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 2009-12-10

00:00:43stripwax_Mm, after a reset, that's possible, good point
00:02:31***Saving seen data "./dancer.seen"
00:05:44B4gder ...
00:09:23JdGordon|oh goody!
00:09:30 Quit LambdaCalculus37 ("Leaving")
00:09:39stripwax_JdGordon - so, auto update does seem to be causing the thrashing. Disk check turned up a couple of lost file fragments which looked a bit like database data.
00:10:08stripwax_One thing I don't really understand, is if the thrashing *IS* the database updating, why there is no progress shown under debug menu -> database info?
00:10:42stripwax_(it says - Initialized: Yes, DB Ready: Yes, RAM Cache: Yes, RAM: 0/1506744 B, Progress: -1% (0 entries)
00:11:08stripwax_disk continues to thrash. I'll backup the database for forensics and just remove/rebuild?
00:12:01stripwax_I confirmed that 'database load to ram' setting on its own does not appear to cause the thrashing; only auto-update
00:12:46stripwax_B4gder - why is that good (or bad) ?
00:13:03*stripwax_ isn't sure if JdGordon's goody is sarcastic or not!
00:13:29B4gderI'm not sure it is either way
00:13:46stripwax_advertising is good; on the other hand instream of not "advanced users"?
00:14:26JdGordon|stripwax: always assume I'm being sarcastic.. its safer that way..
00:14:44JdGordon|Progress: -1% looks odd.. unless that means its not busy?
00:15:05stripwax_certainly sounded busy
00:15:29stripwax_Reinitialising db, let's see if it fixes it.
00:15:56JdGordon|I dont tihnk any active devs know how the DB works...
00:17:48 Join Soap [0] (n=David_Ha@rockbox/staff/soap)
00:17:52 Quit petur ("Zzzzz")
00:17:52JdGordon|the DB whats YOU! (pic of uncle same)
00:20:07 Join Thundercloud [0] (
00:25:03 Quit tarbo (Connection timed out)
00:26:02 Quit domonoky (Read error: 54 (Connection reset by peer))
00:28:02MG_ManI have a problem
00:28:18MG_ManMy H300's battery completely drained and can't even turn on long enough to start charging
00:28:41MG_ManIf I try USB instead of wall charger, it turns on then dies shortly after, repeat, also it disables my mouse in the process
00:29:07MG_ManDo I have to keep having it restart until it can sustain itself?
00:34:45MG_ManOkay, it's stable now
00:34:50MG_ManSo nevermind
00:35:50MG_ManOr not
00:36:27 Quit Soap (Read error: 113 (No route to host))
00:37:29stripwax_MG_Man - can you use the original wall charger? I don't have an H300, but does it need to *boot* in order to be able to charge?
00:37:39MG_ManI was using the wall charger
00:37:45MG_ManDoing that makes it unable to turn on
00:37:55MG_ManAnd I believe so, otherwise it would probably be on
00:41:23MG_ManWell this sucks
00:41:39stripwax_Is it booting to the original firmware? (or, at least, trying to boot to the original firmware)? or booting to rockbox
00:41:52MG_ManIt never gets past the bootloader at all
00:42:10MG_ManIt does show the iriver splash screen if I tell it to boot the original, then it dies
00:42:54MG_ManIt also won't turn on unless a USB cabme is plugged in
00:43:00MG_ManA wall charger won't do it
00:43:16MG_ManI'll try plugging it into a different computer
00:44:25stripwax_Could you try leaving the wall charger connected and plugged in, for a while. Just for a short while.
00:44:28MG_ManIt's on the WARNING! BATTERY LOW!! line
00:44:49MG_ManWith or without the USB in?
00:45:37MG_ManIt isn't on, though
00:46:04stripwax_would it normally turn itself on when charging through a wall charger?
00:46:06MG_ManI'll check it in 5 min or so
00:46:08MG_ManI dont know
00:46:14stripwax_how long have you had your h300?
00:46:24MG_Manabout a month or so
00:46:33MG_ManI've mostly charged it while it was already on
00:47:23MG_ManOh, well what do you know.
00:47:31MG_ManI put the wall charger on it's own socket and it turned on
00:47:34MG_ManCrisis averted
00:47:41stripwax_what does 'on its own socket' mean?
00:47:47MG_ManEr, it's own outlet
00:47:55MG_ManNot on one with something else on the other slot
00:47:55stripwax_what was the wall charger plugged into .. before .. ?
00:48:00stripwax_oh ok
00:48:10 Quit killan_ ("( :: NoNameScript 4.22 :: )")
00:48:21stripwax_could be coincidence, or maybe your outlet is faulty
00:48:26stripwax_glad it worked!
00:48:26MG_ManI used to have it on a power strip a while ago, I doubt that would have work so I took it off
00:48:29MG_ManYeah I think so
00:48:53MG_ManIt is, for some reason a 600w PSU for a computer makes the room die, on that outlet only
00:48:57stripwax_wall chargers should always work regardless of whether in a power strip, or their own socket. unless the outlet is faulty
00:49:00MG_ManBlows the breaker
00:49:10stripwax_ok, your wiring sounds scary
00:49:25MG_ManIt's just a faulty outlet
00:49:46MG_ManWe had a watt meter and it didnt go above 300
00:50:02MG_ManIt probably causes a spike at first which trips the breaker
00:50:20MG_ManOh well, I only keep small things on it now, such as an alarm clock and a lamp
00:50:34stripwax_rockbox cannot help with faulty outlets and cannot offer advice on domestic electrical wiring :)
00:50:49MG_ManI see
00:50:55 Join GeekShadow [0] (n=Antoine@reactos/tester/GeekShadow)
00:50:59MG_ManWell my problem is solved either way
00:51:02stripwax_stick duct tape over the outlet and don't use it would be my advice though. glad your h300 is charging ok now
00:51:37MG_ManDoesn't duct tape conduct electricity?
00:52:07 Join killan [0] (
00:52:16stripwax_get the kind that doesn't..
00:52:24MG_ManI dont think electricity passing from one peg of an outlet to another with little to no resistance is very good for it
00:52:39MG_Man(diverting to community now)
00:52:51stripwax_(fair point)
00:56:31 Join antil33t [0] (
01:07:09 Join fdinel [0] (
01:13:16 Join CaptainKewl [0] (
01:16:06 Quit Lss (Read error: 104 (Connection reset by peer))
01:19:54 Quit PaulJam (".")
01:27:19Unhelpfuldoes anybody happen to know how important the use of a specific prng is for aac pns? there are only two callers for random_int, so 1) it should probably be inlined 2) perhaps there are better algorithms? mersenne twister is supposed to be quite fast, and we have an implementation in core already...
01:31:40 Part froggyman
01:32:12 Quit Rondom (Read error: 104 (Connection reset by peer))
01:32:22 Join Rondom [0] (
01:32:38 Quit stripwax (Read error: 104 (Connection reset by peer))
01:36:38 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier.")
01:40:52stripwax_Unhelpful - any old prng ought to work fine. main requirement would be low autocorrelation and high periodicity, same as any other prng used for audio purposes. mersenne ought to be fine.
01:41:51Unhelpfulstripwax_: that's roughly what i figured. perhaps i'll try giving codeclib its "own" MT with the state in iram, and see how that compares speed-wise.
01:42:31stripwax_good idea! when does pns kick in for aac (only low bitrates, or is it used for regular audio encodings too)?
01:42:32Unhelpfulmight also be worth considering SFMT, which is apparently both faster and has better randomness
01:43:29Unhelpfuli'm not precisely sure there. i guess i'd better figure out which, if any, of our samples will be exercising this RNG :/
01:47:12 Quit hd ("Ω")
01:50:59 Quit stripwax_ ("")
01:51:23 Quit GeekShadow ("The cake is a lie !")
01:51:32 Join HellDragon [0] (
01:55:20 Quit Thundercloud (Remote closed the connection)
01:57:30 Part toffe82
02:02:34***Saving seen data "./dancer.seen"
02:04:15 Quit dmb ("Leaving")
02:19:34 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
02:20:23LambdaCalculus37I'm so close to getting the plugins for the SA9200 fully done.
02:20:51LambdaCalculus37But I need to get a 128x160 backdrop for rockblox: LD rockblox.rock
02:20:51LambdaCalculus37Undefined symbols:
02:20:51LambdaCalculus37 "_rockblox_background", referenced from:
02:20:51DBUGEnqueued KICK LambdaCalculus37
02:20:51LambdaCalculus37 _init_rockblox in rockblox.o
02:20:51LambdaCalculus37 _rockblox_loop in rockblox.o
02:20:52***Alert Mode level 1
02:20:52LambdaCalculus37ld: symbol(s) not found
02:20:54LambdaCalculus37collect2: ld returned 1 exit status
02:20:56LambdaCalculus37But I need a 128x160 backdrop for rockblox:
02:21:06CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
02:21:06*LambdaCalculus37 hates copy-paste
02:22:10 Join krabador [0] (
02:23:54CIA-6New commit by rmenes (r23915): These are the last plugin keymaps needed for the GoGear SA9200. ...
02:24:39 Quit efyx_ (Remote closed the connection)
02:25:56 Quit kugel (Read error: 104 (Connection reset by peer))
02:27:05LambdaCalculus37The best I can do is a dummy file to make it happy for now.
02:29:44 Quit liar (Read error: 113 (No route to host))
02:30:53***Alert Mode OFF
02:35:36 Quit LambdaCalculus37 ("Fwump")
02:44:06 Quit w1ll14m (Read error: 145 (Connection timed out))
02:47:10 Join w1ll14m [0] (
02:47:56 Quit phanboy4 (Read error: 104 (Connection reset by peer))
02:49:03 Quit DerPapst ("Leaving.")
02:55:02 Join FairyKing [0] (
02:55:07FairyKingdoes rockbox have battery charging bugs?
02:55:37FairyKingmy h120 with 2300 mah battery has recently got the habit of not charging past a certain % while left on
02:55:41FairyKingbut when i turn it off while charging it does go up
02:57:22FairyKingcould it be the battery i got off ebay is malfunctioning?
02:57:35FairyKingbulging and about to pour acid all over the insides?
03:11:38 Quit fyrestorm (Read error: 104 (Connection reset by peer))
03:16:11Unhelpfulinteresting the aache sample *does* enter pns_decode but never seems to hit gen_rand_vector or random_int
03:18:45 Quit Kitar|st (Read error: 131 (Connection reset by peer))
03:35:15 Join toffe82 [0] (
03:56:24 Quit MethoS- (Remote closed the connection)
04:00:54 Join grndslm [0] (
04:01:42 Quit krabador ("Sto andando via")
04:02:37***Saving seen data "./dancer.seen"
04:05:11 Join phanboy4 [0] (
04:06:51 Join angelwolf71885 [0] (
04:08:51 Nick fxb is now known as fxb__ (
04:09:26 Quit marek_ (Read error: 60 (Operation timed out))
04:16:12 Join killan_ [0] (
04:17:03 Join angelwolf71885_ [0] (
04:17:04 Quit Rondom (Nick collision from services.)
04:17:07 Join Rondom [0] (
04:19:23 Join FlynDice_ [0] (
04:19:34 Quit angelwolf71885 (
04:19:34 Quit killan (
04:19:34 Quit FlynDice (
04:19:34 Quit Byan (
04:19:34 Quit gevaerts (
04:19:34 Quit solexx (
04:19:34 Quit JdGordon (
04:19:43 Nick angelwolf71885_ is now known as angelwolf71885 (
04:20:48NJoinFlynDice [0] (
04:20:48NJoinByan [0] (
04:20:48NJoingevaerts [0] (n=fg@rockbox/developer/gevaerts)
04:20:48NJoinsolexx [0] (
04:20:48NJoinJdGordon [0] (n=jonno@rockbox/developer/JdGordon)
04:22:15 Quit angelwolf71885 ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
04:23:19 Quit MG_Man (Read error: 54 (Connection reset by peer))
04:30:14 Quit FairyKing ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
04:31:24 Quit FlynDice (Read error: 110 (Connection timed out))
04:36:39 Quit BlakeJohnson86 ("Leaving.")
04:46:01 Join BlakeJohnson86 [0] (
04:46:29Unhelpfulmethod #3 here is intriguing. i think it can be made *quite* efficient on an architecture with a barrel shifter, especially if we use 2/1, 4/3, 16/15, 256/255 as the multipliers, since the numerator can easily be advanced from one to the next by a shift+add
04:56:44 Join tarbo [0] (n=me@unaffiliated/tarbo)
05:01:57 Nick FlynDice_ is now known as FlynDice (
05:04:18 Join Strife89 [0] (
05:22:37 Quit Strife89 ("Bed.")
05:26:11 Quit BlakeJohnson86 ("Leaving.")
05:26:26 Join BlakeJohnson86 [0] (
05:46:16 Quit Horscht ("Verlassend")
05:49:13 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-aygjkdjbdagqjjgq)
05:49:56saratogaUnhelpful: when I made 64kaache.m4a I tried to make it SBR without PS
05:51:09saratogaopps read "PNS" as "PS"
05:59:52Unhelpfulsaratoga: interesting. how do you control that, exactly? nero's encoder doesn't appear to offer many flags...
06:00:16saratogaUnhelpful: Nero only uses PS below 32kbps IIRC
06:00:41Unhelpfuls/P/PN/ ?
06:00:57saratogathey did a public test of PS many years ago and concluded that the feature was useless above 32kbps
06:01:13saratogaPNS is used higher IIRC, but don't remember much about it
06:01:18Unhelpfulhrm, that method's probably not worthwhile, too many LUT hits
06:02:20saratogathough i don't really remember how PNS works, other then that it was some core LC feature that no one else seemed to care about
06:02:41***Saving seen data "./dancer.seen"
06:04:53saratogagenerally in the absence of some evidence to the contrary, when the mpeg group talks up some new feature that no one else on earth seems to care about, I tend to assume that one of their members patented it, and that it serves no functional purpose
06:07:43Unhelpfulheh. lovely. i can't seem to create a sample where the decoder actually enters gen_rand_vector. the debugger is also being a bit of a pain as gcc has optimized away variables i'd like to examine.
06:08:30saratogaUnhelpful: how are you creating these samples?
06:08:57saratogai would test iTunes AAC at 2-3 bitrates and if they dont' hit it assume that its more useless mpeg shit
06:09:09Unhelpfuli don't have itunes :)
06:09:23saratogai can make you some if its easy
06:10:07Unhelpfuli'm using the nero encoder. i'm guessing that i probably shouldn't bother with optimizing the PRNG for libfaad, since it seems it's used only inside a function that is apparently never called
06:12:32saratogathe purpose of the mpeg group is to create standards that impinge upon their members IP, our purpose is to optimize codecs for samples that actually exist in reality, sadly the two goals often have little to do with one another
06:13:18saratogayou may have just found some more of their crap
06:13:28 Quit fdinel ("Miranda IM! Smaller, Faster, Easier.")
06:13:29saratogaon a side note: "PNS is a lossy compression technique that is based on the assumption that all white noise sounds similar to the human being."
06:13:41saratogaisn't quantization error assumed to be white noise?
06:13:43*Unhelpful so wanted an excuse to port SFMT ;P
06:14:07saratogawhy is it needed to impose a method to encode something the decode produces automatically
06:17:36Unhelpfulquantization at which stage? time-domain quantization should produce white-ish noise, but at a fixed volume. i don't know what quantization noise in the transform space "looks" like...
06:18:51saratogaUnhelpful: well presumably quantization in the frequency domain produces a random error in frequency space which is uncorrelated with frequency, which is basically what the definition of white noise is
06:19:05saratogabut i'm probably missing something
06:19:24saratogai'm very biased in favor of WMA, I like simple formats
06:19:58Unhelpfulsaratoga: perhaps this has something to do with encoding "loud" white noise (greater than that introduced by quantization) by storing the noise level?
06:20:23saratogawell you can make quantization error as loud as you want by chopping off bits ;)
06:20:57saratogabut maybe this takes into account ensuring that the noise is really white>?
06:21:19saratogathat might take up some space, but i don't really know much about the statistics of noise
06:23:07saratoga(i ditched random signals and noise in grad school because it was boring)
06:27:59Unhelpfulhrm, i'd rather hoped the MUL_* macros all over sbr_qmf.c were bad... they seem reasonable so far...
06:28:17 Quit CaptainKewl (Remote closed the connection)
06:28:55saratogalibfaad seemed efficient to me in terms of basic fixed point competency
06:29:08saratogathat author clearly knew what he was doing
06:29:23saratogathe problem was that at the algorithm level much of the code was crap
06:29:33saratogafor example the use of a stupidly slow fft
06:29:44saratogain place of a fast split radix one
06:29:58 Join DogBoy [0] (n=john@unaffiliated/dogboy)
06:33:04Unhelpfulah, but that's the level at which i don't know enough to understand the code... which is why i was poking around in the log2 stuff ;)
06:33:47DogBoyis there a way to setup rockbox to stop after each track
06:33:49saratogaif you want to understand how to efficiently decode, looking at the ffmpeg or helix decoders isn't a bad idea
06:33:58saratogathe basic idea is the same
06:34:27saratogapick some thing rockbox does slowly and see if one of them does it faster
06:35:24Unhelpfuli'm somewhat afraid to look at helix, the *player* appears to be tri-licensed including GPL but GPL is not mentioned anywhere in the codec source :/
06:35:26saratogabasic functions like mdct, qmf, iqmf, quantization are essentially independent of implementation and the algorithm can be swapped between decoders
06:36:12saratogaUnhelpful: the code is definitely not GPL, but it is open souce, and algorithms cannot be licensed, so you're welcome to any ideas in the code, if now the implementation of them
06:36:45saratogaffmpeg is of course nicer since it is LGPL, but it is not fixed point
06:43:23Unhelpfulwell, actually, the ARM macros use long multiplies, but that doesn't seem avoidable short of using smaller coefficients. long muls are not free, though, and they also need an extra register vs 32x32->32
06:43:41Unhelpfulnot free vs short ones, that is :)
06:45:39Unhelpfulif it used 16 fractional bits, then multiplies could be done rather more quickly on ARMv5/6 using SMLUWx... also the fractional multipliers could be packed two to a register. i don't know if that level of precision would be acceptable though.
06:46:49Unhelpfulalso these are the CPUs that need the least help :
06:46:55 Join krazykit` [0] (
06:48:29 Part toffe82
07:03:13 Quit krazykit (Read error: 110 (Connection timed out))
07:11:21 Quit bluebrother (Nick collision from services.)
07:11:23 Join bluebroth3r [0] (n=dom@rockbox/developer/bluebrother)
07:15:31 Quit JdGordon ("Leaving.")
07:35:36 Join dmb [0] (n=Dmb@unaffiliated/dmb)
07:39:13DogBoyjust installed on sansa fuze
07:39:20DogBoythank you, very nice
07:40:52pixelmaTorne: why did you close FS #8982? It's correct that the display is only for developers but the actual bug is that initialising the database fails if there are no music files... and as far as I've seen, it hasn't been fixed yet
07:43:01pixelmaif he hadn't mentioned "-1%" all would be fine (?).
07:50:34pixelmaTorne: oh, I mixed it up with another bug report I am watching... nevermind
08:02:45***Saving seen data "./dancer.seen"
08:03:15 Join DerPapst [0] (
08:13:17 Join hebz0rl [0] (
08:14:27 Join Bagder [0] (
08:19:06 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
08:20:06 Quit goffa (Read error: 110 (Connection timed out))
08:36:46 Join einhirn [0] (
08:38:53 Join soap [50] (n=soap@rockbox/staff/soap)
08:51:21*amiconn thinks that Unhelpful stopped halfway
08:51:56amiconn[06:46:50] <Unhelpful> also these are the CPUs that need the least help :
08:52:00amiconnThen nothing...
08:52:27Unhelpfulah... no, it's just that the : didn't belong
08:52:41 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
08:53:27*amiconn re-reads
08:54:03Unhelpfulremove the ":" and read it after the previous line, and i think it actually makes sense :)
08:54:56 Quit faemir ("Leaving")
08:55:54UnhelpfulARMv5/6 have all sorts of nifty (and usually faster-than-smull) ways to do fractional multiplies - as long as you're using ARMv6 or your fractions are 16-bit
08:56:26amiconnBtw, if you just need the to 32 bits of a 32x32->64 multiply, you can even use smmul on ARMv6
08:57:39Unhelpfulwhich i would bet saves cycles vs smull - but again our lone ARMv6 target can crush most of these jobs with raw clock speed :/
08:57:40 Join flydutch [0] (
08:57:45 Join JdGordon [0] (i=836b0050@gateway/web/freenode/x-qqynnmvumoyfhgtu)
08:58:10amiconnIt saves one cycle iirc, and it also saves a register
09:01:59 Join shai [0] (
09:02:49Unhelpfulit looks like SMULWx would be a pretty decent boost on ARMv5 if we can accept 16-bit fractional constants
09:03:38Unhelpfulyeah, trashing a register for result you don't actually need isn't *too* nice, although most of the MUL_* macros in faad appear to use output shifts of less than 32 bits
09:05:13amiconnOn ARM1136, smull is 3 cycles (6 when setting flags) with 4 or 5 cycles latency, smmul is 2 cycles with 4 cycles latency (no flag setting possible
09:09:06Unhelpfultbh i wonder if the ARM targets wouldn't do *better* with C code for these multiplies, or with only the smull in asm, outputting __hi and __lo, and then the shift/merge code in C. the macros as written pretty much force a stall waiting for the smull output, where a C version could at least *try* to schedule other instructions
09:09:31Unhelpfulat least, i'm assuming gcc isn't going to schedule other instructions between ones in an inline asm block
09:12:40amiconnThat's a problem with many of those little asm macros, but not on all architectures
09:12:57amiconnARMv4 doesn't stall iirc. Same applies to coldfire
09:14:06Unhelpfulsurely the next instruction can't execute if it's waiting on a result that's not ready? or can those cores execute out-of-order?
09:14:09amiconnActually, coldfire could stall when using emac
09:14:40amiconnOn those architectures, results are always ready when the instruction ends
09:15:31Unhelpful((int64_t)x*y) >> x generates code equivalent to those asm macros on ARM, except that gcc still "knows" what's happening.
09:15:57amiconnThe downside is that more complex instructions always need several clock cycles. Introducing the concept of result latency speeds things up - at the cost of potential stalls
09:17:15Unhelpfulah, i see what you mean. delay == throughput, so there's no value in being able to reorder instructions for such cases :/
09:18:14Unhelpfulthere are a good number of LUTs not in iram in faad... it's hard to say how often they're read though
09:18:34amiconnBtw, the quite old SH1 already requires that one takes care about result latency, unlike arm7tdmi
09:21:44amiconnIt probably makes sense to let gcc potentially reorder instructions even for ARMv4, as it would help the ARM9 variants with v4 (e.g. Gigabeat F/X)
09:22:59 Join goffa [0] (n=goffa@
09:24:12Unhelpfulhrm, completely weird test_codec result after moving another LUT into iram - it fails to decode :/
09:31:01 Quit hebz0rl ("Ex-Chat")
09:32:21 Join Thundercloud [0] (
09:44:57 Join fyrestorm [0] (
09:45:41 Quit Thundercloud (Remote closed the connection)
09:54:48 Join DerPapst1 [0] (
10:02:01 Join DerPapst2 [0] (
10:02:26 Join kugel [0] (i=kugel@rockbox/developer/kugel)
10:02:49***Saving seen data "./dancer.seen"
10:03:29 Join FOAD_ [0] (
10:04:36 Quit DerPapst (Read error: 110 (Connection timed out))
10:06:06 Quit DerPapst1 (Read error: 60 (Operation timed out))
10:09:58 Join robin0800 [0] (
10:19:36 Quit FOAD (Read error: 110 (Connection timed out))
10:19:36 Nick FOAD_ is now known as FOAD (
10:20:55 Quit goffa (Read error: 110 (Connection timed out))
10:22:20kugelUnhelpful: you said remote aa only needs passing the screen for FORMAT_REMOTE to the decoder?
10:22:29kugelto the scaler*
10:22:41Unhelpfulnot *only*, no.
10:23:26Unhelpfulit also needs an output format function for the remote. this should be fairly trivial for the most part.
10:24:30kugeloutput format function?
10:25:01kugeldoesn't the scren api provide that?
10:28:19Unhelpfulsee pictureflow or test_greylib_bitmaps_scale for examples. no, this is not something the screen API provides. to save an extra store/load from memory, the scaler output is passed as 32-bit RGB values without their final descaling to an output function that descales the values and converts them to screen format.
10:32:21kugelhrm, I may have a look, it doesn't sound trivial right now to me though :(
10:33:17 Join Kitar|st [0] (
10:35:01 Quit phanboy4 (Read error: 104 (Connection reset by peer))
10:35:21 Join phanboy4 [0] (
10:36:30 Join PaulJam [0] (
10:38:53Unhelpfuli'd start, actually, by duplicating output_row_32_native* in resize.c... especially because the M3 remote format is already support there as a main display format, and that remote works with other devices. then you'd want to add any other greyscale remote formats that are supported. i wouldn't bother supporting mono remotes, AA bitmaps aren't really readable in mono.
10:40:22kugelI'm not after greyscale or mono remotes
10:40:43kugelthe X5 has a color remote IIRC, and could show aa in the wps
10:40:45pixelmamonochrome main screens don't support AA in the WPS either, and the Iaudio one is the only greyscale remote (currently)
10:40:55kugel(or greyscale)
10:41:11pixelmakugel: a greyscale one which works on all supported Iaudios
10:41:13kugelthe first greyscale was meant to be "greylib"
10:41:27pixelmaand is the main screen for the Iaudio M3
10:41:58UnhelpfulX5 has a color remote?
10:43:01UnhelpfulDeviceChart says otherwise ;P
10:45:24 Join goffa [0] (n=goffa@
10:46:04kugelanyway, I would like AA on remotes in the wps on greyscale (or future color) remotes
10:49:05 Quit DerPapst2 ("Leaving.")
10:49:15 Join webmind [0] (n=webmind@
10:49:22webmindgood dayu
10:49:49kugelUnhelpful: does that also need a output format function?
10:50:38Unhelpfulyes, how else would the scaler produce bitmaps in the right format?
10:51:07kugelwell, you said the the only greyscale remote is already supported
10:52:26Unhelpfulit's *pixel format* is support on the M3, where it is the main display. output_row_32_native_* only support one format, the native format of the main display.
10:52:49Unhelpfuls/one format/one format on any given target/
10:53:43kugelok. I think I understand
10:54:24Unhelpfulso, as i said, you'd need to clone those functions and make one that outputs M3-remote-bitmaps on other targets that can use it... i guess you'd use output_row_32_remote and output_row_32_remote_from_yuv (on targets where the main screen is color)
10:54:37Unhelpfulthe second will actually be easier, as you can extract brightness directly ;)
10:56:57amiconnRemote AA only makes sense on the X5 and M5 atm. Both use the same remote as the M3 (where it's the main display)
10:57:24amiconnX5 main display is 16 bit colour, M5 main display is greyscale (but another pixel format than the remote)
10:58:43 Quit mc2739 (Read error: 104 (Connection reset by peer))
10:58:48 Join mc2739 [0] (n=mc2739@rockbox/developer/mc2739)
11:00:21*Unhelpful is still not convinced AA makes sense on 4-bit greyscale at all, but won't argue with folks who want to use it :)
11:00:35 Join maruk [0] (
11:01:11gevaertsUnhelpful: some people only listen to music that comes with monochrome sleeves :)
11:01:41kugelsays the one who worked hard for pictureflow on mono :P
11:01:55Unhelpfulgevaerts: your AA is all badly-executed portraits of ugly dead men. ;)
11:02:19Unhelpfulkugel: that's with greylib. you are welcome to code PluginGreylibWPS. ;)
11:02:20pixelmagreylib and 4 grey shades makes some difference
11:03:51Unhelpfulnone of our mono devices has sufficient resolution to present 1-bit dithered AA and look like anything but a bunch of noise
11:03:57pixelmabut 4-grey album art can look ok already too (depends a bit on the picture and the size you are using it)
11:20:08 Quit goffa (Read error: 110 (Connection timed out))
11:26:03 Join DerPapst [0] (
11:37:13 Quit kugel (Read error: 110 (Connection timed out))
11:41:02 Join kugel [0] (i=kugel@rockbox/developer/kugel)
11:48:20 Join angelwolf71885 [0] (
11:48:42 Quit angelwolf71885 (Client Quit)
11:49:13 Join angelwolf71885 [0] (
11:50:04 Quit angelwolf71885 (Client Quit)
11:52:47 Join DerPapst1 [0] (
11:57:08 Join goffa [0] (n=goffa@
12:02:53***Saving seen data "./dancer.seen"
12:08:30 Join hebz0rl [0] (
12:10:04kugelTorne: how's the ata dma stuff going?
12:10:09 Quit DerPapst (Read error: 110 (Connection timed out))
12:10:10kugelpp ata dma thati s
12:10:46Unhelpfulamiconn: actually, i do see a reason arm_mul is constructed the way it is... it saves an add for the rounding value, instead of using mov, orr to produce the shifted 32-bit value from the 64-bit smull output, it uses movs, adc :)
12:11:06Unhelpfulmovs sets the carry bit to the last bit shifted out?
12:11:15Tornekugel: I wrote a new benchmark for raw sector IO and found that not only does it report vastly different performance figures, it also hangs the device midway
12:11:15Tornekugel: i need to do some more experiments
12:12:01TorneI'm gonna wait until after 3.5 to actually submit it anyway
12:12:26kugelof curse :)
12:13:22Tornesince while it seems to be pretty damn stable on ipodvideo it's not really been tested a great deal on all the *other* pp502x devices that use ata
12:13:27Torneand I was intending to enable it for all of them. :)
12:13:43kugelI could run it on my samsung if you like
12:14:02kugelI just need a updated patch (if there isn't already one)
12:14:15Tornean updated patch isn't needed, iirc
12:14:39Torneata-pp502x.c hasn't been changed otherwise :)
12:14:48Torneytou just need to shove the DMA #define in config-whatever.h
12:15:15Torne...actually i've just had a thought; what's the clock speed of that device?
12:15:28Tornethe current patch uses UDMA 1 which is stable on ipod at 24Mhz
12:16:11Torneit seems likely that not all devices will have the same maximum DMA mode usable..
12:16:34kugel30/80 as in svn
12:16:53Torneoh, of course it's the same
12:16:57Tornethat should be ok then
12:17:11Tornei was going to ifdef the DMA speed to look at the defined clock rates
12:17:17Tornebecause UDMA 2 works at 30Mhz
12:17:19Tornebut not 24
12:17:38Torneand the low latency boost patch makes the unboosted speed 24, whcih is why buschel's version of the dma patch uses that :)
12:18:07 Quit Rondom (Read error: 60 (Operation timed out))
12:18:17kugelyea, he mixes his gui boost in everywhere ;)
12:18:26 Join Rondom [0] (
12:18:50TorneOh, that reminds me. When does the system ever run at CPUFREQ_DEFAULT?
12:20:06TorneI don't see the harm in supporting it, tbh :)
12:20:25Tornebut i'll ifdef it to be conditional on CPUFREQ_NORMAL
12:21:39 Quit Topy44 (Remote closed the connection)
12:21:40 Join Topy44 [0] (
12:22:12kugelmaybe you could post a patch which works as is with svn?
12:22:27Tornethe one there does, as i said
12:22:33Tornewell, the combination of the several patches :)
12:22:45Tornethey all apply
12:22:58kugelwhich one is that? I assumed you meant the one that applys is designed for 24MHz
12:23:06Tornebut it makes very little difference
12:24:19kugelit seems we're post poning the release by 2 weeks, which would mean 3.5 is 4 weeks away. it might be enough to commit it before 3.5?
12:24:58Tornei really don't think it's that important :)
12:25:31Torneidle mode only gets set while listening to the radio, pretty much we care if the disk works while in cpu idle mode?
12:27:41Torneif so we should stick with UDMA 1 anyway
12:28:16kugelis it possible that the cpu is idle during disk activity?
12:28:34Tornewell, idle mode only gets set while sitting on the radio screen
12:28:47Torneor while on externally-controlled USB
12:28:54Torne(which doens't happen on the pp players)
12:28:56kugelthere's usually no disk access when listening to radio
12:29:12 Quit xavieran (Remote closed the connection)
12:29:14kugelexcept maybe on entering for loading the .fmr
12:29:18TorneYah, but 1) is that guaranteed and 2) will anyone ever add cpu_idle_mode() calls anywhere else?
12:29:49Tornei was trying to run my own benchmarks to see how much difference it made, but the benchmark makes the device hang which is bad
12:30:09Tornethe other benchmarks run by dreamlayers/buschel suggest that higher DMA modes are not noticably faster
12:30:27Tornebut I distrust test_disk somewhat as it is in svn..
12:30:35Torneat least, for measuring raw ATA performance
12:30:41kugelI'm still not sure if it's even possible to access the disk with cpu being in idle mode
12:31:05Tornenot at the moment
12:31:07kugelIIRC, idle on pp means that the cpu does virtually nothing and runs at a very low clock (32kHz?)
12:31:15Torneno, it just goes down to 24MHz
12:31:23Tornethis is a different thing to actually sleeping the cpu
12:31:34kugeloh, I was confusing it then
12:31:36Torneboost-capable targets have *three* frequencies
12:31:39Tornedefault, normal and boosted
12:31:51Tornethey run at normal when boost count is 0, unless cpu_idle_mode(true) gets called
12:32:06Zagoriirc idle turns off the PLL, which takes it below 1MHz
12:32:23TorneZagor: that's actual sleep mode, which goes to 32768Hz, yes
12:32:33Tornebut that's only when there is no code to execute
12:32:39Tornecpu_idle_mode(true) still runs code
12:33:02Torneit's used for the radio screen (it drops out of idle on button press though) and it's used for targets where USB is handled by an external chip while in USB mode
12:33:40Torneit's not used for anything else atm..
12:34:44Tornei am just wondering if anyone might ever expect to be able to access the disk in future while in idle mode :)
12:35:16Tornethere's no reason why you *couldn't* right now, even though it looks like nothing does.
12:35:29kugelthe ata layer could simply force non-idle mode
12:36:03Torneit could, but that kinda sucks
12:36:10Torneit already forces boosting if the DMA mode set is higher than 2
12:36:24Tornebut that means you add the latency of calling set_cpu_frequency around each request
12:36:25kugelwhich isn't a problem IMO
12:36:32Tornewhich for normal disk access might be just one sector at a time
12:36:46kugelwell, otoh virtually all disk access happen during boosting anyway
12:37:02TorneOnly buffering..
12:37:03amiconnkugel: It is. Boost/unboost must not happen too often
12:38:58amiconnmust not
12:39:31 Join MethoS- [0] (n=clemens@
12:41:19kugelmust/must not is pretty strong. is there any danger?
12:41:43Zagorkugel: it takes time. lots of it.
12:42:20kugelexcept overhead I mean
12:43:00kugeland it's free/takes much less time on some other targets
12:43:12kugelbut wasteful on PP, yes
12:48:08 Nick YPSY is now known as Ypsy (
12:56:27 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:16:38 Quit goffa (Read error: 110 (Connection timed out))
13:24:40 Quit kugel (Read error: 110 (Connection timed out))
13:25:03 Join archivator [0] (n=archivat@
13:31:33 Join fyre^OS [0] (
13:43:59 Quit Tomis ()
13:46:25 Join Tomis [0] (n=Tomis@
13:47:34 Quit fyrestorm (Read error: 110 (Connection timed out))
13:47:41 Join watto [0] (n=watto@
13:50:19 Join huelk_ [0] (
13:51:01huelk_hi, is there a port of rockbox for the a2025 chip
13:51:16huelk_*ak2025 sry
13:51:33 Join goffa [0] (n=goffa@
13:53:35TorneThe front page of the website says exactly which platforms are supported
13:53:53CIA-6New commit by teru (r23916): Rework spacerocks: ...
13:53:57 Join teru [0] (
13:54:07markunhuelk_: which players use that chip?
13:55:22Bagderhuelk_: yes, but no usable code
13:55:25huelk_its a no name device i got as a present
13:55:36Bagderor have you found actual usable driver code?
13:55:40huelk_i disassembeld it a hour ago to write down the chip names
13:55:52Bagderah ak2025, I misread
13:55:57*Bagder hides again
13:56:35markunBagder: which one were you thinking of?
13:56:54BagderI thought about AS, the ams one
13:57:08huelk_in linux lsusb gives me: d-wave 2gb player
13:58:49 Part watto
13:58:58TorneOh, it's one of the s1mp3 thingies
13:59:31huelk_doesnt it work with rockbox?
14:00:29Bagderhuelk_: only targets we list on the front page work with rockbox
14:00:43TorneNo, they won't ever work probably
14:00:49Tornethe s1mp3 devices are Z80 with a DSP
14:00:56Torne8-bit MCU is way smaller than anything we can run on
14:01:16Torne <- these people might be able to help you thouhg
14:01:24huelk_hmm ok, is there something like rockbox for these players
14:01:35huelk_k i will ask them
14:02:12Tornes1mp3 has a firmware replacement for them in general, dunno if it supports that specific one
14:02:20Torneand dunno what it can do compared to the original firmwar.e
14:02:32huelk_are there other projects like rockbox, i googled but dont found anything thats nearly as good as rockbox
14:02:55TorneOh, they have a *project* for a firmware replacement
14:02:56***Saving seen data "./dancer.seen"
14:02:58Tornebut not really any code yet, it seems :)
14:03:07Tornewell, nothing finished
14:03:11huelk_hmm k ^^
14:03:14huelk_thx for your help
14:03:33huelk_the player is good, but the firmware sucks :/
14:03:35Torneand no, there aren't really any other projects like rockbox, at least not that are open source (rather than hacks of OF) and actually work on one or more players
14:08:11 Quit PaulJam (Nick collision from services.)
14:08:16 Join PaulJam_ [0] (
14:08:24 Join kugel [0] (i=kugel@rockbox/developer/kugel)
14:11:45amiconnkugel: It's wasteful on PP, even more wasteful on coldfire, and affects timer precision on coldfire
14:13:19kugelTorne: ipodlinux
14:14:09 Part Bagder
14:18:43 Quit archivator (Remote closed the connection)
14:18:53 Quit TaZzZ ("ZNC -")
14:19:18Tornekugel: does just porting linux really count? :)
14:19:38Tornethat's not really an OS project
14:20:04Torneyou write a couple drivers and hack up the boot process and you're done, and the rest of your development is just "can you write a userspace app which is tolerable to control with a scrollwheel" :)
14:22:47kugel"you write a couple drivers and hack up the boot process and you're done" <- that also describes the progress of porting rockbox (unsurprisingly)
14:23:46gevaertskugel: exactly :)
14:23:56Tornemy point is that the rockbox *project* includes developing the actual OS as well as porting it :)
14:24:56kugelyou could also just port it and be done
14:25:33kugeland you can develope on linux as well after porting. anyway, I don't see how ipodlinux doesn't count. It's a firmware replacement too :)
14:25:39Torneanyway, it's all kinda irrelevant :)
14:25:51 Join TaZzZ [0] (
14:25:54Tornethe hardware he was talking about is not going to run *any* of these things
14:25:59Tornebecuase it's an 8-bit MCU.
14:26:18Tornethat's not likely to ever run anything that wasn't written very specifically for it.
14:26:27Tornesince 8-bit machines can't even support a 'real' standard C environment
14:27:18linuxstbarchopen is/was another project vaguely similar to Rockbox
14:27:21kugelnot? with enough registers and instructions, 16/32bit can be emulated sufficiently, not?
14:27:38kugelwasn't C designed in 8bit times?
14:27:39Tornekugel: well yes, but you wouldn't want to..
14:27:41Torneand no
14:27:49gevaertsTorne: is there something in the C spec that needs more than 8?
14:28:01TorneYes. 'short' must be at least 16 bits.
14:28:06Torneand thus int/long also
14:28:31TorneWhich means that if you have an 8-bit machine then *all* your arithmetic will be crap, except where you explicitly use char :)
14:28:59Tornecompilers for 8-bit machines tend to just ignore a few inconvenient spec details, afaik :)
14:29:08kugeldoes C really define short as 16bit?
14:29:16TorneI think it's defined in range terms
14:29:21Tornebut yes, more or less
14:29:36*Torne looks it up
14:29:59 Join faemir [0] (n=faemir@
14:30:06kugelIIRC the standard permits short, int and long being all of the same size
14:30:19kugelwithout minimum size
14:30:54Torneshort, int and long can all be the same size, yes
14:30:59Tornebut they must all be at least 2 bytes
14:31:09*gevaerts also wants to see a citation :)
14:31:10kugelbut for C99 you probably need to be able to handle int32_t and the like
14:31:21Tornegevaerts: i'm looking for it but you know how the C spec is ;)
14:31:39*gevaerts moves to the other channel
14:33:02 Part LinusN
14:35:01 Join angelwolf71885 [0] (
14:38:51 Quit DerPapst1 ("Leaving.")
14:46:37amiconnTorne: I don't know whether this applies to all 8 bit CPUs, but Z80 has some 16 bit arithmetics instructions, and supporting 16 bit arithmetics on it isn't too difficult
14:46:54amiconn(multiplication needs to be performed by a subroutine anyway)
14:47:52Tornewell yes. but anyway, we are gonna be ay too big to fit :)
14:52:37 Join Omlet [0] (
14:53:40CIA-6New commit by teru (r23917): plugin.c: It wouldn't make much sense to clear screen right before setting back colors to global settings.
14:53:56 Join DerPapst [0] (
14:58:43 Quit robin0800 (Remote closed the connection)
14:59:51 Join Strife89 [0] (n=Strife89@
15:03:24 Quit huelk_ (Read error: 110 (Connection timed out))
15:10:23 Nick Omlet is now known as Omlet^study (
15:12:47 Join krazykit [0] (
15:17:23kugelteru: the whole clearing shouldn't be needed anymore
15:20:35terukugel: ?
15:29:14 Quit krazykit` (Read error: 110 (Connection timed out))
15:31:57kugelteru: w.r.t to you last commit
15:34:25terukugel: why it shouldn't clear display?
15:34:48kugelthe function call below should handle that
15:44:43 Quit Strife89 ("Leaving")
15:45:35gevaertsteru: you saw FS #9380 on arm?
15:47:08terugevaerts: what do you mean by on arm?
15:47:29gevaertswhen I reported it, I could only reproduce it on coldfire
15:47:38pixelmagevaerts: noticed the yellow Player builds?
15:48:15gevaertspixelma: oops, no
15:48:52*gevaerts will fix that when he gets home
15:49:01gevaertsunless someone else is faster
15:59:57terugevaerts: i think i saw the issue on my gigabeat when the task was opened.
16:00:36gevaertsanyway, I'll try your patch tonight
16:02:59***Saving seen data "./dancer.seen"
16:07:51kugelI think I have seen that too once
16:08:26teruit occured when i left the game in start screen for a while. but since start screen was replaced by menu, it may not likely to reproduce.
16:10:30 Join archivator [0] (n=archivat@
16:16:47gevaertshm, indeed...
16:16:54gevaertsI'd just commit and close then
16:23:16 Quit archivator (Read error: 131 (Connection reset by peer))
16:23:30 Join archivator [0] (n=archivat@
16:27:01 Quit goffa (Read error: 110 (Connection timed out))
16:27:59 Quit teru ("Quit")
16:35:24 Join watto [0] (n=watto@
16:39:36 Join Grahack [0] (
16:41:41 Join toffe82 [0] (n=chatzill@
16:42:55 Join goffa [0] (n=goffa@
16:48:33 Quit Zagor ("Don't panic")
16:55:20 Join archivator_ [0] (n=archivat@
16:55:27 Quit toffe82 (Read error: 104 (Connection reset by peer))
16:55:47 Quit archivator (Nick collision from services.)
16:55:54 Nick archivator_ is now known as archivator (n=archivat@
17:04:00 Quit MethoS- (Remote closed the connection)
17:07:02 Join toffe82 [0] (n=chatzill@
17:08:08 Quit einhirn ("Miranda IM! Smaller, Faster, Easier.")
17:09:18 Quit angelwolf71885 ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
17:15:47 Quit toffe82 ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
17:16:36 Quit mc2739 (Read error: 104 (Connection reset by peer))
17:16:48 Join mc2739 [0] (n=mc2739@rockbox/developer/mc2739)
17:18:14 Quit kugel (Read error: 110 (Connection timed out))
17:21:33CIA-6New commit by Blue_Dude (r23918): Fix yellow: add braces
17:22:08 Join Zaba [0] (n=zaba@about/goats/billygoat/zaba)
17:22:46Zabahi guys, how hard is it to port rockbox to a new device?
17:23:10 Join toffe82 [0] (n=chatzill@
17:23:46Zabaand what are the minimum requirements from the device in question? Does it need, for instance, a mmu?
17:23:53TorneNo, it doesn't need an MMU
17:24:15TorneIt's not particularly hard to port to a new device, really, unless the device in question has poor/no public documentation
17:24:22Torneunfortunately this is true for a number of our ports :)
17:24:56 Join dfkt_ [0] (i=dfkt@unaffiliated/dfkt)
17:24:59ZabaI was just entertaining the idea of building myself an audio player based on some microcontroller.
17:27:11Zababut rewriting all software that's involved in that seemed like the most effort-duplicating idea ever.
17:31:57TorneWell, someone recently ported Rockbox to the mini2440 dev board
17:32:07Torneas a stepping stone toward making a custom player..
17:32:12Torneso that kind of thing is certainly possible, yes
17:32:44Torneif you are sourcing your own components you will almost certainly have full docs for them, and thus a rockbox port is probably quite a bit easier than for any commercial DAP
17:33:12Tornewhat kind of microcontrollers are you looking at?
17:33:12Zabahm, why can't I seem to find a source download of rockbox, and only prebuilt versions...
17:33:26Zabaat the moment, various AVRs
17:33:29TorneThe source tarballs are gone
17:33:36Torneuse svn to check out the current code
17:33:38TorneAVRs are too small
17:33:41Zabareally? That sucks, svn is so slow
17:33:48Zababut okay.
17:33:58TorneWe don't currently support anything smaller than 32-bit.
17:33:59Zabahm, I thought it'd be ok with some expansion ROM
17:34:12Zabathere are 32bit AVRs
17:34:36Zabaadmittedly, I had trouble finding any in stores around here, but there were a few
17:34:39Torneno there aren't
17:34:53 Quit archivator (Remote closed the connection)
17:34:59TorneAtmel make other microcontrollers which are based on ARM cores
17:35:02Tornewhich are 32-bit.
17:35:34Tornebut the AVR architecture is an 8-bit system with a 16-bit address space
17:36:06TorneOh, there's AVR32 as well. Whoops.
17:36:10TorneBut that's basically a completely different hting
17:36:16Zabaah, yes, that makes sense
17:36:18Torneit's not really related to AVR at all except in the name
17:36:38TorneThey're also not something we support
17:36:46TorneWe only have toolchains/code for ARM, SH, MIPS and ColdFire
17:37:06TorneFor Rockbox you would probably want something more like the AT91SAM
17:37:18Zabaso it would be indeed a lot more porting work, if I insisted on that?
17:37:33TorneIf you wanted to actually use avr32 then yes, that would be a great deal more work
17:37:41ZabaI see
17:37:52Torneyou would not be able to make use of *any* processor-specific code
17:37:59Torneand you would have to get your own toolchain working
17:38:44Zabadoes "toolchain" refer to a C crosscompiler & friends, or is it anything rockbox-related I'm not aware of?
17:38:45Tornealso, rockbox is generally going to need at least a megabyte or two of RAM
17:38:54TorneJust the compiler and linker, yes
17:39:27Tornean external ROM isn't going to suffice; the AVR32 chips I see on atmel's site have only 32KB of on-chip SRAM or thereabouts
17:39:40Tornewhich is not sufficient to run rockbox even if the code i sin rom
17:39:43 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
17:39:53ZabaI see.
17:40:52Torneyou would need an external SDRAM device
17:43:22 Quit dfkt (Read error: 110 (Connection timed out))
17:44:00Zabahm.. some of the plugins are scary
17:44:13TorneWell, the plugins aren't required
17:44:18Tornemany of them don't work on all targets
17:44:25Zabaif I wanted doom and sudoku, I'd get a laptop :>
17:44:34Tornedoom doesn't work on lower end targets, not enough cpu/ram :)
17:44:54Tornethey're conditionally built for the platforms which are fast enough, have enough memory, and a big enough screen to suport them
17:45:19TorneBut anyway, yah. If you *really* wanted to use AVR32 then it would be a lot of work
17:45:27Tornebut it's not clear why you would want to :)
17:45:45TorneIt's not really related to AVR in any way other htan the name and some of the peripherals; the instruction set is entirely different
17:45:54Torneso if you already have AVR experience then it won't hepl you with AVR32
17:46:22Zabathat makes sense, from atmel's marketing point of view.
17:46:24ZabaI guess.
17:46:24Torneyou might as well learn ARM instead, which is easier for us to support and is more generally useful in the industry anyway ;) (avr32 is pretty obscure at present, they are not doing a very good job of competing with ARM)
17:47:32gevaertsdepending on what you want to run, don't get a too slow ARM either
17:47:38Tornethe line between what's a microcontroller and what's a processor is kinda blurry and vague, and a lot of the manufacturers have odd definitions
17:47:46Tornebut *generally* rockbox can't run on most htings which are described as microcontrollers
17:48:13Tornethe kind of processor our targets use ar enormally described as SoC, system-on-chip, rather than microcontrollers
17:48:41Tornethough we could run on AT91SAM or similar ARM-based devices which Atmel market as microcontrollers, if you provide an external SDRAM
17:49:45TorneIf you know how to develop for AVR then you have a lot of the skills for low level development already
17:49:59TorneMoving that to a new architecture, even one which is rather larger and faster, is not too difficult
17:50:13Torneyou might have to learn a new assembly language, but hey
17:50:29Tornevery little of rockbox is actually written in assembly outside of the early startup code.
17:50:41Zabathat seems reasonable
17:51:05 Quit Sajber^ (Read error: 54 (Connection reset by peer))
17:51:39TorneYou might be interested in which is an ARM devboard that was ported recently
17:52:41Tornethat's probably hte kind of system you are looking at, anyway
17:53:20Torneyou could go quite a bit smaller *physically*, if you got hardware that was less targetted towards development and more towards prototyping/integration, and a bit lower spec too, but that's the ballpark ;)
18:03:03***Saving seen data "./dancer.seen"
18:06:44 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:13:22 Quit FlynDice (Remote closed the connection)
18:13:36 Join kugel [0] (n=kugel@rockbox/developer/kugel)
18:18:05 Join FlynDice [0] (
18:20:08 Quit hebz0rl ("Ex-Chat")
18:27:31 Quit DerPapst ("Leaving.")
18:28:34 Join pamaury [0] (
18:29:30 Quit Grahack (Remote closed the connection)
18:31:24 Quit maruk ("Leaving.")
18:42:04CIA-6New commit by FlynDice (r23919): Sansa AMS: Add extra delay to the uSD init sequence to enhance card compatability, ...
18:44:06 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
18:50:15 Join DerPapst [0] (
18:54:59 Join Tomis2 [0] (n=Tomis@
18:56:26LambdaCalculus37All of the plugin keymaps for the GoGear SA9200 are done. Now I just need artwork for the plugins.
18:56:34LambdaCalculus37And that cabbiev2 theme. ;)
18:57:38 Join robin0800 [0] (
18:58:01 Quit HBK ()
18:58:19 Join Blue_Dude [0] (n=chatzill@rockbox/developer/Blue-Dude)
18:58:55Blue_DudeI was making a manual build and saw a number of overfill and underfill warnings. They're just warnings, but is it worthwhile to fix them?
18:59:20AlexPBlue_Dude: There are millions
18:59:39AlexPBlue_Dude: If you can fix them great, but IIRC bluebroth3r said it wasn't trivial/easy
18:59:56AlexPAnd if they could go it'd make the real errors actually possible to see
19:00:09AlexPer, s/easy/possible/
19:01:45Blue_DudeIs there a way to change the verbosity of the manual build log file? Maybe some could be suppressed...
19:02:04 Quit Tomis (Read error: 110 (Connection timed out))
19:02:05 Nick Tomis2 is now known as Tomis (n=Tomis@
19:05:13saratogalatex spits out a ton of warnings basically always and for all things
19:05:29saratogaits not like c where the warnings are more serious
19:05:50*LambdaCalculus37 begins to work on the SA9200 manual
19:10:55saratogaLambdaCalculus37: whats the status of that port?
19:11:01Blue_DudeI guess I'm just procrastinating and casting about for a short project. I'm absolutely deadlocked on the pcmbuf mixer thing. It's going nowhere. :(
19:12:40LambdaCalculus37saratoga: The plugin keymaps are done. I need the artwork for 128x160 resolution so that some of the plugins compile, and I need a cabbiev2 theme for 128x160 as well.
19:13:02saratogaLambdaCalculus37: is it usable? perhaps making it unstable would be a good idea
19:13:08LambdaCalculus37saratoga: Most everything works, and IIRC lowlight even fixed up the sound output so there's no pops or clicks.
19:13:10 Join linuxguy4 [0] (
19:13:22LambdaCalculus37saratoga: Once I get the artwork in, then yes.
19:13:23 Quit linuxguy3 (Read error: 60 (Operation timed out))
19:13:56LambdaCalculus37saratoga: It's fully usable and even has USB and charging already enabled, since there's no dual-boot for it.
19:14:00pixelmaLambdaCalculus37: does a sim compile?
19:14:12LambdaCalculus37pixelma: No, it craps out because of lack of artwork.
19:14:29LambdaCalculus37It compiles fine without the plugins enabled for the time.
19:14:37 Join HBK [0] (
19:15:05saratogayou could disable nonworking plugins for now and make it 'unstable' then
19:15:15pixelmaok, might have a look but can't tell when (well... except... probably not today)
19:15:23CIA-6New commit by rmenes (r23920): The very tiny beginnings of a manual for the GoGear SA9200. Still ...
19:15:43LambdaCalculus37saratoga: Currently not working plugins are rockblox, and.... rockblox (doesn't get any further than that).
19:15:49 Quit dfkt_ ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
19:16:43saratogawell if everything else works it seems a shame to hold up things over some bitmaps
19:17:31LambdaCalculus37saratoga: I know you can disable certain plugins using apps/plugins/SOURCES, but I'm not completely sure of how to go about that.
19:17:37 Join Tomis2 [0] (n=Tomis@
19:17:54 Quit kugel (Read error: 113 (No route to host))
19:18:21 Quit lostlogic (Read error: 60 (Operation timed out))
19:18:30 Join liar [0] (n=liar@
19:19:05 Join lostlogic [50] (n=lostlogi@rockbox/developer/lostlogic)
19:19:16 Quit liar (Read error: 104 (Connection reset by peer))
19:19:19 Join liar [0] (n=liar@
19:20:48LambdaCalculus37saratoga: I want to get the GoGear SA9200 pushed to unstable as well... would be a nice Xmas gift. :)
19:21:02saratogaLambdaCalculus37: you can add a "#if (LCD_WIDTH == XYZ)" in the SOURCES or SUBDIRS files to disable plugins that don't have the right bitmpas
19:21:31saratogathere are already checks for screen size in the files
19:21:35 Join robin0800_ [0] (
19:22:11LambdaCalculus37saratoga: Okay, I'll take a look at that later on.
19:22:16*LambdaCalculus37 has to go now
19:22:35 Quit LambdaCalculus37 ("Fwump")
19:22:39 Quit Blue_Dude ("ChatZilla 0.9.85 [Firefox 3.5.5/20091102152451]")
19:23:23 Join kugel [0] (n=kugel@rockbox/developer/kugel)
19:24:57 Quit Tomis (Read error: 110 (Connection timed out))
19:24:58 Nick Tomis2 is now known as Tomis (n=Tomis@
19:27:32 Quit amiconn (Nick collision from services.)
19:27:33 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
19:27:51 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
19:29:31 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
19:29:31 Quit pixelma (Nick collision from services.)
19:29:49 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
19:31:09 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:32:51 Quit phanboy4 (Read error: 104 (Connection reset by peer))
19:34:23 Quit robin0800 (Read error: 110 (Connection timed out))
19:34:55 Join Sajber^ [0] (
19:35:32 Join phanboy4 [0] (
19:41:58 Join Thundercloud [0] (
19:43:10 Quit phanboy4 (Read error: 104 (Connection reset by peer))
19:43:13 Quit robin0800_ (Remote closed the connection)
19:44:56 Quit Zarggg ()
19:48:33 Quit kugel (Read error: 113 (No route to host))
19:49:36 Join phanboy4 [0] (
19:53:20 Quit DerPapst ("Leaving.")
19:53:59 Quit Omlet^study ("( :: NoNameScript 4.22 :: )")
19:54:59 Join einhirn [0] (
20:03:04***Saving seen data "./dancer.seen"
20:06:59 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
20:07:44 Join Zarggg [0] (
20:09:28 Join huelk_ [0] (
20:11:47 Join AmimanPL [0] (
20:12:09AmimanPLHi, I've 2 problems with rockbox
20:14:18AmimanPLHi there
20:14:44gevaertsAmimanPL: if you have questions, just ask
20:15:38AmimanPLRockbox doesn't want to update (i'm doing it correctly and I update daily) and it still shows version r23900
20:16:05AmimanPLI've asked but irc is very slow
20:16:13 Join Tomis2 [0] (n=Tomis@
20:16:20gevaertswhich player?
20:17:08AmimanPLsecond one is the tag problem. In the one case, I have solved it manually but in the second I can't do anything
20:18:02AmimanPLalbum is tagged correctly and is showing all the tags in windows but when I browse my collection on ipod I see the cutted album tag
20:18:10AmimanPLand rockbox read that value
20:19:03AmimanPLso I have 1990 - The Dresden Performance (with all 2 discs as one folder) instead of two albums - 1990 - The Dresden Performance [CD 1] and same but with [CD 2]
20:19:34 Quit flydutch ("/* empty */")
20:19:55AmimanPLipod, 5gen, 80gb
20:20:36AmimanPLi think that there is a limit for an album name tag
20:21:08AmimanPLi tried to convert the files from mp3 to mp4 but it didn't worked
20:21:37gevaertsdon't. Transcoding between lossy formats is always a bad idea
20:21:52*gevaerts doesn't see anything wrong with the ipod video downloads
20:22:38AmimanPLI'm doing it rarely and to no lower bitrate than 128kbps
20:23:00AmimanPLi don't hear the difference - i have impaired hearing
20:23:51 Quit Tomis (Connection timed out)
20:23:51 Nick Tomis2 is now known as Tomis (n=Tomis@
20:23:52AmimanPLso I do it very rarely
20:25:13AmimanPLI've updated rockbox - now it works good
20:25:32pixelmaAmimanPL: do you use the Rockbox Utility to update? And when did you try last time?
20:26:03AmimanPLI have to go now, I will be back in some minutes. Just to change the computers
20:26:11 Quit AmimanPL ("CGI:IRC")
20:30:42CIA-6New commit by tomers (r23921): Uncomment lines which were previously commented out by mistake
20:36:46 Join AmimanPL [0] (
20:36:54AmimanPLHi, I'm back
20:37:00*pixelma doesn't understand the \\nopt there at all in this commit
20:38:58pixelmaAmimanPL: shall I repeat my questions?
20:39:21AmimanPLNow, my rockbox updates correctly but I still have the tag problem
20:39:43AmimanPLyes, please
20:41:07AmimanPLthere is about a minute delay in messages from my side
20:42:23pixelmaif you could update now, they don't matter anymore. Abou the tags - can you tell us a bit more about the type of tags (ID3v1 or v2 or both or even Ape tags, and maybe if you have some unusual comment tags or embedded album art)?
20:42:53AmimanPLthe delay is greater
20:43:23AmimanPLok, I wil check them
20:44:11pixelmamaybe try freenode's webchat and join #rockbox by hand if you know how to do that, it's snappier
20:44:57AmimanPLapev2 & id3v1.1
20:45:07AmimanPLno album art
20:45:27AmimanPLalso, Rb doesn't show me it, even if its entered
20:45:51AlexPRemove the ape tags
20:46:01AmimanPLcomments - none
20:46:06pixelmaid3v1 has a length limit and Rockbox doesn't support Ape tags in MP3, I guess your PC app reads those
20:46:20AlexPCopy everything to id3v2
20:46:51AmimanPLso how I can do it?
20:47:21AlexPI don't know, there are plenty of programs to manipulate tags out there
20:47:30AmimanPLI can only edit the id tags
20:48:06pixelmause a tagging program on your PC to copy the content of the Ape tags to ID3v2 and remove the former (maybe even the ID3v1 tags)
20:48:20CIA-6New commit by mc2739 (r23922): tools/configure does not let you select normal build for sansae200r - add similar logic for the simulator build
20:48:26 Part watto
20:48:28AmimanPLi use db poweramp
20:48:57AlexPAmimanPL: This isn't a general tagging support channel :)
20:49:21 Join solexx_ [0] (
20:49:49AlexPAmimanPL: Either have a search for how to do it with that, or have a look for something that will do it. I would recommend something, but I don't know off the top of my head
20:50:19 Join n1s [0] (n=n1s@rockbox/developer/n1s)
20:50:27CIA-6New commit by mc2739 (r23923): Get working again after target rename
20:51:38pixelmathe UsefoolTools in the Rockbox wiki might mention some other tagging programs and I'm quite sure you can find one that can :) I'm not sure what to recommend because it depends also on your operating system (and I can't recommend my favourite tagger for removing Ape tags :\ )
20:56:19 Quit AmimanPL ("CGI:IRC (Ping timeout)")
20:56:46 Join tomers [0] (
20:56:50 Join AmimanPL [0] (
20:56:55CIA-6New commit by tomers (r23924): Manual: Apply dos2unix to replace CR+LF end-of-line with LF
20:57:33 Quit AmimanPL (Client Quit)
20:57:39 Join AmimanPL [0] (
20:57:51AmimanPLi got disconnected
20:57:51pixelmatomers: I don't understand the \nopt in the previous commit, can you explain?
20:58:29AmimanPLi downloaded mp3tag to solve that problem
20:59:51pixelmaI always have trouble finding out how to really delete Ape tags with MP3tag, hopefully that was improved in the meantime
21:00:07tomerspixelma: C200 had no USB HID keypad shortcut for 'Blank screen', so the row in the table had an empty cell, which should have been removed in the manual for that specific target. I added these lines in my last commit, bu they were commented by mistake
21:00:31pixelmabut, it's getting offtopic
21:01:37 Quit solexx (Read error: 110 (Connection timed out))
21:02:29pixelmatomers: by why did you put the \nopt before the "next line" \\ and the close it before the next \\ and even keep the c200 pad inside the whole block?
21:02:40AmimanPLnew version is about 7 days old
21:03:10pixelmatomers: in the \opt list inside the excluded part
21:03:52pixelmathe & in that list is also wrong
21:04:34tomerspixelma: refer to FS #10652 - Manual: Can't use nopt inside btnmap
21:04:40pixelmaafter SANSA_C200_PAD I mean, it needs to be an %
21:04:51tomersThis was the only way it could compile...
21:05:08tomersAlso, I tried to follow the guidelines...
21:07:16 Join JdGordon| [0] (n=Miranda@nat/microsoft/x-qrtotowphfnrhzkx)
21:08:13AmimanPLNow i have configured mp3tag to solve my problem
21:08:25AmimanPLit needs minor optimization
21:08:45pixelmatomers: yes, I noticed which is nice but the placement of the \nopt is really confusing. I also had problems when nopting out complete lines for one target but only if those would be the last line in the table (or then not because the part is excluded)
21:10:08tomerspixelma: I agree this is confusing (Latex is confusing in general). If you have any neater way to do it, you are welcome :-) Or let me know how to do it and I'll do it myself
21:11:05tomerspixelma: BTW thanks for helping me with WPS (RTL issue). I'll probably do it sometime soon...
21:11:19pixelmathe SANSA_C200_PAD inside that still doesn't make sense and & in that list is still wrong
21:11:38pixelmaI'll take a look though
21:11:39 Join FOAD_ [0] (
21:13:42 Quit AmimanPL ("CGI:IRC (Ping timeout)")
21:14:40 Join AmimanPL [0] (
21:15:23 Join Creposucre [0] (
21:15:32AmimanPLthanks for helping
21:15:50AmimanPLWindows is showing me correct tags
21:16:05AmimanPLnow i have to rebuild the database
21:17:52*bluebroth3r doesn't think that LaTeX is confusing in general :)
21:18:10CreposucreI would like to get wiki rights, who should I ask for, please?
21:18:37pixelmatomers: hope you could follow my explanation, thought it could be a bit difficult to do this theoretically
21:18:38*tomers Thinks that Latex has syntax that's not intuitive, and easy to break - and hard to debug
21:18:46 Quit Utchybann_ (Read error: 113 (No route to host))
21:18:52 Join ennui [0] (
21:18:57pixelmaCreposucre: what's your wiki name?
21:19:26tomerspixelma: Yeah, I think I'll have to decipher your explanation...
21:19:43bluebroth3rtomers: have you ever used MusixTeX? *That's* confusing :)
21:20:15bluebroth3rbut sure, LaTeX itself is not trivial if you want to do more advanced things.
21:20:35bluebroth3rthis opt stuff unfortunately is more advanced ...
21:21:19 Join DSDV [0] (i=4fd5aa50@gateway/web/freenode/x-nhomqjmvvlnrtbgl)
21:21:20tomersI wish that at least compilation will be less noisy :-)
21:21:43DSDVanybody german ?
21:22:00DSDVok so in english
21:22:25*bluebroth3r hands out some patience :P
21:22:29DSDVcan some help me to change the bootscreen/loadingscrren in rockbox ?
21:22:41Creposucreit's "LaurentGautier"
21:23:02tomersDSDV: You can change it, but then you have to compile and install manually
21:23:03DSDVsry what ?
21:23:12AmimanPLthanks for all
21:23:17AmimanPLI'm going out
21:23:28 Quit AmimanPL ("CGI:IRC")
21:23:32DSDVk and how i do it (ami bb)
21:23:42bluebroth3rDSDV: please use real words.
21:24:16*tomers checkiung
21:24:23DSDVok thx
21:24:24 Quit FOAD (Read error: 110 (Connection timed out))
21:24:25 Nick FOAD_ is now known as FOAD (
21:24:39 Nick linuxguy4 is now known as linuxguy3 (
21:25:25 Quit toffe82 (Read error: 104 (Connection reset by peer))
21:25:27ennuiDid anyone succeed in enabling timestretch on the iPod Nano G2?
21:25:33 Quit jasio (
21:25:38pixelmatomers: by the way - why doesn't the c200 have that shortcut?
21:26:15tomersDSDV: rockbox/apps/bitmaps/native - you should edit the image that corresponds to your player's dimensions, and recomile
21:26:47DSDV<tomers> k much thankx i´ll try it ^^
21:27:03tomerspixelma: I probably couldn't find an available key for this function (also, I couldn't find a key for right-click in mouse mode)
21:28:02pixelmaare you sure it's the c200 - and only for "screen black"?
21:28:04DSDVlol i dont have a floder called "apps"
21:28:23Creposucrepixelma: my wiki name is "LaurentGautier"
21:28:29tomersDSDV: If you're into writing code, think about writing a utility / script that is given an svg image in any size, and converts (using imagemagik) to all possible dimensions. It would be nice to let people customize the boot screen :-)
21:28:30mc2739Creposucre: you should have wiki update rights now
21:28:52 Join gitster [0] (
21:28:59Creposucremc2739: thank you!
21:29:57pixelmaI had to find out first how that page is called now
21:30:23mc2739pixelma: WikiUsersGroup
21:30:41 Part gitster ("ERC Version 5.3 (IRC client for Emacs)")
21:31:39pixelmayes, found it now but you were quicker - I also wonder if the new wiki system doesn't have the recent changes list anymore, it helped me to quickly find TWikiUsersGroup back then...
21:32:31DSDVhm... i think im to stupit do doo it ^^ i dint real noticed that its not only replaceing an image ......
21:32:42 Join efyx_ [0] (
21:33:52mc2739pixelma: there is RecentChanges, but it only shows the last 10 updates
21:34:10 Quit Creposucre ("CGI:IRC (Ping timeout)")
21:34:19pixelmatomers: the screen black action is defined in the c200 keymap file. Maybe you are confusing it with a different target? My guess it's the H10 as it seems the most restricted/unusual button layout
21:34:38mc2739oh, the is a "more" link, too
21:36:20pixelmaah, good. In Twiki times, that table was on the wiki's main page too (webhome)
21:37:15DSDVhm.... dosent give it an easyer way to change it ???? (the botscreen)
21:39:27 Join toffe82 [0] (n=chatzill@
21:39:38 Quit toffe82 (Read error: 104 (Connection reset by peer))
21:41:13 Join toffe82 [0] (n=chatzill@
21:41:23pixelmatomers: I wouldn't call the black screen "blank screen" though, in my settings a white screen would be blank...
21:42:57 Join froggyman [0] (
21:43:25 Part DSDV
21:46:27tomerspixelma: I'll check it out. I think I saw 'black screen' in the 'PowerPoint keyboard shortcuts' page I've googled... :-)
21:46:31n1sheh, ther is no way to cancel a test_codec run :)
21:46:52 Join huelk__ [0] (
21:46:59pixelmafinally, why is the mouse over action not opted out anymore for the H10 as it was in the patch
21:47:16pixelmanopted, I mean
21:47:37ennuiWhere can I propose changes to a player specific manual?
21:49:14tomerspixelma: The 'Black screen' is for PowerPoint/Impress etc. - as opposed to 'White screen'. hmmm, just noticed it is called 'blank'. I will revert
21:49:56pixelmaI see it in the c200 keymap file, so why the exclusion in the manual?
21:50:08tomerspixelma: Error. Fixing
21:50:20 Join huelk [0] (
21:50:32pixelmaand the mouse over?
21:51:05mc2739FlynDice: ping
21:51:41FlynDicemc2739: pong what's up
21:53:04AlexPennui: If there is something wrong, then a bug on flyspray. If you have the fix, then a patch on flyspray :)
21:53:14mc2739FlynDice: in your work on the AMS sd driver, have you noticed any way to tell how many banks are present on the internal storage?
21:53:35tomerspixelma: Mouse over is mapped for h10 (long right)
21:54:32pixelmasounds like no need for nopting then?
21:54:37FlynDicemc2739: No, I remember funman mentioning that awhile ago but I've never looked at that
21:55:47FlynDicemc2739: are you working on something that needs the info or just curious?
21:55:53mc2739FlynDice: Ok, I'll check with funman then.
21:56:44tomerspixelma: indeed
21:56:54ennuiAlexP, thanks!
21:57:12mc2739FlynDice: On the debug screen, disk info shows block for 1 bank, not the total blocks for all banks
21:57:49AlexPennui: Anything in particular?
21:57:55pixelmatomers: while you're add it, please fix the SANSA_C200_PAD& there (should be a % there
21:58:03pixelmaerr... at it
21:58:05mc2739FlynDice: it just bugs me to see inaccurate info there
21:58:21tomerspixelma: What line?
21:58:40tomersoh, found it
21:58:55FlynDicemc2739: yes, I believe that is correct, funman mentioned perhaps trying to switch banks until it finally failed to count them but I haven't heard much about it since
22:00:33mc2739FlynDice: I've tried switching banks, but it does not appear to give any indication of failure.
22:01:53FlynDicemc2739: So we can switch to a nonexistent bank? Maybe switch banks then and attempt a write to see if it fails?
22:02:06FlynDiceer resd...
22:02:16JdGordon|does anyone know what the plan for cabbieV2 + sbs is for svn? are we going to just take whatever moonscape comes up with in the forums?
22:02:20FlynDiceor even read
22:02:55mc2739FlynDice: that is what I was going to try next
22:03:05***Saving seen data "./dancer.seen"
22:03:41 Quit huelk_ (Read error: 110 (Connection timed out))
22:06:23 Quit huelk__ (Read error: 110 (Connection timed out))
22:09:55 Quit n17ikh (Connection timed out)
22:10:38 Join huelk_ [0] (
22:12:42CIA-6New commit by Domonoky (r23925): Add the possibility edit target definitions to the admin interface.
22:14:10 Join beta_ [0] (
22:14:13CIA-6New commit by tomers (r23926): Manual, USB HID: C200 does have keymap for Black screen in Presentation Mode ...
22:14:23 Nick beta_ is now known as beta2k (
22:20:27 Quit huelk (Read error: 113 (No route to host))
22:24:19FlynDicemc2739: If it helps at all by dividing the number of blocks in my 8gb uSD by the number of blocks reported for the 8gb internal I come up with 7 banks for an e280v2 if that helps
22:26:16 Join kugel [0] (n=kugel@rockbox/developer/kugel)
22:28:57CIA-6New commit by tomers (r23927): Cowon D2: Add some more key mappings
22:34:28 Join LambdaCalculus37 [0] (n=rmenes@rockbox/staff/LambdaCalculus37)
22:34:45 Quit Rondom (Nick collision from services.)
22:34:48 Join Rondom [0] (
22:34:53 Join grant__ [0] (i=439e4f63@gateway/web/freenode/x-zhjwprfsfkvevbuu)
22:37:25 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
22:44:02CIA-6New commit by nls (r23928): Remove inline asm that hasn't been useful for years, no speed difference
22:44:06 Join FOAD_ [0] (
22:46:07 Join MethoS- [0] (n=clemens@
22:46:35 Join TheSeven|Mobile [0] (n=theseven@
22:49:06tomerspixelma: ping
22:53:00 Join flydutch [0] (
22:54:06LambdaCalculus37toffe82: The HDD1630 is pretty much stable and usable for daily use.
22:54:34LambdaCalculus37There's just no manual and the install method is rather funky, requiring you to go into recovery mode.
22:58:08toffe82LambdaCalculus37: can't we find how they do with the original firmware update ?
22:58:25LambdaCalculus37toffe82: lowlight was working on that, I think.
22:59:13toffe82We need more people with gogear 's ...
22:59:22toffe82people who have time ...
22:59:47tomersIn the manual, what exactly is \ActionQuickScreenExit ? I understand what it does, but I can't see it in apps/keymaps/keymap-*
22:59:53 Quit FOAD (Read error: 110 (Connection timed out))
22:59:53 Nick FOAD_ is now known as FOAD (
23:00:00LambdaCalculus37toffe82: And lowlight needs to get his work for the HDD6330 touchpad committed. ;)
23:00:59toffe82LambdaCalculus37: he is still active, there was a message from him in the forum this week or last week I think
23:01:25 Quit ennui (Read error: 110 (Connection timed out))
23:01:36 Quit einhirn (Read error: 104 (Connection reset by peer))
23:01:52grant__hey, theseven, is mine and liars version of rockbox on svn?
23:04:03AlexPyour version of Rockbox?
23:05:00grant__yah, i got a M version of rockbox from liar, bc the other version didnt work
23:05:14AlexPreal words please :)
23:05:33AlexPyou can look at what commits have happened since the version you have
23:05:47LambdaCalculus37toffe82: Yeah, but from what it looks like, he's been rather busy as well.
23:06:17 Quit BlakeJohnson86 (Read error: 104 (Connection reset by peer))
23:06:21 Quit grant__ ("Page closed")
23:06:37 Join BlakeJohnson86 [0] (
23:08:09LambdaCalculus37toffe82: Rockbox also doesn't dual-boot on the SA9200, but I don't know if that's important or not because USB and charging are enabled in the port already, so we can probably forget dual-booting.
23:09:34toffe82if we can charge and transfer files, I don't see the problem
23:09:48toffe82is this something needed to go unstable ?
23:10:27LambdaCalculus37toffe82: For the SA9200, we can go unstable already but we just need to not build any plugins that require external artwork for the time.
23:14:44toffe82can we find some information on how it is done in the original update program looking at the usb ? If I log all the transfer, can we find something?
23:15:02LambdaCalculus37toffe82: Yes, definitely so. :)
23:15:29toffe82I will try to do it, if I find my sa9200...
23:19:58 Part domonoky
23:22:12 Nick Ypsy is now known as YPSY (
23:25:07LambdaCalculus37toffe82: Do you happen to have another HDD1630?
23:25:18LambdaCalculus37Otherwise I can try the same with the one I've got.
23:25:25toffe82LambdaCalculus37: yes but no hdd,
23:25:42toffe82I have to try with a cf
23:25:55 Quit TheSeven|Mobile ()
23:25:57toffe82but the only one I have is 512mb
23:26:45LambdaCalculus37toffe82: I tried putting a 4GB CF into my HDD1630, but it doesn't seem to be recognized by the Philips updater app.
23:27:32toffe82if you install first on the hdd and make an image to the cf, is it working ?
23:27:41LambdaCalculus37Haven't tried that.
23:28:14toffe82the hdd was working on this one ?
23:30:06LambdaCalculus37The hdd still works.
23:30:36toffe82you should try this, first install the hdd and then make an image to your CF
23:31:26LambdaCalculus37toffe82: Shall do.
23:31:41*LambdaCalculus37 wants to get to the hdd6330 and get work done for that as well...
23:31:47LambdaCalculus37And the Gigabeat T. :)
23:32:51toffe82and the V...
23:33:38 Quit bmbl ("Bye!")
23:36:43*LambdaCalculus37 doesn't have a V though :()
23:37:39*toffe82 has one wiating...
23:39:11LambdaCalculus37toffe82: I need to go over my disassembly of the Gigabeat T OF to see where (and if!) that digital signature exploit exists still.
23:44:12 Quit n1s ("Lmnar")
23:44:24LambdaCalculus37Besides when you, me, and saratoga tried using beastpatcher to see what happens when you patch a Gigabeat T OF.
23:46:55 Quit shai ("Leaving")
23:48:39 Join akur [0] (
23:53:34LambdaCalculus37toffe82: AFAIK the Gigabeat V doesn't seem to have the signature exploit that the beast has. Can you confirm for me?
23:53:56toffe82LambdaCalculus37: I don't know..
23:54:15 Quit Llorean ("Leaving.")
23:54:20LambdaCalculus37toffe82: Hmmm...
23:54:24*LambdaCalculus37 would like to try that
23:55:09toffe82I remember trying to load rockbox on it and end up with a blank partition
23:55:24toffe82but I have to try agian
23:56:31 Join Strife89 [0] (
23:58:17 Join BlakeJohnson861 [0] (
23:58:23LambdaCalculus37toffe82: Blank partition? It wiped the hard drive on you?
23:58:40toffe82like on the S
23:58:57toffe82the hidden partition

Previous day | Next day