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 2007-10-13

00:00:14LloreanOnce USB code is finished charging should be at least comparable in speed to the original firmware, if still a little bit slower.
00:00:34 Join stevenm [0] (
00:04:03 Quit hcs (Read error: 113 (No route to host))
00:05:47 Quit davina (Remote closed the connection)
00:07:01*keanu waits for netsplit
00:10:03 Quit kugel ("ChatZilla [Firefox]")
00:10:15 Join matsl_ [0] (
00:12:13 Quit matsl (Read error: 110 (Connection timed out))
00:15:29n1spixelma: I have an almost finished "Theme Settings" chapter now but i will re-read it in the morning before I commit :-)
00:20:05 Quit amiconn (
00:20:05 Quit w0rd54 (
00:20:05 Quit linuxstb_ (
00:20:05 Quit karashata (
00:20:05 Quit _pill (
00:20:05 Quit ze (
00:20:05 Quit Thundercloud (
00:20:05 Quit keanu (
00:20:05 Quit qweru (
00:20:05 Quit iamben (
00:20:05 Quit tchan (
00:20:05 Quit Chronon (
00:20:05 Quit mozuch (
00:20:05 Quit Toki (
00:20:05 Quit jmspeex (
00:20:05 Quit Zom (
00:20:05 Quit DataGhost (
00:20:05 Quit Presence (
00:20:05 Quit Galois (
00:20:05 Quit crashd (
00:20:05 Quit hannesd (
00:20:05 Quit Toxicity999 (
00:20:05 Quit toffe82 (
00:20:05 Quit Gnu47 (
00:20:05 Quit idnar (
00:20:05 Quit bluebrother (
00:20:05 Quit Variable (
00:20:05 Quit krazykit (
00:20:05 Quit Soap (
00:20:05 Quit sslashes (
00:20:05 Quit annulus (
00:20:05 Quit Nimdae (
00:20:05 Quit jepler (
00:20:05 Quit marcosource (
00:20:05 Quit dionoea (
00:20:05 Quit lostlogic (
00:20:05 Quit matsl_ (
00:20:05 Quit BigBambi (
00:20:05 Quit Arathis (
00:20:05 Quit Pyromancer (
00:20:05 Quit tedrock (
00:20:05 Quit shodanX (
00:20:05 Quit GodEater_ (
00:20:05 Quit GodEater (
00:20:05 Quit newbyx86 (
00:20:05 Quit tumu (
00:20:05 Quit chandlerc (
00:20:05 Quit Bjoern-Erik (
00:20:05 Quit signuts (
00:20:05 Quit rjg (
00:20:05 Quit bb (
00:20:05 Quit J (
00:20:05 Quit solatis (
00:20:05 Quit bumper (
00:20:05 Quit lee-qid (
00:20:05 Quit miepchen^schlaf (
00:20:05 Quit Zagor (
00:20:05 Quit [AFX] (
00:20:05 Quit advcomp2019 (
00:20:15 Quit zicho (
00:20:15 Quit RaRe` (
00:20:15 Quit ivan` (
00:20:15 Quit linuxstb (
00:20:15 Quit ATravelingGeek (
00:20:15 Quit Solskogen (
00:20:15 Quit gtkspert (
00:20:15 Quit Bagder (
00:20:15 Quit rux (
00:20:15 Quit lids (
00:20:15 Quit BHSPitMonkey (
00:20:15 Quit webmind (
00:20:15 Quit scorche (
00:20:15 Quit Kohlrabi (
00:20:15 Quit blithe (
00:20:15 Quit DiDjCodt (
00:20:15 Quit Weiss (
00:20:15 Quit stevenm (
00:20:15 Quit bertrik_ (
00:20:15 Quit Robin0800 (
00:20:15 Quit atsea-34 (
00:20:15 Quit spky (
00:20:15 Quit alienbiker99 (
00:20:15 Quit Llorean (
00:20:15 Quit animeloe (
00:20:15 Quit Bitter (
00:20:15 Quit troxor (
00:20:15 Quit Xerion (
00:20:15 Quit in-jane (
00:20:15 Quit luckz (
00:20:15 Quit maxkelley (
00:20:15 Quit pixelma (
00:20:15 Quit n1s (
00:20:15 Quit Rick (
00:20:15 Quit joshin (
00:20:15 Quit midkay (
00:20:15 Quit BrianHV (
00:20:15 Quit Ave (
00:20:15 Quit lymeca (
00:20:15 Quit andrewg877 (
00:20:15 Quit rvvs89 (
00:20:15 Quit bagawk (
00:20:15 Quit pabs (
00:20:15 Quit PaulJam (
00:20:15 Quit ender` (
00:20:15 Quit n17ikh (
00:20:15 Quit kclaf (
00:20:15 Quit feisar (
00:20:15 Quit Tanuva (
00:20:15 Quit guyzmo (
00:20:15 Quit [omni] (
00:20:44NJoincrashd [0] (
00:20:44NJoinGalois [0] (
00:20:44NJoinPresence [0] (
00:20:44NJoinZom [0] (
00:20:44NJoinDataGhost [0] (
00:20:44NJoinjmspeex [0] (
00:20:44NJoinToki [0] (
00:20:44NJoinmozuch [0] (n=mozuch@
00:20:44NJoinChronon [0] (
00:20:44NJointchan [0] (n=tchan@lunar-linux/developer/tchan)
00:20:44NJoiniamben [0] (
00:20:44NJoinqweru [0] (
00:20:44NJoinkeanu [0] (
00:20:44NJoinThundercloud [0] (
00:20:44NJoinze [0] (
00:20:44NJoin_pill [0] (
00:20:44NJoinkarashata [0] (
00:20:44 Join linuxstb_ [0] (n=chatzill@rockbox/developer/linuxstb)
00:20:44NJoinw0rd54 [0] (
00:20:44NJoinamiconn [0] (n=jens@rockbox/developer/amiconn)
00:20:44NJoinBjoern-Erik [0] (
00:20:44NJoinchandlerc [0] (n=chandler@
00:20:44NJointumu [0] (
00:20:44NJoinnewbyx86 [0] (
00:20:44NJoinsignuts [0] (n=sig@
00:20:44NJoinrjg [0] (
00:20:44NJoinGodEater [0] (n=bryan@rockbox/staff/GodEater)
00:20:44NJoinGodEater_ [0] (n=bryan@rockbox/staff/GodEater)
00:20:44NJoinshodanX [0] (
00:20:44NJointedrock [0] (
00:20:44NJoinPyromancer [0] (
00:20:44NJoinArathis [0] (
00:20:44NJoinBigBambi [0] (n=alex@rockbox/staff/BigBambi)
00:20:44NJoinmatsl_ [0] (
00:20:44NJoinbumper [0] (
00:20:44NJoinsolatis [0] (
00:20:44NJoinJ [0] (
00:20:44NJoinbb [0] (n=bb@unaffiliated/bb)
00:20:44NJoindionoea [0] (
00:20:44NJoinlostlogic [0] (n=lostlogi@rockbox/developer/lostlogic)
00:20:44NJoinmarcosource [0] (
00:20:44NJoinjepler [0] (n=jepler@emc/developer/jepler)
00:20:44NJoinNimdae [0] (
00:20:44NJoinannulus [0] (
00:20:44NJoinsslashes [0] (n=rmf@
00:20:44NJoinSoap [0] (n=Soap@rockbox/staff/soap)
00:20:44NJoinkrazykit [0] (
00:20:44NJoinVariable [0] (
00:20:44NJoinbluebrother [0] (i=ppg1HDBX@rockbox/staff/bluebrother)
00:20:44NJoinidnar [0] (i=mithrand@unaffiliated/idnar)
00:20:44NJoinGnu47 [0] (i=Gnu47@
00:20:44NJointoffe82 [0] (
00:20:44NJoinToxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
00:20:44NJoinhannesd [0] (
00:20:44NJoinKohlrabi [0] (
00:20:44NJoinDiDjCodt [0] (
00:20:44NJoinblithe [0] (
00:20:44NJoinWeiss [0] (
00:20:44NJoinscorche [0] (i=Blah@rockbox/administrator/scorche)
00:20:44NJoinwebmind [0] (
00:20:44NJoinBHSPitMonkey [0] (n=stephen@unaffiliated/bhspitmonkey)
00:20:44NJoinlids [0] (
00:20:44NJoinrux [0] (
00:20:44 Join Bagder [0] (n=daniel@rockbox/developer/bagder)
00:20:44NJoingtkspert [0] (
00:20:44NJoinSolskogen [0] (
00:20:44NJoinATravelingGeek [0] (n=ATG@pdpc/supporter/student/ATravelingGeek)
00:20:44NJoinlinuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
00:20:44 Join ivan` [0] (n=ivan`@unaffiliated/ivan/x-000001)
00:20:44NJoinRaRe` [0] (
00:20:44NJoinzicho [0] (
00:20:44NJoinadvcomp2019 [0] (n=advcomp2@
00:20:44NJoin[AFX] [0] (
00:20:44 Join Zagor [0] (n=bjst@rockbox/developer/Zagor)
00:20:44NJoinmiepchen^schlaf [0] (
00:20:44NJoinlee-qid [0] (
00:20:44NJoinmaxkelley [0] (
00:20:44NJoinluckz [0] (
00:20:44NJoinin-jane [0] (
00:20:44NJoinXerion [0] (
00:20:44NJointroxor [0] (n=boyo@unaffiliated/troxor)
00:20:44NJoinBitter [0] (n=Nocebo@unaffiliated/helldragon)
00:20:44NJoinanimeloe [0] (n=animeloe@unaffiliated/animeloe)
00:20:44 Join Llorean [0] (n=llorean@rockbox/administrator/Llorean)
00:20:44NJoinalienbiker99 [0] (
00:20:44NJoinspky [0] (
00:20:44NJoinatsea-34 [0] (i=atsea-@gateway/tor/x-cdd6ad3bc300af5b)
00:20:45NJoinRobin0800 [0] (
00:20:45NJoinbertrik_ [0] (
00:20:45NJoinstevenm [0] (
00:20:45NJoinpabs [0] (
00:20:45NJoinbagawk [0] (n=lee@unaffiliated/bagawk)
00:20:45 Join rvvs89 [0] (n=rvvs89@unaffiliated/rvvs89)
00:20:45NJoinandrewg877 [0] (
00:20:45NJoinlymeca [0] (i=lymeca@unaffiliated/lymeca)
00:20:45NJoinAve [0] (
00:20:45NJoinBrianHV [0] (
00:20:45NJoinmidkay [0] (n=midkay@rockbox/developer/midkay)
00:20:45 Join joshin [0] (n=joshin@unaffiliated/joshin)
00:20:45 Join Rick [0] (i=rick@unaffiliated/rick)
00:20:45NJoinn1s [0] (
00:20:45NJoinpixelma [0] (i=pixelma@rockbox/staff/pixelma)
00:20:45NJoinguyzmo [0] (
00:20:45NJoin[omni] [0] (
00:20:45NJoinTanuva [0] (n=tanuva@
00:20:45NJoinfeisar [0] (
00:20:45NJoinkclaf [0] (i=kclaf@
00:20:45NJoinn17ikh [0] (
00:20:45NJoinender` [0] (
00:20:45NJoinPaulJam [0] (
00:20:49***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! ( )'
00:20:52***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! ( )'
00:20:55***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! ( )'
00:20:58***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! ( )'
00:21:01***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! ( )'
00:21:27 Quit Bjoern-Erik (SendQ exceeded)
00:21:29n1sstevenm: I have a new midi patch if you would like to have a look
00:21:30 Join Bjoern-Erik [0] (
00:21:30stevenmn1s, ooh cool! What's it do?
00:21:32n1scan't type :(
00:22:02n1s here if you're interested
00:22:09stevenmsure, I can throw it on. I've been looking at the IRC changes pretty often
00:22:15stevenmer nor IRC, SVN
00:22:21stevenmthanks! will put on now
00:22:56 Nick bertrik_ is now known as bertrik (
00:23:35n1smy monkey island test file now plays with only one buffer miss at 45 mHz
00:24:08stevenmn1s, ????! WOAH
00:24:26stevenm(also, I see I'm not the only one using monkey island)
00:24:38n1swell, it's pretty simple other files miss a lot more
00:25:03stevenmis this good for just coldfire, or PP as well?
00:25:19LloreanStill, Monkey Island is the important one.
00:25:32n1sI only have coldfire to test on so I don't know...
00:25:37*Llorean is not ashamed to admit the whole reason he wants MIDI is to put some Lucasarts adventure game music into his shufle.
00:25:38stevenmdare I say.. 44100 ?
00:25:58stevenmLlorean, I mean, MIDI is nice and all.. but there's the whole monkey island soundtrack
00:26:00 Quit lee-qid ("aufwiederbyebientotsayonara")
00:26:30stevenmLlorean, it's like, 2 CDs per game. Got it off worldofmi or scummbar or one of those
00:26:38stevenmn1s, building..
00:26:44n1sI have found some pretty nice ones from the Japanese version of ff2 :-)
00:26:45Lloreanstevenm: I had the CD edition of MI1 with redbook audio as it is.
00:27:07stevenmLlorean, aah, audio straight as audio tracks ?
00:27:24stevenmI know Sega Saturn did that
00:27:34jhMikeSdual-core MIDI on PP for sure. emac on cf for sure. those two things help about the same for each respectively.
00:27:43stevenmwe actually bought earthworm jim 2 for saturn just to rip the 'Puppy Love' track
00:27:44 Join jurrie [0] (
00:28:23n1sjhMikeS: or 22kHz on pp if that's possible at all
00:29:38n1sbut if this patch iproves performance as much on pp as it did on cf think some very basic midis should play ok now
00:29:38 Quit hannesd ("Client suicide")
00:29:42jhMikeSn1s: H10 dc works, iPod Color dc works. I need to clean it up and maybe get some PP5022 tests and it's good to go. Why shoot for 22kHz? It's possible though.
00:31:04n1sjhMikeS: 22kh seems simpler to me (but I have never done any multi threaded/cored coding so...)
00:31:44n1salso btw congratz on getting that dc stuff working :-)
00:31:58jhMikeSsure it's simpler. All the worst SPCs play with echo on PP dual core. I don't know how this compares in load.
00:32:26n1smaybe I should get some cheap pp player to play around with
00:32:52LloreanYou can get a Sansa for fairly affordable prices. And I'm sure 1st gen Nanos are fairy cheap these days.
00:32:56stevenmn1s, with midi, we have have sample n and sample n+1. Maybe we can send n, (n+n+1)/2 at 44100?
00:33:04jhMikeSit's always good to have at least one of each processor type to mess with
00:33:06n1sjhMikeS: well, preglow wanted to add reverb so I'm sure the extra horespower will help :-)
00:33:20jhMikeSto midi?
00:33:29stevenmhow would we add reverb?
00:34:39n1sI don't know if he meant just for midi but I guess it sgould be done as dsp, dunno much about it though
00:34:42jeplerI see that usb is getting closer .. I enabled it in config-e200.h and the device identifies to linux (but doesn't do anything useful)
00:34:43stevenmI've got a wish list of things to add... pitch bend depth would be good, and vibrato would sound real nice
00:34:44 Quit Soap (Connection timed out)
00:34:45*jhMikeS needs to cleanup the cd code a bit. some stuff is sort of messy still
00:35:19stevenmvibrato would mean offsetting the deltas by a sign wave on every sample. that could get ugly
00:35:42[AFX]i like rockbox
00:36:00[AFX]can i request something?
00:36:02jhMikeS[AFX]: i sort of like it too :D
00:36:14n1sstevenm: we can use a precalculated sin table and stuff it in iram and it should be doable
00:36:56n1sor didn't you mean sin?
00:37:12stevenmn1s, yes, sin
00:37:25jhMikeSn1s: want one? It's 32-bit fractional fixedpoint though and uses emac on coldfire. Easy rescale or precision change.
00:37:30pixelman1s: considering the sound of my 32kHz mp3 on swcodec I don't know if the resampler is such a nice idea... ;)
00:37:33stevenmn1s, yes, that would work. we just have to know how much to scale it depending on the needed vibrato.
00:37:46[AFX]is there any way you can add a sleep/powerdown feature instead of holding the buttons down (kind of like podzilla)? that way i dont have to boot into rockbox every time
00:38:19stevenmn1s, this is running at 45MHz now?
00:38:32[AFX]or even if it cant powerup back to rockbox, just make the feature available from the main menu?
00:38:37Llorean[AFX]: I don't understand what you're asking. Rockbox is designed to be a replacement firmware, and the intended behaviour is to boot Rockbox by default.
00:38:43n1sstevenm: no I just use 45 mHz to test, it won't miss otherwise
00:38:51stevenmn1s, aah ok
00:39:05[AFX]i see
00:39:18n1sjhMikeS: not today, but maybe one day :-) or maybe stevenm would want to do it
00:39:50 Join Soap [0] (n=Soap@rockbox/staff/soap)
00:39:56jhMikeSjust offering code that's ready made. It's used for a fixepoint blit synth so should be pretty quick.
00:40:17n1spixelma: true :-)
00:40:58n1sjhMikeS: how does that compare to fetching precalculated values from an iram table?
00:41:19jhMikeSn1s: it does that and interpolates them
00:41:46n1saha, sounds sweet :-)
00:42:12pixelman1s: nice about the theme settings chapter, looking forward :)
00:42:18 Join webguest60 [0] (i=4a469ecd@gateway/web/cgi-irc/
00:43:02[AFX]is there a gameboy plugin that exists for rockpod?
00:43:23n1spixelma: it was mostly a bit of cut and paste and adding the few new options
00:43:34stevenmn1s, wow, you re-wrote a lot of the synth loop
00:43:38webguest60instead of using ubuntu, i am using gparted. Now, how do i download the files i need when i can't access the internet through gparted?
00:44:06 Join Canar [0] (
00:44:27stevenmn1s, did it change a lot of the ramping/adsr stuff that much?
00:44:29Llorean[AFX]: There is no such thing as "Rockpod"... If you mean "Rockbox", please simply read the manual before asking questions.
00:44:46[AFX]sorry man
00:45:02n1sstevenm: that was preglows idea, it now synths one voice for the whole tick at a time so it will only load voice stat etc once per tick but also enabled me to break some stuff out of the innermost loop
00:45:42n1sstevenm: according to my testing it should not change the output at all
00:46:01n1sexcept for fixing a small bug i introduced earlier
00:46:08 Quit matsl_ (Remote closed the connection)
00:46:15stevenmn1s, oh, ok. The ramping/rates/whatever stuff is truly a pain in the neck
00:47:13stevenmI don't think anyone out there really knows how it is supposed to be done
00:48:03stevenmn1s, what are the magic numbers 0x7FFF80 ?
00:48:14n1salso I got rid of the cpShifted variable and do the shifting of the new cp_temp (which caches so->cp) in place where cpShifted was used before
00:48:44jhMikeSisn't MIDI a standard that should have info all over the place by now?
00:49:13stevenmjhMikeS, MIDI, yes. although I can find surprisingly little info on pitch bend depth
00:49:49stevenmjhMikeS, the real issue is finding correct specs for the Gravis patch format. There are two documents out there, and from what I can tell, they sort of copied each other
00:49:56stevenmand they are both incorrect in the same ways
00:50:02jhMikeSI suppose I'd compare results with a MIDI sound card or something
00:50:09n1sstevenm: those values were shifted right 7 and then masked by 0xffff and then the left sample were shifted left 16 before, masking like that makes the first shift go away
00:50:28stevenmwhy 7?
00:50:48 Join RCM123 [0] (
00:50:55stevenmor is that from the crazy rate/offset code?
00:50:56n1sthat I dunno, it was like that when I started
00:51:32stevenmn1s, aah, ok. I think it is probably the interpolation code. sorry, i am not all that awake at the moment
00:51:52jhMikeSit's an address mask of sorts?
00:53:25n1sjhMikeS: basically we want to get those 16 bits out of the 32 bit variable
00:53:48 Quit [AFX] ()
00:54:38 Join Rincewind [0] (
00:55:38jhMikeSI just looked, now I see.
00:57:06 Join chris__ [0] (
00:57:36 Quit billytwowilly (Read error: 113 (No route to host))
01:00:36 Quit bluebrother ("leaving")
01:00:44Zagorwhat's the difference between CPU_INT_STAT and INT_STAT on pp5020?
01:00:53Zagor(in pp5020.h)
01:01:20Zagordon't say "different addresses" :-)
01:02:27jhMikeSWhat's INT_STAT?
01:02:49jhMikeSah, forced interrupts
01:03:22Zagorforced by software?
01:03:37jhMikeSoh, hehe. not sure but I'm guessing it's perhaps a source mask and perhaps shows status regardless of masking on core
01:04:00jhMikeSforced ones are force by software, yes
01:04:20bertrikwhere can datasheets about the PP be found?
01:04:25Zagorthey can't
01:05:07jhMikeSI never could see interrupts forceable on COP though. Just CPU.
01:05:13 Quit scorche|w ("CGI:IRC (EOF)")
01:05:18bertrikso basically all info about this chip is from reverse engineering?
01:05:33Zagorbertrik: yes
01:06:24barrywardellZagor: with the USB code, will you add a charging device class, so that we can get the full 500mA when chargning?
01:06:46bertrikyou need a class for that? I don't think so
01:07:05Zagorcharging device class? you simply say you want 500mA and then you get it. all devices declare their power need.
01:07:36Zagorah right, it doesn't work today because we don't identify
01:07:37barrywardellyeah, but can you do that without going into msc mode?
01:07:42Zagorso yes, we'll get charging
01:07:44bertrikminimum device for charging has zero endpoints (just EP0) and 500 mA in it's descriptor
01:07:48Zagoryes you can
01:08:01barrywardellyeah, that's what I meant
01:08:50bertriki think most hosts will just supply any current you draw, until some kind of built-in the fuse gives up
01:09:20bertriki mean, not many hosts enforce the current limit
01:09:24Zagoryeah I think few hosts really throttle the power
01:09:41***Saving seen data "./dancer.seen"
01:09:42bertrikso there must be some logic inside the device
01:09:51barrywardellyou still need to set a device class in the descriptor
01:10:15barrywardelle200 only draws 100mA in rockbox, but 500 mA in OF
01:10:17Zagorbarrywardell: no, you can set it 0 and leave the class to the endpoints
01:10:22LloreanCould charging mode be "Fixed" relavitely soon?
01:10:36barrywardellah, yes. and then just have ep0. i see
01:11:12ZagorLlorean: yeah I have it working now so I guess we could add it pretty soon.
01:11:25 Join einhirn [0] (
01:11:31LloreanZagor: Good to hear. For people who don't change songs often, this would mean almost never needing to boot the OF already.
01:11:33jhMikeSThe AS3514 is responsible for setting current so I'm glad it's just running gently in rockbox.
01:12:08*Llorean wishes more devices were like the gigabeat with the ability to charge from mains or from USB.
01:12:35bertrikso as an dirty ugly hack, you could just ask the as3514 to draw more power even when USB has not been properly enumerated, don't you think?
01:12:50jhMikeSOnly the PP targets seem to not have mains charging without adapters
01:13:03Zagorsounds like usb isn't a factor in the charging equation then
01:13:10Zagori mean "proper" usb
01:13:22Lloreanbertrik: I do think hosts do the 100mah throttling though.
01:13:26jhMikeSbertrik: Possibly. I don't want an overcharge risk.
01:13:40 Join mokkurkalve [0] (
01:13:47Lloreanbertrik: I know that my iPod at least seems to charge faster from the wall adapter than from my laptop, both while in Rockbox.
01:13:51barrywardellwhat about the other pp targets. don't some of them not draw enough power
01:14:03 Quit webguest60 ("CGI:IRC (Ping timeout)")
01:14:17Lloreanbarrywardell: As far as I know, all the portalplayer targets suffer slow charging in Rockbox while attached to a PC or USB hub.
01:14:18barrywardellI thought I remembered something about that with the ipods
01:14:43 Quit ze (Read error: 104 (Connection reset by peer))
01:14:44jhMikeSIf I plug USB in the emulator, retailos automatically tries the charging but gives up after a few seconds.
01:15:35bertrikZagor: i read you were writing an USB driver for the sansas. I'd like to help if possible
01:15:49bertrikI have an electronics degree and experience in embedded software
01:15:52 Nick Gnu47 is now known as ChannelComedian (i=Gnu47@
01:15:57ZagorjhMikeS: I could use some help with something I got relayed from mr.h about the usb interrupt: "Well just in case nobody else has figured it out yet, at least on Sansa it's the bit 20 on the lower interrupt controller (0x60004010)."
01:16:27jhMikeShehe. Gotta love the way he puts things. :)
01:16:30bertrikAlso I wrote an USB driver for the USB device controller in LPC214x microcontrollers
01:16:41Zagorbertrik: sure. it's in quite an early state still, but some bits will hopefully be committed soon
01:17:01bertrikwhat kind of controller is in there anyway?
01:17:05Zagormost of the core stuff is done though, but not much yet on the scsi/ata translation
01:17:11Zagorbertrik: ARC OTG
01:17:35jhMikeSZagor: I don't think anyone was poking too much.
01:18:05bertriki have no experience with OTG, but getting the device side is a good start
01:18:09 Nick ChannelComedian is now known as Gnu47 (i=Gnu47@
01:18:23ZagorjhMikeS: I've been trying a few variants, but I can't get that interrupt activated. which port should enable the interrupts on INT_STAT?
01:18:24jhMikeSFrom the sound of it, it seems the INT_STAT must do something like I mentioned.
01:18:36jhMikeSCPU_INT_EN = (1<<20)
01:18:52Zagorbertrik: yeah I'm doing device only for now.
01:19:48ZagorjhMikeS: yeah that's what I've been trying. though I don't get any interrupts.
01:19:51 Nick RCM123 is now known as lightbulb (
01:20:00 Nick lightbulb is now known as RCM123 (
01:20:03jhMikeSTo MrH: Why didn't you tell me there were two GPIO banks? Then I wouldn't have had to figure that out. :P
01:20:09 Part RCM123
01:20:23barrywardellI thought I had usb interrupts working at one stage
01:21:11barrywardellbut maybe they were already enabled on h10
01:21:13jhMikeShmmm...will simply plugging USB fire them? Maybe a register needs to ACK it.
01:21:14Zagorhmm, I think I actually set the wrong bit...
01:21:44ZagorjhMikeS: reception of status message should fire interrupts.
01:22:08bertrikwhere do guys get hardware info if there's no datasheet? is there a repository of reverse engineered info somewhere?
01:22:33Zagoreuh, setup messages I mean
01:23:07 Quit chris__ (Read error: 113 (No route to host))
01:23:13 Quit Gnu47 (Excess Flood)
01:23:15Zagorbertrik: the ipodlinux project did the first bits, and we've simply continued from there
01:23:33 Join Gnu47 [0] (i=Gnu47@
01:23:33Zagor(the information, not the code)
01:24:21 Join ze [0] (
01:25:21Zagoryess, irq works
01:25:41jhMikeSZagor: Well, if you didn't set bit 20, I guess you wouldn't see them.
01:25:44Zagoreverything works a lot better with proper irq than polling
01:26:13Zagorcorrect. I mistakenly used the wrong USB_MASK define, which I thought I corrected. (I wonder which file I did that in then...)
01:26:14jhMikeSpolling is nasty in general. /me has aversion to tick tasks
01:27:47 Quit Arathis ("Bye, bye")
01:28:20*jhMikeS sees the H10 scrollpad GPIOD _input_ isn't used to detect when the pad is touched and avoid reading it.
01:28:46 Join FrododorF [0] (n=balgjolo@
01:28:47jhMikeSah, scratch that.
01:29:16jhMikeSI misread
01:29:26 Quit ender` (" Just because I have a short attention span doesn't mean I")
01:29:46*jhMikeS what ender` said :p
01:30:06 Part toffe82
01:30:20 Quit Robin0800 (Read error: 104 (Connection reset by peer))
01:30:40 Part FrododorF
01:31:28jhMikeS...pretty much sums up...^
01:31:58Zagorbertrik: do you know if it is allowed for usb devices to present MaxPacketSize 512 even when connected to usb 1.1?
01:32:28Zagori.e. do I have to bother with the alternative interfaces?
01:32:36bertrikno, i don't know for sure. i don't think so
01:32:46 Quit Gnu47 (Excess Flood)
01:33:01bertriki only have experience with full-speed devices
01:33:06 Join Gnu47 [0] (i=Gnu47@
01:35:07Zagorall the high speed devices I've looked at has the qualifier descriptor for full speed, so I guess I have to add it too
01:35:33 Quit Toxicity999 (Remote closed the connection)
01:36:09Zagorthough it does feel a bit silly. MaxPacketSize is _max_ packet size. if you want smaller, use smaller...
01:36:53barrywardelljhMikeS: since you mention it, have you looked at improving the h10 scrollpad at all?
01:37:39SoapI keep waiting until the Europeans are asleep, but I have a Nano with the crashing problem, and I'll pay round-trip postage to/from any developer interested. - i.e. amiconn .
01:37:40jhMikeSbarrywardell: I'm in the thinking stage. It seems it's sort of a button and a scrollwheel and would probably use both types of interface at once.
01:38:03barrywardellthere's also the problem of unstable readings from the adc
01:38:21barrywardellthere are a few patches for scrolling in the tracker, but none of them work too well
01:38:47jhMikeSproper filtering and some quantization of the values I think
01:38:52 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
01:39:52 Quit linuxstb_ ("ChatZilla [Firefox]")
01:39:52jhMikeSIt's a far more complicated control than a scrollwheel in any case.
01:41:25*bertrik is European and almost asleep
01:42:00barrywardellit would be great to get ADC_STATUS working as it should, so there would be no need for filtering
01:45:03jhMikeSADC reading will still have higher-f components than are needed. Does the status bit just not get set?
01:45:40barrywardellno, but it should, and my disassembly of the of shows if even waits for it to be set
01:46:37jhMikeSdoes it set anything before that? perhaps clear it?
01:49:01barrywardelli'll try to find the disassembled code i had
01:50:05barrywardellthere it is:
01:50:08 Join billytwowilly [0] (
01:50:26jhMikeShmmm...perhaps this can run background like meg-fx?
01:53:16barrywardellwhere does the meg-fx do that?
01:54:09barrywardellnvm, I see it now
01:54:47 Join jba_ [0] (
01:55:02barrywardellthat would work if the bit got set in ADC_STATUS
01:55:05jhMikeSwhat disassembler puts out nice code like that?
01:55:52 Part signuts ("Leaving")
01:55:54jhMikeSor is it your own decompilation?
01:56:27barrywardellmy own
01:56:39barrywardellso there could be mistakes in there too
01:56:57 Part pixelma
01:58:21 Quit billytwowilly (Remote closed the connection)
02:05:08 Quit Nico_P (Remote closed the connection)
02:05:23barrywardelljhMikeS: I'm also not sure what 0x70000010 is for. any ideas?
02:06:19ZagoridVendor 0x0781 SanDisk Corp.
02:06:25ZagoriProduct 2 Rockbox
02:06:26jhMikeSactually I was looking at what 0x60006008 really might be. I called it DEV_OFF_MASK but there may just be another reset and enable reg
02:06:59bertrikyou just enumerated and configured successfully?
02:07:29*jhMikeS feel like he should know this 70000010 from somewhere
02:07:32 Join billytwowilly [0] (
02:08:12 Join webguest60 [0] (i=4a469ecd@gateway/web/cgi-irc/
02:08:15webguest60instead of using ubuntu, i am using gparted. Now, how do i download the files i need when i can't access the internet through gparted?
02:08:40Zagori'm off to bed
02:08:41 Quit Zagor ("Client exiting")
02:09:05iambenuse a usb flash drive
02:09:26iambenhell, use your ipod or whatever you've got
02:09:35webguest60oh, that's a good idea : )
02:10:07webguest60well, i have a card reader and a memory card, will that work?
02:10:43jhMikeSbarrywardell: it's used in the e200 lcd controller init too
02:12:14iambenwebguest60: any usb storage should work, i tested it earlier w/ my microsd card + reader
02:12:24jhMikeSalso for i2c init...maybe it configures pins?
02:12:56webguest60awesome, thanks for the info.
02:13:35*barrywardell checks the e200 lcd code
02:14:18 Quit bertrik ("sleep")
02:15:03 Join ashes [0] (n=ashes@2001:5c0:8fff:ffff:0:0:0:4d)
02:16:53barrywardelljhMikeS: it does look a bit like it configures pins. it uses different values for different sources
02:17:52 Quit webguest60 ("CGI:IRC (EOF)")
02:20:01 Quit Gnu47 ("You know you'll miss me :P")
02:20:57jhMikeSyeah. just has that look about it. :)
02:22:01barrywardelland 60006008 has the look of a reset reg
02:24:54 Join Gnu47 [0] (i=Gnu47@
02:25:54jhMikeSI think I said that :). it's part of the main dev reset. I named it but not quite appropriately I think.
02:26:36barrywardellyeah, I was agreeing with you :)
02:29:37 Quit jhMikeS (Nick collision from services.)
02:29:43 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
02:31:36jhMikeS0x60006008 = DEV_RS2, 0x60006010 = DEV_EN2 if I read well (repost - lost connection)
02:34:25barrywardelland 0x60006094?
02:36:33barrywardellsomething to do with clocking?
02:37:07jhMikeSthat's a mytery but is used on e200 as well
02:37:31jhMikeSit's used right near other clocking regs
02:39:41 Quit n1s ()
02:40:51barrywardellwhere is it used on e200?
02:41:22jhMikeSnoowhere in our source. 60006080 is used though to set the iis speed
02:44:03jhMikeS0x600060a4 sets i2c speed
02:45:48barrywardellthese things should probably be added to pp5020.h, even if we don't use them yet
02:47:02barrywardellhmm, we call 0x70000010 ADC_ENABLE_ADDR
02:48:00 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
02:49:37jhMikeSBut it's not limited to ADC
02:54:57 Quit jba_ (Read error: 110 (Connection timed out))
02:58:57 Join psycho_maniac [0] (
02:59:35 Quit Rincewind ("Verlassend")
03:04:21 Quit PaulJam (".")
03:09:45***Saving seen data "./dancer.seen"
03:15:26 Quit hcs ("Leaving.")
03:16:18 Quit einhirn (Read error: 104 (Connection reset by peer))
03:17:06 Join Davide-NYC [0] (
03:17:29 Join webguest63 [0] (i=4a469ecd@gateway/web/cgi-irc/
03:17:43webguest63is there anyway to use linux and windows at the same time?
03:18:37LloreanThat has nothing, at all, to do with Rockbox...
03:20:09webguest63i was just wondering to see if anyone would answer this quick question like a yes or no
03:20:19Davide-NYCyes (VMWare)
03:20:52Davide-NYCAnyone up for a gigabeat cross sensor discussion?
03:21:08Lloreanwebguest63: In the future, please don't bring off-topic questions to this channel. The guidelines exist for a reason.
03:21:21LloreanDavide-NYC: What about the cross pad?
03:21:35Davide-NYCI met Markun yesterday. He mentioned that there are 13 sensors in the crosspad of which five are currently being used.
03:21:45webguest63sorry, and thanks
03:21:51 Quit webguest63 (Client Quit)
03:22:15Davide-NYCBefore I was made away of this I felt a 'wierdness' in the feel of the crosspad. I think I know why that was.
03:22:51Davide-NYCHe mentioned that he had written some code, plying around with all 13 sensors, taking pressure values etc.
03:23:40LloreanI don't even know what sort of "weirdness" you're talking about, nor can I guess why it is without a little more description.
03:23:47LloreanI personally haven't found the crosspad odd at all for use.
03:24:15Davide-NYCThe tips of the 'cross' and the center are the only parts of it that register input.
03:25:01LloreanFor me there's only a very small space between an "arm" of the cross, and the center, that is dead.
03:25:14Davide-NYCthere are two more sensors per 'leg' that could be put to great use in term of 'feel'. I'm having trouble expresing myself here. Not for lack of trying.
03:25:22LloreanAnd I think a deadzone is very important to keep people from accidentally pressing "Right" and "Center" simultaneously for exampe
03:26:47Davide-NYCI believe there is substantial gain to be had in the 'feel' for the users. That's an opinion.
03:26:51LloreanFor me at least, approximately 50% of the area on an arm/leg responds to touches.
03:27:10LloreanI don't know what you mean by "feel" though.
03:27:18LloreanAre you having problems getting the Gigabeat to respond to touches?
03:27:46Davide-NYCMy wife couldn't scroll through items in a list until I showed he just how to press.
03:27:57LloreanWhat was she doing wrong?
03:28:59Davide-NYCTHe crosspad is a bit 'particular' or 'touchy'. You have to press it just right for it to register. Once I told her to stay on the tips of the crosspad it was fine. I presume she was using it like a D-Pad.
03:29:26Davide-NYCI presume that she assumed that any pressure along the legs would register.
03:29:59LloreanWell, it helps to know what was being done wrong. The statement "there is a problem" is a lot less useful than "here's where users have problems"
03:30:24Davide-NYCRight. Like I said I'm having trouble expressing myself on this one for some reason.
03:30:25LloreanFor example, from day 0 I never had any problems using it, and I've never heard anyone else complain of which sensors are on or off, or the feel, except Soap at DevConW
03:30:26 Join TNTales [0] (
03:30:33LloreanI use it very much like a D-pad
03:30:46LloreanI rest my finger on the leg I want to use, and apply pressure when I want it to move.
03:31:16LloreanHow many Gigabeats have you tried with?
03:31:24Davide-NYCJust the one I own.
03:31:28Davide-NYC Good point.
03:31:51LloreanI'm not saying it's certain, but it's *possible* that yours has a touchy touchpad.
03:32:00Davide-NYCTis possible
03:32:33LloreanBecause honestly, I don't need to be anywhere near the tip to get responses.
03:32:56Davide-NYCMaybe my crosspad is 'low gain'.
03:33:03*Soap still thinks the gigabeat touchpad is "tricky". At least trickier than he desires.
03:33:10LloreanAs long as I'm approximately halfway down the leg (from the center, not from where it meets other legs), I can get it to respond
03:33:27Davide-NYCMine too. (there's two folks)
03:33:45LloreanSoap: Maybe you can describe what sort of trickiness you feel of it, I never quite understood at DCW
03:33:58Davide-NYCThis is going to be very difficult to describe (for me at least)
03:34:44TNTalesI am having trouble dual-booting rockbox with the OF for my sansa e260. I've found several forum discussions about dual-booting the e200r, but not the vanilla. I get a typical error: I press the <- when powering up to boot to the of (same thing happens when I plug in the usb without pushing anything) The screen fades up from black to white and gets caught rebooting. Is there a fix for the e200 (not Rhapsody)?
03:34:55LloreanDavide-NYC: Why not explain what you expect out of it, and how it fails to meet that expectation?
03:35:04LloreanRather than describing what it does wrong, describe what it doesn't do right.
03:35:22LloreanTNTales: Which version of sansapatcher did you use to install?
03:35:40TNTaleslet me check, it was from one of the tutorials
03:36:03LloreanAs in "not the one from the manual"?
03:36:33psycho_maniacTNTales: are you the one that tried installing on your "R" but then realized that it wasnt an R..or was that somebody else?
03:37:06TNTalesyeah.. that was me <.< >,>
03:37:30TNTalesI know just enough about all this to break something
03:37:56TNTalesI downloaded the bootloader from this site the online manual
03:38:05SoapLlorean, the iPod touchpad is equally tricky, but - and this is what I feel is the important distinction - the ipod touchpad does not control - either in Rockbox or in the original firmware - the "important" functions. You can not accidentally skip tracks on the ipod, or change contexts (WPS to menu for example) with the touchpad - the physical buttons do that. On the gigabeat you can - as the touchpad is used /instead/ of physical buttons, not in adju
03:38:05Soapnct to. I don't have to use the keylock on the ipod to prevent accidental button presses, I do on the gigabeat.
03:38:19LloreanTNTales: What version number does the sansapatcher program report?
03:38:34TNTalesWhere do I see that, on the player?
03:38:39LloreanYou don't see that on the player
03:38:44LloreanYou see that when you run the program
03:38:52TNTalesjust a second, let me check
03:39:08LloreanSoap: Ah, so it's not so much the touch-sensitive areas, for you, as the lack of tactile feedback for critical controls?
03:39:38TNTalesit says "sansapatcher v0.4 with v2.0 bootloaders" when I run the program
03:39:54Soapyes, and it doesn't help that the touch-pad is the opposite of recessed on the gigabeat.
03:40:08LloreanTNTales: You need to be using 0.5
03:40:32LloreanSoap: So, while a relevant problem, a different one from what Davide-NYC seems to be bringing up, I think.
03:40:40TNTalesok, where do I find it. PS: the online version (HTML) of the manual links to taht old version
03:40:41LloreanYours is more simply "Toshiba made a silly design decision."
03:41:07LloreanSoap: BTW, on the S (and I assume the X) it's actually a four-way cross of buttons, with a center button, rather than a touchpad.
03:41:19LloreanI guess Toshiba realized it too
03:41:23Soapbut not just tactile feedback - you barely get that on the ipods, heck you barely get that on the H300. The issue is that non-intentional pushes are too easy. Yes, not at all what Davide-NYC was talking about. Just defending my dislike of the gigabeat implementation - sorry to imply otherwise.
03:41:36LloreanTNTales: No, the manual links to the current version. When did you download it?
03:42:03TNTalesok, I will try again, but I'm pretty sure I got it there yesterday. But let me download it again in case I am wrong
03:42:04LloreanSoap: Well, I'm just clarifying as he's already added you to the number of people bringing up his problem. ;)
03:42:21Davide-NYCLlorean: Soap: I have to backtrack a bit. It is possible that my "right' sensor is failing.
03:42:49TNTalesthis is the link I am clicking to download from the online version of the rockbox manual
03:43:11LloreanTNTales: Yes, that's where we put the newest version when a new version comes out.
03:43:18SoapDavide-NYC, since there are no moving parts in the touchpad, I would suspect a bad connection @ the ribbon cable before I would suspect a physical failure.
03:43:29Davide-NYCIt's not failing. (at least I don't think it's failing)
03:43:41Davide-NYCwhen I tap the crosspad it works fine.
03:44:46Davide-NYCMaybe it's just the timing parameters I don;t like.
03:45:12TNTaleshmm, I guess I must have downloaded it somewhere else. I will try this patcher. I have the current build already on the player, do I need to make any other changes?
03:45:14Davide-NYCI just watched my wife press 'right' ten times and nothing happened. Then I try it and it works fine. It's as if I'm losing my mind.
03:45:22TNTalesI have the right one now
03:45:24LloreanTNTales: When you download that one, it reports 0.5?
03:45:30TNTalesyes, llorean
03:45:36LloreanYou should restore the Sandisk firmware first
03:45:42LloreanThen install the bootloader with sansapatcher
03:45:57LloreanYou don't need t do anything else
03:46:14TNTalesOk, I have the player with the original firmware running but rockbox is on the player (just not patched to boot)
03:46:29 Join jba_ [0] (
03:46:35LloreanTNTales: Then run sansapatcher on it, and you're done.
03:46:47 Quit miepchen^schlaf ("Verlassend")
03:51:26TNTalesok, It went to the load screen, faded to white and then loaded the OF, is that how it's supposed to work?
03:52:25TNTalesOk, thank you guys very much. Do you mind if I lurk awhile in case I break something as I get it set up? I usually try the forum first though
03:55:05 Quit barrywardell ()
03:56:23 Quit Bitter (Client Quit)
03:57:31 Join HellDragon_ [0] (n=Nocebo@unaffiliated/helldragon)
03:57:34TNTalesif a theme requires a custom build, do you have to install a full rockbox zip or just the files for that WPS?
04:00:04TNTalesare custom builds dependent on their author for upgrades? like you can't update by getting the most recent daily build, correct?
04:00:21 Nick HellDragon_ is now known as Bitter (n=Nocebo@unaffiliated/helldragon)
04:00:36 Join saratoga [0] (i=9803c6dd@gateway/web/cgi-irc/
04:00:37Davide-NYCI no longer have an H300 and was wondering if the unit still goes to OF when plugged into it's charger?
04:00:56saratogaLlorean: I was going to upload the cabbie stuff i've been working on to somewhere
04:01:04saratogawould you recommend the patch tracker or the wiki?
04:01:42LloreanDavide-NYC: As there hasn't been a new bootloader release yet (as far as I know), nothing dependent on bootloader behaviour has changed.
04:01:57Lloreansaratoga: I guess the patch tracker for now.
04:03:02Davide-NYCThere was a meetup here in NYC last night and the conversation went to discussing a 'release'. AFAIK the h1x0 series is fully supported.
04:03:27Davide-NYCI wasn't sure about the h3x0 or the iaudios
04:04:12Davide-NYCbut I guess the h3x0 needs another bootloader update before 'release'. What is the deve consensus on the concept of 'official releases?
04:04:45Llorean"We'd like to have one, one day, but there's still quite a bit of work to be done."
04:05:24LloreanVoice is unreliable, there are still some odd playback bugs, at least.
04:05:47Davide-NYChmm, is there documentation of these playback bugs?
04:05:50psycho_maniacmaybe look at the release todo page or the iriver status pages. . i thought there would be a release for the irivers as well till i looked at that page.
04:06:55TNTalesWhen installing a WPS, the instructions say to place it in the root folder of the player. Is this the root of the driver or the root for rockbox? Or does root always mean "device root"
04:07:04saratogathere was a todo list about 18 months ago when people thought we'd be ready to do a 3.0 release, however i don't think its been updated much
04:07:10TNTalesroot of the device*
04:07:16LloreanDavide-NYC: Just look at open bugs in the tracker, I think many of them have bug reports. Especially relating to seeking or track changes in the last few seconds of a track.
04:07:36LloreanTNTales: Always device root
04:07:45TNTalesthanks :)
04:09:35ATravelingGeeksaratoga: what player is the cabbie 'stuff' for?
04:10:15LloreanATravelingGeek: It's an attempt at a default theme
04:10:39ATravelingGeekIt'd be a nice default theme
04:12:33saratogayes i hope so
04:12:55saratogaif anyone is interested, i'll post the filespray number in a minute, and you're welcome to help port it to your player
04:13:08ATravelingGeeksaratoga: are you the original author of it?
04:13:16 Quit Thundercloud (Remote closed the connection)
04:14:16saratogaATravelingGeek: no, the original author is helping though
04:14:24ATravelingGeekAh, alright
04:19:12ATravelingGeeksaratoga: I'd be interested. How's it different from the current cabbie theme for the gigabeat series?
04:20:51saratogaATravelingGeek: very little different, i started with the gigabeat version
04:21:28saratogai just removed the gigabeat logo since technically we don't have permission to use it, and I removed the unofficial rockbox logos
04:21:42saratogaalthough i must admit i like the one it comes with
04:22:05ATravelingGeekI like the rockbox logo that's used in it
04:22:40saratogaits very nice, but for now we're not going to have a logo in there
04:22:40 Join ddalton [0] (
04:22:53saratogathough eventually i'd like to ask for submissions for one, and maybe we could choose that one
04:23:20ATravelingGeekIs the 'rockbox' text still left in?
04:23:32saratogathe two colored one?
04:23:50saratogano i took that out too
04:24:39saratogawhen i asked around in IRC a couple days ago, most people seemed to think we shouldn't use unofficial graphics in the official wps
04:24:46saratogaso for now i'm just not going to deal with it
04:24:51ATravelingGeekOh ok
04:24:53ATravelingGeekMakes sense
04:24:55saratogawe can always update it later once we agree on it
04:25:12saratogathe challange will be getting everyone to agree to changing the wps at all
04:25:20 Quit jba_ (Read error: 110 (Connection timed out))
04:25:41LloreanPretty much everyone agrees to changing from the current default
04:25:53LloreanThe problem is getting everyone to agree on what a new default should or shouldn't be.
04:26:03saratogathats my concern
04:26:15saratogaall i need is a couple people who want to hold out forever waiting for the perfect wps
04:26:28 Quit midgey ()
04:27:11Soapwhy not just do a monochrome version of the official logo in the corner?
04:27:28saratogaSoap: yeah i was thinking about that
04:27:38saratogajust never really got around to it
04:27:51saratogaand secretly i'm hoping someone else will take interest in the aesthetics for me
04:28:04 Quit psycho_maniac (Read error: 113 (No route to host))
04:30:20 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
04:31:31ATravelingGeekHeh, the Sansa actually comes with some decent music
04:33:52 Quit Gnu47 (Nick collision from services.)
04:34:00 Join Gnu47 [0] (i=Gnu47@
04:34:10saratogado we have any logos besides the official?
04:37:47saratogaalso, is there a prefered font?
04:38:02 Join psycho_maniac [0] (
04:38:04saratogacabbie seems to default to one it includes called "chicago"
04:38:41 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
04:38:43ATravelingGeekWhich is no longer included in the font pack
04:38:48 Join z35 [0] (
04:39:18Soapcopyright issues?
04:39:28ATravelingGeekI think I read that
04:39:34saratogaok then i won't use that one
04:39:41saratogais there a prefered rockbox font?
04:41:08Lloreansaratoga: There should be an appropriate sized Nimbus to replace chicago
04:41:22LloreanNimbus 13?
04:41:42Davide-NYCthere's a Nimbus 13?
04:42:21ATravelingGeekNimbus-12 seems to work well
04:42:44Davide-NYCI must have an outdated font pack!
04:43:44LloreanI dunno
04:45:03TNTalesdoes rockbox always show the contents of a micro sd (sansa e200) in a separate file listing instead of the main database of tunes?
04:47:03LloreanI don't understand the question. The "Database" and the filetree are two separate things
04:47:27TNTalesok, when I use rockbox it has the option to scroll through the songs by album, artist, etc
04:47:45 Join sdoyon [0] (
04:47:54TNTalesbut the music I add to the microSD card doesn't show up there, but I can find it under a separate menu for the card
04:48:11LloreanThere shouldn't be a separate menu for the card.
04:48:18LloreanIt should show up as a folder under the "Files" menu
04:48:53TNTalesright, but is there a way to make it show up as a unified "library" with the other music on the main player drive?
04:49:12LloreanJust update the database while the card is inserted.
04:50:00TNTalesok, this is probably a stupid question, but, how do I update the database on here? I can't find it in the settings and I'm not sure where to look.
04:50:21LloreanThat's one you can find in the manual...
04:50:44TNTalesfair enough :) Thanks for all your help and patience, I'm going to spend some quality time and RTFM.
04:51:16Davide-NYCIs there ever any Twiki page culling done?
04:51:24Davide-NYCI have a suggestion for a deletion.
04:52:13LloreanPages do get deleted on occasion
04:52:39 Join homielowe [0] (
04:52:40psycho_maniacsometimes they dont but are posted on top in red that there obsolete.
04:53:26LloreanIf there's any value to being able to look at old revisions of the page, it's usually just best to delete the contents of the page, and add a note like "This page is outdated. Please go to (something) instead."
04:53:30 Join midgey [0] (
04:53:46Davide-NYCI am going to do that.
04:53:55psycho_maniaccan you tell the difference between nimbus13 and chicago 12? ,
04:55:04ATravelingGeekThey're close enoguh
04:55:09ATravelingGeekThe line height is just about the same
04:55:14 Quit Toxicity999 (Read error: 104 (Connection reset by peer))
04:55:20 Join Toxicity999 [0] (n=bryan@unaffiliated/Toxicity999)
04:55:31ATravelingGeekI like the first one slightly more though
04:55:57psycho_maniacfirst time i loaded it i couldnt tell untill i looked at them one after another like a slide show
04:56:29psycho_maniaci think thats nimbus-13. i cant remember though.
04:57:06LloreanNimbus was, iiuc, designed to be a replacement for Apple's Chicago font.
04:57:47 Quit karashata ("Leaving.")
04:58:16Lloreanpsycho_maniac: Are you sure you have the right chicago12 though?
04:58:18 Join karashata [0] (
04:58:22LloreanFrom the old font pack that we used to provide?
04:59:11psycho_maniacLlorean: i have the chicago 12 that came with that theme.
05:00:16psycho_maniacLlorean: i have a chocago font in the font back from aug 06. if you want me to try that.
05:00:43LloreanIt just seems odd that the characters look to be pixel for pixel identical, with the only difference being spacing
05:00:57ATravelingGeeksome of the characters are slightly different
05:01:14ATravelingGeeky / j / g
05:01:38Davide-NYCthe straighter one of the two is better IMO
05:01:55LloreanATravelingGeek: True, but somehow I doubt that's enough to avoid copyright infringement
05:02:10psycho_maniaci think the one with the J that goes below the other letters is chicago. is this correct?
05:02:28ATravelingGeekThey do look alike, though, other than the characters with bits that hang down (can't think of the correct term for it)
05:03:23 Join donutman25 [0] (n=chatzill@
05:04:01psycho_maniaci have them zoomed in and showing like a slide show and the only difference really seems to be the spaceing. the W is different also
05:04:12LloreanBut I swear that the old chicago font we used was visibly at least decently different from nimbus, but I guess I could be wrong
05:04:41psycho_maniacLlorean: you want me to try the font from aug of 06?
05:05:21TNTalesOk, I'm updating the database, but when I go to the file menu I see the (microSD) card directory and the video and TMP directory but not the directory that contains the music on the player drive, do I need to make a directory or have I missed something in the manual
05:05:34saratogaposted as FS #7946
05:05:37saratoganow i'm going to get some sleep
05:06:08LloreanTNTales: The "File" menu, shows you the files on the disk
05:06:15LloreanTNTales: It has nothing, at all, to do with the database
05:06:29LloreanBecause "Music" is a hidden folder, you need to set file view mode to "all" to see it.
05:06:38ATravelingGeekI'll work on that for the e200 unless someone else tackles it first
05:06:44TNTalesah. ok, thanks again
05:06:48saratogaATravelingGeek: that'd be great
05:06:56ATravelingGeekEr, one exists for it, but I can clean it up
05:06:59saratogatheres already a port to the e200, so it should be quite easy
05:07:11saratogatake a look at the gigabeat one i posted in the sim
05:07:14saratogashould give you a good idea
05:07:17ATravelingGeekI actually started on porting the original gigabeat theme until the author pointed me to the already ported one
05:07:30*ATravelingGeek will just load it onto his gigabeat
05:07:31saratogaalso, if you have any ideas regarding fonts, graphics, etc post them in that FS task
05:07:38ATravelingGeekEr... wait
05:07:45ATravelingGeekIs the only change removing the rockbox logo?
05:07:54saratogaand switching fonts
05:08:07saratogamostly i've been working on the ipod video version
05:08:17saratogathe aspect ratio change is harder then you might think
05:08:20ATravelingGeekIt looks lacking though now at the top of the screen
05:08:21saratogaanyway, good night
05:08:34saratogayeah we need a graphic or something i guess
05:08:45ATravelingGeekWell, for smaller screens it's not as much of an issue
05:08:46ATravelingGeekbut oh well
05:09:03 Quit saratoga ("CGI:IRC (EOF)")
05:09:17 Join RoC_MasterMind [0] (
05:09:49***Saving seen data "./dancer.seen"
05:10:18TNTalesok, well, good night everybody, I'm going to go experiment with rockbox, this is an awesome bit of software, thanks to all the devs and people who were so helpful to me, I appreciate you guys taking the time to help me.
05:10:30 Quit RoC_MasterMind (Client Quit)
05:11:25Davide-NYCjhMikeS: you ther buddy?
05:11:35*TNTales is AFK, ZzZZzzZZzzzZzzzzz —I-n-v-i-s-i-o-n—
05:11:48psycho_maniacone thing i dont like when people make themes. they change settings that dont seem to be in a theme. (ie: status bar: no)
05:11:48spkyhow can i get an older version of source with svn?
05:12:12 Quit hcs ("Leaving.")
05:13:57ddaltonHow does int_setting work?
05:15:29 Quit midgey ()
05:16:09 Quit bb (Nick collision from services.)
05:16:14 Join bb_ [0] (n=bb@unaffiliated/bb)
05:16:20 Quit Gnu47 ("You know you'll miss me :P")
05:18:00 Join Gnu47 [0] (i=Gnu47@
05:18:22 Quit Davide-NYC ("ChatZilla [Firefox]")
05:21:20 Quit karashata ("Leaving.")
05:21:43 Join karashata [0] (
05:22:42jhMikeSDavide-NYC: yes sir
05:22:48jhMikeSoops, left
05:24:41 Join zajacattack [0] (i=42617c5c@gateway/web/cgi-irc/
05:28:33psycho_maniaci have another comparison this time i labled them correctly and more leters to compare. and
05:31:03 Join Mouser_X [0] (
05:31:34 Join webguest03 [0] (i=62ce3370@gateway/web/cgi-irc/
05:32:12 Nick Gnu47 is now known as Gnu47|LandOfNod (i=Gnu47@
05:32:45Mouser_XHow do I copy a directory in cygwin? I ask, because I haven't been able to get Rockbox to compile (with patches. I'm compiling right now w/o patches to see what happens), and it'd be easier to simply copy the directory, and work with the copies than to add a patch, compile, remove the patch, compile, add a different patch, compile, etc.
05:33:08webguest03hi, i cant figure out how to get rockbox on my 1st gen ipod nano, when i run the rockbox utility and i set the thing to look in the correct drive for a 1st gen ipod nano, but i says no ipod found
05:33:32ddaltonMouser_X: why do you have to do that? what about svn?
05:33:54ddaltonyou copy a dir in cygwin like you do in linux. "cp dir to copy location"
05:34:34Mouser_Xddalton: What do you mean what about SVN? I've got the SVN here, but I think the patches I'm using are screwing it up, and I'm getting tired of downloading everytime it breaks.
05:34:49Mouser_X*downloading it again.
05:35:00Mouser_X(I've tried to revert, and it's still not working.)
05:35:05ddaltonOh ok svn revert -R . reverts your changes to as they were before you started adding patches.
05:35:27Mouser_X^ Didn't work last time I tried it. At least, it didn't compile.
05:35:28ddaltonsvn up updates to the latest only downloading the files that have been modified.
05:35:50ddaltonwhat happens when you type svn revert -R .
05:35:57ddaltonafter applying some patches?
05:36:10Mouser_XStuff. A list of files.
05:36:19psycho_maniacMouser_X: dont forget the "." its "svn revert -R ."
05:36:21 Join midgey [0] (
05:36:30Mouser_XAh. Well then.
05:36:44psycho_maniacsame thing happened to me.
05:36:47ddaltonsvn revert -R . should say what files it reverted
05:37:07ddaltontry compiling a clean one first to make sure your path and everything is setup correctly
05:37:18Mouser_XThat's what I'm doing right now.
05:37:41Mouser_X(I did it earlier, but my laptop crashed, so I had to "resume" a few hours later.)
05:39:54webguest03i open the utility from the extracted file, i choose the correct mount point, i choose ipod nano 1st gen, i go back to first screen, choose install complete, starting bootloader installation searching for ipods, no ipods found......anyone no what to do?
05:41:11*Mouser_X does not have an iPod.
05:41:38psycho_maniacwebguest03: i would suggest to use this, , this is how i installed rockbox when i got that same problem.
05:41:40Mouser_XFor that matter, I've never used the RockboxUtil or any other tools either...
05:42:04webguest03'ight, i'll try that psycho
05:42:47webguest03i would try to do it manually but i have trouble figuring out how to do all that stuff, lol
05:43:19Mouser_X:( @ compiling... It takes about 1.5-3 hours to compile Rockbox on my laptop... (My desktop needs a new power supply)
05:43:26ATravelingGeekIs it possible to turn off the Sansa e250 when it's connected via USB?
05:43:32ddaltonYeah cygwin is slow. :-(
05:43:39webguest03whats cygwin?
05:43:50ddaltongoogle will tell you
05:44:17Mouser_Xcygwin is a development environment. It's used for compiling Rockbox (this way, you can add your own patches and such).
05:44:22ddaltonbasically it is like a linux console for windows. Just command line
05:45:02webguest03ic i think i no what you mean, does it let you put your own stuff in rockbox?
05:45:14ddaltoncompile it from source
05:45:15Mouser_XWell, w/o patches, Rockbox appears to have compiled.
05:45:34ddaltondo you get hunk errors on the patches?
05:45:38ddaltonwhen applying
05:45:44psycho_maniacMouser_X: what patches are you trying to use?
05:46:02Mouser_XFor the GBS one, there's a few errors. It complains about some unused variables.
05:46:34Mouser_XI'm using the GBS and MOD patches. I'd like to try the NES emu patch, but I think it's got issues.
05:46:55Mouser_X(I'd give you FS#s, but I forget what they are.)
05:47:30Mouser_X5241 for the MOD patch.
05:47:41Mouser_X7331 for the GBS
05:48:20spkyi cant seem to patch with customline ;[
05:48:25Mouser_X2911 for the NES emu patch (sorry about the white-space)
05:49:43Mouser_X(The MOD patch appears to be recieving maintenance still. The GBS patch needs a lot of work.)
05:50:35 Part zajacattack
05:51:18 Join webguest05 [0] (i=62ce3370@gateway/web/cgi-irc/
05:51:21 Quit webguest03 ("CGI:IRC (EOF)")
05:51:42webguest05psycho_maniac, thanks for that other rockbox utility, it worked perfectly
05:52:50 Quit webguest05 (Client Quit)
05:53:09*karashata is looking at the IriverH10Port wiki page, thinks it's in dire need of an update...
05:55:12psycho_maniacstrange how the old utility worked for him while the new one rockboxutilityqt didnt.
05:55:39Mouser_XMaybe he wasn't using the new one? (I don't know. Just saying...)
05:59:48 Join jba_ [0] (
06:00:10jhMikeSkarashata: what in particular is out of date there?
06:00:26 Quit stevenm ("Connection reset by beer")
06:02:26karashatawell, for one, I believe the port's been around a lot longer than "a few months", but that's a small thing, really... and I don't think I've *ever* run into any of the issues listed in the "Common Bugs & Workarounds" section since I started using it a while ago...
06:02:45karashatathen again, I may just have been a bit luckier than some people
06:04:40 Quit alienbiker99 ("( :: NoNameScript 4.02 :: )")
06:04:41jhMikeSIt should be stable and useable now I think
06:04:57karashatait's been stable and usable for ages AFAIK
06:05:33karashatathough there was a brief period of time that it ended up unstable again, I was lucky to get a working build once out of every four I tried...
06:06:21karashataI ended up getting into the habit of keeping a build I knew worked until I had a chance to test the latest for random crashes and hangups
06:06:39Mouser_XI do that anyway.
06:06:56karashatait's probably a good idea considering the nature of the code
06:07:01jhMikeSthere were some recent changes that caused trouble. now it seems to survive pretty much any code change.
06:07:09*Mouser_X has been using the same build from June/July for months...
06:07:14karashatait's too bad I didn't learn that until the "flaky" period as I think of it...
06:07:36karashataI currently have the most recent build on mine, works beautifully
06:07:44jhMikeSit shouldn't be flaky now. the reasons for flakiness were identified and fixed
06:08:02Mouser_X(This is also why I'm attempting to compile my own builds. I feel I should update [lots of good stuff has been fixed/added], but I *really* want my MOD and GBS support...)
06:08:11karashata*nods* it's been stable since somewhere around the 14800's or something...
06:08:35karashataI don't remember when exactly it stabilized again, all I remember is I was happy it did
06:08:41ddaltonwhat MOD?
06:08:52 Join massiveH [0] (
06:09:06Mouser_XFS #5241 for MOD patch. It's a codec.
06:09:24Mouser_X(I actually listed all the patches I wanted earlier.)
06:09:42jhMikeSclocking work, changing thing it doesn't like...mine will not crash unless I make a coding mistake.
06:09:55ddaltonWow I had my voice file and my build building at the same time. It wasn't a clean build all that had changed was a / to a * And the voice finished first.
06:10:25Mouser_XMine crashes... When I throw it out the window while driving. (I've never done that though.)
06:10:25ddaltonOh ok can you tell me all the patches you want?
06:10:35ddaltonI will tell you if they are likely to work
06:10:48jhMikeSwell...unintended use aside :)
06:11:37Mouser_XFS#s: 5241 for MOD patch; 7331 for GBS (codec); and 2911 for NES emulator.
06:12:06ddaltonhow old are they (the latest update)
06:12:44Mouser_XMOD is very recent. GBS was very recently updated, but the update itself is from awhile ago (though, I know the author, and it still works for them).
06:13:23Mouser_XNES emu, I don't even know if it works. I'd like it, but I currently have low expectations for it.
06:14:43*Mouser_X is currently compiling Rockbox+GBS patch.
06:15:04Mouser_X(So far, it's working. Who knows if that will last all the way through though?)
06:15:19psycho_maniaca nes emulator would be nice. i have over 700 nes games.
06:15:40Mouser_XI have the GoodNES set.
06:16:13Mouser_X(And the GoodGB/C set, and the N64, and SNES... You get the idea.)
06:20:39 Quit ddalton (Read error: 104 (Connection reset by peer))
06:20:53 Join ddalton [0] (
06:21:05ddaltonMouser_X: Did you get hunk errors?
06:21:15ddaltonMy connection dropped out
06:21:30ddaltonwow voice building is fast
06:21:33Mouser_XNot yet. Or would I have gotten those when I patched it in? Either way, I am not aware of any errors happening.
06:22:07psycho_maniacstrange i dont see the gbs codec in the SoundCodecs wiki page.
06:22:20Mouser_XIt's like NSF.
06:22:28Mouser_XExcept for the Gameboy/color.
06:23:07psycho_maniacyes but its still a different codec so why isnt it on that page?
06:23:45Mouser_XI don't know. Not only do I not have a Wiki account, but I don't know "wiki language" stuff either.
06:23:58LloreanI didn't think GBS was in SVN?
06:24:03Mouser_XIt's not.
06:24:11LloreanSo why should it necessarily be on that page?
06:24:25psycho_maniactheres a lot of codecs on that page thats not in svn
06:24:29Mouser_XIt only runs on the Gigabeat (currently). It needs a lot of work before it's in SVN.
06:24:44Lloreanpsycho_maniac: Well, you *can* add it you know.
06:24:50Mouser_XPSF being one of them (though, that'd be *REALLY* nice to have...)
06:24:57LloreanMy point was just, if it's not in SVN, I don't see any reason one should *expect* for it to be there.
06:25:22psycho_maniacyeah it would be nice but i doubt it will happen as it says its not possible. maybe thats because nobody knows how to port it?
06:25:39psycho_maniacLlorean: you could add it too if you wanted.
06:25:48Lloreanpsycho_maniac: Yes, but I don't care.
06:25:50LloreanYou asked about it.
06:26:12 Quit sslashes (No route to host)
06:26:24psycho_maniacLlorean: i agree. its confusing to see all those codecs on that page when some arent even in svn.
06:27:19ddaltonis HZ/2 2 milliseconds
06:27:20Mouser_Xpsycho_maniac: PSF *is* possible. However, it would be difficult. Especially to get it running on "lower" targets.
06:27:47Lloreanddalton: Shouldn't it be half a second?
06:28:10psycho_maniacor atleast have them listed under something that shows that there not in svn. or being developed.
06:28:11ddaltonLlorean: but didn't I say two?
06:28:23ddaltonwhat would hz*2 be?
06:28:28ddaltontwo seconds?
06:28:54jhMikeSHZ = 100 since there 100 ticks/second
06:29:00jhMikeSso yes
06:29:32 Join sslashes [0] (n=rmf@
06:30:01ddaltonjhMikeS: So HZ/0 is no MS? and hz/9 is 9 ms and hz*9 is 9 seconds?
06:30:32jhMikeSHZ/9 is one-ninth of a second. HZ/0 would be a divide-by-zero error
06:30:42jhMikeSHZ*9 = 9 seconds
06:31:06ddaltonSo what would happen if you divide hz by 0?
06:31:32jhMikeSthe compiler would error since division by zero is not defined
06:31:46ddaltonwhat about a variable that can equal 0?
06:31:59jhMikeSHZ/variable ?
06:32:31jhMikeSthe system will crash if it does at runtime with a divide-by-zero error.
06:32:48ddaltonok thanks for telling me that
06:33:23sdoyonjhMikeS: Hi. I hear a kind of stuttering effect with the voice that seems to be caused by left-over data in the pcmbuf when shutting up the voice.
06:34:07sdoyonjhMikeS: I'm not sure how it's happening, but if I move the stuff in voice_stop() to mp3_play_stop(), the stuttering goes away...
06:34:20 Join SirFunk [0] (
06:34:34sdoyonjhMikeS: because that does a pcmbuf_stop() obviously...
06:34:51ddaltonsdoyon: I didn't realise you were here. But did you see p7910?
06:35:40psycho_maniacjust curious. why was the wiki table on the main page removed? it was between svn activity and mail.
06:36:00sdoyonddalton: I read the comments, not the code yet. Hold on a moment.
06:36:26ddaltonsdoyon: ok does it sound like a nice feature or not? Wasn't that one of your ideas?
06:36:34sdoyonjhMikeS: Does this mean anything to you?
06:37:03karashatapsycho_maniac: the Wiki table is still on the main page, its located above the SVN table, to the right of the Project news table
06:37:04jhMikeSsdoyon: I haven't heard what might be a stuttering effect yet.
06:37:30 Quit homielowe ("ChatZilla [Firefox]")
06:37:33karashatathey changed the layout of the main page some time ago
06:37:52psycho_maniackarashata: oh i see it now. i didnt notice that. quick look made it seem that that was all one table on top.
06:38:04karashatanope, it's two tables
06:38:05jhMikeSsdoyon: any recipe to make it happen?
06:38:13sdoyonjhMikeS: It's when I flick through several menu entries or files in very quick succession, like with the e200 scroll wheel. Say I'm on recent bookmarks, then I give it a spin downwards, I'll hear "fi..." (beginning of "files") then "system".
06:38:17karashataI think they changed the layout to save page space
06:38:37sdoyonjhMikeS: the up: "plu..." "recent bookmarks".
06:39:08ddaltonsdoyon: that sounds like a simular problem to my talking alarm patch
06:39:15jhMikeShmmm...yeah, some old data sounds like it being left from a clip that started but didn't play
06:39:17 Quit jba_ (Read error: 110 (Connection timed out))
06:39:27ddalton(Where the voice couldn't keep up with the scroling)
06:39:53 Quit massiveH ("Leaving")
06:40:18jhMikeSif the PCM buffer isn't flushed that could be the reason
06:40:33psycho_maniacwell as long as they keep the svn activity on there. im happy. the other stuff i like also but svn is the most i look at.
06:41:05sdoyonjhMikeS: mp3_play_stop() normally sens Q_VOICE_STOP, and I would think that causes voice_on_voice_stop() to be called and do pcmbuf_stop(), but apparently not always. As I said if I put the pcmbuf_stop() in mp3_play_stop(), this problem vanishes.
06:41:06karashataI suspect everything on the front page is going to stay there
06:42:21jhMikeSsdoyon: it might be overflowing the queue
06:42:24Mouser_Xddalton: The GBS codec compiled, but there was lots of warnings.
06:43:42Mouser_X(Rockbox itself isn't done compiling, but a gbs.codec file exists now)
06:43:42sdoyonjhMikeS: Hmm I kind of doubt it, I can make it happen by moving down only 3-4 items...
06:45:19jhMikeSthe problem of forcing a pcmbuffer stop there is it could stop playback
06:46:19sdoyonjhMikeS: No it's guarded by if (!playing).
06:46:37sdoyonjhMikeS: and if mixing then the problem doesn't occur anyway apparently...
06:47:04jhMikeSright, you moved the check too.
06:47:40sdoyonjhMikeS: Yes. I've been using that for several days actually.
06:47:44ddaltonsdoyon: Would the case_action_none: case be useful? You could put the code to voice the menu option there once scroling has stopped
06:48:57jhMikeSIt's not really safe to put that code there since there no threading sync. The reason it's not flushing the old clip should be found.
06:49:10ddaltonjhMikeS: My setting for a .talk clip delay is 1 ms and it goes up to 5 ms. But 5 ms is the shortest delay and 1 ms is the longest. How do I fix this?
06:49:18sdoyonjhMikeS: And the while (voice_is_playing || !queue_empty(&voice_queue) || ci_voice.new_track) ... although I'm unsure why it needs this many conditions...
06:49:42ddaltonI go #define hover_delay(hz/global_settings.talk_clip_delay)
06:49:50ddaltonor how ever the sintax is
06:50:59sdoyonjhMikeS: Does pcm_is_playing() become true as soon as we start putting data into it? Or is there some sort of low water mark?
06:51:08jhMikeSsdoyon: me too. the playback engine is messy right now but the MoB work will probably help that out.
06:51:43jhMikeSsdoyon: the pcm buffer does allow forcing playback even if the low water mark isn't met
06:52:25sdoyonjhMikeS: MoB?
06:52:48jhMikeSpcm_is_playing is true when the pcm driver has started playing samples
06:53:01jhMikeSMetadata on Buffer
06:53:48sdoyonjhMikeS: In voice_on_voice_stop(), it'll do pcmbuf_stop() only if pcm_is_playing(). But the left-over data sound like the beginning of the utterance. So if the shutup can come early enough that there is data in the pcmbuf but it hasn't started playing, then it won't get cleared.
06:55:49jhMikeSthat's true. it's a race condition. does removing the pcm_is_playing() help too?
06:56:58sdoyonjhMikeS: compiling, I'll tell you in a few minutes :-)
06:59:02sdoyonjhMikeS: Yes!
06:59:17sdoyonjhMikeS: That'd be safe, right?
06:59:47jhMikeSthat's a much better way and removes code to fix a bug which is good :)
06:59:54ddaltonsdoyon: can you help me with one thing?
07:00:13sdoyonjhMikeS: OK if I commit that?
07:01:00sdoyonddalton: what's that? Oh I've been meaning to ask: how useful is it to speak when the charger is inserted/removed?
07:01:06jhMikeSI think. You did test everything else right? :)
07:01:39jhMikeSIt is safe to force stop the pcm buffer even if not playing so I don't see a problem off hand
07:03:27ddaltonsdoyon: I guess not very but it is kind of nice to have so you know the player is charging.
07:03:33sdoyonjhMikeS: Well, it is 01:00 in the morning for me, and that's perhaps not the best hour to commit stuff :-). I'll just put that in my build and run with it for a few days. I bet it's good though.
07:03:44ddaltonnot very useful I guess what do you think?
07:04:31jhMikeSsdoyon: 1:04AM here :)
07:04:52ddaltonalso I am making the .talk clip delay configurable On 5 ms in the setting it is quicker than 1 ms. do you know how to make this work the other way round?
07:05:15sdoyonddalton: I'd probably want it in my build, but there's a bit of sensativity against adding stuff just because it's nice if it doesn't help substantially.
07:05:36ddaltonsdoyon: are you saying your considering using one of my patches? :-)
07:05:42sdoyonddalton: But it's kind of useful to be able to confirm whether there's power in your cable or not.
07:06:03sdoyonddalton: Well I might reformat it slightly :-).
07:06:05ddaltonyes that's what I thought. JdGordon and I argued about it for about 20 minutes.
07:06:21ddaltonsdoyon: ok what needs to be done?
07:06:52ddaltonsdoyon: do you know the answer to my question about the talk clip delay setting?
07:07:19sdoyonddalton: How did you get JdGordon and Pondlife so interested in that?! :-)
07:07:35 Quit Mouser_X (Nick collision from services.)
07:07:44sdoyonddalton: re .talk clip hover delay: the answer is to use FS #7774.
07:07:52 Join Mouser_X [0] (
07:08:05ddaltonsdoyon: does that make it configurable?
07:08:12sdoyonjhMikeS: Thanks.
07:08:50ddaltonwell jdgordon wasn't interested. But pondlife helped me because I asked him if he would and linusN gave me help as well
07:09:14*ddalton will be back
07:09:44sdoyonddalton: No. That's the list voice callback patch I've been asking you to test for weeks :-). It makes the delay unnecessary unless you're actually moving through items quickly.
07:09:54***Saving seen data "./dancer.seen"
07:10:37 Quit karashata ("Leaving.")
07:11:02 Join karashata [0] (
07:12:28 Part psycho_maniac
07:14:35*ddalton is back
07:15:00ddaltonsdoyon: now i understand and sorry...
07:15:31ddaltonSdoyon: you didn't like me implementing the voicing of disk size?
07:15:34ddaltonin p6138
07:15:48ddaltonit didn't seem like you did anyway
07:16:44ddaltonsdoyon: can you send me your most updated set?
07:16:53ddaltonSome of mine have been committed.
07:19:20ddaltonsdoyon: around?
07:20:12sdoyonddalton: re disk size: Same consideration: how useful is it really. That screen already takes forever to speak, and you're not likely to forget what your disk size is. Might be good for the SD card though.
07:21:19ddaltonsdoyon: I guess your right.
07:21:58sdoyonddalton: I'll send you another set soon, but it's quite late for me and I'm dead tired, and I'll regret it yet again this morning when my daughter will wake up at 06:00!
07:22:05*sdoyon is going to bed.
07:22:08 Quit sdoyon ("ircII EPIC4-2.4 -- Are we there yet?")
07:32:06 Quit jhulst (Remote closed the connection)
07:33:11 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
07:34:38 Join n17ikh|Lappy [0] (
07:36:18 Quit courtc (Read error: 104 (Connection reset by peer))
07:36:28 Quit Mouser_X (Read error: 104 (Connection reset by peer))
07:37:30 Join Mouser_X [0] (
07:38:05 Join courtc [0] (
07:40:01 Quit z35 (Read error: 110 (Connection timed out))
07:46:28 Quit Mouser_X (Read error: 104 (Connection reset by peer))
07:47:36 Join yoiuryuieroewyr [0] (
07:47:52 Quit n17ikh (Nick collision from services.)
07:47:58 Quit n17ikh|Lappy (Nick collision from services.)
07:47:59 Nick yoiuryuieroewyr is now known as n17ikh (
07:48:06 Nick n17ikh is now known as n17ikh|Lappy (
07:48:10 Quit n17ikh|Lappy (Client Quit)
07:48:28 Join n17ikh|Lappy [0] (
08:11:57ddaltonjhMikeS: around?
08:12:14amiconn[01:12:51] <jhMikeS> Only the PP targets seem to not have mains charging without adapters <== what did you mean?
08:13:18ddaltonOk do you know the problem you were helping sdoyon with before? well would making use of the action_none function work?
08:15:22jhMikeSamiconn: don't all the others have some sort of AC charging but PP tends to only come with USB charging unless you buy more equipment?
08:15:58amiconnThe H10 comes with an usb charger.
08:16:13ddaltonjhMikeS: it works good here but a few things need to be done: -I need to work out how the menu voicing code works -stop it from repeating -make the action_none delay shorter or make a new action
08:16:14 Join kubiix [0] (
08:16:16jhMikeSddalton: for what? in action.c? I'm not too familiar with action codec but can look.
08:17:04amiconnAnd the 1st and 2nd gen ipods came with a firewire charger, although they're perhaps not a fair example as they have no USB.
08:17:06jhMikeSamiconn: the H10 I have has USB charging but only a little plug-in on the side for an adapter. I don't know if I just didn't receive it.
08:18:36amiconnI dunno whether the small & big H10 are different in that respect
08:18:38jhMikeSddalton: you mean the voice stutter?
08:18:54ddaltonjhMikeS: Hang on I can upload my patch it is very messy. It still repeats it self when stopped and doesn't check for talk_menus I removed that when I didn't need to
08:19:04ddaltonand I need to do something about action_none.
08:19:19amiconnThe small one comes with an usb cable that has a dc socket on a short, secondary wire
08:19:20ddaltonI just added a call to the talk_menu function in the action_none if in menu.c
08:20:01amiconnMy H10 didn't come with a dc adapter for this socket, but with an usb charger (standard usb A socket)
08:20:05ddaltonjhMikeS: where if you are on bookmarks and you hold down down it might say "fi" "pl" and so on until you stop
08:20:07jhMikeSddalton: I think sdoyon just removed the pcm_playing() call in voice_on_voice_stop and that killed any started clip
08:20:33 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
08:20:56ddaltonjhMikeS: did he I thought he was still having trouble with it. Anyway I will ask him when he wakes up and if he wants he can see my patch
08:21:04jhMikeSamiconn: yeah. I'm saying the PP targets don't come with a wall adapter. none I know of anyway.
08:21:35jhMikeSddalton: We examined things further and it seemed a better and more thread-safe solution.
08:21:38karashatamine came wirh a wall adapter
08:21:44amiconnMy H10 did...
08:21:52*karashata nods
08:22:00*jhMikeS guesses he got swindled horribly :P
08:22:06amiconn...even though it doesn't plug into the socket I would expect
08:22:28jhMikeSWhat's the voltage and current rating on it? I probably have something handy.
08:22:42 Quit midgey ()
08:22:49ddaltonjhMikeS: Ok but it worked just was a bit annoying the delay but I guess that wouldn't be to hard to fix. Anyway I will just get rid of all of this stuff
08:22:55ddaltonmy stuff
08:23:04amiconnjhMikeS: Standard USB charger, i.e. 5V DC
08:23:45jhMikeSddalton: well, I suppose some sort of hover before playing any voice would actually be reasonable
08:24:01ddaltonjhMikeS: Have you got some time to help me with the "properties" plugin?
08:24:06jhMikeSlong enough not to trigger tons of clips and stop them when moving quickly
08:24:24jhMikeSddalton: that's petur's baby but I can try
08:25:03amiconnCurrent rating on this one is 1.2A but I would expect others to work as well
08:25:17ddaltonjhMikeS: Ok thanks basically I want to make it talk. Do you know what variables I should speak. I just want to find out: -file name -path -date created -date last modified -size and that is all
08:25:27*amiconn usually charges from the usb hub
08:25:29ddaltondoes it even display all of this?
08:25:46jhMikeSamiconn: My H120 adapter fits it
08:26:14amiconnjhMikeS: eh? The H120 charger isn't usb...
08:26:42jhMikeSddalton: it shows metadata too on music files
08:26:52 Join Mouser_X [0] (
08:27:05ddaltonjhMikeS: I guess that would be hard to do and the spelling would be annoying.
08:27:22jhMikeSamiconn: my connector is both USB and a charger dongle on the side with the usual plug
08:27:22ddaltonI have a better idea for metadata if it is possible
08:27:55amiconnjhMikeS: Yes, mine too, but the charger it came with plugs to the usb plug
08:28:13*karashata nods, is the same here
08:28:13Mouser_Xddalton: Rockbox compiled with the GBS patch (my laptop kept crashing, so it took awhile). I'm now attempting with the MOD patch (I removed the GBS patch).
08:28:48Mouser_X(Just in case you wondered.)
08:28:56amiconn...and iirc it's shown that way in the H10 manual. Nobody seems to know what thie dc socket on the cable is for, including iriver itself ;)
08:29:02jhMikeSamiconn: hmmm...I should see if it has diodes or pluggin both usb and wall charger wouldn't be safe
08:30:04spkycan someone please help me out, i'm trying to patch with customline #5900 and just cant get it to work ;[
08:30:09jhMikeSiRiver doesn't know why the DC socket is there? lol
08:30:40jhMikeSWell, it has the red charging led on it so I can only suppose....
08:31:28jhMikeSthen again it could be an output and not an input =)
08:31:37ddaltonjhMikeS : I found the line " rb->lcd_puts(0,1,dps->dirname);" so what is the variable that holds the directory name?
08:32:48jhMikeSddalton: it's in the typedef struct {} DSP;
08:33:29jhMikeSit's done that way so recursion doesn't use much stack space
08:33:58ddaltonjhMikeS: So when I use talk_file what should I give it what variable to speak?
08:34:15ddaltonbut wouldn't I need a talk_id to say "directory name" then speak the dir name
08:35:16jhMikeSsure: doesn't matter if it's displayed with "Director Name:"
08:36:33ddaltonjhMikeS: so is directory name or something displayed on the screen?
08:36:40jhMikeSthe thing is, that is rapidly updated during scan with the currently scanned directory.
08:36:40ddalton(I am blind that's why i am asking)
08:36:48jhMikeSddalton: yes
08:37:20ddaltonjhMikeS: So where does it display "directory name"?
08:37:23ddaltonin the code
08:37:27jhMikeSddalton: I'm aware of that. I'm looking at code I've never even seen before.
08:37:44amiconnjhMikeS: Hmm, I remembered wrong, the H10 manual shows an AC adapter connecting to that dc socket on the cable
08:37:49ddaltonok so do you know where I should start voicing things?
08:37:53amiconnBut mine came with an usb charger??!
08:37:56 Part karashata
08:37:57jhMikeSddalton: It just shows the path but no header
08:38:31n17ikh|Lappymy H10 (20gb) came with a cable that splits off into USB and a charging jack
08:38:33jhMikeSamiconn: I just charge from the connection with it turned off so it's all the same
08:38:34ddaltonjhMikeS: So for talk_file what should I give it as the dir to speak?
08:38:40n17ikh|Lappythe H10 will charge from USB
08:38:43n17ikh|Lappyor from the jack
08:38:43 Nick bb_ is now known as bb (n=bb@unaffiliated/bb)
08:38:44ddaltonso it plays the dirs .talk clip
08:38:50n17ikh|Lappyreally, the ac adapter is sort of pointless
08:39:03n17ikh|Lappyshould have just been a USB power adapter like the ipod has
08:39:32jhMikeSddalton: starting on line 208 is the final info
08:39:40amiconnjhMikeS: , unzips to a pdf
08:39:48amiconnPage 10 as adobe reader counts
08:40:10n17ikh|Lappyalso, jhMikeS: the power jack goes straight into the iriver
08:40:13n17ikh|Lappyseperate from the USB
08:40:23n17ikh|Lappywhere it's internally seperated
08:40:25jhMikeSddalton: you definitely don't want to put talking in the scanning code :)
08:40:43ddaltonyou don't
08:40:44n17ikh|Lappyand the power for the LED charging light on the jack comes back out of the iriver
08:41:00ddaltonso should I put my talking stuff on line 114?
08:41:16ddaltonor is that in the scanning code
08:42:06jhMikeSddalton: looking...hold on.
08:42:12ddaltonok thanks
08:43:04jhMikeSfile_properties and dir_properties both need it
08:43:50jhMikeSso yeah, 214 in dir properties
08:44:47 Join petur [0] (n=petur@rockbox/developer/petur)
08:44:49ddaltonso what variable do I speak
08:46:34ddaltonand what if it was a file?
08:47:22jhMikeSthere's str_dirname and str_filename in file_properties. you could speak both
08:48:26jhMikeSddalton: do you want the header included in the string or the path only?
08:48:27ddaltonok I will just try this before I go and write a whole lot of code. I want to make sure it will voice ok.
08:48:35*ddalton Needs quilt!
08:49:11jhMikeSvoicing if file_properties should probably go after line 119
08:50:52*jhMikeS apologizes for his terrible typing that just must sound awful
08:51:31amiconnEh? Did I miss something?
08:51:47amiconnProperties are shown in a plugin. How would that be voiced?
08:52:07amiconnThat needs some fundamental work regarding voice file structure and handling first
08:52:38ddaltonI know but I just want the properties to work for my own use. I might put it on the tracker if I get it to work :-)
08:53:09amiconnHalf baked solution :(
08:53:25amiconn(re fs #7704)
08:53:52 Join desowin [0] (
08:53:58*jhMikeS will check out using the H120 charger. What's H10 DC socket polarity?
08:54:45*jhMikeS likes the lack of any usual diagram on it :p
08:54:48ddaltonjhMikeS: what were you saying about line 119?
08:56:18jhMikeSddalton: I'm not even sure where it writes the stuff to the screen
08:56:49ddaltonok I will try this but it might keep talking
08:56:53jhMikeSah synclist
08:58:32jhMikeSline 119 is when it's done will obtaining and formatting the file's data
09:01:05jhMikeSddalton: it also lets you select the lines in the list so perhaps each line could talk when selected?
09:02:17jhMikeSthat could happen in get_props
09:02:19Avefyi, my problematic nano works 100% with the notorious 2 line diff against head revision
09:02:28Avebased on hour of testing
09:02:54jhMikeSAve: what diff would that be?
09:03:29AveB#7510 and related discussion
09:03:43LloreanAll it does is reduce it back to 78mhz instead of 80
09:03:45peturjhMikeS: the H10 5/6GB charger is center-positive
09:03:47LloreanIt's not a solution, it's a workaround
09:04:03LloreanIt doesn't fix the problem with timing, it goes back to when we weren't agitating the issue.
09:04:20jhMikeSLlorean: do they normally run at 80 on retailos?
09:04:24AveI'm content with the solution
09:04:30LloreanAve: It's not a solution.
09:04:43AveI'm content with the workaround then :P
09:04:46LloreanjhMikeS: We don't know, we just know that the chip is supposed to be specced to 80mhz, and the problem seems to be with ATA timing and the flash
09:04:55jhMikeSpetur: thanks. I think the H120 charger will work then.
09:05:00LloreanAve: We've known that would work since about the day the problem became evident.
09:05:12amiconnpetur: You have a charger with the dc plug?
09:05:25Avenice, but imo the workaround should be enabled until it gets sorted out for good
09:05:27LloreanAve: As a point of interest, for some people the problem shows up even at 78mhz
09:05:35Aveyeah I read the messages
09:05:40LloreanSo it shouldn't be enabled.
09:05:44LloreanIt's not a workaround for everyone anyway
09:05:55LloreanThe problem should be investigated and fixed.
09:07:05peturamiconn: yes
09:07:15peturit plugs into the cable
09:07:31peturbut the pin is much smaller than the h300 charger
09:07:49*jhMikeS wonders about the clockskipping during switching perhaps changing that. That's what H10 does when using the COP and scaling and the clockskipping avoids it.
09:07:52amiconnpetur: Interesting... mine came with an USB charger. But it's a used one, so this might not be the original one after all.....
09:08:29LloreanjhMikeS: changing "that" <−− what is "that" in that statement, I'm missing a noun to associate it iwth.
09:08:40amiconnpetur: The (coldfire) iriver chargers are strange. The H100 and H300 series chargers look exactly the same, but have different plugs
09:08:45peturamiconn: the connector that goes into the H10 has two cables: one USB and one that ha this DC socket
09:08:49jhMikeSLlorean: the FS #7510 stuff :)
09:08:56amiconnpetur: I know...
09:09:15LloreanjhMikeS: Okay, that's what I was guessing, but I wasn't sure.
09:09:57***Saving seen data "./dancer.seen"
09:10:08LloreanjhMikeS: If you have any ideas, there's plenty of people with problematic Nanos, among them Soap.
09:10:19 Join barrywardell [0] (
09:10:21LloreanI'm one of the few whose Nano shows no sign of problems whatsoever.
09:10:57jhMikeShmmm....doesn't connect property. just a little blink when inserting it. well, I've got a whole crate of wall warts.
09:11:55jhMikeSLlorean: I don't know it could change after I commit DC...can't promise that though. Maybe Soap could try the patch on his.
09:13:29jhMikeSMaybe I'll just make a patch with that alone to see.
09:13:59amiconnjhMikeS: Nano is PP5022 (well, PP5021)
09:14:11 Join pixelma [0] (i=pixelma@rockbox/staff/pixelma)
09:14:49jhMikeSamiconn: Llorean ran my DC patch on his without trouble (with SWP being used).
09:15:01LloreanI did?
09:15:26jhMikeSI believe you did like a month or more ago...different version but the PP5020 troubles don't seem to plague those
09:15:28LloreanThis wasn't recently, at least. Unless you've got me confused with someone else?
09:15:30LloreanAh, yes.
09:15:40LloreanI did back then
09:15:47LloreanI was thinking a more recent test of some sort
09:16:02jhMikeSas far as the scaling...well maybe PP still had that all buggy still
09:16:35jhMikeSIt's fine on PP5020 and PP5024. If it ran before on 5021 I think everything is fine.
09:20:29jhMikeSI'll post the clocking mod alone just to see if it clears up.
09:20:33pixelmaLlorean: reading the logs a bit - I _guess_ many of the Sansa users that talk about letting built the database and talk about "music" folders and whatnot actually mean the stupid database refresh of the OF - but don't tell...
09:20:45 Quit ompaul (Read error: 113 (No route to host))
09:22:43 Join n1s [0] (
09:22:49Lloreanpixelma: I think some of them may not realize that the "Files" option is really the filetree, and not a specialized menu or something
09:24:10pixelmaprobably as you can see the OF's database folders there
09:24:16barrywardellamiconn, jhMikeS: the H10 comes with a wall charger which should connect into the dongle. 5V, 2A, centre positive
09:29:05jhMikeSbarrywardell: I didn't get mine, but then I got a nice price so I can't complain. Just the player and USB cable.
09:29:54barrywardellit charges a good bit faster with the wall charger, but otherwise it's unnecessary
09:31:38jhMikeSthe H120 tip doesn't seem to quite make full contact on the +, otherwise it would be fine.
09:38:13 Part Llorean
09:38:39 Join Llorean [0] (
09:40:04jhMikeSwell, a patch is posted. I guess someone will be willing to try it out. :)
09:41:18 Join DraX [0] (n=alex@xmms2/developer/DraX)
09:43:24LloreanjhMikeS: Honestly, you might want to attach a build. I'm not wholly sure how many of those people are willing to learn to compile in order to help out.
09:44:31DraXi met some stellar rockbox guys at the gsoc mentor summit and promised I'd share my tagcache python implementation and xmms2 medialib -> rockbox syncing client, so if anyone is interested:;a=summary
09:44:57jhMikeSLlorean: how many builds?
09:45:12LloreanjhMikeS: Well, as it's only a Nano problem, just current SVN with your patch for Nano ought to be enough. :)
09:45:47LloreanFor the people who experience the problem, it's apparently consistent across builds, so all they should have to do is play audio for 3-5 minutes and see what happens.
09:46:20LloreanAnd thanks
09:46:58jhMikeSthe problems literally sound the same as the H10 had with COP and scaling so it really seems a reasonable check
09:47:12LloreanCorrupted reads from, and writes to, disk?
09:47:35jhMikeSscaling without the delays corrupted ram
09:47:57amiconnjhMikeS: Then it's most probably not the issue on nano
09:48:23amiconnOtherwise it should affect all PP5022/5024 targets
09:48:34jhMikeSyou said it's 5021 though
09:48:46LloreanIt could be a '21 quirk.
09:49:00LloreanI'm curious about the apparent heat-sensitivity of it though
09:49:08amiconnThe 5021 is most probably just a 5022 that's not specced up to the full 100MHz
09:49:10jhMikeSIt's possible they fixed SWP but not the clocking
09:49:23amiconnIt identifies itself as PP5022C
09:49:52jhMikeSSo does e200, right? does a nano have an AS3514 too?
09:50:00amiconn...and the PP5022 in the mini G2 identifies itself as PP5022B
09:50:28amiconnThe B should be older than the C, and the mini G2 has no trouble, not even at 96MHz (the highest I tried)
09:51:06jhMikeSno hurt to check this. anything I succeed on with PP is just trying stupid things and seeing if they work.
09:52:13amiconnWell, my experience is that if you need to do stupid things to make code work on PP, most probably something else is wrong
09:52:18jhMikeSI never would have found all the packed sample stuff being reasonable or the interrupts or whatever
09:52:45ddaltonamiconn: do you know how to install quilt to cygwin?
09:53:17pixelmaLlorean: and btw. yes chicago-12 and nimbus-13 (which is just the 12 pixels tall one with additional spacing) are quite similar but there are small differences here and there and the greatest difference is that chicago only had ascii characters and nimbus is more complete with a lot more characters
09:53:19jhMikeShah. Just look at's full of stupid things. SWP is also evidence of that coupled with the fact retailos has no SWP use.
09:53:21amiconnE.g. the workaround for freezing with clock scaling enabled on PP5020 used to be to re-enable the timer interrupts. Turned out that the problem was actually caused by improper handling of the clock setup
09:54:30jhMikeSso bugs definitely exist in some chips and not others regardless of how they id themselves. heck even whole codecs exist it seems.
09:55:26*amiconn doesn't even know what quilt is
09:55:43 Quit Mouser_X (Read error: 110 (Connection timed out))
09:55:44ddaltonits a package to manage lots of patches
09:56:25*jhMikeS gets some idea of why PP went bankrupt. Their design stall obviously cared more about 0xC0EDBABE's than solid hardware.
09:56:48amiconnjhMikeS: swp failing is probably a bus controller bug and so we need to work around it. I don't know of any other bugs except the cache bug on PP5002 (which doesn't affect stability, just speed)
09:58:22jhMikeSIt seems when using cop, it doesn't like immediate memory access when switching clocks either. The H10 fw seems to indicate that. perhaps Apple's design doesn't have the problem but skipping doesn't hurt other chips anyway and it only a couple uS more.
09:59:00jhMikeSit = pp5020 of couse
09:59:12amiconnI wonder how apple does it, then...
09:59:23jhMikeSdifferent RAM chips?
09:59:44amiconnI also wonder whether PP5002 also has that problem, and whether it also has the clocck skipping feature
10:00:28jhMikeSyou can at least sleep a core and do a udelay and reenable it. I'm sure it has it's own weirdness but probably conquerable.
10:00:30Avehow large percentage of the current supported devices are pp-based?
10:00:43jhMikeSAve: Way more than I'd like
10:00:58Aveand if that thing went belly up, does that mean that rb will die eventually due to no hardware avail
10:01:30Aveit doesnt exactly look like the hw markets are "opening up"
10:01:41AveDRM and NDA ever more
10:01:53LloreanRockbox will only die if people give up on it.
10:01:54jhMikeSnot will live on
10:02:21amiconnABout half of the targets are PP based
10:02:28Aveyeah I like to believe so. but its hard reverse-engineering stuff, true hackerdom
10:03:43amiconnjhMikeS: Maybe apple sleeps the cop during the whole clock switch, somewhere in the caller.
10:04:19amiconnI guess that would work too, instead of using the special skipping features?
10:04:20jhMikeSamcionn: for 500 uS? I only disable for about 2 at 30Mhz
10:04:56jhMikeS1) sleep other core 2) udelay(1) 3) wakeup other core
10:05:13amiconnI dunno. It looks like several OFs use a different approach for selecting cloks, leading to way less switching than in rockbox
10:06:43jhMikeSH10 uses no long udelay loops at all. Just a clock skip of a few uS.
10:07:20amiconnThey select a clock from a big table, based on what format is currently decoded, whether eq or other dsp features are enabled etc, and don't change the clock unless any of these conditions change
10:07:54jhMikeSSo it's tuned in advance with pre-canned tables? ick.
10:08:01amiconnThe drawback is lower efficiency than rockbox, as the clock selection has to include a safety margin
10:08:48amiconn...and less flexibility when adding cpu-intensive features
10:08:55jhMikeSIn rockbox, we may lower the clock frequency, but if it stays high, we end up sleeping the processor more often anyway. During the 30Mhz, it running 100% of that time.
10:09:45amiconnYes, but sleeping at high clock is less efficient than running it at lower clock
10:10:22jhMikeSPP's performance for using both cores is dismal. During playback on e200, there's huge lag in setting just the volume in retailos.
10:10:54Aveis rb using both cores atm?
10:10:55amiconnThis is because (1) the clock dependency of cpu current consumption is almost linear when not sleeping and (2) the current consumption in sleep mode is not zero
10:11:32jhMikeSAve: just for mpegplayer atm.
10:12:20jhMikeSI guess because the lower PLL frequency will help save some gate charging too.
10:15:55 Join sin613 [0] (i=d06b7b36@gateway/web/cgi-irc/
10:16:51sin613does the station scanning feature for models with fm radios use signal strength to locate stations?
10:17:03 Quit jhulst (No route to host)
10:18:26Aveoh yeah, regarding the sansa targets, are they actually useable? when my nano bites it I might have to go sansa way
10:18:47Aveeven tho I hate their wheel implementation...
10:18:52jhMikeSwe need a CPU loadmeter in the kernel :) should be pretty easy actually.
10:20:45BagderAve: the sansas are pretty neat
10:21:13LloreanAve: Have you tried the test build jhMikeS posted to the patch tracker?
10:21:24jhMikeSsin613: that's mostly it atm. the philips chip uses the IF counter. none of them really have as sophisticated an implementation as is needed.
10:21:51jhMikeSsin613: and doing it properly seems to be rather region dependent.
10:22:34sin613jhMikeS: would it be feasible to 'search' through the available spectrum for the lowest signal strength to determine what would be best to use for, say, an fm transmitter?
10:23:06AveLlorean: not yet
10:23:21jhMikeSyou could do that I suppose. finding nothing seems pretty easy.
10:23:35 Quit kubiix (Read error: 110 (Connection timed out))
10:23:44sin613but is finding the least nothing easy? or would it be the most nothing?
10:24:32jhMikeSlowest strength and most out of tolerance IF counter reading should be a hit then
10:25:32sin613is it possible to use fm through a plugin?
10:25:50sin613supposing the player supports it, naturally
10:25:57sin613(fm, that is)
10:26:14jhMikeSnot right now but putting radio_set/radio_get in there isn't hard
10:26:54jhMikeSthe API to power it up is already there
10:27:06sin613okay, i've got a project on my hands, i guess. i'm just so sick of switching through all the stations when i'm on a roadtrip trying to find something that's open
10:27:57Avenow running the mikes-build, no problems off the bat
10:27:57jhMikeSuse the audio_set_XXX_source routines for the input/output selection
10:30:03sin613thanks, jhMikeS
10:30:13 Part sin613
10:30:21jhMikeSAve: does it usually crash fast?
10:31:32Aveif I start "cold" I mean litereally a cold room temperature device I can do around 10-15 minutes with mine before it starts to glitch
10:32:07Avebut if the device is warm the problems manifest immediately
10:32:21Aveusually just a data abort few seconds into playback
10:33:30jhMikeSjust microwave it for a few seconds first :p
10:33:54LloreanHe's joking.
10:33:59LloreanThat is a very, very bad idea.
10:34:00Aveme too :P
10:34:10Avedebug shows rather frequent switches from 30 to 80 megs during vorbis playback
10:34:13Aveseems fine so far
10:35:09 Join SkinInd95 [0] (
10:35:26AveI think I better test this after reset and completely wiped .rockbox dir
10:37:50jhMikeSAve: it only affects the core firmware
10:38:08jhMikeSso no other change from current SVN
10:39:39barrywardelljhMikeS: quick arm asm question
10:40:12barrywardelldoes that wait for bit 0x40 to be set or cleared?
10:40:40jhMikeSargh...comp being slow...too much open crud
10:40:59jhMikeSto be set
10:41:10barrywardellok, then I had it backwards
10:41:23barrywardellI think I have the H10 ADC working properly now :)
10:41:32barrywardellwith ADC_STATUS working
10:43:51jhMikeSactually, see the gigabeat code. it never actually loops in the interrupt handler because the conversion that was started on the last tick is complete by the next tick. it just makes sure the last conversion started is done by testing that.
10:43:59 Quit desowin (Read error: 113 (No route to host))
10:44:15Aveah here we go
10:44:18AveI got problems
10:44:24Avebut its not so severe as before
10:44:58jhMikeSAce: what is happening?
10:45:01Avethe funny thing is, mp3 playback glitches, but vorbis playback skips or rather omits bits here and there
10:45:21Aveno data aborts so far, just little audio skips
10:45:50barrywardelljhMikeS: I have yet to change the code to the way the gigabeat does things
10:45:59LloreanAve: Check to make sure the audio files from your Nano play properly on your computer.
10:46:02barrywardellI've just got the status polling working
10:46:05AveLlorean: they do
10:46:27AveI've used these same files going back to r14004 build or what was it
10:46:33jhMikeSbarrywardell: that's of course a nescessary part of using the background method
10:46:38barrywardellit only has to loop around 3-4 times before it's done though, so is it really necessary?
10:48:09jhMikeSIt's that fast a conversion?
10:48:27barrywardellwell, occasionally up to 8 times, but yeah
10:49:39jhMikeSI dunno, it's cycles better spent decoding audio or something. *shrug*
10:49:55barrywardelltrue, 0 is better than 3-4
10:49:57amiconnSounds like a similar thing as the 1st and 2nd gen battery adcs
10:50:10barrywardellthe H10 adc is internal to the PP5020
10:50:29jhMikeSbarrywardell: but meg-fx only does one conversion per tick (round robin)
10:50:36barrywardellso maybe it's exactly the same?
10:50:51amiconnThey both use serial transfers (bitbanged). While the one used in the 1st gen is fast (needs just a few dozen cycles per bit), the one in the 2nd gen is dead slow
10:51:18 Join hcs [0] (n=agashlin@rockbox/contributor/hcs)
10:51:21amiconnbarrywardell: I don't mean polling for conversion, but for bit transfer to be ready. They're not internal
10:51:42amiconnOn 2nd gen, I use gpio interrupts in the driver
10:51:43jhMikeSThis of course only reads it once/second
10:52:28barrywardellamiconn: ah, I see
10:52:34amiconnjhMikeS: Yes, but the transfer can take several thousand cycles per bit on 2nd gen, and there are 8 bits...
10:52:46 Join linuxstb_ [0] (
10:53:06barrywardelljhMikeS: the H10 has to read it for the scrollpad, so sometimes a lot more often than that. will that cause problems with the gigabeat method?
10:53:09jhMikeSamiconn: well, that's that one so doing background with interrupts is very reasonable in that case.
10:53:12amiconnOn sh1 we also use adc interrupts
10:53:20 Quit linuxstb_ (Client Quit)
10:53:48jhMikeSbarrywardell: why would it? you won't get nested interrupts since they're disabled in the exception mode.
10:54:11barrywardellthe h10 buttons don't use interrupts anyway, yet
10:54:41jhMikeSthose would just be using the GPIO ones...same priority as the tick so I don't think so
10:54:47 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
10:55:21 Join bluebrother [0] (i=2mM3HGWQ@rockbox/staff/bluebrother)
10:55:24jhMikeSFIQ is allowed when in IRQ, none when in FIQ
10:55:32barrywardellbut the h10 doesn't use interrupts yet, so the reading wouldn't finish until the next tick? too slow for scrollpad?
10:55:42amiconnThe sh1 adc can convert 4 channels at once, and it has 8 in total (almost all of which we need). So the adc tick starts converting the first 4 channels and enables the interrupt. First time the interrupt fires, the 4 results are stored, and the conversion is started a second time for channels 5..8, enabling the interrupt again. Second time the interrupt fires, results 5..8 are stored
10:56:03amiconnSo we get all 8 adc values read once every tick, with very little overhead
10:56:16 Join mf0102 [0] (n=michi@
10:56:46 Quit BigBambi (Remote closed the connection)
10:57:17jhMikeSbarrywardell: you'll get an IRQ for the highest priority interrupt first, then other ones after you've serviced those. The arbiter determines that. If two are pending, you just get them in succession.
10:58:07 Join BigBambi [0] (n=alex@rockbox/staff/BigBambi)
10:58:26Avewrote a comment to #7510
10:58:30jhMikeSA GPIO button driver wouldn't make it wait. An easy thing to cook up quickly.
10:58:48barrywardellamiconn: if I could figure out the interrupt for the adc, then that method would work well
10:59:09barrywardelljhMikeS: i did add button interrupts for the h10, but the scrollpad caused problems back then
10:59:20jhMikeSI think the INT_STAT reg may tell you
10:59:23barrywardellmaybe now that the adc is working better i'll do it again
11:00:22barrywardelljhMikeS: I'll check that
11:00:40 Quit SkinInd95 ("ChatZilla [Firefox]")
11:01:28jhMikeSThe buttons are just GPIO lines on GPIOA and B so shouldn't have anything to do with the scrollpad which I'd use the GPIOD interrupt to enable the scan during tick I suppose.
11:01:42 Join costas_c [0] (
11:01:52barrywardellyeah, I did that. the problem was with jumpy readings from the ADC
11:02:38jhMikeSI guess a few loops is better than udelay(250) :P
11:05:16*jhMikeS still wants to try to make a clockskipping udelay instead of a cycle-burner one
11:06:35barrywardellhmmm. INT_STAT stays fixed at 0x00900000, although maybe I should check it when an interrupt happens?
11:07:50 Join FOAD [0] (
11:07:55 Part costas_c
11:08:00 Quit FOAD (Client Quit)
11:08:41 Join FOAD [0] (
11:08:53 Quit FOAD (Client Quit)
11:09:06barrywardelljhMikeS: is it ok with you if I rename DEV_OFF_MASK to DEV_RS2
11:09:30 Join FOAD [0] (
11:09:39jhMikeSI think that's ok. Don't forget DEV_EN2 as well.
11:09:58barrywardellyeah, I've got that
11:10:00***Saving seen data "./dancer.seen"
11:10:08amiconnI wonder why the PP502x seems so much more complex than the 5002
11:10:33barrywardellany suggestions for what to call 0x70000010?
11:10:40jhMikeSlater design? they made sure they used all their patents in the design?
11:10:53amiconnThe 5002 only has one DEV_EN, where just 16 bits are used. The 502x obviously has 2, using all 32 bits each...
11:11:02 Quit FOAD (Client Quit)
11:11:47 Join Thundercloud [0] (
11:11:47jhMikeSbarrywardell: I'm not sure what that does, functional/vs. GPIO pin or internal vs. internal connection.
11:11:51 Join FOAD [0] (
11:11:51amiconnIn other places, the design seems very similar, with many controllers just moved around in the address space
11:11:58jhMikeS*internal vs. external
11:12:28 Quit Nico_P (Remote closed the connection)
11:12:46barrywardelland also 0x60006094?
11:13:02jhMikeSThen setting up PP5002 for DC shouldn't be that terrible. Easy to select which mutex method to use (one #define).
11:13:13 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
11:13:17jhMikeSbarrywardell: ADC_CLK_DIV?
11:13:19barrywardellI currently have them as ADC_ENABLE_ADDR and ADC_CLOCK_SRC
11:13:24 Join desowin [0] (
11:13:31amiconnjhMikeS: Cache flush/invalidate needs to be researched
11:13:36barrywardellbut they're both used for more than ADC
11:13:53jhMikeS60006094 is?
11:14:09barrywardelli thought you said is was used in the sansa of?
11:15:02jhMikeSfor something or maybe nothing. it came from an emu dump which thinks it's running on something else.
11:15:26barrywardellah, ok. i'll leave them with the adc names for now
11:15:59jhMikeSI only found the second GPIO bit by poking. The OF running in the emu ignores it when obviously it must use it running natively.
11:18:13amiconnbarrywardell: Did you also check HI_INT_STAT?
11:18:13 Quit idnar (Nick collision from services.)
11:18:15jhMikeSThe emu must not set some HW revision properly.
11:18:16 Join idnar_ [0] (i=mithrand@unaffiliated/idnar)
11:18:25barrywardellamiconn: no, I'll check now...
11:18:41 Join ender` [0] (
11:19:09 Quit FOAD ("I'll be back")
11:19:38amiconnbarrywardell: Also, iirc it's necessary to clear interrupts, otherwise the bit in INT_STAT / HI_INT_STAT stays forever
11:19:39*jhMikeS likes that the e200 lcd driver design allows all kinds splashing during ISRs without crashing anything.
11:20:43 Join lazka [0] (
11:20:48barrywardellHI_INT_STAT is 0x00001000, so there's another possibility
11:21:29amiconnbarrywardell: ..meaning that since in your reported INT_STAT, bits 20 and 23 are set. Bit 23 is IDE, so it could be bit 20
11:21:40jhMikeSI guess it's as simple as poking each 1 bit
11:21:46jhMikeSbit 20 is USB
11:22:10jhMikeSZagor got success earlier and it connected
11:22:38amiconnSo just bit 44 remains...
11:23:06amiconn...unless the adc interrupt needs to be enabled explicitly in the adc config reg(s)
11:24:23*jhMikeS was just going to mention that I suppose setting all the bits OF does and checking any change in INT_STATs could do the trick.
11:24:51barrywardellI don't see any enabling in the adc init in the of
11:25:16 Join FOAD [0] (
11:25:40 Join bertrik [0] (
11:26:23*jhMikeS wonders how the interrupt gets acked now
11:26:39barrywardellno interrupt enabling
11:27:32jhMikeSI mean does reading the result ack it similarly to filling the IIS FIFO acks the I2S interrupt or reading the timer value acks the timer. Seems typical PP design.
11:28:59jhMikeSConstrast that with explicit registers that must be written with '1' on gigabeat
11:30:02amiconnGPIO interrupts need to be acknowledged explicitly on PP
11:31:05 Join hannesd [0] (
11:31:46jhMikeSthat's rather atypical for PP though. GPIOx_INT_CLR just acks, not actually masks it like CPU_INT_CLR
11:32:51jhMikeSCPU/COP_(HI_)INT_CLR should probably be renamed since they don't do that
11:33:55barrywardellthat jumps to 0x4e774 if the bit is not set?
11:34:59jhMikeSthat's the relock bit
11:35:37jhMikeShas to be
11:35:52barrywardellthe adc increases it's reading by 0x14 if that bit is set
11:36:09 Quit hannesd ("Client suicide")
11:36:11barrywardellit's pll enable
11:36:12jhMikeSit's checking if the PLL is enabled
11:36:52barrywardellso if PLL is enabled, jump to 0x4e774, or the opposite?
11:37:05*barrywardell neads to understand arm asm better
11:37:16jhMikeSjump if disabled
11:37:38jhMikeStst sets the zero flag if the result of the and is zero
11:37:43barrywardellcool, thanks
11:38:40 Join JdGordon [0] (
11:39:10 Quit hcs ("Leaving.")
11:40:09jhMikeSthat reminds me I could have used "tst r12, %[mask]" in the FIQ handler instead of "ands r12, r12, %[mask]" :P
11:44:22Nico_Phmm my gigabeat'c CPU frequencey is 17 MHz according to the debug screen. That can't be right
11:45:46jhMikeSit's fast enough at that speed :)
11:47:18*jhMikeS reminded of wanting to try a very slight overclocking experiment there
11:48:04amiconnNico_P: That's because cpu scaling is disabled on gigabeat, so it uses the #define in config-gigabeat.h, which is wrong
11:48:05Nico_PjhMikeS: problem is in system-target.h for the gigabeat, CPUFREQ_DEFAULT == CPUFREQ_NORMAL == 98784000
11:48:24amiconnI guess this frequency is actually the base frequency when scaling is enabled
11:48:29 Nick parafin|away is now known as parafin (
11:49:09jhMikeSthe 17MHz is just the appromate frequency of the X-tal. 16.9384 or whatever.
11:49:23 Join merbanan [0] (n=banan@
11:49:30amiconnYes. Debug menu rounds towards full MHz
11:49:42Nico_Pwhat's the X-tal ?
11:49:56jhMikeSthe crystal clock source
11:50:32Nico_Pah I see the #define CPU_FREQ 16934400 in config-gigabeat.h... but then what's the acutal freq ? surely not 17 MHz
11:50:57jhMikeSthe crystal is 16934400 Hz
11:51:55jhMikeS16934400/44100 = 384 = DAC OS rate
11:53:16Nico_Pthat doesn't mean anything to me :/
11:54:31 Quit gromit` (Read error: 104 (Connection reset by peer))
11:54:50jhMikeSthe DAC uses that X-tal as its clock source directly. Oversampling is 384 times that.
11:54:56 Join davina [0] (
11:56:20 Join gromit` [0] (
11:57:24 Quit JdGordon ("Konversation terminated!")
11:58:38Avehey mike, is there a way I can easily test using "longer clock skips", does that mean increasing the time slept/delayed between changing the pll values?
11:59:02 Join JdGordon [0] (
11:59:31jhMikeSAve yes.
11:59:34 Quit gromit` (Client Quit)
12:00:20jhMikeSwhere is says 0x480000xx you can go up to 0x480000FF
12:02:33 Quit Nico_P (Read error: 113 (No route to host))
12:03:08 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
12:03:47 Join obo [0] (n=obo@rockbox/developer/obo)
12:05:20jhMikeSAve: I suppose just max out the delays to start. If that won't do it, then other things would need to be found.
12:05:53 Join kugel [0] (
12:06:18 Quit Nico_P (Remote closed the connection)
12:08:58 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
12:11:27 Join gromit` [0] (
12:13:05*jhMikeS goes and gets some zzzzz's
12:13:43 Join jba_ [0] (
12:23:05n1swake up alarm works for ipod now, right?
12:26:51 Quit Gnu47|LandOfNod ("You know you'll miss me :P")
12:28:37 Join homielowe [0] (
12:30:38linuxstbn1s: It should do, yes.
12:30:57 Quit JdGordon (Read error: 104 (Connection reset by peer))
12:31:24peturargh... what's up with the forum?
12:31:41n1sgood, I closed some tracker task with a plugin that would pause playback and wait for a set time and resume, It was described as awork-around for proper wake-up so :-)
12:31:52linuxstbpetur: Works fine for me...
12:32:21peturyes now it does, 1 minute ago it only loaded half, and the post I made is just gone :(
12:32:42*petur posts again
12:33:06 Join JdGordon [0] (
12:34:00*petur copies text before posting
12:34:14peturbah, forum gone again :(
12:35:14peturlinuxstb: does it still work for you?
12:35:26peturcrap, back again
12:35:55peturha... managed to post after 3 tries :/
12:37:16 Join Frazz [0] (
12:38:44 Quit desowin ("use linux")
12:39:24 Join lee-qid [0] (
12:42:21 Nick fxb__ is now known as fxb (
12:42:58pixelmapetur: couldn't you get your previos attempts back with the help of your browser cache? I often do because many times it happens to me that the forum sessions times out when I want to post (or get the preview)
12:44:12peturmaybe firefox does it better, I happened to use IE now... and the forum timed out and when trying to get back it gave me a session timeout error and a blank form
12:45:22pixelmaah ok, maybe that would be different here too
12:45:59peturanyway, I copied the text before hitting send the second time ;)
12:47:41n1swhat!, there's another pause-and-wait plugin in the tracker, sight
12:48:19 Quit homielowe (Read error: 110 (Connection timed out))
12:48:42n1samiconn: a hacky alarm based on pausing the music and simply waiting while the player is still on
12:50:22n1ssince it apparently requires an rtc I see no point, reject? or should we leave it with all the other strange hacks that will never be comitted?
12:53:45JdGordonnot all the rtcs have alarms..
12:54:04amiconnAfaik they have
12:54:15amiconnJust that rockbox doesn't make use of them
12:54:35pixelmaisn't there an "alarm mod" for the recorders?
12:54:38amiconnOnly on recorder v1 it requires a hw mod to use the rtc alarm
12:54:53JdGordonalso, i tihnk at east on of those alarm hacks make use of the tick counter to fake a time, so it wakes up in x min from now instead of a time
12:56:47Nico_Ppixelma: I've made some changes to the buffering, would you have time to try them out ?
12:58:59pixelmaif someone would be so kind as to build it for me
12:59:54Nico_PI can. M5 ?
13:03:08Nico_Pit still releases tracks as soon as they're finished playing, but it should spin the disk less often and hopefully make the most of each time it spins
13:08:24 Quit Zom ("leaving")
13:08:38 Join Zom [0] (
13:08:43barrywardellhmmm. I think we'll need a new battery_bench with this new ADC code
13:09:08barrywardellcan anyone run one on a small H10?
13:10:03***Saving seen data "./dancer.seen"
13:10:37 Join Lear [0] (i=chatzill@rockbox/developer/lear)
13:13:47 Join Rob2222 [0] (
13:14:32 Join rvvs89_ [0] (n=rvvs89@unaffiliated/rvvs89)
13:15:14 Quit rvvs89 (Nick collision from services.)
13:15:21 Nick rvvs89_ is now known as rvvs89 (n=rvvs89@unaffiliated/rvvs89)
13:21:30 Join spiorf [0] (
13:21:31 Quit obo ("bye")
13:33:50amiconnbarrywardell: I could start one later, just before leaving
13:34:02amiconnDon't forget to deactivate low-bat shutdown temporarily
13:34:12barrywardellthanks, this patch needs to be applied first:
13:35:24barrywardellthe problem is that we have to artificially increase the value from the adc when pll is enabled
13:36:37pixelmaNico_P: seems to fill the buffer completely now, spun of the disk for rebuffering after the expected time and takes longer :)
13:36:39amiconnI doubt that's correct
13:37:36amiconnMaybe the OF does this for battery voltage reading in order to compensate for the higher load when running at higher frequencies
13:37:37barrywardellthat's what the of does
13:38:03barrywardelland turning off pll does indeed change the reading by ~0x14
13:39:03amiconnFor battery?
13:39:37Nico_Ppixelma: so better than before, and how is it compared to SVN ?
13:40:01amiconnSo it's probably just the compensation. Did you try different CPU clocks, all using the pll?
13:40:14Nico_Pwhat I was thinking is that it could be possible to have different buffering strategies and let the user choose, depending on his usage patterns
13:40:44barrywardellit's easy to check - check the value in the i/o ports screen, then change frequency to 24MHz (default, no pll) in the cpu freq screen, the check i/o ports again
13:41:52amiconnDoes it change between 30 and 80 MHz?
13:41:52rasherNico_P: I filed a request for that once..
13:42:42pixelmaNico_P: I _think_ it's the same as svn though haven't compered exactly yet but from the bitrate of the tracks I'm listening to I'd expect the refill of the buffer at the same amount of tracks as it was with your build. Should I compare to svn more exactly?
13:43:00barrywardellamiconn: it's slightly (~4) lower at 30 than at 80
13:43:08rasherNico_P: closed as "invalid" by JdGordon for som reason.. FS #3008
13:43:51barrywardelland much higher (~0x24) at 24 (pll off)
13:44:12amiconnIt becomes lower with lower clok, but higher when disabling the pll?
13:44:34amiconnIs it an offset or a correction factor?
13:44:38barrywardellthe of just checks for pll
13:44:51barrywardellI'm not sure, but in the OF it's just an offset
13:45:05Nico_Prasher: ah, so I'm not the first to have thought about it :) It should be rather easy now that buffering is well separated
13:45:07Nico_Pnot one of my primary objectives though
13:45:11amiconnEven more weird... unless it's adc calibration
13:45:25JdGordonwhat am i being blamed for this time?
13:45:27amiconnWhat about the other adc channels?
13:45:44rasherJdGordon: you closed my tracker entry with a rather strange reason
13:45:56 Quit Nico_P (Broken pipe)
13:46:07 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
13:46:09rasher(invalid rather than say, rejected)
13:46:46*JdGordon can change it to rejected if youd prefer that?
13:47:19barrywardellamiconn: I'm checking now...
13:47:53*Nico_P reads the log and noticed he missed several messages
13:48:11rasherJdGordon: well so could I.. nothing important of course, I just noticed it
13:48:21Nico_Ppixelma: no need for an exact comparison
13:48:59 Join webguest09324 [0] (i=29e1f734@gateway/web/cgi-irc/
13:49:00n1slinuxstb: did you kill arm-elf-gcc on your server?
13:49:30 Join desowin [0] (
13:49:45webguest09324Can I record with rockbox on my ipod 2g?
13:49:47Nico_PJdGordon, rasher: that request could be reopened soon
13:49:49webguest09324I mean 3g
13:49:54 Quit webguest09324 (Client Quit)
13:50:12JdGordonNico_P: really?
13:50:39Nico_PJdGordon: it would be quite doable with the buffering code
13:50:49JdGordonif you say so...
13:50:57Nico_Pand it would be a nice option for the skippers
13:52:03 Quit jhMikeS (Nick collision from services.)
13:52:09 Join jhMikeS [0] (n=jethead7@rockbox/developer/jhMikeS)
13:52:18JdGordonbarrywardell: have you seen 7913? needs a osx tester
13:52:32barrywardellamiconn: it's hard to tell, but I think it's only the battery that's affected
13:52:38barrywardellJdGordon: haven't seen that yet
13:53:01rasherNico_P: Probably needs to be slightly smarter than what I described, but yeah, for skippers, the current strategy isn't terribly convenient
13:53:56JdGordonhow could it work?
13:54:40JdGordonx seconds of a track could be any random byte value of the file...
13:55:09JdGordonand then you need to worry about making sure there is enough room to buffer the whole track if a skip doesnt happen
13:58:18linuxstbn1s: Yes, sorry it was broken temporarily, unfortunately during a build. It's OK again now.
13:58:23 Quit Nico_P (Remote closed the connection)
13:58:41n1slinuxstb: just checking ;-)
13:58:45 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
13:59:20linuxstbn1s: Do you know about gcc and multilib?
13:59:34n1snot really, no
13:59:43linuxstbOK, I'll post to the dev list...
14:00:25linuxstbIt seems we all need to rebuild arm-elf-gcc to support armv6 (Gigabeat S) and armv5 (Telechips)
14:01:01 Quit Nico_P (Remote closed the connection)
14:01:15GodEaterlinuxstb: how do we go about that ?
14:01:22 Join Crash91 [0] (n=evil91@
14:01:34n1slinuxstb: doesn't the gigabeat s people use 4.1?
14:01:58linuxstbGodEater: I'll post to the mailing list. You need to edit a file in the gcc source to build different versions of libgcc - it's already done for thumb (that's what the thumb patch does).
14:02:08Crash91can anyone help me set up a rockbox dev evnironment on linux? a wiki page or something would be useful
14:02:42GodEaterlinuxstb: isn't this something we should do in ?
14:02:47linuxstbGodEater: Of course.
14:03:12GodEaterCrash91: have you looked in our wiki at all yet ? There *is* a search feature...
14:03:29Crash91oh yeah....thats what that box does!
14:03:41barrywardellamiconn: I guess we could just ignore the problem since we always run with pll on?
14:03:51linuxstbGodEater: already downloads and applies a patch to enable a thumb version of libgcc to be built, so the solution will probably be to just modify that patch to enable the other required versions as well.
14:03:53JdGordonany objections to ?
14:03:57amiconnbarrywardell: Before testing that thoroughly, we need to find out whether it's an offset or a factor
14:04:17GodEaterlinuxstb: perfect - let me know when that's in and I'll update my build server :)
14:04:20amiconnComparing the change with almost empty battery and then with full battery should tell
14:04:56 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
14:05:13*Crash91 pokes bertrik
14:05:45linuxstbn1s: I just assumed the S people used the same as every other ARM target. But tools/configure isn't specifying the correct CPU - it's using arm9tdmi (same as the F - armv4), rather than arm1136jf-s (armv6).
14:06:30n1siirc they had some problem with 4.0 but I'm not sure
14:06:41linuxstbMaybe it was this problem.
14:06:54barrywardellok, it's fully charged now and the difference is 0x14, maybe a little more
14:07:01barrywardelltime to run my battery down
14:07:07Nico_PJdGordon: sorry for disappearing... we could buffer eg the first half of each track
14:07:13GodEaterarm1136jf-s just trips off the tongue doesn't it /
14:07:40 Quit zicho (Read error: 110 (Connection timed out))
14:07:41JdGordonNico_P: really doesnt sound like its worth the hassle...
14:07:57GodEaterNico_P: had some odd moments with my MoB build last night
14:08:13Nico_Pmaybe not buffering only part of a track, but maybe spinning the dik a little more often
14:08:14rasherNico_P: I'd say less. Perhaps closer to 20%
14:08:16GodEatergot complete silence in the middle of a track a fair way into my play list
14:08:29Nico_PGodEater: strange
14:08:36rasherNico_P: careful with your typos..
14:08:37GodEaterhappened a couple of times
14:09:26Nico_PGodEater: did audio come back or did you need to stop playback for it to work again ?
14:09:43GodEaterI had to stop playback
14:09:52GodEateror more accurately
14:09:55GodEaterskip to the next track
14:10:08GodEaterit never resumed in the one I was listening to
14:11:27Nico_PGodEater: do you remember what's the latest commit on your build ? I think I might have fixed something related... also was the disk starting to spin when the stalling occured ?
14:12:06GodEaterNico_P: don't know to either question I'm afraid
14:12:11GodEaterI can maybe find the commit though
14:12:15GodEatergimme a minute
14:12:21Nico_PGodEater: usong git ?
14:13:58Nico_PGodEater: git log then
14:14:19Nico_Pbtw, I rebased recently so pulling won't do
14:14:20GodEateryeah - had to see if I could still login to my work PC though which is where I built it :)
14:14:36*Crash91 curses his stupid ISP/DNS for not being able to load rockbox directly
14:14:41 Join zicho [0] (
14:14:43GodEaterFri Oct 12th, 00:12:44 - "Fix warning"
14:14:59pixelmaGodEater, Nico_P: I had that with the last build I tried on thursday (not today), couldn't figure out how to reproduce
14:15:16GodEaterno I can't make it happen on demand either
14:15:34GodEatercan give you the hash too if you want Nico_P
14:15:47Nico_PGodEater: thanks but no need. I think it may be fixed but I'm not sure... I also added a debug screen
14:16:25GodEaterI'll update then
14:16:33Nico_Pand improved the buffering strategy too
14:17:03Nico_PGodEater: to update I suggest deleting your mob branch and checking it out again. pulling will most likely fail with conflicts
14:19:44*GodEater hunts through the logs to find out how to pull the mob branch again
14:19:48Crash91so do i just svn co and then do ./ is in the tools folder, but otherwise, yes.
14:20:27Nico_PGodEater: git checkout master; git branch -D mob; git checkout -b mob origin/mob
14:21:22Crash91ok thanks,
14:21:36GodEaterNico_P: found it anyway :)
14:21:57Nico_Pbah I typed for nothing ;)
14:22:45 Quit Crash91 ("Bye Bye!")
14:22:48GodEaterI only slightly misremebered it
14:22:59GodEaterI'd done git branch mob origin/mob
14:23:12GodEaterwhich would have worked too if I'd then switched to it and pulled I think
14:24:31Nico_Pno need to pull after doing the checkout
14:25:50GodEaterindeed not
14:25:55GodEateryour command is quicker :)
14:26:27*bertrik pokes crash91
14:28:42SoapjhMikeS, amiconn: either one of you are welcome to my Nano, or I could drive home today and pick it up if you have tests you want performed.
14:34:04Nico_PGodEater: I found how to pull from a rebased remote branch without conflicting: "git pull origin +mob:mob", where the first "mob" is the remote one and the second is the local one (which can have another name)
14:35:49 Join Wiwie [0] (
14:40:35 Quit Lear ("ChatZilla [Firefox]")
14:41:07 Join Arathis [0] (
14:42:37amiconnHmm, the sim doesn't simulate lcd_yuv_blit() properly
14:42:59amiconnlcd_yuv_blit() should be non-destructive to the framebuffer. In the sim it isn't.
14:46:23 Quit zicho (Read error: 110 (Connection timed out))
14:46:42 Nick fxb is now known as fxb__ (
14:48:25amiconnSoap: hmmmm....
14:55:25 Join DogBoy [0] (
14:58:57 Quit Wiwie (Read error: 110 (Connection timed out))
14:59:02JdGordonzzzzzzzzzzzzzz booooooooooored
15:03:57 Quit petur ("more shopping....")
15:05:41 Part pixelma
15:10:05***Saving seen data "./dancer.seen"
15:11:23 Quit ddalton ()
15:13:18 Join Rob222241 [0] (
15:15:38 Quit spiorf (Remote closed the connection)
15:22:48 Quit BigBambi (Remote closed the connection)
15:23:58 Quit tumu ("~")
15:26:11 Quit bertrik ("bye")
15:26:34 Join bertrik [0] (
15:29:53 Join bryan__ [0] (
15:30:10 Quit Toxicity999 (Nick collision from services.)
15:30:16 Nick bryan__ is now known as Toxicity999 (
15:30:39 Join ompaul [0] (n=ompaul@freenode/staff/gnewsense.ompaul)
15:31:51 Quit Rob2222 (Read error: 113 (No route to host))
15:34:45 Join df___ [0] (
15:42:38 Quit jba_ (Read error: 110 (Connection timed out))
15:45:41 Join bryan__ [0] (
15:46:00 Quit Toxicity999 (Nick collision from services.)
15:46:06 Nick bryan__ is now known as Toxicity999 (
15:51:55 Join Robin0800 [0] (
15:52:51 Join advcomp2019_ [0] (n=advcomp2@
16:04:36 Join BigBambi [0] (n=Alex@rockbox/staff/BigBambi)
16:05:25 Nick idnar_ is now known as idnar (i=mithrand@unaffiliated/idnar)
16:06:45 Join spiorf [0] (
16:06:47 Quit advcomp2019 (Nick collision from services.)
16:06:51 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@
16:06:54 Quit spiorf (Read error: 104 (Connection reset by peer))
16:07:42 Quit Frazz ("Leaving")
16:10:54 Join spiorf [0] (
16:17:22 Nick fxb__ is now known as fxb (
16:17:31 Nick fxb is now known as fxb__ (
16:20:17 Join midgey [0] (
16:25:52 Nick kugel is now known as laurence (
16:25:54 Nick fxb__ is now known as fxb (
16:26:56 Nick laurence is now known as morpheus (
16:27:11 Nick morpheus is now known as kugel|afk (
16:27:42 Nick kugel|afk is now known as kugel (
16:28:17 Nick kugel is now known as kugel|afk (
16:31:08 Quit Rob222241 (Read error: 104 (Connection reset by peer))
16:31:30 Nick kugel|afk is now known as kugel (
16:31:55 Quit GodEater_ (Read error: 110 (Connection timed out))
16:33:32 Quit donutman25 ("ChatZilla [Firefox]")
16:36:54 Part kugel
16:37:17 Join kugel [0] (i=kugel@unaffiliated/kugel)
16:37:25 Join kkurbjun [0] (
16:42:45 Quit midgey ()
16:51:54 Quit qwm (Remote closed the connection)
16:52:05 Join qwm [0] (
16:52:17 Join Rob2222 [0] (
16:58:58 Join daurn|iphone [0] (n=daurn@
17:02:14 Quit billytwowilly (Read error: 113 (No route to host))
17:10:09***Saving seen data "./dancer.seen"
17:12:27 Quit Rob2222 (Read error: 104 (Connection reset by peer))
17:17:40 Join Rob2222 [0] (
17:19:48 Join Frazz [0] (
17:25:38 Join marc_ [0] (
17:27:56 Quit JdGordon ("Konversation terminated!")
17:28:59barrywardellbluebrother: I tried your rbutilqt mac os x patch, but i get compile errors
17:29:40 Join Klevi [0] (
17:29:59bluebrotherbarrywardell: is it related to the patch or is it problems with libusb?
17:30:22barrywardellnot libusb, I got that working...eventually!
17:30:32 Quit Arathis ("Bye, bye")
17:30:34barrywardellI can build rbutilqt fine without the patch
17:30:49bluebrotherhmm. Any idea how to fix the problems?
17:30:56 Quit daurn|iphone (Read error: 110 (Connection timed out))
17:32:07barrywardellnot really, I haven't looked at the code
17:33:42barrywardellwhere does getmntinfo come from?
17:33:59bluebrotherfrom BSD.
17:34:14KleviIf recovery mode fails, is there another way to reflash a sansa c200?
17:35:11barrywardellah, missing include
17:35:35barrywardellnow it just complains about struct statfs
17:36:10*bluebrother slaps head
17:36:52bluebrotherI did a short test of getmntinfo on bsd (but without Qt due to space reasons). How could I forget the includes part?
17:37:42barrywardellwhat should mntoname and f_mnfromname be?
17:38:20 Join zicho [0] (
17:38:43bluebrotherargh, another typo. it's f_mntoname, missed the f_
17:39:15barrywardellthat's what I suspected
17:39:20 Join pfuender [0] (i=5288762a@gateway/web/cgi-irc/
17:39:28barrywardellnow it builds:)
17:39:32KleviIf recovery mode fails, is there another way to reflash a sansa c200?
17:39:51barrywardelldetected my H10 :)
17:39:59bluebrothercool :)
17:40:00pfuenderany clue why the recording with the internal mic not works on the iAudio X5 (60gb)?
17:40:18pfuenderI can't hear anything, but i can choose the microphone as the source
17:40:33 Join alienbiker99 [0] (
17:40:35barrywardellbluebrother: should I post a patch? or just commit the change?
17:40:37kugelKlevi: on the e200 there is also a manufacture mode, dunno if it exists for c200
17:40:45bluebrotherbarrywardell: feel free to just commit it.
17:40:46kugelask google
17:41:02barrywardellbluebrother: there were some changes necessary to the pro file too...
17:41:11KleviIve tried, it doesnt work. Holding record while turning it on with the hold switch
17:41:27kugelisnt this recovery mode?
17:42:34 Join ToHellWithGA [0] (
17:42:37barrywardellthe /usr/local stuff is just because that's where I installed libusb
17:42:39KleviThe only thing that google pulls up is the e200tool, (if i look for manufacturers mode)
17:42:47barrywardellbut the framework bit is necessary
17:42:54bluebrotherah, ok.
17:43:05kugelsure, the e200tool is the only tool that can access the manufacture mode
17:43:09barrywardelland I couldn't build for ppc because I don't have a universal binary of libusb
17:43:33bluebrotherwell, it's nice to have it building at all.
17:43:44Klevibut my sansa isnt recognized by windows or linux as a disk at all
17:43:53 Join raphi [0] (
17:43:56KleviWindows says its a USB device, and nothing more
17:44:01kugelthats fine
17:44:18kugelthat's normal when you are in the manufacture mode
17:44:18 Nick raphi is now known as pfuende1 (
17:44:19KleviHow would I use the 3200 tool under cygwin?
17:44:26 Quit pfuender ("CGI:IRC (EOF)")
17:44:37 Nick pfuende1 is now known as raphi (
17:44:40 Nick raphi is now known as pfuender (
17:44:57kugele200tool isnt made for windows
17:45:18Kleviif i run linux, (sabayon mini ed. 3.3) I lack an internet connection
17:45:42pfuenderanyone got an idea? recording for iAudio X5 doesn't work......$
17:45:57Kleviwireless and linux dont go hand in hand too well for weak wifi signals
17:45:59kugelsome people reported they got e200tool to work under win
17:46:10Klevibut you have no idea how to set that up?
17:46:43Klevithough, your saying that it should work, the current state my sansa is in, using the e200 tool?
17:47:20kugelim not familiar with the e200tool
17:47:26KleviThats okay.
17:47:38KleviBack to google I go!
17:47:38kugelespecially not when combining it with a non-e200
17:48:13kugeleven if you get it to recognize, i doubt it has a file to flash a c200
17:48:41Kleviif i do
17:49:00Kleviwouldnt the sansa updater get the original firmware working?
17:49:38kugelthat would require the c200 being in the recover mode
17:50:06kugelthe sansa update can't access the manufacture mode
17:51:21kugelKlevi: please note that i might tell you big shit, since i used neither the e200tool nor a c200 ever
17:51:31KleviI'll try what google just gave me, and if it doesnt work, i can spare 50$ off of my next paycheck
17:51:36KleviKugeel, its fine.
17:51:42Klevino worries.
17:51:43kugelcan't you RMA your player?
17:51:59kugelbtw i wonder why recovery mode isnt working
17:52:11Kleviit wont do anything..
17:52:24kugelsent it back to sandisk/your seller and get a new/repaired one
17:52:38KleviWont they get me for software hacking?
17:52:45KleviIt is still warrentied
17:53:13kugelit's worth a try
17:53:25kugela not working player isn't useful at all anyway
17:53:34Kleviill try the e200 tool
17:53:50Kleviif that doesnt work it cant possibly make things any worse
17:54:01kugeli think at there was a guide for using the e200tool
17:54:11kugelyou have to look in the e200 sections ofc
17:54:19KleviYeah, ive looked
17:54:32Klevithough i didnt think you could use the e200tool for a c200
17:54:39Kleviso, i didnt bother
17:56:04 Join petur [0] (n=petur@rockbox/developer/petur)
17:58:59kugelthe e200 and c200 are very similiarin the software level
18:00:31 Quit alienbiker99 ("( :: NoNameScript 4.02 :: )")
18:03:44*barrywardell has a universal binary of libusb:)
18:05:32KleviI see, kugel
18:06:56 Join Crash91 [0] (n=evil91@
18:07:24Kleviall libusb picks up is my mouse..
18:07:31 Join webguest02 [0] (i=4a469ecd@gateway/web/cgi-irc/
18:07:32Kleviit wont detect my sansa.
18:07:39Kleviis it supposed to?
18:07:54kugelmy your pc is the problem, not the sansa
18:08:04kugeltry a different usb slot
18:08:15Klevihold on
18:08:22kugela reboot might help as well
18:08:55Kleviwindows detects my sansa being connected
18:09:00barrywardellbluebrother: do we statically link with Qt?
18:09:27bluebrotherbarrywardell: yes, but I don't know if linuxstb did this for the mac binary
18:10:18barrywardellIt doesn't look like he did
18:10:18kugeli dont think so
18:10:36bluebrotheryou need to have build Qt statically.
18:10:48bluebrotherargh, misread the last line.
18:10:55Crash91kugel: you dont?
18:10:58barrywardellah, so I'll have to build qt again!
18:10:59kugelwait, are you saying that being gay means being bad?
18:11:04Kleviweird, libusb only detects the mouse, stil
18:11:26Crash91gay as in mostly unreliable
18:11:34Klevilol, yeah
18:11:39*Crash91 doesnt know how he made that conenction, but he did
18:11:39kugelcan you access the sansa with windows now?
18:11:41Klevii agree with you on that one
18:11:47KleviNo not at all
18:11:55webguest02i'm trying to use gparted to do the bootloader patching and i put the files i need on a memory stick hooked up to a card reader. When i look through everything i can't find the two devices i have hooked up (sansa in manufacturing mode and card reader). I mounted both devices but still can't find anything. Has anyone ever tried this using gparted or know what mistake i am making?
18:12:02KleviI get USB DEVICE in device manager
18:12:11kugelyou've said that win shows your sansa
18:12:14Kleviwhich is entirely useles
18:12:44kugeldid you try normal boot or recovery mode now?
18:12:53bluebrotherwebguest02: gparted for bootloader patching? What exactly are you trying to do?
18:13:00KleviYes, neither worked.
18:13:10kugeli say, try a reboot
18:13:19Klevikugel, the blue lights are all that come on
18:13:27Klevithe screens backlight doesnt even kick in
18:13:35webguest02well, i tried using the ubuntu livecd but it wasn't working for me
18:13:44kugelhow did you brick the player?
18:13:51bluebrotherand why do you need gparted?
18:14:04Klevikugel, me?
18:14:05Crash91webguest02: what are you doing exactly??
18:14:32Kleviit formatted itself
18:14:49*bluebrother pulls crystal ball out
18:14:55KleviI had it in my pocket and i guess the hold switch was on.
18:14:55Crash91new FW??
18:15:01bluebrothermaybe e200r bootloader installation?
18:15:20webguest02bluebrother: yeas
18:15:21kugeli highly doubt that it formatted itself
18:15:30Kleviit shutoff after saying 'format complete;
18:15:44Crash91isnt hold menat to erase the RB CFG file?!?!
18:15:49pfuenderwhere can I download an rockbox version out of the january 2007?
18:15:56Kleviand since, hasnt turned back on
18:16:00bluebrotherpfuender: from svn
18:16:02pfuenderthe built versions are just 1 month old...
18:16:06kugelyou where listening to some music, and occasionally it formatted itself?
18:16:13bluebrotherwe don't keep older builds for that long
18:16:16Crash91wow, how the hell did that happen
18:16:30pfuenderbluebrother: can you tell me with which parameters? I'm not so familiar with svn...
18:16:33Kleviit was off, kugel
18:16:46bluebrothersvn co -r <revision>
18:16:55pfuenderbluebrother: thanks!!!
18:16:58kugelfrom what i know from the e200, there is no way to format the player without a pc
18:17:08Kleviapparently theres a way
18:17:14Crash91except with the new firmware
18:17:23Kleviyeah i had the latest svn at the time
18:17:26bluebrotherthere is a way. Using this empty file format.fmt (or something like this)
18:17:34Crash911.03.07P and 1.03.04H contain a "format"option
18:17:42kugelwhat do you mean with svn?
18:17:42webguest02has anyone tried this using gparted
18:17:47Crash91apparently nver tried em
18:17:49kugelsvn has nothing to do with the OF
18:17:54KleviI had 00.01.01p
18:18:11Crash911) why such and old FW
18:18:13KleviWell ,i didnt realize that was the OF and not rockbox that formated my player
18:18:19bluebrotherwebguest02: how should gparted patch the bootloader? Thats two completely different things
18:18:29Crash912) what were you doing at the time of format
18:18:45Klevii was in class sitting... i guess i moved and buttons were pressed
18:18:46bluebrotherah, it's "sansa.fmt"
18:18:49KleviI have no idea what
18:19:51Klevithat was precisely my reaction to seeing my sansa
18:20:09Crash91ok...are you in linux?
18:20:24KleviUsing linux kills my internet connection
18:20:26Crash91hold on
18:20:27kugellet me sum up
18:20:39KleviWireless and linux = no
18:20:51kugelthere is a "feature" in the OF which completely bricks the player while just sitting around
18:20:54Crash91can happen, depends on support and distro
18:21:05KleviSabayon 3.3 mini ed
18:21:14Klevikugel What????
18:21:15kugelw-lan is bad anyway
18:21:18Crash91fedora supports it, ubuntu too
18:21:21barrywardellbluebrother: do you think it's OK to add -L/usr/local/lib in the .pro file for macx?
18:21:30Crash91ok, offtopic
18:21:36kugelw-lan is for those who cannot achieve cables!
18:21:56bluebrotherbarrywardell: I guess only few people will build on mac, and those need to take care of libusb anyway so I thing it's ok.
18:22:10kugelKlevi: I meant, the OF bricket your player without you having done something
18:22:22barrywardellit only produces a warning if the dir doesn't exist anyway
18:22:26Kleviso, it killed my player after realizing that it was hacked?
18:22:46 Join Mouser_X [0] (
18:22:51bluebrotherdoesn't /usr/local/lib exist on mac per default?
18:23:02Crash91Klevi: Thats ridiculous, the FW isnt that smart,..
18:23:10Klevi*shrug* I dont know.
18:23:17barrywardellI don't think so anyway...or maybe, I can't remember
18:23:29kugeli guess, the rockbox bootloader is your problem
18:23:37kugelsince it doesnt find a rockbox installation
18:23:40Crash91okay, so overall, your player has been formatted by ghosts and you cant connect it'
18:23:59Klevilol.. yeah, or access anything thats said to recover it.
18:24:05kugelbut actually you should be able to boot into the of
18:24:06bluebrotherwell, I don't think it would be an issue adding it. And if we find a better solution we can always improve it ;-)
18:24:06Crash91press on and flick hold, then hold the REC switch
18:24:14Klevitried that many times
18:24:32Crash91try pressing select instead of REC
18:24:40Crash91 scratch that
18:24:40 Quit webguest02 ("CGI:IRC (EOF)")
18:24:53Crash91get your cable and plug it in to your computer
18:25:14Crash91connect sansa
18:25:29Crash91goto disk manager
18:25:42Klevi"found new hardware, USB DEVICE"
18:26:04Crash91ok, what does your sansa look like atm
18:26:11KleviBlack screen
18:26:17Kleviconsant blue lights
18:26:30kugelCrash91: i just want to let you know, that he has an c200
18:26:35Crash91ok i think somehow, the hold switch may have got stuck
18:26:37KleviI dont have the scroll wheel , i have a c200
18:26:45Klevi240, rather
18:27:23Crash91i thought u had a e200
18:27:30Klevino, sorry
18:27:42 Join karashata [0] (
18:28:22kugelit seems that only the e200 can repair your player, but again, i don't know if it's working on the c200
18:29:02kugeli suggest you use the RMA
18:29:14Crash91thats what i thought, but "E"200 tool is unlikely to work
18:29:43kugeli allready told you what RMA is
18:29:59Klevi.. mightve mised that, scrolls up***
18:30:31 Part df___
18:31:31Crash91Klevi, i found this on a SanDisk thread
18:31:31kugelCrash91: i think it can work, since c200 and e200 are not too different in the software level
18:31:31Crash91sounds like a lemon. Take it back to the place of purchase or call sandisk @ 866.sandisk to get it replaced.
18:31:58Klevilol.. ill cal sandisk.. brb
18:32:20kugeli remember
18:32:38kugeli had to call sandisk 1 day after i got my e200
18:33:29 Join My_Sic [0] (
18:33:34kugelwould be off topic
18:33:47Crash91kugel: #crash?
18:33:48My_Sichie all
18:34:49My_Sicwhen do you think Rockbox will work on ipod 6G ? one month ? two month ?
18:35:05Crash91we dont give estimates
18:35:08Crash91when its ready
18:35:22Klevilol.. "the number you have dialed is not in service"
18:35:31Kleviare you guys SURE thats the number?
18:35:33Crash91lol look on their site
18:35:38KleviI am now.
18:35:41Crash91Klevi: this is the thread
18:35:42Klevihavent found it yet
18:35:43 Quit marc_ ()
18:38:04 Quit My_Sic (Client Quit)
18:40:27 Quit ompaul (Remote closed the connection)
18:46:46 Part kkurbjun
18:51:56 Nick fxb is now known as fxb__ (
18:53:33Klevithat was
18:53:41Klevireally easy to get a new player
18:54:08 Join parafin|away [0] (
18:55:03 Quit Mouser_X (Read error: 104 (Connection reset by peer))
18:55:04 Quit parafin (Nick collision from services.)
18:55:09 Nick parafin|away is now known as parafin (
18:56:15 Join Mouser_X [0] (
18:57:39 Part pfuender
19:05:02 Quit lee-qid (Read error: 104 (Connection reset by peer))
19:06:11 Join lee-qid [0] (
19:06:33 Join ilgufo [0] (
19:08:18 Join obo [0] (n=obo@rockbox/developer/obo)
19:10:13***Saving seen data "./dancer.seen"
19:10:34barrywardellbluebrother: do you have a 128x128 version of the rbutil icon? I want to add an icon for the mac version
19:12:20KleviQuestion: Will rockbox be in some form of legal trouble if by someodd chance they can prove it was rockbox that caused my sansa to die when I sendit in?
19:12:27bluebrotherI think the icon used is the one rasher also has on his page.
19:13:01bluebrotherKlevi: no, why should it? They will only tell you that you've void your warranty.
19:13:01kugelafaik it does not boid the warranty
19:13:12BigBambiKlevi: They may refuse warrenty repair, but it won't cause Rockbox any issues
19:13:23Klevikay, was worried thats all
19:14:00Klevi=) Dont want you guys to suffer becasue of me.
19:14:09BigBambiDon't worry
19:14:23 Join alienbiker99 [0] (
19:14:31Klevikugel, it doesnt void the warranty?
19:14:38Kleviyour sure?
19:14:45barrywardellbluebrother: I can get by with just the ones in the ico file for now
19:14:54kugelnot 100% sure
19:15:24kugelbut, sandisk never rejected an RMA, even if rockbox was installed
19:15:28Klevi*shrugs* im fairly certain someone else already did this with Sandisk
19:15:31bluebrotherbarrywardell: gtg now. See you later, but I don't think someone did a larger rockbox icon.
19:15:43KleviWhat does RMA mean..
19:16:09barrywardellbluebrother: it's ok. I don't need one, it would just be nice.
19:17:06Kleviok thank you
19:17:19Kleviall i needed was the literal definition of the acronym
19:17:28Klevi*points to URL*
19:19:05 Join lee-qid_ [0] (
19:19:14 Nick lee-qid_ is now known as lee-qid__ (
19:20:08 Quit lee-qid (Read error: 104 (Connection reset by peer))
19:22:14 Quit Nico_P (Remote closed the connection)
19:27:46 Quit lee-qid__ (Read error: 104 (Connection reset by peer))
19:28:29 Join lee-qid [0] (
19:34:24 Quit petur ("gone taping!")
19:40:55 Join strid [0] (
19:43:00stridhi, I'm having issues with the interface freezing using the latest rockbox build for a 30gb video ipod; it keeps playing music by the playlist like nothing's wrong, but the ipod isn't responding to physical keypresses. No, the hold switch is not toggled.
19:44:17stridis this a common issue or did I somehow mess up my ipod or rockbox installation?
19:45:03 Quit zicho (Remote closed the connection)
19:48:22 Join zicho [0] (
19:55:05 Join jhulst [0] (n=jhulst@unaffiliated/jhulst)
19:55:44parafinusually after a couple of clicks it starts responding
20:01:17stridI'm getting nothing
20:01:31stridthe backlight turns on, but no reaction no matter how much i click
20:01:41stridthis doesn't happen in the original firmware, though
20:03:36alienbiker99its a bug, reset your ipod if you can't get rockbox to run
20:04:40stridyeah, that's what i do. it's just really annoying since it happens within five minutes of starting rockbox, every single time.
20:05:05stridguess i better find an older version that doesn't have this bug
20:05:20barrywardellwhy does sansapatcher not work on the e200r?
20:08:14 Quit Crash91 ("Bye Bye!")
20:08:39 Join delYsid [0] (
20:11:45 Quit Rob2222 ()
20:12:31 Join linuxstb_ [0] (n=linuxstb@rockbox/developer/linuxstb)
20:12:51barrywardellah, invisible hidden firmware partition
20:15:43 Join Rob2222 [0] (
20:17:26AveI'm trying out a build with maxed out clock skips, as per mikes suggestion (FS #7510)
20:17:44Avesvn head with his patch + the skip mods
20:18:30 Join Nico_P [0] (n=nicolas@rockbox/developer/NicoP)
20:19:04 Join Crash91 [0] (n=evil91@
20:19:48 Part Crash91
20:24:22 Quit linuxstb (Read error: 113 (No route to host))
20:27:09 Quit Rob2222 (Read error: 104 (Connection reset by peer))
20:28:15Aveno dice, I even got a data abort..
20:28:39Aveundefined isntruction.. sounds like mem corruption or some pointer hoopla?
20:28:58 Quit Nico_P (Read error: 104 (Connection reset by peer))
20:31:40 Quit SirFunk (Read error: 110 (Connection timed out))
20:31:50 Join ptw419 [0] (
20:33:55 Join Owner__ [0] (
20:34:05 Quit spiorf (Remote closed the connection)
20:34:35 Join spiorf [0] (
20:34:59 Quit Klevi (Nick collision from services.)
20:35:04 Nick Owner__ is now known as Klevi (
20:36:10 Join foot [0] (i=88df3002@gateway/web/cgi-irc/
20:36:14footwhats up
20:36:34Mouser_XFrom the foot? I'd say the head.
20:37:05footok - so obviously the 2nd generation nano is unsupported, right
20:37:14footare there any alternatives that i can throw on my 2nd gen?
20:37:32Aveyou can throw it into lake, for example
20:37:34foot(2nd gen nano)
20:37:36Aveuseful huh
20:38:17AveI'm rather peeved that 2nd gen isnt avail to rb. my 2g nano is getting rather tight and old
20:38:40Aveand the opsions are.. weak, apple has sexy hw but the software, meh
20:38:58footipodlinux looks like it sucks
20:39:08footthe mp3 player is slow −− kinda defeating the purpose
20:39:22Aveipl is more like "the linux port to wooden leg" not much use as music player
20:39:31Avebut they support 2nd gen nano?
20:40:42footyes sir
20:41:03 Quit Klevi ("Quit messages. The unintentional channel spam of people closing IRC, realizing they have a life.")
20:41:09Avewhat the hey... if so, then there is possibility that rb would one day as well
20:41:20Avethere is some code sharing between the projects afaicr
20:41:44footthat's queer
20:41:48 Join Davide-NYC [0] (
20:41:58 Join Ebert [0] (
20:42:03Avestraight software for the queer hardware
20:42:58 Quit ilgufo (Remote closed the connection)
20:43:16Avewikipedia states that ipl does NOT support 2nd gen (or 3rd gen) nano
20:43:31Avewith a footnote: 7 - Will probably never be supported due to encrypted firmware.
20:43:44 Quit barrywardell ()
20:43:47footi said that
20:44:04Aveok my bad then
20:45:14 Quit Mouser_X (Nick collision from services.)
20:45:26 Join Mouser_X [0] (
20:46:43 Join JunkheadSDK [0] (
20:47:16AveI'm also a bit interested in cracking my nano open and replace or add memory chips to it
20:47:30AveSMD soldering is fun
20:47:45rasherAve: good luck with that - they're probably glued to the board as everything else
20:47:56Avenothing that little ehat wouldnt solve
20:48:19Avenah wel that is really distant option
20:48:52Aveor, try to score 2nd hand 4G nano from local "ebay"
20:49:01Avedid they make 8G first gens even
20:50:52 Join Wiwie [0] (
20:54:43 Join barrywardell [0] (
20:55:42 Join ilgufo [0] (
21:01:56 Join billytwowilly [0] (
21:04:41 Join TMM [0] (
21:06:53 Join linuxstb [0] (
21:08:35Mouser_XWhen building Rockbox, should I clean out the directory I'm building to? Specifically, I've been trying out different patches, but building into the same directory. Are those additional files going to be a problem? I mean, not all the builds I've done have used or output the same files.
21:09:12desowinI delete build dir, and do new one
21:09:17Avethe what what? try renaming build-dir
21:09:21Aveor that
21:09:45Avemy build script is like so:
21:09:46Avecd build
21:09:46DBUGEnqueued KICK Ave
21:09:46Avenice make clean all
21:09:47Avenice make zip
21:10:12Mouser_XWell then... I may need to redo some of these builds I have...
21:10:15***Saving seen data "./dancer.seen"
21:10:22Aveor if I'm doing little changes, Make does the magic and I trust it
21:10:37Avejust run make in the build dir after changes
21:10:51Mouser_XWhat does "nice make clean all" do?
21:11:10Mouser_X(I've just been doing "make")
21:11:17Avenice just lowers priority but clean cleans the build tree and all builds all but doesnt make the zip
21:11:24Mouser_X(And of course "make zip")
21:11:42AveI cant even remember the rationaly behind those commands really
21:12:16Aveyou have to say "all" target if you have other targets as make parameters..
21:12:45Avealternatively you could say: make clean; make
21:13:36 Join Rob2222 [0] (
21:15:05 Quit foot ("CGI:IRC (EOF)")
21:18:40jepleron e200, where does logf() "go"?
21:19:02jeplerI read logf.c but was unenlightened
21:19:24 Quit barrywardell ()
21:22:24 Quit Rob2222 (Read error: 104 (Connection reset by peer))
21:22:43linuxstbjepler: Into a memory buffer, which you can view/save via the debug menu.
21:24:03 Join BHSPitLappy [0] (n=steve-o@unaffiliated/bhspitmonkey)
21:25:27 Quit Davide-NYC ("ChatZilla [Firefox]")
21:26:44 Quit ilgufo ("So Long, and Thanks For All the Fish -")
21:27:26 Quit TMM ("Ex-Chat")
21:27:50jeplerdo I have to define ROCKBOX_HAS_LOGF to get that? otherwise I'm overlooking it on the debug menu
21:27:56CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
21:27:56*jepler rebuilds again
21:27:56 Quit Wiwie (Read error: 110 (Connection timed out))
21:29:22 Join midgey [0] (
21:29:33 Join TMM [0] (
21:31:09 Join Wiwie [0] (
21:31:22 Quit bertrik ("rebooting to linux")
21:32:34 Join ilgufo [0] (
21:34:45 Quit midkay ("Leaving")
21:35:10Mouser_XSo, with all these "extra" files (codecs mostly), will running this build of Rockbox (the one I did, in an unclean directory) cause any problems?
21:36:37jepleraha there's the logf debug option
21:36:59 Quit billytwowilly (No route to host)
21:37:31 Quit FOAD ("I'll be back")
21:37:46 Join FOAD [0] (
21:42:08 Join bertrik [0] (
21:47:23 Quit zicho ("Lämnar")
21:47:31 Join zicho [0] (
21:47:46 Quit Wiwie ("Miranda IM! Smaller, Faster, Easier.")
21:50:59 Quit midgey (Read error: 110 (Connection timed out))
21:51:16 Join wookey_ [0] (
21:51:52wookey_any progress on the sansa USB support recently?
21:52:03wookey_I just looked on the mailing list and see no mention, so I presume not
21:52:36linuxstbRead the IRC logs - there's been new work on it.
21:52:58wookey_I fell off this channel so don;t have logs here - are they stored somewhere?
21:53:30linuxstbYes, on the main website - follow the IRC link
21:53:37wookey_ah, yep just got there - cheers
21:53:45wookey_you are organised !
21:54:53 Quit animeloe ("Leaving")
21:55:22 Quit Thundercloud (Remote closed the connection)
21:58:01 Join animeloe [0] (n=animeloe@unaffiliated/animeloe)
22:01:59 Part JunkheadSDK
22:10:56 Join Zagor [0] (
22:12:35 Quit amiconn (Nick collision from services.)
22:12:42 Join amiconn [0] (n=jens@rockbox/developer/amiconn)
22:27:02 Join intruz [0] (
22:34:28 Quit ilgufo (Read error: 110 (Connection timed out))
22:35:52 Quit bertrik ("Leaving.")
22:40:19*jhMikeS notices the I2S additions :)
22:40:25ZagorBagder: it's ironic that your dblog backdrop graphics includes a PP processor
22:41:03Bagderbut rather subtle if I may say so...
22:41:26jhMikeSI hate their logo
22:42:33zichoAfter using it for a week and a day now, i must conclude; Rockbox kicks ass.
22:43:46 Quit ptw419 (Read error: 110 (Connection timed out))
22:44:02Mouser_Xzicho: Using a Gigabeat (if I recall correctly)?
22:44:11Mouser_XEven if you're not, I completely agree.
22:44:27amiconnHaha, Gnireenigne Esrever
22:44:35zichoMouser_X: Yes! The F60. Awesome piece of hardware.
22:44:54Mouser_XAwesome. Glad you got one.
22:45:10zichoGlad you told me to get one, and glad that I did too =P
22:45:14jhMikeShow are reg values being checked while retailos is playing music? have to think about that one.
22:45:14Mouser_XI've got the F40. F60 would have been nice, but 40 is fine for me.
22:45:41zicho40 would have been fine too, but i got the F60 for like 1000 SEK.
22:46:03Bagderah zicho is a Swede
22:46:06zichoCheap as hell, my 1 GB mp3-player costed 800 SEK =P
22:46:11zichoYes i am.
22:46:56Mouser_XAlso, I figured as much. 1000 SEK = ??? to me. Seeing that, I was able to conclude that it wasn't USD, and therefore, zicho wouldn't be from the US.
22:47:22Mouser_XStill though, 800 SEK compared to 1000, not bad at all.
22:47:58zichoHahaha no. Actually its cheap as hell.
22:50:39amiconnMouser_X: Google can convert currencies. In this case "1000 SEK in USD"
22:50:47Mouser_XSo, did you ever bother trying out the OF? Or did you go straight to Rockbox? (I went straight to Rockbox, and really have no idea what the OF does, or doesn't, have to offer. I'm sure I'm not missing anything.)
22:50:49 Quit jhulst ("Konversation terminated!")
22:51:06zichoMouser_X: Straight to RockBox.
22:51:12Mouser_XGood idea.
22:51:15zichoThe OF was too ugly, and i wanted it gone.
22:52:22Mouser_XIf you never need to worry about the OF again, you can check the patch tracker. Someone on there has a way to speedup boot time. I haven't implemented it, but if I recall correctly, it makes booting to Rockbox less than 10 seconds.
22:52:42jhMikeShmmm...if MrH isn't careful I might just show up for dinner one day :p
22:53:05 Quit delYsid (Connection timed out)
22:53:06 Quit Frazz (Read error: 110 (Connection timed out))
22:53:34 Join miepchen^schlaf [0] (
22:53:49zichoThats nice, although i dont find the boot time to be that bad, its always nice if it is faster
22:54:05Mouser_XSame about boot time. It's why I didn't bother.
22:55:01Zagorspeaking of that, do we have a good reason not to display the logo in the bootloader?
22:55:19jhMikeSZagor: which logo?
22:55:26Zagorthe rockbox logo
22:55:31linuxstbWe don't display anything in the bootloader any more...
22:55:31 Join Frazz [0] (
22:56:03Zagorlinuxstb: I know, I get the ugly OF boot image instead. that's why I'm asking
22:56:20jhMikeSit's possible to stop it?
22:56:32*linuxstb hasn't seen a Sansa boot
22:57:17linuxstbSomeone should try replacing the bootloader on a Sansa with (the BL_whatever.rom file) with the Rockbox bootloader...
22:57:38Zagoror maybe that flicker isn't our booloader initalizing the lcd, as I assumed. if so it wouldn't matter.
22:57:42jhMikeSIt shows "San Disk" then shows "Rockbox"
22:58:24 Quit Frazz (Client Quit)
22:58:41amiconnOn c200 it shows that ugly something
22:59:05jhMikeSthe mangey lil' monta thing?
22:59:08Zagorif I boot OF (holding 'left'), the screen is cleared (and text starts printing) just after the second showing of the sansa logo. when loading rockbox, that second showing is much longer.
22:59:31amiconnMeanwhile I tend to think that our bootloader should show a _small_ logo in "silent mode"
23:00:03ZagorI think it should display the same logo as when starting rockbox, just without version number
23:00:09amiconn...builtin of course, loading from disk would be pointless
23:00:52amiconnThe full width rockbox logo is rather large, data wise
23:01:03amiconnI'm not sure whether it would fit e.g. on H300
23:01:25amiconn(in flash rom)
23:02:29 Join andrewbassplaya [0] (i=4c0796df@gateway/web/cgi-irc/
23:02:29Zagoron targets where the full logo doesn't fit I'd say it's probably better not showing any logo than showing two different sizes in quick succession
23:02:32 Join BigMac [0] (
23:03:06LloreanIn "silent" mode, doesn't the bootloader only last a second or two?
23:03:19amiconnBut showing nothing isn't nice either, as you noticed. Either there's a manufacturer logo or screen, or nothing at all
23:03:37 Part andrewbassplaya
23:04:01ZagorLlorean: yeah, so it's definitely no big deal
23:04:28LloreanZagor: Well I seem to recall there being a (subjectively) noticeable speedup when the bootloader stopped drawing, on the iPods.
23:04:29amiconnLlorean: On the irivers the bootloader is quite slow....
23:04:30Zagoramiconn: well personally I think nothing is better than flickering btwn different sizes. at least if the boot is as quick as a second or two
23:05:02Lloreanamiconn: True, I'm used to either "really slow" (gigabeat) or "quite fast" (flash targets)
23:05:08amiconn8 seconds from tapping Play to the rockbox logo...
23:05:10LloreanI haven't booted my H120 in a sadly long time.
23:05:18amiconn(on H300)
23:05:28 Join pickscrape [0] (
23:05:37Zagoramiconn: at what time in that boot can we draw to the screen?
23:05:37LloreanCould we simply draw the word "Booting" or even "Rockbox" non-graphical to the center of the screen?
23:05:55LloreanInstead of using a logo, just something really, really simple that gives the user feedback that the boot as started?
23:05:58Nimdaehas there been a bug report where rockbox locks up (at least on video ipod) when a power only connection stops providing power (the condition that would trigger a pause, such as when turning off car)?
23:06:09Nimdaei'm searching, but i don't see this specific problem
23:06:12amiconnRight now the H300 bootloader is always verbose, so at least you notice that something's going on
23:06:15ZagorLlorean: yeah, that's better than the current at least imho
23:06:22LloreanNimdae: I think there is a bug report for car mode causing a freeze.
23:06:34jhMikeSDidn't LinusN commit a fix?
23:06:36Nimdaeok, i didn't want to submit a dupe
23:06:43LloreanZagor, Amiconn: I'm not a fan of a completely blank screen, at least.
23:06:55LloreanjhMikeS: Dunno
23:07:09Zagorotoh on c200 the boot is so fast that anything other than the logo would probably feel like noise
23:07:30Nimdaealso, i need to patch to latest to see if it's still a problem, but i hadn't seena mention in the patch notes
23:07:36amiconnHmm, maybe I should test how well the logo compresses
23:07:51amiconnWe could use our familiar ucl compression for it
23:08:11LloreanOr even use a grayscale logo in the bootloader, so instead of switching sizes you're just switching to a colored version?
23:08:18amiconnThe decompressor is just ~1KB code
23:08:38LloreanGoing from gray->color is a similar "feel" to a backlight coming on.
23:08:44pickscrapeHi, I just updated rockbox on my ipod for the first time in ages, when when it boots it just reaches "Rockbox loaded" and hangs there.
23:09:08Lloreanpickscrape: Try installing the latest bootloader. No special instructions, just follow the manual for it.
23:09:18pickscrapeLlorean: thanks, I'll try that
23:09:30jhMikeSsounds like the 30GB Video problem reported
23:10:20***Saving seen data "./dancer.seen"
23:10:36amiconnLlorean: A greyscale logo would still be a native bitmap, hence the same size as a coloured one on colour targets
23:10:50amiconnOtherwise we'd need a special drawing routine
23:10:52Lloreanamiconn: Well, grayscale probably compresses further than color, doesn't it?
23:10:56pickscrapeLlorean: how do I get the ipod to reboot so it will mount again ready to have the bootloader updated? Currently holding MENU+select makes it boot rockbox again
23:11:06Lloreanamiconn: Assuming the use of UCL as you suggested.
23:11:31Lloreanpickscrape: Just follow the instructions for booting into the original firmware, also in the manual I believe.
23:11:41amiconnI'd expect minimal differences. Apart from the orange the rockbox logo is already greyscale
23:11:51pickscrapesweet, been looking for that but I'll look a bit harder now. :)
23:11:54LloreanjhMikeS: Ah, I see it's fixed. I was really only commenting on "I think it's been reported recently" anyway. :)
23:12:12Nimdaeah, that's more recent than what i'm running on my ipod
23:12:24Nimdaeso i'll just continue building and install it and test it out
23:12:30Lloreanamiconn: Ah, never mind then. I haven't really looked closely at the small ones, so I didn't know whether some of the colors were solid, or a bit more detailed in similarity to the large one.
23:13:29 Quit lazka (Read error: 110 (Connection timed out))
23:14:16 Join lazka [0] (
23:14:54Nimdaejust want to make sure i understand correctly...the bug report is about the iriver player, i'm assuming the fix affects pretty much all players that had similar problems?
23:15:32Nimdaei guess i should just build and test
23:17:17amiconnLlorean: A quick test (using the .o which has some extra cruft in it) shows that the rockbox logo compresses to <30% of its original size
23:17:28amiconn(29.8KB->8.5KB, H300)
23:19:21Nimdaemeh, looks like i didn't catch the last time it locked up so it killed the battery...
23:19:47jhMikeSwhy couldn't rarely used images like the rockbox logo and the USB logo be kept compressed?
23:20:11BagderjhMikeS: indeed a good question
23:20:24amiconnOn colour targets that might in fact make sense
23:20:41LloreanSounds like a good idea to me.
23:20:44amiconnIt doesn't reduce ram usage though, as you need a buffer to compress into
23:20:52jhMikeSIn both cases the audio buffer is available
23:21:00Bagderand it could use the same bufffer for that
23:21:05Nimdaei have a theme i made about a year ago that i haven't touched the code for since i finished it...should i be concerned about the tags changing? i've noticed some updates for the theme tags and i always worry if when i boot rockbox it'll be broken
23:21:18amiconnjhMikeS: Not always, e.g. credits.rock also shows the rockbox logo
23:21:22LloreanNimdae: The RTC tags have changed.
23:21:35Lloreanamiconn: But in that case doesn't it use the plugin buffer anyway?
23:21:35Nimdaeit doesn't use rtc tags
23:21:43jhMikeSamiconn: but it is a plugin though
23:21:59amiconnjhMikeS: Yes, but it calls the core function to show the logo
23:22:33jhMikeSsure, but can't it just keep the buffer to decompress into itself?
23:22:37amiconnWe could use the plugin buffer, that should be available in all cases
23:22:56amiconnThe ucl decompressor doesn't work in-place
23:23:04pickscrapeHow do you turn the ipod off from the bootloader screen? It's not recognising the Long Play keypress...
23:23:17jhMikeScan it stream or be tweaked to do so?
23:23:22amiconnI'm not sure, perhaps it's possible to decompress while drawing
23:23:42amiconnThat requires to modify the output to write non-sequential
23:23:56Lloreanpickscrape: You can't turn it off from the bootloader screen, you can only reboot it.
23:24:08amiconn(a special lcd_cbitmap() )
23:24:24amiconn..just for colour targets; for greyscale or mono it doesn't make sense
23:24:35*jhMikeS really wants the bitmaps to have small headers so only one external call is visibile for any format
23:24:36pickscrapeLlorean: Yes, I'm doing that, but I'm struggling to get it to go into the original firmware on reboot.
23:25:14Nimdaeshouldn't switching hold on make it boot to original?
23:25:27amiconnjhMikeS: I have plans for that. It would also allow for nice handling of transparent bitmaps (using a mask) on non-colour targets
23:25:28LloreanYes, turning on hold immediately after the screen clears from holding Menu+Select, then leaving hold on
23:25:28pickscrapeBootloader version is 20060404-0258
23:25:44LloreanBut if it's a very old bootloader, you may have to hold "Menu" down, instead of switching on Hold.
23:25:57Nimdaeit sounds like the case here, heh
23:26:10Nimdaei'd switch it to disk mode instead and update the bootloader
23:26:11jhMikeSamiconn: and the saving by not specifying it's dimension and soforth for every invocation
23:26:24linuxstbpinkscrape: Why do you want to start the original firmware, rather than just enter disk mode?
23:26:24pickscrapeStill loads rockbox even when I hold menu after rebooting... :(
23:26:37Lloreanpickscrape: Use disk mode then.
23:26:39 Join Rob2222 [0] (
23:26:41 Join bertrik [0] (
23:26:50LloreanHold "Select+Play" immediately after the screen clears from holding "Menu+Select"
23:26:55jhMikeSamiconn: and "transparent" use of alpha masks :)
23:27:00amiconnjhMikeS: Yup, although for the _part version we won't get below 5 arguments
23:27:40amiconn4 (or less) arguments are preferable on SH and arm (repgparms)
23:27:54pickscrapeLlorean: yay, went into disk mode! Thansk, I'll proceed with the bootloader update now
23:28:45jhMikeSI think you'll see a nice binsize drop if using a nice compact but descriptive header. Icon lists can just be image lists with a common header to all sections.
23:29:04amiconnBlah, alpha .(
23:29:16Nimdaeactually, pickscrape reminded me of an issue i have with the bootloader and original firmware detection...
23:29:39Nimdaei don't use the original firmware at all, so when i have hold on and i boot my ipod, it stalls stating it couldn't find the original firmware
23:29:49Nimdaecouldn't it just boot rockbox at that point with hold enabled?
23:30:05 Join Thundercloud [0] (
23:30:09SoapjhMikeS, did you say there was something you wanted me to test on my problematic Nano?
23:30:46LloreanSoap: It got tested last night on someone else's, and was unsuccessful.
23:30:47jhMikeSSoap: I put a patch up to try some clockskipping and see if that helped things but apparently not
23:31:08Soapdang, I was more behind in my log reading than I thought.
23:31:19LloreanThe interesting thing, I think, is that some of the people with problematic Nanos have gone back to 75mhz builds and still seem to occasionally experience the problem.
23:32:22jhMikeSLlorean: you mean old builds or builds with the 75MHz patch? Perhaps DEV_TIMING1 needs different settings there? That reg affects H10 but e200 doesn't seem to care how it's dunno.
23:32:26amiconnHmm, I just had an idea
23:32:58LloreanjhMikeS: Old builds.
23:33:14LloreanjhMikeS: They've gone back to a revision prior to the changes, then said "It's much more stable, but it still sometimes happens"
23:33:21LloreanUsually with a "when the device is hot"
23:33:31amiconnMaybe that after all the rule that the PLL needs to run at >=96 MHz on PP5022 and to use the post divider isn't correct?
23:33:43jhMikeSLast I saw was new builds at 75MHz seemed ok
23:33:55amiconn(the apple OF does it that way - that's where I got it from)
23:34:04LloreanjhMikeS: There's somewhere between three and five forum threads on it as well
23:34:05jhMikeSamiconn: again worth a check
23:34:14LloreanPlus about a thousand scattered posts.
23:34:39LloreanNew builds at 78mhz seem mostly okay for most of the people who've tried them.
23:35:01 Quit Mouser_X (Read error: 110 (Connection timed out))
23:35:01jhMikeSIf I owned a Nano myself I probably would have just done something about it by now. =)
23:35:22LloreanjhMikeS: If you owned a Nano, you'd probably have a "working" one just to spite everyone
23:35:44jhMikeSjust like I have an e200 that never experiences problems...heh
23:36:30jhMikeSmy H10 was no such beast, fortunately
23:37:01LloreanjhMikeS: Soap has offered to attempt to get his non-working Nano to a developer interested in doing some on-target investigation into the issue.
23:37:09LloreanRather than bumping ideas off willing testers and waiting for feedback.
23:37:23LloreanWell, "non-working" as in "has the problem"
23:39:21pickscrapeThanks for your help everyone: I have a fully-functional and up to date ipod again :)
23:40:05 Quit desowin ("use linux")
23:40:54amiconnSoap: Wanna try a patch?
23:40:57jhMikeSpickscrape: and mean another reporter of the same problem may only have a simple fix to do
23:41:11jhMikeSLlorean: I could try messing with it">
23:41:26Soapamiconn, sure
23:41:46amiconnI verified on mini g2 that it sets the correct clocks
23:42:02 Join LambdaCalculus37 [0] (
23:42:10 Join marc_ [0] (
23:42:30amiconn...just without the rule to run the pll at >=96 MHz, using the same pll control values as for 5020
23:42:40LambdaCalculus37Hello, everyone!
23:43:04Soapdoing the compile thing
23:43:47LambdaCalculus37I'm working on the beginnings of a new language file for Rockbox.
23:44:46rasherPlease let it be lojban
23:44:47LambdaCalculus37I wanted to find out something about the way the data is entered into the source.
23:45:00LambdaCalculus37rasher: Actually, it's Tagalog.
23:45:20rasherAh, that'll do
23:45:24rasherAsk away
23:45:24amiconnMaybe that also fixes the occasional ui freeze on g5.5 ... would be weird in fact :\
23:46:17LambdaCalculus37rasher: Okay, in the .lang files, we have <dest>, <source>, and <voice> entries for each phrase that is used in the interface.
23:46:36LambdaCalculus37I have to change the entries for the <source> fields, correct?
23:46:52jhMikeSamiconn: maybe _some_ have to be at >= 96MHz?
23:46:56rasherNo, <dest> (and <voice>)
23:47:21rasher<source> should be left as the english phrase
23:47:34LambdaCalculus37rasher: Good thing I just changed only one to start. :)
23:47:47amiconnjhMikeS: I can't imagine a sane reason for such special-casing...
23:48:24LambdaCalculus37rasher: This one may be quite difficult to finish, however.
23:48:32rasherSome languages have skipped the <voice> part. I suppose mostly due to laziness on the part of the translators
23:48:57jhMikeSHeh. By this time and after all the crud, I sure can see it...what's "sane" got to do with PP?
23:48:58rasherLambdaCalculus37: translation is rarely easy
23:49:01LambdaCalculus37rasher: Actually, Tagalog has several dialects, so I need to find the most common dialect and translate using that.
23:50:17LambdaCalculus37It's not easy, believe me. :)
23:50:59rasherSounds like a lot of work, good luck on that
23:51:35LambdaCalculus37rasher: Thanks. I also have my girlfriend to thank for this. She's doing the bulk of the translation work.
23:52:38rasherI trust you'll be using latin?
23:52:42 Quit davina (Remote closed the connection)
23:52:46rasherCharacters, that is
23:53:06LambdaCalculus37rasher: Yes. Tagalog uses Latin characters.
23:53:27Soapamiconn, still glitching with the patch
23:54:06 Quit marc_ ()
23:54:22rasherLambdaCalculus37: yeah, just reading the wikipedia page on it, and it mentioned a (disused) script called baybayin
23:54:28Soapmind you - my Nano would always glitch for a while before freezing, many of the other people never seemed to be able to play at all.
23:55:12LambdaCalculus37rasher: We don't have to worry about Baybayin.
23:55:31LambdaCalculus37rasher: According to my girlfriend, she's using Metro Manila dialect, which is the most common.
23:55:55rasherYeah, sounds like it's about as relevant as hieroglyphics
23:56:35 Part pickscrape
23:58:44LambdaCalculus37rasher: I'm also going to add the <voice> fields as well. I want to see if it can generate voice files for the language.
23:58:50Soapdo I need UTF-32 to display hieroglyphics?
23:59:00 Join lfs-livecd2 [0] (n=guest@

Previous day | Next day