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 2005-09-30

00:11:22 Join Ickyb0d [0] (
00:12:05 Quit Ickyb0d (Client Quit)
00:14:03 Quit RotAtoR ()
00:15:33 Quit dpassen1 ()
00:22:10 Quit Moos ("Glory to Rockbox")
00:27:15 Quit ashridah ("out seeing serenity")
00:35:29 Join Naked [0] (
00:36:50 Quit Hadaka (Read error: 104 (Connection reset by peer))
00:36:50 Nick Naked is now known as Hadaka (
00:47:06 Quit DangerousDan ("Miranda IM! Smaller, Faster, Easier.")
01:02:15 Join ilikedirt [0] (
01:02:48 Join bagawk [0] (i=1000@unaffiliated/bagawk)
01:05:44 Quit matsl ("Leaving")
01:09:03 Part ilikedirt
01:13:39***Saving seen data "./dancer.seen"
01:22:40rasherXavierGr: Does your "fm preset file loading" patch ignore lines preceded by #?
01:26:07XavierGrin the presets?
01:26:56XavierGrbecause I havn'e changed anything in the way that the presets are saved inside a file.
01:27:18rasherDo you know if comments are allowed?
01:27:30XavierGrI think not, let me check
01:27:54 Join Mirthoneist [0] (
01:28:02MirthoneistHey BirdFish, you there?
01:28:30XavierGrno rasher I don't think so.
01:28:44rasherOkay. Oh well, nevermind.
01:28:50 Nick TiMiD[zzzZZ] is now known as TiMiD (n=TiMiD[
01:28:52rasher(shouldn't be too hard to add anyway)
01:29:13XavierGrwhat do you want to do?
01:29:33rasherI'm just writing a script to scrape presets from the wikipage into a zip file
01:29:59XavierGrand what the # has to do with it?
01:30:03rasherand I wanted to add whichever comments were on the page (author, notes etc.) into the file
01:30:34rasherIt should be a 2 line patch to wherever presets are loaded
01:30:37MirthoneistOff topic of Rockbox, but does anybody know if I burn files, normally found on a CD, using Nero, would it boot like a normal CD?
01:30:49rasherif (line[0] == '#') continue;
01:30:53rasherthat's one line, really.
01:31:04XavierGryou want me to add it?
01:31:29XavierGrbecause it reads per line not per character?
01:31:44rasheroh don't know
01:32:04rasherI was just assuming it used the readline function
01:32:19XavierGryes that the function
01:33:40XavierGrhere is the saving routine
01:34:09rasherwell, the comments will just be gone if you change the file
01:34:51rasherIt's mostly just to keep the notes from the wiki
01:37:44XavierGrto keep or it is tiresome to delete all this notes before making the script?
01:40:50MirthoneistYou guys working on Rockbox?
01:43:51rasherXavierGr: Neither, really
01:44:00rasherXavierGr: I could just make the script strip the comments
01:44:14rasherI just thought it'd be a nice touch if the files included the comments
01:44:30XavierGrI don't think it matters so much...
01:45:44 Join JoeBorn [0] (
01:45:58linuxstb_Rasher:How about putting all the comments in a seperate text file in the zip file ?
01:46:07XavierGrhey rasher do you know why the devs chose to make all the files that need loading in boot (font, lang, wps) 20 character long?
01:46:21rasherlinuxstb_: I guess that'd work
01:46:38XavierGrI mean 20 * 5 = 100
01:46:48XavierGr260 * 5 = 1300
01:47:15XavierGrit is not so memory waste!
01:47:17linuxstb_XavierGr: I think it's because So they fit in the 512-byte config sector.
01:47:36XavierGrahh the hole config file is in that sector?
01:48:19XavierGrthat's why they use the set_file and they strip the path and extension...
01:48:48XavierGrit could be super to load and remember any file from every where though.
01:49:07 Quit Mirthoneist (Read error: 104 (Connection reset by peer))
02:15:47 Quit einhirn (Read error: 113 (No route to host))
02:20:16 Join Yono [0] (
02:34:40 Nick TiMiD is now known as TiMiD[zZZ] (n=TiMiD[
02:37:02XavierGrTiMiD[zZZ] how does your patch going?
02:37:10XavierGrIf you are awake?
02:46:52*preglow discovers the exciting world of snes emulators on xbox
02:48:05rasherI guess we just lost a valuable developer then.
02:48:28XavierGrmaybe he will try to port that on rockbox
02:49:40*linuxstb_ summons preglow back to remove mallocs from FLAC
02:49:43preglowanyone know if mono streams have been tested with the playback/dsp engines?
02:49:58linuxstb_I haven't tested any.
02:54:08XavierGrwhats a dsp?
02:54:15XavierGrI keep hearing about it
02:54:28 Quit bagawk ("Leaving")
02:56:58preglowdigital signal processing
02:57:35preglowin rockbox it's the part that's responsble for converting all audio to the common internal format, dithering, resampling, ec
02:59:26XavierGrso even wave needs to be converted?
02:59:52linuxstb_If it isn't 44.1KHz or 16-bit.
03:00:19XavierGr(44.1 or 16) or (44.1 and 16)?
03:00:45linuxstb_not(44.1 and 16)
03:00:56XavierGrah not
03:01:23XavierGrso any other sampling rate doesn't need converting? stange
03:01:58zeif ((khz != 44.1) || (bits != 16)) convert();
03:02:22rasherEverything but 44.1KHz, 16-bit needs converting.
03:02:59XavierGrif(radio_get_status() == 0 || radio_get_status == 1)
03:03:04XavierGris this expression wrong?
03:03:19rasherI assume you mean radio_get_status()
03:03:29rasherBut.. wouldn't that always be true?
03:03:40XavierGrI want to enter the "if" if it is 1 or 0
03:03:43 Quit void ("EPIC4-2.2[730] - oblivion : brb")
03:03:44rasheror is there a 2?
03:03:54XavierGrthere is a 2 too! :p
03:04:04linuxstb_What are the possible range of values? It looks bad to call the same function twice.
03:04:25linuxstb_Test for != 2
03:04:29rasherif (radio_get_status() != 2
03:04:35XavierGryou say then call it first save it to a value and then compare?
03:04:46XavierGroops I am an idiot!
03:05:12XavierGrit is so obvious oh my god, where are my programming skills!
03:05:26linuxstb_But yes - if there was a large range of values, I would save the return value to a variable, and then test that variable.
03:05:40XavierGrthank you very much
03:09:18XavierGrHmm loading and saving to presets works almost perfect.
03:12:16XavierGrI will try to solve that pop whenever someone leaves the radio (while playing) and then returns
03:12:48rasherWasn't that fixed?
03:12:53rasheror was it just identified
03:13:04XavierGrI said it to Linus the other day.
03:13:15XavierGrI think I know what is causing it.
03:13:31XavierGrthe radio_screen() sets the radio on
03:13:41***Saving seen data "./dancer.seen"
03:13:52XavierGrbut every time this function is called the boot of the radio appears too.
03:14:32XavierGrso I will put an if statement where it calls radio_get_status to see if the radio is already started.
03:27:01XavierGrno I was wrong it wasn't that! Though it must be it!
03:27:24rasheramiconn said he knew why it happened
03:29:27XavierGr#if CONFIG_CODEC != SWCODEC
03:29:31XavierGrthat is !iriver?
03:30:31preglowi'm starting to wish i could specify shift factor and sample size independently
03:32:47preglowwavpack passes even 8 bit samples in longs
03:33:41 Join random_man [0] (
03:34:48random_manhello i have a small problem with Compiling Rockbox. yes i know there is a simpalifed post out there and i am reading that but this problem is not in there
03:35:34preglowbed now, later
03:35:35 Quit preglow ("leaving")
03:36:12rasherrandom_man: speak
03:36:21rasherWhat's the problem?
03:39:17random_manwell i did a work around but wondering why it does this
03:39:55random_mani go to download the daily build source code and my screen justs sits there adter i hit enter after it prompts me to input password
03:41:22rasherWhere's this? What prompts you to input password?
03:41:48random_manuhh on cvs screen hold on ill type it
03:42:29rasherAh, cvs login?
03:42:47random_manuhh oh i was sopposed to type the 2 commands. i thougth you were sopposed to type 1 of the 2 commands
03:42:53rasherAh, no
03:43:05rasherfirst the login one (which doesn't produce any output), then the other one
03:43:37random_manbut i am having a problem now
03:43:52rasherWhat's up?
03:44:20XavierGrto get the cvs you need to type into the shell:
03:44:24random_man Now, type daily filename.tar.gz, where filename is the name of the file... obviously
03:44:24XavierGrcvs -z3 co rockbox-devel
03:44:43random_mani type that (yes the actually file name) and it says error
03:44:53rasherWhich error?
03:44:55XavierGrtype what I said and tell us if you can get all the source
03:45:56random_manok just a min
03:46:07 Quit markun ("leaving")
03:47:40random_manok its working now
03:52:54random_mani should read all of this. i have a habit of not reading everything
04:05:42 Join QT [0] (i=as@madwifi/users/area51)
04:18:30 Quit QT_ (Read error: 110 (Connection timed out))
04:20:31XavierGrFound it! Now radio will not pop when entering again!
04:21:48BirdFishIf you don't mind my asking, is it generally accepted that there will be an iaudio x5 port?
04:22:02BirdFishI wasn't sure how everything was going with that whole situation.
04:22:21rasherIt's hard to tell
04:22:34BirdFishSo the project could go down at any time then?
04:22:37rasherThere was only one developer working on this, and we haven't heard from him for close to two months
04:22:47BirdFishOh, I see
04:23:16BirdFishWhat languages are generally used when creating firmware?
04:23:42BirdFishI'm in the process of learning rigth now
04:23:53BirdFishSo I guess I couldn't be of any help there
04:23:54rasherAnd for low-level things, assembly language of whatever the device uses
04:24:00fuzziecombined with liberal amounts of architecture-specific assembler
04:24:06fuzzienot something for beginners :-)
04:24:36XavierGrBirdFish if you learn good VB you are a step closer to programming
04:24:46XavierGrIt is the preogramming mind that matters
04:24:58XavierGrThen you could jump to C far more easily
04:25:25BirdFishSo which would you guys probably recommend? The iRiver H3xx or the IAudio X5? (with and without concerns for Rockbox) Because I'm definitely not happy at all with my H10 :/
04:25:36BirdFishXavierGr: that's what I plan to do
04:26:04XavierGrwell I think that iAudio is a little better though I have never used it.
04:26:15XavierGrI have never used H300 but I have an H100
04:26:58BirdFishI've looked at those also. But for $400 bucks, I think I probably want a color screen (though I'll most likely never use it for videos or pictures)
04:26:58XavierGrIf you ask me I would take and H300 only for the suspicion that it might run rockbox
04:27:21rasherH300 port is quite a lot more certain than X5
04:27:27random_manok i have another probme for compiling my own rockbock could someone help?
04:27:46XavierGrOh and for the VB matter I was just like you. I learned VB6 and then started for a little C.
04:27:49rasherrandom_man: probably, just ask your question - don't ask to ask
04:27:58BirdFishAlso, I was wondering what was so special about rockbox (right now I'm interested because I like to customize my stuff and rebel against companies) But what does Rockbox really do for the player?
04:28:13random_manwhen i go to add a patch i get an error
04:28:17XavierGrNow that I program a little for Rockbox I must say I have extend my knowledge in C
04:28:20 Join amiconn_ [0] (
04:28:31XavierGrhi amiconn!
04:28:41random_mani type Type the command patch -p0 < patchfilename −−binary and i get a missing error
04:28:45XavierGrBirdFish: Many things!
04:28:59 Quit Yono (Read error: 110 (Connection timed out))
04:29:00BirdFishXavierGr: :) Elaborate a little bit please :)
04:29:02rasherrandom_man: the −−binary should go after patch
04:29:24random_mani type patch -p0 < fm-screen-remote.patch −−binary
04:29:35random_manand get.....just a min
04:29:41XavierGrthere you can see many differences
04:29:57random_manmissing header for unified diff at line 3 of patch cant fid file to patch at input line 3
04:30:13XavierGrtry another -p lets say -p1
04:30:23random_mandid that and even 5
04:30:45rasher−−binary shouldn't go after the <
04:30:56BirdFishBattery Lifetime is definitely a plus. You'd say it was very much so noticeable, or just statistics?
04:31:00XavierGrdoes it need the binary
04:31:04XavierGrtry without it
04:31:29random_mani dont know
04:31:59XavierGrOn the archos it is, but I am not sure for iriver
04:32:20random_manso type i type patch -p0 < fm-screen-remote.patch
04:32:20XavierGrit is sure the same though
04:32:51XavierGrI never tend to watch the battery I have changed it with one for 24 hours of music
04:33:28XavierGramiconn: I just solved the poping of the radio when someone re-enters the FM screen.
04:33:38XavierGrI suppose this error is iriver specific?
04:33:50XavierGror can the archos leave ths screen too?
04:34:26rasherI'm sure it acn
04:34:51BirdFishReplayGain is nice too
04:36:28 Quit random_man ("CGI:IRC")
04:37:52 Quit Nibbler (Read error: 110 (Connection timed out))
04:38:51XavierGrOk that was a productive day. Time to sleep. It is 5.40 am here!!!
04:39:58 Quit XavierGr ()
04:43:14 Join k-man [0] (n=jason@unaffiliated/k-man)
04:45:06 Quit amiconn (Read error: 110 (Connection timed out))
04:45:07 Nick amiconn_ is now known as amiconn (
04:59:11 Join Paul_the_Nerd [0] (
05:13:45***Saving seen data "./dancer.seen"
05:32:48 Quit rasher ("leaving")
05:39:16Paul_the_NerdDoes the tag database have to rebuild entirely every time, or does it have a switch of some sort that lets you update it?
05:46:53Paul_the_NerdHrm... the TagDatabase generated with the .jar is... well... very incorrect.
05:48:42k-manis there a list of known hardware rockbox runs on somewhere?
05:50:10Paul_the_NerdNot really just a list
05:50:37Paul_the_NerdBut the information seems mostly there
06:15:34k-manso whats the cheapest hardware one can run rockbox on?
06:17:00 Quit BirdFish (Read error: 110 (Connection timed out))
06:19:53Paul_the_NerdThat's a tough one, because generally speaking you're going to be looking for it on ebay.
06:20:24Paul_the_NerdI mean, what do you want it for?
06:24:03k-manto record live audio
06:25:51k-manand i wouldn't be recording a huge amount of audio either
06:25:58k-manjust 5-10 minutes max
06:30:04 Join LinusN [0] (
06:34:14Paul_the_NerdI seem to recall hearing the OndioFM supports recording.
06:34:54Paul_the_NerdLinus, you're more likely to know this
06:35:03Paul_the_NerdCheapest Rockbox platform that supports recording?
06:35:12*Paul_the_Nerd has no idea about Archos prices.
06:35:39LinusNi think ondio might be cheaper, but they are not as common
06:35:51LinusNthe plain Recorder 20 is probably the cheapest
06:36:07LinusNsearch ebay
06:37:24k-manand that has mic?
06:37:33k-manthats the archos recorder 20?
06:38:02k-mandamn, no one sells the archos recorder in Australia
06:38:06 Join BirdFish [0] (n=bradbox8@
06:38:21LinusNnoone sells it anywhere
06:39:07LinusNyou have to buy used ones on places like ebay
06:40:32k-mani mean, no none sells used ones on ebay in australia
06:40:41k-mani'll have to buy it internationaly
06:41:17k-manare you the Linus who does the development?
06:42:12LinusNyes, i am the linus, but i'm far from the only one who develops rockbox
06:59:16 Join amiconn_ [0] (
07:13:46***Saving seen data "./dancer.seen"
07:16:41 Quit amiconn (Read error: 110 (Connection timed out))
07:16:41 Nick amiconn_ is now known as amiconn (
07:30:07 Quit webguest95 ("CGI:IRC (Ping timeout)")
07:42:39 Join Paul_the_Nerd_ [0] (
07:56:33 Join B4gder [0] (n=daniel@
07:57:23B4gderguten morgen
07:58:19amiconnLinusN: Yesterday I tested how much of a slowdown this format conversion in lcd_blit() will cause on H1x0
07:58:30LinusNtell me
07:59:46 Quit Paul_the_Nerd (Read error: 110 (Connection timed out))
08:01:16amiconnIt was only around 4% at all 3 clock freqs
08:01:42amiconn...but it depends on whether the stack lies in iram or not
08:02:14 Nick Paul_the_Nerd_ is now known as Paul_The_Nerd (
08:02:34amiconnLinusN: For an isr, you can't tell which stack it uses, correct?
08:03:21LinusNyou can of course change the stack pointer temporarily
08:03:27LinusNa special isr stack
08:03:57amiconnThe conversion needs 2 buffers of 160 bytes (LCD_WIDTH)
08:04:05LinusNwith all the weird problems that goes with that (nested irq's etc)
08:04:46amiconnThere are 2 potential issues with that
08:05:14amiconn(1) Depending which stack it uses, it will run slower or faster
08:05:35amiconnmain, codec and voice_codec stacks are in iram, the others are not
08:05:55amiconn(2) The stack it uses may overflow...
08:08:57 Join ashridah [0] (
08:13:09amiconnI could use static iram buffers, but I'm hesitant to set aside 320 bytes of iram just for lcd_blit()
08:15:41LinusNno, that would perhaps not be economical
08:16:30LinusNon the other hand, 320 bytes out of 96k of iram isn't a big deal
08:17:07LinusNwe could probably shrink one of the stacks by 320 bytes without any problem
08:24:00amiconnIt's 320 out of 48K
08:24:54amiconnI'll leave it as it is for now, it's simple to change that later
08:26:11B4gderthey do have wild USB ideas in that Neuros camp ;-)
08:26:32Vladomanbut Ethernet I think would be nice
08:26:59B4gderyes, lots of things could be nice
08:27:29Vladomandepends on whether it will be "pure" rockbox or running on top of Linux
08:27:55B4gderwhen considering ethernet etc, I would guess using linux will start making more sense
08:29:00B4gderI run linux 2.6.12 on a ARM920 at work
08:29:17Vladomanwhat chip?
08:29:33Vladomanwhat does it do :-)
08:29:33B4gderAtmel thing
08:29:57B4gderwe build an RFID "reader"
08:30:33Vladomanno good on me, I wear my tin foil hat :-)
08:30:59*B4gder points his scanner in Vladoman's direction
08:31:15 Quit k-man ("Client exiting")
08:31:39Vladomanif I knew how the IRC works i'd say something like "* Vladomann ducks away" :-(
08:31:51LinusN /me bla bla
08:31:59*LinusN bla bla
08:32:07*Vladoman understands
08:32:08B4gderanyway, a full-fledged kernel is around 950Kb compressed
08:41:31VladomanOops, this NetChip2272 is a "peripheral" only, not a host, right?
08:41:37Vladomanso no ehternet :-(
08:41:56B4gderwell, they were previously talking about a separate host chip
08:42:06B4gderbut I'm not sure where they stand on that now
08:42:24Vladomanquote: The main reason we chose the NET2272 route is to enable other than direct-to-disk applications in high speed. Such applications include Janus (MTP), Ethernet-over-USB (RNDIS), fast serial as well as mass storage.
08:42:57B4gderyou found any details on that chip?
08:43:04Vladomanoh, I get it, ethernet works by simulating a USB2Ethernet device, not by connecting one!
08:43:29Vladomangoogle did actually :-)
08:43:45B4gderthat google guy sure is friendly
08:48:13B4gder"The NET2272 supports Microsoft?s Media Transfer Protocol
08:48:13B4gder(MTP), which allows playback of WMA and WMV files with
08:48:13B4gderDigital Rights Management (DRM)."
08:49:14LinusNwhat a blessing
08:49:26B4gderhow can the USB chip do this?
08:49:38B4gderbut ok, I now nada about MTP
08:54:20 Join einhirn [0] (
08:57:48 Join Vld0man [0] (
09:00:05 Quit Vladoman (Read error: 110 (Connection timed out))
09:01:15 Nick Vld0man is now known as Vladoman (
09:02:58VladomanThe USB chip does not do any DRM, it just allows to run any protocoll over it
09:03:15Vladomanso, you can do MTP + Janus, not just MSC
09:04:34B4gderright, I just found it weird that the chip datasheet says it supports MTP
09:04:54VladomanI guess they want that keyword to appear in the google index :-)
09:05:07B4gderyes, that's very probable
09:13:48***Saving seen data "./dancer.seen"
09:15:05Paul_The_NerdThis is a purely hypothetical question. (I'm reading a few articles on Hymn/Playfair and it was just an odd thought I have.) Wouldn't it be possible to add support for encrypted m4ps on Rockbox? (At least for windows users.) You could write a utility that gets the key info and stores it into a file, which could then be copied to the player, then when an encrypted file runs, it could use a...
09:15:06Paul_The_Nerd...version of those algorithms to decode the file for playing, without writing a decrypted version. Again, not a feature request, since I'm not a big fan of DRM at all, but just sorta an odd thought I had.
09:16:15linuxstb_Paul_The_Nerd: There's no point. The user can (and should!) run hymn on their PC to permanently decrypt the files once.
09:16:27Paul_The_NerdAs I said, just a random hypothetical
09:16:41Paul_The_NerdI avoud iTunes like the plague personally.
09:16:53linuxstb_I guess it's possible in theory, but IMO, there is no need to.
09:17:04*Paul_The_Nerd agrees on the "no need to."
09:17:50Paul_The_NerdIt'd definitely qualify as "feature bloat" anyway.
09:17:56linuxstb_I did "buy" one file from the iTunes store the other day - partly for research purposes to see how it (and hymn) worked, and partly because it was an exclusive-to-itunes-store track by one of my favourite bands.
09:19:04Paul_The_NerdHow much do lossless-format songs cost on iTunes? Or are they even available?
09:19:44linuxstb_No - 128kbps AAC only. In the UK it is 79p (about 1 EUR) per track.
09:21:07linuxstb_The iTunes store makes it scarily easy to buy songs. You register, enter your credit card details, and then just click on the "buy now" button next to a track. You are then asked to confirm the purchase, and the track is then downloaded into your iTunes library.
09:21:14Paul_The_NerdSee, I just can't stand not having a lossless version of my music
09:21:30Paul_The_NerdI'm even very wary about buying CDs now because of all these issues people have had with the way a lot of newer albums are mastered.
09:21:45linuxstb_The scary thing is that you can disable the "confirm purchase" dialog - so pressing the "buy now" is equivalent to a download link.
09:23:08Paul_The_NerdI can see a lot of people easily spending a lot of money there
09:23:12linuxstb_And all of this is within iTunes, not your web browser.
09:25:23Paul_The_NerdIf you have confirm disabled, you could accidentally spend money.. that kinda sucks.
09:26:16B4gderyes, it sucks your money ;-)
09:26:19linuxstb_Exactly. And the worse thing is that you can only download each file once. If you accidentally delete the file, you have to buy it again.
09:26:45linuxstb_Apple's support site has a useful guide to backups as part of the iTunes Store help.
09:26:50Paul_The_NerdApparently, that's imposed by the software though.
09:27:08Paul_The_NerdDVDJon has an alternative iTunes store interface.
09:27:19Paul_The_NerdThat... doesn't encrypt them after downloading (since apparently that's client side)
09:27:27Paul_The_NerdAnd... seems to let you redownload purchases.
09:27:34Paul_The_NerdAt least... so this page seems to claim.
09:27:51linuxstb_That seems a big loophole in security if the encryption is done on the client.
09:28:00Paul_The_NerdI would agree.
09:28:04linuxstb_But it's been hacked anyway....
09:28:14Paul_The_NerdBut then, storing the decryption key permanently on the client is just as big, I'd say. they have a deal with amazon :-)
09:28:46Paul_The_NerdOf course LinusN.
09:28:50linuxstb_How can you have DRM if the decryption key isn't permanently on the client?
09:28:57*LinusN wants to patent 2, 3, and 4-click shopping
09:29:13*linuxstb_ wants to patent 0-click shopping
09:29:17LinusNand infinite-click surfing
09:29:32B4gderyou can buy my "shop-before-you-click" patent
09:30:37Paul_The_NerdI wonder if you could patent "User points to something in the store, which the clerk then retrieves"
09:31:24linuxstb_Anyway, back to Rockbox, I've almost got a bootloader (compiled using the Rockbox build system and linked to rockbox.a) running on the iPod.
09:33:01linuxstb_Once I squash the last bug, I can then start on the drivers. LCD and buttons should be relatively easy, and then the ATA.
09:33:20B4gderyou on top of your lcd then?
09:33:25*Paul_The_Nerd points to his talking about the ITMS DRM decryption idea.
09:33:41Paul_The_NerdYou could actually have a method to sync with iTunes on the iPod port with something like that, couldn't you?
09:33:55B4gderI mean, wasn't that still not really figured-out?
09:33:57linuxstb_B4gder: Not exactly. But I've worked out how I can write to the part of the display the Apple flash firmware uses to display the Apple logo during startup.
09:34:00Paul_The_NerdAnd maybe convince "normal" itunes users that a better way exists..
09:34:28linuxstb_The problem with the new LCD is that the "set update region" commands are not the same. But the protocol for actually sending data to the LCD is the same.
09:34:41B4gderI see
09:34:50Paul_The_NerdThat seems kinda odd.
09:34:51linuxstb_So I can use the update region set by Apple's firmware - a 78x100 window in the middle of the display :)
09:35:23linuxstb_Hopefully the ipodlinux guys will sort it out. In the meantime, I can at least continue with the other parts of the port.
09:35:46B4gderyou in contact with any ipl guys during your work on this?
09:36:01linuxstb_I've had some brief chats on #ipodlinux.
09:36:31B4gderok, I was just curious
09:36:41 Join k-man [0] (n=jason@unaffiliated/k-man)
09:36:51B4gderI'm puzzled on their lack of devoloper community
09:36:57linuxstb_But I haven't discussed Rockbox - just specific questions about the hardware.
09:37:33linuxstb_I think it's just a lack of developers. "leachbj" seems to be the only kernel hacker.
09:37:41k-manB4gder, who's lack of developer community?
09:37:53B4gderk-man: ipodlinux
09:38:22linuxstb_Putting Linux into the equation makes it even harder to find low-level hackers who can make things happen.
09:38:52B4gderlinux is quite lines of source to figure out ;-)
09:38:59B4gder"quite a few"
09:41:24linuxstb_How far would you want me to get before I start committing patches? I'm in no rush, just curious.
09:41:49B4gderin my view, you could commit already
09:41:57linuxstb_I want someone else to take the target #13 slot first :)
09:42:13B4gderI don't see that happen anytime soon
09:42:35B4gderyou could easily take 20 or so ;-)
09:42:49linuxstb_OK, once I get the bootloader working, I'll clean things up and start committing.
09:43:16B4gderI've put my order on hold for a little while longer, due to personal stuff and general work load
09:43:23B4gderbut I hope to get one later on
09:43:57linuxstb_No problem. I'm happy to keep going.
09:44:29k-manone what?
09:44:53B4gderk-man: linuxstb_ works on rockbox for ipod
09:45:44k-mani see
09:53:03k-mandid i read somewhere about a project for an open hardware mp3/media player?
09:57:29k-manthat might have been it
09:58:15k-manoh geez, its not cheap
09:58:20k-manoh well
09:58:56B4gderwell, the current discussions are for their upcoming models
09:59:07B4gderbut I don't think they'll be cheap either ;-)
10:02:02*Paul_The_Nerd kinda wishes more devices were like computers: There's a standard, and I can easily replace any parts I feel like.
10:02:18Paul_The_NerdI mean, there's clearly not just one standard among personal computers
10:02:22Paul_The_NerdBut there's few enough.
10:04:55zePaul_The_Nerd: mp3 players are like the opposite extreme... they seem to be generally considered unservicable for the most part, and afaict standard waranty/service-plan practice is just replacement
10:05:16Paul_The_NerdI know, it drives me nuts.
10:05:27Paul_The_NerdAt least with *some* of them you can effectively replace an HD.
10:05:42B4gderand they sell just like TVs or VCRs, nobody cares about the SW in them
10:07:55Paul_The_NerdI don't care about it in my TV, because there's really not much going on in there.
10:08:07Paul_The_NerdBut say, my set top cable box... I'd like to have some more freedom there. =/
10:08:24HClwhats a set top cable box?
10:10:20Paul_The_NerdCable Jack on wall -> Cable box (upon which channel guide information is parsed, and which you use to change channels etc) -> Television.
10:11:21Paul_The_NerdIt also decrypts the signal.
10:19:18 Quit linuxstb_ ("Leaving")
10:19:19HClwhat would you want to modify about it?
10:20:11Paul_The_NerdWell, the program guide is fairly limited.
10:20:21Paul_The_NerdIt presents a list, and I can flag a program to remind me to change channels
10:21:44Paul_The_NerdIt'd be nice to have some customization options for the whole thing. Maybe the ability to remove channels that I will never watch from the main list, or at least list them after my favorites, as opposed to only listing in numerical order.
10:22:14Paul_The_NerdThings that more complex boxes do, but mine from a hardware perspective probably isn't incapable of.
10:22:20 Join _FireFly_ [0] (
10:22:54Paul_The_NerdI don't really feel the need for a PVR system. But it'd be nice to have some flexibility within the bounds of being able to tell what shows are on.
10:22:58Paul_The_Nerd'alo FireFly
10:25:54 Join Nibbler [0] (
10:28:13HClaha o-o
10:28:23HClwell thats more luxery than me, i suppose :p
10:29:39Paul_The_NerdI suppose so.
10:31:33k-manso will rockbox ever be able to run on the small flash iriver players?
10:32:06Paul_The_NerdFrom what I understand such things aren't necessarily planned. They depend on the right people having the interest in it.
10:32:31Paul_The_NerdAs it is, it looks like they've made some headway into decrypting the firmware, but it's not even really in the realm of "likely" that there'll be a build for it at some point.
10:32:36*Paul_The_Nerd doesn't know *that* much though.
10:33:07B4gderk-man: if enough (skilled) people work on it, then sure why not
10:33:50k-mani wish i had the time to work on something like that
10:33:53k-manoh well
10:36:02Paul_The_NerdI'll be happy when pretty much any software codec flash player is supported.
10:37:49LinusNthe problem is that most flash players are built on proprietary chipsets
10:40:37Paul_The_NerdThat's why I'm kinda excited by the iPod port even though the HD players don't interest me much.
10:43:11 Join webguest95 [0] (
10:45:15amiconnLinusN: I did some more measurements. (1) With intermediate buffers in iram, I'm down to 2% slowdown :)
10:45:43amiconn(2) Intermediate buffers in dram mean ~ 75% slowdown....
10:51:23 Join ender` [0] (i=ychat@
10:58:25LinusNamiconn: wow!
11:00:34 Join markun [0] (
11:03:58 Join ep0ch [0] (n=ep0ch@
11:05:26ep0chi'm really really really ****ed off, me thinks my IHP's HD has died :( it is a sad sad day. boo hoo.
11:06:33ep0chdoes anyone have any spare 20 gig drives i can take off you?
11:08:01 Quit ashridah (Read error: 110 (Connection timed out))
11:08:57Paul_The_NerdI'll have one as soon as the MK4007GAL is purchaseable.
11:09:15ep0choh hang on, i've wacked the player a few times and its booted up
11:09:26ep0chtime to backup my tunes while i can
11:10:00 Quit ender` (" HydraIRC -> <- IRC with a difference")
11:10:05ep0ch MK4007GAL? that a single platter 40gig?
11:10:05 Join ender` [0] (i=ychat@
11:10:41B4gderan advice: never use a DAP as the primary storage of anything
11:10:57Paul_The_NerdSingle platter 40
11:11:05Paul_The_NerdIt's out in some players now, but only the Toshiba brand ones, I hear.
11:11:30Paul_The_NerdToshibal claimed that the perpendicular technology would allow *at least* double data density, so that implied a 60 gig at *some* point.
11:11:50ep0chany idea when the 40 gig might be available?
11:11:55Paul_The_NerdNo clue
11:12:17Paul_The_NerdIt exists physically, but the only option for acquiring one as far as I know is purchasing certain other players and gutting them. Hardly cost effective.
11:12:25ep0chno no no CRC error, grrr
11:13:05 Quit ender` (Client Quit)
11:13:52***Saving seen data "./dancer.seen"
11:14:36 Join ender` [0] (i=ychat@
11:15:27 Join ashridah [0] (
11:20:11ep0chhmm the 30gb one is tempting at £81
11:20:23ep0choh no 95 inc vat :(
11:20:37Paul_The_NerdI'm waiting for at least 40gb.
11:20:45ep0chi hear ya :)
11:20:50Paul_The_NerdI'm going to reserve 10gb for applications anyway.
11:21:09ep0chyeah? what kind of applications?
11:21:18ep0chlike, Word?
11:21:25Paul_The_NerdA variety of things. Most of them are relatively small. Miranda, stuff like that.
11:21:30Paul_The_NerdWord? Ew... openoffice
11:21:45Paul_The_NerdInstallation binaries for Linux, Windows and OSX.
11:21:53ep0chhehe, yeah sorry Word was the first thing that came to mind
11:22:05Paul_The_NerdIDEs and compilers of various sorts.
11:22:09ep0chopenoffice is very good
11:22:14Paul_The_NerdBasically, all the tools that I like to have readily available.
11:22:27Paul_The_Nerd10gb is way too much for just that
11:22:39Paul_The_NerdBut it's also for backups of my writing, projects, etc.
11:22:44Paul_The_NerdIt's more of "10gb for non-music data"
11:22:58 Join linuxstb [0] (n=linuxstb@
11:23:16ep0chi use a 1gb pen drive for that
11:24:02 Join Lynx_ [0] (
11:24:05Paul_The_NerdI think my current data size that I'd want is about 1.3-1.4gb
11:24:13Paul_The_NerdBut that's like, bare minimum
11:24:18Paul_The_NerdThat's data, not applications or anything else.
11:25:59ep0chok I think just one area of my HD is knackered, besides fscking it, is there anything else i could do to help repair it?
11:26:43ep0chclean format comes to mind
11:27:54*Paul_The_Nerd shrugs.
11:28:08Paul_The_NerdThere's a whole world of disk scanning and "repairing" utilities out there, I imagine.
11:28:23*Paul_The_Nerd doesn't have a lot of experience with HD failure of the non "it's melting" sort.
11:28:35Paul_The_NerdWhen my hardware fails, it's very good at doing it in generally spectacular ways.
11:34:29_FireFly_now my remote-patch isn't dependig on patches to applied first :)
11:35:31_FireFly_and it inclueds all additional remote-patches which i have made after it
11:43:03*Paul_The_Nerd cheers
11:51:50_FireFly_now i think i can it post to the tracker
11:58:33_FireFly_bagder do you here??
12:13:21 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
12:13:39_FireFly_ups there was an bug in the fm-screen, remote control of volume didn't work
12:18:21 Join preglow [0] (
12:26:04_FireFly_hi preglow
12:27:49_FireFly_bagder ??
12:35:05 Join XavierGr [0] (
12:37:04_FireFly_hi XavierGr
12:38:05B4gderI'm here
12:38:06ep0choh GCC 4.0.2 is out
12:39:46_FireFly_B4gder you want say me an example where i can remove safely some of the #ifdefs in my updated version can be found in my forums-thread on
12:39:54XavierGrHow often do they update it!?
12:40:17XavierGrI mean GCC!
12:40:25B4gder_FireFly_: for example, you can make a function of the "too big to fit" text
12:40:34ep0chno idea :)
12:40:41B4gderand have that function defined to nothing on models without remote
12:41:52B4gderalthough of course the best fix would be that all screens would fit
12:41:58B4gderjust allow scrolling or similar
12:42:18B4gderhowdy preglow
12:42:50B4gderI'm asking you to use your imagination, I haven't solved all the issues in my head
12:43:12B4gderin general we should strive to decrease the amount of #ifdefs in the code
12:45:23_FireFly_the text "too big to fit" is currently only an place-older in the debug-menu
12:46:36XavierGrHave you thought that keeping the same code for all platforms will make the code humongous in future?
12:46:44_FireFly_until someone makes the output fit on the remote
12:46:48B4gderXavierGr: ?
12:47:39XavierGrI mean I know the advantages of keeping the same code for all platforms, but in future when rockbox moves to many targets it will be a hell with #ifdefs!
12:47:48B4gderXavierGr: no
12:47:55B4gderjust as I explaine
12:48:03B4gderwe should decrease the amount
12:48:18LinusNhaving all #ifdefs in the code will be a problem, yes
12:48:20B4gderby being clever
12:48:30LinusNbut there are other ways
12:48:51XavierGrI think that the low level code must be programed in a such clever way
12:49:20B4gderand that's what I am saying
12:49:25XavierGrSo that when e.g tree.c comes to play, there will be no if defs
12:49:28B4gderfor example regarding the remote patch
12:49:46B4gderwell zero ifdefs might not be possible
12:49:53B4gderbut that's the aim of course
12:50:03XavierGryeah I know screen sizes and other factors
12:50:28XavierGrBut some parts of the code already contain a lot of ifdefs
12:50:57 Join Febs [0] (
12:50:57B4gderyou feel like fixing? ;-)
12:51:16XavierGrNonono I am not so advanced in this, yet... :p
12:51:39B4gderA general approach in Rockbox: do not do things for the future now, wait until we face the probem for real
12:52:13B4gderwe had many people saying we couldn't ever make rockbox on anything except archos
12:52:18B4gderdue to how the code worked
12:53:00XavierGrLazy type approach but yeah I know what you mean.
12:53:09B4gdernot lazy, reality-bound
12:53:21XavierGrYou could say that too.
12:53:42B4gderit actually prevents us from doing stupid things now
12:53:45B4gderthat won't be used
12:53:58_FireFly_with the widget system on which timid currently works, it would be easier wouldn't be ?
12:54:21XavierGrOh Linus I nearly finished the load/save preset patch, and I found the cause of the pop bug when reentering FM screen. It should no longer pop now!
12:55:47_FireFly_and i said some days ago that my patch is only a kind of place-holder until the widget system is ready to get into rockbox
12:56:09XavierGrAlso I increased presets to 64. My city Heraklion in Crete (Greece) has 48 stations and has 200.000 citizens so...
12:58:18XavierGrNo I will set the Preset mode. Because searching presets is not so intuitive now. And Myabe I will add a Make new preset option, because if someone wants to make a new clean preset he will have to delete every preset and then save it to another file (spoiling the first file)
12:58:24XavierGrno = now
12:59:56ep0chdo you mean left/right will cycle through presets instead of inc/decrementing frequency?
13:00:12XavierGrmake a choice for it
13:00:57linuxstbDoes anyone have what is suggested to be the H300 LCD datasheet - e789r_101.pdf ?
13:01:02ep0chany chance it can be sorted by 'name' instead of frequency?
13:01:06ep0ch(fm presets)
13:01:26linuxstbIt's missing from here:
13:01:50XavierGrhmmm currently there is no sorting system in either way.
13:01:59XavierGrname or frequency
13:02:13ep0chyeah i had to edit the file by hand :(
13:03:10B4gderlinuxstb: try this
13:03:15ep0chXavierGr: are you also making the buttons consistant with the rest of rockbox?
13:03:30XavierGrit could be tricky to implement. Also a new option would be needed because someone might wanted the way he wants.
13:03:52XavierGrI haven't changed a single button yet.
13:04:21ep0chi have an outrageous idea, regarding rockbox and the radio...
13:05:28ep0chcould we have a file extension that calls the radio? so I could have a directory in root called 'Radio' and in there are radio files that contain the name and frequency of the stations
13:05:49XavierGryes it set to .fmr
13:05:57linuxstbB4gder, That contains a link back to the Wiki page :) That PDF doesn't seem to be in the Forum.
13:06:01ep0chahh ok
13:06:04B4gderhehe, sorry
13:06:26XavierGrBUT if you want a preset to be remembered after a boot you will have to store your presets to /.rockbox/presets/
13:06:37XavierGrelse in the next boot up it will be forgotten.
13:07:19XavierGrBut yeah currently when selecting a fmr file the radio will pop and load the preset. (and it will remember it untill you shut down)
13:07:55ep0chso i can have 20 fmr files each with an individual station
13:08:18ep0chbut left and right would not cycle through each fmr file?
13:08:32ep0chlike an mp3
13:08:37XavierGrno I mean cycling the presets inside the preset file
13:09:03XavierGryou have a filepreset with 40 stations and left/right cycles between those stations.
13:09:35XavierGrI dont think what you mention as a good idea.
13:09:38ep0chhmm, any chance of if only one station is in the .fmr file then it will look at the next file?
13:09:57XavierGrehmm why have one station in a file?
13:10:08XavierGrIt could get too complex to do that
13:10:22XavierGrJust select a different file.
13:10:39XavierGrThat's the normal and most intuitive way I have thought.
13:10:40ep0chbasically, i don't like the idea of the radio being treated in a special way - having to go to the menu and selected FM Radio.
13:10:53ep0chit would be nice (for me) if the radio was treated like any other file
13:10:58ep0chbut hey, that's just me :)
13:11:23ep0chdo you understand what i mean?
13:11:23B4gderyou mean any other music file
13:11:26XavierGrYou don't have to do that. You can select your favorite preset file (inside or outside /.rockbox/presets) and the radio will load and pop up
13:12:03XavierGrso there it is just selecting an mp3
13:12:29XavierGryeah if you press right it will go to another station not another preset file
13:12:44XavierGrbut that's how a playlist works remember?
13:12:53XavierGran mp3 file is a single mp3.
13:13:09ep0chif i have 5 radio stations (FM1, FM2, etc), i want FM2, i would have to go to the .fmr file, which would load up FM1, and then navigate to FM2?
13:13:10XavierGrBut a playlist or a fmr file is a holder of many mp3s, stations
13:13:28ep0chno worries, i can live without it
13:13:38ep0chi'll shut up :D
13:13:50XavierGrno I think that the radio remembers your last station.
13:13:55***Saving seen data "./dancer.seen"
13:14:16ep0chwhat if next time i want FM5?
13:14:41XavierGrI think that you ask too much!
13:14:59XavierGrIhave never heard such approach for the radio before.
13:15:00ep0chyou think?
13:15:33ep0chexactly, i just think it would be consistent to have the fmr files treated like any other audio file
13:15:33XavierGrSo you say enable 3 modes? Frequency, preset and filepreset? Dont you think it is just too much?
13:16:04XavierGrbut an fmr file is to hold many stations, not one.
13:16:45XavierGrOthers: Which is your opinion?
13:18:02linuxstbA presets file is something you change very infrequently. I think a simple load option on the menu is sufficient,
13:18:31 Join tucoz [0] (
13:18:36XavierGrand save to be able to make one on the go.
13:18:47XavierGr(As it is now on my patch)
13:18:54linuxstbYes, that would be useful.
13:19:24tucozHi, sorry for jumping in to this. Was reading the logs. But imho, I think linuxstb idea of having a base menu will take care of some of the navigation issues in rockbox
13:19:27B4gderbtw, I love your Sudoku-plugin linuxstb
13:19:42linuxstbB4gder, Thanks :).
13:19:49tucozme too, it is really fun
13:20:01ep0chbase menu? is that like the feature request i made a while back?
13:20:01LinusNit rocks
13:20:14linuxstbtucoz, I'm starting to be less convinced about having a base menu, but I'm going to continue implementing a demo build.
13:20:49linuxstbB4gder, Any luck getting your friend's Sudoku site to generate .ss files?
13:20:55B4gdernot yet
13:21:05LinusNwe're threatening him every day
13:21:12LinusNbut he's stubborn
13:21:23tucozBut after all, it is a dap, and not a pc. I think it is a quite nice idea. Imagine having the base functions plus settings in that menu
13:21:25B4gderhe has a really nice "difficulty" level too
13:21:31ep0chlinuxtsb: same as:
13:21:38XavierGrGuys: Do you think that it would be would to: In the fmr file add a last line (after the sations) that will hold the last preset position of the user?
13:21:42linuxstbSomeone should get the Rockbox Sudoku generating random games. I've got other things I would prefer to do at the moment.
13:22:03XavierGrThat way when loading different fmr files it would jump automatically to the last preset position.
13:22:08tucozlike, browse files, record, fm-radio, settings in the base menu
13:22:22linuxstbep0ch: Yes.
13:23:58tucozI posted a link to a c implementation of a sudoku program with a generator in the PluginSudoku page on the wiki
13:24:45tucozI am not sure if it is easy to just steal that function though. Haven't really looked at the code yet. Just a glimpse
13:24:49 Join webguest50 [0] (
13:25:24webguest50I'm trying to save a cfg file with upside down screen, but it wont let me create it
13:25:36webguest50or save it rather
13:26:28webguest50anyone else have this problem?, or is it just me
13:27:21LinusNshould work
13:27:25tucozlinuxstb: what makes you rethink the idea of a base menu?
13:27:36LinusNwebguest50: which model?
13:27:57 Join webguest54 [0] (
13:27:58 Quit webguest50 (Client Quit)
13:28:16webguest54uggh sorry linus, its a H140
13:29:17LinusNwebguest54: i just tried, it works for me
13:29:43ep0chis it just me or is the 'quick menu' a bit annoying if you accidently hit 'shuffle', it shuffles immediately, and so you have to change it back and it unshuffles immediately. IMHO actions should take place when you exit the quick menu.
13:29:43 Quit webguest54 (Client Quit)
13:29:53 Join webguest54 [0] (
13:30:28webguest54thanks, I'll try again
13:30:47tucozep0ch: maybe not when you exit, but if that is a problem, perhaps a confirm should be in place
13:31:03tucozI do not use shuffle, so what do I know ;)
13:31:24ep0chi dont either, but i just hit it accidently every so often
13:31:25XavierGrLike tucoz, I don't shuffle either
13:33:19tucozMight be a good idea to have stuff like shuffle with a confirm. The other options just sets a flag, right?
13:34:04ep0chthat would slow down the quickness of the quick menu, why not just perform that action on exit?
13:34:05webguest54is there any way I can display cfg files in root without having show files set to all ?
13:34:35tucozwebguest54: not that I know of
13:34:49webguest54hmm, thanks
13:35:14tucozif the configs are under /.rockbox/*.cfg then you can use browse config in the settings menu
13:35:31tucoz...and call the viewer I think
13:35:35ep0chcan you associate the .cfg in viewers.config?
13:35:49linuxstbtucoz: My initial idea was to simplify the Rockbox UI by making everything start from the main menu. But I'm thinking that it will become too cumbersome to have to go through the menu.
13:35:49webguest54yep, just trying to figure a short way to access cfg files
13:36:10linuxstbBut that can probably be solved with shortcuts. I'll see how my prototype develops.
13:36:14ep0chlinuxstb: let the user set a default screen?
13:36:22tucozlinuxstb: ok, I'll look forward to testing it out.
13:37:02linuxstbI want something that is easy for a new user to use (i.e. every feature available in nice menus), but also something that is quick to navigate for advanced users.
13:37:08tucozIn my mind it sound better though. More intuitive at least
13:37:32tucoznot "hmm, what button should I use for that". At least for beginners
13:37:50webguest54why nit put quick access to cfg files in the 12 o'clock quick menu position ?
13:39:03LinusNwebguest54: everybody wants their favourite feature in the quick menu
13:39:05ep0chI believe Cassandra had something else in mind for the 12 oclock
13:39:05 Quit tucoz ("CGI:IRC (EOF)")
13:39:19ep0chheh, i wanted crossfade there :)
13:39:23webguest54heh, I guess so,
13:39:56webguest54an option for an option to have it there :)
13:42:26amiconnwebguest54: .cfg files *are* visible in the root if you have set 'show files' to 'supported' (or 'all')
13:43:05amiconnThat's how I manage my various configurations - much faster than 'Show config files' in the mneu
13:43:47*webguest54 tries
13:43:48markunlinuxstb: Instead of the root menu you could also have a menu that pops up (like the one in rockboy) to do the same.
13:44:55linuxstbmarkun: I think Rockbox needs both. My idea is that the root menu is the "home" - i.e. you always go back to it when leaving a screen. So the WPS and browser screens for example are "on top of" the menu. Not the other way around.
13:45:02Slasheriamiconn: hmm, do you think the stack overflow error message is still working or was it that works only for codecs/plugins?
13:45:24markunlinuxstb: I know what your idea is. It was just another suggestion.
13:45:49linuxstbOK, but that's not what I want :).
13:46:06amiconnSlasheri: Stack overflow detection should work for all threads, but of course it can't be bullet-proof
13:46:47amiconnStack overflow means something gets overwritten, and rockbox might crash before being able to detect the overflow
13:46:54Slasheriamiconn: i had a too deep recursion in the code (now completely removed) a day ago, but the system only crashed or displayed I00 or something like that (can't remember) but not the stkoflw
13:46:56markunamiconn: Can you explain how stereo width works for values greater than 100?
13:47:29ep0chlinuxstb: when rockbox starts up it will go to your root menu? Do you plan on adding an option so that on startup the user to default to 'browser mode', and from there he can left click to go to the root menu?
13:48:26Slasheriamiconn: ah, so it works only if the process can yield before it has used too much of the stack?
13:48:43linuxstbep0ch: What I'm planning is just to develop a prototype that makes the main menu the "root" of Rockbox. What actually gets implemented will be the results of the subsequent discussions.
13:48:55ep0ch:) ok
13:49:25webguest54ah supported works well, what is the purpose of the Rolo" feature?
13:51:10LinusNto manually load and start a firmware binary
13:52:06webguest54thanks, so there's no harm if I accidently acitvate it like I just have
13:53:49markunSlasheri: Do you have a new version of the dir chache patch?
13:54:19 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
13:54:24 Join DangerousDan [0] (
13:54:36Slasherimarkun: i will put the new version online soon (maybe today or tomorrow)
13:54:51Slasheriit's much better than the last one
13:55:35markunSlasheri: Thanks! How much faster is the hd scanning?
13:56:06SlasheriLinusN: or what do you think about committing that code soon? (with the option to skip caching at boot disabled)
13:56:18Slasheriit's much faster
13:56:20markunNow it takes about 8 seconds here I think.
13:56:21amiconnmarkun: The stereo width formula is in firmware/sound.c, lines 432ff
13:56:25XavierGrI think it is great slasheri!
13:57:21LinusNSlasheri: gimme da patch
13:57:29LinusNi want to try it first
13:57:43SlasheriLinusN: ok, i will put it to the web soon :)
13:57:47amiconnmarkun: The basic idea is simple. For stereo width <100%, the left signal is mixed partially into the right channel and vice versa
13:58:22amiconnFull amplitude will yield mono of course
13:59:04amiconnFor stereo width >100%, the left signal is also partially mixed into the right channel and vice versa, but with inverse phase
14:00:05amiconnFull amplitude will give a virtual stereo width of infinity, meaning the mono signal part is completely cancelled
14:00:12amiconnThat's what karaoke mode does
14:00:44markunamiconn: ok, but what is the full amplitude? It's more than 200%?
14:01:19ep0ch*** ok my iHPs HD is now officially dead, does anyone have a single platter HD I can buy from them? ***
14:02:39amiconnmarkun: Of course the total amplitude has to be scaled down in order to avoid clipping least for widths <100%
14:03:27XavierGreP0ch: are you sure? what happened?
14:04:12ep0chwell i've just about recovered most of tunes from it, so i now decided to do a full format on the drive...
14:04:14XavierGrIf you want to buy an HD for you H100 then try I got my 60Gb HD from there and no worries.
14:04:27ep0chbut the format is not going well
14:04:31amiconnThat's what the formula in sound.c does. Only thing to note is that the MAS mixer also inverts, so 100% is -1, not +1
14:04:32XavierGrAnd it never got back?
14:04:47ep0chXavierGr: thanks for the link i'll take a look
14:05:14markunamiconn: I think I will give it a go. Some albums (Beck, Hendrix, Beatles) don't sound so good with headphones because of the extreme use of stereo.
14:10:39amiconnHehe. My standard stereo width setting for earphones is 120%...
14:10:53amiconn(doesn't work on iriver yet)
14:14:30preglowand wont until someone figures out how to fix the latency issue :/
14:16:33preglowunless a five second latency is acceptable for sound changes, which i really dont think
14:16:55XavierGrdoes anyone know a quick way to print an integer with splash?
14:17:03_FireFly_ok i could remove some #ifdefs in my patch it can be reviewed hier:
14:18:14Slasheripreglow: one way to "fix" that could be temporarily limiting the pcm buffer length while browsing the sound settings
14:19:07preglowdon't know what i think of temporary fixes like that
14:19:14preglowmakes it far too easy to ignore the fact that there's a real issue
14:19:35ep0chor have another buffer albeit very small that reads from the pcm buffer?
14:20:13linuxstbIs the only solution to move the DSP closer to the DAC? i.e. have a large buffer for unprocessed PCM data, and a tiny buffer the other side of the DSP?
14:20:16Slasheriep0ch: hmm, that could be one possible way to do it
14:20:42preglowlinuxstb: would be a good solution, but wasting memory, of course
14:20:47ep0chsnap :p
14:21:02linuxstbWhy wasting memory? Because we store 32-bit samples?
14:21:02LinusNXavierGr: splash(HZ, true, "%d", myint);
14:21:21ep0chits not a waste if it fixes the problem
14:21:29Slasheripreglow: and the other solutions of telling codec the exact position where it should start decoding pcm data again would be very complex and hard to get working properly
14:21:30XavierGrso it works like the printf? I though it can take only chars and strings!
14:21:42LinusNjust like printf
14:23:34ep0chwouldn't this small DSP buffer have to have a mega high priority though?
14:24:22Slasherithen it couldn't do crossfading
14:24:38ep0chthat could still be done in the pcm buffer? :s
14:24:47Slasherino if it's unprocessed
14:24:53Slasheri(no dsp at all)
14:25:18Slasheribut maybe processed pcm buffer and post-processed final output buffer
14:25:23ep0chok two buffers, one big one small, the DSP can affect either one it chooses
14:25:36Slasherihmm, maybe
14:25:55preglowthen we'll have to write dsp functions for both shorts and longs
14:26:09pregloweverything will have to be duplicated
14:26:44ep0chwill all functions need access to both buffers?
14:27:36 Quit Febs (Read error: 110 (Connection timed out))
14:28:23XavierGrLinus: Do you remember if you have a value that holds how many presets exist in the list?
14:28:54XavierGrforget it I found it num_presets
14:29:43amiconnI think the dsp might need to be split in 2
14:30:03preglowlike one part where precision isn't a problem?
14:30:06preglowlike channel width?
14:30:29preglowthe usual archos settings
14:30:39amiconnFirst part being the dsp that we have now, doing resampling, unify sample format etc, where latency isn't a problem
14:31:12amiconnSecond part would do the time critical things, rendering to a very small output buffer
14:31:29webguest54preglow, hows the eq comming on ?
14:31:33preglowi don't see why this seperation is necessary
14:31:41preglowwebguest54: not at all
14:31:44amiconnIt could run in the isr setting up audio dma, but then it must not take too much cpu time
14:31:57preglowwebguest54: i've got a prototype going, but not ported to target yet
14:32:00_FireFly_preglow how about the wavpack codec ??
14:32:20preglow_FireFly_: might look at it later, i've got some other coding to do first
14:32:27webguest54is it a comprehensive eq?
14:32:37preglowwebguest54: as comprehensive as you want it
14:32:49XavierGrcoprehensive meaning?
14:32:50preglowi haven't thought about a gui
14:33:07preglowamiconn: too late, i don't want 16 bit samples for precision reasons
14:33:25preglowamiconn: i want all dsp to run on the full precision buffer
14:33:38webguest54XavierGr, comprehensive= full featured
14:35:05preglowamiconn: and again, i don't see why this seperation is needed, shouldn't all audio adjustments be low latency?
14:35:41preglowamiconn: as for the resampler having no issues with latency, that's wrong, pitch adjustments will need to adjust the resampler on the fly
14:35:58 Join rasher [0] (
14:36:22Slasheripreglow: Hmm, what about possible echo and reverb-plugins? I think they need more buffer
14:36:34preglowin what way?
14:36:43preglowthey'll need additional ring buffers, yes
14:36:47preglowbut i don't see how that enters into this
14:36:55Slasheriah, ok..
14:37:16Slasheribut crossfading at least needs a long buffer somewhere
14:37:54preglowwe can't dispense with the long main buffer
14:38:37preglowand the simplest possible solution to this is keeping this buffer full precision, and applying dsp in the other end of it
14:38:55preglowbut other solutions not requiring the wasted memory would be nice
14:39:27linuxstbI can't think how there can be any other solutions.
14:39:33preglowno, me neither, really
14:39:50preglowthe traditional low latency way is just keeping a very small buffer
14:40:03linuxstbAnd as ep0ch said, it's not wasted if it means very low latency and high audio quality.
14:40:21preglowwe could of course have a switch
14:40:25preglow'low latency on/off'
14:40:39preglowif it's off, the buffer goes back to being 16 bit, and dsp is applied straight after decoding, like now
14:40:54Slasheripreglow: and it's also possible to do just that, we could set a max buffer usage limit at anytime we need
14:41:05linuxstbAs long as you call it a switch, and not an option, it will be OK :)
14:41:15preglowas far as i know, this setting would not add much complexity at all
14:41:43preglowSlasheri: sure, but then crossfading and low latency would not work at the same time
14:42:03Slasheripreglow: hmm, true
14:42:36ep0chif you dont want to waste memory reduce the main buffer a bit to compensate ;)
14:43:06linuxstbDo we need 32-bits? Would making it 24-bit be more trouble than it's worth?
14:43:25Slasheripreglow: but it's very rare user wants even listen to crossfading while trying different sound settings
14:43:45XavierGrOk here is a interface wuestion: If radio is on preset mode (left/right cahnges presets instead of frequencies) what long left/right would do. Frequency scanning or Preset Scrolling?
14:43:57preglowlinuxstb: yes indeed it would
14:44:15Slasherilinuxstb: i doubt we can make any audio buffer 24-bit because that would dramatically affect the performance
14:44:30preglowlinuxstb: or not really, we'd need to extract the samples before we can use them, would just need a bit more cpu
14:44:41ep0chXavierGr: how about nothing? :)
14:44:53webguest54can I delete the rockbox file which has the 'Rolo' stuff
14:45:27Slasheriwebguest54: hehe, if you delete that you can no longer boot in to rockbox :D
14:45:34preglow24 bit would need another costly stage just for saving and retrieving from the main buffer, though
14:45:39XavierGrep0ch: it seems like an option, but nothing it seems lazy.
14:45:48 Quit webguest95 ("CGI:IRC (EOF)")
14:45:49Slasheriwebguest54: if the file name is rockbox.iriver
14:45:55XavierGrLinusN; any feedback
14:46:17ep0chXavierGr: i think in preset mode you shouldn't need anything but next/previous preset
14:46:30XavierGrwebguest54 dont delete that file, then rockbox will not boot.
14:46:53 Quit t0mas (Read error: 104 (Connection reset by peer))
14:46:54 Join Tomas_ [0] (
14:47:02 Nick Tomas_ is now known as t0mas (
14:47:03amiconnSlasheri: Why does crossfade need a large pcm buffer?
14:47:05ep0chXavierGr: Frequency Scanning makes the mode inconsitent
14:47:10XavierGryes but let's say you have 64 presets and you are in the middle. If you want to go at the first station holding left is more right
14:47:35XavierGrep0ch: I agree on that, that why I say preset scrolling.
14:47:43XavierGr^that is
14:47:49ep0chXavierGr: ok ok, so preset scrolling takes you to the first or last preset?
14:48:02preglowamiconn: so we dont need to run two codecs in parallel
14:48:18preglowwhich would entail even more codec swapping on top of what we do now for voice ui
14:48:21amiconnSlasheri: Also, how large (in terms of seconds) is the pcm buffer if crossfade is disabled?
14:48:22XavierGrit scrolls but more quickly. 25-24-23-22-21-20 e.t.c fast and with one button
14:48:23Slasheriamiconn: it needs to access the samples from previous song to do the crossfading/mixing
14:48:25LinusNXavierGr: i'd say preset scrolling
14:48:37Slasheriamiconn: it should be 2s
14:48:39amiconnSlasheri: Ah, yes. Maybe I have a solution for that :)
14:48:40LinusNslowly, so you can hear each preset
14:48:45Slasherihmm :)
14:49:28XavierGrHmm slowly eh? I will have to think for that first before implementing?
14:49:35Slasheriamiconn: what kind of solution are you thinking about?
14:49:40amiconnSlasheri: Could that be lowered?
14:49:48amiconn(the 2 seconds I mean)
14:49:53XavierGrDoes the HZ in splash can be used for time latency?
14:49:54Slasheriamiconn: not much without experiencing possible problems
14:50:15amiconnMy idea regarding crossfade is as follows:
14:50:23Slasheriat least 1s is not long enough for the highest bitrate files
14:50:26LinusNXavierGr: for what?
14:50:36amiconnThe (standard) pcm buffer is always small, i.e. the 2 seconds
14:51:12amiconnIf crossfade is enabled, there would be a second pcm buffer with the length of the crossfade setting
14:51:51Slasheriamiconn: ah, hmm that sounds good in fact :)
14:51:54LinusNXavierGr: the first argument to splash() is the number of ticks to wait before returning from splash()
14:52:04LinusNit can be 0
14:52:05amiconnThe end of the old track would be decoded into that, and the dsp would take the old-track samples from there to mix with the new-track samples, then put that into the short main pcm buffer
14:52:42preglowexactly how does this differ from what we do now?
14:52:47preglowapart from using more memory, that is
14:53:39amiconnpreglow: It uses *slightly* more memory (2 seconds worth of pcm audio), but it would keep the lower latency of non-crossfade mode for the dsp
14:54:28XavierGrLinus: But is there a function to delay the scrolling? Because if I call splash, a blank screen will appear between calls.
14:55:13XavierGrYeah you said scroll the presets slowly, I will need a time delay for that.
14:55:15preglowas a matter of fact, proper low latency can't be done for crossfading unless we decode two tracks in parallel
14:55:42preglowforget me
14:55:44LinusNwell, i didn't know you wanted to use splash for the preset scrolling
14:56:16XavierGrwell i asked if there is something similar with the delay of splash.
14:56:23LinusNmy idea was to switch to the next preset
14:56:37 Nick TiMiD[zZZ] is now known as TiMiD (n=TiMiD[
14:56:39LinusNalbeit more slowly than the regular key repeat
14:56:52XavierGrhi TiMiD
14:57:09LinusNa normal delay is done with sleep(), but that is not recommended in this case
14:57:12_FireFly_hi TiMiD
14:57:24XavierGrSo what do you suggest?
14:57:41XavierGrI must use something for delay.
14:57:51LinusNpretty much the same think as the frequency scanning does
14:58:54LinusNbut use TIME_AFTER() to pace the switching
15:00:27 Join Xavi [0] (
15:00:33XaviSorry for that.
15:01:03amiconnLinusN, XavierGr An example of repeat-pacing can be found in the sudoku plugin
15:01:35XaviAs I was saying: Linus: The frequency scanning routine uses the sleep function which you previously said that it is inapropriate.
15:02:40preglowbut anyway, no better alternatives than a full precision buffer for low latency?
15:04:38Xavihmm on my disconnection my image is still on the server.
15:06:52TiMiDXavierGr: about the work I'm doing on remote : everything should be here :
15:07:43 Join Sucka [0] (
15:07:49TiMiDI didn't started to integrate it to the rockbox gui interface since I didn't have a lot of time this week
15:08:23TiMiDbut in the test program, list is working fine, and status bar as well
15:08:32TiMiDtested on all models with simulator
15:09:15Xaviso what are these files?
15:09:28XaviYou will include them if there is a remote for the target?
15:10:01TiMiDthese files are a rewrote / adaptation of the code for handling display
15:10:24TiMiDit does the same work as the code that handles lists and statusbar
15:10:36TiMiD(list is a rewritten fros scratch)
15:10:59 Quit paugh ("Leaving")
15:11:07_FireFly_when he is ready with his work than my patch will be obsolete :)
15:11:07TiMiDthen all programs in rockbox that wants to use a list should call this code
15:11:38TiMiD_FireFly_: your patch has a long time doing then :P
15:12:39TiMiDbtw, you have done a huge work on it ! (grep gives me more than 4000 calls to lcd_* functions in the whole code ...)
15:13:58***Saving seen data "./dancer.seen"
15:14:04_FireFly_bagdar mention that there are too many #ifdefs :) i could reduce it a little bit but most of this #ifdefs are currently needed imho
15:14:31ZagorTiMiD: are both menus and file browser the same list?
15:14:39XaviTiMiD: what the devs say? Are you on a good approach?
15:15:09_FireFly_it would make it a lot easier to manage more displays
15:15:12TiMiD_FireFly_: I've got only 2 or 3 ifdefs for HAS_REMOTE :)
15:15:30_FireFly_in your code ??
15:15:54 Part ep0ch
15:16:00TiMiDZagor: yes itis handled the same way (that means menus could eventually have icons too :P)
15:16:04 Quit XavierGr (Read error: 110 (Connection timed out))
15:16:25ZagorTiMiD: how do you handle the big file buffer?
15:16:26TiMiDXavi: I don't know, I think nobody looked at the code
15:16:30 Nick Xavi is now known as XavierGr (
15:17:01TiMiD_FireFly_: the only ifdefs are here : and here : :P
15:17:14TiMiDZagor: I don'tduplicate it
15:17:41_FireFly_i had short look at the code from TiMiD and it looks good for me:)
15:17:43Zagoryeah I mean how do you get it in the list widget? a list of pointers?
15:17:48TiMiDthe list is filled with callbck functions (to avoid eating the preciousssssss memory :) )
15:18:22 Quit actionshrimp (Read error: 110 (Connection timed out))
15:18:48XavierGrBetter someone exprerianced dev look at your code to see if you heading the right way, I can't help you there.
15:18:51TiMiD(a 9999 files dir eats even with ptrs 80k (and on players with 2M memory it's a lot)
15:19:21B4gder9999 files dirs are insane, on any model ;-)
15:19:50TiMiDI don't have more than 50 in one dir :p
15:20:03TiMiD(and it's the max :) )
15:20:14_FireFly_TiMiD: it is easier to reduce the #ifdefs when rewrite the code complete :) than adding a functionality to a existing code especially if the new functionality only used from some supported devices :)
15:20:23 Quit ender` (Read error: 110 (Connection timed out))
15:21:03TiMiDyes, but writing new code is not very easy
15:21:10_FireFly_but i do my best to use only there #ifefs where it is necessary
15:21:24_FireFly_that is the other side :)
15:21:49TiMiDthe problem is that the drawing codeis everywhere in the current code
15:22:07ZagorTiMiD: I know it's early to complain, but if you could make the code 80 columns wide that would be nice. and split very long lines.
15:22:33B4gderotherwise the Code Police will strike down! ;-)
15:22:46TiMiD(Iwork in 1280x800 so I didn't noticed it was that long)
15:23:07XavierGr1280*800 are you sure?
15:23:08B4gderI work in 1600x1200
15:23:13LinusNme too
15:23:16XavierGrit is 1240*1024
15:23:17B4gderI never go beyond 80 cols
15:23:19TiMiDmy laptop ...
15:23:22LinusNbut my emacs window is 80 columns
15:23:38TiMiDargh emacs :( vim :)
15:23:45LinusNwar! war!
15:23:49TiMiD(don't feed the troll please :p)
15:24:15TiMiD(but vim remains the best editor of the galaxy ^^)
15:24:20_FireFly_i like my jdit ;)
15:24:31markuncopy con rules ;)
15:24:55TiMiD(btw, I'm woking with notepad2 on windows because under linux the simulator doesn't want to compile)
15:25:04TiMiDyes I know this sucks hard :(
15:25:08XavierGrnotepad2 rules!
15:25:32B4gderyou are insane
15:25:49TiMiDno :( I miss my middle click copy-past so much :(
15:25:55TiMiDand I hate windows interface :(
15:25:57rasherTiMiD: what's up with the simulator?
15:26:07TiMiDyou want the error :)
15:26:24TiMiDkevin@harkonnen /mnt/obelix/rockbox/rockbox-devel/build $ make
15:26:24TiMiDmake[1]: Nothing to be done for `all'.
15:26:24TiMiDmake[1]: Nothing to be done for `all'.
15:26:24DBUGEnqueued KICK TiMiD
15:26:24TiMiDmake[1]: Nothing to be done for `all'.
15:26:24***Alert Mode level 1
15:26:24TiMiDLD rockboxui
15:26:27TiMiD/usr/lib/gcc/i686-pc-linux-gnu/3.4.4/../../../../i686-pc-linux-gnu/bin/ld: errno: TLS definition in /lib/ section .tbss mismatches non-TLS reference in /mnt/obelix/rockbox/rockbox-devel/build/librockbox.a(errno.o)
15:26:31TiMiD/lib/ could not read symbols: Bad value
15:26:33TiMiDcollect2: ld returned 1 exit status
15:26:36TiMiDmake[1]: *** [/mnt/obelix/rockbox/rockbox-devel/build/rockboxui] Error 1
15:26:38TiMiDmake: *** [all] Erreur 2
15:26:39rasherAh, that thing.
15:26:41TiMiDkevin@harkonnen /mnt/obelix/rockbox/rockbox-devel/build $ gcc -v
15:26:43TiMiDLecture des spécification à partir de /usr/lib/gcc/i686-pc-linux-gnu/3.4.4/specs
15:26:46TiMiDConfigur√© avec: /var/tmp/portage/gcc-3.4.4/work/gcc-3.4.4/configure −−enable-version-specific-runtime-libs −−prefix=/usr −−bindir=/usr/i686-pc-linux-gnu/gcc-bin/3.4.4 −−includedir=/usr/lib/gcc/i686-pc-linux-gnu/3.4.4/include −−datadir=/usr/share/gcc-data/i686-pc-linux-gnu/3.4.4 −−mandir=/usr/share/gcc-data/i686-pc-linux-gnu/3.4.4/man −−infodir=/usr/share/gcc-data/i686-pc-linux-gnu/3.4.4/info −−with-gxx-include-dir=/usr/lib/gcc/i686-pc-linux-gnu/3.4.
15:26:53TiMiDModèle de thread: posix
15:26:56TiMiDversion gcc 3.4.4 (Gentoo 3.4.4, ssp-3.4.4-1.0, pie-8.7.8)
15:26:58TiMiDkevin@harkonnen /mnt/obelix/rockbox/rockbox-devel/build $ ld -v
15:26:59rasherI blame the French.
15:27:01TiMiDGNU ld version 2.16.1
15:27:03TiMiDyes :(
15:27:05B4gderI blame gentoo
15:27:10rasherOr that, yes.
15:27:16TiMiDI love french and gentoo :)
15:27:17LinusNi blame glibc
15:27:23TiMiDI fuck gcc
15:27:32B4gderwell, gentoo _did_ actually screw up
15:27:40B4gderand TiMiD might be using that screwed version
15:27:44markunI can't even configure because FreeBSD's sed takes different arguments.
15:28:06B4gdermarkun: we could remove the sed use and replace it with pure perl instead...
15:28:21TiMiDI have another gentoo box with the same versions and the same compile options for glibc and the simulator is working fine
15:28:33_FireFly_im running gentoo too ;) and i had no problems to compile the simulator
15:28:57B4gderright, the problem was with some specific version
15:29:05B4gderI don't recall which
15:29:14TiMiDyaa 2 gentoo integrists here =(_O)=
15:29:25preglowB4gder: now there's a good idea
15:29:50TiMiDsome people on the net have the same problem with other distribs and other programs
15:30:02LinusNperl? nah, let's do it in java!
15:30:06CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
15:30:06*LinusN hides
15:30:18TiMiDthe error message means that errno is no safe with multithread
15:30:31Zagorlet's use ant instead of make. ant is revolutionary!
15:30:42LinusNZagor: yes! yes! yes!
15:30:46markunOr c# ;)
15:30:47_FireFly_TiMiD: have you compiled glibc with following use-flag set nptl and nptl-only ??
15:31:06_FireFly_or scons ;)
15:31:24TiMiD_FireFly_: yes, and since it wasn't working, I recompiled it with ntpl+linuxthreads
15:32:02ZagorI'm afraid scroll speed on archos may suffer terribly with your approach, TiMiD :-(
15:32:03_FireFly_jupp the simulator doesn't work with a glibc which has only nptl support
15:32:46TiMiDZagor: because I redraw the screen at each refresh ?
15:33:10Zagorno, because each redraw does so much work
15:33:26TiMiDeach redraw draws all the lines
15:33:43Zagorseveral callbacks via struct pointer lookup for each line
15:34:06TiMiD2 callbacks
15:34:17TiMiDone for icon and one for label
15:34:50TiMiDbut since I don't have an archos I can't test :(
15:34:58Zagorcan you compile a test version for archos so we can test?
15:35:10Zagorrec20 for me
15:35:29TiMiDrecorder v2 ?
15:35:39Zagorno sorry, option #2: archos recorder
15:36:06LinusNZagor: but isn't the most cpu time spent on rendering the text anyway?
15:36:21B4gderI would assume so
15:36:25***Alert Mode OFF
15:36:29Zagori hope so
15:37:31TiMiDI tested on iriver with a loop that scrolls (each loop turn scrolls a line) and it was not slow at all
15:37:42TiMiDbut since archos have aslower cpu :(
15:38:14TiMiDoops some compiles erors with my demistify plugin :)
15:39:01ZagorI just need the ajz file anyway
15:41:12TiMiDrombox.ucl ?
15:41:26TiMiDoops .ajz u said :)
15:43:12Zagoroh, it's only test code still? no dir loading?
15:43:31Zagoralso keyrepeat doesn't work so scrolling is difficult to test :-)
15:44:06TiMiDi will make you a build which scrolls without keypresses :)
15:44:18Zagorpressing DOWN repeatedly does strange things. no scrolling, but seems to make the list shorter
15:44:31Zagorthat won't help me compare with the old code.
15:44:59TiMiDup and down are in test progra used to add and removedummy items
15:45:36TiMiDit's left / right (because the player doesn't have up/down and I was bored to change the keys each time I tested on it
15:45:50Zagorwell let's run more tests when the code does a little more :-)
15:46:08Zagorwhat did you change?
15:46:18TiMiDit should scroll as fast as it cans
15:46:25TiMiDwithout a keypress
15:46:34Zagorcan't you make it recognize key repeat instead?
15:46:45Zagorthat would make it more comparable
15:46:53LinusNZagor: don't push him :-)
15:47:19LinusN"why doesn't it work yet?"
15:47:32TiMiDthe code for that is in tree.c and I wanted to do this when integrating the code to tree.c
15:47:34Zagorthat's not what I'm saying
15:47:47ZagorTiMiD: ok, no problem
15:48:02TiMiDhere it's just a small test program so that I can see everything is working as expected
15:48:02XavierGrZagor: Does it have any lag?
15:48:17ZagorXavierGr: can't test that yet
15:48:31TiMiDjust dl
15:48:49TiMiDand see how fast is the max speed
15:49:06XavierGrOk another interface question: Do we want somewhere in the screen to declare that Preset Mode is on (preset scanning instead of frequency)?
15:49:10Zagorwell just scrolling really fast doesn't test what needs to be tested. we need to verify it scrolls as fast as the button queue is filling up with keypresses.
15:49:37Zagorso that when the user lets go of the button, the scroll stops immediately and doesn't continue a couple of lines
15:49:45Zagorthis has been an issue before
15:49:48TiMiDZagor: but if it scrolls fast the it should be fast enough
15:50:09Zagori hope so, yes. we'll simply have to wait and see.
15:50:13TiMiDwhy would scroll continue ?
15:50:29TiMiD(the list code doesn't handle buttons)
15:50:35Zagorbecause the button queue still has unprocessed button events
15:50:37LinusNbecause ot scrolls slower than the key repeat
15:53:38TiMiDbtw, do you think the approch /api is good or should it be something else .
15:56:28ZagorI'm not sure. On one hand I think it limits what can be done, since everything must now be a widget.
15:57:39Zagoron the other hand it's a good solution.
15:58:02TiMiDI dont want to call this a widget :p
15:58:17TiMiD(a widget is lot more generic than this)
15:58:48Zagorwell if it handles menus and file tree, that's plenty generic for us
15:59:23TiMiDsure it would avoid duplicating code (even if menus and filetree are not exactly the same)
15:59:31Zagorbtw, simply add "case BUTTON_RIGHT | BUTTON_REPEAT:" after "case BUTTON_RIGHT:" to enable key repeat.
15:59:43TiMiDoh :)
16:00:20TiMiDI recompile it
16:08:26Zagorspeed looks good
16:10:37TiMiDkewl XD
16:12:25TiMiDwell I'm going to school right now (and I don't think I will connect this evening)
16:13:30 Part LinusN
16:13:31TiMiDso think about that and tell me if I can start doing smth usefull (like filetree)
16:13:55XavierGrOk another interface question: Do we want somewhere in the screen to declare that Preset Mode is on (preset scanning instead of frequency)?
16:14:34 Nick TiMiD is now known as TiMiD[farAway] (n=TiMiD[
16:14:34DBUGEnqueued KICK TiMiD[farAway]
16:14:38XavierGrI ask again and also: Do all the archos targets are much different from choice 3 of configure?
16:16:14XavierGrWhich targets have FM radio besides: iriver, ondio FM and FM recorder?
16:16:43rasher ?
16:17:15XavierGrah yes thanks that's what I was looking for. I had forgotten the page.
16:24:25 Quit ashridah ("Leaving")
16:32:05rasherB4gder/Zagor: Bleeding sources are missing a lot of files
16:32:33rasherparticularly in codecs, but also elsewhere
16:32:40B4gderI need to run
16:32:42 Quit B4gder ("time to say moo")
16:32:57rasherHow convenient
16:39:42 Join jborn_ [0] (
16:40:08linuxstbrasher: What is missing? Can you add them to apps/FILES ?
16:41:26XavierGrGreat! The load/save fm preset patch is finally done and workable!
16:41:40rasherlinuxstb: if that's how to fix it, sure
16:42:02rasherall c files in libmad for example
16:42:27rasheronly changes, copying, copyright and credits is included
16:42:58rasherwell, FILES has codecs/libmad/*
16:43:02rasherI suspect something else is up
16:46:27linuxstbLooks that way. Lets wait for Bagder to come back.
16:47:48rasherOn a related note, libalac isn't there
16:48:24_FireFly_in the daily-builds there is libalac
16:49:27rasherI meant in apps/FILES
16:49:28preglowthe source packer has been having some problems lately
16:49:53rasherMaybe it should just do a cvs checkout and delete the cvs control files?
16:50:58_FireFly_rasher also in apps/FILES in the daily-builds
16:51:14_FireFly_there is libalac
16:51:15rashermaybe I'm just not using my eyes
16:51:36rasheroh yes indeed, I'm not
16:52:04_FireFly_it seems that only the bleeding-sources are "corrupted"
16:53:14rasherso the daily sources work, but bleeding doesn't?
16:53:15rasherthat's bizarre
16:54:33XavierGrto add some LANG names I must edit the english.lang?
16:54:41_FireFly_jupp because i using the daily-build 20050930 to add new "features" to my remote-patch
16:54:51_FireFly_XavierGr japp
16:55:22rasherindeed, libmad contains everything from a daily
16:55:23_FireFly_and after that you must use binlang (is in tools) to convert the *,lang to lng
16:55:25amiconnrasher: iirc, 'cvs export' just does that what you want
16:55:41XavierGrOk then enough for now, when I come back I will add the strings to english.lang and make that create new file preset menu.
16:55:50XavierGrlater all!
16:55:56 Quit XavierGr ()
16:56:00 Quit JoeBorn (Read error: 110 (Connection timed out))
17:13:59***Saving seen data "./dancer.seen"
17:27:07 Join jborn__ [0] (
17:37:24 Join ghode|afk [0] (
17:40:53 Join muesli- [0] (
17:44:50 Quit jborn_ (Read error: 110 (Connection timed out))
17:45:27 Join JoeBorn [0] (
17:48:14_FireFly_ho :)
17:52:10 Part linuxstb ("Leaving")
17:56:16muesli-hi feuerfliege ;)
17:58:11muesli-muuuuchas gracias for your works on the remote
17:58:42 Quit jborn__ (Read error: 110 (Connection timed out))
18:01:15muesli-when i'd like to build a rwps just click it to activate it?
18:02:31_FireFly_or use the .rwps menu :)
18:02:45muesli-rwps menu?
18:03:22_FireFly_a similar menu as the .wps menu
18:03:52_FireFly_is in General-Settings->Display->Browse .rwps files
18:04:41muesli-aah, i c ;)
18:05:12muesli-are conditional images possible?
18:05:42_FireFly_i have limited the size to 800 bytes i don't think that more will be necessary
18:06:10_FireFly_yes the size limit is the only thing that i have changed
18:06:55_FireFly_the rest is the same as for the normal wps
18:07:15muesli-800bytes are quit tight. you can easily reach that size in a wps
18:07:29_FireFly_my one doesn't ;)
18:07:59_FireFly_but it can be easly change if necessary
18:08:57muesli-ie. if you use 4images as a battery indicator and 4-5 for volume its almost over
18:10:20_FireFly_i have 13 images in my remote wps and it has only 606bytes
18:10:24solexxare there any pre-built versions available for download?
18:10:38solexx(with remote patch)
18:11:04_FireFly_based on 20050930 :)
18:11:14solexxthanks a lot!
18:11:37_FireFly_the zip has only the rockbox.iriver and the updated english.lng
18:11:56_FireFly_the rest can be used from the official daily-build
18:14:24solexxthats ok
18:14:37_FireFly_muesli: you can save a bit of space when you reduce the filename-length of the bitmaps
18:15:13muesli-e4k.bmp reckon barely space for savings ;-)
18:16:55muesli-56bmp's ;)
18:18:33_FireFly_wow that are many bitmaps
18:19:00muesli-but i guess there are some rudiments
18:19:10CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
18:19:10*rasher is thinking that could be reduced a lot
18:19:34_FireFly_think about your remote display isn't so big as the main display ;=
18:19:57muesli-true too ;)
18:20:08muesli-45images are in use *just counted*
18:20:15 Join webguest41 [0] (
18:20:27muesli-but not at the same time
18:20:49muesli-as i said... 4 for codecs
18:20:53muesli-4 for playmode
18:21:00muesli-4 battery level
18:21:06muesli-4 volume level
18:23:08muesli-thats the hidden trap of conditions..many choices require many assumptions..
18:24:03 Join ender` [0] (i=ychat@
18:32:00amiconnmuesli-, _FireFly_: I just want to remind of my suggestion to allow bitmap-regions, and put many images used in the .wps into one .bmp files
18:32:37amiconnThis would both help to reduce file system clutter, and keep the .wps files smaller
18:33:29_FireFly_that is a good idea but i have currently no clue how to implement it
18:33:33muesli-err..could you describe it..just dont have any pic of your idea
18:33:51_FireFly_you have one bitmap file
18:34:06amiconnMany skinning systems for pc programs use that
18:34:35_FireFly_the bitmap file have all stages e.g. volume-bar
18:35:12_FireFly_the iriver firmware use it also
18:35:22 Quit webguest41 ("CGI:IRC (Ping timeout)")
18:36:06muesli-nope, afaik irivers fw uses for any event a specific bmp
18:36:34rasherIt uses a couple of combined bmps
18:36:59amiconnmuesli-: The idea is to put multiple, or even all, images used in the wps into one .bmp
18:37:24amiconnThe individual images can be put side by side, or benath each other, in the .bmp
18:37:54rasherin the case of a volume meter, you could even just use one image, and then use half of it
18:37:55amiconnThe wps code would then allow to clip a region from the whole .bmp for display
18:38:04muesli-and when you want a specific satge for i.e battery put some "spacers" about it?
18:38:44amiconnrasher: Yes, provided the higher stages are simply extending some kind of graph
18:40:00amiconnSometimes one may want en extra image for volume==0, a muted speaker symbol or such
18:40:56amiconnrasher: The iriver firmware allows changing of images?
18:41:21rasherBut internally it uses combined bmps
18:41:31_FireFly_someone has striped it off
18:41:39_FireFly_from the firmware
18:43:03muesli-amiconn your idea sounds good..but doesnt kill it the freedom of different wps?
18:43:17amiconnWhy should it?
18:43:51amiconnFirst, we could keep the ability to load multiple .bmp files *in addition* to allow clip regions
18:44:26amiconnI'd rather not want that, but even the one could rearrange available images with an image editor
18:44:35amiconn*even then...
18:45:04muesli-how do want to place different codec icons at the same spot when using only one bmp?
18:45:16amiconnNot at the same spot
18:45:40amiconn(in the bmp)
18:45:49amiconnThat's of course impossible
18:46:12muesli-how do want to solve this?
18:46:35amiconnYou clip a region from the .bmp, and draw it at a specified location on the screen
18:46:50amiconnBoth positions are independent from each other
18:47:11_FireFly_here is an example bmp from the iriver firmware
18:47:27amiconnHmm, that might in fact not save that much on the .wps file size...
18:47:57amiconn_FireFly_: That's.. stupid. Repeating 'CH' over and over...
18:48:33_FireFly_i haven't done this bmp ;)
18:48:48_FireFly_it is a bmp from the frimware stripped off
18:49:08amiconnI know... I mean, iriver is stupid.
18:49:24_FireFly_or it was easier to make it so
18:50:08muesli-dont be so harsh with them ;-) just see from their point, develop it fast and easy ;)
18:53:39 Join guillaumh [0] (
18:55:42 Join linuxstb [0] (
19:00:13linuxstbI'm about how an iPod colour LCD driver will work in Rockbox. It's 220x176 (same as H300) 16-bit colour: 5R, 6G, 5B, (the H300 is listed as 18-bit) and the data is written to the device as 32-bit words containing the values of two adjacent (in the x direction) pixels.
19:00:49linuxstbHas anyone thought about how the H300's colour LCD will fit into Rockbox?
19:01:29Bagderamiconn might have
19:01:36*Bagder runs off again
19:04:14muesli-gotta go
19:06:47_FireFly_oh poor Bagder ;)
19:06:48 Quit muesli- (Read error: 104 (Connection reset by peer))
19:09:24SlasheriUpdated dircache:
19:09:52Slasheriplease report any problems you will find
19:14:01***Saving seen data "./dancer.seen"
19:22:11 Quit guillaumh (Remote closed the connection)
19:43:56markunSlasheri: It works quite nice. The scanning is a bit faster than before, and a lot faster than the iriver firmware I think.
19:44:26rasherIt was a lot faster to begin with
19:45:38markunThe hd just started spinning when I entered a dir.. but that was just the unicode patch :)
19:50:20ender`how's the unicode patch progressing? i noticed i can't play some files (rockbox just skips over them)
19:51:26markunender`: does the skipping have anything todo with unicode?
19:51:58ender`file names
19:52:37ender`they contain japanese characters, and while iRiver's firmware displays (and plays) them fine, rockbox simply skips to the next file
19:53:28markunHow strange, but it's been a long time since I used a non unicode version of rockbox so I wouldn't know.
19:53:30fuzzieoh, i was wondering earlier if anyone was working on fixing that
19:54:04 Quit Maxime (Read error: 104 (Connection reset by peer))
19:54:57ender`markun: are your binaries available anywhere? (i'm not in the mood to set up compile environment ATM)
19:57:12markunender`: yes, but the font with japanese is very large (12x13)
19:57:32ender`no problem, i'd just like to test if the file plays
19:57:33 Join Maxime [0] (
19:58:06ender`btw, would it be possible to extract the font from original firmware and convert it to a format that rockbox understands?
19:59:20markunender`: Do you want the bdf versions of the fonts so you can convert them yourself?
19:59:41ender`i could probably do that
20:00:40markunhttp:// and
20:00:55markunender`: The are from X11
20:03:28markunCan you convert them or do you want a converted font?
20:05:15ender`is there a converter, or should i write one?
20:06:22ender`btw, the files seem to play now
20:13:35markunender`: Use rockbox/tools/convbdf
20:14:19ender`ok, i'll look at that, thanks
20:17:00markunender`: Here ya go laddy :)
20:21:24 Join mrelwood [0] (
20:22:34mrelwoodhow come the daily build is of a later date than bleeding edge? for iRiver H120 that is.
20:22:36ender`haha, thanks
20:22:56 Join Colddy [0] (
20:23:37rashermrelwood: because the bleeding edge build is the latest - daily is only done once a day
20:24:56mrelwoodthe bleeding edge is for yesterday.
20:25:07amiconnrasher: I think mrelwood means the current daily is newer than bleeding edge
20:25:10_FireFly_currently the daily-build-20050930 is also the latest because the last cvs commit was on 20050929
20:25:16mrelwoodi ment the daily is later than the bleeding, ofcourse.
20:25:33rasherThe daily is built each midnight. Bleeding is built after each change.
20:25:41rasherThere haven't been any changes since midnight.
20:25:48amiconnThe dailies are built once a day, the bleeding edge is built whenever something is checked into cvs
20:25:50mrelwoodokay, so they are the same, even though different date?
20:26:15mrelwoodokay, so no prob then. I'll stich with the bleeding... :)
20:26:25mrelwoodstick I ment. :P
20:30:54 Join ep0ch [0] (n=ep0ch@
20:32:09ep0chSlasheri (or anyone): in order to apply your second dircache.diff do i have to apply the first patch you made?
20:32:09 Quit mrelwood ("CGI:IRC (EOF)")
20:33:59Slasheriep0ch: no, you shouldn't have it installed
20:34:22ep0chmeh, i messed up somehow then
20:34:34Slasheriwhat was the problem?
20:34:42Slasheriit could be my patch also
20:35:07ep0chfirst i put the patch into apps/
20:35:17ep0chapplied the patch with -p0
20:35:30Slasherino, put it to the rockbox root
20:35:41Slasheriit needs to access firmware/ part also
20:37:08 Join solexx_ [0] (
20:39:25 Join paugh [0] (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
20:43:12webguest54Slasheri, I change the contents of my player continuously, will the file update with your system be automatic/transparent,
20:43:28Slasheriwebguest54: yes
20:43:32Slasheriautomatic and transparent
20:43:43webguest54sounds good
20:43:58 Quit Sucka ("a bird in the bush is worth two in your house")
20:44:16BirdFishTransparent update? As in it updates behind the scenes rather than a slowed boot?
20:45:45 Join Moos [0] (
20:49:26 Quit solexx (Read error: 110 (Connection timed out))
20:52:35ep0chis there an easy way to see how much memory & diskspace the dircache uses up?
20:53:01Slasheriep0ch: not a separate debug menu entry yet, but you can use logf for that
20:53:42ep0chwhen does the actual dir cache file get written to disk?
20:53:50ep0chon shutdown?
20:54:04Slasheriwhen you shutdown the unit. But it has no real use currently
20:54:14Slasherionly headers are read on next boot from that cache file
20:54:40ep0chso everytime rockbox boots up now it does a scan? whereas before the user had the option?
20:55:21Slasheriyes, the option is disabled at least until we can get few problems with bootloader usb mode solved
20:55:45Mooslet's hope in Linus :;)
20:56:02Slasheriand only reason for that option was increased battery performance. However, the current implementation gives fast rebuilding time so there should be no need for that option anymore
20:56:43ep0chhow come irivers scanning takes 20 times longer ?? :s
20:56:58MoosSlasheri: thanks for this again
20:57:04ep0chdoes it read all the tag data too or something
20:57:10Slasheribootup time shouldn't be affected at all all whether cache is enabled or not
20:57:25rasher_FireFly_: where's your latest patch?
20:57:55MoosTimid where is your last patch :D
20:57:59Mooshi rasher
20:58:10amiconnSlasheri: I'd vote for removing the option and the need for a cache file altogether
20:58:12*BirdFish wonders why people beg and beg for DRMed music... Napster and such services are such a waste of money. For maybe 2 dollars more, one could get the entire cd in pure format rather than 128 kb/s (and it wouldn't be protected)
20:58:47fuzzieBirdFish: right, but 99c for the only track off it you want is a rather better deal
20:58:49ep0chrasher: quick question about the File Properties patch you use, any idea how do i use it? :s
20:58:55amiconnSlasheri: You can store the needed cache size value in the config sector
20:59:01fuzzieand jhymn makes it a matter of seconds to strip the DRM off iTMS files. :)
20:59:16rasherep0ch: Use "open with" on a file.. but it's really stupid
20:59:26BirdFishfuzzie: I realize, but it's still encoded at 128 kb/s which sux b@lls
20:59:27ep0chduh like me :D
20:59:53Slasheriamiconn: that option is already removed (ifdeffed out)
21:00:14MoosHCl: are you aware, it's about your runtime db ?
21:00:14fuzzie128kbps AAC isn't that bad, but, yeah, if it's not just single tracks then the CD is obviously the best option by far
21:00:35Slasheriamiconn: hmm, yes. i think i will move that cache size value to the config then
21:00:40*amiconn is fiddling with the grayscale lib
21:01:16amiconnWish coldfire wouldn't split the registers into dedicated address & data registers :(
21:01:22ep0chi personnally would like a compromise with the cache and have a dynamic on the fly cache, that still starts the harddisk up even if something is cached or not.
21:01:48rasherep0ch: why would you spin up the harddisk if something is already cached?
21:02:13ep0chso when you navigate to something that isnt in the cache there is little delay (as the HD is already started)
21:02:17amiconnI'm short of data registers, so I have to store values in address registers and swap everytime, as address registers can't be used for logical operations and bitshifts :/
21:02:51Slasheriep0ch: everything is cached so there will be no delays (with my patch)
21:03:23ep0chSlasheri: yes i know, but i'd like the compromise of not wasting too much memory and getting a speed boost
21:03:23Moosall is instantatly, browsing your files, plugins, config...
21:03:58rasherYou can't really use any less memory
21:04:09ep0chwhy not?
21:04:11rasherIt wouldn't make sense to cache half the disk, when it's such relatively small amounts
21:04:12_FireFly_rasher the latest version can be found here
21:04:21rasher_FireFly_: thanks
21:04:43_FireFly_it has also the updated wps-sb-tag patch included in the zip
21:04:46Slasherithat wasting memory is something around 200-500 KiB depending how much you have files.. under 2% of the iriver memory size
21:04:46ep0chjust have a cache size limit, last one used gets kicked out
21:05:06_FireFly_and some other so that they all can be cleanly applied after the remote patch
21:05:14*amiconn had an idea how to fix the problem that the file browser doesn't update the list when returning from a plugin
21:05:27_FireFly_the order are given in the filename beginning
21:05:30rasheramiconn: the plugins hould just return PLUGIN_USB_CONNECTED
21:05:41rasherAt least, that's what Linus said, and it works
21:05:54amiconnrasher: No, that would make the browser re-read the dir *everytime* when returning
21:06:04amiconn(from such a plugin)
21:06:17amiconnMy idea would only make it re-read when necessary
21:06:32amiconn...without touching any plugin source file :)
21:06:58rasherah, taking notice on when the plugin struct file-functions are used?
21:07:07_FireFly_i have added some options for the remote: seperate scrollbar, statusbar, line selector
21:07:19amiconnThis would be a general approach:
21:07:33rasherdammit, rejects
21:07:48_FireFly_which patch ??
21:08:00rasherThe first. I'm not applying to a clean tree
21:08:02amiconnThe file system code would maintain a "revision counter" that is increased every time a file operation changes the directory structure
21:08:13amiconn(creating/renaming/moving a file)
21:08:20_FireFly_it based on 20050930
21:08:46amiconnEvery module that caches directory data (like the file browser) would read that counter and compare it to a private copy
21:09:17amiconnIf the browser sees the counter has incremented compared to its own copy, it would re-read the dir
21:09:53amiconnThis counter would also be incremented when leaving USB mode
21:10:28ep0cheven if no files change during USB mode?
21:10:39rasherthere's no way to tell
21:10:41amiconnep0ch: It's impossible to decide
21:10:49ep0chthat;'s what i thought :)
21:11:03amiconnNo need to excessively pass around info about returning from usb mode, as it is done now
21:11:33amiconnThe return codes of the menu functions, plugins etc could be reused for different purposes, or dropped completely
21:14:04***Saving seen data "./dancer.seen"
21:14:55_FireFly_rasher: whenn you applay both wps-sb-tag patches now it is possible to have the tags also in the wps for the remote
21:15:05amiconnCurrently there are many places where such status info is passed around, not only in menus but also in recording screen etc
21:17:19ep0chanyone know which file/files the quick menu code is in?
21:17:20rasher_FireFly_: well, I think I broke my tree pretty much
21:17:45_FireFly_hmm maybe i have forgotten a patch
21:18:48_FireFly_no all patches are included
21:20:07rasherwell, I've applied other patches before applying yours
21:20:54_FireFly_i have used the plain daily-build :)
21:21:14rasherof course
21:33:02 Quit Colddy ("PG = Boy gets girl, R = Villain gets girl, X = Everybody gets girl .gx.")
21:33:31 Join Paul_The_Nerd [0] (
21:34:15 Join Ismo_ [0] (
21:41:33 Quit Ismo (Read error: 110 (Connection timed out))
21:48:40 Quit Paul_The_Nerd ("Chatzilla 0.9.68a [Firefox 1.0.7/20050915]")
21:49:04 Join XavierGr [0] (
22:04:36 Join dpassen1 [0] (n=dpassen1@resnet-233-61.resnet.UMBC.EDU)
22:14:09 Quit JoeBorn (Read error: 104 (Connection reset by peer))
22:14:48 Join JoeBorn [0] (
22:17:09 Quit JoeBorn (Read error: 104 (Connection reset by peer))
22:17:11 Join jborn_ [0] (
22:41:55 Join webguest95 [0] (
22:43:08 Quit XavierGr ()
22:52:29 Part ep0ch
22:56:33 Join linuxstb_ [0] (
23:00:10 Nick paugh is now known as AliasAle (n=kickback@2001:5c0:8fff:ffff:8000:0:3e03:6822)
23:08:07_FireFly_night everybody
23:08:33 Quit _FireFly_ ("Leaving")
23:12:40 Join RotAtoR [0] (
23:12:58 Quit linuxstb (Read error: 110 (Connection timed out))
23:13:40 Part webguest54
23:14:07***Saving seen data "./dancer.seen"
23:19:22 Nick linuxstb_ is now known as linuxstb (
23:23:36solexx_kudos to FireFly for his LCD builds!
23:24:02solexx_if only my friend who used his remote a lot hadn't got his iriver stolen...
23:27:28solexx_Rockbox another user now...
23:28:13 Join matsl [0] (
23:28:32 Quit Maxime (Read error: 104 (Connection reset by peer))
23:28:47 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
23:28:55 Join Maxime [0] (
23:29:21 Quit webguest95 ("CGI:IRC (EOF)")
23:35:39 Join webguest19 [0] (
23:36:33webguest19hi guys
23:37:17webguest19hello, is there anybody in there? ;-)
23:44:16webguest19midi doesn't really work on the iriver, does it?
23:44:28markunNo, not yet.
23:45:32webguest19alright, just wanted to know if im too stupid or if it really doesn't work...
23:46:22markunThe plugin produces the audio data, but it is not playing it. Someone is working on it.
23:47:42webguest19it seems like it freezes somewhere..
23:47:59webguest19at least i have now a remote that displays something useful

Previous day | Next day