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

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

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

#rockbox log for 2005-06-12

00:10:04 Join webguest47 [0] (
00:10:36webguest47Hello guys, quick question. Is there a way to make my own GB ROMS with the disks I still have?
00:10:42webguest47(a simple way)
00:11:21yngwii guess no, because you would need a "game reader" or transfer device to the pc
00:11:43yngwibut downloading games which you physically own is perfectly legal i think
00:11:45thegeek_just download them
00:11:52thegeek_they are just gb roms
00:11:59thegeek_I doubt anyone would care
00:12:19yngwii guess not, but if he owns them anyways..
00:12:21 Join RED_M_CHIU [0] (
00:12:35webguest47lol ok, so finding sites will be easy? I have not attempted yet
00:12:41thegeek_try using emule
00:12:44thegeek_it has everything
00:12:48webguest47ok thanks
00:12:49thegeek_and for something as small as a gb rom
00:12:54thegeek_it will be fast enough
00:12:57yngwithis is the best option
00:13:07webguest47All I really want, is some mario ROMS which I own the disks anyway
00:13:07 Quit webguest47 (Client Quit)
00:13:26 Join webguest47 [0] (
00:13:38webguest47all I want is the mario roms which I own anyway
00:13:58webguest47just another note, how big are the .gb files usally so I know when looking on emule?
00:14:07yngwia few kB
00:14:15yngwior so
00:14:16thegeek_max 32 kb?
00:14:17thegeek_I think
00:14:20webguest47double or triple?
00:14:23thegeek_by most are a lot smaller
00:14:24webguest47oh ok they are small then
00:15:12thegeek_I searched for "mario .gb"
00:15:13thegeek_on emule
00:15:16thegeek_and got lots of results
00:15:27webguest47thanks, Ill check it out as soon as I get home
00:15:29thegeek_you should sort by "availability"
00:15:33thegeek_for faster download
00:15:52webguest47is the rockboy fully complete? or still being built?
00:15:58webguest47I havent heard any progress in a while
00:16:24thegeek_I think Hcl said he would work a bit more on it
00:16:30thegeek_when he was done with the database
00:16:33thegeek_right now
00:16:40thegeek_what it needs is optimization
00:16:44thegeek_and that's boring;)
00:16:49thegeek_so it takes time
00:16:52webguest47ahh, so otherwise it works fine?
00:16:54thegeek_(not to mention it's very hard)
00:16:56thegeek_it's slow
00:17:00thegeek_but it works ok
00:17:16webguest47i feel like my iriver is a whole new device now
00:17:24webguest47good work guys
00:17:25yngwii tried it with super mario world and yeah, its slow but it works already
00:18:04preglowwebguest47: be aware that rockboy needs work
00:18:36 Quit ShockerEngr (Read error: 60 (Operation timed out))
00:18:39webguest47well I understand whole of rockbox is still in prealpha
00:18:55preglowbut gameboy emulation is pretty slow now
00:19:18webguest47there is no additinal things I need to install other than the .gb files though right?
00:19:18preglowthere's tons of room for improvement, someone just has to do it
00:19:18 Quit webguest47 ("CGI:IRC (EOF)")
00:19:28 Quit Bger (Read error: 110 (Connection timed out))
00:21:07*preglow finds a bug
00:22:00yngwiyou can keep it :-)
00:22:09yngwibut thanks anyways
00:22:20 Quit Stryke` (Read error: 110 (Connection timed out))
00:24:53preglowthe enc_padding i read is sometimes wrong
00:26:52preglowno it isn't
00:31:31 Join Stryke` [0] (
00:32:03Stryke`freenode hates me
00:35:59 Join Bger [0] (~Bager@
00:36:35preglowstatic packet_decoder_factory<packet_decoder_mp3> EHEHEHEHE;
00:36:46preglownot often i laught while reading source code
00:37:45 Quit yngwi ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
00:42:13 Quit [solid] ("leaving")
00:43:58 Quit lodesi (Read error: 113 (No route to host))
00:47:37webguest65preglow: take a break before you turn into Dr Jeckyle/Hyde
00:52:04preglowoh, i will
00:52:10preglowten more minutes
00:52:25preglowbut it's so hard stopping when i've nearly got the bug ;)
00:53:16*webguest65 hands preglow a frosty beer
00:57:15*preglow appreciates
01:00:26 Quit Querty (Remote closed the connection)
01:05:52 Quit Shagnar ("( :: NoNameScript 3.81 :: )")
01:06:49 Quit Stryke` ("Friends don't let friends listen to Anti-Flag")
01:08:46 Quit webguest65 ("CGI:IRC")
01:21:40preglownot very easy to debug codecs, i'd say
01:26:39MoosCamarocoura ;-)
01:31:44preglowno, sleep
01:31:46 Quit preglow ("leaving")
01:37:01***Saving seen data "./dancer.seen"
01:45:43 Join jpburton5150 [0] (
01:49:04 Quit markun ()
01:51:54jpburton5150where is this crossfading feature in the menus?
01:51:59jpburton5150haha i cannot find it for the life of me...
01:52:24jpburton5150(im using a bleeding edge build from about an hour ago)
01:52:37jpburton5150(on an iriver)
01:54:23 Quit Bger (Read error: 113 (No route to host))
01:54:26 Join Bger [0] (~Bager@
01:56:07]RowaN[jpburton5150: general settings > playback > crossfading
01:56:21]RowaN[although it does nothing for me when enabled (ihp-120)
01:57:00 Join stripwax_ [0] (
01:57:42jpburton5150hmm well its mysteriosly not there fore me...
02:01:33jpburton5150hey look its there (with another download)
02:01:38jpburton5150hmm, nevermind
02:04:29jpburton5150yeah it doesnt seem to do much with me either...
02:07:01HClits under development.
02:07:06HCldon't expect anything.
02:14:56 Join yngwi [0] (
02:21:40 Quit Sucka ("a bird in the bush is worth two in your house")
02:27:04 Quit jpburton5150 ()
02:50:35RoriWar of the Worlds pulled amid pirate fears
02:54:06yngwimy benchmark is now well over 16 hours... (17 in fact)
02:54:18bipak_hrhr :>
02:54:26Rorifor what?
02:54:26bipak_still awake ? :>
02:54:29 Join ashridah [0] (
02:54:51Rorifor what?
02:55:04yngwimp3 on normal iriver firmware
02:55:11bill2or3it's *still* going?
02:55:28Roriah not Rockbox fw
02:55:32yngwibut i don't understand why
02:55:37yngwione bar left
02:55:53yngwiI would say It's too long
02:56:07yngwiBut I couldn't make an error??
02:56:11bipak_yngwi: your iriver is corrupted ;)
02:56:33yngwi128k, a folder with 12 files on repeat directory.. sounds ok?
02:56:50bipak_yeah :)
02:56:52yngwimaybe they shipping now with the 2200 batteries :-)
02:57:28yngwioh, and vol. 20
02:57:35yngwithe standard setup
02:57:57yngwiim curious how long it will last with rockbox
02:58:10yngwimaybe try out tomorrow??
02:58:29yngwibut it's boring, maybe let slasheri do it :-)
02:58:41bipak_do it both :)
03:15:28 Part MoosCamaro
03:16:00 Quit yngwi ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
03:17:01 Join yngwi [0] (
03:32:08 Join amiconn_ [0] (
03:32:41 Join webguest98 [0] (
03:33:42 Quit webguest98 (Client Quit)
03:33:47 Join webguest63 [0] (
03:37:04***Saving seen data "./dancer.seen"
03:37:12webguest63are there any developers here?
03:37:31yngwiok, i don't think so
03:37:42yngwihaven't heard of any for a few hours
03:38:14yngwibut i don't know
03:38:20webguest63is it supposed to be this quiet?
03:38:25HClits 3:30 am.
03:38:28HClwhat do you expect.
03:38:46bipak_same time here
03:38:48webguest63what? it's 9:38pm here
03:38:48*HCl is only still here cause his gf hasn't gone to bed yet even though he told her to go to bed :/
03:38:57yngwiyeah 3:30
03:39:03ashridahwebguest63: the world is round :)
03:39:07webguest63where do you guys live?
03:39:10yngwiand im going down... but my player not
03:39:27bipak_lol yngwi
03:39:43ashridahwebguest63: a lot of the developers are in northern europe.
03:39:46bipak_you got a fake ihp :p
03:39:54webguest63didn't know that
03:40:20bipak_it plays too long :>
03:40:37webguest63i thought they were americans
03:40:38 Quit amiconn (Read error: 145 (Connection timed out))
03:40:38 Nick amiconn_ is now known as amiconn (
03:41:00yngwioh i'm tired, i read ip not ihp... :-)
03:41:20yngwi17 hours and one bar left...
03:41:24ashridahwebguest63: the original rockbox site was if i recall. is sweden
03:41:48ashridahit's not so obvious now that they've got access to
03:42:28webguest63they speak good english then
03:42:28ashridahbut yeah, the names of the developers can also be a dead giveaway.
03:43:13ashridaheurope often uses a few major languages as common languages.
03:43:16ashridahenglish is one of them
03:43:25yngwilatin the other :-)
03:44:00ashridahdo people even speak latin anymore?
03:44:19yngwino :-)
03:44:38yngwiexcept myself of course.... :-)
03:44:52yngwino, thats not true, i'm lying
03:45:01bipak_yngwi is a latin lover ;)
03:45:11yngwirather not
03:45:22yngwiat the moment a latin hater
03:45:44*bipak_ is happy that he never had to learn latin
03:46:01bipak_such a nonsense language :)
03:46:21yngwii wish i had in school so i didn't have to now
03:46:35ashridahwebguest63: yeah, so anyway, give it about 5-6 hours. people should have woken up then
03:47:01HCli'm gonna be dead tired tomorrow :( gnight..
03:47:02bipak_yngwi: why do you have to learn it now?
03:47:03webguest63thx ashridah
03:47:13bipak_night HCl
03:47:18yngwifor university, i study history
03:47:23yngwigood night
03:47:37webguest63good night, guys
03:47:39 Part webguest63
03:47:48yngwii should hav learned it for 5 years, but now is my last chance
03:47:51bipak_dann geh ich jetz auch ma :p
03:47:54yngwiexam in 5 days
03:47:57yngwiich auch
03:48:06bipak_dein ihp rennt hoffentlich noch, wenn ich aufwache ;)
03:48:07yngwiauf den rest vom test pfeif ich
03:48:11bipak_hehehe :>
03:48:21yngwi17 stunden reicht
03:48:22bipak_n8 :)
03:48:29yngwier soll ja morgen wieder frisch sein
03:48:38yngwigute nacht
03:48:38bipak_dann kommt der grosse rockbox test
03:48:42bipak_jo nacht
03:48:50 Quit yngwi ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
03:51:21*HCl slaps putty
03:52:50 Quit stripwax_ (Read error: 110 (Connection timed out))
03:53:39Rorimaybe one day
04:05:49 Join QT_ [0] (as@area51.users.madwifi)
04:15:41 Quit gromit` (Remote closed the connection)
04:20:09 Quit QT (Read error: 113 (No route to host))
05:37:06***Saving seen data "./dancer.seen"
07:37:09***No seen item changed, no save performed.
07:55:47ashridahthere we go, finally upgraded my m68k cross compiler to 3.4.4
09:18:32 Join textchimp [0] (
09:18:33 Join Harpy [0] (
09:23:18 Join t0mas [0] (
09:27:14t0masmorning :)
09:31:59 Join StrathAFK [0] (
09:37:11***No seen item changed, no save performed.
09:46:06 Quit Strath (Read error: 110 (Connection timed out))
10:02:12 Join markun [0] (
10:02:26markunmorning t0mas!
10:47:34 Join niobos [0] (
10:47:38niobosnorning all
10:48:59Slasherihi :)
10:49:13Slasherimp3 battery test: ~11h 30 min.
10:49:31t0masthat's almost iRivers 14 hours
10:49:33nioboson H120?
10:49:49Slasherii think vorbis should do same results or better
10:50:06amiconnhi Slasheri
10:50:14Slasheriamiconn :)
10:50:21amiconnBattery runtime isn't bad :)
10:50:43nioboshmm, H120 vs H140 shouldn't differ much, I guess
10:51:06amiconnH-120 will probably run a little longer regardless of firmware (iriver or rockbox)
10:51:34niobosHDD uses less power???
10:51:37amiconnThe disk needs less power to spin up (1 platter instead of 2, so less inertia)
10:52:10*niobos learns something new every time he comes around here...
10:52:22niobosnow the number of platters in my player...
10:53:17niobosSlasheri: that test was just play and keep playing, I guess, no next/pause/... things
10:54:26amiconnSlasheri: Unfortunately I have to report a strange bug :(
10:55:05Slasheriamiconn: Hmm, please do so
10:55:28amiconnI started playing a dir of oggs, and while playing the 3rd track, my iriver suddenly powered off (no battery problem)
10:55:32Slasheriniobos: Yep, just play. No user interaction at all
10:55:47Slasherihmm, strange
10:56:06amiconnThen I restarted the iriver, accepted resume, and it started playing the 1st track from the beginning(!)
10:56:06niobosSlasheri: and all on rockbox, i suppose
10:56:34Slasheriniobos: of course, test was never stopped/resumed
10:56:40amiconnI skipped forward to the 3rd. Further within the track, it suddenly stopped again (no poweroff this time)
10:57:29SlasheriHmm, what if you start playing the 3rd track without skipping?
10:58:00amiconnAnd btw, the bitrate display for oggs seems to be a bit off (probably something for linuxstb). All these files show 256 kbps, although the real bitrate is lower (around 210)
11:00:47 Nick QT_ is now known as QT (as@area51.users.madwifi)
11:01:57amiconnI just found something very unusual - the line output level depends on the volume setting (!)
11:02:19SlasheriThat's same with iriver's firmware
11:02:40SlasheriBoth headphone jack and line out outputs the same level
11:02:56SlasheriBut i think there is some filtering between those two jacks
11:03:13amiconnHmm, perhaps a hardware restriction. Imho then this is not true line out
11:03:38amiconnThe archos player (only archos with a separate line out) does have true line out
11:05:03SlasheriIf you want use line out as true line out on iriver, you have to put volume level near 100%
11:06:51amiconnYes, but I'd still not call this true line out. True line out would allow to use headphone and line out at the same time, without one influencing the other
11:07:14amiconnBtw, there was a pop in the headphone when I plugged the line out
11:07:15markunamiconn: I also had the power-off problem. Turned off "Idle Poweroff".
11:07:59markunnow I don't have any power-off's anymore, so maybe it just thought rockbox was idling while it was in fact playing music.
11:08:48markunThe plop is still the 'dac is not reset' problem I think.
11:09:26amiconnSlasheri: Hmm, now it played that 3rd track. I'll let it continue playing...
11:09:57Slasheriamiconn: ok, i am also playing some oggs now
11:10:03ashridahmm, i've had cases where it drops to the menu as tho someone hit 'stop' with oggs
11:10:15ashridahjust haven't played anything with ogg in it in a few days tho
11:10:21amiconnmarkun: I mean something different. I had the iriver playing already, with headphone. Inserting the line out caused a plop
11:10:56amiconnThat's caused by the hardware design
11:11:14markunamiconn: Do you get the same with the original firmware?
11:11:14Slasheriamiconn: Hmm, maybe the line out is not separated from headphone jack at all :/ There might be only some filter capasitors on the line out
11:12:43markunLast night I've been reading on how to insert tracknumer and CD text into the spdif output. Still not completely clear to me.
11:13:07ashridahweird. i'm noticing some static added if i plug headphones into the line-out jack that isn't there when i use the headphone jack
11:13:48amiconnmarkun: I dunno, just noticed it. I didn't use the original firmware much
11:14:09markunsome people complained about center-scrolling. I would like to make a menu option for it under general -> navigation -> center-scrolling.
11:14:56amiconnPerhaps there is an even better way
11:15:05markunAnd I would also like to put the patch that stops scrolling at the end (if you keep the joystick pressed) as an option there too.
11:16:03markunamiconn: what do you sugest?
11:16:06amiconnIf the scrolling upper/ lower boundary is calculated with the same method as the cursor positioning in keyboard.c it would even allow to adjust the 'center' scrolling
11:16:31amiconnIt would be more like a 'scoll look-ahead' setting
11:16:45amiconnSetting it to 0 would give the old behaviour
11:16:53markunyes, that would be better.
11:17:01markuncenter-scrolling is not a good name anymore.
11:17:50markunamiconn: Do you have a spdif receiver that can display tracknumer + elapsed time and/or cd-text?
11:18:16amiconnI don't have _any_ optical spdif receiver
11:18:26 Join LinusN [0] (
11:18:36LinusNmarkun: maybe a scroll-margin setting?
11:18:55amiconnLinusN: Sounds like a good name
11:19:05LinusNwhere 0 means the previous behaviour
11:19:11markunLinusN: Yes, I like the relative margin we now have.
11:19:12LinusNperhaps in %
11:19:37amiconnmarkun: I only have one sender (my dvd player) and a coax sender/receiver (my archos recorder)
11:20:24markunI think I will borrow a friend's md player to play around with tracknumer changing
11:20:48amiconnI doubt that I can read all this spdif info from the mas, and I don't have an optical cable that fits into the iriver
11:21:47 Join yngwi [0] (
11:22:13amiconnSlasheri: Now it repeats a track (track 6) :(
11:22:49Slasheriamiconn: Hmm.. :/ Maybe i should do more testing with vorbi
11:23:28amiconnHaha, track title is 'Never-ending story'
11:23:51yngwigood morning
11:24:44ashridahamiconn: rofl
11:25:01bipak_good morning yngwi :)
11:25:02niobosamiconn: well, you asked for it with that track ;-)
11:25:19yngwi:-) should be awake for hours...
11:26:19markunIf I insert a directory in the playlist all the tracks are inserted right after the current track, but I thought the manual said they would be inserted at the end..
11:28:49yngwithere is something wrong with the resume function..
11:29:36yngwibut i guess you know that already
11:29:46Slasheriyngwi: resume (seeking to correct file position) works currently only with mp3 files
11:30:30amiconnSlasheri: I think this behaviour was observed already: after track 6 was played twice, track 7 was skipped
11:30:53Slasheriamiconn: Hmm.. With ogg files?
11:31:30Slasheriok, that's needs then some debugging
11:31:39yngwii mean for mp3; if I switch the player of with resum on, on startup of the resumed mp3 there is a nasty scratch. it always startles me.. :-)
11:32:15Slasheriah, that is a known problem :)
11:32:37yngwiok, thanks
11:34:56*LinusN is going on a vacation
11:35:20*LinusN packs his laptop :-)
11:36:42SlasheriLinusN: have a nice vacation ;)
11:37:15***Saving seen data "./dancer.seen"
11:37:40yngwihave fun and come back with full strength :-)
11:37:57LinusNwill try
11:38:10markunI will try not to make any stupid commits while you are away :)
11:38:22amiconnSlasheri: The rest of the album played fine.
11:38:30amiconnGtg now, back in the evening
11:38:38LinusNimmediately after the vacation i'll work for 3 days as a chef on a sailing boat :-)
11:39:00LinusN3 days without internet...shrug
11:44:12niobosLinusN: don't worry, internet will survive :-p
11:44:41yngwisounds like a movie title: "Chef for three days" starring LinusN
11:45:18 Join preglow [0] (
11:47:06preglowlooks like we'll have wavpack encoding
11:47:10yngwiif it was me going to be chef, i'd say "Chef for three days, fool for a lifetime" :_)
11:48:45*niobos tries to get his concentration back together and start learning those damn courses...
11:48:51 Nick niobos is now known as niobos_study (
11:52:01preglowSlasheri: here?
11:53:54*LinusN - the swedish chef goes sailing
11:54:56t0masanybody here has latin in school?
11:55:03t0mas(or had)
11:55:15*t0mas has a little problem with the following line:
11:55:17t0mas"Oggi alle 12 e mezza ant cessava di vivere Everdine. Dopo breve e dolorosa malattia."
11:55:27t0mas(Everdine = a name)
11:56:55yngwii'm currently learnin latin, but it seems like i can't help :-(
11:57:22t0mashm... I've dropped the subject after 2 years :)
11:57:32t0masand that was 3 years ago...
11:57:37yngwii have to for university...
11:57:51LinusNt0mas: i had latin in school
11:58:02t0masyes, me too... but I wasn't really good at it :)
11:58:05t0mascan you read this LinusN?
11:58:17LinusNlooks more like italian to me
11:58:31t0masyes, i tought so..
11:58:39t0masbut someone told me I should be able to read it...
11:59:11yngwithe words are pretty similar in latin and italian
11:59:13LinusNyeah, i had good use of my latin when i learned other languages
11:59:19 Join matsl [0] (
11:59:25t0masjup... even french is like it a lot...
12:00:01preglowa lot of western languages are
12:00:12preglowsame concepts reused everywhere
12:00:31t0masjup... but I'm bad at languages... so dropped latin and greek long ago
12:00:39yngwithey are not called "romanic" languages for nothing.. :-)
12:01:12preglowi think languages are really interesting stuff
12:01:27t0masprogramming languages ;)
12:01:46*t0mas is dyslectic (is that the correct English spelling?)
12:01:55preglowthose too, but they usually aren't too interesting
12:02:02preglowwith the notable exception of perl!
12:03:48t0mas"Oggi alle 12 e mezza ant cessava di vivere Everdine. Dopo breve e dolorosa malattia." -> something with cessava di vivere = end of life
12:03:51LinusNA sad story, btw, poor little Everdine died at 00:30 after a short period of painfull illness
12:04:28*LinusN had latin, english, french and italian
12:04:44LinusNi sucked at italian though
12:04:46*t0mas latin, greek, english, german and french
12:04:56t0masbut dropped latin and greek
12:05:06LinusNi learned german by reading 64'er
12:05:24yngwii only know german, english, soon latin and the babys language..
12:05:28t0masit's obligated here... for VWO
12:05:41t0mas(VWO = highschool like.. highest level... needed for university)
12:05:54*LinusN packs his car
12:06:19preglowooh, i'll soon have that
12:06:29t0mas1 week for me :D
12:06:37t0masand then 8 weeks of doing nothing :)
12:06:54t0maswell... working... but that's not to bad :)
12:07:12preglowcouple of weeks here, then i'll be moving
12:07:16preglowwhich i'm really looking forward to
12:07:45*t0mas goes to a little town right next to St. Tropez
12:07:55t0maswith my girlfriend and her parents
12:08:39yngwiyou're lucky, i'm going nowhere..
12:08:55t0masmy parents aren't going anywhere either...
12:09:03preglowi'm going to england for a while
12:09:07preglowwhich i'm too looking forward to
12:09:13preglowohh, the beers i will drink
12:09:15yngwi...except back to my latin books :-)
12:09:15t0maspreglow: by plane?
12:09:22preglowt0mas: plane, yes
12:09:34t0masthen wave when above Amsterdam ;)
12:09:54preglowlike mad
12:09:56preglowi'll yell too
12:09:57t0masghehe :)
12:10:08*LinusN is going to a small island on the west coast of sweden, lots of fishing
12:10:53yngwii'd be happy as a happy cow if there would be any decent weather in austria
12:11:00yngwiraining sucks...
12:11:53t0masyngwi: ah... you have great winters
12:12:03t0massnowboarding in your home country
12:12:16yngwiyeah, but the winter is over for even 4 months..
12:12:20yngwion calendar
12:12:31preglowi so bloody badly want to go fishing
12:12:49t0maspreglow: you don't have water around?
12:13:01yngwibut sometimes i think the guy who makes the weather doesn't have a calendar
12:13:08preglownot much that is fishable
12:13:37yngwiwith the right bait, everything is fishable :-P
12:14:01preglowi don't want seeweed biting
12:14:05preglowseaweed even
12:14:18preglowanywho, that'll come later
12:14:32preglowright now i wish i just could get gapless mp3 working quickly, so i don't have to think about it
12:14:38*preglow produlates Slasheri
12:15:01*t0mas continues writing about this book...
12:15:15t0mas(still reading french... and looking up 33% of the words :P)
12:15:40t0masI'm so happy I don't have to write the essay in French
12:15:55 Nick yngwi is now known as yngwi_away (
12:16:00linuxstbamiconn: I know the Ogg bitrate is not accurate - Rockbox is currently displaying the "nominal bitrate" from the headers. The right thing to do is to calculate it based on data size and total length. I'll get there eventually.
12:18:10preglowi've got codecmpa knowing exact track length
12:18:15preglowand what amounts to skip here and there
12:18:24preglowbut it somehow ends up generating less data than it should
12:19:22 Join Sucka [0] (
12:37:33 Join hubbel [0] (
12:38:09hubbelfirst working sound recording from iriver!
12:39:40preglowwooot, god damnit!
12:39:57Slasheripreglow: now i am here :)
12:39:58hubbeli was lucky and think i've got freqency and clock settings right directly..
12:40:47hubbelSlasheri: when you're done with playback, are you up to integrate recording? =)
12:40:51preglowSlasheri: codecmpa gets exited before samplescount == 0 here
12:41:06hubbelSlasheri: great work btw!
12:41:17preglowSlasheri: and please don't rip out mp3_get_info!
12:41:23preglowget_mp3_info, wtf
12:41:27Slasherihubbel: yes, i think so :)
12:41:38Slasheripreglow: ah, you can remove the whole samplecount stuff from there
12:41:52preglowSlasheri: well, no, i need it
12:42:09Slasheripreglow: the mp3_get_info was too slow for some files..
12:42:14preglowfor gapless
12:42:15SlasheriIt caused pcm buffer to run out of data
12:42:18SlasheriAh, ok
12:42:20preglowi also need get_info
12:42:24preglowit has to be fixed, then
12:42:37preglowwithout get_mp3_info we can kiss gapless goodbye
12:42:41Slasherii think that some yields should be added to mp3data.c
12:42:46SlasheriThen it would work better
12:42:54preglowi'll try it afterwards
12:42:55Slasheriok, then you have to re-enable it
12:43:01preglowi already have
12:43:07preglownot commited, though
12:43:11preglowbut yes
12:43:18preglowi should be having gapless sound now
12:43:41preglowfor some files it is gapless, for some others codecmpa doesn't render the final samples
12:43:47Slasherioh, thats great :)
12:43:49t0mashm? my mp3 playback was already gapless
12:44:08t0mas(without crossfade)
12:44:17preglowmine surely wasn't
12:44:21t0masSlasheri cut the silence of the file wasn't it?
12:44:38linuxstbt0mas: Do you encode with lame −−nogap ?
12:44:39Slasherit0mas: yep, beginning of file. But that was not enough
12:45:08t0maslinuxstb: yes, and with all wav's in 1 dir
12:45:36t0masbut the album I tested with as 1 large mp3 file... cutted with a cue-sheet
12:45:47preglowi don't get this, codecmpa should never exit unless samplescount == 0, but somehow it happens anyway
12:45:58t0maszo there were no empty last frames (same as −−nogap does)
12:46:06 Quit textchimp (Read error: 60 (Operation timed out))
12:46:24linuxstbt0mas: Yes, that's always worked. preglow is working on removing the silence added at the end of MP3s.
12:48:35t0maslinuxstb: not in iriver fw ;)
12:49:07hubbelLinusN: I wonder why the original firmware uses interrupts to copy from PDIR2 to their input buffer and not DMA, which should be possible
12:49:44preglowhave you tried?
12:49:48hubbelnot yet
12:49:55preglowthen you'll find out ;)
12:50:31Slasherihubbel: I wouldn't be suprised at all. The dma controller is very buggy and has caused a lot of trouble with playback already
12:51:01LinusNhubbel: i don't have the faintest idea how the original firmware does it
12:51:56hubbelSlasheri: in what ways is it buggy?
12:52:16Slasheriit's hard to use the dma without making system to crash :)
12:52:55 Join MoosCamaro [0] (
12:53:03hubbelHehe, yeah, I noticed that until I got it to work the first time :)
12:53:16 Join DMJC [0] (
12:53:18MoosCamaroMorning all
12:53:40 Part DMJC ("Leaving")
13:02:38 Quit edx (Read error: 145 (Connection timed out))
13:04:36preglowlibmad just doesn't want to render the final samples, it seems
13:05:12linuxstbpreglow: Maybe you need to flush the internal Mad buffers or something.
13:05:54preglowmad has no internal buffers of that kind
13:06:04preglowi ask for a frame, it gives me a frame
13:06:25linuxstbSo what does it give you for the final frame in a file?
13:07:23preglowi haven't looked at it per frame yet
13:07:50preglowi just know that according to calculations based on encoder delay and padding, which coincides with the ones foobar2000 uses, i should have more data incoming
13:08:17linuxstbDoes foobar2000 also use libmad?
13:08:25preglowif you want it to
13:08:34preglowso yes, libmad should handle this
13:08:41preglowbut i really, really have no time to work in this any more
13:08:47preglowanyone else wanna have a shot?
13:08:51linuxstbYou said that 24 hours ago :-).
13:08:55preglowi know
13:08:58preglowand now i'm really short on time
13:09:17linuxstbWhy don't you commit what you've done so far - with a comment indicating the bug?
13:09:22preglowi might
13:09:26preglowbut brb first
13:13:26 Join stripwax_ [0] (
13:13:37stripwax_i don't get it. when I build fwpatcher from source, it always complains that the firmware doesn't match a known checksum but when I run the bin downloaded from it works fine. is fwpatcher really picky about how it's compiled or something? (i'm using cygwin, gcc -mno-cygwin)
13:19:33linuxstbSlasheri: Is there anywhere I can get a large (8K or 16K) buffer from when I'm parsing the Ogg metadata? Can I use some of the audio buffer temporarily?
13:19:47preglowwell, it of course depends on whether the bootloader.bin you use corresponds to checksums.h
13:20:58preglowexport CVSROOT=""
13:20:58preglowexport PATH=/home/thomj/m68k/bin:$PATH
13:21:15preglowsorry bout that
13:23:21stripwax_preglow - hm, well I'm just using the one at (2June). is that not right?
13:23:32preglowshould think so
13:23:46stripwax_preglow - if you'd like to review the diff, should i mail you ,rockbox mailing list, or .. ?
13:23:53stripwax_(as I don't have cvs write access)
13:24:07stripwax_^ this is the command line handler for fwpatcher. thanks
13:28:49Slasherilinuxstb: Hmm, you can use malloc buffer
13:29:19preglowstripwax_: looks good to me, but can't test it since i'm in linux
13:29:36SlasheriBut be aware: if you use malloc, free does not free the data. Only codec_init does that
13:29:52linuxstbI need to use it inside playback.c
13:29:54stripwax_preglow - which bootloader.bin should I be using then
13:30:16Slasherilinuxstb: ah, hmm..
13:30:39preglowstripwax_: well, just try patching a firmware manually with the bootloader.bin in the wiki, and see if the md5 you get coincides with the one fwpatcher contains
13:30:44Slasherimaybe you can use some free portion of file buffer or guard buffer
13:30:54SlasheriBut you should verify the area is free
13:30:55preglowstripwax_: checksums.h contains the md5s fwpatcher uses
13:31:15 Quit hubbel ()
13:31:24stripwax_yep. but really i meant - which bootloader.bin does the fwpatcher.exe binary currently use :) (if it isn't the latest one)
13:31:52linuxstbSlasheri: How can I check that?
13:32:06preglowstripwax_: it should contain the latest bootloader.bin put in the IriverBoot wiki page at all times, i don't know, i'm not the one building it
13:32:54stripwax_preglow ok..
13:33:05preglowstripwax_: you are aware fwpatcher has to be built with the bootloader.bin lying in the build dir, yes?
13:33:13preglowi assume windres would give you an error otherwise
13:33:39Slasherilinuxstb: Hmm, that is harder. You could check codecbufused and if it's less than AUDIO_BUFSIZE, there are free space that much from buf_widx index (but the buffer might wrap)
13:34:02SlasheriI think the guard buffer maybe your best choise, but you need to add some flags to determine if it's used or not
13:34:13SlasheriAnd if it's being used, just wait until it's not
13:34:17preglowthat's a hack...
13:34:26preglowyou COULD just use the stack :)
13:34:37Slasheri(btw, in vorbis guard buffer is always free but you should not rely on that)
13:34:50linuxstbSlasheri: OK, I'll forget about that then. I don't _need_ that large a buffer, but it would have made it a lot simpler.
13:35:46linuxstbThe problem is to search for the last occurrance of the string "OggS" in the Ogg file. I think this will be somewhere within the last 8K-16K.
13:36:33stripwax_preglow yes.. hmm, maybe something else is up. thanks
13:37:18***Saving seen data "./dancer.seen"
13:37:28preglowi can't guarantee fwpatcher contains another bootloader, as i'm not the one who built it
13:40:13LinusNtime to go
13:40:15stripwax_weirdness. must be some hiccup, I rebuilt all and it works just fine now. ok so, it works. wanna cvs it?
13:40:16LinusNcu all
13:40:19preglowlinuxstb: have nice
13:40:23preglowLinusN even
13:40:30 Part LinusN
13:40:33*preglow pats his nick completer
13:43:52markunShould the scroll-margin (center-scrolling) be set in percents? I think that's overkill. I was thinking of "off", "center" and "2/3"
13:44:24preglowhell, if you've first got an option, why not?
13:44:46markunpreglow: why not percents?
13:45:29zedoesn't parsing some tokens and converting them into numbers that you have to place them with anyway just add more code?
13:45:52zeseems like thats the overkill to me, and reduces flexibility in the process
13:46:28markunok, I'll just use a percentage then.
13:49:57stripwax_preglow - you ok to check it in?
13:56:26preglowmight as well
13:56:37preglowcan't see any possibility of your changes breaking md5ing
14:02:02stripwax_ah, doesn't seem to handle spaces in filenames properly..
14:07:42preglowthat does need to be fixed, yes
14:13:26 Join Adam [0] (
14:13:48 Join DaKi][er [0] (
14:14:31Adamquick (quite possibly ill-researched) question: how far along is the remote work? (my main LCD screen and buttons are stuffed)
14:21:13 Join gromit` [0] (~gromit`
14:22:50preglowhow far along?
14:22:55preglowpretty far
14:23:06preglowyou can control the unit with the remote now, at least
14:23:18preglowat least i think you can, hehe
14:23:43stripwax_preglow - win32 sucks. just mailed you the working versino
14:25:50 Quit Adam ()
14:28:14stripwax_preglow - thanks!
14:29:02preglowwin32 sucks indeed
14:30:24linuxstb.Does anyone else have a problem when playing Oggs - If I play an Ogg, and start skipping immediately to the next track, then after about 4 ot 5 skips, playback stops.
14:31:13markunlinuxstb: Yes, I had similar problems.
14:31:40markunAnd the unit still shuts down during playback sometimes.
14:32:10linuxstbI just want to make sure it wasn't caused by my metadata parsing. I don't think it is, because I commented out my Ogg parser, and still got the problems.
14:32:40preglowstripwax_: i'll just fix your brace placement to adhere to mine, i don't like inconcistencies like that in sourcecode
14:33:12linuxstbI can now calculate an accurate bitrate and total length for Ogg files, but I'm sure that code is buggy as well (it's too complicated). I just don't want to confuse things by introducing more playback bugs.
14:33:40stripwax_[preglow sure
14:33:44markunlinuxstb: was the lisp source of use?
14:34:30linuxstbmarkun: A little. It showed me that there is no easy way - you just need to a brute force search for the string "OggS" near the end of a file.
14:35:23linuxstbWe can't use Tremor to do it, because (ATM), we don't allow Tremor to do seeks in the file.
14:36:15linuxstbI also think lack of seeking is causing the Musepack playback to fail.
14:37:48preglowstripwax_: btw, doesn't the command_line parameter to WinMain point to the command line? i don't think it's necessary to call GetCommandLine
14:38:48stripwax_preglow - pretty sure command_line is non-unicode only...
14:39:38preglowi'll commit it
14:40:27preglowyour name again? dave hooper?
14:41:14 Quit ]RowaN[ ()
14:42:01 Join ]RowaN[ [0] (
14:45:50 Join DomZ [0] (
14:51:49preglowok, i'll commit the gapless thing
14:51:55preglowso i've got it out of my hands
14:53:06 Join zezayer [0] (~chatzilla@
14:53:08 Quit zezayer (Client Quit)
14:53:26RoriI will do some testing
14:53:46 Join edx [0] (
14:57:45linuxstbIt seems I can reliably break Ogg playback if I press NEXT before the current track has started to play.
14:59:36 Join Lost-ash [0] (
15:11:05 Join Chamois [0] (
15:18:51 Quit ashridah (Read error: 113 (No route to host))
15:18:52preglowSlasheri: could you try making get_mp3_info work? i don't know what bug it's triggering, and it's needed for the gapless code i'm about to commit
15:21:00crwlhmm, sometimes playing doesn't start when i choose a song from the filebrowser and it's playing something else already
15:21:14crwlthe WPS comes up for a second, music stops and then it goes back to filebrowser
15:21:20crwlat next try playing starts
15:21:53crwloh, and sometimes there can be briefly seen something like "couldn't load codecs/blahblabh" too
15:22:04crwlhaven't noticed this before yesterday or so
15:22:54markunDoes anyone feel like trying out my patch for the 'navigation' menu?
15:23:26markunIt's not finished.. but it works
15:23:27RoriI have a problem with some MP3's cutting off short and not playing the next MP3 in the folder. It ends short on some MP3's and then stops
15:23:40Roriknown bug?
15:23:48crwlmarkun, what does that do?
15:24:08RoriThese are 320k MP3 files
15:24:12markunIt makes center-scrolling and scroll-stop configurable
15:24:30crwli'll try it
15:24:43markunok, thanks
15:24:53markunHope it doesn't break anything
15:25:29Roripreglow is this playback stopping short anything to do with the last commit of gapless in progress?
15:26:14RoriI can't get pass a certain playback time on some tracks. If I fast forward too it only goes so far into the track then thinks the track is finished but it's not
15:26:36Roriit's miscalculating song/file length
15:26:42 Join tucoz [0] (
15:27:08tucozRori: happened to read the logs :). Is that by any chance the last song in a directory?
15:27:25tucozAnd you have repeat off?
15:27:34Rorirepeat is off I think...checks
15:27:56RoriOK it was on let me try again with repeat off
15:28:26crwlmarkun, it seems to work fine, but what does scroll stop do? :)
15:28:59tucozwell, I experienced earlier on, with the last track in a directory, and with repeat off, it stops playback about 20 seconds from end of song
15:29:06markunIt stops at the end of the list if you keep 'down' pressed (same for beginning and 'up')
15:29:20preglowRori: it's cutting short for me as well, but i'm just talking about something like 100ms of music here
15:29:42Rorisong length seems incorrect in the file
15:29:49tucozI guess it's just an unhandled event.
15:29:56preglowrockbox relies blindly on id3 info
15:30:02preglowwhich is completely wrong behaviour
15:30:08markuncrwl: Do you notice the difference?
15:30:10tucozOr was, haven't tested that lately
15:30:55RoriI don't get it
15:31:01Rorithe song length says 4:20
15:31:12Roribut it's cuttonf off several seconds too short
15:31:16crwlmarkun, ah, true. works.
15:31:16Roricutting off
15:31:36Rorithe song is cutting off by several seconds which is insane!
15:32:10markuncrwl: Do you think I should have put them in an existing menu or is navigation fine?
15:32:10tucozGreat work anyways you all. Just listening on a flac for the first time in my life.
15:32:17crwlshould rockboy work from cvs, btw?
15:32:49crwlmarkun, it would be kind of logical to put them in scrolling, but then not... because it's different scrolling altogether
15:33:00RoriI play the same MP3 in winamp and it says 4:21 and plays right to the end of the track fine
15:33:21RoriRockbox is not counting right. Like it's counting 4:20 too fast
15:34:03RoriIt literally is not playing several seconds from the end. It's way off
15:34:13preglowRori: what does id3 info say?
15:34:21crwlhum, gtg ->
15:34:29 Quit Chamois ("Leaving")
15:34:43Rorinot v2
15:34:48Roricould that be the problem?
15:35:31Roriwinamp gives length 261 seconds
15:36:05Roridisplays 4:21 but Rockbox displays 4:20 and cuts off several seconds short. Not just 1 second.
15:36:42RoriThese are 320kbps MP3 files
15:36:49preglowthen rockbox reads the incorrect time somehow
15:36:51preglowi have no idea
15:37:01Roriwho to ask?
15:37:19***Saving seen data "./dancer.seen"
15:37:30preglowdunno, i don't know what part of the code figures out the song length
15:37:42RoriI am going to play the track without fast forwarding to see if ff is causing the problem in track seek
15:37:43preglowmy new patch will calculate the song length exactly if the mp3 file supports it
15:37:55preglowseeking is not very well at the moment
15:37:59preglowdon't fast forward
15:38:07Rorilet me play it all the way through
15:38:10RoriI will let you know
15:38:43RoriBTW your gapless works great for Lame files. Not tested other MP3 yet
15:38:51Roriand you have not committed the final code yet?
15:40:11RoriI read back you are waiting on Slasheri for MP3_info or something?
15:40:11preglowi haven't commited it yet
15:40:34Roriso the commit up at present giving gapless is experiemental
15:40:53preglowit's experimental, i'm still having some data cut away at the end
15:40:57preglowbut i know what's wrong, at least
15:41:08preglowit's not gonna work perfectly, i don't have time, i just need to get it off my hands
15:41:24RoriI am sure it will get fixed eventually
15:41:43preglowt0mas: fixed point conversions can get really tricky very fast...
15:43:52Roriok ff is the cause
15:44:01Roriworks perfectly if you don't ff
15:44:13Rorialso I can't hear any joins
15:44:23Rorieven though you say it's not perfect. I hear no glitches
15:44:38preglowdepends on how it's encoded
15:44:42preglowseeking needs a lot of work
15:44:44Roriwhich is good enough for me. Lets try non-lame as a comparison :)
15:44:46preglowit's currently just based on guesswork
15:47:35 Nick Lost-ash is now known as ashridah (
15:47:47Rorigood guess
15:48:50 Join [solid] [0] (~solid@
15:49:34t0maspreglow: hm? shift 8 bits left... and remember to check every division/multiplication?
15:49:44t0maslot of work... but not really impossible?
15:49:46[solid]the vorbis endurance test just started... the tremor opt commit was perfectly timed:)
15:50:05RoriMiika "Fixed MP3 forward seeking." Oh no you didn't! :D
15:51:51RoriNon-Lame playback cuts off too much silence from the end of each track. thus transitions between tracks jump ahead too much
15:51:57preglowt0mas: you think porting floating point code to fixed point is that simple? :PPP
15:52:11RoriI know it's nowt to do with your code preglow so don't jump on me :)
15:52:24preglowmy code isn't even commited
15:52:41preglowsince i don't even know how things are working now, my code might even make things worse
15:52:48preglowdon't kill me if it does, rockbox is work in progress
15:52:57Rorilet me try it :)
15:53:05preglowit should at least make things better
15:53:06RoriI can always go back
15:53:11preglowin the long run
15:54:02t0maspreglow: name some problems on floating -> fixed conversion?
15:54:02RoriI wonder if Slasheri can do anything about not being so harsh cutting the silence off non-lame MP3's
15:54:14t0mas*conversion -> porting
15:55:42 Quit tucoz ("CGI:IRC 0.5.4 (2004/01/29)")
15:55:48preglowt0mas: for one, you might need to do varying levels of precision throughout the file. you'll also need a 64 bit multiply, divisions need to be avoided completely, you must always make sure overflow is impossible
15:57:53t0mashm.. still I think it is possible...
15:57:58preglowof course it's possible
15:58:13preglowit'll just be hard unless you have fixed point experience or are a bloody genious ;)
15:58:35*t0mas starts reading "how to be a genious in 10 minutes."
15:58:43Roriask preglow
15:59:08preglowbut by all means, if you think you can do it, start coding
15:59:48preglowwma support would be great
15:59:49t0masjup... I'll at least try to find out how difficult it really is in my vacation
16:00:15Slasheripreglow: ok, i can try to fix that :)
16:00:24SlasheriI need to add few yields to right place
16:00:59preglowgimme a sec and i'll commit a new one
16:01:03preglowbefore you start working
16:01:12Slasherisure :)
16:02:55Roriso what does the code look for when determining it's the end of a non-lame mp3 file, a db drop off or just a guess based on the running time or what?
16:04:37RoriI just wondering because there is no gap but it does jump too soon to the next track so some silence cutoff is going on
16:04:52stripwax_Rori yes i think it looks for silence
16:05:05Roriit's off by a few 100ms
16:05:27Roriit always jumps too soon rather than too late from what I can tell
16:06:14Roridoes it cut off anything from the beginning of a file or just the ends?
16:07:16preglowmy patch cuts off in the beginning
16:07:23preglowyou really want to do that for gapless playback
16:07:57Rorihmmm. it's a bit harsh. is there a number in the code you can adjust to set how harsh it is?
16:08:28preglowas a matter of fact, we want to skip a fixed factor at the start of the file for all mp3s
16:08:31preglowabout 581 samples
16:08:49Roriwhy fixed?
16:09:05preglowencoder delay
16:09:05preglowit's fixed
16:09:11Roriah ok
16:09:34stripwax_wtf is encoder delay?
16:09:43RoriI realise the glitch may never go away but I am sure it can be tightened alittle
16:10:35preglowstripwax_: no biggie, it just delays the sound by a fixed amount
16:10:42preglowstripwax_: as a result of the encoding process
16:11:24stripwax_preglow ;-) i meant, why is it there? what purpose does it have
16:11:42Rorishove it in and see how it sounds
16:11:44stripwax_preglow - don't need to answer that ; i'll google ;-)
16:11:54Roriooer :)
16:11:59preglowno purpose, i guess, all encoders have some latency, and they just didn't remove it
16:13:30Roriit may just be enough to not have to worry about gapless so much on non-lame encodes
16:14:06RoriMy problem is always with 4/4 beats that skip a beat in mixes ya see
16:14:40RoriI would not even mind a small gap of empty silence as long as the beat stays in time lol
16:16:03Roriheck you could even cheat and tell it to fill the empty gap with so many samples from the previous frame lol
16:16:12preglowthat'd sound ouchy
16:16:29RoriI don't mind hacks as long as they make it sound better :)
16:17:18preglowSlasheri: i guess there's no easy way for me to see if i'm decoding the first frame in a stream?
16:17:36RoriI want to know what Nullsoft's output plugin does to get seamless transitions. Is it purely the silence detection algorithm they use do you think?
16:18:03Slasheripreglow: id3.h has first_frame_offset
16:18:31Rorishame Nullsoft don't have open source :P
16:18:46RoriAOL I hate you guys
16:19:37RoriOK gonna grab your commit
16:19:44preglowit doesn't do squat yet
16:19:49preglowi'm still working on the mp3 codec
16:20:35 Join hicks [0] (~hicks@35562086c255368c.session.tor)
16:20:50RoriOK I will wait some more :)
16:21:19RoriTis very exciting and all ;)
16:22:17*Rori can't wait for his new PSU to arrive
16:22:26Rorias low as 14db noise levels
16:22:39hicksOn the wiki it says you can always hold rec when starting the iriver to load the original firmware. But it also says if the bootload fails you might not be able to restore the firmware. Is there no way to recover from a bad flash? I wouldn't mind trying this out but don't want to totally mess things up :P
16:23:15Rorivery rare to get a bad flash
16:23:22Roriif you patched correctly
16:23:44Rorionce the bootloader is on it's simple to boot back to iRiver
16:23:45hicksIf it did though, is there any way to restore the factory default?
16:23:57preglowa bad flash is a bad flash
16:24:10stripwax_hicks - just don't upgrade fw when running on low batteries :-) bad flash is very unlikely
16:24:14yngwi_awaybut can also happen with the original firmware...
16:24:23preglowit can happen with everything
16:24:24markunYou could send it to LinusN maybe if you have a bad flash
16:24:32ashridahhicks: no, if you nerf the firmware on the device, you're SOL, unless you send it to someone with a BDM (which, given that linus is on vacation, is likely to be impossible atm :) )
16:24:51Roriwiggle it! just a little bit!
16:25:07Rorisorry bad joke
16:25:27hicksok cheers.
16:25:27stripwax_Rori bad song :-)
16:26:15RoriMy friend came round yesterday and let me copy his Galactica 2004 boxset. I am so bad.
16:26:34 Join textchimp [0] (
16:30:05preglowSlasheri: what's the id3.length field based on anyway?
16:31:09Slasheripreglow: i don't know if that's accurate..
16:32:03SlasheriBut now the mp3 gapless seems to work much better. I was almost unable to hear any gap :)
16:32:17preglowgimme a sec, and i'll upload codecmpa.c
16:32:23preglowthat'll put an end to that!
16:36:45Slasherii found that reading one byte from file in mp3data.c is a _really_ bad idea
16:36:56Slasheri*one byte at time
16:37:49Slasheriif i add a yield before each read, that will work but will cause serious decrease in reading performance
16:38:19 Join belgarath [0] (
16:38:54preglowok, commited
16:39:07preglowread one byte at a time?
16:39:15preglowi thought it read everything to a buffer
16:39:25preglowor do you mean the search for the first frame?
16:39:25Slasherifileread function in mp3data.c
16:39:36preglowahh, yes
16:39:37Slasherithat causes the problems
16:39:38preglowthat's not very clever
16:40:02preglowbut yoink, i can't spend any more time on rockbox today, unless something important happens
16:40:05Slasheriwith a few files i have, the searching might take even a few seconds and that will run pcm buffer empty
16:40:38preglowbut ok, lowdown on codecmpa.c:
16:40:50preglowwhen i have a frame count from a vbr header or something
16:40:56preglowi get accurate song length
16:41:06preglowthis figure seems to be accurate, i have cross checked with foobar2000
16:41:19Slasheriah, good
16:41:32preglowso samplecount should be accurate if i have a frame count
16:41:47preglowit should be sample accurate if i also have enc_padding and enc_delay
16:41:57preglowhowever, i just can't squeeze enough data out of libmad
16:42:11preglowi always get cut a few thousand/hundred samples short
16:42:41preglowthat is, not always
16:42:46preglowbut in the test tracks i've used
16:42:48preglowwhich decode perfectly in foobar
16:44:53Slasheriyes, libmad is a bit strange..
16:45:08preglowfoobar does however also use libmad, and seems to work fine with it
16:45:15preglowthat is, it's optional at least
16:45:42Rorihave you thought about asking the guys who work on it over at hydrogenaudio?
16:47:03preglowyes, i but i thought i'd give it a shot myself first
16:47:08RoriI wouldn't worry about it too much. These things you can always come back to at a later date
16:47:15Rorifine tune
16:47:16preglowso i wont have rtfms thrown all around the place
16:47:27preglowand now i must do other things
16:47:33Rorik l8tr
16:47:55preglowSlasheri: i removed your silence detection, btw
16:47:57RoriSlasheri will no doubt be poking around anyhow
16:48:11*Rori will test that
16:48:15Rorigets commit
16:48:33Slasheripreglow: ok, np :)
16:48:59Roriif anyone fancies fixing the FF bug ;)
16:49:00Slasheripreglow: I think that silence detection should be moved somewhere else like separate dsp functions
16:49:07preglowyes, indeed
16:49:15preglowit is generic
16:49:36preglownothing wrong with it, it'll just make lame gapless harder to debug
16:49:55preglowand we should have that working
16:52:38Roriseems better off without it....lets see what the next mix transition is like though
16:53:02Rorineed to play the track all the way through because of the ff prob :P
16:54:13preglowit should be fine
16:54:14preglowwith a little click
16:55:26Rorifinally decided to superglue my rubber nippple onto the joystick since I use it all the time
16:57:17Rorihmm still seems to jump ahead too much
16:58:08preglowdoes the file have a lame header?
16:58:12preglowdo you have foobar installed?
16:58:42textchimpi was wondering about the feasability of being able to make the functions of different buttons in different screens completely configurable - can anyone say if this would be possible?
16:58:48Rorino it's not lame Rockbox cuts either the beginning or end of the track though
16:59:01Roriby half a beat
16:59:04preglowit's highly probable the end
16:59:14preglowit cuts a bit of the start now too, though
16:59:41preglowno, you want that
16:59:46 Quit matsl (Remote closed the connection)
16:59:51Roriyeah there is no gap
17:00:14Roriit just jumps too far ahead. not sure how you would get around that
17:00:34preglowfirst i need to know what kind of info the file has
17:00:46preglowis it a vbr file?
17:00:52RoriI could send you 2 test tracks that mix together
17:01:22Rorithe ones that give me trouble then you could see perhaps
17:01:51RoriI dunno let me see what they are
17:01:52preglownot now, please
17:01:53preglowno time
17:03:17textchimpno one?
17:03:57Rorihow do you tell if it's vbr in winamp?
17:03:57preglowplease do check if they're vbr, though
17:04:09preglowand if you've got foobar, tell me if it has mp3_accurate_length
17:04:21preglowwell, does the bitrate jump around when you play it?
17:04:24Roriok lets try fb2k
17:04:48preglowclick on properties for the file in fb2k
17:04:49RoriI see no bitrate in winamp. used to tell me but now it don't display it
17:05:02preglowit should tell you if it's vbr and if it's got mp3_accurate_length
17:05:27Rorifixed bitrate
17:05:39preglowok, i don't handle those yet
17:05:45preglowdoes it say mp3_accurate_length?
17:05:58Roriloadin foobar
17:06:47Roriwhere do I find that info in foobar?
17:06:56preglowright click
17:06:58preglowpress properties
17:07:03preglowon the file in the playlist
17:07:22Rorinothing about accurate
17:07:23Roribitrate = 192
17:07:23Roricodec = MP3
17:07:23DBUGEnqueued KICK Rori
17:07:23Rorichannels = 2
17:07:23Rorisamplerate = 44100
17:07:23***Alert Mode level 1
17:07:23Rorimp3_stereo_mode = joint stereo
17:07:24***Alert Mode level 2
17:07:24Rori −−−−−−−−−−
17:07:26Rori12931823 samples @ 44100Hz
17:07:28RoriFile size: 7 027 753 bytes
17:07:46preglowthen my newest patch will definitely not do anything for it
17:07:52preglowit doesn't handle cbr yet
17:07:58preglowand you're file isn't gapless
17:08:49Roriwould have to be based on buffering and silence detection similar to nullsoft to work unfortunately
17:09:02Rorilets see how foobar handles the gap btw
17:09:40Roridoes a better job
17:10:02Roriit glitches with a very small gap but it stays nearly in time with the beat
17:10:19Rorithe small gaps put the beat out but not as bad as Rockbox does jumping too far ahead
17:10:42yngwi_awayhow do I know if my file is gapless from the foobar info??
17:10:52Roriyou don't I don't think
17:11:15yngwi_awayi just downloaded it to try
17:11:31preglowyngwi_away: it says mp3_accurate_length
17:11:41RoriNullsoft out is totally seamless
17:11:48yngwi_awaythen its gapless?
17:11:51Rorican't hear the join or any beat delay
17:11:56yngwi_awaythen my files are :-)
17:12:01yngwi_awaysurprise surprise
17:12:17RoriI wish Rockbox could do that trick
17:12:38RoriI am not a virtious person ;)
17:13:05RoriNullsoft send me your source code!
17:13:41Rorijust for the in and out plugs at any rate
17:14:14RoriI should do a wav capture and look at the sinewave
17:14:37preglowwhat would that tell you?
17:14:41preglowyou already know it's gapless
17:14:43Slasheripreglow: I committed some fix for the mp3 metadata reading problem
17:14:47Roridunno. not a lot I guess
17:14:49SlasheriNot a best one but it works
17:14:52preglowSlasheri: excellent
17:15:20Roriwhat does that do? does that fix the ff per chance or is that something else?
17:15:48Roriwoosh! it went past my head. whoosh!
17:16:42preglowit doesn't do anything new atm
17:16:43hicksIs my understanding correct that once the bootloader firmware is flashed to the iriver, it hands control over to rockbox.iriver on load? So if a future update to rockbox.iriver went wrong, it wouldn't really matter, as you can just copy over a new version. Its only the firmware going wrong on the initial flash that could be a problem?
17:17:00stripwax_hicks yep
17:17:01preglowthat is, apart from actually enabling my code to work
17:17:16Roripreglow was you off to do something else? :)
17:17:21preglowi am
17:17:25***Alert Mode OFF
17:17:33preglowi'm multitasking
17:17:37Rorioh ok I thought that meant leaving lol
17:17:45stripwax_me too. i'm watching Kill Bill 2 :-)
17:17:52RoriI am gonna watch some anime
17:18:26RoriTsubasa Chronicle. Anyone familiar with Clamp will know of it.
17:25:48 Join TROGDORRRRR [0] (
17:25:51 Join thegeek [0] (
17:26:31 Join elinenbe1 [0] (~elinenbe_@
17:26:58 Join Bagder_ [0] (
17:27:08 Join silencer [0] (
17:27:32 Quit belgarath (
17:27:32 Quit textchimp (
17:27:32 Quit hicks (
17:27:32 Quit [solid] (
17:27:32 Quit DomZ (
17:27:32 Quit MoosCamaro (
17:27:32 Quit niobos_study (
17:27:32 Quit t0mas (
17:27:32 Quit Harpy (
17:27:32 Quit QT (
17:27:32 Quit RED_M_CHIU (
17:27:32 Quit thegeek_ (
17:27:35 Quit silencer_ (
17:27:35 Quit Bagder (
17:27:35 Quit bill20r3 (
17:27:35 Quit elinenbe (
17:27:35 Quit C-Keen (
17:27:35 Quit Rick (
17:27:39 Join Harpy [0] (
17:27:40 Join niobos_study [0] (~niels@
17:27:40 Join Rick [0] (rick@
17:27:43 Nick elinenbe1 is now known as elinenbe (~elinenbe_@
17:27:43NJoinC-Keen [0] (
17:27:45NJoin[solid] [0] (~solid@
17:27:45 Join t0mas [0] (~Tomas@
17:27:47NJoinQT [0] (as@area51.users.madwifi)
17:28:16NJoinbill20r3 [0] (
17:28:34 Join hicks [0] (
17:37:21***Saving seen data "./dancer.seen"
17:38:06markunShall I commit basic navigation with the remote control in the tree viewer?
17:39:45preglowwhy not
17:40:31markunAnd also implement remote control in the menu?
17:41:27markunDon't know why it's not implemented
17:42:57preglowif it doesn't break anything and works decently, i don't see why not
17:47:45yngwi_awaydoes any of you hav an idea why foobar is not reading the id3tags from my files?? winamp and rockbox for example have no problems..
17:47:58preglowbecause they're id3v2, probably
17:48:11preglowthat plugin might not be bundled by defaylt
17:48:17preglowthe foobar people hate id3v2 with a passion
17:49:13yngwi_awayhmm do they?
17:49:29 Join phil- [0] (
17:51:32yngwi_awayi think i have the solution; i use mp3gain and that seems to cause foobar to ignore id3v2
17:57:22markunpreglow: While I'm at it, shall I move the button defenitions in wps.c to wps.h?
17:57:59markunMaybe I should stop asking questions and just do it..
17:59:32preglowi'm not much of an authority on what to do and what to don't unless it comes to codecs anyway
18:00:28 Quit phil- ()
18:12:17markunok, committed. One more thing, don't you think that pressing 'menu' in the menu should leave the menu?
18:12:17 Quit elinenbe_ (Read error: 104 (Connection reset by peer))
18:14:55 Join bumi [0] (
18:21:20 Join ShockerEngr [0] (
18:22:34 Join webguest55 [0] (
18:22:42bumihey everyone
18:22:52bumii whold like to donate a trout to the rockbox project
18:23:03zemmm trout
18:23:04CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
18:23:04*bumi donates a HUGE trout
18:23:22*bumi slaps ze around with Windows Me
18:23:30bumiwhats up?
18:24:53 Quit ashridah ("Leaving")
18:25:05zewindows me
18:25:06zethats harsh
18:25:34bumithe DDay wepeon
18:27:29[solid]vorbos endurance test update... after 2.5hrs the battery is at 87, might end not bad after all...
18:27:56t0massounds better then I thought it would be
18:28:05[solid]even if it'll tie with flac in terms of time it'll be as good as the original firmware
18:28:27[solid]and it looks to to me that it'll at least tie ^^
18:29:28 Quit crwl (
18:29:56[solid]i'll do all test i can (ie except ac3 since i don't have any dvd's) since the finals are starting and i no longer have to go to the uni everyday... just an exam here and there :p
18:30:01NJoincrwl [0] (
18:30:18zegoogle for ac3s im pretty sure theres some free ones out there
18:30:43[solid]it would be perfect to test that with the same album tho
18:30:53[solid]and that's not really doable i'm afraid
18:31:15zewell you got the album in lossless/cd?
18:31:27zeget faac and make your own ac3 then
18:31:30[solid]do ac3 encoders exist at all?
18:31:38[solid]didn't know that :D
18:31:50zei think faac is whats appropriate
18:31:50zei dunno
18:31:54zebut yeah ac3 encoders exist for sure
18:32:17[solid]i thought faac does aac only
18:32:31zeer maybe
18:32:38zeis aac something diff from ac3?
18:32:49[solid]yes ;)
18:33:02bumiguys, i think i found a bug in rockbox for iriver
18:33:03zecause it seems like faad is what stuff wanted when it wanted to support ac3
18:33:10bumiwhen i change the fonts to hebrew fonts
18:33:13bumiall works well
18:33:16zeeither way, ac3 encoders do exist :p
18:33:23bumiand then when i restart the player
18:33:30bumiit goas back to default
18:33:33[solid]ac3 basically equals dvd-v audio and aac basically is mpeg4 audio (very basically)
18:33:56zeer mpeg4 audio? hmm
18:34:03zei thought mpeg4 was a video codec
18:34:11[solid]mpeg4 is a container
18:34:17zeavi is a container
18:34:20bumican anyone test what i said and see thats its realy a bug
18:34:22[solid]that too
18:34:27bumiand not just something am doing wrong
18:34:32zedivx is mpeg4
18:34:43zeand they're both in avi's
18:34:56[solid]apple sells aac in .mpa
18:34:57zeand often paired with ac3 or mp[23] or vorbis audio
18:35:06[solid]which basically is .mp4 with a bad extension
18:35:47 Join Stryke` [0] (
18:35:48[solid]i'm back to tha books anyway now :)
18:40:16 Nick niobos_study is now known as niobos (~niels@
18:41:30niobosI need some help in C-coding: how do I iterate over a directory (recursively)?
18:41:37niobosI checked the CVS and found this:
18:41:47niobosif (entry->attribute & ATTR_DIRECTORY) {
18:42:12nioboshowever, gcc complains that "structure has no member named `attribute'"
18:42:32nioboshow should I check if an entry is a directory?
18:43:27niobosit's for songdb, so it should run on both: in rockbox and on a regular PC
18:45:57Rorimarkun yes pressing menu button again should exit menu
18:46:10Roricoz it pee's me off keep having to hit back
18:46:58markunyes, me too
18:47:25yngwi_awayshould hitting the "play" button always take you back to the wps screen? or only when in the filetree?
18:48:55 Quit bumi ("( :: NoNameScript 3.81 :: )")
18:48:58RoriI'm confused as to why hitting play/pause does not play the currently selected file
18:49:17Roriwould make more sense to me
18:49:51markunHm, I'm now wondring how you should get to the filetree from the wps using the remote. There is a menu button that I now use to enter the menu.. but also a mode button.
18:50:04Roributton assignments need to be sorted out or user configureable :)
18:50:50yngwi_awayrori: the pause function has been temporarily disabled today
18:50:54yngwi_awayor yesterday
18:51:03Roriyeah I know that
18:51:17yngwi_awayoh, sorry, a missunderstanding
18:51:23Roribut hitting the play button on a file or folder should play it
18:51:25yngwi_awayshould read posts better
18:51:38t0masmarkun: I guess we need to make all buttons changeable... I have those wps buttons mostly just copied from the simple archos remote code...
18:51:49t0masmaybe someone needs to get a right interface doc sometime
18:51:58Roriactually hitting play on a folder and having it play the contents sounds good to me
18:52:00t0masiriver buttons are named wrong too... in some places
18:52:09markunRori: how would you then go to back to the wps?
18:52:20t0maswould be a nice choise
18:52:23Roriit just shows the first file played
18:52:35Rorilike it normally does
18:53:46Roriwhy bother with album playlists when you can just select the album folder and hit the play button? :)
18:54:05Rorias long as they are in sequential order no problem
18:54:21yngwi_awayyou can actually make a playlist with the currently selected folder and subfolders :-)
18:54:32RoriI know
18:54:36Roribut there is no need
18:54:39yngwi_awaysorry again
18:54:44Rorijust play contents of folder
18:54:55yngwi_awayyeah i would like that
18:55:09yngwi_awaywould be more comfortable
18:55:11Rorithe folder becomes the album playlist because playing tracks in a folder plays in sequence anyhow
18:55:35Rorirather than digging down to the files in the folder and going to the first track and hitting the joystick
18:55:39Roritoo much effort ;)
18:55:59Rorifolders become a kind of easy album database that way
18:56:06markunt0mas: I think the 'mode' button on the remote would be a good choice to switch between wps and files
18:56:40yngwi_awaywhy not the joystick?
18:56:45Rorimarkun good idea
18:56:52t0masmarkun: ok, I can add that
18:56:54Roribecause joy you can sometimes hit wrong ;)
18:57:04niobosIs there a portable way to recursively iterate over a directory in C?
18:57:06t0masbut file browsing should be possible on the remote too
18:57:09RoriI dislike center buttong on joystick to be honest
18:57:13niobosportable = linux, windows and rockbox
18:57:31t0masniobos: what do you want to iterate? filenames?
18:57:42markunt0mas: I added WPS_RC_MENU here, but somehow it doesn't work.
18:57:48niobositerate over filenames and descend into directories
18:57:54Rorileft/right/up/down is fine but push in sometimes you don't push dead center and bosh you do something you did not intend
18:57:59t0masmarkun: I can fix that
18:58:04t0maswhat button did you use?
18:58:32markunt0mas: did you see my latest commit?
18:58:34Roriplay button on folder select. must have function IMO :)
18:58:44markunI moved the button defines from wps.c to wps.h
18:59:01t0masmarkun: jep
18:59:10t0masmarkun: I saw it... why?
18:59:18yngwi_awayhmm i'd really like a Buttons.cfg file were I can assign any function i want *dreams*
18:59:31 Nick yngwi_away is now known as yngwi (
18:59:42markunt0mas: because it's also done in menu.h and tree.h
19:00:02t0mashm.. it has been in wps.c for a long time
19:00:29t0masbut it sounds logic to mee
19:01:48RoriI added the option that when hitting play on a folder if there is a playlist in the folder it overrides sequential playback to play the playlist instead
19:01:59RoriI mean the suggested option ;)
19:02:20Rorisounds like a logical approach to me
19:02:22yngwihehe good idea
19:02:33markunRori: What if there are multiple playlists?
19:02:40Roriit plays those in sequence
19:02:49yngwijust enter the directory and let choose
19:02:50Roriif in the same folder
19:02:58t0masmarkun: you defined WPS_RC_MENU in wps.h
19:03:01Roriyngwi this is a shortcut
19:03:03t0masand added it to wps.c?
19:03:07t0masor just that define?
19:03:17Roriplay on folder rtaher than having to enter folder every time and select files
19:03:21markunt0mas: yes, that's what I did
19:03:26Rori1 step instead of 2
19:03:36yngwiyeah, i mean wehen there are more than one playlists then enter the folder
19:03:43markunt0mas: Just like I did in tree.h and tree.c
19:03:56t0masmarkun: ok, I'll take a look at it
19:03:59Roriwhatever. I just want play folder that will make me very happy
19:04:13markunt0mas: But I didn't commit it because it didn't work
19:06:07t0masyou forgot WPS_MENU_PRE I guess?
19:06:25Rorianyone working on fixing the FF issue?
19:06:56t0masSlasheri does that...
19:07:32RoriFF is not accurate at the moment. It thinks it's the end of a song when it isn't. It's jumping too far forward even though it reads the correct time
19:08:14Rorior rather it's not moving far enough but thinks it's further forward and ends the track prematurely
19:08:18markunt0mas: this is what I did
19:08:51t0masI'll fix it
19:08:51t0masand commit
19:09:02Roriit goes "Oh only 10 seconds left to the end of the song" and ends at 10 seconds but the song is still 20 seconds longer
19:09:11Rorirough exapmple
19:13:28markunt0mas: Hmm, BUTTON_RC_MODE is not defined..
19:13:51t0masthat's not the only problem
19:14:04t0masI have to eat now... then I'll commit it
19:18:41 Join MoosCamaro [0] (
19:22:11 Quit webguest55 ("CGI:IRC (EOF)")
19:23:44 Join Aison [0] (
19:29:12 Quit Rori ()
19:37:24***Saving seen data "./dancer.seen"
19:38:56[solid]this is weird... i reached for my player to check the runtime and found that... it's off
19:39:14[solid]the vorbis test is continued but that is very very weird
19:39:25[solid]if it crashed it wouldn't turn off, right?
19:39:51[solid]wouldn't have turned*
19:41:56[solid]luckily the runtime is saved, hope it's accurate
19:42:20crwlit doesn't seem to be
19:44:20preglowi think someone else mentioned this happening yesterday
19:45:04t0masmaybe it's the off timer?
19:45:13t0maswhen it's stopped it times down 10 minutes
19:45:33t0masso when OGG playback fails to load next codec (it does something like that sometimes) and it's stopped
19:45:40t0masit waits for 10 minutes, and then shuts down
19:49:13[solid]that explains the shutdown
19:49:31[solid]so there must be something causing it to stop... hm
19:49:45[solid]it's the same files over and over again, all oggs
19:50:19[solid]i know it'd be better to do this with headphones on but i'd go crazy listening to the same album over and over again :D
20:02:46t0masmarkun? do you know where the filetree is drawn? (I don't see much drawing code in filetree.c?)
20:06:42 Nick StrathAFK is now known as Strath (
20:17:08markunt0mas: no idea
20:17:30markun[solid]: I also had vorbis playback turning my iriver off, even today while I was listening.
20:18:02Slasherimarkun: what kind of songs you was playing when that happened?
20:18:05 Quit silencer (Nick collision from services.)
20:18:10Slasherionly oggs or other formats also?
20:18:12markunt0mas: Are you looking for tree.c?
20:18:24markunSlasheri: I only have q3 oggs
20:18:28 Join silencer_ [0] (
20:18:33markunSlasheri: and no mp3
20:19:12markunI turned the idle timer off the last time it happened, but obviously that didn't help.
20:19:41SlasheriHmm.. How long the songs were?
20:19:42preglowpretty strange behaviour
20:20:26markunNot very long, can't remember which album it was. I've since then listened to a lot of music without it turning off.
20:21:05markunI think amiconn reported the same thing
20:21:08Slasheriok, but any of them weren't over 10 minutes or something like that?
20:21:14preglowmarkun: with mp3s then, i guess?
20:21:49markun10.55.28 # <amiconn> I started playing a dir of oggs, and while playing the 3rd track, my iriver suddenly powered off (no battery problem)
20:22:35SlasheriHmm, yep. Can anybody reproduce the problem with same set of songs?
20:22:50markunSlasheri: I'm not sure, but I've idle poweroff turned off anyway..
20:23:15markunmaybe amiconn knows which songs he was playing
20:23:16Slasherimarkun: yes, i think the problem is caused somewhere else.. Some buffer overflow or something like that
20:23:55markunSlasheri: Maybe amiconn's new overflow detection? Or wasn't that committed yet?
20:24:21SlasheriI think it's committed now, but it can detect only stack overflows
20:25:36markunt0mas: I tried to make BUTTON_RC_VOL switch between wps and tree, but it doesn't do anything.
20:26:51markunHm, probably the _PRE problem..
20:26:58t0masyou can switch with BUTTON_RC_ON
20:27:56t0masthat would be more like the real device does it?
20:28:01t0masand yes... there's a WPS_PRE too
20:28:08t0masjust do like I did it in wps.c
20:28:08markunwhich button is RC_ON?
20:28:14t0masthe play button
20:28:24markunplay trigers pause here..
20:28:42markunwhen in wps
20:28:46markunhi HCl!
20:28:57HClhi :)
20:29:02 Join MrStaticVoid [0] (
20:33:09 Join DomZ [0] (
20:34:59t0masmarkun: I mean play from tree should return to wps
20:38:11preglowany news on grayscale?
20:39:28markunt0mas: The how do you play a new file from the tree?
20:39:40 Quit stripwax_ (Read error: 131 (Connection reset by peer))
20:39:57markunpreglow: no, not yet. I've made it a lot faster, but it's not finished.
20:43:52markunI would like the 'mode' button on the remote (called BUTTON_RC_VOL) to switch between tree and wps.
20:45:22preglowa lot faster = good
20:49:16markunamiconn told me it was more or less what he had in mind. Maybe he should finish it because he wants to change the graphics api anyway.
20:58:17 Join Chamois [0] (
21:00:14 Quit Chamois (Client Quit)
21:08:13 Quit DomZ ("CGI:IRC (Ping timeout)")
21:10:46 Join matsl [0] (
21:15:04[solid]markun: you play it with 'next' on the main unit, and since it's pressing the joystick right, it makes sense and is quite a good idea from my experience
21:17:31 Nick Naked is now known as Hadaka (
21:19:55markun[solid]: Then how do you get to the file-menu from the remote control?
21:20:13markunright now it's not possible
21:20:46markunI thought keeping 'play' on the remote pressed could give you the menu.
21:22:38[solid]with the original fw it was with navi/menu
21:23:09[solid]not a bad concept, since you open the menu with the a-b/mode button
21:23:14[solid](on the main unit)
21:26:26[solid]vorbis test update - after 5 hours the battery is at 65 :D
21:35:02[solid](same as with the flac test - remote plugged in, headphones plugged into line-out, volume at 50%)
21:35:51preglowthat's cheating, really
21:36:01preglowthe headphones need to be in the proper output to loud the amplifier
21:36:42[solid]i don't understand how could that be, the perceivable volume is exactly the same
21:37:19[solid]and that's the way i always use it, since i carry my h120 in a bag and this way it's far more comfortable for me..
21:37:28***Saving seen data "./dancer.seen"
21:40:13[solid]but i'm open to argumentation, if a luser and a dev have contradictory opinions usually the dev is right lol
21:44:13preglowdepends on the earphones, i guess
21:44:14 Quit gromit` (Read error: 54 (Connection reset by peer))
21:44:23preglowthe line out isn't by far capable of driving my earphones
21:45:11[solid]well 50% isn't too loud either, it's good for listening at home and not outside
21:46:16 Join gromit` [0] (~gromit`
21:51:13markunpreglow: Is it very expensive to scale evey sample in Tremor for clipping prevention, maybe replaygain?
21:52:05[solid]\o/ for replaygain
21:52:28crwlimplement the replaygain with some volume trick ;)
21:53:11markunYes, but we still have to scale the samples if we don't want clipping
21:53:15markunThe replaygain can be done with the volume control
22:00:13preglowmarkun: no
22:00:19preglowmarkun: not very, you need to do a mul on each sample
22:02:09markunmaybe the peak value computed with libvorbis differs from the peak you get with Tremor..
22:02:22preglownah, not by much
22:02:47preglowan integer decoder is in some cases more exact than a floating point decoder
22:02:52preglowi think they're pretty equal
22:02:55preglowdoing a test shouldn't be hard
22:04:07 Join ^koniu^ [0] (
22:08:49 Join bipak [0] (
22:10:36markunI made the remote code a bit more complete and would like to commit it. The actual button assignmet can be changed later easily.
22:11:21 Join einhirn [0] (
22:13:14 Join F1^Aison [0] (
22:14:19preglowi wanna do the resample code now
22:14:59markunBut you don't have time right>
22:15:39preglowif i do that, i wont get anything else done today
22:17:04preglowmy god, how i look forward to finishing this school nonsense
22:19:37markunYes, me too..
22:20:33markunAnd then we'll get proper jobs and be too tired at night to write programs for fun..
22:21:52preglowof course
22:22:10preglowoptimally money would just appear in my mailbox every day, so i could spend all my time on hobbies
22:22:14preglowbut anywho
22:22:17preglowi'd rather try working than school now
22:22:22preglowsince i'm so incredibly tired of it
22:22:49markunI've been studying now for almost 6 years..
22:23:11markunAnd all the years of school before it of course..
22:23:30preglowfive years here
22:23:34preglowplus the school before that, yes
22:23:41preglowwhich amounts to thirteen years
22:23:49markunYes, for me too
22:23:59preglowbut yeah
22:24:02preglowi'm done in two weeks
22:24:11preglowthen relocation and work is next
22:24:27markunwhere will you move?
22:25:36 Quit bipak_ (Read error: 110 (Connection timed out))
22:27:59preglowthat is, in august
22:28:08preglowdunno what i'll do in july
22:28:14preglowprobably stay at my father's
22:28:23preglowwith a 28k8 modem, expect heavy work to be done!
22:29:28 Quit Aison (Connection refused)
22:42:20 Nick F1^Aison is now known as Aison (
22:50:38 Join Querty [0] (
22:51:35Quertyhmm, seems the latest commit to tree.c has broken CVS builds
22:51:37Querty make[1]: *** [build-recorderv2/apps/tree.o] Error 1
22:52:17preglowa ton of reds
22:52:23preglowmarkun: feex
22:52:28*t0mas is innocent :P
22:53:17[solid]t0mas: any news on remote wps?
22:54:10[solid]vorbis test update: 6.5hrs play time, battery level: 62
22:54:18preglowstrange error
22:54:51preglowstraight in the middle of ifdef hell
22:57:16markunI'll fix it right now
22:58:22t0mas[solid]: no, I have to finish a lot of reading this week
22:58:39t0masbut friday my vacation starts
22:58:49t0masso then I can start coding whole nights :D
22:59:09 Join david88 [0] (
23:00:00 Part david88
23:00:58 Join david88 [0] (
23:01:58markunHm, don't know what the best way to fix this would be.. RC_ENTER is defined the same as RC_RUN, but not on iriver
23:05:19markunI could disable RC_ENTER on all the devices that are red now, but that would probably make other people angry
23:06:13markunHm, no, it shouldn't matter
23:06:18 Quit MrStaticVoid ("leaving")
23:08:28 Join thegeek_ [0] (
23:08:28 Quit thegeek (Read error: 54 (Connection reset by peer))
23:10:46 Join webguest23 [0] (
23:12:15webguest23Is HCL around?
23:13:37webguest23People great work I have been following your development and now have Rockbox playing my collection of a couple thousand oggs
23:14:15webguest23HCL is there an easy fix to the issue with the songdb? issue?
23:15:23 Nick Bagder_ is now known as Bagder (
23:17:13webguest23In the wiki page on creating the songdb there are some perl scripts to run but one generates an error because of the HCL has noted it on the wiki page
23:17:38webguest23page is
23:18:00preglowstrict is a file?
23:18:07preglowi thought it was a compiler hint
23:18:10Bagderit is
23:18:40Bagderwebguest23: so remove the use strict line
23:18:41 Quit Aison ("( :: NoNameScript 3.72 :: )")
23:18:48Bagderand what makes HCl claim this?
23:19:07Bagderit works fine for me
23:19:29Bagderbut I don't use windows of course
23:19:45preglowsounds really strange
23:20:21 Quit matsl (Remote closed the connection)
23:20:52BagderI removed that line now
23:20:52webguest23Okay I will look at the code and see what is happening, but I am a real noob
23:21:11Bagderthere might be a problem with it, but it is _not_ a missing
23:22:03webguest23I am amazed at the speed that things have been going Great work
23:23:09markunThe red builds are gone :)
23:24:25webguest23Can't locate in @INC (@INC contains: .) at line 12.
23:24:25webguest23BEGIN failed−−compilation aborted at line 12.
23:24:25webguest23Compilation failed in require at line 10.
23:24:25DBUGEnqueued KICK webguest23
23:24:25webguest23BEGIN failed−−compilation aborted at line 10.
23:24:46webguest23Above is the error generated
23:24:48 Join Rori [0] (
23:25:06Bagderand if you remove the 'use strict' line?
23:26:56webguest23Now it is cannot locate in @INC ... at line 13
23:27:43Bagderso install perl properly, then try again
23:27:46[solid]webguest23: your perl seems fux0red
23:27:58BagderI guess that sneak way suggested is bad
23:28:10^koniu^[solid]: czytasz sredniowiecze, aye? =)
23:28:17webguest23Okaay thanks I will do that and see what happens.
23:28:39webguest23It is good to see the buttons on the Iriver remote working
23:28:52[solid]^koniu^: aye
23:29:04webguest23Is there going to be a wps for the remote as well to choose your own config?
23:29:09^koniu^(ops, everybody let me apologise for not using the common language of this channel. this won't happen again)
23:29:54[solid]erm... after 7 hours of playing q6 vorbis the battery indicator is still >50%
23:30:28markun[solid]: that's incredible!
23:30:36[solid]you guys were really too pesimistic about your code :)
23:31:18[solid]wonder if this time it'll drain *before* dawn
23:33:41[solid]maybe the volume is too low, my fones are extremely low-power driven, or using the lin-out is a dirty cheat indeed as preglow mentioned...
23:33:43Bagderwebguest23: according to plan, yes
23:33:47[solid]but i really don't think so xD
23:34:08nioboswebguest23: I'm a bit late... but anyway: I'm (re)writing the songDB-program in C
23:34:55 Quit david88 ()
23:35:02webguest23Excellent news niobos thanks
23:35:14niobosso sooner or later the (old) perl script will become outdated
23:35:35Bagderits not that old
23:35:39niobosthe DB-writing is done, the tag-reading still needs to be done
23:35:43webguest23Bagder that is great, I expect to have some time in July, but at the rate you all are going things will be finished
23:35:56niobosBagder: no, but it has bugs that don't get fixed
23:36:10Bagderniobos: really? I didn't know
23:36:15Bagderand besides
23:36:16[solid]niobos: and what about the java version?
23:36:35BagderI don't see how rewrites fix such problems
23:36:44niobosI did some tests and for example the <no artist tag> doesn't work... it's just empty etc
23:36:52webguest23So long keep up the good work
23:36:55niobosthe Java version is fine, but won't work on the device itself
23:37:02niobosHCl is working on the Java
23:37:04[solid]niobos: right, good point
23:37:04BagderI'll never use the java version
23:37:12niobosand Java is horribly slow
23:37:31***Saving seen data "./dancer.seen"
23:37:34Bagderanyway, a C version will be handy
23:37:42niobosBagder: you're right rewrites won't fix those problems, but C will run on the device
23:37:53Bagderbut veeeeeery slow
23:38:12 Quit ^koniu^ ("leaving")
23:38:19Bagderand you need to do major magic to deal with large databases
23:38:26[solid]Bagder: it's better to start with C and them optimise using asm, no? :D
23:38:34niobosWell, I won't excpext my Iriver to built a complete DB in a few seconds... it's ok if it takes a minute or two...
23:38:41Bagderits not the language, its the cpu/disk
23:39:47[solid]vorbiscomments might be very slow indeed
23:39:48 Quit webguest23 ("CGI:IRC (EOF)")
23:40:24 Quit hicks (Remote closed the connection)
23:40:31 Join tvelocity [0] (
23:40:46Bagdermy 6000 takes minutes to index on my host
23:40:48niobosdoes anyone know of a portable (=linux, windows, rockbox) way to get all files from a directory, and recursively descend into subdirs?
23:40:49Bagder6000 songs
23:41:06Bagderthere is none
23:41:11Bagderunless you count cygwin
23:41:32niobosI don't think we can force window-users to use cygwin to make their DB?
23:41:35 Join hicks [0] (~hicks@215d868496cb3a16.session.tor)
23:41:45preglow[solid]: what's the status?
23:41:46Bagderno, they won't like that...
23:42:03nioboswhere can I find/use/try the rockbox-way?
23:42:23Bagderrockbox uses posix opendir/readdir style
23:42:27niobossince I tried some things, but GCC complains that some attributes don't exist
23:42:27[solid]preglow: 7.5hrs, battery at 50
23:42:46nioboseg: to find out if an entry is a directory
23:43:08niobosthe CVS code uses entry->attributes & DIRECTORY (or something similar)
23:43:23niobosbut on my linux system there is no ->attributes
23:43:34Bagderyes, that is not portable
23:43:35preglow[solid]: niceness
23:43:44[solid]preglow: loveliness ;)
23:43:52niobosok, long live the #ifdef's, I guess?
23:44:05Bagderniobos: indeed, that's what portable programming is all about ;-)
23:44:13niobosk, thx anyway...
23:44:28niobosI'll concentrate to write linux-code, and #ifdef the rockbox/windows in later
23:44:53niobosthe "special" things about rockbox, are they in the wiki?
23:44:59niobos(like that readdir stuff)
23:45:15preglowwhy didn't you make it portable?
23:45:15niobosjust ask here?
23:45:30Bagderniobos: yes, we have almost no internals documented
23:45:39Bagderpreglow: I don't remember ;-)
23:45:51Bagderposix doesn't define that anyway
23:45:53niobosBagder: hmm... no internals documented... how recognizable...
23:46:29Bagderso rockbox _is_ posix compliant, just not doing it like linux
23:46:36Bagder(in regard to those functions that is)
23:46:50nioboswell, Windows XP is also (kind of) POSIX compilant... onlt totaly different
23:46:59Bagderit is _not_
23:47:32nioboshmm, could be... I tought I read that somewhere
23:47:46Bagderit is only similar to posix occasionally
23:47:51CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
23:47:51*niobos is getting tired and is off to bed
23:47:59niobosBagder: thx for the info
23:48:20Bagderniobos: good luck with your work
23:48:33niobosYou'll see me around if I have questions...
23:48:40niobosa lot :-D
23:48:56 Quit niobos ("off to bed")
23:49:21]RowaN[woo toshiba gigabeat coming to the uk ( i almost bought one from japan instead of my iriver
23:50:26 Quit Querty ("Leaving")
23:50:52markunWhat kind of HD is in it? The same as in the H140?
23:51:40Bagder"Approximately 264,000 colours"
23:51:51Bagder2^18 is 262144
23:52:03]RowaN[no idea, i just wanted one coz it looks pretty
23:52:27]RowaN[kinda cool that its screen is vcd resolution
23:55:05preglowmore than enough to play doom in its original resolution, if you tilt the player :P
23:55:38Bagderdoesn't play ogg
23:55:46preglowindeed it doesn't
23:56:04preglowi hope my h120 will survive for aeons
23:56:14preglowi'll be seriously depressed if i have to get a player with no rockbox support
23:56:24preglowi'll then start porting
23:58:30Bagderthere's a 60GB version too
23:58:53preglowi want an arm based player, mommy

Previous day | Next day