't even manage to make a-b exit
11:47:28amiconn#define FM_EXIT BUTTON_MODE
11:47:29rasherwhich is confusing me
11:47:40amiconn(and remove #define FM_EXIT_PRE)
11:48:02rasherI thought I tried that
11:48:18amiconnThis does of course require to change FM_MENU as well, as otherwise there would be 2 identical case: values
11:48:43amiconn#define FM_MENU (BUTTON_SELECT | BUTTON_REPEAT)
11:49:14rasheroh jesus christ
11:49:28rasherWell I forgot that
11:51:14rasherThink I'll try again, and actually think this time
11:52:05rasher#define FM_PRESET_PRE BUTTON_SELECT
11:52:10rasher#define FM_PRESET (BUTTON_SELECT | BUTTON_REL)
11:52:11rasherlook fine?
11:53:08rasherand then add the check
11:53:19amiconnOf course that requires to add a piece of code to case FM_PRESET:
11:54:00amiconn(#ifdefed properly)
11:55:36Zagoramiconn: would you say Vaga Bund is a real german name?
11:55:37rasherThink I have it now
11:56:14amiconnZagor: Not really
11:56:29Zagorthat's what I felt too
11:56:30rasherDammit, no!
11:56:41amiconnZagor: Btw, could you check what went wrong with building the 2.5 installer?
11:56:53amiconnIt contains binaries from 2005-06-07 ...
11:57:33rasherrecorder/radio.c:843: error: duplicate case value
11:57:33rasherrecorder/radio.c:820: error: previously used here
11:57:43amiconnZagor: "Vagabund" in german is vagabond in english
11:58:30rasherIs it using MENU_ defines from elsewhere to confuse my poor brane?
11:58:40amiconnSeems so...
11:58:55rasherAnd what does that case even do?
11:58:59amiconnThis callback function looks messy...
11:59:47 Join webguest29 [0] (
12:01:32rasherI think I'll leave this in more competent hands
12:01:54rasherI could give you a patch of what I have now, but it's not really worth it
12:02:11rasherit's either been written by you or pasted by me, so
12:10:45 Quit webguest47 ("CGI:IRC")
12:13:04linuxstb_Morning all. So there are problems with iTunes 5 and ALAC?
12:13:18rasherYeah, probably minor
12:13:47rasherfuzzie wants to use the metadata parser from faad
12:14:01linuxstb_As long as it is small and efficient.
12:14:01rasherand use that for alac as well
12:14:23***Saving seen data "./dancer.seen"
12:14:27linuxstb_I don't have a problem with that. We will have a problem with two codecs using the .m4a extension though.
12:15:22rasherI guess using .aac can be used until that's put in place?
12:15:54linuxstb_Or changing ALAC to .alac - considering AAC will be by far the most used of the two.
12:16:14 Join noC|andY`fRa [0] (
12:16:18linuxstb_But we should probably try and fix it properly instead.
12:16:24rasherOf course
12:16:34rasherI just meant so as to not halt work
12:17:25rasherI dropped by the channel the alac author mentions on his webpage and mentioned the itunes5 thing
12:17:29rasherdon't think he's there though
12:18:01 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
12:18:10linuxstb_fuzzie could just disable the ALAC decoder for his testing - i.e. map .m4a to AFMT_AAC
12:21:49rasheramiconn: if anything, reception seems worse now?
12:23:35rasherfrankly, I'm underwhelmed by the iriver's reception
12:23:35amiconnIt shouldn't
12:23:44rasherI agree
12:23:56rasherMight just be coincidence
12:23:58amiconnShould be the same, or slightly better, depending on the station
12:24:15amiconnZagor: Any special reason why you removed the "Index" link from the wiki?
12:24:39 Join Tom101 [0] (
12:26:45rasherAh, charging creates worse reception
12:27:35Zagoramiconn: i didn't feel it was very useful. do you use it regularly?
12:28:06ZagorI reasoned that most people will probably rather search than see the monster index
12:30:02rasherThat's what I told him, but he wouldn't listen!
12:30:43 Quit Tom101 ()
12:30:52amiconnZagor: I used it quite often. Guess I have to resort to the search function...
12:31:22Zagoramiconn: or simply type WebIndex in the Go to box
12:33:16amiconnrasher: Why is there a ReleaseNotes-25.txt attached to the ThingsTodo topic?
12:33:53Zagorit was a test, can be safely removed
12:34:21rasherforgot about that
12:35:45*rasher adds very unhelpful fm presets
12:46:15amiconnlinuxstb_: For your ipod port, I think we shouldn't detect everything at runtime, but only some things
12:46:31rashervariations within single generations?
12:46:44rasherso we at most get a version for each gen
12:47:11amiconnI think it would make sense to have different builds per main type (plain ipod, mini, color/photo, and nano)
12:47:27amiconnThe generations should be detected at runtime, imho
12:47:35rashereurgh runtime display adaption :-\
12:47:55rasherdon't the newer greyscale have different displays than the old ones?
12:47:57amiconnSame way as we do for archos player (old vs. new lcd), Ondio FM (samsung vs. philips tuner)
12:48:15rasherBut different sizes, I mean? (I could be wrong)
12:48:27rasherah, that's good
12:49:02rasherit's not like I'll let facts get in the way
12:49:17rasherBut yeah, sounds sane then
12:49:25amiconnThere are a few more places where we adapt to hardware revisions at runtime on archos
12:49:38rasherDifferent keypads though
12:49:41rashermight not be fun
12:50:05amiconnOkay, if the unit is physically different, then a different build makes sense
12:50:26rashermight have to split "plain ipod" in 1-3gen to 4gen
12:51:18rasheractually, 3rd gen is quite different from 2nd
12:51:28rasherand 4th
12:51:30rasherit seems
12:51:38rasherfrom looking at the pictures
12:52:46amiconnMy point is that it's not user friendly if there are different versions for hardware that looks the same to the user
12:53:10amiconnIf the user is able to tell the difference, different builds are no problem
12:53:47rasherEach gen is physically different I believe
12:54:46 Quit Vladoman_ (Read error: 104 (Connection reset by peer))
12:58:42amiconnFrom the images on ipodlinux, it seems that 1G/2G look the same, and 3G/4G too
12:59:12rasher4g doesn't have the 4 buttons under the display
12:59:59rasher1g/2g is right though
13:00:03rasherhadn't checked that
13:00:47 Join Vladoman [0] (
13:04:01rasherso 1g/2g, 3g, 4g, mini, color/photo and nano?
13:06:03linuxstb_amiconn: I agree about the ipod - my plan at the moment is to split the targets by LCD and keypad.
13:06:54rasherwon't that end up like I put it?
13:07:20rasherwith perhaps 1g/2g and 4g merging, depending on how the controls work
13:11:56rasheramiconn: Seen this: ?
13:13:49amiconnIt seems this is related to an older rockbox (only hint is "the 2.4 software" in the middle of the text)
13:14:45rasherAh, missed that
13:16:24 Join cYmen [0] (
13:24:40rasherlinuxstb_: you around?
13:25:01rasherOh wait..
13:25:16rasherWho was building rockbox for iriver with gcc4?
13:25:33linuxstb_rasher: Yes.
13:25:45rasherwell, I misremembered
13:25:50amiconnpreglow (iirc)
13:25:58rasher might be interesting
13:26:06rasherI believe it ICEd out?
13:26:30amiconnIirc the binary crashed on iriver...
13:26:42rasherlinuxstb_: I just saw "gcc" "m68k" and "ice" and thought of you because of the deal with faad
13:26:48rasherwrong connection though
13:29:32 Quit linuxstb_ ("Leaving")
13:39:33 Join linuxstb [0] (
13:40:58linuxstbI've just committed a patch which fixes the metadata reading for iTunes 5 files - so ALAC is working fine again.
13:41:22linuxstbI'll have a look at the "crashing on AAC files" problem reported in the Forum.
13:41:35linuxstbI think I know an easy fix.
13:42:28Zagor /* FIXME not 8 */ ? :-)
13:45:13amiconnZagor: Any news concerning the installer problem?
13:45:29Zagorhaven't had time to look yet. removed the 2.5.exe from download though
13:45:43amiconnAhem, just noticed the daily installers aren't built since 2005-09-17
13:48:48linuxstbZagor: That's taken from the alac decoder itself. I think Quicktime files can be either 32-bit or 64-bit, and the alac decoder is hard-coded to only deal with the 32-bit case.
13:49:21linuxstbI think there's a plan to use the faad m4a parser - which should be more reliable and deal with more cases.
13:51:18amiconnZagor: I think the wiki deletion also removed something necessary for building the installer, and so both the daily installers aren't built and the 2.5 installer build failed
13:51:31Zagoramiconn: seems like it, yes
13:51:48amiconnAs you alreaedy prepared a 2.5 release in June, maybe you already made an installer
13:52:07amiconnSince the real 2.5 build failed, you put up the old one (w/o checking the file date...)
13:52:38amiconnI could build a 2.5 installer locally, from the 2.5 release .zip s
13:53:13Zagorthat sounds like a plausible theory
13:53:27Zagorsure. but I still need to fix the builder.
13:56:57linuxstbOK, I've hopefully stopped the ALAC decoder crashing on AAC files as well.
14:00:23rasherthat looks nice
14:00:35rasherand like boring work
14:00:48rasherstepping back from where errors happen
14:03:17linuxstbIt should be done everywhere in demux.c - but I'm hoping we will replace that code with the faad parser.
14:04:36rasher"adding bass just squashes out alot of the other mid and top values it seems" just shows how much people imagine things
14:04:47rasher(when at 100% volume)
14:07:01rasherHoly crap, that's a lot of fm presets
14:12:53rasherOh dear, I forgot about the yield thing
14:14:27***Saving seen data "./dancer.seen"
14:18:45linuxstbYes, I think yield() needs to be called from inside libalac itself.
14:20:05linuxstbI think the only way I can do that is to pass a pointer to the rb->yield() function to the decode_frame() function.
14:27:34rasheralac is indeed fixed
14:33:21linuxstbYes - adding more yields from inside libalac has done the job. The UI during ALAC playback doesn't quite feel as responsive as other codecs yet, but it's very close.
14:35:09 Join jonash [0] (
14:35:55linuxstbOK, that's my last ALAC fix committed - it's now much more co-operative in its yields.
14:37:06*linuxstb just read that pause doesn't work for ALAC files...
14:37:32linuxstbI'll have to look at that later. Got to go now.
14:49:32amiconngotta go now
14:50:01amiconnZagor: I'll keep my dyndns web server running
14:50:10 Part amiconn
14:54:26 Quit rasher (Read error: 110 (Connection timed out))
14:54:48 Join XavierGr [0] (
14:56:16 Join Lear [0] (
14:59:24 Join Sucka [0] (
15:01:23 Quit jonash (Read error: 110 (Connection timed out))
15:05:21Zagoramiconn: thanks, grabbed now
15:10:06 Quit Lear (Excess Flood)
15:15:02 Quit Febs (" HydraIRC -> <- State of the art IRC")
15:15:12 Join Lear [0] (
15:15:14 Join webguest09 [0] (
15:16:17 Quit webguest09 (Client Quit)
15:25:15 Quit XavierGr ()
15:36:46 Join tucoz [0] (
15:37:20tucozHi, is there a reason to why .fmr files do not have the open-with entry in the context menu? all other files I think have this.
15:38:34LearThere's no viewer defined for it?
15:38:57Zagortucoz: it's simply not implemented. there is currently no support for multiple fmr files
15:41:26tucozYes, but I mean, one would like to view it with the viewer?
15:41:30tucozlike text
15:42:27tucozAs it is possible to open a binary file with the viewer. For instance a mp3 file.
15:45:00 Quit Sucka ("a bird in the bush is worth two in your house")
15:46:09tucozOk, better example. It is possible to open up a .jar file (which is def. unsupported) with the viewer.
15:46:54tucozThat is why I thought there were some special reason to why .fmr files doesn't have the open with entry
15:47:29 Join ShacharLiberman [0] (
15:47:58ShacharLibermanguys, I'm getting a:
15:47:59ShacharLibermanTopic save error
15:47:59ShacharLibermanDuring save of file Main.PluginChip8 an error was found by the version control system. Please notify your Rockbox administrator.
15:47:59DBUGEnqueued KICK ShacharLiberman
15:47:59ShacharLibermanSave attachment error /usr/bin/ci -q -l -m%COMMENT|U% -t-none -w%USERNAME|S% %FILENAME|F%
15:47:59ShacharLibermanci: missing message for -m option
15:48:03ShacharLibermanGo back in your browser and save your changes locally.
15:48:11ShacharLibermanwhen I'm trying to upload the chip 8 games to
15:49:03ShacharLibermanthey are missing because of the wiki bug
15:49:41ShacharLibermanthe wiki bug? -yeah.
15:49:51tucozAnd that you cannot upload the files
15:50:03ShacharLibermanthat too. what 2do?
15:50:17tucozHmm, Zagor is the man for this I guess
15:51:08ShacharLibermanwell, he's not saying anything.
15:51:17tucozHe was around like 15 minutes ago
15:51:48tucozIf you got the time, I suggest you just wait some more.
15:52:12ShacharLibermanI'll be connected, just not around. I'll try to find my registered nick as well :(
15:52:33tucozOk, could you mail them to rasher perhaps?
15:53:35ShacharLibermannp. what's his address?
15:53:37tucozI think that he keeps track of what is missing and not.
15:54:15tucozok, it is rasher at
15:54:50tucozrasher: I just thought that you wanted it done this way. :)
16:00:31 Quit ashridah (Read error: 110 (Connection timed out))
16:01:31ShacharLibermanwill you mind helping me in another problem? i've installed cygwin from downloaded packages on a computer, and then found out that i was missing the Bash shell (so cygwin won't start at all). now i've downloaded bash & bash completion packages (version 3.0-11 & 20050721). now cygwin starts but the syntax is "BASH-3.00$" and not whatever it used to be, and no installed packages are running like mc or xemacs, i can't even dir or ls. should i
16:02:23ShacharLibermanI would love and answer I will be off for the next 20 minuts thanks!
16:02:52 Join ashridah [0] (
16:07:49lImbusShacharLiberman: I've been very successful with the DevKit BlueChip is providing:
16:08:03lImbusIt's a completely packed up cygwin
16:14:28***Saving seen data "./dancer.seen"
16:26:05 Quit ashridah ("Leaving")
16:27:02ZagorShacharLiberman: sounds like you didn't provide a description for the attachment. the new version of twiki requires it.
16:29:07 Join pike [0] (
16:29:56 Quit Hadaka (Read error: 110 (Connection timed out))
16:31:02 Join hd [0] (
16:32:07LearIf so, I'd say that's a pretty bad error message. :)
16:35:25tucozInteresting to read about the reorganization of Vista:,,SB112743680328349448,00.html?mod=todays_us_page_one
16:35:49tucozMakes me think that windows might end up quite good.
16:36:00LearYou mean like how they move much of the drivers out from kernel space again?
16:36:31tucozYes, to make a smaller kernel
16:36:56LearLess likely to crash, rather.
16:37:25tucozBut I have to say, windows xp doesn't crash that often
16:37:28LearBut I'm not sure about all the DRM and related "security" stuff though.
16:37:33tucozOr rather, almost never
16:38:06tucozBut isn't apple also in love with DRM and stuff like that?
16:38:10LearWhen it does for me, it's likely to be a certain driver-related thing (IRQ less than, or something).
16:38:29tucozAnd I think that osx looks like a nice OS
16:39:00LearNot into the OS core like what appears to be the case in Vista.
16:39:17tucozOh, didn't know that.
16:39:19t0masLear: "IRC_NOT_LESS_OR_EQUAL"
16:39:26t0masa popular one ;)
16:40:09LearExactly. I had problems with that for a while, then I realized I needed to install some SATA RAID drivers from within windows (not just during install), after that, it's been much better.
16:40:28t0maswell... windows still sucks on installing/updating drivers
16:40:38t0masspecially videodrivers
16:40:43tucozmy experience of SATA has been a PITA
16:40:49t0mas(doesn't matter if it's ATI or Nvidea)
16:41:47t0masand security is just almost impossible for windows workstations.... if it's your own workstation, and you want security from the outside world it's possible with some thirdparty software...
16:42:26t0masbut security from the inside (eg. the pc's at my vacation job needed to be used by visitors, but shouldn't be crackable)
16:42:30t0masthat's just impossible
16:43:00tucozt0mas: I heard Vista will incorporate some sort of SUDO when installing programs.
16:43:10t0masyou can't work without admin rights (lots of software doesn't work without it)... and you can't secure a workstation agains someone with admin rights (eg. disable registry editing, and people will use the reg import tool on the commandline :P)
16:43:35t0masit's just coded the wrong way around...
16:43:51t0masif you disable the commandprompt (cmd.exe) from tweakui... you can still run batch files...
16:44:06Zagorit's a single-user system. always has been.
16:44:15t0masand if you disable registry editing (policy) then you can use the commandline tool "reg import file.reg" to change whatever you want
16:44:47t0masand if you can edit the registry, all oter policysettings are useless... you can disable them from a .reg file
16:45:17tucozt0mas: are you talking about vista or xp?
16:45:35t0maswe haven't tested if it's still this way in vista
16:45:55t0masit would be a good reason to upgrade if vista is better at this...
16:46:44t0masat this moment we have linux terminals, 1. they are more secure if you don't have root access... 2. people don't know that much about linux, so it takes longer to find leaks :P
16:47:00 Quit goa (Read error: 113 (No route to host))
16:47:01 Nick hd is now known as goa (
16:47:21tucozBy reading that article, I got the feeling that Vista is a much better os in terms of how the core system is arranged
16:47:31ShacharLibermanLimbus: how have you managed with the devkit?! without the setup program of cygwin I can't download new packages so all i can do is build rockbox up.
16:47:33tucozmuch better that xp at least
16:47:50ShacharLibermanZagor: i gave a description but i will try again and let you know.
16:48:14tucozDid you check the check box?
16:48:22tucozor whatever it is called
16:48:40ShacharLibermangimme a minute
16:48:41lImbusShacharLiberman: you can restart the cygwin-setup afterwards. it will recognise the bluechip-thing as cygwin and allow new packages to be installed
16:49:02t0mashaha... today I had something funny
16:49:03ShacharLibermansomone has uploaded them already :P
16:49:16t0massomeone asked my dad if he knew what rockbox was...
16:49:16 Join Naked [0] (
16:49:22t0mashis name was in the releasenotes of 2.5
16:49:29 Nick Naked is now known as Hadaka (
16:49:46t0mashe was wondering what the other guy found...
16:49:52t0masand found out it was my name
16:50:17ShacharLibermanlImbus: but now that I already have all the packages except the bash working properly it's a shame :/ but thanks i'll try that
16:54:19tucozt0mas: just curious, did your dad know what rockbox is?
16:58:14t0masno he didn't
16:58:31t0masbut when he read it was an mp3 player firmware he did know it was me...
16:58:44t0masbecause I showed him my iriver sometimes... showing new features
17:09:08CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
17:09:08*fuzzie yawns.
17:13:17*tucoz leaves
17:13:19 Part tucoz
17:13:51ShacharLibermanLimbus (or anyone): where's "which" & "mc" packages are hiding?
17:14:21ShacharLibermanwoop! found which
17:14:51ShacharLibermanbut where's midnight commander?
17:16:47ShacharLibermanhelp, someone?
17:18:26lImbusdunno, sorry
17:19:23ShacharLibermani'm just asking where can you find midnight commander at :/ what else could you use? plain shell?
17:27:30 Quit ShacharLiberman ("CGI:IRC (EOF)")
18:14:29***Saving seen data "./dancer.seen"
18:22:49 Join einhirn [0] (
18:29:25 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
18:31:53 Quit Lear ("Chatzilla [Firefox 1.4/undefined]")
18:32:18 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
18:52:23linuxstbfuzzie: How is the AAC decoding going?
18:53:16fuzzieI got it to work fine using the mp4ff library I mentioned, wondering exactly how/where to integrate that.
18:54:00fuzzieIdeally it'd be accessible from both metadata.c and my codec, which means I can't just link it to one of them.
18:54:17fuzzieWell, 'work fine' means 'it seems to work with one test file in the simulator'
18:54:44fuzzienothing beyond just playing, yet.
18:54:51fuzziei assume most of the work will be optimisation
18:55:32 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
18:56:55 Join einhirn [0] (
18:57:43 Join webguest00 [0] (
18:59:02 Quit einhirn (Client Quit)
19:01:18webguest00good afternoon - is it likely patch 1298779 will get committed soon? (corrects ato power off)
19:05:01 Quit webguest00 ("CGI:IRC")
19:08:38linuxstbfuzzie: That's good news. I'm happy to have a look at it if you want.
19:10:03linuxstbDoes the existing metadata.c m4a parser work with your AAC files?
19:13:38fuzzieit worked on the test file i'm using with some minor modifications
19:13:55fuzzienot on others, though
19:14:40linuxstbDo you know how much memory mp4ff uses when parsing the metadata?
19:15:12fuzzienope, i'm not sure if it's an option
19:15:18fuzzie <- if you want to take a look
19:15:42linuxstbThanks, I'll have a look. Have you tried playback on a real device, or just the simulator?
19:15:47fuzzieit might be best just to fix the code which is present, that's what i'm looking at it now
19:15:58fuzzieand, just the simulator.
19:17:50linuxstbI see there are lots and lots of mallocs in mp4ff - we want to avoid those if we can.
19:21:06fuzziea shared m4a metadata parser would probably work, it just needs to be capable of skipping all kinds of odd atoms
19:22:49 Join DangerousDan [0] (
19:32:53 Join bagawk [0] (n=lee@unaffiliated/bagawk)
19:54:00linuxstbSlasheri: Are you around? It seems we have reached the stage where we need to solve the problem of two codecs using the same extension (m4a).
19:55:00 Join Moos [0] (
19:55:32MoosHello folks
19:56:40MoosI just bought an ondio SP :)
19:57:25Moosfor preseving a bit my iriverh140 for the nights
19:57:46MoosHello lImbus, thanks
19:58:01Moosit look very fine
20:00:44Slasherilinuxstb: sorry but almost completely away before the next week :/
20:01:44linuxstbI don't mind looking into it. It will be a chance for me to finally try and completely understand playback.c
20:09:05 Part Pieter_
20:10:17 Quit bagawk ("Leaving")
20:14:32***Saving seen data "./dancer.seen"
20:31:08 Join solexx_ [0] (
20:37:08 Quit solexx (Read error: 104 (Connection reset by peer))
20:40:55 Join rasher [0] (
21:01:16 Join kurzhaarrocker [0] (
21:01:52 Quit kurzhaarrocker (Client Quit)
21:04:13 Join Paul_The_Nerd [0] (
21:09:32Paul_The_NerdI'm sure it's been discussed before somewhere, but I'm curious if once encoding is enabled whether transcoding files on the HD would be doable?
21:09:53Paul_The_NerdSorry, by "doable" I meant "a feature" rather than "physically possible"
21:16:48ZagorPaul_The_Nerd: no it won't be added
21:16:54 Join Sucka [0] (
21:16:59Paul_The_NerdFair enough.
21:17:25Zagorit's one of those things that's possible but rather pointless to do on the player.
21:17:44rasher(currently you can do wav->wavpack)
21:19:06Paul_The_NerdI thought of a few reasons I personally might want to do it. How hard would you say it might be for someone to create their own patch to do such a thing?
21:22:16Zagorit can be pretty tricky, I'd say
21:26:30Zagoryou'd probably have to solve it by loading codec A, uncompressing to ram, unloading A, loading B, compressing to disk, repeat
21:27:12Paul_The_NerdI was thinking of actually uncompressing to disk, since ram may not always be big enough.
21:27:19Paul_The_NerdBut yeah
21:27:25Zagorand, unless it isn't obivous: it will be slooow :-)
21:27:39Paul_The_NerdIn *most* of the cases I imagine I'd want it, I'd be taking uncompressed audio and compressing it anyway
21:27:59Paul_The_NerdI can just see myself recording say, the first day of something, realizing I didn't have as much space free as I thought, and not having someplace to offload it overnight.
21:28:24Paul_The_NerdSo as long as the compressing was realtime or better, and it was going from WAV to whatever, no biggy.
21:28:40Paul_The_NerdI imagine I'll be able to work something out.
21:28:48ZagorI doubt you'll get realtime from any codec
21:29:56Zagorhowever I have no numbers on that, so that's just speculation
21:30:12Paul_The_NerdFair enough.
21:30:32Paul_The_NerdBut don't they have to reach realtime to compress while recording?
21:30:41Zagorafaik no codec has been optimised for recording (compression) yet
21:31:04Paul_The_NerdI thought the goal was to have at least MP3 working for that (assuming the "can we use mp3" argument ever resolved)
21:31:47Zagoryeah, it'd be nice to record in all formats. but I don't know how far away that is.
21:33:14Zagori still find the need for transcoding pretty hypothetical. but I'm not arguing with your right to choose your own project :-)
21:33:16Paul_The_NerdYeah, but as soon as they get any one lossy format working for compression, it'd suit my purposes.
21:33:31Paul_The_NerdYes, well, it wouldn't be a project any time soon.
21:33:46 Quit webguest29 ("CGI:IRC")
21:34:11Paul_The_NerdI just know that any time I think of a "wouldn't it be great if X did Y" for *anything* I own, I eventually find myself in a situation if X did do Y, I'd be much better off.
21:35:42Zagoryeah, maybe i've grown over-dismissive over the years. open firmware tends to bring out many "hey, wouldn't this be cool?" ideas...
21:35:47 Nick solexx_ is now known as solexx (
21:36:10Paul_The_NerdI agree though, transcoding on-box seems pretty special-case useful.
21:37:27Paul_The_NerdI was just wondering how difficult it seemed likely to be to implement once encoding codecs were in.
21:40:18Zagorif you decompress completely to disk first (or only compress from wav), it's probably not that hard actually.
21:40:33Zagoroh well :-)
21:41:10linuxstbAll we need to do is given plugins access to the codecs. We can then write a generic "decode to wav" plugin, and eventually "encode from wav" plugin.
21:41:21 Join _DangerousDan [0] (
21:42:12linuxstbA wav-writer plugin is useful for a) optimising the codecs and b) testing that codecs (especially the lossless ones) are giving bit-perfect output.
21:42:44 Part rasher ("Ex-Chat")
21:42:53Zagorlinuxstb: yeah. I was making it too complex in my head, running dual codecs in parallell and such...
21:42:57 Join rasher [0] (
21:46:10Paul_The_NerdComplex can be a synonym for "fun" if you're the right sort of crazy, y'know.
21:47:05Paul_The_NerdI'm *usually* that sort. But it's been forever since I programmed anything, and I keep delaying learning about the internals of rockbox.
21:50:01 Join hd [0] (
21:52:36 Quit goa (Connection timed out)
21:52:36 Nick hd is now known as goa (
21:57:06 Quit DangerousDan (Read error: 110 (Connection timed out))
22:05:07 Quit rasher ("Ex-Chat")
22:05:47solexxmy 2 cents: if the hw is too slow for realtime encoding to ogg/mp3, I'd still vote for flac :9
22:06:18solexxI just compared encoding times from wav to ogg and flac
22:06:45solexxand flac (default compression) is ~3 times faster than oggenc (quality doesn't matter much)
22:07:49solexxi think especially for people doing long recordings, encoding would be a very cool thing.
22:08:08 Quit noC|andY`fRa (Read error: 110 (Connection timed out))
22:12:56linuxstbI could be wrong, but I think David Bryant has got his wavpack encoder (lossless mode) working in realtime on the H1x0.
22:13:35linuxstbSo we just need a proper "recording codecs" architecture to plug it in to the iriver recording.
22:13:37solexxgreat! (never heard of wavpack except through rockbox, though)
22:14:33***Saving seen data "./dancer.seen"
22:14:49linuxstbIn fact, there's a "wav2wv" (wv is the extension for wavpack) plugin already in CVS.
22:15:14Paul_The_NerdDoes anyone know of webpages where codecs are compared?
22:15:20linuxstbI'm just testing it now, and it's showing that it is encoding at 176% realtime :)
22:15:36Paul_The_NerdWould that just various forums, and like, HA?
22:16:02linuxstbYes, HA is probably the best source. I think wavpack is highly regarding.
22:16:10Paul_The_NerdI know it is.
22:16:15Paul_The_NerdI haven't used it lossy at all yet though.
22:16:35Paul_The_NerdI have enough space on my iRiver right now for my entire CD collection in lossless.
22:16:42linuxstbGiven how well David Bryant optimised it for the iriver, it's probably the best lossless codec to use in Rockbox at the moment.
22:17:01Paul_The_NerdThough one or two of my songs still seem to skip.
22:17:31Paul_The_NerdLike... I think I have maybe two that don't quite decompress fast enough.
22:18:07linuxstbFile a bug report if you can replicate it - I'm sure wavpack can run in realtime with the CPU at 48KHz most of the time, so it shouldn't skip on any file.
22:18:28Paul_The_NerdI need to check to make sure it's not the file too.
22:18:55linuxstbIf you go to the Debug menu, and select "view audio thread", you can see what is happening during decoding.
22:19:19linuxstbThat will tell you if it's a CPU issue, or another problem.
22:19:21Paul_The_NerdI've been there, but I'm always driving when it happens, and forget which song it was.
22:20:50solexxPaul_The_Nerd: see to compare yourself
22:20:54Paul_The_NerdIt's also possible that the player's getting shaken just enough.
22:21:13solexxdoesn't contain wavpack, though
22:21:24Paul_The_NerdThat's more what I was looking for. WV and MPC
22:21:47solexxthe original wav is there, so if you have an encoder...
22:21:54Paul_The_NerdVorbis and various mp3 I've got enough experience with
22:21:57Paul_The_NerdActually, I might do that.
22:22:22solexxhow "free" is wavpack compared to flac/ogg?
22:22:40Paul_The_NerdEqually, if I recall.
22:23:53Paul_The_NerdThey describe it as "completely open"
22:25:11solexxThanks, sounds really interesting. Especially the hybrid mode.
22:25:30 Quit Sucka ("a bird in the bush is worth two in your house") even lists rockbox under "supported hardware" :)
22:26:48Zagorhmm. the wavpack license includes an advertising clause...
22:26:55Paul_The_NerdWhat sort?
22:27:28Paul_The_NerdYeah solexx, but if I recall Rockbox doesn't detect if a hybrid file is present, and just reads the lossy portion.
22:27:36Zagormust redistribute entire license in docs "and/or other materials provided with the distribution"
22:29:17Zagornot as bad as bsd, but the gpl-compatibility is not guaranteed
22:30:01solexxdo plugins need to be gpl-compatible?
22:30:33Paul_The_NerdI *think* it's okay though
22:30:41Paul_The_NerdAll it says is the copyright notice *must* be reproduced.
22:30:48Paul_The_NerdDoesn't the GPL have a clause like that?
22:30:55solexxhm. debian doesn't ship any wavpack software...
22:31:01Zagornot just the copyright notice, the entire license with terms and disclaimer
22:31:23Paul_The_NerdI seem to recall the issue with Faad2 wasn't that it said that the reproduction of the notice was necessary, but the fact that it stated *how* it must be reproduced (IE, forced it to be displayed in the software)
22:31:28ZagorI wonder if that could be defined as "additional requirements"
22:31:43ZagorPaul_The_Nerd: yes, they did the old bsd mistake
22:33:27Paul_The_NerdActually, the third term, the "neither..." is the one I'd be worried about.
22:33:49Paul_The_NerdBecause the first two terms "must be reproduced."
22:33:57Paul_The_NerdBoth of them are present in the GPL, I think.
22:34:09Paul_The_NerdSo, they are reproduced *by* the GPL.
22:34:48Zagori don't know. it's not obvious.
22:35:06Paul_The_NerdBut the third one "Neither the name..." etc, is very definitely an additional requirement.
22:35:14Paul_The_NerdIf it has to be reproduced as well.
22:35:39Zagor"must reproduce the above copyright notice, this list of conditions and the following disclaimer"
22:35:52Zagorit doesn't say "the first two conditions"
22:35:59solexxStrange. This issue doesn't seem to have been discussed on debian-legal yet
22:36:22Paul_The_NerdSo, the first two could be argued over, but the third is clearly an addition requirement I think.
22:36:28ZagorI think this is just a home-cooked license nobody really gave much thought about
22:36:32Paul_The_NerdYeah, probably
22:36:41Paul_The_NerdI think he wanted it to be less restrictive than GPL
22:36:47linuxstbI've got a feeling wavpack used to be GPL'd.
22:36:54Zagorprobably would be easy to persuade them to change to BSD
22:37:05solexxIs "wavpack" a "name of Conifer Software"?
22:37:06Paul_The_NerdI *bet* if you said "Hey, you want wavpack to be in Rockbox right? Will you license a version to us under GPL?"
22:37:14linuxstbI also don't think it's "them" - just David Bryant.
22:37:26Paul_The_NerdYou'd get a "sure"
22:37:43Paul_The_NerdJust because it's released under one license, doesn't mean it can't also be released under another simultaneously.
22:37:44Zagoraha, david is the sole genius
22:37:57ZagorPaul_The_Nerd: right
22:38:27Paul_The_NerdAnd I'd get good odds you'd get a "fine" or a "I don't really like it, but maybe we can settle on another license that's compatible"
22:39:22linuxstbZagor: I'm not 100% sure if David Bryant is the lone genius, but I haven't heard any other names mentioned.
22:40:13ZagorI'll fire off a friendly email
22:42:18 Quit fuzzie ("leaving")
22:47:44Paul_The_NerdWell... I don't know of the wavpack problem is simply not reproduceable, or just gone.
22:48:18Paul_The_NerdThough, I suspect it may have something to do with codec swtching. I've noticed sometimes when I next track, it plays a fraction of a second of the next track, then some from the track I'm leaving, then starts playing the next track properly, which seems odd behaviour as well.
22:48:58solexxPaul_The_Nerd: I have this behaviour quite often with rockbox
22:49:00Zagorman, I'm out of my head today. that *IS* the BSD license. bah!
22:49:06ZagorI should get more sleep
22:49:16linuxstbAny bugs you can find, add to the IRiverStatus wiki page.
22:49:41linuxstb(Sorry, IriverStatus)
22:50:02solexxlinuxstb: didn't think of it as a bug until now, just "unclean" behaviour
22:50:25linuxstbAnything that isn't perfect, mention on that page. We can decide later if it's important or not.
22:50:49Paul_The_NerdI still can't reproduce it consistently.
22:51:10Paul_The_NerdI was waiting to really bring it up until I could figure out how to definitively tell someone a way to cause it.
22:51:41Paul_The_NerdIt *seems* to be switching tracks from WV to MP3 when the next song isn't buffered. Maybe.
22:52:32solexxPaul_The_Nerd: not in my case - I only have mp3 and ogg
22:52:55Paul_The_NerdI figured it might happen on others
22:53:04solexxand it occurs even when switching tracks from the same album (==same codec)
22:53:11Paul_The_NerdI haven't noticed that
22:53:47solexxwill try to narrow it down.
22:53:47Paul_The_NerdDoes that happen with MP3s, Oggs, or both?
22:53:57solexxdon't know
22:55:25solexxcannot test it now, I have a friend over here
22:55:40solexxHe's already irritated that I am IRC'ing all the time :)
22:55:56Paul_The_Nerdlinuxstb: Where would it fit on that page?
23:09:47 Join DrMoos [0] (
23:09:47 Quit Moos (Read error: 104 (Connection reset by peer))
23:15:25linuxstbPaul_The_Nerd: Not sure. Maybe add a table called "Other Issues" after "Tasks Remaing" and put it in there.
23:16:17 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
23:17:03Paul_The_NerdFair enough.
23:17:34 Quit paugh (Client Quit)
23:19:32 Nick DrMoos is now known as Moos (
23:26:36 Join Sucka [0] (
23:40:09 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")

Previous day | Next day