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

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

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

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

#rockbox log for 2007-11-22

00:00:06preglowthe copies are almost certainly negligible compared to the rest of the stuff, so sure
00:00:13preglowi wasn't really looking for a huge speedup anyway
00:01:56scorche|wZagor: ah....forget about it then...not worth the effort :)
00:02:26roolkuNico_P: no, my mistake. Looks good so far, playcount is updated correctly - will add some debug messages
00:09:30 Join saratoga [0] (i=47e276af@gateway/web/cgi-irc/
00:10:19saratogalinuxstb: is there a way for a codec to get the length of a file (for seeking)?
00:10:56preglowisn't that part of its id3 struct?
00:11:48roolkuNico_P: I can't fault your fix, commit it :)
00:12:08saratogapreglow: no idea, never occured to me that it would be in there
00:12:54Nico_Proolku: I think I'll merge audio_release_tracks and audio_clear_track_entries before
00:12:57roolkuNico_P: if you look at - line 24 and 25 (the data for track 6) were missing previously
00:13:14preglowsaratoga: yeah, it's there, called filesize
00:13:50saratogai spent all afternoon trapped in an air port with nothing but wma seeking to work on
00:14:20saratogaand i didn't really get far, some files still just randomly don't work, and each in their own special way
00:14:31roolkuNico_P: one thing I am surprised to see is that you read the metadata for certain tracks twice or three times - will this effect the elapsed time?
00:15:27saratogai'm thinking of just commiting it with a check to make sure you don't seek out of the current file, and making it give up and skip to the next track if a seek fails
00:15:33Nico_Proolku: what do you mean?
00:15:35roolkuNico_P: I am guessing here, but does maybe the second copy override the first and eradicate any elapsed time or other runtime data?
00:16:23roolkuNico_P: example track 7 or track 14 (the lines starting with −−−− are get_metadata)
00:19:11preglowsaratoga: hmm
00:19:17preglowsaratoga: and no idea _why_ the files fail?
00:19:44saratogapreglow: some fail because the binary search gets stuck in an infinate loop (somehow)
00:19:54saratogaothers just sort of die and skip to the next track for no obvious reason
00:19:56preglowwell, that has to be debuggable
00:20:20 Quit przemhb ("ChatZilla [Firefox]")
00:20:24saratogathe loop thing is bizzare
00:20:43saratogasay the correct packet is 100, it'll ping back and forth between 99 and 101 forever
00:21:09preglowfancy bug, but should ultimately be debuggable with enough printfs :>
00:21:10saratogai couldn't debug it mostly because i could reproduce it
00:21:24saratogait would just happen rarely
00:21:40preglowanyway, that bug is kind of a showstopper
00:21:48preglowseeking not working is ok, but seeking hanging is not
00:21:58saratogawell its easy enough to hack away
00:22:12saratogajust limit the search to 5 packets and quit where you are
00:22:17saratogabut i'd like to fix it
00:22:25preglowwell, then commiting it is ok with me, just stuff a warning in the commit message that it's not 100%
00:22:27 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
00:22:41saratogai guess i should hack up the sim some to allow me to do programatic seeks, since presumably that would be reproducible
00:22:56preglowoh yes, definitely the way to
00:23:04 Quit advcomp2019 (Nick collision from services.)
00:23:05preglowjust make the sim dump the seek params, then find one that shits itself
00:23:06 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
00:23:23saratogathis whole problem is infuratingly simple
00:23:32preglowmany of them are :>
00:23:35saratogai have a gettimestamp() function that works perfectly
00:23:41saratogaand has since August
00:23:45saratogaand yet, here we are
00:24:11scorche|wbillenium: stop that
00:29:23 Join H10_007quick [0] (
00:32:05H10_007quickHello, I am trying to compile in Ubuntu and have been succesful except that I get tonnes of warnings that read "recorder/recording.c:1579: warning: pointer targets in passing argument 3 of ‘screens[i].puts’ differ in signedness" Does anyone know what is wrong? Am I missing a package
00:32:25H10_007quickoh and not just errors in that file
00:32:31H10_007quickthey are all over the place
00:33:14LloreanAre you using any patches?
00:33:42H10_007quickjust the standard svn version
00:33:49XavierGrcan someone build voice menus under windows anymore?
00:33:50H10_007quickand there are no errors on the servers
00:34:01XavierGr(without vmware or cygwin)
00:34:49H10_007quickalways get argument errors
00:35:11 Quit H10_007quick ("ChatZilla 0.9.79 [Firefox]")
00:35:20 Quit ompaul (Client Quit)
00:35:31 Join H10_007quick [0] (
00:36:03H10_007quickoops I hit X one too many times did I mis something?
00:36:50H10_007quickOh and some files do not have warnings at all
00:37:39 Join BigBambi_ [0] (n=Alex@rockbox/staff/BigBambi)
00:37:53H10_007quickis it possible that I have the wrong C compiler?
00:37:59 Quit BigBambi (Read error: 104 (Connection reset by peer))
00:38:35 Nick BigBambi_ is now known as BigBambi (n=Alex@rockbox/staff/BigBambi)
00:40:47ZagorH10_007quick: how did you build it?
00:40:54n1sH10_007quick: which version is your gcc and are you compiling a sim or for a target?
00:41:23H10_007quickand I just ran the shell script
00:41:33H10_007quickand compiled arm-elf-gcc
00:41:38H10_007quickand added it to my path
00:41:51H10_007quickand it compiles fine now just that there are tonnes of warnings
00:41:56n1stry arm-elf-gcc -v
00:42:37H10_007quicknope didn't help
00:42:49n1sbut what did it print?
00:43:06n1swhich version?
00:44:03XavierGrHow can someone use SAPI with linux? On the voicebuilding wiki page it says to you select your TTS engine on linux, but I thought that SAPI5 was windows only
00:44:12n1sgreat, did you perchance run configure before building the compiler?
00:44:31LloreanXavierGr: espeak, festival, flite?
00:44:32H10_007quicki did
00:44:35n1sXavierGr: cygwin users can
00:44:42n1sH10_007quick: rerun configure then :)
00:45:25H10_007quickah thats better
00:45:28H10_007quickthanks alot
00:45:44XavierGrn1s: bah that's lame that means that I have to use both vmware and cygwin now :(
00:45:51n1sH10_007quick: the cause of those warnings is that gcc is a lot more picky since 4.0 and we simply disable them in config :)
00:46:15n1sXavierGr: or as Llorean suggested any narive linux engines...
00:46:47XavierGrn1s: yes but native linux engines don't do many languages except the most popular ones
00:49:08rasherXavierGr: you're free to invent a way to use sapi voices on Linux, but until it exists, it's hard to support them.
00:49:17 Quit ender` (" It's all fun and games until someone loses an eye. Then it's fun and games you can't see.")
00:50:04 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
00:50:27 Join miepchen^schlaf [0] (
00:50:51 Quit H10_007quick ("ChatZilla 0.9.79 [Firefox]")
00:51:13XavierGrrasher: would it be possible to modify the old vbs scripts to make them work with langv2?
00:52:01preglowXavierGr: what scripts?
00:52:10pregloweverything is possible, it just needs to be done
00:52:22XavierGrpreglow: under windows
00:52:39Nico_Proolku: sorry, I was on the phone for a while
00:52:40rasherIt'd take a fair amount of work
00:52:53preglowit's not trivial, let's just say that
00:52:55preglowbut very possible
00:53:08rasherAnd need some support from the build server (to save apps/features in a publicly available file)
00:53:15Nico_Proolku: the duplicate get_metadata calls for tracks 7 and 14 are because the first is to populate lasttrack_id3 and second is for the real buffering
00:53:19XavierGrok I'll see what I can do
00:54:26rasherI'd say it's better to do the work to support running standalone
00:54:47 Quit saratoga ("CGI:IRC (Ping timeout)")
00:54:53roolkuNico_P: okay, probably no point in re-using the struct in memory? Why are there 3 for track 14?
00:54:55rasherWhich again would require some help from the build server (perhaps it could generate a suitable .bat file during nightly builds)
00:55:22XavierGrit was a cool way to build voice menus nonetheless, quick and very easy, maybe I will start poking either these or
00:55:25Nico_Phmm 3 calls is suspicious
00:56:35 Join einhirn [0] (
00:58:09 Quit scorche|w ("CGI:IRC (Ping timeout)")
01:00:48XavierGrrasher: btw is there a reason why some languages on your page cant be clicked?
01:01:02rasherXavierGr: Because they're fully translated
01:01:29XavierGrrasher: ah yeah silly me, sorry :P
01:01:33 Quit spiorf (Remote closed the connection)
01:05:47jhMikeSpreglow: what was done in libmad to make it output 28 bits anyway?
01:06:13 Quit idnar (Nick collision from services.)
01:06:15 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
01:07:14preglowjhMikeS: nothing, that's how it's coded
01:07:22 Join radinp [0] (
01:07:30preglowjhMikeS: but i'm quite sure simple modifications to the qmf filter data can make the precision less
01:08:13jhMikeSI'm asking since it seems to say it outputs 24-bit
01:09:33preglowwell, that's bullshit
01:09:35preglowgimme a sec
01:10:18preglowjhMikeS: mpa.c sets sample depth to MAD_F_FRACBITS, that's defined as 28 in fixed.h
01:10:28preglowjhMikeS: where does it seem to say it outputs 24 bits?
01:10:46jhMikeSyeah, the source from the site says 28 too
01:11:10***Saving seen data "./dancer.seen"
01:11:16preglowjhMikeS: btw, do you need any more iram from speex? there's still stuff left that can be reclaimed
01:12:00jmworxpreglow: does the fixed-point wb encoder work for you on Coldfire?
01:12:08jhMikeSAs far as IRAM I think every unneeded byte should be reclaimed. Too valuable.
01:12:48jhMikeSpreglow: frankly I'm not seeing a problem with the dithering itself but it's possible to overflow since the additions are performed before clamping
01:12:57preglowjmworx: never tried it
01:13:27preglowjhMikeS: yep, know, and that's the next point i need to check out; does it really overflow
01:13:58preglowjhMikeS: if crossfeed alone doesn't overflow on arm, i'm somewhat skeptical
01:14:21preglowjhMikeS: and it didn't but the second i turned on dithering, clipping started on negative peaks
01:15:24preglowjmworx: i really haven't tried any of the encoders, there used to be libm stuff in there and i haven't had time since
01:15:26jmworxpreglow: Can you try it? Someone's reporting a problem I can't reproduce, so I'd be curious to see whether you get it.
01:15:50preglowjmworx: a problem on coldfire?
01:16:03jmworxNo, Blackfin, but using generic C code
01:16:17preglowjmworx: well, i'd need to actually enable the encoder bits and write a small encoder application, i don't really have time right now
01:16:28jhMikeSpreglow: clipping or wrapping?
01:16:31preglowjhMikeS: wrapping
01:16:48preglowjmworx: but it might be interesting anyway, i'll see tomorrow
01:18:24 Quit linuxstb ("ChatZilla 0.9.79 [Firefox]")
01:18:30jhMikeSpreglow: mpa won't pass DC will it?
01:18:49 Quit OlivierBorowski_ (Remote closed the connection)
01:18:57 Quit kugel (Read error: 104 (Connection reset by peer))
01:19:19preglowjhMikeS: hmm, i don't know, i've never seen it do it
01:19:29preglowjhMikeS: and from a compression standpoint, it'd be silly to even encode the dc coefficient
01:19:30 Quit n1s ()
01:19:43 Quit barrywardell ()
01:20:15markunXavierGr: if greek support in espeak is not perfect, perhaps you could help out with fixing it
01:20:50Nico_Proolku: committed... sorry for being so slow
01:22:24roolkuno problem, I am just glad it is fixed
01:24:21 Quit Zagor ("Client exiting")
01:25:16 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
01:25:20 Join barrywardell [0] (
01:25:31preglowjhMikeS: i'll check out what else can be removed from iram tomorrow, but i think another 1kb can go
01:26:13jhMikeS1KB? rather a massive amount imo.
01:26:16linuxstbBah, just missed Zagor...
01:26:22XavierGrmarkun: greek support in espeak? First time I hear that, I will definitely try this, thanks for pointing it
01:26:25 Quit petur ("Zzzzz")
01:26:35markunXavierGr: it might not support it, let me check
01:27:01linuxstbDoes this commit look wrong to anyone else?;r2=15746 Shouldn't the if() be testing for (settings_fd >= 0) ?
01:27:37markunXavierGr: yes, language "el"
01:27:52XavierGrmarkun: nice!
01:27:57jhMikeSlinuxstb: uh, yeah, it should
01:27:58markunit also supports use with a greek voice of mbrola
01:28:07preglowjmworx: nb_celp.c:200, you init a state member twice
01:28:19XavierGrmarkun: I will make sure to try it
01:28:22linuxstbjhMikeS: OK, I'll fix
01:28:24markunXavierGr: I hope it's not old greek :) And it will probably not sound as good as the english voice
01:29:01jmworxpreglow: oops
01:29:25markunjmworx: speexenc doesn't support resampling, right?
01:29:30XavierGrmarkun: haha you mean "ancient greek"? :P Well the loquendo SAPI5 voice file I have is the best tts I've ever heard, much better even from the english ones
01:29:38preglowmarkun: correct
01:29:49XavierGrmarkun: it seems to me that the greek language is quite tts friendly
01:30:31jmworxmarkun: No, it doesn't. Waiting for your patch :-)
01:30:39 Quit PaulJam (".")
01:30:56markunjmworx: well, all the code is already there, right?
01:31:02preglowjmworx: should be fairly quick too, with resample.c
01:31:11markunthat's what I mean
01:31:21 Nick Soap_ is now known as Soap (n=Soap@rockbox/staff/soap)
01:31:31preglowmarkun: it's very quick, i use that resampler in rbspeexenc.c, and it needs very few lines of code to do a good job
01:31:42jmworxmarkun, preglow: yes, there's already a resampler and a ring buffer, so it wouldn't be hard indeed.
01:31:56preglowmarkun: go go!
01:32:08markunnot now, must sober up a bit first :)
01:32:10jmworxAnd should probably be the default to prevent idiots from encoding at 44.1 kHz despite the warning (and then complaining it doesn't sound good).
01:32:52 Part briantumor ("Leaving")
01:33:17markunjmworx: why did you write the resampler?
01:33:33rasherpreglow: rbspeexenc build problem on the mailinglist
01:33:37XavierGrbah tried apt-get install espeak but the links that apt-get tried to resolve returned 404 :(
01:33:49rasherXavierGr: apt-get update
01:33:50Lloreanrasher: More of a dev-environment problem from the looks of it?
01:33:57rasherYEah, I just noticed
01:34:02jmworxmarkun: To get rid of linear interpolation resamplers once and for all.
01:34:23preglowjmworx: ubuntu distributes a wb 44.1khz spx file as part of its example content :P
01:34:25XavierGrrasher: yes trying that now :)
01:34:33markunjmworx: ... which we still use for rockbox :)
01:34:35jmworxpreglow: Damn idiots. Where?
01:34:55Lloreanjmworx: When you install, there's examples linked to from your desktop. Media files and such.
01:34:57preglowjmworx: in the Examples/ folder it puts in your home dir
01:35:14preglowand it does sound rather horrible
01:35:40 Quit Mouser_X (Nick collision from services.)
01:36:37preglowouch indeed
01:36:39 Join Mouser_X [0] (
01:37:01markunproject gutenberg uses wideband for librivox audiobooks, but the quality is pretty low
01:37:04LloreanI hadn't listened to it before, but yeah... ick
01:37:10 Nick fxb is now known as fxb__ (
01:37:12 Quit Nico_P (Remote closed the connection)
01:38:26preglowmarkun: this is one of their files
01:38:54LloreanIt's something like 38kbpsish 44.1khz audio
01:39:32markunpreglow: these ones are 16kHz
01:39:34jmworxI think in the mean time, I'll add a −−i-am-an-idiot option without which speexenc will refuse to encode at 44.1 kHz
01:39:46preglowjmworx: :D
01:39:57Lloreanjmworx: −−codec-abuse-time ?
01:39:59preglowsounds excellent to me
01:41:47jmworxI think that should be good enough as a deterrent
01:42:33jhMikeSwhat! it won't do 96kHz? ;)
01:42:33preglowi suspect it will drive the point home, if nothing else
01:43:03preglowsoon you'll find people encoding at 44099 hz to avoid this
01:44:16jhMikeSwhy would that get around it? it should proceed to format the HD if attempted.
01:45:19preglowmarkun: i'm not getting any bugreport on rbspeexenc build
01:45:25 Quit roolku ()
01:45:39markunpreglow: wrong person..
01:46:07jhMikeSmarkun: preglow types the wrong person alot :) (an autocomplete bug)
01:48:10 Quit lee-qid ("aufwiederbyebientotsayonara")
01:50:06preglowoh, that was just a memory error
01:50:09preglowrasher: i'm not getting any bugreport on rbspeexenc build
01:50:25XavierGrerr I got an error while I tried to make voice:
01:50:27preglowrasher and markun sound alike in my head :)
01:50:49rasherpreglow: It was on the -user list, but it's just another user lacking cc1.exe
01:50:50XavierGr"No rule to make target `inttypes.h', needed by `bits.o'. Stop."
01:51:12preglowrasher: can't help with that
01:51:59 Quit radinp (Read error: 113 (No route to host))
01:52:03preglowrasher: well, i hate cygwin, so can't fix that for him
01:52:23XavierGrIt showed that error trying to compile rbspeexenc as it seems
01:53:15 Join mschneider [0] (i=42444ee1@gateway/web/cgi-irc/
01:54:17LloreanBagder: do you have shell access to the forums server, or is it only LinusN and Jeff?
01:57:52*jhMikeS was getting that error on Debian VMWare early on
01:59:01 Join K-Man666 [0] (
02:00:21jhMikeSXavierGr: Will is make from within tools/rbspeex?
02:00:22XavierGrdoes rbspeexenc.c has any dependencies?
02:00:35LloreanIt seems the mysql version being used by our forum server needs upgraded.
02:00:45LloreanThe bug is fixed in 4.0.28, we're 4.0.27. =/
02:00:54XavierGrjhMikeS: you mean if I run the make command from tools/rbspeex?
02:01:08jhMikeSit works here
02:01:10XavierGrno I did configure then make voice on the target folder
02:02:17XavierGrI don't have speex installed though
02:02:24XavierGr(if that matters)
02:03:16mschneiderSorry to interrupt, but is there any way for me to continually run a program in the background while running rockbox?
02:03:24XavierGrnope still the same error after apt-get install speex
02:03:34jhMikeSno, it's nothing to do with that
02:04:03Lloreanmschneider: Look at the battery_bench plugin
02:04:31markunXavierGr: speex or espeak?
02:04:40XavierGrwow! now I get errors just trying to build tools on make bin!
02:05:08XavierGrdamn it, it seems that apt-get update and then apt-get install espeak messed somehow my compiler?
02:05:33jhMikeSXavierGr: I'd just try 'make' from tools/rbspeex and see if it actually works from theere
02:06:37XavierGrjhMikeS: I will try that too, but I think that I just messed up my vmware image
02:06:48XavierGrpitty and it was working quite nice :\
02:07:49preglowjmworx: is DISABLE_VBR only meant as a temporary measure to allow building fixed point encoders, or is it supposed to stay?
02:07:49jhMikeSapt-get is a constant pile
02:07:52preglowi sassume the first
02:08:35*jhMikeS wonders why no build is triggering a rbspeexenc build
02:08:38XavierGrI just hope that I had a backup of the whole image
02:08:38 Quit mschneider ("CGI:IRC (EOF)")
02:08:52XavierGrit was a custom one with the sims cross compiled for windows
02:09:09jmworxpreglow: It's meant to stay for code size reduction purposes, even when the VBR code is converted
02:09:49jhMikeSoptions (A)+(V) should make voice stuff right?
02:10:18preglowjmworx: okiedokie
02:10:24preglowjhMikeS: aye
02:10:55jhMikeSI don't see rbspeexenc even coming up in make tools
02:10:55preglowjmworx: same with DISABLE_FLOAT_API, i take it?
02:11:42jmworxpreglow: DISABLE_FLOAT_API is meant to stay exactly as it is, because those parts just cannot be converted to fixed-point
02:11:56jhMikeSit's not getting built...hmmm
02:12:56rasherjhMikeS: make voicetools
02:14:41 Join webguest53 [0] (i=cb3b81c3@gateway/web/cgi-irc/
02:15:03webguest53Forums say "Database Error: Please try again. If you come back to this error screen, report the error to an administrator."
02:15:33jhMikeSaha...make help doesn't show that :)
02:15:42rasherjhMikeS: Blame linuxstb
02:15:45 Quit einhirn (Read error: 104 (Connection reset by peer))
02:15:57jhMikeSsvn blame linuxstb
02:16:29 Join erchino [0] (n=erchino@
02:16:39jhMikeSok, it did build ok for me though
02:16:47erchinohoe do i uninstall the bootloader???
02:17:14Lloreanerchino: The way the manual says?
02:17:17rashererchino: That's described in full detail in the manual.
02:17:34preglowjmworx: lingering mention of EPIC_48K in include/speex/speex.h
02:18:27jmworxpreglow: thanks
02:23:21 Part pixelma
02:24:02erchinoit says unsupported uninstall?atioin...
02:24:38erchinoon rbutil
02:24:45 Join Thus0 [0] (n=Thus0@
02:26:24erchinohelp please...
02:26:33erchinoi installed manually
02:26:40*Mouser_X has never used rbutil.
02:27:06*preglow wonders if rockbox will ever need jitter.c
02:27:43erchinoit says unsopported bootloader uninstall method...
02:28:29 Quit homielowe (Read error: 104 (Connection reset by peer))
02:28:47 Join homielowe [0] (
02:29:31Mouser_XWhat target do you have?
02:29:36*Mouser_X assumes iPod.
02:29:50jhMikeSpreglow: is that for network transmission?
02:29:50XavierGryay! fixed it
02:30:10XavierGrafter installing espeak I had to install again gcc and g++
02:30:27XavierGrthen make sure to build all the rockbox tools, then try to make voice
02:30:44XavierGrelse you will get errors
02:31:38preglowjhMikeS: yes
02:31:43preglowjhMikeS: for reordering packets and such
02:32:00preglowjhMikeS: don't get me wrong, it's not linked in, i just wonder if i should bother to continue syncing it
02:33:21 Quit barrywardell ()
02:33:42Lloreanerchino: What did you use to install Rockbox?
02:34:23erchinoi think it was terminal on mac
02:35:12LloreanWhat program?
02:35:47erchinoconsole on mac
02:36:38LloreanYes, but you can't install it without certain utilities.
02:36:42LloreanWhich instructions did you follow?
02:37:16erchinosome on in this channel helped me
02:37:21erchinoto install it
02:37:26erchinobut now I want to uninstall
02:37:29erchinoand I cant...
02:37:41LloreanYes, and to offer you advice, I need to know what you actually *did* to install it.
02:37:57preglowwoo, two green deltas for me today
02:38:01erchinoi gotr in to console...
02:38:18erchinoand he told me get in to the ipod root somethihng like that
02:38:42erchinoand told me to type something manually to extrac the rockbox files to the ipod manually
02:39:15LloreanIt's really not helping if you're just telling me "I did something, and something."
02:39:23LloreanDid you use ipod_fw? ipodpatcher? dd?
02:39:32erchinoI download it
02:39:41LloreanThe answer to all of those questions can't be "yes"
02:39:43erchinobut I couldnt extract it to the root of the ipod
02:41:27rasherXavierGr: it's generally a good idea to pay attention to what (if anything) gets removed when you apt-get install stuff..
02:41:45XavierGrrasher: indeed
02:42:07amiconnThat's why I prefer a graphical tool like synaptic
02:42:21rasherI don't see the difference
02:42:28rasherapt-get presents a list of what it's going to do
02:42:32erchinothat is totally true...
02:42:37erchinonow im stuck...
02:43:19rashererchino: Did you use ipod_fw, ipodpatcher or dd?
02:43:35erchinoi dont even know that is those things...
02:43:52Lloreanerchino: You never answered my question. Which of those programs did you use?
02:44:02erchinono idea...
02:44:10erchinoI think it was ipodpatcher
02:44:16LloreanThen try using ipodpatcher to uninstall it
02:44:24LloreanIf you run ipodpatcher it should provide you that option.
02:45:22erchinolet me try
02:45:40erchinoyes I did use that
02:45:48erchinobut hhnow do i uninstall it?
02:46:05LloreanIf you run ipodpatcher it should provide you that option.
02:47:27preglowjmworx: speex_memset() in os_support.h currently reuses OVERRIDE_SPEEX_MOVE, that's got to be wrong
02:48:17XavierGrwhat the hell...
02:48:46XavierGrit just spells out the letters in the worst voice quality I've ever heard
02:48:56XavierGrand no not in folders or files
02:49:04XavierGrmenu itens
02:49:47XavierGrwell not too weird that the greek.voice file is 4.69MB. If the entries are spelled out each sentence is HUGE
02:50:35XavierGror maybe I just set up the voice file with the wrong options
02:52:24 Quit erchino ()
02:52:25 Join MacSob [0] (i=442ea717@gateway/web/cgi-irc/
02:53:44MacSobhey all, I just installed ROckbox on my iPod video, everything looks and workds great... one problem I do have is my iHome dock remote control doesnt seem to work. Any ideas?
02:54:09MacSobsmall thing, I can live without it, just wondering if I missed something
02:54:13preglowjmworx: somewhat updated patch,
02:55:00 Join donutman25 [0] (n=chatzill@
02:56:01MacSobuser manual has instructions to apply?
02:56:13 Join psycho_maniac [0] (
02:56:39*preglow bed
02:56:52krazykitMacSob, there's a wiki page that have various docks and peripherals listed. it's linked from the ipod page.
02:57:20MacSobthanks, ill go check it out. I knew I missed something
02:58:03MacSobi must say that this is tons better than the original firmware :) but im sure you all know that
02:58:57 Quit MacSob ("CGI:IRC 0.5.7 (2005/06/19)")
02:59:11 Join Llorea1 [0] (n=llorean@
03:06:02 Join mud-rb [0] (
03:08:02 Join keanu [0] (n=keanu@unaffiliated/keanu)
03:09:46 Quit Llorean (Read error: 110 (Connection timed out))
03:10:12 Nick Llorea1 is now known as Llorean (n=llorean@
03:10:56 Quit psycho_maniac (Read error: 113 (No route to host))
03:11:05 Join psycho_maniac [0] (
03:11:11***Saving seen data "./dancer.seen"
03:11:16K-Man666Hey, Im Kaspar Rothenfusser (which is also my wikiName) I just wanted to ask for a writing permission to the wiki as I am willing to put me in the RockboxTesting list for the sansa c200v1 and e200v1 which I've both been owning now for a (rather) short time
03:13:26 Quit Rob2222 (Read error: 104 (Connection reset by peer))
03:15:35tkoodahow the heck do I clear the playlist?
03:15:43LloreanPress stop?
03:15:56tkoodawhat "stop"?
03:16:03tkoodano stop button on gigabeat F40
03:16:06keanuquestion: since people can't agree on an e200 keymap, would it be worth having a patch to allow users to switch between keymaps (assuming it'd be possible)
03:16:11LloreanNo, but there's a button that stops playback
03:16:20LloreanSeparate from the button that pauses playback
03:16:21XavierGrwell espeak on greek works decently but too computerized for my taste
03:16:24LloreanAs is covered by the manual
03:16:36XavierGrLoquendo has such a nice tts, pity that it isn't free
03:17:00Lloreankeanu: There really aren't a large number of people complaining.
03:17:45LloreanNot to mention the somewhat long standing no-option-to-change-buttons policy.
03:18:44tkoodaLlorean, tapping the power button on a gigabeat F40 stops playback, but I still can't find any way to clear the playlist (whilst playback is stopped either)
03:20:05keanuLlorean, I understand how devs wouldn't want to change buttons, but I was thinking something similar to kbd files for the text editor. Granted, those two are different, but simply having the option for users to change keymaps would prevent everyone from complaining
03:20:50mud-rbtkooda: once playback is stopped, and music you insert starts a clean playlist with only that music
03:21:00mud-rbtkooda: it's all pretty clear in the manual, by the way
03:21:16jmworxpreglow: I'm wondering whether a SPEEX_MOVE() macro would be more appropriate. Could be something like:
03:21:29 Join DM| [0] (
03:21:35tkoodamud-rb, aah thanks.
03:21:48tkoodamud-rb, I found the manual to be unclear, btw.
03:22:06jmworx#define SPEEX_MOVE(dst, src, N) memmove(dst, src, N*sizeof(*dst)) and even check that the pointers are compatible
03:22:26Lloreankeanu: It would also add a very significant amount of complexity to the cod.e
03:23:01Lloreankeanu: If people want to create their own whole keymap, the file keymap-e200.c in /apps/keymaps is there for them.
03:23:11keanuLlorean, ok, so while possible, it wouldn't be worth it?
03:23:14amiconnThe main reason why we do not want configurable buttons is support
03:23:17LloreanEditing it is honestly about as simple as you can make it anyway
03:24:28 Quit donutman25 ("ChatZilla 0.9.79 [Firefox]")
03:24:36Lloreanamiconn: Not to mention the fact that you'd either need to dumb it down to the point where people just pick which button is going to replace "Select" and then it gets all select's old functions, or you'd need to make it complex enough to be able to pick what a key does in every single possible situation in which you can press that key
03:25:28keanuYes, but for users who aren't familiar with building/compiling may get annoyed. I'm not trying to start a huge argument, but it seems that some think Rockbox keymaps should be similar for all targets, while others think keymaps should be similar to the OF. It seems that there isn't much of a compromise otherwise, which is why I was wondering if it'd be worth having they keymaps changeable
03:25:30 Join Febs [0] (
03:26:10amiconnOne problem would be users who misconfigured their buttons and then are unable to revert, but the biggest problem would be to explain all the interdependencies and hardware limitations
03:26:17Lloreankeanu: Honestly, is it so much trouble to spend a little while getting used to a new keymap, when you've got about a thousand new features as well?
03:28:20krazykitor a little time learning to build rockbox and work svn enough to revert the keymap.
03:28:28 Join BHSPitLappy [0] (n=Me@unaffiliated/bhspitmonkey)
03:29:15keanuLlorean, I can agree with that - since Rockbox supported sound on the e200 I've only been using the OF for USB. I can't think of anything to compare it to at the moment, but I think it boils down to older users being frustrated about how what keymaps once worked no longer do
03:29:38Lloreankeanu: So basically, I should've changed it sooner?
03:29:50LloreanI'm all for mapping the keys based on the restrictions and or benefits of specific hardware, but I don't think that taking it to the point of attempting to duplicate the original hardware is a good idea. I think using it for inspiration isn't bad, but I really like to try to keep the keys consistent between screens to how they go between screens in Rockbox rather than the OF
03:30:07keanuLlorean, I think if it was changed sooner, it would have been easier for people to adjust
03:30:16LloreanThis makes supporting it easier, as well as designing new parts of it cross-platform
03:30:51LloreanYou know that the ACTION_STD_CANCEL is likely to be in a good place for a stop or cancel key, so you can depend on it being "good" a little more
03:31:29LloreanAnd of course, when someone's old Rockbox player dies, and they buy a new one, the learning curve is a lot less steep. Once they learn which button does ACTION_STD_CANCEL, they know how to use everything that uses button actions for its keymap
03:36:45keanuHmm...good points. To developers, having keymaps consistant makes perfect sense (and to those with multiple DAPs) but to casual users, not so much. I think I may have something to compare the keymaps to - the terminology differences between browsers (like IE and Firefox - Temporary Internet Files vs Cache for example) - I think if there were a chart/table comparing the new keymaps to the old, it'd help some users. I personally had quite a bit of tro
03:36:45keanuuble finding the context menu with the new keymaps
03:37:17Lloreankeanu: I updated the manual
03:37:21LloreanIt shouldn't have been hard at all.
03:37:30LloreanI liken it to the left and right mouse buttons
03:37:46LloreanAny time you buy a new mouse, you know if it has two buttons on it, one will be left, and one will be right
03:37:58LloreanOnce you learn which is which, then you know what they'll do in most programs, if you already know the program too
03:38:22LloreanIf you buy a new dap, there will be a "Select" (or equivalent), a "Menu" (A-B on irivers, etc, but a button that summons the menu), a Left, a Right, a Play/Pause, etc.
03:38:30LloreanOnce you know what those are, the rest of Rockbox works as expected.
03:38:37LloreanYou don't have to learn which does what in every single screen.
03:38:45LloreanI know a lot of people only own one DAP
03:38:54LloreanBut an awful lot of Rockbox users buy a new Rockboxable DAP when their old one dies.
03:39:05 Join EternalRains [0] (
03:39:13LloreanThe only people mirroring the OF really helps is people who expect to only own one Rockbox DAP ever.
03:39:48LloreanAnd even then, I think in some cases mirroring the OF means you have to make compromises on some functionality. Like the nearly-missing "Stop" button on the Sansa previously.
03:39:49keanuLlorean, I'm having trouble finding it in the manual - what chapter is it in?
03:40:07Lloreankeanu: I do believe the keymap is in "Getting Started"
03:40:20alienbiker99your definately right Llorean. i like having similar keymaps throughout rockbox, so switching between daps is easy to remember
03:40:39LloreanOr maybe quickstart
03:41:01*Llorean is really not sure.
03:41:20LloreanThe platform file for the manual that you edit to update the keymaps in it doesn't actually map to any specific page, because it's just macros to replace every mention of that key
03:41:50LloreanIt's referred to as the "File Menu" in the manual, which I'm not terribly fond of.
03:42:20LloreanHm, there's an error there
03:42:47keanuok, thanks
03:43:00LloreanBut the error is for "Stop"
03:43:04LloreanThe context menu seems to be correct
03:43:16keanuyeah -
03:43:26keanu"Stop" is menu, not long play
03:44:25keanuthat definitely helps, though I'm thinking about comparing everything between the keymaps and making a table with the changes
03:44:25LloreanStop is Power
03:44:58 Part webguest53
03:46:38 Quit EternalRains ("Konversation terminated!")
03:46:51 Join EternalRains [0] (
03:47:48LloreanI still wonder about the person who said the menu button is the button he presses the most.
03:47:57 Quit ryanakca (Remote closed the connection)
03:48:02 Join ryanakca [0] (n=ryan@ubuntu/member/ryanakca)
03:49:48 Join TheBudMan [0] (
03:50:43TheBudManHello room....Where can I download the latest rockbox installer x?
03:51:05LloreanThat's unofficial, unsupported, and outdated software.
03:51:11LloreanTry using the official install method.
03:52:06TheBudManwhich is?
03:52:14LloreanIn the manual?
03:52:39TheBudManSorry I'm new to this
03:52:46 Join [1]psycho_maniac [0] (
03:52:59LloreanI would suggest visiting the Rockbox website, then
04:02:18 Quit psycho_maniac (Nick collision from services.)
04:02:18 Nick [1]psycho_maniac is now known as psycho_maniac (
04:03:32 Quit EternalRains ("Konversation terminated!")
04:03:34 Quit TheBudMan ()
04:18:16 Join roxfan [0] (
04:20:04 Join saratoga [0] (i=47e276af@gateway/web/cgi-irc/
04:20:24saratogai haven't tried it on a device yet, but the new wma seeking patch I put on the tracker works pretty well on the sim
04:21:33psycho_maniacflyspray number please.
04:24:25psycho_maniacnow i just have to find some wma files
04:24:53saratogai don't have my cable handy so i didn't try it on device
04:25:02saratogaso no promises it will work as well there
04:25:04saratogathough it should
04:25:13psycho_maniacill try it on my ipod video
04:27:51 Quit Llorean ("Leaving.")
04:28:00 Quit mud-rb (Remote closed the connection)
04:30:44 Quit roxfan[zzz] (Read error: 110 (Connection timed out))
04:30:56 Join Llorean [0] (
04:33:41psycho_maniacwow i havnt updated svn in a long time i guess.
04:34:36 Quit saratoga ("CGI:IRC (EOF)")
04:40:06 Quit miepchen^schlaf (Read error: 110 (Connection timed out))
04:40:14 Join miepchen^schlaf [0] (
04:42:32 Quit XavierGr ()
04:43:06 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
04:53:13 Quit z35 ("Leaving")
04:55:13 Quit stewball`ghost (Read error: 104 (Connection reset by peer))
04:59:22 Join sarixe [0] (
05:00:54 Join TwilightInZero [0] (
05:02:59 Join webguest58 [0] (i=cb3b81c3@gateway/web/cgi-irc/
05:05:01 Quit webguest58 (Client Quit)
05:08:10TwilightInZeroMkay, before I ask my question... is this channel strictly for Rockbox itself, or does it pertain also to the hardware that runs it, like iPods?
05:08:34LloreanIt only pertains to the hardware that runs it in the context of running it and developing it for that hardware.
05:08:50TwilightInZeroSo if I'm having problems with the hardware, I can't ask for assistance here, then?
05:09:24LloreanDo the problems have anything to do with Rockbox?
05:09:39TwilightInZero...I will move to the other channel.
05:10:00TwilightInZeroSee you there... maybe?
05:10:10DogBoylinuxstb, yea I "misspoke" when I said better cook support, I meant any
05:11:13***Saving seen data "./dancer.seen"
05:12:23 Quit billenium ("Ex-Chat")
05:16:11 Join XavierGr [0] (
05:17:12psycho_maniacsaratoga: this patch is messed up on my player.
05:18:45 Quit keanu (Read error: 104 (Connection reset by peer))
05:21:15DogBoylinuxstb, though I have used realplayer on a palm tungsten e2
05:24:13 Quit K-Man666 ("ChatZilla 0.9.79 [Firefox]")
05:24:32 Quit Thundercloud (Remote closed the connection)
05:33:19 Quit jhulst_ ("Konversation terminated!")
05:55:11 Quit DM| ("*bashes head against keyboard*")
05:58:10 Quit XavierGr (Read error: 110 (Connection timed out))
06:06:11 Nick ionvas__ is now known as ionvas (
06:12:26 Join Jon-Kha [0] (
06:30:21 Quit TwilightInZero ("ChatZilla 0.9.79 [Firefox]")
06:35:12 Quit scorche (Read error: 104 (Connection reset by peer))
06:37:41 Quit psycho_maniac (" HydraIRC -> <- In tests, 0x09 out of 0x0A l33t h4x0rz prefer it :)")
06:39:32 Join scorche [0] (n=scorche@rockbox/administrator/scorche)
06:40:40 Join billenium [0] (
06:45:36 Quit joshin (Read error: 110 (Connection timed out))
06:58:39 Nick ivan` is now known as ivan (n=ivan`@unaffiliated/ivan/x-000001)
07:08:28 Quit tchan (SendQ exceeded)
07:10:02 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
07:11:15***Saving seen data "./dancer.seen"
07:12:18 Quit tchan (SendQ exceeded)
07:13:42 Join DM| [0] (
07:22:41 Quit krazykit ("leaving")
07:25:18 Join jhulst_ [0] (n=jhulst@unaffiliated/jhulst)
07:32:10 Join LinusN [0] (i=linus@rockbox/developer/LinusN)
07:37:01 Nick fxb__ is now known as fxb (
07:40:07 Join Rob2222 [0] (
07:43:06 Quit sarixe ("(EE) Failed to load "quit" module")
07:55:05 Nick fxb is now known as fxb__ (
07:59:22 Join J-23 [0] (
08:04:12 Nick billenium is now known as BilleniumZzZ (
08:17:11 Quit jhulst_ ("Konversation terminated!")
08:21:34 Join CaptainSquid [0] (
08:31:04 Quit BigBambi ("Please insert girder")
08:31:24 Quit J-23 (Remote closed the connection)
08:36:08 Join qweru [0] (
08:38:38GodEaterwhy did the DRM thread get locked ?
08:39:23LinusNit is?
08:40:13LloreanAsk Soap
08:40:44GodEaterhe locked it ?
08:41:02*GodEater has been trying to work out how you tell who locks threads,
08:41:13LloreanI have a button
08:41:18GodEaterand when are you going to ban ONE_HUMAN again ;)
08:41:19Mouser_XIf someone figured out how to play DRMed music, and posted a patch on the tracker, would it be removed by admins? (I assume that it would, for legal reason.)
08:41:26LloreanIt tells me all the "moderations" that have been performed
08:41:31LloreanThreads deleted, moved, or locked.
08:41:34GodEaternice ubtton :)
08:41:38GodEaterer, button
08:41:44LloreanI also have one that tells me all the errors that have been logged.
08:41:51LloreanApparently mistyping your password qualifies as an error.
08:41:59LloreanSo I know who are particularly bad at typing their own passwords
08:42:18GodEaterMouser_X: yes it would - it would be software to violate a license agreement, which we're against.
08:42:57scorcheLlorean: do tell! :P
08:43:10LloreanWe don't want people violating the GPL so we should do our best not to go violating other peoples' licenses
08:43:42 Join Rob222241 [0] (
08:43:50GodEaterscorche: I have so many variations on the same password from place to place I often get it wrong when I sign into the forums from a new computer
08:44:22LloreanBesides, in all honesty, if we don't like DRM, the right way to fight it is to go out of our way *not* to support it, so people fed up with it stop paying for it. In my opinion.
08:44:36Mouser_X^ Good point.
08:44:45LloreanCircumventing it takes the edge off the restrictions they put on the user
08:44:47GodEaterLlorean: indeed - which is the point I was trying to make to Hef
08:44:54GodEaterit appeared to fall on deaf ears :(
08:45:39GodEaterhe didn't seem to understand the irony of his complaint
08:45:41*scorche already knew it was soap >_>
08:46:08 Quit DM| ("*bashes head against keyboard*")
08:46:09scorchesilly PM...
08:46:12GodEaterI didn't think the discussion was done
08:46:38Mouser_XAm I supposed to be able to access "" yet?
08:46:41LinusNi'm surprised it was locked
08:46:52scorcheMouser_X: no
08:46:53Mouser_X(It's not working, but I might have typed it wrong.)
08:47:04GodEaterLniusN: me too - it wasn't offensive, and was quite informative I thought
08:47:06scorcheno it isnt...
08:47:14scorcheDNS has had plenty of time to propagate..
08:47:23Mouser_XAh. I saw some minor discussion about it when I logged into IRC, so I wasn't sure.
08:48:40LloreanGodEater: Go ahead and unlock it?
08:49:05GodEaterwell I was going to - but I wanted to find out why it got locked in the first place in case I was stepping on someone's toes
08:50:43 Join webguest31 [0] (i=cb3b81c3@gateway/web/cgi-irc/
08:51:06 Join petur [0] (n=petur@rockbox/developer/petur)
08:55:12GodEatercan't bloody unlock it now - forums are in "swimming through pitch" mode
08:55:53 Nick fxb__ is now known as fxb (
08:59:42 Join freader [0] (i=c27f0812@gateway/web/cgi-irc/
08:59:42 Quit webguest31 ("CGI:IRC (EOF)")
09:00:44freaderIs the latest implementation of file_exists() correct? I think it will also return true if the specified path refers to a dir (and not to a file). What is the intended semantics of the func?
09:01:32amiconnjhMikeS: around?
09:01:40 Quit Rob2222 (Read error: 110 (Connection timed out))
09:02:40GodEaterfreader: that sounds like correct behaviour to me
09:02:57 Join ender` [0] (
09:06:30amiconnfreader: Why would it?
09:06:52amiconnopen("/thats_a_dir") should fail with EISDIR
09:11:16***Saving seen data "./dancer.seen"
09:12:43 Quit BHSPitLappy (Remote closed the connection)
09:16:50 Quit nanok (Read error: 113 (No route to host))
09:22:37linuxstbHas anyone tried WMA seeking? It seems to freeze whenever I seek by more than a couple of minutes in my 2h (110MB) test track...
09:22:59GodEateris it at least an interesting 2 hour long wma ? :)
09:23:02 Join saratoga [0] (i=47e276af@gateway/web/cgi-irc/
09:23:31linuxstbSure, a John Peel show...
09:24:04*GodEater googles John Peel
09:24:14*linuxstb hopes GodEater is joking...
09:24:20GodEatersadly not
09:24:22GodEateroh him
09:24:27*scorche joins in for the googling
09:24:47*GodEater suspects linuxstb will let scorche off for googling
09:25:03linuxstbOnly because I'm still in shock you didn't know him ;)
09:25:17GodEateryou expect scorche to know a Radio 1 DJ (deceased) ? english thing :)
09:26:00saratogalinuxstb: I only had tracks < 5 minutes for testing
09:26:17linuxstbsaratoga: Seeks fail consistently with my test track.
09:26:20GodEaterlinuxstb: sounds like you need to spread the good word of John Peel then ;)
09:26:44saratogawhat happens when it fails?
09:27:35scorcheBagder: around?
09:27:40 Quit saratoga (Client Quit)
09:28:01linuxstbsaratoga: (hoping you read the logs) It just freezes, sometimes with "seeking" displayed, but not always. Buttons are unresponsive.
09:35:03 Quit qweru ("moo")
09:37:01*linuxstb also notices a post in flyspray reporting problems with the last WMA seeking patch and wonders if that patch was what was committed
09:38:42freaderamiconn: I think, it's possible to open a dir with open(). In UNIX, everything is a file, even dirs.
09:40:10freaderamiconn: that's why at some places in code, not open() but opendir() and then iteration over entries is used (with attrs testing)
09:40:18linuxstbYes, you can open a directory read-only with open()
09:40:40linuxstbNot sure about Rockbox's implementation though...
09:46:10 Quit atsea-63 (Remote closed the connection)
09:50:14freaderlinuxstb: I think it's different in the sim and on target. In sim, you can't, on target you can :-/
09:51:21 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
09:51:40linuxstbThen it sounds like a bug (or at least, variation from POSIX) in Rockbox's open implementation - in the sim, I would expect it to use the system open()
09:53:43 Join t0mas [0] (n=Tomas@rockbox/developer/t0mas)
09:54:02Bagderlook, a t0mas!
09:54:13*t0mas hides
09:54:29 Quit jhMikeS (Nick collision from services.)
09:54:33scorcheBagder: i flushed him out =>
09:54:35 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
09:54:51scorche(have a look at /cs info #rockbox)
09:55:24t0masbtw.. Bagder and Zagor had accesslevel 29 too?
09:55:47scorchei plan on shuffling that around a bit
09:55:55scorche(not removing theirs)
09:56:40 Quit GodEater (Read error: 104 (Connection reset by peer))
09:57:58 Part ionvas
09:58:33 Join GodEater [0] (n=bryan@rockbox/staff/GodEater)
09:59:19 Part freader
10:00:27 Join DerPapst [0] (
10:01:00DerPapstscorche: my sansa arrived today :-)
10:01:11scorchegoody :)
10:01:16 Join radinp [0] (
10:01:21Bagdernot sansa view by any chance? ;-)
10:01:34DerPapstsmaller than i thought ;)
10:02:28markunhi t0mas!
10:02:30DerPapstwill try to install rb on it this evening.. and then complain to you guys if i "broke" it ]:->
10:03:42 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
10:05:54scorchethere shall be many ops around here soon..
10:06:19peturwas there a shortage recently?
10:06:27*petur checks the log
10:06:53Bagderwe shall all be ops and kick everyone!
10:07:10*petur kicks Bagder first
10:07:35scorchepetur: it has been requested many times and been on my list of things to do, but all devs shall be ops (we already trust em to not delete svn ;)
10:07:50peturnext, we can start throwing empty beer bottles :)
10:08:05JdGordonscorche: deleting svn and random kick bans need different trust...
10:08:09JdGordonsvn can be reverted :p
10:08:20scorcheJdGordon: there will be access control
10:08:21*JdGordon slaps petur for 2 weeks ago....
10:08:32JdGordontripel karmilat sucks...
10:08:39peturno way
10:08:43JdGordonkarmelet even
10:08:50peturkarmeliet even
10:09:02pixelmathen it maybe was a fake ;)
10:09:08JdGordonit tastes the same as the cheap german crap thats 1/3 the price!
10:09:17JdGordonthat could be it!
10:09:22peturno way at all
10:09:30JdGordonbut goign to the belgium beer garden was fun anyway
10:09:52scorcheanyway, i am off to can highlight me if you want to call my attention to something..
10:10:11JdGordonscorche: oy! ignore this... im being mean :D
10:10:22*petur lights torch on scorche
10:10:36scorchepetur: i knew JdGordon would, but you too? ;)
10:10:56*scorche was waiting for JdGordon to do something before he went for real..
10:11:03JdGordonsure... sure....
10:11:33*petur hands torch to JdGordon and goes back to work
10:11:39scorcheJdGordon: face are predictable :)
10:12:30 Quit Nico_P (Remote closed the connection)
10:14:01*JdGordon hopes this patch gives a big green delta... getting sick of fixing these errors
10:14:51 Join atsea- [0] (i=atsea-@gateway/tor/x-6cf5506901e88fee)
10:18:05 Quit atsea- (Read error: 104 (Connection reset by peer))
10:21:19 Join lee-qid [0] (
10:22:48t0masmarkun! :D
10:23:30*t0mas has missed coding-rockbox :)
10:26:55 Quit TMM ("Ex-Chat")
10:27:21amiconnhi DerPapst
10:28:35amiconnDerPapst: Are you running current rockbox on your 3rd gens?
10:31:19 Quit Isolinear ()
10:32:07 Quit radinp (Read error: 113 (No route to host))
10:34:31 Part tkooda ("thanks")
10:39:05 Join atsea- [0] (i=atsea-@gateway/tor/x-09a1319f67aed03e)
10:46:25DerPapstamiconn: about 2 weeks old i guess.
10:46:31DerPapstbut i can update
10:49:44*DerPapst compiles
10:51:11amiconnThere seems to be a problem with stability on PP5002
10:52:21amiconnIt *almost* looks like a stack overflow, as I get all sorts of bad things happening: Data aborts in DRAM or IRAM, undefined instructions all over the place, Panics, and freezes
10:52:43DerPapsteh.. sounds fun :-/
10:53:10DerPapstbah.. getting compiling errors :-/
10:53:53amiconnThe strange thing is that it doesn't happen with every build. Looks like some alignment covers the problem from time to time
10:54:30amiconnUnfortunately I can't tell precisely when this problem was introduced. Somewhere between August and now
10:54:48 Join Isolinear [0] (
10:54:52amiconnI first thought it might be a hardware problem of my 2nd Gen
10:55:33DerPapsti can't compile... this is odd.
10:56:54pixelmaDerPapst: after a "make clean" and reconfigure?
10:57:09DerPapstgetting errors in firmware/target/arm/system-arm.h: In function `queue_enable_queue_send'
10:57:31DerPapsti even hat "deleted" firmware/ and checked it out again
11:00:26amiconnWrong toolchain? (ipl maybe?)
11:01:11LinusNDerPapst: which compiler version do you have?
11:01:14*DerPapst hides
11:01:39DerPapsti thought the ipl toolchain is prefixed arm-uclinux-elf-gcc
11:01:48DerPapstbut obviously it's not :P
11:02:53jmspeexpreglow: ping
11:09:25linuxstbDerPapst: The kernel toolchain is, but the userland one is arm-elf-gcc as well...
11:10:06DerPapstcompiling works now.
11:10:21DerPapsthave to figure out how to switch between those two
11:10:39linuxstbRemove the IPL one and come and join Rockbox full-time ;)
11:11:18***Saving seen data "./dancer.seen"
11:11:19Bagderisn't ipl dead and abandoned anyway? ;-]
11:11:50DerPapstprogress is slow... though progress is there :P
11:12:37linuxstbIs anyone seriously looking at the new ipods?
11:12:59DerPapstno. but fixing old bugs
11:13:33DerPapstor enhancing podzilla and modules :)
11:13:47 Join spiorf [0] (
11:14:40DerPapstwe have a new version of mikmodule wich plays up to 50 voices on pp502X including nice visualizers :)
11:17:30Bagdercowon Q5W =>
11:17:48Bagdermonster 5" 800x400 touch screen
11:18:11amiconnI think the ipl situation is even worse than rockbox', regarding the "working on all kinds of fancy features, while the most basic things (kernel in case of ipl) is buggy, and no-one seems to be working on fixes & enhancements"
11:18:13DerPapsthave to run now. amiconn: will throw the new build on my 3G later. but i'm going to berlin today (530km) so can do some testing this evening i think
11:18:18Bagderclaims ape and musepack in the OF
11:19:02DerPapstamiconn: we only have one kernel guy currently.. and he works on some kernel fixes.
11:19:03LinusNBagder: woo
11:19:14Bagderamiconn: I agree with that perception, I think the kernel hackers are all gone and only apps hackers are left
11:20:11amiconnjhMikeS: ping
11:40:16 Quit crwl ("ropelihommat")
11:40:41 Join barrywardell [0] (
11:45:15 Nick fxb is now known as fxb__ (
11:46:55 Quit lee-qid (Read error: 110 (Connection timed out))
11:47:31Bagderit seems the e200 v2 series is out now
11:47:45Bagderpeople in the abi forums pop up with them...
11:55:25markunBagder: do we know what's inside yet?
11:57:45preglowjmspeex: pong
11:58:15jmspeexpreglow: #define SPEEX_MOVE(dst, src, n) (((dst)-(src)), memmove((dst), (src), (n)*sizeof(*(dst)))
11:58:23jmspeexwhat do you think of that instead?
11:59:04preglowjmspeex: i actually did think about that, and it's not a bad idea
11:59:17preglowmakes the code clearer
11:59:32preglowand we should definitely check that the pointers are of the same type, or at least type size
12:00:03 Quit rasher ("leaving")
12:00:12jmspeexactually, it really checks the type. (int*)-(float*) produces an error
12:03:20 Quit linuxstb ("Leaving")
12:04:57preglowoh yeah
12:04:58preglowthat's true
12:08:32 Join rasher_ [0] (n=rasher@rockbox/developer/rasher)
12:09:26 Join spiorf_ [0] (n=spiorf@
12:13:21 Join crwl [0] (
12:15:35pixelmathe swcodec recording screen does not tell exactly how many seconds are already buffered if you set prerecording (only showing "Prerecording...")? I can see that on the Ondio...
12:17:10preglowjmspeex: do the same for memset, i assume?
12:17:28 Quit spiorf (Read error: 110 (Connection timed out))
12:17:35preglowthat can't check pointers, but at least the free sizeof can be done
12:18:00jmspeexpreglow: same for memset/memcpy
12:18:10jmspeexWhat do you mean about checking pointers?
12:18:53preglowsince there's no src...
12:19:45preglowbtw, i had a look in TODO, you can't use an iir filter instead of a qmf filter instead of making things incompatible, right?
12:20:16preglowthat last "instead of" should be "without"
12:20:21 Join Bagder_ [0] (
12:20:25 Quit spiorf_ (Remote closed the connection)
12:20:30 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
12:20:37 Quit jhMikeS (Nick collision from services.)
12:20:37jmspeexfor memset, it's not a problem because there's no potential incompatibility
12:20:43 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
12:21:08pixelmaah, no. I can only see it in the Ondio's radio screen where you actually can record from, much more convenient. But it makes me wonder why it isn't shown in the recording screen as well...
12:22:49 Quit midkay ("Leaving")
12:23:03preglowjmspeex: and do you process the upper qmf band in its frequency reversed state, or do you modulate it somehow first?
12:24:14jmspeexpreglow: in freq-reverse state mostly. Though in one of the mode, I'm playing with modulation (I think, but I'm not sure what I settled on).
12:24:14 Nick fxb__ is now known as fxb (
12:24:52preglowand that has no real effect on lpc analysis?
12:25:37jmspeexwell, it has an effect on the *interpretation* of the LPC coefs, but LPC itself works fine
12:26:28preglowyeah, but you don't really need to handle the whole lpc part any differently, do you? after all, you do apply the lpc coefs on the frequency reversed excitation too in the decoder, i assume
12:28:30 Join midkay [0] (n=midkay@rockbox/developer/midkay)
12:29:05preglowso, you can't really just jump in and replace the qmf filter without something else, then
12:29:05jmspeexSo it's business as usual. There's only one place where I need to take that into account because I'm converting LPCs to freq response
12:29:10preglownot without introducing new modes
12:29:33preglowand disallowing vbr...
12:29:38jmspeexpreglow: I could replace by an iir filter at the cost of a tiny bit of aliasing
12:29:58preglowwould be tons cheaper to calculate
12:30:15preglowbut you'd also have to do freq reversal to maintain compatibility, then?
12:30:34 Join spiorf [0] (n=spiorf@
12:30:35jmspeexyou'd get the same reversal with any filter
12:30:43preglowahh, of course
12:30:55preglowqmfs can use iirs just fine
12:31:03preglowi don't really have much experience with those things
12:31:50jmspeexdoesn't need to be a QMF
12:31:57jmspeexactually, better not be a QMF
12:32:08jmspeexjust a brickwall elliptic IIR would work fine
12:33:00jmspeexor cheb with the ripple in the passband
12:33:11preglowhow much ripple could you use?
12:35:27 Quit barrywardell ()
12:36:21 Quit Bagder (Read error: 110 (Connection timed out))
12:51:03 Quit t0mas ("lunch break time! :D")
12:53:17 Join stewball`ghost [0] (n=WTFOMGBB@
12:56:50 Join webguest84 [0] (i=7cb74c05@gateway/web/cgi-irc/
13:03:04 Join Arathis [0] (
13:03:57 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
13:06:31 Quit webguest84 ("CGI:IRC (EOF)")
13:10:07 Join Phill [0] (
13:10:25 Nick Bagder_ is now known as Bagder (
13:10:57jmspeexpreglow: yours to test. My guess is around 0.5 dB
13:11:21***Saving seen data "./dancer.seen"
13:16:44peturpixelma: good point, should be added to the recording screen....
13:17:38 Quit atsea- (Remote closed the connection)
13:28:54 Quit Phill ()
13:33:22 Quit hcs ("Leaving.")
13:52:48 Join XavierGr [0] (
13:55:16 Join PaulJam [0] (
13:57:07marazI had a song skip to the next one totally randomly today - may be related to FS7510 not being completely fixed (ie. if both of the flags need to be set, 0x30000000 instead of 0x10000000). Could be totally random, too. Time will tell, I guess. If anyone has similar glitches using the current build on nano, do pipe up.
14:00:13peturdid you skip back to that song?
14:02:14 Join Thundercloud [0] (
14:03:16preglowjmspeex: what order filter would you expect to use? just tried an eight order elliptic one, and i actually think the aliasing might be audible with that
14:03:50jmspeexpreglow: picture?
14:04:02preglowno picture, i'm just doing some quick listening
14:05:12preglowbest thing would just be integrating it in speex and trying, of course, it should be a quick enough job anyway
14:05:32preglowunless i'm missing something vital
14:05:32 Quit Nico_P (Read error: 104 (Connection reset by peer))
14:06:25jmspeexpreglow: what's mostly important is that the low frequencies don't cause much aliasing, which is why the stop band should go down as a function of freq instead of having ripples.
14:06:34jmspeexWell, that's what I assume anyway.
14:07:11jmspeexI don't think the aliasing near the cutoff freq matters that much
14:07:36preglowok, so then type 1 cheb should be the way to go
14:07:57jmspeexI never remember which is type 1 and which is type 2...
14:08:04preglowtype 1 is the one with passband ripples
14:08:19 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
14:08:22jmspeexthen type 1 would be my best guess
14:08:36 Quit Thundercloud (Remote closed the connection)
14:08:49jmspeexpreglow: BTW, can you update your meme copy/move/set patch?
14:09:24preglowjmspeex: with the new macro stuff?
14:09:33GodEaterNico_P: is crossfade still broken currently ?
14:09:47jmspeexpreglow: yes
14:09:48Nico_PGodEater: did it use to be?
14:09:55preglowjmspeex: sure, but later
14:09:58GodEaterit was initially with MoB
14:10:05GodEaterI assume that means you fixed it
14:10:08preglowjmspeex: btw, binary is smaller on arm with mem*, but slightly bigger (100b) on coldfire
14:10:28preglowjmspeex: i'll see if i can do some speed tests on at least those platforms
14:10:28Nico_PGodEater: I never directly fixed it, but I think it's OK
14:10:34jmspeexpreglow: maybe it depends on how memmove is implemented
14:10:46GodEaterNico_P: there's a post in the forums which would indicate otherwise...
14:10:52preglowjmspeex: i think it pretty much just depends on how much overhead passing parameters has, it's free on arm
14:10:59Nico_PGodEater: oh? which one?
14:11:09jmspeexpreglow: isn't memmove inline or something?
14:11:13preglowjmspeex: amd64 also uses regs for param passing, and that explains why the binary is smaller here
14:11:25preglowjmspeex: gcc sometimes inlines it, but only newer ones
14:11:27jmspeexif it's a function call, then we should check that it's not used on really small arrays
14:11:54preglowthere's tons of lpcSize memcpy going on, and i agree we might want to cut out those
14:12:00preglowbut small size memset should be ok
14:12:04GodEaterNico_P: although as you say - it appears to be working ok on my Gigabeat
14:12:13Nico_PGodEater: I wasn't aware of any problems
14:13:14GodEaterNico_P: hence I've taken back what I said
14:13:18jmspeexpreglow: I'm seeing some of the memset you changed use N. In most (all?) functions, N is complexity-dependent, and usually 1 or 2.
14:13:41preglowjmspeex: i changed any to use N? really, now, i didn't intend to do that
14:13:48Nico_PGodEater: :) his report isn't exactly detailed either... I think I'll just wait for more details
14:14:12GodEatergood plan
14:14:26jmspeexpreglow: in cb_search.c there are loops that do (for j=0;j<N;j++) ... that you changed to memcopy
14:15:15preglowjmspeex: right, right, i thought i had used the wrong size
14:15:23preglowso yeah, there's not much point in converting those
14:15:38jmspeexIn general, you have:
14:15:42preglowi always think N is some kind of frame size
14:15:43jmspeexN=1 or 2
14:15:58preglowyeah, nsf copies are good
14:15:58jmspeexord=lpcSize=8 or 10
14:16:29jmspeexframe_size = 160
14:16:31 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
14:18:58preglowand the lpcSize copies can go
14:19:11preglowcall overhead and memcpy setup probably swallows the benefit there
14:19:19preglowbut lpcSize memsets might still be beneficial
14:19:28 Quit tchan (SendQ exceeded)
14:20:10 Quit spiorf (Remote closed the connection)
14:21:19 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
14:21:26jmspeexpreglow: what's the difference?
14:22:08preglowa good memset does fewer branches for the same write, with bigger transfers
14:22:11preglowshould make it faster
14:22:28preglowand memset doesn't have much setup to speak of
14:22:39preglowbut it's not a major point, of course
14:22:44preglowperformance gains would be meagre
14:23:38 Quit tchan (SendQ exceeded)
14:24:32 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
14:27:27 Quit tchan (SendQ exceeded)
14:28:07jmspeexpreglow: did you find many copies that were useless in the first place?
14:29:03 Join atsea- [0] (i=atsea-@gateway/tor/x-60f9a1abb14c695f)
14:31:16preglowjmspeex: three
14:31:22preglowall in decode, afaik
14:32:45preglowjmspeex: anyway, leave all lpcSize and smaller copying as loops?
14:34:08jmspeexlet's do that for now anyway
14:34:23preglowokies, what about clearing?
14:36:00preglowand should i let speex_move() and co stay, or should i just allow SPEEX_MOVE and co?
14:36:08jmspeexcan you check whether it makes a difference in code size. I expect the performance difference to be really small
14:36:15 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
14:36:28jmspeexjust SPEEX_MOVE, no need to keep speex_move() at all
14:43:52 Join mf0102 [0] (n=michi@
14:48:07 Join Thundercloud [0] (
14:49:37 Nick parafin|away is now known as parafin (
14:52:09marazpetur: Actually, it was the same song (repeat one).
14:52:46peturso yes, you skipped back. Same issue as I have, it's in the tracker already
14:53:02marazHmm, so it didn't really skip forward but to the start of the song?
14:53:06marazWhat's the FS#?
14:54:45peturrelated to (or the same as) 8092
14:58:36marazActually, I think that I stumbled upon 8160 today, too - just before the glitch I described. After powering up, resuming and rewinding, I couldn't hear anything but I had the volume off from my headphones... it played for a few seconds without me hearing anything, so I skipped back to the start and it started playing okay, but skipped to start after a minute or so.
15:04:08 Join MethoS- [0] (
15:09:24 Join MethoS-- [0] (
15:11:23***Saving seen data "./dancer.seen"
15:12:24 Quit ctaylorr (Read error: 110 (Connection timed out))
15:17:10 Join Toki_ [0] (
15:18:53 Quit Toki (Read error: 104 (Connection reset by peer))
15:20:53 Join joshin [0] (n=josh@unaffiliated/joshin)
15:24:59 Join OlivierBorowski [0] (
15:26:41 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
15:27:53 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
15:29:00 Quit MethoS- (Read error: 110 (Connection timed out))
15:32:53 Quit Llorean (Read error: 110 (Connection timed out))
15:33:01 Join Llorean [0] (
15:35:08linuxstbHas anyone (preglow, Llorean?) tested the current SVN bootloader (with amiconn's backlight changes) on the Nano?
15:39:41preglowcan ipodpatcher install an arbitrary bootloader?
15:40:14CaptainSquidit seems to boot (i installed bootloader and rockbox today with rbutilqt) or what else needs to be tested?
15:41:00preglowlinuxstb: bootloader.bin is what i write, yes?
15:42:43preglowlinuxstb: seems to work
15:43:03preglowstill doesn't exactly look like 100% if you ask me, but that's because of the multiple light levels
15:44:00preglowdark, dim, brighter, final setting
15:45:47CtcpPing from CaptainSquid!
15:45:47***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
15:46:13CtcpTime from CaptainSquid!
15:46:13***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
15:46:14CtcpUserinfo from CaptainSquid!
15:46:15***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
15:46:16CtcpVersion from CaptainSquid!
15:46:16***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
15:46:57DerPapstheh... today someone in my univerity saw my all new and shiny sanse and inforemed me that there is a project called "rockbox" and that i have to check it out. :)
15:47:03amiconnshould just be dark -> dim -> final
15:48:05 Join krazykit [0] (
15:48:17preglowamiconn: looks like rockbox switches to default before it switches to final, or something
15:50:57 Part LinusN
15:54:27amiconnYes, and default == minimum
15:55:20amiconnI wouldn't expect an extra switch to the actual settings default (which is halfbright), and on G5.5 I know that it doesn't
15:55:38 Join pondlife [0] (n=Steve@rockbox/developer/pondlife)
15:58:34 Join desowin [0] (
16:05:54linuxstbpreglow: Normally you would do "ipodpatcher -a bootloader-ipodnano.ipod" - I'm assuming you used the -ab option to install bootloader.bin?
16:15:55 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
16:16:18 Join miepchen^schlaf [0] (
16:16:35 Join J-23 [0] (
16:17:40 Quit Arathis ("Bye, bye")
16:18:05pondlifeHmm, anyone know where I might find instructions on taking my H340 apart? I thought they were on the wiki, but I can't find them now :/
16:19:41 Quit DerPapst ("So Long And Thanks For All The Fish!")
16:20:14alienbiker99yike. the only other place i know is mistic river but i dont think they have it up either
16:20:33krazykitlucky for you, grabbed most of the images, too
16:20:56 Nick BilleniumZzZ is now known as Billenium (
16:22:38*petur knows them by heart :/
16:24:41pondlifeHmm, doesn't have the text :/
16:24:52pondlifeAnd misticriver's forums are shot
16:25:06peturwhat do you need to knwo?
16:25:20pondlifeHow to get an H380
16:25:25pondlifeI have a drive
16:25:39pondlifeI've opened it before, a while ago
16:25:50pondlifeSo far I've just removed the 5 screws
16:25:59pondlifeNow I take the back cover off, right?
16:26:16peturremove external screws, pry open, remove blue stuff, switch drive, put blue stuff back, screw back together
16:26:49peturwatch those fingernails ;)
16:27:14pondlifeThe battery is attached to blue stuff; does it just fold back, or can I remove it more?
16:27:42peturyou can separate them all
16:28:12peturfold back battery, then remove blue stuff
16:28:43pondlifeWoops, bit of metal came out too...
16:28:46linuxstbpondlife: Where did you manage to find a drive?
16:28:49peturno worries
16:28:59*petur guesses ebay
16:29:08linuxstbJust a private seller?
16:29:10peturwhere amiconn and I got ours
16:29:16pondlifeYes. $80, so I hope it works!!
16:29:35*petur paid 200+ euro
16:29:45pondlifeWhich end is the drive connected? Top or bottom?
16:30:23peturpondlife: just get on with it, it's not that hard :P
16:30:35pondlifeI'm not very hardware-y
16:30:47pondlifeHmm, drive seems very wedged in
16:31:11peturyes, pull it out in the direction of its lengh
16:31:39peturthe connector is where that metal part came from
16:31:43pondlifeOK, new drive time...
16:33:55 Quit MethoS-- (Read error: 110 (Connection timed out))
16:34:39pondlifeHmm, which way up? HD connector is longer than H340 one (on both old + new drives)
16:35:18peturthere's a platic pin next to the connector
16:35:28peturit fits in a hole in the drive
16:35:45*petur goes to typing class too
16:36:13pondlifeDrive is label side up, right?
16:36:41peturflat side down
16:37:08peturbumpy side matches blue stuff to give battery a flat spot
16:37:33peturyou're lucky you're not replacing the battery
16:38:01pondlifeSafe to power up with new drive in, but bit of metal out?
16:38:09peturthe drive takes me < 5 min, the battery about 30 min
16:38:49peturyes, the bit of metal makes sure the connector stays in, however
16:38:57peturbut to test it's ok
16:39:00pondlifeJust want to check
16:39:23pondlifehaha, drive came out of a Vaio!
16:39:42pondlifeFormat time
16:39:43*petur wonders if pondlife is booting windows on his h380
16:39:46*Nico_P managed to send a firmware file to his S !
16:40:56lostlogicsweet, my ipod froze overnight.
16:41:29peturthat cold already outside?
16:41:50lostlogicaye, it's below freezing atm
16:42:00JdGordonlostlogic: what he heck you doing up so early? its like 7.40am there yeah?
16:42:06peturno wonder then... take it inside tonight ;)
16:43:02Nico_Pmarkun: ping
16:43:28lostlogicJdGordon: yeah −− I get up at 6 normally :-P
16:43:37Nico_Pmarkun: I think I've managed to send a firmware update to my S through libtmtp :D
16:43:54Nico_Pmarkun: but now I'm hesitant to send custom code to it
16:44:01lostlogicI blam album art for the freeze, since that's the only feature I recently started using.
16:44:37markunNico_P: wow! great news!
16:44:40Nico_Pmarkun: do you have an idea of what could possibly go wrong ?
16:44:56markunno, I don't know anything about the S, maybe linuxstb has some ideas?
16:44:56lostlogicJdGordon: go to sleep, it's nearly 3 there? :-\
16:45:42pondlifeHaha 80MB is too big for FAT32, allegedly..
16:45:52pondlife"QuickFormatting 76316M The volume is too big for FAT32."
16:46:19peturuse swissknife
16:46:27 Nick fxb is now known as fxb__ (
16:46:27Nico_Pmarkun: I've only tried with the original firmware file for now, and the unit reboots after I disconnect the cable, so I guess it's a good signe :)
16:46:56*linuxstb doesn't know anything about the S's firmware upgrade process
16:47:26Nico_Phmm the only ones who know are probably ptw and aliask
16:48:45 Quit pondlife (Read error: 104 (Connection reset by peer))
16:50:22Nico_Phmm should I try it ?
16:52:12 Join pondlife [0] (n=Steve@rockbox/developer/pondlife)
16:52:21GodEaterNico_P: I've done it a couple of times
16:52:29GodEaterin order to get a usb log
16:52:50Nico_PGodEater: did you send a custom firmware or the original one?
16:53:28Nico_PGodEater: also, how does the player behave during and after the update?
16:53:58GodEaterI sent the original one
16:54:08GodEateri.e. the hacked V one
16:54:37Nico_Pyou mean using the hacked V updater?
16:54:54GodEaterall the files I used were from the wiki page
16:55:03GodEaterI didn't use a rockbox "infected" firmware ;)
16:55:07Nico_Pand what did the player display during the process? "connected"?
16:55:28GodEateras I recall, the player shows a sort of three stage screen during the process
16:55:35JdGordonlostlogic: yeah, just got hoe from a night on the piss
16:55:38GodEaterbut it's been several weeks now, and I can't remember the details :(
16:55:51*JdGordon hopes that saying is understood in the funny parts of the planet :p
16:55:53GodEaterI *do* recall the process not working at all from one PC I tried
16:55:58GodEaterand there being NO obvious reason why not
16:56:02 Join MethoS- [0] (
16:56:06Nico_Phmm... mine just displayed "connected" as for any other operation... then when I disconnected the cable it rebooted
16:56:07GodEaterthe updater just refused to detect that the S was present
16:56:07pondlifepetur: OK, almost there, but which way up does the metal bit go? I assume the two metal bits are on top of the ends of the HD connector...
16:56:23Nico_PGodEater: that happened to me too when I tried
16:56:29GodEaterNico_P: you did the steps on the player to put it into firmware restore mode ?
16:56:29pondlifeBut I can't seat the blue stuff... :/
16:56:41Nico_PGodEater: no... which steps? :p
16:56:47GodEaterthey're in the wiki too :)
16:56:59GodEatersomething like hold power, the windows button, and left
16:57:18Nico_Pyeah, just saw them... I'll try that but it didn't complain when I didn't do that
16:57:21GodEaterthis will format your players drive apparently :)
16:57:28GodEaterno? Curious
16:57:29peturpondlife: can't open mine now to check, it realy only fits in one way
16:57:31Nico_PI don't really care
16:57:32GodEaterI didn't try it without doing that
16:57:40peturcheck the screwholes
16:58:33peturpondlife: so the drive is working fine? lucky b*****d :)
16:58:52 Quit CaptainSquid (Remote closed the connection)
16:59:03pondlifeSecond hand, of course
16:59:20pondlifeI haven't formatted it yet
16:59:26pondlifeWell, not fully
17:00:25 Quit parafin ("restart")
17:00:27Nico_PGodEater: success!
17:00:34 Join parafin [0] (
17:00:45GodEaterI mean \o/
17:00:57Nico_PI entered recovery mode, it asked me for an update, I ran my updater and then it booted fine :)
17:01:56 Quit JdGordon ("Konversation terminated!")
17:02:05Nico_PI'm feeling confident enough to try running custom code but I have no idea what to run :/
17:02:08GodEaterso make it run rockbox ;)
17:03:40pondlifeHmm, my eyesight's not good enough to get the screws back in :/
17:04:39peturthe long ones go in the bottom iirc
17:04:56pondlifeYup, but my fat fingers can't even hold them
17:06:05Nico_Phmm nothing's happening
17:07:36Nico_Pafter updating with the rockbox bootloader, the player stays stuck on the blue gigabeat screen
17:07:36GodEaterNico_P: the crossfade man is back in the forums Nico_P
17:07:42Nico_Pthe progressbar doesn't move
17:07:48GodEaterhow annoying
17:08:12Nico_PI've been able to restore the original firmware though
17:08:32Nico_P... a good sign that my updater works, I suppose :)
17:08:47markunNico_P: well, it's a good start
17:09:03markunis this the same code that is used for updating the creative zen?
17:09:18Nico_Pactually it is
17:09:21GodEaterit is ?
17:09:30Nico_Pwell it's what was supposed to, but it hadn't been tested
17:09:39pondlifeH380 boots!
17:09:47markunpondlife: congrats!
17:09:53Nico_PI corrected only one minor mistake to make it send the firmware (a lucky guess), and a segfault
17:10:11GodEaterNico_P: hang on - I think I've missed something here - you've written your own updater ?
17:10:51pondlifehehe the default theme sucks so much, it's a real shock when the nice USB graphic appears
17:10:54Nico_PGodEater: no, mcuelenaere did by changing a few very small things in libmtp's sendfile.c example
17:11:19GodEaterchecked into svn ?
17:11:21Nico_PGodEater: that was for the ZVM, but seeing that I asked him if it could work with the S... he looked at the logs and said it should
17:11:27***Saving seen data "./dancer.seen"
17:11:29pondlifeSo...does anyone have a spare back cover for an F40?
17:11:38 Join midgey [0] (
17:11:41markunpondlife: toffe?
17:11:52Nico_PI then tried with no success (the day before yesterday), and today I changed one thing and it worked!
17:12:25Nico_PGodEater: before putting it into svn I need to make it standalone
17:12:33GodEaterah ok
17:12:33markunNico_P: do you overwrite nk.bin?
17:12:47Nico_Pmarkun: what do you mean? on the player?
17:12:58Nico_PI think so, yes
17:13:09Nico_Pthe name of the dest file is set to nk.bin
17:13:31Nico_Pthe updater should also work for the ZVM too
17:13:59markunNico_P: how strange that you can access the first partition over mtp
17:14:12GodEaterwhy is that strange ?
17:14:22Nico_Pmarkun: I tried making a rockbox bootloader and sent it to the player but that didn't boot
17:14:34markunGodEater: because it's supposed to be hidden
17:14:41Nico_Pmarkun: maybe not to mtp?
17:15:03markunbut otherwise a firmware update would not be possible I guess
17:15:53GodEaterI thought the whole point of mtp was that it was filesystem agnostic
17:16:09GodEaterso provided the *device* understands where stuff is supposed to go, it doesn't really matter
17:16:35 Join TwilightInZero [0] (
17:21:23preglowlinuxstb: i used -a
17:22:51Nico_Pnow I need to find out how to make that program compile against system libmtp
17:23:38GodEater-lmtp ?
17:25:54 Join n1s [0] (
17:28:56*Nico_P downloads a test bootloader made by ptw419
17:30:00 Join MouFou [0] (
17:31:21Nico_Phello world !
17:31:40Nico_P(this means it worked)
17:32:16GodEaterso what's different about his bootloader to the one you built yourself ?
17:32:30qwmGod knows.
17:32:42Nico_PGodEater: it was made by ptw419 from the original one
17:32:51GodEaterah - it's not a rockbox one ?
17:32:59Nico_Pthe one I built myself was made with mknkboot
17:33:20GodEaterI'm sure I read aliask had rockbox one working
17:33:26Nico_PGodEater: I took it from there:
17:33:27GodEaterit just wouldn't load the main rockbox image
17:35:07linuxstbpreglow: That shouldn't have worked...
17:38:33linuxstbNico_P: Nice work on the S - when you say you want to make it "standalone" before committing to SVN, what do you mean?
17:38:39 Join Frazz [0] (
17:42:10Nico_Plinuxstb: ATM it's just a modded version of one of libmtp's example files... it's not even a standalone binary, as it's being called through a main program. Also I'm compiling it by running make && make install for the whole libmtp tree
17:42:47Nico_Pso I want to make it standalone and compile against the system libmtp, which I don't really know how to do yet
17:43:29Nico_PI can pastebin a patch to the libmtp tree if someone is willing to help me
17:46:42 Quit MouFou (Remote closed the connection)
17:47:47linuxstbDo you know if libmtp relies on libusb?
17:48:50markunlinuxstb: it does
17:49:13markunat least it depends on it in the FreeBSD port
17:50:49linuxstbNico_P: Is the example program just one .c file, or is it more complex than that?
17:51:02GodEaterhe said it was based on sendfile.c
17:51:17linuxstbSo "gcc -o sendfile sendfile.c -lmtp" should work?
17:51:33Nico_Plinuxstb: it's basically one c file, although the actual binary that is called is made from connect.c
17:51:38linuxstbOr maybe -lmtp -lusb
17:51:40Nico_PI'll try that
17:52:24*GodEater guesses the latter option
17:53:05Nico_Phmm it complains about "config.h"
17:54:27markunNico_P: can you upload the file somewhere?
17:54:28Nico_Pthe makefile runs gcc -DHAVE_CONFIG_H -I. -I.. -I../src -g -O2 -D_FILE_OFFSET_BITS=64 -Wall -Wmissing-prototypes -MT sendfile.o -MD -MP -MF .deps/sendfile.Tpo -c -o sendfile.o sendfile.c
17:54:56linuxstbSo it complains about config.h, even if you don't add -DHAVE_CONFIG_H...
17:55:47Nico_Plinuxstb: and even if I add it... although I have no idea what it does
17:55:58linuxstbI guess you should find the config.h it is using, and copy it into the same directory as your source, then try "gcc -DHAVE_CONFIG_H -o sendfile sendfile.c -lmtp -lusb'
17:56:22GodEaterautoconf nastiness :(
17:56:40Nico_Plinuxstb: gcc -DHAVE_CONFIG_H -o sendfile sendfile.c -lmtp -lusb didn't work
17:56:55GodEaterNico_P: even with the config.h it's using ?
17:57:04linuxstbWhat is complaining about config.h ? Your file isn't using it...
17:57:28*GodEater shuts up and gets out of linuxstb's way
17:58:00Nico_Pah, I've copied the config.h in the current dir (which I hadn't done before)... now the error is different
17:58:18Nico_Pa few undefined references
17:58:50Nico_P... the ones that are defned "extern" in the files
18:00:06Nico_Premoving the extern keword solves these but there is one left
18:00:18OlivierBorowskiHi guys, I installed rockbox on my h340 two weeks ago and it works well. Thanks for your job! I'd just like to point out some graphics artifacts in rockdoom (some kind of blinking). Is it a known bug? (poor quality video here :
18:00:20markunNico_P: btw, what was the lucky guess?
18:01:05Nico_Pmarkun: mcuelenaere had set parent_id to 0x10001... I set it back to 0 and it worked
18:01:09markunOlivierBorowski: never seen it before
18:01:41GodEaterdo we have devs that still care about rockdoom ?
18:01:49Nico_Pmarkun: I did that because I had no idea why he put 0x10001 and 0 seemed to be ok in the rest of the source
18:02:42 Join Lear [0] (i=chatzill@rockbox/developer/lear)
18:02:43OlivierBorowskiI've seen this subject on the forum : but I don't know if it's related
18:03:00 Quit petur ("*plop*")
18:03:55Nico_Pha, I got it to compile
18:04:05linuxstbNico_P: After removing a few #includes, it's compiling for me, but "device" and "progress" are undeclared. I'm guessing progress is a call-back function (so can be replaced by NULL), but I've no idea where "device" comes from, or how the program is supposed to know which device to use...
18:04:16linuxstbNico_P: This is without the config.h
18:05:12 Join ribx_ [0] (
18:05:22Nico_Plinuxstb: I took progress from pathutils.c and removed the include pathutils.h
18:05:38Nico_Psome things need to be taken from connect.c
18:06:10markunNico_P: where is "progress" defined?
18:06:21Nico_Pmarkun: in pathutils.c
18:08:07 Nick fxb__ is now known as fxb (
18:08:07linuxstbSo just device = LIBMTP_Get_First_Device(); from connect.c ?
18:08:37 Join MethoS-- [0] (
18:08:38linuxstbAnd LibMTP_Init() maybe...
18:08:52linuxstbAnd LIBMTP_Release_Device(device);
18:08:53linuxstbat the end...
18:09:22Nico_Pyeah I'm making a complete file
18:10:13linuxstbSo basically, it looks like MTP has a standard way to transfer firmware upgrade files, and libmtp supports it...
18:10:40Nico_Pyes, we're lucky :)
18:11:33Nico_POK, I have a c file that compiles fine with "gcc -o sendfile sendfile.c -lmtp" and works too
18:12:17linuxstbNice. Maybe commit to "utils/gigabeat-s" ? Or maybe utils/MTP/ as it looks like it will be useful for other devices?
18:12:39Nico_Pyeah, there is some cleaning up left to do though
18:14:00Nico_Pwhat should I put in there? the rockbox header with the original copyright plus mine?
18:14:13linuxstbDo you need those "files = ... " and "folders = ..." lines?
18:14:33Nico_Pprobably not
18:14:38linuxstbAnd also the lang check?
18:15:04 Quit TwilightInZero ("ChatZilla 0.9.79 [Firefox]")
18:15:13 Quit XavierGr (Remote closed the connection)
18:15:28 Join XavierGr [0] (
18:16:02 Join japc [0] (n=japc@
18:16:32Nico_Pyup, works fine without all that
18:17:37Nico_Plinuxstb: do you think common.h and config.h are needed?
18:18:32preglowlinuxstb: i did ./ipodpatcher -a ../../build/nano-boo/bootloader-ipodnano.ipod, that's not supposed to work?
18:19:56 Join thegeek [0] (
18:20:15Nico_Pcommon.h doesn't appear to be needed, and thus neither config.h
18:21:37amiconnpreglow: That is the standard method and should work fine
18:24:13linuxstbpreglow: Yes, but you said you used bootloader.bin....
18:25:59 Nick fxb is now known as fxb__ (
18:26:16 Quit MethoS- (Read error: 110 (Connection timed out))
18:29:11preglowlinuxstb: i say lots of things :)
18:29:54linuxstbpreglow: Have you tried wma seeking?
18:31:11preglownot the current one
18:32:06linuxstbIt consistently freezes my ipod when attempting long seeks in my 2hr test file...
18:33:37*Nico_P is discovering indent... a nice tool
18:33:51advcomp2019linuxstb, i can test it on the sansa if i had a 2hr wma file
18:34:11 Quit thegeek_ (Read error: 110 (Connection timed out))
18:34:40linuxstbI'm sure there must be podcasts or something available to download in wma format somewhere...
18:34:43preglowlinuxstb: yeah, saratoga mentioned it had problems, but he also said he thought he'd hacked around most of them
18:36:41pixelmalinuxstb: isn't that long enough that it doesn't fit completely in the buffer? Maybe there are interactions?
18:36:56linuxstbYes, it's about 110MB
18:37:05Nico_Plinuxstb: so what should I do about the header?
18:37:18linuxstbNico_P: Which header?
18:37:39Nico_Pthe header for the firmware updater source file
18:37:47linuxstbYou mean the (C)/license header?
18:38:12Nico_Palso I've called the file sendfirm
18:38:49linuxstbI would add your name at the top, and add a line saying something like "based on the example application sendfile.c from libmtp", followed by the existing (C) and license info.
18:40:20linuxstbAlthough maybe say "Modified by Nico_P" rather than (C) - claiming copyright when you've just copied and pasted code is perhaps going too far...
18:40:51advcomp2019linuxstb, found one but it is mp3 but i will convert it to wma for testing
18:41:08Nico_PI'll also add mcuelenaere, as it was his idea
18:41:25linuxstbadvcomp2019: Just make sure it's larger than the RAM in your player - so at least 30MB, preferably bigger.
18:42:28 Quit mf0102 ("Verlassend")
18:44:40 Quit Billenium ("Ex-Chat")
18:44:54 Join Billenium [0] (
18:45:26mokkurkalveI've ordered a e280 to add to my H340 as rockboxed players. I think this is a straight e280, but what was significant with the "Rhapsody" models?
18:46:50 Join evert [0] (n=evert@
18:47:26 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
18:48:22Nico_Plinuxstb: I have no idea what to do for the makefile though
18:48:39krazykitmokkurkalve, the fact that the firmware partition can't be seen over usb, meaning the install is kind of a pain.
18:48:39Nico_PI guess I'll commit the file and take care of the makefile later, I need to go do some shopping
18:49:08evertwhat installation method would you suggest me to get rockbox on my sansa ? The automated or the manual ?
18:49:41Learevert: If on Windows, manual for now.
18:49:54evertLear: it's on linux
18:50:08everti'm not afraid for the cli (in fact i'm not even using X at the moment)
18:50:19everti was just wondering what the 'best' way was :)
18:50:35mokkurkalveis there anything on the package that indicates if it's a "Rhapsody" or not?
18:50:54everti couldn't find anything what said 'rhapsody'
18:51:05evertotherwise i hadn't bought it, but i'm not 100% sure
18:51:10 Quit Billenium ("Ex-Chat")
18:51:48krazykitthe best way to tell if it's a rhapsody sansa is that you have rhapsody channels in the main menu
18:52:06evertwhat are raphsody channels ?
18:52:23krazykitthey indicate that you have an e200r.
18:52:35krazykitit's Real's DRM scheme, basically
18:53:08evertkrazykit: and how can i see that in the main menu ?
18:53:18krazykitin the OF, you spin the wheel
18:53:34evertyeah, then i get all the menu items ? Nothing what says 'raphsody' ?
18:54:01krazykitif nothing says rhapsody, then you have a regular e200, meaning you get the easy install
18:54:16evertok, that's good thus :)
18:54:28evertso i can use the automated installer without an problem ?
18:54:35krazykityou should be able to, yes
18:54:37 Quit Febs ("ChatZilla 0.9.79 [Firefox]")
18:54:38evertit hasn't got a 'higher brick' chance or such .
18:54:48krazykityou can't brick your sansa this way.
18:55:04evertok, thanks, i'm going to try it then :)
18:55:12markunkrazykit: does the e200 use MTP?
18:55:31krazykitmarkun, only if you tell it to. it has both options, MTP and MSC
18:55:36evertmarkun: i can choose between mtp and msc
19:02:42 Quit Llorean ("Leaving.")
19:02:54everthmm, i've mounted my sansa on '/mnt/mp3' but the rockbox installer won't see it
19:03:21evertin the configure rockbox utility i've set 'select your device in the fs' to that point
19:03:29evertbut it won't recognize it
19:03:46krazykitcan you access the files as the user rbutil is running as?
19:04:44evertwith mplayer i can play music from my sansa
19:04:52evertso it's mounted correctly
19:05:39evertthe strange thing is, is that the player says 'disconnected'
19:05:52krazykityeah, the LCD messages are completely useless, really
19:06:05krazykitmeh, just do a manual install. not really much harder
19:06:45evertIt seem your Bootloader is already up to date.
19:06:46evertDo really want to install it?
19:06:57krazykitsounds like rbutil managed to that
19:07:03 Join spenc [0] (i=472542c4@gateway/web/cgi-irc/
19:07:24evertignore that ? (sounds riskfull to me)
19:07:33krazykitshould be fine
19:07:47evertit's installing :)
19:08:32 Join nanok [0] (n=nanok@
19:11:00 Part pondlife ("Gone")
19:11:29***Saving seen data "./dancer.seen"
19:12:30evertnot working, still the default sandisk firmware when rebooting
19:12:32 Quit spenc ("CGI:IRC (Ping timeout)")
19:12:40advcomp2019linuxstb, it happens on the sansa with wma seeking
19:13:00krazykitevert, ah. run sansapatcher again, and choose whatever option will install the bootloader
19:14:31 Quit tchan (Read error: 110 (Connection timed out))
19:14:37evertnow i get rockbox, i thinks it needs some configuring, the font is a bit too small, i need glasses to read it lol
19:15:45marazthen change the font.
19:16:00everti'm already happy it works, now time to configure it :)
19:17:24 Join TMM [0] (
19:20:38evertthe rockbox menu's are much better than the default firmware menu's :)
19:24:18 Quit J-23 (Remote closed the connection)
19:30:52evertwith rockbox i can't mount my mp3 any more
19:31:21 Join spenc [0] (i=472542c4@gateway/web/cgi-irc/
19:31:40 Nick fxb__ is now known as fxb (
19:31:57 Join lazka [0] (
19:33:35Learevert: You must boot the original firmware for that.
19:33:45evertLear: how can i do that ?
19:34:16spencWhen I transfer music to my E250R, it shows up in the original firmware, but not in Rockbox. Any suggestions?
19:34:17LearEasy way is to turn it off, then plug in the connector. Or press left when turning it on.
19:34:59Learspenc: Set it to display all files (which includes hidden files), or use the database.
19:35:16evertLear: ok, thanks
19:36:21 Quit joshin (Read error: 113 (No route to host))
19:36:44 Join zack [0] (i=59a8103d@gateway/web/cgi-irc/
19:36:53zackHi everyone
19:37:47zackI have a small problem
19:37:59zacklooking for help
19:38:14zacki can't have my themes work right sometimes
19:38:51 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
19:39:24 Quit spenc ("CGI:IRC (Ping timeout)")
19:40:16 Join spenc [0] (i=472542c4@gateway/web/cgi-irc/
19:40:33zackanyone can help me please
19:41:05spencLear: I'm not sure what you mean by that. Sorry, I'm new to this. I transferred the music by putting it in the MUSIC folder in the root of the drive. Was this not what I was supposed to do?
19:41:13 Quit tchan (SendQ exceeded)
19:41:26 Join mikill [0] (
19:41:56 Quit BigBambi (Read error: 113 (No route to host))
19:41:57zackfor your music put it into playlist
19:42:03 Part mikill
19:42:06zackthat should work
19:42:06LearThe OF marks that folder as hidden. Rockbox only displays that if "show all files" is enabled.
19:42:33LearLong press menu, and then use menu (I think) to change that setting. Exit the quick menu by pressing the off button.
19:43:13everthmm, strange, just for a test i've copied a avi file to the /vidoe dir of my sansa, but the file browser even can't see that file
19:43:15 Quit zack ("CGI:IRC (EOF)")
19:43:58spencBut, I can see the preloaded music with Rockbox, and that is in the MUSIC folder...
19:47:25spencI'm assuming that long press just means to hold it down... When I hold down menu, it turns the player off
19:47:59LearWith menu I mean the button below the wheel, not the power button.
19:50:59 Join FOAD_ [0] (
19:52:29krazykitevert, that's probably because you're browsing in "supported" mode.
19:52:47 Quit spenc ("CGI:IRC (Ping timeout)")
19:53:12evertkrazykit: yeah i see now rockbox can only play mpeg1/2 :)
19:53:13Nico_PGodEater, linuxstb: care to try sendfirm?
19:53:29evertkrazykit: what happens if the resoultion of my vidoe is bigger than the screen resolution ?
19:54:29 Join spenc [0] (i=472542c4@gateway/web/cgi-irc/
19:54:41 Quit spenc (Client Quit)
19:54:50evertkrazykit: do you know wheter it recharges in 'rockbox' mode or not ? (i've to use the original firmware to mount it ..à
19:55:10Nico_PGodEater, linuxstb: I just tested and updating while in the OF works
19:55:30krazykiti believe it will charge, but very slowly. the OF is still recommended for charging, too, until Zagor commits the charging part of usb
19:55:48evertkrazykit: ok, thanks :)
19:56:08*evert is very happy with RB, sadly the usb stuff won't work but that doesn't matter that much :)
19:57:16 Join spenc [0] (
19:57:24 Join przemhb [0] (
19:57:33linuxstbNico_P: So you don't need recovery mode - you can just run it in the normal firmware?
19:57:52linuxstbNice... Now we just need a dual-booting bootloader...
19:57:59 Quit Frazz (Read error: 104 (Connection reset by peer))
19:58:06Nico_Pof course once you put a rockboxed nk.bin, you'll need to enter recovery mode to update
19:58:28spencI can see the music now if I go into files, but if I go into database all I can see is the preloaded content, which I deleted
19:58:35spencWhy does it still show up?
19:58:42linuxstbRecycle bin?
19:59:30*linuxstb goes home
19:59:32 Quit linuxstb ("Client Exiting")
20:00:05LearUpdate the database?
20:02:21przemhbCould someone take a look at FS #8203 ? It fixes a bug that caused rec_split_type not to be loaded properly. As a result RB did not stop recording on split even if it was to do so.
20:02:28 Join Zagor [0] (
20:02:37spencLear: Sorry, but how do I do that?
20:07:36 Join advcomp2019_ [0] (n=advcomp2@unaffiliated/advcomp2019)
20:07:48 Quit advcomp2019 (Nick collision from services.)
20:07:54 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
20:08:19LearThe manual explains all that. You can also browse the menus... (I don't use the DB myself.)
20:08:22 Quit FOAD (Read error: 110 (Connection timed out))
20:08:22 Nick FOAD_ is now known as FOAD (
20:08:24 Join saratoga [0] (i=47e276af@gateway/web/cgi-irc/
20:09:23 Join Bagder_ [0] (
20:10:09 Join Frazz [0] (
20:11:40spencI'm reading the manual right now. It just talks about going to the database menu, but when I go into database, and hold menu, I don't get any of the options that the manual talks about
20:11:40 Quit Thus0 (Read error: 104 (Connection reset by peer))
20:12:32 Quit japc (Read error: 110 (Connection timed out))
20:13:07markunspenc: there is another database menu in settings -> general settings
20:13:18spencThank you markun
20:13:21markunmaybe that's the one they are talking about
20:13:31*Zagor aims to commit charge-only usb tonight
20:13:35spencI found it just as I read your message
20:13:45 Quit spenc ("ChatZilla 0.9.79 [Firefox]")
20:13:53markunZagor: what does that mean?
20:14:12nanokZagor: yes!!! :)
20:14:30Zagormarkun: the full stack as-is, but with a dummy interface active instead of the storage one. so it only asks for full power, and does nothing else.
20:14:32*nanok hails Zagor
20:15:28markunZagor: great, I hope it will motivate more people to work on it
20:15:29nanokZagor: how is the progress on the mass storage stack, any leads on craking that 64B block "limitation"?
20:16:01nanokmarkun: that was part of my sketchy theory also ;)
20:16:02Zagornanok: no, nothing yet
20:17:07 Join lee-qid [0] (
20:20:47nanokZagor: i assume mi childish questioning of usb 1.1 instead of 2.0 proved to be ..just that, childish :)
20:21:53ZagorI wouldn't say childish, but it was wrong. but I bet in the end there is something stupid like that that's causing the problem.
20:22:12Zagorso keep firing silly suggestions :)
20:22:35desowinZagor: :-)
20:23:11nanokZagor: you can bet on that, if i come up with anything like that, i'll say it out loud. i'm not ashamed, i know i am useless as a coder, you know it, everybody does :)
20:23:42nanokseriously, it's true, usually it's just in front of you but cannot see it, because you've been looking at it too long
20:23:51nanokit happened to me more than once..
20:24:13saratogaZagor: I found that disabling the USB hardware when its not actively being used saved a bit of power
20:24:28n1sNico_P: I have'n been studying how to load firmware on the gigabeat S but how big is the risk with this new tool of yours?
20:24:32saratogahow hard would it be to switch off the USB when done checking for a connection
20:24:38Zagornanok: yeah. hopefully when I commit it, someone will look at it and ask "why are you doing like that?"
20:24:53Nico_Pn1s: apparently not big...
20:24:54Zagorsaratoga: not hard at all
20:25:02Nico_Pn1s: I've tried it quite a lot and haven't had any problems
20:25:16n1sNico_P: is there any way to break the recovery mode?
20:25:19 Quit Bagder (Read error: 110 (Connection timed out))
20:25:31Nico_Pn1s: I don't know
20:25:51Nico_Pn1s: even when the update I sent didn't boot I was still able to access recovery mode and sent the OF
20:25:56saratogaZagor: good, was actually going to disable it all the time on the sansa, but maybe i won't if you're making progress
20:25:57n1sNico_P: so have you been able to load a rockbox bootloader?
20:26:21Nico_Pn1s: only the hello world one that was posted on the forums... current SVN didn't boot
20:28:31saratogais this because of the trouble with interupts on the S?
20:29:12nanokZagor: if it would be bash, i would probably have a chance looking at it, but c is chinese to me, sorry :(, all i can do for now is come up with unsubstantiated stupid ideas and make you angry :)
20:29:28nanok(that, actually, has been known to work sometimes)
20:29:29Nico_Psaratoga: no idea
20:29:57n1sNico_P: but it did boot at some point right?
20:30:21Nico_Pn1s: it stopped doing anything right after showing the progressbar
20:30:35Nico_Pusually the progressbar progresses... it didn't
20:30:56n1sNico_P: I mean the rockbox bootloader (if you can call it that) for someone at some point in time :)
20:32:18Nico_Pn1s: no I don't know
20:34:41everti was looking what video i could use with rockbox, where can i find the exact list of what rockbox plays ? I can't come further then the mpeg1/2 page
20:36:20PaulJamrockbox only plays mpeg1/2 video. see the PluginMpegPlayer wiki page.
20:36:21 Join class37 [0] (
20:36:27evertPaulJam: i'm on that page
20:36:29evertbut what exact ?
20:36:33evertand what for the audio ?
20:37:03PaulJamevert: 44.1 kHz mp3 audio.
20:37:05class37bought ipod nano 8 gig today
20:37:10evertPaulJam: ok
20:37:12class37rockbox avail for it?
20:37:19krazykitclass37, the fat one?
20:37:24evertand for the video ? I'm using avidemux and i'm seeing lot's of 'mpeg' stuff
20:37:27Zagorsaratoga: do you mean DEV_EN &= ~DEV_USB ?
20:37:39class37thin krayzykit
20:37:53krazykitclass37, no, i mean what generation? the newest one?
20:38:04krazykitthen no rockbox.
20:38:09krazykitas it states on the front page.
20:38:21class37going to be?
20:38:31krazykitonly if someone writes it.
20:38:43krazykitwhich no one is right now.
20:39:01class37that sucks
20:39:07krazykitthey have 8 gig sansas, you know ;-)
20:39:16class37are they good?
20:39:29DogBoyI like mine
20:39:50krazykiti'm happy with mine. radio, voice recording, and microSD(HC) expansion are great features.
20:40:07everti like my new sansa (8gb, yesterday bought) very much :)
20:40:26*desowin joins the sansa e280 fanclub
20:41:04class37like ipod... spent $200 approx
20:41:16class37maybe i'll get another 30 gig video
20:41:38class37dropped one
20:41:40desowinbut tbh. sansa wouldn't be that nice if not rockbox port to it
20:41:44class37broke screen
20:42:02DogBoyI just use my ipod video for video now
20:42:12evertdesowin: sansa is oout of the box all right, i only got lost in the menu's
20:42:38 Quit TMM ("Ex-Chat")
20:42:38desowinit doesn't even have proper volume control and you call it all right
20:42:54everti was able to control the volume ?
20:43:02advcomp2019class37, the sansa e280 is $150 new the last i new
20:43:28desowinevert: the volume range it offers is well, very tight
20:43:59class37TY advcomp
20:44:02desowinI wouldn't recommend sansa for anyone for OF, but if one plans to put rockbox on it - well, it gets my recommendation ;-)
20:44:12evertdesowin: i hadn't noticed that, but you could be right ;)
20:44:24 Join Domonoky [0] (n=Domonoky@
20:44:40DogBoyheh, I thought rockbox was what we were talking about
20:47:25 Quit desowin ("use linux")
20:50:51 Join bertrik [0] (
20:52:32saratogaZagor: assuming that constant does what I think it does, that should do it
20:53:29 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
20:53:33pixelman1s: a bit of yellow...
20:53:47n1spixelma: ah, thanks
20:56:22 Quit tchan (SendQ exceeded)
20:56:34 Join perrikwp [0] (
20:56:47perrikwp /msg NickServ IDENTIFY kp0431688
20:57:17 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
20:57:19n1sperrikwp: hint, change your po
20:57:35*n1s kills enter key grrr.
20:57:41perrikwpo well, i messed that up
20:58:55nanokperrikwp: don't worry, shit happens
20:59:51nanoki know of someone who once entered his passphrase in broadcast mode on a multiterminal , which included some irc terminals
21:00:27nanoki wonder who it was..
21:00:31 Quit perrikwp (Nick collision from services.)
21:00:33 Join linuxstb [0] (n=chatzill@rockbox/developer/linuxstb)
21:00:53 Join perrikwp [0] (
21:01:27 Quit perrikwp (Nick collision from services.)
21:01:50 Join perrikwp [0] (
21:02:09 Quit perrikwp (Client Quit)
21:02:20 Join perrikwp [0] (i=421af228@gateway/web/cgi-irc/
21:02:31 Quit perrikwp (Nick collision from services.)
21:04:11 Join perrikwp [0] (i=421af228@gateway/web/cgi-irc/
21:06:18 Quit perrikwp (Nick collision from services.)
21:09:51*linuxstb wonders what hardware is inside here -
21:10:48 Quit tchan (SendQ exceeded)
21:11:32***Saving seen data "./dancer.seen"
21:12:15 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
21:15:31 Join hannesd_ [0] (
21:15:44Zagorlinuxstb: here's the manufacturer:
21:16:09saratogathat thing looks huge
21:16:24Zagor115mm x 62mm x 26mm (w = 145g)
21:16:31linuxstbYes, I found that - no firmware downloads that I could see though...
21:16:56linuxstbPerhaps the largest ever flash player?
21:17:35 Join billenium [0] (
21:18:53 Join kfazz [0] (
21:19:56kfazzZagor: i tried the usb patch on a usb 1 and 2 port
21:20:03 Join spiorf [0] (n=spiorf@
21:20:07kfazzon the 2 i got the same dmesg entries everyone else did
21:20:15kfazzon the 1, just 4 device resets
21:20:23kfazzis that expected?
21:20:36Zagorkfazz: ah, thanks for reminding me. I have yet only tested on usb2 ports.
21:20:36 Nick Bagder_ is now known as Bagder (
21:20:51ZagorI'll do some checks before committing.
21:21:48 Quit tchan (SendQ exceeded)
21:22:04*ender` yawns
21:23:16 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
21:23:44 Quit miepchen^schlaf (Read error: 104 (Connection reset by peer))
21:24:07 Join miepchen^schlaf [0] (
21:24:26Zagorkfazz: works well for me on linux
21:24:48Zagor"new full speed USB device", "not running at top speed; connect to a high speed hub"
21:25:07kfazzhmm.. i'll try it again
21:25:14kfazzit was on my linux box
21:25:24Zagorwell the patch is quite old so there may be a difference
21:26:39Zagorespecially since my current code only does charging, so there is less to fail...
21:27:17bertrikWhat are the thoughts on using 'assert' on rockbox?
21:28:31Zagorbertrik: personally I'm not a fan of it. it tends to either replace better checks or be used unnecessarily
21:28:52linuxstbZagor: Is there anything Sansa-specific in your patch, or should it work on other PP targets?
21:29:02pixelmalinuxstb, Zagor: don't trust the weight specs, when I searched for the weight of the c200, I found a few sites with the right number and quite a few claiming it was 200 and something grams. The fun thing that it was always the same number and in they even specified tenths of a gram...
21:29:21linuxstbpixelma: Must depend how many songs are on it...
21:29:21bertrikoh, I very much like it, I don't see many checks at all in the current code
21:29:55everthow can i see the fps in rockbox's mplayer ?
21:29:59bertrikIt can be used to catch a bug early and avoid confusion by side-effects
21:30:00Zagorlinuxstb: in the current state only sansa and h10 is activated in their config files. I'd like some more testing before enabling it for ipod in svn.
21:30:40linuxstbZagor: Sure. A simple approach would be to just commit it, and then we can enable it for each target as it gets tested.
21:30:44 Quit hannesd (Read error: 110 (Connection timed out))
21:30:45 Nick hannesd_ is now known as hannesd (
21:30:50Zagorlinuxstb: exactly
21:31:14linuxstbDoes it affect the reboot-to-disk-mode ?
21:31:50everti'm trying video encoded on 600kbit, it seems to play well ,but i'm wanting to see the fps of it :)
21:31:57 Quit tchan (SendQ exceeded)
21:32:06linuxstbevert: There's a "show fps" option in the mpegplayer menu.
21:32:07Zagorlinuxstb: I hope not :-) I have stayed away from modifying the ipod behaviour yet.
21:32:55evertlinuxstb: where can i find that ? When i klick on a mpeg file it only shows a few options (play , resume, play from minute xx and quit ?)
21:33:16 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
21:33:18linuxstbThat's the "start menu". Once you play the video, you can access the main mpegplayer menu.
21:33:22Zagorlinuxstb: for that reason the code is not yet fully "clean". i.e. there are some hw init code left in usb-fw-pp502x.c that should probably only be in my new usb-drv
21:33:53Zagormeh, I'll just commit this monster and you'll all see
21:34:42*linuxstb does the replace-one-broken-usb-stack-with-another dance ;)
21:34:57evertlinuxstb: i've enabled the display fps, but it's flickering like every second on the screen ?
21:35:20linuxstbevert: Yes. It's just a debugging feature, so isn't overlayed on the video.
21:35:40evertlinuxstb: ok, it says 24 fps constantly as far as i can see, that's good enough i think :)
21:35:44bertrikany other opinions on assert? or are people not familiar with it at all?
21:35:53evertwhat's the third number it's showing ? fps-frame number- ?
21:35:57 Quit tchan (SendQ exceeded)
21:36:14Zagorbertrik: do you want to use it for "code contracts"?
21:36:18Bagderbertrik: I like assert()s
21:36:45bertrikZagor: no
21:37:13 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
21:37:24bertrikI use them in my own code and I feel it helps to contain a bug to the earliest possible occurance
21:38:22 Join GodEater__ [0] (n=bryan@rockbox/staff/GodEater)
21:39:25 Quit amiconn (Nick collision from services.)
21:39:34 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
21:40:55Zagora problem with assert is there is no natural limit to them. you can dozens of asserts to every function. (yes I've seen it happen!)
21:41:20 Join roolku [0] (
21:41:20Zagoralso people tend to use assert instead of proper error handling
21:41:22bertrikI don
21:41:30bertrikt see a problem with that actually
21:41:52Zagorthere is a problem if you intend to have them compiled-in in normal builds
21:41:55bertrikyou can make asserts disappear completely in a non-debug build
21:42:05Bagderasserts should only be used in debug builds
21:42:16Zagorif you are talking debug-only then I have no objections
21:42:17bertrikI agree that asserts should not be used as an error handling mechanism
21:42:36Bagderno, one need to keep regular error checks apart from asserts
21:45:06 Nick _pill is now known as pill (
21:45:39LearBtw, would it be possible to send debugf:s over USB? Not sure what would be used on the PC to receive it though... (Some sort of serial port driver?)
21:46:16 Join kugel [0] (
21:46:17bertrikLear: it should be fairly easy to make a virtual com port, at least much easier than MSC
21:46:18ZagorLear: yes it is possible
21:46:18Domonokyserial over usb could be possible..
21:46:41linuxstbDoesn't the iFP port already do that?
21:46:58bertrikwindows xp has a usbser.sys driver built-in and linux supports it with cdc-acm kernel module IIRC
21:47:27Zagorlinuxstb: at least there is a usb_serial.c for ifp
21:47:31 Join ompaul [0] (n=ompaul@gnewsense/friend/ompaul)
21:48:09 Join BeChris [0] (i=51fb6c8c@gateway/web/cgi-irc/
21:48:34BeChrisGood evening everybody
21:49:28Nico_PBagder: are assert calls compiled in by default in a non debug build?
21:49:51Bagderno, normal behaviour is to only have assert on debug builds
21:50:51bertrikBagder: AFAIK normal behaviour of assert.h is to have it enabled by default and disabled if NDEBUG is defined
21:51:02bertrikBut that can be changed the other way around ofcourse
21:51:23BagderI think you're right
21:51:23 Quit BeChris (Client Quit)
21:51:43BagderI think it's mostly me who have adapted my own "I-use-assert-in-my-debug-builds" style
21:52:10*Bagder lives alone in a cave and builds his own universe in here
21:52:43Nico_Passert.h behaves the way bertrik said, but maybe we could define NDEBUG by default and undefine it on debug builds
21:52:53 Quit Lear ("ChatZilla 0.9.79 [Firefox]")
21:54:26Zagorkfazz: ok. I just committed my code, please try it and see if you get the same problem.
21:54:31 Join stripwax [0] (
21:55:38kfazzZagor: ok, as soon as turkey dinner's done
21:56:19GodEater__I should really setup my build server on this thing, and let my PC die in peace
21:58:15stripwaxGosh, a new USB stack's' been committed.
21:58:20*stripwax goes reading the logs
21:58:29*Zagor swats some reds
21:58:48 Join BeChris [0] (
21:59:47BeChrisGood evening every rockbox users
21:59:57pixelmaZagor: the deltas look weird too
22:00:44Bagderwhy? it's disabled on ipods but the old stack was removed
22:01:03rasher_Zagor: is this expected to work on sansas?
22:01:13 Quit billenium ("Ex-Chat")
22:01:19ZagorBagder: yes, some apps code was removed for the old stack so green deltas should be correct
22:01:26Zagorrasher_: yes
22:01:45Zagorrasher_: "work" as defined as "enables charging"
22:01:54pixelmaBagder: ok, I thought it still would be enabled on Ipods because the H10s still have it too
22:01:58rasher_Ah, so no full-speed msc yet?
22:02:05Zagorrasher_: no
22:02:15 Quit class37 ()
22:02:28Zagorpixelma: I wanted more ipod testing before enabling it
22:02:39pixelmaI see
22:02:52CtcpPing on #rockbox from BeChris!
22:02:53***Server message 505: 'logbot :Private messages from unregistered users are currently blocked due to spam problems, but you can always message a staffer. Please register! ( )'
22:02:55bertrikProper charging needs to code to configure the max current and voltage for the AS3514 charger I think
22:03:16bertrikOr is that part of your commit too?
22:03:39GodEater__linuxstb, you here ?
22:04:04linuxstbGodEater__: Yes
22:04:21GodEater__did you ever get your PS3 set up as a rockbox build server ?
22:04:23 Join moos [0] (
22:04:54 Quit tchan (Read error: 113 (No route to host))
22:04:57rasher_Zagor: would it be possible to always enter "charge without usbmode" (which usually requires you to hold down a button when inserting usb) as long as msc isn't working?
22:05:04linuxstbIt's not an official build server, no. I don't think it would help much...
22:05:18GodEater__but it does work though ?
22:05:30linuxstbSure, did its job IIRC.
22:05:35saratogarasher_: its possible, my own builds do that
22:05:55Zagorrasher_: absolutely. I just didn't want to change a lot of behaviours with this initial commit.
22:05:58GodEater__linuxstb, I might set mine up as one in that case, and see if it does better than my PC
22:05:58 Quit miepchen^schlaf ("Verlassend")
22:06:18 Quit Isolinear ()
22:06:20 Join miepchen^schlaf [0] (
22:06:23rasher_Zagor: I think it doesn't make much sense to actually enter usb mode as long as there's no msc.
22:06:43 Join donutman25 [0] (n=chatzill@
22:07:20Zagorrasher_: I agree
22:08:19rasher_Can anyone explain this: ?
22:08:39rasher_Would be nice to have at least the manual indexed...
22:09:22Zagorrasher_: historical reasons, when we had a single dl server with tiny bandwidth. I'll remove it.
22:09:53Nico_PGodEater__: have you run builds on it yet?
22:10:26GodEater__no - but apparently linuxstb has on his
22:10:36Nico_PBagder: have you had time to think about turning rev numbers into links in svn logs?
22:10:45Nico_Pmaybe URLs too
22:12:03rasher_Zagor: in that case, a search field on that does a google search with " inurl:manual/$target" would be nice
22:12:43linuxstbGodEater__: I've just checked the IRC logs, and I reported that a full swcodec build took 3m 30s (on 3rd April) on my ps3
22:13:03GodEater__perhaps I won't then
22:13:43Zagorrasher_: what would $target be then? a drop-down selector?
22:13:48 Nick parafin is now known as parafin|away (
22:13:56rasher_Zagor: yeah
22:14:14Zagorcould be nice
22:16:00preglowjmspeex: ahoy?
22:18:55 Join darkapostrophe [0] (
22:19:07darkapostropheI installed a new revision this morning
22:19:15darkapostropheI've been having USB mounting problems all day
22:19:19darkapostropheIs this a known bug?
22:19:34 Quit stripwax ("Miranda IM! Smaller, Faster, Easier.")
22:19:42krazykitwhat player, darkapostrophe
22:21:07 Join JdGordon [0] (
22:23:14krazykiti don't think there have been any changes that should affect the gigabeat..
22:23:40krazykitbut you know there's a bootloader USB mode, so you can still transfer files
22:25:41 Join krabador [0] (
22:26:33 Join billenium [0] (
22:26:39 Quit billenium (Read error: 104 (Connection reset by peer))
22:26:55darkapostrophekrazykit: I'm using that now
22:27:01darkapostrophekrazykit: But I'd prefer not to
22:27:16krazykitupdate to the latest build, see if that fixes it
22:28:14BeChrisI'm looking for someone involved in rockbox Makefiles (to propose some modifications)
22:29:17GodEater__BeChris, just make your suggestions - even if the people you want to address specifically aren't paying attention - the channel is logged
22:29:52 Part przemhb
22:29:55BeChrisok GodEater, indeed I noticed some parts in Makefiles that could be highly improved to reduced compile time
22:30:18BeChrisespecially on cygwin systems which have very poor disk I/O performances
22:30:29GodEater__that would be most welcome :)
22:31:16linuxstbZagor: USB detection no longer works on my ipod...
22:31:33Zagorlinuxstb: ouch
22:31:36amiconnZagor: red...
22:31:42Zagoramiconn: I know
22:31:47BeChrisI don't want to annoy you with that but I don't find normal to link again and again the sound codecs at every make invocation
22:32:05linuxstbZagor: Too much red here... -;r2=15758
22:32:31advcomp2019Zagor, i thought you said that the charging would work
22:32:35BeChrisand 'make install' for a simulator build create a zip file to unzip it right after : what a waste !
22:33:15Zagoradvcomp2019: it does for me
22:33:26linuxstbBeChris: We would welcome fixes...
22:33:41GodEater__BeChris: yes - please submit patches :)
22:34:05BeChrisOk, I've redone 20% but I'll post a patch
22:34:08advcomp2019Zagor, mine still acts like the old builds when charging in rockbox
22:34:11BeChriswhen done
22:34:31Zagoradvcomp2019: what player? and what does "like the old builds" mean?
22:34:59 Join keanu [0] (n=keanu@unaffiliated/keanu)
22:35:04BeChrisGoing back to real life
22:35:13BeChrisI wish you all a good end of day
22:35:20Nico_Phmm, an assert call with a false condtition just causes a crash without any useful indication in my sim...
22:35:28Zagorlinuxstb: could you simply enable usbstack and see how it works on ipod?
22:35:43linuxstbI've tried that - detection doesn't work, so nothing happens...
22:35:49darkapostrophekrazykit: It still doesn't react to USB insertion
22:35:57 Quit BeChris ("Quit")
22:36:20Zagorlinuxstb: aha
22:36:53advcomp2019Zagor, i have the sansa e280r... it hits 100% in seconds from about 85%
22:37:17linuxstbZagor: usb_detect() has no code for ipod usb detection left....
22:37:27Zagorlinuxstb: oops...
22:37:34linuxstb(and it's not as simple as checking a GPIO pin)
22:38:08linuxstbThat complex code that was there before is the only way we know to distinguish between a usb connection to a PC and usb connection to an AC charger.
22:38:24kfazzzagor: current build works fine, i think the problem was i applied the patch incorrectly
22:39:56Zagorlinuxstb: ah, I remember now. my plan was to implement that in the driver instead of there because it depends on the usb registers.
22:41:26 Quit bertrik ("bye")
22:42:07keanuZagor, saw some usb code in SVN - what's happened?
22:42:31Zagorkeanu: I simply committed my code so everyone can see it instead of asking me about it ;)
22:42:57Zagorthe commit had some mistakes that we're cleaning up right now
22:42:59krazykitkeanu, read the irc logs :P
22:43:00keanuZagor, ah, ok. I thought maybe something super major happened
22:43:08keanukrazykit, 'k
22:43:24advcomp2019Zagor, is my issue still ok or not
22:44:03Zagoradvcomp2019: I haven't changed anything in the charging code, all I do is ask the hub for full power.
22:44:07 Quit GodEater__ (Remote closed the connection)
22:44:45advcomp2019Zagor, ok
22:44:53keanuis there anything that needs testing? ;)
22:44:54 Join Daniel_S [0] (i=57b0ff03@gateway/web/cgi-irc/
22:45:14Zagorkeanu: first I need to fix the reds and the broken ipod code
22:45:27 Join tchan [0] (n=tchan@lunar-linux/developer/tchan)
22:45:43keanuoh....those reds ;)
22:46:10Bagderthat cowboy talk... :-)
22:46:23Nico_PGodEater, GodEater_: yes?
22:47:19Bagderthe god is eaten, he said he'd be back tomorrow
22:48:00Nico_PBagder: seen my question above?
22:48:12Bagderyeah, but I ignore it for now ;-)
22:49:37Nico_PI'd do it myself if I knew perl
22:51:59Zagorlinuxstb: I think the quick solution is to bring all those lines back but replace the register read with a driver function call
22:52:19 Quit krazykit (Read error: 104 (Connection reset by peer))
22:52:56 Join krazykit [0] (
22:53:35 Join billenium [0] (
22:57:33 Quit saratoga ("CGI:IRC (Ping timeout)")
23:00:18 Quit linuxstb ("ChatZilla 0.9.79 [Firefox]")
23:01:06 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
23:02:37 Quit J3TC- (".UPP.")
23:03:28Zagorlinuxstb: please refresh my memory: that code was to choose wether to stay and charge or reboot to OF?
23:07:00Nico_Plinuxstb: how is mknkboot called to produce an nk.bin file at the end of the S bootloader uild?
23:07:26Zagorhmm, I got the impression that wasn't the desired way to work
23:09:51linuxstbWell, usb_detect() returned either USB_INSERTED, USB_EXTRACTED or USB_POWERED
23:10:20linuxstbUSB_POWERED was returned if either an AC charger was detected, or the user was holding the button to charge.
23:10:25 Quit MethoS-- (Read error: 113 (No route to host))
23:10:57 Quit ompaul (Client Quit)
23:11:10 Join J3TC- [0] (
23:11:33***Saving seen data "./dancer.seen"
23:11:45linuxstbNico_P: IIRC, something like "mknkboot nk-original.bin bootloader.bin nk-merged.bin"
23:12:09Nico_Plinuxstb: yes, but isn't it called by the make process? do I have to run it myself?
23:12:23linuxstbNo, it requires the original firmware, so isn't part of the build system.
23:12:42Nico_Pah well then that's why I wasn't able to boot it :)
23:12:46linuxstbSo you run it yourself.
23:12:56jmspeexpreglow: pong
23:13:49linuxstbNico_P: I'm not sure how the upgrade process works, but I have a feeling it may be possible to transfer code to the Gigabeat that isn't flashed, but just executed. That would be perfect for testing...
23:13:53 Join Robin0800 [0] (
23:13:58preglowjmspeex: got a new patch, if you're interested
23:14:11Nico_Plinuxstb: that would indeed be perfect, how would it work?
23:14:56linuxstbNico_P: That's for you to find out ;) I think I read something about transferring very large files (e.g. 16MB) would prevent flashing - probably in the description of the nk.bin file format.
23:15:11Robin0800I see usb stack has just been commited what does this do?
23:15:25linuxstbNico_P: (the file format docs I read)
23:15:35Nico_Plinuxstb: thanks
23:16:32Nico_Plinuxstb: the bootloader file that is required (bootloader.bin)... is it eboot.bin from the OF?
23:16:35ZagorRobin0800: currently all it does is ask the hub for full power
23:16:55markunlinuxstb: is part of the nk.bin file flashed?
23:17:29linuxstbNico_P: bootloader.bin is Rockbox code - the aim is to inject this into the firmware update so it's run. I don't understand the details though, mknkboot just does what ptw419's (I think) merge0.cpp did, but cleaner.
23:17:42linuxstbmarkun: I've no idea what happens to it...
23:17:53jmspeexpreglow: I'm interested
23:17:54markunI suspect that nothing gets flashed
23:18:04Robin0800Zagor,so improves charging from usb?
23:18:07linuxstbmarkun: Although there seems to be a firmware FAT partition, so maybe it's just written there?
23:18:19markunlinuxstb: yes, the nk.bin file is located there
23:18:24ZagorRobin0800: yes, that's the idea
23:18:48preglowjmspeex: this time i've replaced all the non-overlapping SPEEX_MOVEs with _COPY
23:18:56preglowjmspeex: and no small copies/sets
23:19:05Nico_Plinuxstb: the bootloader build produces an nk.bin file...
23:19:11preglowjmspeex: btw, how to handle the bfin speex_move?
23:19:25Nico_Pshould I insert it as the bootloader?
23:19:30 Quit animeloe ("This computer has gone to sleep")
23:20:09Robin0800Zagor,does it still use apples disc mode or has it got its own?
23:20:24ZagorRobin0800: still OF disk mode
23:20:30jmspeexpreglow: I think the bfin function should go until it's replaced with one that complies with the definition
23:20:58 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
23:21:49jmspeexpreglow: I think it's good to only use mem* for 40 and above. That way, we know we can be a bit smarter at optimising without slowing down short copies
23:22:20Zagorlinuxstb: you have an ipod video?
23:22:35linuxstbZagor: An ipod Color and an ipod Video
23:23:55preglowjmspeex: yeah, agree
23:24:19 Join austriancoder [0] (n=austrian@rockbox/developer/austriancoder)
23:25:17Nico_Phmm... nothing happens after the progressbar finishes filling
23:26:52*amiconn wonders about the delta differences between PP502x targets
23:27:52Zagoramiconn: beats me
23:28:31Zagorunless you mean the big red/green variations? those are because the new stack is only included in sansa and h10
23:29:16 Quit krabador (Read error: 110 (Connection timed out))
23:29:43amiconnThere's ~900 bytes difference between sansas and h10
23:29:53amiconnAnd why is the new stack not included for ipod?
23:30:18preglowjmspeex: btw, is it just me or can't the bfin speex_move copy chars?
23:30:20Zagoramiconn: because afaik nobody ever tested the patches in ipod
23:30:34preglowjmspeex: what's the smallest bfin data type? 16 bits?
23:30:49Zagornow it turns out I broke usb for ipods anyway, but the idea was not to :)
23:31:43Zagorthe h10 build needs to be tested. the delta difference could be due to missing code.
23:31:50 Quit Domonoky (Read error: 104 (Connection reset by peer))
23:32:02jmspeexpreglow: Blackfin has chars
23:32:14jmspeexbut I made sure not to use that function on chars
23:32:24jmspeexunless they'd have to be 16-bit aligned anyway
23:33:03Zagoris the ipod auto-boot a holy feature? it's a rather ugly hack imho
23:33:52 Join qweru [0] (
23:34:02n1sZagor: you would have to put a whine filter on irc/forums/ml/fs/wiki if you kill it :)
23:34:10amiconnZagor: huh?
23:34:15Zagorn1s: yeah I guess
23:34:24preglowauto-boot? boot on usb?
23:34:45Zagorpreglow: reboot-to-of-on-usb-insert
23:34:49preglowwe don't want to lose that, no, not until we have usb
23:35:00preglowdeveloping would be a nightmare without it
23:35:19preglowno keys presed on usb insert, no keys pressed on usb unplug
23:35:26Zagormanual reboot = nightmare? welcome to sansa land...
23:35:48amiconnpreglow: Well, the reboot doesn't always work
23:35:49preglowand i don't envy you
23:35:59preglowamiconn: it works plenty more times than it doesn't, depends on the build
23:36:27pixelmaZagor: the boot to OF from off works for me... sometimes (maybe 50% of the time)
23:36:50Zagorpixelma: from off it works every time for me. I was referring to from rockbox.
23:36:51amiconnZagor: This reboot isn't a hack, btw. On early ipods, it's exactly what the OF does when it detects storage connection (though that's firewire)
23:37:11pixelmaZagor: oops, that's what I meant too
23:37:13amiconnAll ipod ROMs have this magic-string mechanism
23:37:17Zagorit's still a hack, even if other people do it
23:37:29 Quit Daniel_S ("CGI:IRC")
23:37:35amiconnIt just tells the apple loader what to do
23:37:48Zagorwe must be talking about different things. I mean the "start stack and see if a computer responds" hack.
23:38:58Robin0800preglow,I find the disconnect from usb to always be ok if mode started from rockbox but mode fom apple disc mode always crashes player on disconnect
23:39:29Zagorpreglow: that's how it detects if its' a charger or a computer connecting. it starts the stack and looks for traffic for a while. if no traffic, it's a charger...
23:39:53Zagors/it/we :)
23:40:37preglowZagor: i didn't even know about that
23:40:42preglowbut then i don't have a charger
23:42:58 Quit lee-qid ("aufwiederbyebientotsayonara")
23:44:23 Quit moos ("Rockbox rules the DAP world")
23:45:28linuxstbZagor: Yes, it's a hack, but it works nicely. You insert USB, your ipod reboots into disk mode, you extract usb, your ipod reboots back into Rockbox...
23:46:44preglowthat's not what he's talking about
23:46:54Zagoractually it is
23:47:05Zagoror, partly at least
23:47:06preglowi thought you were talking about charging now :/
23:47:54Zagorwell yes, specifically I'm talking about the detection of charger/computer. I have no objection to the rebooting, but the charger detection is rather messy.
23:48:06linuxstbIf you can suggest a better way....
23:48:38amiconnlinuxstb: In 50% of all cases, at best...
23:49:05linuxstbamiconn: You're talking about the bug that stops the reboot?
23:49:07Zagorhow about we tell the charger users to hold down a button too?
23:49:14preglowwhere'd we get that detection code from?
23:49:16linuxstbZagor: That's not a hack? ;)
23:49:52Zagorlinuxstb: is it a hack for computer users?
23:49:53linuxstbamiconn: Do you have dircache enabled on the targets where you experience that? On my ipod color, dircache enabled -> freeze, dircache disabled -> reboot nicely.
23:50:06amiconnDircache doesn't matter
23:50:17Zagorthe separation is silly imho, and causes much ugliness in our code
23:50:17linuxstbIt reliably does for me...
23:50:21jmspeexpreglow: I think the memset you've used in bits.c is actually useless (i.e. the loop it replaces was useless). Can you confirm since you use it?
23:50:33amiconnThe target that freezes most often is the ipod mini - and there I don't use dircache because of the quick spinup
23:50:57linuxstbZagor: There were _lots_ of complaints before about ipods rebooting when you insert the charger - ipod's don't have a separate AC socket, they can only charge via usb.
23:52:00advcomp2019Zagor, sorry but my e280r is not charging in rockbox as far as i can see
23:52:00 Quit hannesd ("Client suicide")
23:52:16linuxstbZagor: It also causes reboot cycles - if you insert a charger when the ipod is switched off, it would boot Rockbox, Rockbox would detect USB, so it would reboot to disk mode, disk mode would detect no computer, so reboot back to Rockbox, and so on....
23:52:57Zagorlinuxstb: true. alright, I'll add the darn thing
23:53:24amiconnlinuxstb: They can also charge via firewire - even those which don't support firewire data transfer :)
23:53:25Zagoradvcomp2019: and it did before my commit?
23:54:08advcomp2019Zagor, yea, it was slow but it did
23:54:10 Join hannesd [0] (
23:54:13Soapif I'm reading this battery_bench right, the Sansa E200 in question does not appear to charge while attached to USB and in Rockbox.
23:54:23 Join darkapostrophe__ [0] (
23:55:42Soapnote the old SVN number, if that has any bearing on the topic at hand.
23:57:35amiconnlinuxstb: Where is this detection method from, btw?
23:57:51ZagorSoap: previously we used whatever power we got from the hub. I just committed a patch that requests full power from the hub. that *should* help
23:58:03amiconnMaybe there's an input bit on GPO32 and it's actually GPIO32 ??
23:58:29preglowjmspeex: i'll have a look

Previous day | Next day